Embed Size (px)
Citation preview
APJCN 2004: 13, Number 4: 318-414 ISSN 0964-7058
2004 Volume 13 Number 4
A
Asia Pacific
Journal of
Clinical Nutrition
Editors
Mark Wahlqvist, Melbourne Akira Okada, Osaka
ww
PCNS � IUNS � NSNZ
HEC PRESS w.healthyeatingclub.org/APJCN/
APJCN 2004: 13, Number 4: 318-414 ISSN 0964-7058
2004 Volume 13 Number 4
APCNS � IUNS � NSNZ
Asia Pacific Journal of Clinical Nutrition
2004, HEC PRESS, Melbourne, Australia ISSN 0964-7058
Visit the journal website at: http://www.healthyeatingclub.org/APJCN
Asia Pacific
Journal of
Clinical Nutrition
THE JOURNAL OF THE ASIA PACIFIC CLINICAL NUTRITION SOCIETY
APJCN 2004: Volume 13 (Number 4) : 318-414 ISSN 0964-7058
EDITORS: Mark Wahlqvist MD, Asia Pacific Health and Nutrition Centre, Monash Asia Institute, 8th Floor, Menzies Building, Monash University, Wellington Road, Clayton, Melbourne, Victoria 3800, AUSTRALIA . Fax: (+61) 3 9905 8146; Email: [email protected]. Akira Okada MD, President, Osaka Medical Center and Research Institute for Maternal and Child Health, 840 Murodo-cho, Izumi City, Osaka, 594-1101, JAPAN. Fax: (+81) 725 56 5682; Email: [email protected]. Managing Editor: Antigone Kouris-Blazos PhD, HEC PRESS, Suite 4, 2 Elm Grove, McKinnon,, Melbourne, Victoria 3204, AUSTRALIA; Fax: (+61) 3 95154544. Email: [email protected] Editorial Office: Asia Pacific Journal of Clinical Nutrition, Asia Pacific Health and Nutrition Centre, Monash Asia Institute, 8th Floor, Menzies Building, Monash University, Wellington Road, Clayton, Melbourne, Victoria 3800, AUSTRALIA Fax: (+61) 3 9905 8146. Email: [email protected]. Administration officer: Wendy Yu
Asia Pacific Journal of Clinical Nutrition (APJCN) is a peer-reviewed journal published for the Asia Pacific Clinical Nutrition Society (APCNS). The journal publishes original research reports, short communications, reviews, and letters to the editors. Letters to the editors may be either very short articles (500 words) containing new material, case reports, or comments on previous papers or on other topics of current interest. News, book reviews and other items may also be included. All articles are peer-reviewed by at least two researchers expert in the field of the submitted paper. For detailed instructions concerning the submission of manuscripts, please refer to the Instructions to Authors at the back of this issue or visit the journal web page, listed below.
The aims of the Asia Pacific Clinical Nutrition Society (APCNS) are to promote the education and training of clinical nutritionists in the region and to enhance the practice of human nutrition and related disciplines in their application to health and the prevention of disease.
The journal is also an organ of The Nutrition Society of New Zealand and of The Nutrition Society of Australia, whose members receive APJCN as part of their society membership. The International Union of Nutritional Sciences (IUNS) is acknowledged as a non-financial co-sponsor of the journal.
Abstracting and Indexing Services: The Journal is indexed by Australasian Medical Index, CAB Abstracts, Chemical Abstracts Service, Current Contents/Clinical Medicine, Current Opinion, Index Medicus/MEDLINE, ISI Alerting Services, Science Citation Index, SciSearch and SUBIS Current Awareness in Biomedicine. The Journal is also indexed selectively in APAIS. Business Correspondence and Advertisement Booking: All matters relating to orders, subscriptions, back issues, offprints, advertisement booking and general enquiries should be addressed to: HEC PRESS, Suite 4, 2 Elm Grove, McKinnon, Melbourne, Victoria 3204, Australia; Tel: +61 (0) 3 95767987, Fax: +61 (0) 3 95154544; [email protected]; http://www.healthyeatingclub.org
Subscription Rates: The journal is published four times per year and the prices per annum for 2004 are as follows:
Personal: Hard copy (including on-line access) AU$135; On-line access only AU$120. Institution: Hard copy (including on-line access) AU$400; On-line access only AU$170.
The journal is despatched to subscribers outside Australia by surface air-lifted delivery.
Copyright and Photocopying: © 2004 HEC PRESS, Healthy Eating Club Pty Ltd. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by HEC PRESS, Healthy Eating Club Pty Ltd for libraries and other users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service, provided that the base fee of US$15 per copy is paid directly to: CCC, 222 Rosewood Drive, Danvers, MA 01923, USA. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works or for resale. Special requests should be addressed to the Publisher.
APJCN online & CD-ROM Issues prior to 2004: Full papers are available free-of-charge via the website http://www.healthyeatingclub.org. Volumes 1-11 can also be purchased on CD-ROM for AU$170 (to order CD-ROM use APJCN subscription form on last page of this issue). Issues after 2003: subscribers have on-line access to full papers on the website www.healthyeatingclub.org (password access).
Disclaimer: The Publisher, the Society and the Editors cannot be held responsible for errors or any consequences arising from the use of information contained in this journal; the views and opinions expressed do not necessarily reflect those of the Publisher, the Society or the Editors, neither does the publi-cation of advertisements constitute any endorsement by the Publisher, the Society or the Editors of the products advertised.
APJCN Editorial Advisory Board
Australia Professor Paul J Nestel
Baker Institute, Alfred Hospital, Commercial Rd, Prahran, VIC 3181 Tel: +61-3-9522-4333 Fax: +61-3-9521-1362 Email: [email protected]
Professor Boyd Swinburn Physical Activity and Nutrition Research Unit, School of Health Sciences, Deakin University, 221 Burwood Highway, Burwood, VIC 3125 Tel: +61-3-9251-7096 Fax: +61-3-9244-6017 Email: [email protected]
Professor A Stewart Truswell Human Nutrition Unit, Biochemistry Dept, University of Sydney, Sydney, NSW 2006 Tel: +61-2-9351-3726 Fax: +61-2-9351-6022 Email: [email protected]
China Professor Gu Jing-Fan
Professor of Nutrition, Institute of Hygiene and Environment Medicine, 1 Da Li Dao, Tianjin, 300050 Tel: +86-22-846-55402 Fax: +86-22-846-55402 Email: [email protected]
Fiji Professor Sitaleki Finau
Head, School of Public Health & Primary Care, Fiji School of Medicine, Private Mail Bag Suva, Fiji Islands Tel: +679-311-700 Fax: +679-303-469
Jimaima Veisikiaki Lako c/o Professor Subramaniam Sotheeswaran, Department of Chemistry, University of the South Pacific, Suva, Fiji Email: [email protected]
Guatemala Dr Noel W Solomons
Centre for Studies of Sensory Impairment, Aging and Metabolism, PO Box 02-5339 Section 3162/Guatemala Miami, FL 33102-5339 Tel: +502-473-3942 Fax: +502-473-3942 Email: [email protected]
Germany Professor Peter Fürst
c/o Institut für Ernährungswissenschaft Universität BonnEndenicher Allec 11-3 D-53115 Bonn, Germany Tel: +49-228-73-38-14 Fax: +49-228-73-64-83
Indonesia Assoc. Professor Widjaja Lukito
SEAMEO-TROPMED* National Centre Indonesia, Regional Center of Community Nutrition, University of Indonesia, Jalan Salemba Raya 6, Jakarta 10430, Indonesia Tel: +62-21-330-205 Fax: +62-21-390-7695
Email: [email protected] Professor Soemilah Sastroamidjojo
(for SEAMEO-TROPMED Nutrition Centre) Faculty of Medicine, University of Indonesia, 6 Salemba Raya, Jakarta 10430
Japan Professor Kazuo Kondo
Institute of Environmental Science for Human Life, Ochanomizu University, 2-1-1 Ohtsuka, Bunkyo-ku, Tokyo 112-8610 Tel: +81-3-5978-5812 Fax: +81-3-5978-2694 Email: [email protected]
Korea Professor Sook He Kim
Department of Foods and Nutrition, Ewha Women’s University,
11-1 Dae-hyun Dong, Seo-dae moon Ku, Seoul 120-750 Tel: +82-2-393-0051 Fax: +82-2-393-5903 Email: [email protected]
Malaysia
Professor B Abdul K Khalid Professor of Medicine, Universiti Kebangsaan Malaysia, Jalan Raja Muda Abdul Aziz, 50300 Kuala Lumpur Tel: +60-3-440-5000 Fax: +60-3-291-2659 Email: [email protected]
New Zealand Professor Jim Mann Department of Nutrition, University of Otago, PO Box 56, Dunedin Tel: +64-9-795-780 Fax: +64-9-770-956 Email: [email protected] Philippines Dr Rodolfo Florentino
Food & Nutrition Research Institute, Pedro Gil Street, PO Box EA-467, Emita, Manila 1000 Tel: +63-2-823-8071 Fax: +63-2-823-8934
Email: [email protected] Singapore Dr Paul Deurenberg
135, Serangoon Avenue 3, #10–01, Chiltern Park, Singapore 556114 Tel: +65- 91251425 Fax: +65-68585985 Email: [email protected]
Taiwan Professor Wei-Jao Chen
National Taiwan University 1 Section 4, Roosevelt Road Taipei, Taiwan, ROC 106 Tel: +886-2-23634090 Fax: +886-2-23621877 Email: [email protected]
Thailand Assoc. Professor Prasong Tienboon
Division of Nutrition, Dept of Paediatrics, Faculty of Medicine, Chiang Mai University, Chiang Mai 50002 Tel: +66-53-221-122 Fax: +66-52-217-144 Email: [email protected]
Vietnam Professor Ha Huy Khoi
National Institute of Nutrition, 48 Tang Bat Ho, Hanoi Tel: +84-4971-7090 Fax: +84-4971-7885
Email: [email protected]
* SEAMO-TROPMED = South East Asia Ministry of Education, Tropical Medicine
APCNS Editorial Advisory Board
Asia Pacific Clinical Nutrition Society
President: Professor Widjaja Lukito, Indonesia
Immediate Past President: Professor Mark Wahlqvist, Australia
Vice President: Professor Kazuo Kondo, Japan
Secretary: Associate Professor Prasong Tienboon, Thailand
Treasurer: Professor Osman Ali, Malaysia
Councillor: Professor Zhu-ming Jiang, China Asia Pacific Clinical Nutrition Society is established to create links between clinical nutritionists in the Asia Pacific region. The Society seeks thereby to promote the contribution of nutritionists to the health of the populations in Asia Pacific countries. Asia Pacific Clinical Nutrition Society will encourage continuing nutrition and training in the region so as to promote the highest possible level of research and practical innovation. Asia Pacific Journal of Clinical Nutrition, along with the sponsorship by Asia Pacific Clinical Nutrition Society of regional and local clinical nutrition meetings, is expected to assist greatly in the achievement of these aims. Membership of Asia Pacific Clinical Nutrition Society, for which there is a modest annual fee and which includes on-line access to Asia Pacific Journal of Clinical Nutrition, is open to all clinical nutritionists in the region. The Society will consider application for membership based on submission of a curriculum vitae and a statement of support from one of the officers and the individuals listed below. The Asia Pacific Clinical Nutrition Society representatives should state: 'On behalf of Asia Pacific Clinical Nutrition Society I support the application of [Name and address] for membership in the Society'. After the applicant has obtained this endorsement from the representative it should be forwarded with the applicant's curriculum vitae for consideration and acknowledgement to: The Secretary, Asia Pacific Clinical Nutrition Society, Asia Pacific Health and Nutrition Centre, Monash Asia Institute, 8th Floor, Menzies Building, Monash University, Wellington Road, Clayton, Victoria 3168, Australia.
ASIA PACIFIC CLINICAL NUTRITION SOCIETY (APCNS)
Society Membership
AU$100 per annum
Asia Pacific Clinical Nutrition Society (APCNS) was established to create links between clinical nutritionists in the Asia Pacific region and to promote their contribution to the health of the populations of Asia Pacific countries. APCNS encourages continuing nutrition education and training in the region in order to support the highest possible levels of research and practical innovation. APCNS sponsorship of local and regional meetings along with the publication of Asia Pacific Journal of Clinical Nutrition, is expected to assist greatly in the achievement of these aims. President: Professor Widjaja Lukito, Indonesia Immediate Past President: Professor Mark Wahlqvist, Australia Vice President: Professor Kazuo Kondo, Japan
Membership Application You will need your membership application endorsed by one of the members of the Advisory Board listed at the front of the journal. Please use the Endorsement Form below. Send the completed endorsement form, your curriculum vitae and bankdraft for AU$100 (made payable to APCNS) or credit card details (use the form below) to the society. Membership Application & Endorsement I wish to apply for membership of the Asia Pacific Clinical Nutrition Society Title ………………. Name……………………………………………………………………………………………………
Mailing Address……………..…………………………………………………………………………………………………
……………………………………………………………………………………………Postcode…………………………..
Tel: ………………………………………………………..Fax………………………………………………………………..
Email ……………………………………………………………………………………………………………………………
I support the application of ………………………………………………………………for membership of the Asia Pacific Clinical Nutrition Society.
Name of Endorser…………………………………………Signature of Endorser……………………………………... CREDIT CARD PAYMENT I wish to pay by Visa MasterCard (Other cards are not accepted) AmNOU SanM
receive ON-LINE ACCESS
to Asia Pacific Journal of Clinical Nutrition as part of your membership
ount $ …………….. Expiry Date ……./……….Signature …………………………………………………………. ame on Card ………………………………………………………………………………………………………………. n-line access: Please provide preferred sername..................................................…………......Password...........................................…………………….
end your membership application to: The Secretary, Asia Pacific Clinical Nutrition Society, Asia Pacific Health d Nutrition Centre, Monash Asia Institute, 8th Floor, Menzies Building, Monash University, Wellington Road, Clayton,
elbourne, Victoria 3800, AUSTRALIA Fax: (+61) 3 9905 8146. Email: [email protected].
Asia Pac J Clin Nutr 2004;13 (4): 318-323 318
Original Article Can a food frequency questionnaire be used to capture dietary intake data in a 4 week clinical intervention trial? Pauline Xie Xinying MND
1, Manny Noakes PhD2 and Jennifer Keogh MSc
2 1Department of Nutrition and Dietetics, School of Medicine, Flinders University, Adelaide, South Australia 2CSIRO HSN, Adelaide, South Australia
Collecting dietary data in the clinical research setting is labour intensive and can be burdensome for study participants. The aim of this study was to assess the agreement between data obtained from 2 different dietary assessment methods, a 74-item semi-quantitative food frequency questionnaire (FFQ) and 3-day weighed food records (WFR) used to estimate dietary intake over the preceding month. One hundred and fifty nine subjects, aged between 31 and 74 years (53 males, 65 females), enrolled in a clinical trial at the Commonwealth Scientific and Industrial Research Organisation, Division of Health Sciences and Nutrition, (CSIRO HSN) Adelaide, Aus tralia. Group mean intakes and individual mean intakes estimated by the two measures were compared. One hundred and eighteen (91%) three-day WFR and their corresponding FFQ were analysed. Pearson correlation coefficients ranged from 0.22 for cholesterol to 0.78 for alcohol (median 0.41). Mean energy and nutrient intakes were within ± 20% difference. The FFQ gave lower carbohydrate intake estimates, percentage energy from carbohydrate (P <0.001) and dietary fibre (P <0.05) and gave higher percentage energy from saturated fat estimates, poly -unsaturated fatty acids (P <0.001) and mono-unsaturated fatty acids (P <0.05). Subjects were also ranked into quintiles and the quintiles cross-tabulated. The FFQ classified more than two thirds of the subjects within ±1 quintile difference for all nutrients. We conclude that this FFQ can capture similar information as WFR and may be used for estimation of dietary intakes over a relatively short time in clinical intervention trials.
Key words: food frequency questionnaire, validity, weighed food record, dietary intake, nutritional analysis, Australia
Introduction Dietary assessment tools are used to obtain information on individual or group dietary intakes and commonly used methods are weighed food records (WFR), food frequency questionnaires (FFQ) dietary recall and diet histories.1 The method chosen depends on the objectives of the study, the resources available and the demands of the technique2 and should be validated in the context in which they are used. WFR provide accurate data on dietary intake3 and thus compliance to a research protocol. However, WFR are time-consuming, requiring highly skilled interviewers and hence are resource intensive and expensive. They are bur-densome for study participants who may be have diffi-culties complying with the rigors of daily weighing of food and may underreport their intake.3 FFQ are retrospective and elicit information on the frequency of consumption of a specified list of foods and drinks, and may or may not include estimates of serving sizes. There has been much debate on the validity and reliability of FFQ as a measure of nutrient intake, and the situations in which it is appro-priate to use them.4-6 FFQ are much less invasive, can achieve higher response rates and are relatively inex-pensive.7
In Australia, the Anti-Cancer Council of Victoria (ACCV) has developed a 74-item semi-quantitative self
administered FFQ which can be optically scanned to provide analysis of nutrient intake data and thereby reduce intensive dietetic input. It is quick and easy to use. It was designed to sort individuals into quintiles based on estimated usual intake of food and nutrients over preceding 12 months and has been validated relative to seven-day weighed food records.8 The aim of this study was to assess the use of the ACCV FFQ in a clinical trial population by comparing data obtained using the FFQ and data from 3-day WFR which were being used to estimate dietary intake over the preceding month.
Correspondence address: Dr Manny Noakes, CSIRO HSN, PO Box 10041 BC, Adelaide SA 5000, Australia Tel: 0883038827; Fax 0883038899
Email: [email protected] Accepted 28 May 2004
319 PX Xinying, M Noakes and J Keogh
Methods Subjects The subjects in this study were enrolled in a dietary intervention study (N =159) comparing the effects of dose and frequency of consumption of phytosterol-containing yoghurt, on serum lipids, carotenoids and phytosterols. Subject selection criteria were: age 20-75 years, body mass index (BMI) <35 kg/m2, total cholesterol 5.0-7.5 mmol/L, triglycerides <4.5mmol/L, cholesterol-lowering medication was allowed if the type and dosage was maintained constant throughout the study. Exclusion criteria were: persons considered by the investigator to be unwilling, unlikely or unable to comprehend or comply with the study protocol and restrictions, subjects taking any supplements which could interfere with the bio-chemical parameters of interest, presence of diabetes, known lactose intolerance and untreated hyper/hypo-thyroidism. The study had ethics approval from the CSIRO ethics committee and subjects gave informed con-sent. The study design was a single-blinded parallel study with 4 interventions over a period of 4 weeks; subjects were matched according to their baseline cholesterol level and randomised to 1 of 4 interventions (yoghurt con-taining 1 or 2g phytosterols every day; 2g phytosterols on alternate days; control yoghurt with no phytosterols every day). The subjects were required to consume 140g low fat fruit yoghurt per day which provided 448kJ, 7g protein, 18g carbohydrate and 250mg calcium but not otherwise change their eating habits. Weighed food records Subjects were given detailed instructions on how to weigh and record their dietary intake, and an opportunity to practice before the commencement of the study; weighing scales were provided for those who did not possess one. Subjects were required to complete two 3-day WFR; each done two weeks apart. Each record was checked, in the presence of the subject, for accuracy and clarifications by a qualified dietitian. Food frequency questionnaire The 74-item semi-quantitative ACCV FFQ was admin-istered at the end of the trial. The subjects were not informed when it would be administered in order to minimise recall bias. The subjects were given clear in-structions to recall their dietary habits over the previous 4 weeks of the trial. The ACCV FFQ was checked for completion by a member of the clinical trial staff. The first page of the FFQ consists of 1) simple instructions on completing the questionnaire, 2) the date completed, 3) questions on the quantity of fruits, milk, bread and sugar taken daily, 4) types of vegetables consumed daily, 5) types of milk, cheese, bread and spread usually used, and 6) number of eggs taken per week. With the questions on the types of food eaten, more than one answer can be selected (e.g. question 10 asks about the type of cheese usually consumed, the subject may select more than one option if they consume more than one type of cheese) in which case, the nutrients are computed with the assum-ption that equal quantities of each type were consumed. The second page of the questionnaire consists of four sets
of photos depicting three different serve sizes for potatoes, vegetables, steak and casserole. Each photograph shows the 25th percentile (photo A), median (photo B) and 75th percentile (photo C) of serving sizes reported by Ireland et al.9 Subjects may select from 7 serving size portions: less than A, A, between A and B, B, between B and C, C, and more than C. There is also an option to select nil intake, e.g. “I never ate steak”. For items that showed consistent differences in serving sizes between genders the portion size will be scaled down or up using a factor automatically used by the nutrient analysis package developed by the ACCV. The 3rd and 4th pages of the FFQ list 74 items with 10 frequency options ranging from “never” to “3 or more times per day”. The list is cate-gorised into 4 sections 1) cereal foods, sweets and snacks, 2) dairy products, meat and fish, 3) fruits and 4) vege-tables. Three questions on alcohol intake are also included to find out 1) how many times, 2) how much, and 3) the maximum amount of alcohol consumed at any one time. Nutrient analysis The WFR were computed at CSIRO using Diet 1™ (version 4.2, 1996, Xyris® software, Brisbane) software and the NUTTAB95 food composition database. FFQ and subject barcodes were obtained from the ACCV and the completed FFQ questionnaires sent to the ACCV for analysis using software based on the NUTTAB95 food composition database. Statistical analysis All statistical analysis was performed using Statistical Package for Social Sciences™ for Windows (version 10.0.7, 1999, ©SPSS Inc.). The means and standard deviations (SD) of nutrient intakes were computed from the FFQ and the WFR. Pearson product-moment corre-lation coefficients were used to compare the questionnaire with the records. Because most nutrient intakes were skewed, all values were loge transformed to improve normality; alcohol intake values were square-rooted to improve normality, to conform to the assumptions of tests required for Pearson correlation. As statistical significance might not be appropriate for assessing agreement between different dietary assessment methods, a technique described by Bland and Altman was applied.10 It involves calculations of the mean and SD of the difference between the two methods, and the 95% limits of agreement i.e. 95% of the difference of the esti-mated nutrient intakes are expected to lie between the limits. Interpretation of the results relies on determining an acceptable difference between the two measures. Quin-tile rankings were used to classify subjects into categories and cross-tabulated. This was done to show the agreement between the classification of subjects in quintiles from the FFQ and the WFR. Under-reporting was addressed using the Goldberg cut-off ratio (energy intake: basal metabolic rate/physical activity level – EI: BMR/PAL).11,12 A blanket PAL of 1.2 was used to calculate the individual Goldberg ratio to identify the under-reporters – under-reporters were those with a ratio of less than 0.76. Other statistical tests included paired t test, 1-way ANOVA, and chi-square tests, all of which were applied as appropriate.
Use of a food frequency questionnaire in a clinical trial 320
Results Of 159 subjects who completed the study, 145 completed the FFQ. 5 FFQ were incomplete and were rejected. Due to the time constraints, not all the WFR were computed. One hundred and eighteen 3 day WFR were paired with their corresponding FFQ and analysed. Gender distri-bution was 53 males and 65 females, 55% and 45% respectively and mean age was 58 years (± 9), range 31 to 74 years, with a mean BMI of 26.1 (± 3.3). Table 1 shows the means and the corresponding SD estimated by the FFQ and the 3-day WFR for energy intake and for 10 selected nutrients. Pearson correlation coefficient and
significance testing from paired t test are also presented. All nutrient estimates by the FFQ are within ± 20% of the estimates produced by the mean of the 3-day WFR. The group means obtained for all nutrients were comparable with the exception of carbohydrate and percent energy from carbohydrate. The inter-individual variability, as measured by the SD, was higher for the FFQ than the corresponding values given by the WFR method. The only exceptions were dietary fibre and percentage energy from total fat, which showed lower variability in the FFQ. The Pearson correlation coefficient, r, ranged from 0.22
N = 118
WFR Mean
SD
FFQ Mean
SD
r†
P*
Energy MJ 8.2 1.9 7.9 2.7 0.39
Protein g 91.0 21.0 90.9 36.3 0.27
Carbohydrate g 241.8 61.3 210.8 75.6 0.48 <0.001
Total Fat g 63.7 21.5 68.0 29.0 0.32
Saturated Fat g
PUFA a g
MUFA b g
23.2
10.7
24.1
9.4
5.4
9.2
25.2
11.9
24.8
12.3
6.0
11.4
0.42
0.32
0.29
Cholesterol mg 231.2 100.7 242.6 114.7 0.22
Alcohol g 10.9 13.8 10.5 15.2 0.78
Dietary Fibre g 25.9 9.9 23.9 9.7 0.56 <0.05
β-Carotene µg 2080.1 1773.0 2682.4 2000.0 0.44
% E from Protein 19.1 3.4 19.5 3.4 0.42
% E from Carbohydrate 47.2 6.4 42.9 6.4 0.43 <0.001
% E from Total Fat 28.3 5.8 31.3 5.5 0.34
% E from Saturated Fat
% E from PUFA a
% E from MUFA b
10.3
4.8
10.7
2.9
2.1
2.9
11.6
5.6
11.3
3.0
2.2
2.3
0.49
0.30
0.42
<0.001
<0.001
<0.05
% E from Alcohol 3.6 4.2 3.8 5.4 0.77
Table 1. Group mean nutrient intake (mean ± SD) from 3-day WFR and FFQ
† Values were loge transformed or square rooted (for alcohol) to reduce skewness and improve normality, as required by the statistical assumption of tests related to the Pearson correlation coefficient; aPolyunsaturated fatty acids; bMonounsaturated fatty acids; *Paired t test
Table 2. Cumulative Percentage Agreement between nutrient intakes derived from the 3-day WFR and the FFQ
Percent Agreement Exact +/- 1 Fifth +/- 2 Fifths +/- 3 Fifths
Energy 34 69 91 97 Protein 33 73 86 97 Carbohydrate 34 70 91 98 Total Fat 31 62 88 97
Saturated Fat PUFA MUFA
35 21 26
66 64 60
86 86 86
98 99 97
Cholesterol 28 65 89 97 Dietary Fibre 34 78 95 100 β-Carotene 26 55 79 100 % E from Protein 28 67 87 97 % E from Carbohydrate 35 69 87 97 % E from Total Fat 28 63 87 94
Saturated Fat PUFA MUFA
37 27 24
78 64 62
95 86 91
97 97 97
321 PX Xinying, M Noakes and J Keogh
Table 4. Percentage of under-reporters distinguished by the FFQ and the WFRs
FFQ WFR % Under-reporters (using PAL 1.55)
45 31
% Under-reporters (using PAL 1.2)
16 6
for cholesterol to 0.78 for alcohol (median = 0.41). There were significant differences between estimates of carbo-hydrate (P <0.001), dietary fibre (P<0.05), percent energy from carbohydrate (P <0.001), percent energy from satu-rated fat (P <0.001), polyunsaturated fatty acids (PUFA) (P <0.001) and monounsaturated fatty acids (MUFA) (P <0.05) from the two methods. Across the 4 groups, there were no significant differences in the nutrient intakes measured by both methods. Table 2 shows the cumulative percentage agree-ment between nutrient intakes estimated from the WFR and the FFQ. The percentage allocated to the same quintile varied from 21% for PUFA to 35% for saturated fat and energy from carbohydrate. Less than 6% of subjects were grossly misclassified. The FFQ was able to classify more than two thirds of the subjects within ±1 quintile difference. According the Bland and Altman, the 95% limits of agreement between the FFQ and the WFR are presented in Table 3. The mean nutrient intakes varied by less than 20%, but the inter-individual variation was very large. The difference in the group mean energy intake estimated by both methods, for example, was only 3.8%, but at the individual level, the difference ranged from –4.9 to 5.5 MJ in 95% of the population. Energy intake difference versus mean energy intake estimated by the 2 methods is shown in Figure 1. The limits of agreement were around 5MJ on either side of the mean, a figure too large to suggest use of the FFQ for individual dietary assessment.
The percentage of under-reporters identified by the FFQ and the WFRs are shown in Table 4. Using the Goldberg cutoff ratio,11,12 the FFQ and the WFR reported 16% and 6% under-reporters respectively. There were no signi-ficant differences in gender, age and BMI in under-reporting in this population (data not shown). Statistical analysis performed after exclusion of under-reporters in both methods showed no significant differences. Discussion The key findings of this study were that all nutrient estimates by the FFQ are within ± 20% of the estimates produced from the mean of the 3 day WFR and that the group means obtained for all nutrients were comparable with the exception of energy, carbohydrate and percent energy from carbohydrate. In the present study mean energy and nutrient intakes were within ± 20% difference, which is similar to the findings of a previous validation study of the same ACCV FFQ in a study of 63 premeno-pausal women.8 The correlations observed were also si-milar to the present study. Pearson correlation coefficients
Table 3. 95% limits of agreement between WFRs and FFQ according to Bland and Altman8
WFR Mean FFQ Mean Mean difference (WFR-FFQ)
95% limits of agreement
Energy MJ 8.2 7.9 0.3 -4.9 5.5 Protein g 91.0 90.9 0.1 -75.2 75.4 CHO g 241.8 210.8 31.0 -112.0 174.1 Total Fat g 63.7 68.0 -4.3 -101.7 15.9 Sat Fat g 23.2 25.2 -2.0 -25.5 21.5 PUFA g 10.7 11.9 -1.2 -14.3 12.0 MUFA g 24.1 24.8 -0.7 -25.5 24.0 Cholesterol mg 231.2 242.6 -12 -277 254 Alcohol g 10.9 10.5 0.4 -23.0 23.8 Fibre g 25.9 23.9 2.0 -17.0 20.9 β-Carotene µg 2080 2682 -602 -2103 2990 % E from Protein 19.1 19.5 0.4 -7.7 6.8 % E from Carbohydrate 47.2 42.9 4.3 -8.9 17.4 % E from Total Fat 28.3 31.3 -3.0 -16.0 10.0 Saturated Fat 10.3 11.6 -1.3 -7.1 4.6 PUFA 4.8 5.6 -0.8 -5.8 4.2 MUFA 10.7 11.3 -0.6 -6.9 5.6 % E from Alcohol 3.6 3.8 -0.2 -8.2 7.8
Mean Energy Intake
1600014000 12000 10000 8000 6000 4000
Energy Difference (WFR-FFQ) 8000
6000
4000
2000
1000
-2000
-4000
-6000
-8000 -10000
Figure 1. Energy intake difference against mean energy
- 2SD
+ 2SD
Mean
Use of a food frequency questionnaire in a clinical trial 322
of all nutrient intakes in this study were comparable to those found in studies conducted in the Italy, Japan and Denmark.13-15 Tjønneland et al., (1991) in a study of 144 subjects comparing a self administered FFQ (92 foods and 40 portion-size photographs) and two 7 day WFR, reported correlations ranging from 0.17 for vitamin A to 0.71 for calcium, for a selected group of 14 nutrients. On average, 70% of subjects were classified in the same (+/-1) quintile.15 In a study of 395 subjects Declari et al., (1996) compared a 77 item FFQ with two 7 day dietary records and found higher correlation in all nutrient intakes, compared to the present study, with the highest and lowest correlations found in percent energy from fat (r = 0.35) and percent energy from alcohol (r = 0.78) respectively.13 Similar to our findings, the correlation for β-carotene was low and for alcohol was high. Shimizu et al., (1999) in a study of 117 subjects comparing a 169 item FFQ with 3 day food records and four 24hr recalls reported correlations comparable to our findings.14 The German part of the EPIC study compared twelve 24hr dietary recalls with values from two FFQs (158 food items) and found higher correlations compared to our findings.16 The inter-individual variation in almost all nutrient intakes was higher with the FFQ than with the WFRs. This is similar to findings by Tjønneland et al and Decarli et al.,13,15 suggesting that perception of intake may add additional variability to the FFQ data. The under-estimation of carbohydrate observed is of concern parti-cularly given the comparable results observed for other nutrients suggesting that some key foods may be missing from this FFQ. It has a truncated upper range of fre-quency categories (3 or more times) which may have reduced the intakes of some high carbohydrate foods e.g drinks, rice, pasta, potatoes and biscuits. It does not include some common food items such as soft drinks or some popular low fat snack items e.g muesli bars which may also have influenced the results seen for carbo-hydrate. Serve size used in data analysis may also be an influential factor. The design of the FFQ was such that there were photographs for serve size information for potatoes, vegetables, steak and casserole, but no serve size information was obtained for cereals, snacks and sweets. The FFQ did not allow subjects with the same frequency of intake but different portion sizes to choose from a variety of portion sizes; hence reducing the sensi-tivity of the FFQ. All of these factors may have contri-buted to the underestimation of carbohydrate. The data-base that both the FFQ and WFRs were analysed with was developed more than 7 years ago, and since then portion sizes of some foods have changed. For example, a slice of bread in the database weighs 28g, while a slice of commonly available bread weighs 35-45g. Because the FFQ was optically scanned and the results computer-generated, the serve size for a slice of bread would be significantly smaller than what would have been recorded in the WFRs. This may also have contributed to the lower estimated intake of carbohydrate by the FFQ. This under-estimation of carbohydrate resulted in an overestimation of percent energy from saturated fat, PUFA and MUFA when absolute intakes of fatty acids were comparable to that estimated by the WFRs.
Overall, the FFQ was able to classify more than two thirds of subjects within ±1 quintile difference, a finding that is similar to that reported by Hodge et al., 2000 in a validation study using the same FFQ, and also studies conducted by Tjønneland et al., 1991 (>70%) and Pietinan et al.,(72%).15,17 This implies that FFQs are good tools to use for classifying subjects into quintiles of intake. It must, however, be born in mind that this result does not show the agreement between the absolute values esti-mated by the two methods. To measure the agreement between the two methods, the Bland and Altman method was applied.10 The analysis makes no assumption that one method is superior to another; it merely measures the level of agreement. From Figure 1, the variation (shown by the SD) around the mean was very large, as much as 5MJ, although the mean difference was near zero. Table 3 shows the 95% limits of agreement for all nutrients – all of which have variations too large to suggest the use of the FFQ to evaluate individual dietary intake. This means that the FFQ cannot replace the WFR for the assessment of an individual’s intake in this population. This was similar to the findings from a validation study carried out by Hodge et al.8 In order to test the ability of the ACCV FFQ to assess group nutrient intake in the context of the present study, the subjects’ mean nutrient intakes compared were com-pared according to the 4 dietary intervention groups. No significant differences were found between the groups, suggesting that this FFQ was comparable in assessing group intake when compared to WFR. The results remained the same after exclusion of under-reporters. It is interesting to note that although the correlation of alcohol intake from the two methods was the highest (r = 0.78) among the other nutrients, the ACCV FFQ identified 20% more subjects who drink alcohol than did the WFR. Subjects who do not drink alcohol on a regular basis (e.g. only on social occasions) could account for this finding. This suggests that a FFQ may be more appro-priate for nutrients that are not consumed on a regular basis, such as alcohol.18 Another nutrient that might be better captured by the FFQ is β-carotene. Studies have shown that the longer the WFRs are kept, the better the correlation between β-carotene estimated by the WFRs and the biochemical measurement. The FFQ in this case may give a more accurate figure as it covers a greater time period. If plasma β-carotene was available it would be possible to see which gave better correlation. One of the strengths of this study is that the subjects were not required to adhere to prescribed diets; the nutrient intakes thus reflect their usual diet. However the need for regular consumption of yoghurt may have altered their dietary intake somewhat. Also, the act of recording or weighing may in itself introduce dietary changes by increasing consciousness of what is being eaten, so it is likely that a FFQ may be a better tool to assess usual dietary intake.19 On the other hand, FFQ rely on perception of intake rather than actual intake which could potentially introduce errors. Efforts were made to ensure accurate recording of the food records – weighing scales were provided for those without accurate weighing apparatus, a 1-day practice record was conducted before the actual recording, and the
323 PX Xinying, M Noakes and J Keogh
records were checked by either a dietitians or student dietitian with the subjects for accuracy and clarification. From observation, none of the subjects had any difficulty completing the questionnaire. Few validation studies have attempted to identify under-reporters,20 as did this study, although no significant differences were found after exclusion of under-reporters.
In conclusion all nutrient estimates by the FFQ are within ± 20% of the estimates produced from the mean of the 3 day WFR and that the group means obtained for all nutrients were comparable with the exception of energy, carbohydrate and percent energy from carbohydrate. It is appropriate to use this FFQ to estimate group intake in clinical trial populations however it cannot be used instead of WFR for estimation of an individual’s dietary intake. Acknowledgements We wish to acknowledge the support of Associate Professor Lynne Daniels and Dr Elaine Bannerman from the Department of Nutrition and Dietetics, School of Medicine, Flinders University of South Australia. References 1. Baghurst KL, Baghurst PA. The measurement os usual
dietary intake in individuals and groups. Transactions of the Menzies Foundation 1981; 3:139-160.
2. Marr JW. Individual dietary surveys: purposes and methods. World Rev Nutr Diet 1971; 13:105-164.
3. Bingham SA, Cassidy A, Cole TJ, Welch A, Runswick SA, Black AE, Thurnham D, Bates C, Khaw KT, Key TJ, Day NE. Validation of weighed records and other methods of dietary assessment using the 24 h urine nitrogen technique and other biological markers. Br J Nutr. 1995 Apr; 73 (4): 531-50.
4. Baghurs t KI. The food frequency technique and its relevance to population surveys in Australia - a commentary. Aust J Nutr Diet 1992; 49: 101-3
5. Horwath CC. Food frequency questionnaires: A review. Aust J Nutr Diet 47, 71-76. 1990.
6. Smith W, Mitchell P, Reay EM, Webb K, Harvey PW. Validity and reproducibility of a self-administered food frequency questionnaire in older people. Aust NZ J Public Health 1998; 22: 456-63.
7. Wheeler CE, Rutishauser IHE O'Dea K. Comparison of nutrient intake data from two food frequency question-naires and weighed records. Aust J Nutr Diet 1995; 52: 140-148.
8. Hodge A, Patterson AJ, Brown WJ, Ireland P, Giles G. The Anti Cancer Council of Victoria FFQ: relative validity of nutrient intakes compared with weighed food records in young to middle -aged women in a study of iron supplementation. Aust NZ J.Public Health 2000; 24:576-83.
9. Ireland P, Jolley D, Giles G, O'Dea K, Powles J, Rutishauser I, Wahlqvist ML, Williams J. Development of the Melbourne FFQ: a food frequency questionnaire for use in an Australian prospective study involving an ethnically diverse cohort. Asia Pac J Clin Nutr 1994; 3: 19-31.
10. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307-10.
11. Black AE. The sensitivity and specificity of the Goldberg cut-off for EI: BMR for identifying diet reports of poor validity. Eur J Clin Nutr 2000; 54:395-404.
12. Goldberg GR, Black AE. Assessment of the validity of reported energy intakes - review and recent developments. Scand J Nutr 1998; 42: 6-9.
13. Decarli A, Franceschi S, Ferraroni M, Gnagnarella P, Parpinel MT, La Vecchia C, Negri E, Salvini S, Falcini F, Giacosa A . Validation of a food-frequency questionnaire to assess dietary intakes in cancer studies in Italy. Results for specific nutrients. Ann Epidemiol 1996; 6:110-8.
14. Shimizu H, Ohwaki A, Kurisu Y, Takatsuka N, Ido M, Kawakami N, Nagata C, Inaba S. Validity and repro-ducibility of a quantitative food frequency questionnaire for a cohort study in Japan. Jpn J Clin Oncol 1999; 29: 38-44.
15. Tjonneland A, Overvad K, Haraldsdottir J, Bang S, Ewertz M, Jensen OM. Validation of a semiquantitative food frequency questionnaire developed in Denmark. Int J Epidemiol 1991; 20: 906-12.
16. Bohlscheid -Thomas S, Hoting I, Boeing H, Wahrendorf J. Reproducibility and relative validity of energy and macronutrient intake of a food frequency questionnaire developed for the German part of the EPIC project. European Prospective Investigation into Cancer and Nutrition. Int J Epidemiol 1997; 26 Suppl 1: S71-S81.
17. Pietinan P, Hartman AM Haapa E. Reproducibility and validity of dietary assessment instruments. A self-administered food frequency questionnaire with a portion size picture booklet. Am J Epidemiol 1998; 128: 655-666.
18. Giovannucci E, Colditz G, Stampfer MJ, Rimm EB, Litin L, Sampson L, Willett WC. The assessment of alcohol consumption by a simple self-administered questionnaire. Am J Epidemiol 1991; 133: 810-7.
19. Margetts BM, Cade JE, Osmond C. Comparison of a food frequency questionnaire with a diet record. Int J Epidemiol 1989; 18: 868-73.
20. Brunner E, Stallone D, Juneja M, Bingham S, Marmot M. Dietary assessment in Whitehall II: comparison of 7 d diet diary and food-frequency questionnaire and validity against biomarkers . Br J Nutr 2001; 86: 405-14.
Asia Pac J Clin Nutr 2004;13 (4):324-329 324
Original Article Body mass status of school children and adolescents in Kuala Lumpur, Malaysia Foong Ming Moy MSc, MMedScPH, Chong Ying Gan MBBS, MPH, MD and Mohd Kassim Siti Zaleha BSc
Department of Social & Preventive Medicine, Faculty of Medicine,University of Malaya, 50603 Kuala Lumpur, Malaysia.
Lifestyle and disease patterns in Malaysia have changed following rapid economic development. It is important to find out how these changes have affected the nutritional status and health behaviour of the population, especially school children and adolescents. Therefore a survey on school childrens' and adolescents' health behaviours and perception in Kuala Lumpur was initiated. This paper only reports the observed body mass status of the school children. A total of 3620 school children were selected in this survey using the method of multi-stage sampling. The students were surveyed using pre-tested questionnaires while weight and height were measured by the research team in the field. Using the cut-off of BMI-for-age > 95th percentile and <5th percentile for overweight and underweight respectively, there were a total of 7.3% of overweight students and 14.8% of underweight students. When analysed by gender; 7.5% of boys and 7.1% girls were overweight, while 16.2% of the boys and 13.3% of the girls were underweight. The youngest age group (11 years old) had the highest prevalence of underweight as well as overweight. With increasing age, the prevalence of underweight and overweight decreased and more children were in the normal weight range. The overall prevalence of overweight among the three ethnic groups was similar. However the prevalence of underweight was highest among the Indian students (24.9%), followed by Malays (18.9%) and Chinese (9.5%) (P <0.001). The results showed that both the problems of under- and over-nutrition co-exist in the capital city of Malaysia. The promotion of healthy eating and physical activities is required to address the problems of under- and over-nutrition in order to build up a strong and healthy nation in the future.
Key Words: body mass status, school children, BMI-for-age, overweight, underweight, Kuala Lumpur, Malaysia Introduction In the last decade, there has been very rapid economic and industrial development in Malaysia and living standards have risen to those of developed countries in many areas. This is particularly true in the case of Kuala Lumpur, Malaysia’s capital city. With rapid socio-economic deve-lopment, lifestyle changes and problems related to phy-sical, behavioural and mental health, become common. Malaysia has a large proportion of young people and school-age children (5-20 years) which constitute about 20 percent of its total population.1 With such a large pro-portion of young people, it becomes important to know how economic development has affected their lifestyles and health behaviours. For example, exposure to mass media coupled with economic affluence could well influ-ence the attitudes, social activities, physical activities and food habits of school aged children.2 The body mass status data presented in this article are part of the findings of a larger survey conducted to gain insight into the health behaviours and perceptions of adolescents in the schools of Kuala Lumpur. Three groups of school children were surveyed, they were in Standard 5 (estimated average age 11 years), Form 2 (estimated average age 14 years) and Form 4 (estimated average age 16 years).
Materials & Methods Formal permission to conduct the survey was obtained from the Ministry of Education of Malaysia. Participation by the school children was voluntary and the self-administered questionnaires were answered anonymously where the children were not required to write their names in the questionnaires. Teachers were asked to leave the classrooms and not allowed to distribute or collect the questionnaires to ensure confidentiality, anonymity and absence of intimidation. This process was conducted by the research team members and the questionnaires were taken back to the research office as soon as the survey in the schools completed.
Sampling method The students were selected using the method of multi-stage sampling.3 All schools in Kuala Lumpur (both government and private) were listed. There were a total of 17 private schools and 212 government schools (figures from the Correspondence address: Department of Social & Preventative Medicine, Faculty of Medicine, University of Malaya, 50603, Kuala Lumpur, Malaysia Tel: + 603-79676657; Fax: + 603-79674975 Email: [email protected] Accepted 28 May 2004
325 FM Moy, CY Gan and MK Siti Zaleha
Ministry of Education). These private schools in general, catered for children from higher socio-economic back-ground and were not included in the sampling frame. Schools for the handicapped were also excluded. All government funded schools (153 primary and 59 secon-dary) scattered in all the six administrative zones of Kuala Lumpur were included in the sampling frame. In order to obtain a sample of school children reflecting as close as possible the population characteristics of Kuala Lumpur, three major factors were taken into consideration in sampling: gender, ethnicity and socio-economic status. As the number of primary and secondary schools differed, primary and secondary schools were sampled separately, but in a similar way. Schools were stratified into ‘Malay’, ‘Chinese’ and ‘Indian’ schools based on the majority of children of that ethnicity in the school. The locations of the schools in each zone were noted and stratified into those in “higher income areas” and “lower income areas”. To ensure that boys and girls were in-cluded in almost equal proportions, schools of each gender were selected in equal proportions. Coeducational schools were also selected on the assumption that boys and girls were in equal proportions. Schools in higher income and lower income areas were selected in approximately equal numbers from each zone. Schools of each ethnicity were also selected to reflect the ethnic proportions of Malaysia. Sample size estimation The programme Epi Info 6 was used to estimate sample size for the population survey. Estimating that the pre-valence of obesity to be 10% in the population and with an allowance of a 5% precision of the estimate and a confidence level set at 95%, the total sample sizes required for all 3 ethnic groups together for Standard 5 was estimated to be 1240 taking into consideration the ethnic proportions in the country. Using the same method, the sample size required for Form 2 and Form 4 were calculated at 1070 each. It was estimated that a total of 3380 school students would be required for the survey. Data collection A self-administered questionnaire was designed to collect the data. The national language (Malay) was used. The questionnaire was pre-tested among some school children in the various grades in the adjacent towns of Petaling Jaya and Kajang. To ensure that the questionnaires were well completed, the investigators were present in the class and read the questions aloud in sequence so that the
questions were fully understood and any queries could be answered. For Indian and Chinese schools, the questions were also translated aloud. The weights of the students without shoes were taken using a bathroom scale while heights were measured using a microtoise. The weighing scale was calibrated daily with standard weight before the trips to the schools. Trained field workers took both the measurements of weight and height. In this study, BMI-for age (Body Mass Index) was used as the anthropometric indicator for the nutritional status of the students. The values of BMI-for-age used were based on the reference data of the WHO report.4 A child was considered underweight or having low BMI-for-age when his BMI-for-age was <5th percentile and overweight when his BMI-for-age was > 95th percentile. Statistical analysis All the variables were coded and entered into SPSS for Windows version 10.0. Appropriate statistical analyses were performed using the same software. The significant level was preset at 0.05. Chi-square test for categorical data was used in the analysis. Results Demographic characteristics A total of 3620 students from 29 schools from all 6 administrative zones of Kuala Lumpur responded with only 13 students declining to participate. Of the 3620 respondents, there were 1871 (51.7%) Malays, 1244 (34.4%) Chinese, 441 (12.2%) Indians and a small group of 64 (1.8%) were of other ethnic groups. Only the three main ethnic groups in Peninsula Malaysia totalling 3556 students were included in the analysis as the numbers in minority groups were too small for comparisons to be made. The sample was reflective of the demographic characteristics of Peninsular Malaysia in terms of ethnic distribution and gender proportions. Table 1 shows the distribution of respondents by ethnic group, age groups (indicated by the level of education) and gender. The students had come from a range of socio-economic backgrounds (as indicated by their fathers' occupations5) as shown in Figure 1. Most of the younger students had professional fathers compared to the older age groups. This reflects the recent development of Kuala Lumpur city. The majority of students (43.7%) came from families with 4 or 5 siblings, approximately 20.8% had more than 5 siblings in the family, and 35.5% had 3 or fewer siblings.
Table 1. Number of respondents by ethnicity, gender and age group
Malay Chinese Indian Others Total
Class Boys Girls Boys Girls Boys Girls Boys Girls
Std 5 308 279 294 235 113 78 6 7 1320
Form 2 322 368 182 170 58 69 10 15 1194
Form 4 276 318 205 158 62 61 14 12 1106
Total 906 965 681 563 233 208 30 34 3620
Body mass status of school children and adolescents in Kuala Lumpur, Malaysia 326
Figure 1. Fathers’ occupations by students’ schooling level Body mass status Based on the cut-offs of < 5th and > 95th percentile of the BMI-for-age, it was found that the problems of under-weight and overweight appeared side by side among the children. A total of 7.3% of the students were overweight and 14.8% were underweight. When analysed by gender; 7.5% of the boys and 7.1% of the girls were overweight, while 16.2% of boys and 13.3% girls were underweight. The youngest age group (Standard 5) had the highest prevalence of underweight (16.1%) as well as overweight (10.1%). As the children’s age increased, the prevalence Table 2. Students’ body mass status by schooling levelsFigure 2. Body mass status of students by gender and schooling levels of underweight and overweight decreased and more of them were in the normal weight range (Table 2). When the students’ weight status was further analysed according to gender, there were about similar proportions of Standard 5 (age 11 years) boys and girls who were under-weight, but more boys were overweight. There was a descending trend in this prevalence as their age increased. However, fewer girls were over- and under-weight. At Form 4 (16 years), the boys and girls had similar pre-valences of overweight, but there were still more under-weight boys as compared to girls at that age (Fig. 2). The overall prevalence of overweight among the three ethnic groups was similar: 7.8% for the Malay students, 6.7% and 7.0% for the Chinese and Indian students respectively. However the overall prevalence of under-weight was highest among the Indian students (21.1%), followed by Malays (15.9%) and Chinese (10.9%); and the difference observed was statistically significant (P <0.001). When these rates were further analysed by gender, it was found that the prevalence of overweight for the boys was 7.6% among Malays, 7.6% among Chinese and 6.9% among Indians; while for the females it was
��������������������������������������������������
����������������������������������������������������������������
����������������������������������������������������������������������
�������������������������������������������������������
���������������������������������������������
�����������������������������������
���������������������������
������������������������
����������������������������������������
24.6
8.6 9.6
16.2
28.2 23.7
32.4 35.9 37.2
19.1 17.7 16.1
7.6 9.6 13.4
0 5
10 15 20 25 30 35 40
Standard 5 Form 2 Form 4 Schooling levels
Percent
Professional ����� Skilled Semi-skilled ��� Unskilled ������Pensioners/others
0 2 4 6 8
10 12 14 16 18 20
Std 5 Form 2 Form 4
Percentage
Underweight Boys Underweight Girls Overweight Boys Overweight Girls
Underweight
N (%)
Normal weight N (%)
Overweight
N (%)
Std 5
221(16.1) 964 (73.8) 132 (10.1)
Form 2
176 (15.1) 927 (79.3) 66 (5.6)
Form 4
138 (12.8) 880 (81.5) 62 (5.7)
Total
535 (14.8) 2771 (77.9) 260 (7.3)
�����������������������������������
��������������������������������������������
16.9
17.8
14.3
������������������������
������������������������������������
9.5 7.7
9.4
��������������������
2
2
0
10
20
30
40
50
60
70
p e r c e n t a g e
Malay Chinese Ind
Underweight* Std 5
����
������������������������
���������������������������������������������
9.2
0.5
15.4
����������
12.7
��������������������3.7 6.5 ������
�����������������
11.6
����������������������3.8 5.4 �������
�������������������������������
7.1
������������������������6.9 6.5
ian Malay Chinese Indian Overweight
���Form 2 Form 4
Figure 3. Body mass status of boys by ethnicity and schooling levels * denotes statistical significance among races (P<0.001)
327 FM Moy, CY Gan and MK Siti Zaleha
8.0% among Malays, 5.5% among Chinese and 7.2% among Indians. There were no outstanding differences among the three ethnic groups. On the other hand, the overall prevalence of underweight among boys was 18.9% among Malays, 9.5% among the Chinese and 24.9% among Indians. The overall prevalence of under-weight among girls was 13.1% among Malays, 12.4% among Chinese and 24.9% among Indians. Indians had the highest prevalence of underweight among both boys and girls, however this difference was only statistically significant (P<0.001) among the boys. Analysis of the rates of underweight and overweight by race, gender and schooling levels was also carried out. It was found that the Indian boys had the highest prevalence of underweight in all age groups while there was no consistent trend in the prevalence of overweight among the boys of all age groups (Fig.3). The Malay girls
food supply may have caused overeating and over-consumption of nutrient dense food such as high calorie and high fat food.9 The aggressive promotion of fast food to children and adolescents may have been another cause of regular or frequent consumption of fast food which is high in fat and high calories. Families with double incomes have greater buying power and less time for home cooking. This may result in more frequent eating out and higher intakes of energy dense, high fat, nutrient poor (especially saturated fat) foods.9,10 In this study where students were between 11 and 16 years, 7.5% boys and 7.1% girls were considered over-weight under the specified criteria of the study. The prevalence rates of overweight among the three ethnic groups were quite similar which fell in the range of 6.7% to 7.8%. A similar study conducted by Kasmini et al.,11 on school children aged 7 to 16 years old in Kuala Lumpur, gave an overall prevalence of overweight of
Figure 4. Body mass status of girls by ethnicity and schooling levels had the highest prevalence of overweight and the Chinese girls had the lowest prevalence of overweight in all age groups while the Indian girls had the highest prevalence of underweight in Forms 2 and 4 (Fig. 4). Discussion Rapid socio-economic development is associated with changes in the lifestyle of the community especially those from the urban areas.6 These include changes in physical activities and food consumption patterns.6,7 Kuala Lumpur as the capital city of the country is well developed, how-ever the influx of migrants from the rural areas seeking employment opportunities has created areas of squatters in certain areas within the city. Therefore problems of under-nutrition and over-nutrition emerged side by side in our findings. This problem is commonly faced by deve-loping or newly developed countries in Asia.7 The problem of overweight might be caused by physical inactivity where the urban children indulged more on inactive leisure time activities such as television viewing, playing computer games and surfing the net. Spending more time sedentarily seemed to have a positive relationship with obesity.8 In addition, the abundance of
9.5% with Indian students being the group with the highest percentage (12.0%), followed by the Chinese (9.9%) and the Malays (8.9%). This difference could be due to the different age groups selected in Kasimini’s study where his sample was younger (mean age of 11.8 years). The prevalence of overweight for the Standard 5 students (average age 11 years) was 11.3% for boys and 8.6% for girls and our results are comparable to another study conducted among the 7 – 10 year old students from the primary schools in Kuala Lumpur12 with 9.7% of boys and 7.1% of girls being overweight. There were less girls in the overweight category; this could be due to their earlier growth spurts compared to the boys (WHO technical report).4 For adolescents aged 14 (Form 2) to 16 years (Form 4), the prevalence of overweight was lower at 5.6% and 5.7% respectively. Similar results were reported in Kasmini’s study11 with the rates of 3.3% to 9.5%. This could be due to the fact that before puberty, these overweight/obese children were able to ‘grow into’ their desirable weight as they had their growth spurts. Overweight and obesity among school children has been reported in many Asian countries. For example, the Singapore School Health Survey13 reported that the prevalence of obesity in the year 2000 was 14.7% in those aged 12- 13 years and 13.1% in those aged 15-16 years. In addition, a study conducted among the 7 – 9 years old school children in the Northeast of Thailand, urban Khon Kaen14 reported an obesity prevalence of 10.8%. The prevalence of overweight in our study was considerably lower than among Singapore school chil-dren13 but this rate is higher than among Malaysian rural children.15 If action is not taken to address the problem, these rates are likely to escalate. Existing data indicates that childhood and adolescent obesity tends to predict adult obesity, and overweight children are more likely to become obese adults.16,17 Epidemiological data has shown that after adjustment for parental obesity, the odds ratio for obesity in adulthood associated with childhood obesity at 15–17 years of age was 17.5 (95% confidence interval, 7.7 – 39.5).18 A compilation of body mass index (BMI) of Malaysian adults by Ismail et al.,19 in urban areas showed that 29% of males and 26% of females were overweight and 5% of
������������������������������
������������������������������������������������
13.3
15.5
10.1 ���������������
���������������������������������
19.6
��������������������6.5
8.2
��������������������
���������������������������������������������
19.2
15.9
14.8
���������
���������������������������
10
8.2
6
������������
������������������������
6.8 4.7 4.4
��������������������
���������������������������������
9
7.2 4.9
0 5
10 15 20 25 30 35 40 45 50
P e r c e n t a g e
Malay Chinese Indian Malay Chinese Indian Underweight Overweight
Std 5 �����
Form 2 Form 4
Body mass status of school children and adolescents in Kuala Lumpur, Malaysia 328
males and 8% of females were obese. In the rural popu-lation, 21.4% were found to be overweight and 6.5% obese.20 These figures suggested that a large proportion of Malaysian adults are overweight and so at higher risk for chronic diseases such as type 2 diabetes, hypertension and hyperlipidemia.21 These di-seases are also more common among children who are obese.22-25 The pre-vention of obesity in Malaysian adults could well begin with making sure that Malaysian children and adolescents do not fall into the overweight category. Our findings also showed that 16.2% of boys and 13.3% of girls were underweight. Underweight was most prevalent in the 11 years age group (Standard 5) and became less prevalent in the older age groups. The 11 year old girls and the 14 year old boys were found to have the highest rates of underweight, possibly because of the different timing of their adolescent spurts. However, low socio-economic status (with lower income and more children in the family) may also have contributed to the proportions of underweight students as poverty still exists in the midst of mainstream affluence in Kuala Lumpur. The Indian students were found to have the highest prevalence of underweight among the 3 ethnic groups and this might be due to genetic factors. In view of the co-existence of underweight and over-weight in Kuala Lumpur, the authorities should address the problems through the education of parents and chil-dren in healthy lifestyles via the mass media and school settings. Healthy lifestyle such as healthy eating habits and the encouragement of physical activities in school or leisure time should be promoted among all age groups in the community. Measures to reduce poverty, such as the provision of better or cheaper housing, skill training for the unemployed or unskilled workers, promotion of family planning and the spacing of children, etc should be provided for the socio-economically deprived people. These measures should be implemented since both under- and over-nutrition will give rise to related diseases which will increase health care costs. School children are the future citizens and their health is essential to the country. Conclusion Under and over-nutrition among the school age children is currently a health problem faced by Malaysia. Appro-priate steps need to be taken to address these problems in order to build up a strong and healthy nation in the future. Acknowledgement The authors gratefully acknowledge the financial support of the Vote F research funding from the University of Malaya. The authors would also like to thank all the school children, and the staff of the Department of Social & Preventive Medicine, who participated in this survey. The Ministry of Education’s granted permission for the conduct of the survey is greatly appreciated. References 1. Department of Statistics, Malaysia, Vital Statistics,
Peninsular Malaysia, 2000. 2. Berkey CS, Rockett HR, Field AE. Activity, dietary intake
and weight changes in a longitudinal study of pre-adolescent and adolescent boys and girls. Paediatrics 2000; 4: 105.
3. Abraham JH. Sampling. In: Survey methods in community medicine. Epidemiological studies, programme evaluation, clinical trials. New York: Churchil Livingston, 1999; 89-103.
4. World Health Organisation. Adolescents. In: Physical status: the use and interpretation of anthropometry. Technical Report Series no.854. Report of a WHO Expert Committee, Geneva 1995; 263-309.
5. Abraham JH. Defining the variables. In: Survey methods in community medicine. Epidemiological studies, programme evaluation, clinical trials. New York: Churchil Livingston, 1999; 123-131.
6. Tee ES. Obesity in Asia: prevalence and issues in assessment methodologies. Asia Pac J Clin Nutr 2002; 11 (Suppl): S694-S701.
7. Rodolfo FF. The burden of obesity in Asia: challenges in assessment, prevention and management. Asia Pac J Clin Nutr 2002; 11 (Suppl): S676-S680.
8. DiPerto L. Physical activity, body weight and adiposity: an epidemiologic perspective. Exer Sport Sci Rev 1995; 23: 275-303.
9. Tee ES. Nutrition of Malaysians. Mal J Nutr 1999; 5:87 – 109.
10. International Life Sciences Institute. Overweight and obesity in European children and adolescents causes and consequences – prevention & treatment. Belgium: ILSI. 2000.
11. Kasmini K, Idris MN, Fatimah A, Hanafiah S, Iran H, Asmah Bee MN. Prevalence of overweight and obese school children aged between 7 to 16 years amongst the 3 ethnic groups in Kuala Lumpur, Malaysia. Asia Pac J Clin Nutr 1997; 6 (3): 172-174.
12. Tee ES, Khor SC, Ooi HE, Young SI, Zakiyah O, Zulkifli H. Regional study of nutritional status of urban primary school children. Food Nutr Bull 2002; 23 (1): 41-47.
13. School Health Services, Annual Report. Singapore: 2000. 14. Langendijk G, Wellings S, Van Wyk M, Thompson S,
McComb J, Chusilp K. The prevalence of childhood obesity in primary school children in urban Khon Kaen, Northeast Thailand. Asia Pac J Clin Nutr 2003; 12 (1): 66-72.
15. Khor GL, Tee ES. Nutritional assessment of rural villages and estates in Peninsular Malaysia. II. Nutritional Status of children 18 years and below. Mal J Nutr 1997; 3: 21-47.
16. United States Department of Health and Human Services. National Centre for Health Statistics. Centres for Disease Control & Prevention. Prevalence of obesity among adults aged 20 years ad over: United States, 1997 -2001. Hyattsville, MD: 2002.
17. Guo SS, Huang C, Maynaud LM. Body mass index during childhood, adolescent and young adulthood in relation to adult overweight and adiposity: The FELS longitudinal study. Int J Obes Relat Metab Disord 2000; 24: 1628 – 1638.
18. Sugimori H, Yoshida K, Mitakawa M, Izuno T, Takahashi E, Nanri S. Temporal course of the development of obesity in Japanese school children: a cohort study based on the Keio study. J Pediatr 1999; 134: 749 – 754.
19. Ismail MN, Zawiah H, Chee SS, Ng KK. Prevalence of obesity and chronic energy deficiency (CED) in adult Malaysians. Mal J Nutr 1995; 1: 1 –9.
20. Mohd Aminuddin MS, Yusoff K, Osman BA, Khalid BAK. Is coronary heart disease potentially a serious public health problem in rural Malaysia? Quarterly Scientific Meeting, Academy of Med Malaysia, Penang 1993.
329 FM Moy, CY Gan and MK Siti Zaleha
21. Wellman NS, Friedberg B. Causes and consequences of adult obesity: health, social and economic impacts in the United States. Asia Pac J Clin Nutr 2002; 11(Suppl): S705-S709.
22. Gidding SS, Bao W, Srinivasan SR, Berenson GW. Effects of secular trends in obesity on coronary risk factors in children: the Bogalusa Heart Study. J Pediatr 1995; 127: 868-874.
23. Clarke WR, Woolson RF, Lauer RM. Changes in ponderosity and blood pressue in childhood: the Muscatine Study. Am J Epidemiol 1986; 124: 195-206.
24. Shinha R, Fisch G, Teague B, Tamborlane WV, Banyas B, Allen K, Savoye M, Rieger V, Taksali S, Barbetta G, Sherwin RS, Caprio S. Prevalence of impaired glucose tolerance among children and adolescents with marked obesity. N Engl J Med 2002; 346: 802-810.
25. Pinas-Hamiel O, Dolan LM, Daniels SR, Standiford D, Khoury PR, Zeithle P. Increased incidence of non-insulin-dependent diabetes mellitus among adolescents. J Pediatr 1996; 128:608-615.
Asia Pac J Clin Nutr 2004;13 (4):330-335 330
Original Article
Haematocrit levels and anaemia in Australian children aged 1-4 years Dorothy EM Mackerras MPH, PhD1, Susan I Hutton MSc(Med)1 and Philip R Anderson BSc(Hons), PhD2 1 Menzies School of Health Research, and Institute of Advanced Studies, Charles Darwin University, Darwin Northern Territory, Australia 2 Australian Institute of Health and Welfare, Canberra, Australia
The aim of this study was to describe the prevalence of anaemia, mean haematocrit levels, and the risk factors influencing haematocrit in participants of the 1995 National Survey of Lead in Children. A nationally-representative cross-sectional survey of children aged 1-4 years inclusive was done. Mean haematocrit and the proportion with anaemia using both the US and WHO haematocrit-based criteria were calculated. Multivariate regression was used to identify factors associated with haematocrit. Mean haematocrit level was 38.8% (95% CI: 38.6 - 39.1%) and varied with age of child, state/territory of residence and whether the child was taking supplements. It did not vary by sex, Aboriginal identification, maternal birthplace, whether the child ate meat or any other selected characteristic. The factors identified explained only 4% of the variation in haematocrit levels. The prevalence of anaemia was 3.3% (95% CI: 2.4 - 4.5%) based on the US criteria and 2.0% (95% CI: 1.3 - 3.1%) based on the WHO criteria. The prevalence of anaemia in this national survey was lower than the prevalence of iron deficiency anaemia reported in several more localised studies.
Key words: anaemia, haematocrit, national survey, children, Australia Introduction World-wide, iron deficiency anaemia is the most common nutritional deficiency. Generally, the highest prevalence is found in pre-school aged children, adolescents and women of reproductive age.1 Several localised studies have been conducted in Australia in recent years. Karr et al., found a prevalence of 1.1% (95% CI: 0.1-2.1%) in a representative sample of children aged 9-62 months living in the central and southern Sydney areas, after excluding 0.8% with thalassaemia.2 This was highest (3%) in 2-year old children, but fell to 0% in children aged 3 years and older. A subsequent study of children from the same area who had mothers born in Arabic countries, found a prevalence of 6% after excluding 5.5% with haemoglobinopathies.3 By contrast, a study from Adelaide reported that the pre-valence of iron deficiency anaemia was 6% in Caucasians aged 6-24 months. Although this study also reported a higher prevalence in Asian than Caucasian children, the children were a non-representative sample and, in parti-cular, the Asian children were recruited if the “workers were concerned that they may have a high risk of iron deficiency anaemia”.4 A small number of studies have found a higher prevalence of anaemia in Aboriginal children in rural areas.5-7 As each of the above surveys was conducted in a single location, their generalisability to the wider Australian pre-school-aged population is uncertain. None of these studies examined risk factors for iron deficiency anaemia owing to its low prevalence. Instead, the risk factors for iron depletion, which has a higher prevalence, were examined. In the general Sydney sample,
there were significant associations with age, non-use of supplements and eating meat less than four times per week.2 An earlier case-control study from the same area found associations with low intake of haeme iron and high intake of cow milk.8 In the Arabic sample, the risk factors were preterm delivery, mother having migrated to Australia within the previous eight years and high intake of cow's milk.3 In the Adelaide group, the risk factors were short duration of breastfeeding, early introduction of cow's milk and high intake of cow's milk.4 Thus there is some con-sistency, but also disagreement regarding the risk factors for iron depletion in these studies. Some of the incon-sistency may be related to the differing definitions of iron depletion that were used. To date there has been no national survey of anaemia or iron status in young Australian children. In 1995, the Australian Institute of Health and Welfare conducted a nationally representative survey of Australian children aged 1-4 years inclusive to examine lead exposure.9 Hae-matocrit levels were measured as part of this survey so that statistical analysis could be carried out on both corrected and uncorrected lead levels to maximise opportunities for comparison with other studies.9 Anaemia can be defined Correspondence address: Dr D Mackerras, Menzies School of Health Research, Building 58, Royal Darwin Hospital, Rocklands Drive, Tiwi, NT 0811 Tel: 08-8922-8283; Fax: 08-8927-5187 Email: [email protected] Accepted 23 July 2004
331 DEM Mackerras, SI Hutton and PR Anderson
using haematocrit levels.1,10 Therefore, we took advan-tage of these data to describe the prevalence of anaemia, the average haematocrit level and the factors which influ-ence haematocrit in a representative sample of Australian children. Materials and methods The 1995 National Survey of Lead in Children (NSLIC) is described elsewhere.9 Briefly, sampling frames were constructed by the Australian Bureau of Statistics for non-remote and remote areas separately based on the census collector districts so that all children in Australia aged 1-4 years would have an equal probability of being sampled. In February and March 1995, 175 trained interviewers visited households in the selected areas. The initial inter-view included an extensive questionnaire about factors which might affect lead levels. At a second visit, a blood sample was drawn from all children in the target age range (1-4 years inclusive) for whom parental consent was given. Haematocrit was measured by inductively coupled plasma mass spectrometry at Royal North Shore Hospital, Sydney. In the context of the current analysis, it is worth noting that the children were not selected because they might have high lead levels. The NSLIC dataset contains many variables relating to demographic, environmental, occupational, residential history and household characteristics collected by questionnaires available in English and seven other languages. To avoid finding spurious associations re-sulting from performing multiple tests, a subset of cha-racteristics was selected on a priori grounds. The socio-demographic variables were: child age, sex, jurisdiction of residence, Indigenous identification, highest levels of education and income in the household, mother’s country of birth, presence of a smoker in the household and lead
level in the child. Maternal country of birth was classi-fied as Australia, an English-speaking country (New Zealand, North America and the British Isles) or a non-English-speaking country. The only nutrition-related questions asked were whether the child ate “beef, lamb and/or pork”, whether the child ate “chicken and/or fish”, whether the child took vitamin or mineral supplements and whether the house had a vegetable garden. We excluded data for all blood samples that had any degree of haemolysis or contained clots on arrival at the laboratory. Anaemia was defined in two ways: the World Health Organisation (WHO) definition is haematocrit <33% for children aged 6-59 months1; the United States (US) definition is haematocrit <32.9% for children aged <2 years and <33% for children aged 2-<5 years.10 Haematocrit was normally distributed and so its pre-dictors were examined using multiple linear regression. All variables, including age, were entered as indicator variables, overall 2-sided F-tests were performed and the distribution of the residuals from the final model was checked. However the prevalence of anaemia was so low that only classification within demographic cate-gories was done. As indicated below, the age and jurisdiction of residence distribution did not quite match the 1996 Census,11 and so post-hoc sampling weights were cal-culated to correct this. Analyses were done using Stata 7 (StataCorp, College Station, TX) and allowed for sample weights and clustering within households. Ethical clear-ance for the NSLIC was given by the Australian Institute of Health and Welfare Ethics Committee and for the current analysis by the Joint Institutional Ethics Com-mittee of the Royal Darwin Hospital and the Menzies School of Health Research.
Table 1. Demographic characteristics of the survey population with usable samples compared to the 1996census population aged 1-4 years (unweighted results). _________________________________________________________________________________________________ Characteristic 1996 Census 1995 National Lead Survey in Children Boys Girls % N % N % Age (years) 1 24.9 160 22.9 126 18.8 2 25.0 172 24.6 204 30.4 3 25.1 178 25.5 197 29.3 4 25.0 189 27.0 145 21.6 Total 100 702 100 672 100 Residence New South Wales 33.8 196 28.0 181 26.9 Victoria 24.4 161 23.0 152 22.6 Queensland 18.7 122 17.5 117 17.4 South Australia 7.5 53 7.6 67 10.0 Western Australia 9.7 65 9.3 71 10.6 Tasmania 2.6 60 8.6 47 7.0 Northern Territory 1.4 12 1.7 8 1.2 Australian Capital Territory 1.7 30 4.3 29 4.3 Identification@ Non-Indigenous 97.8 630 90.1 628 93.5 Indigenous 2.2 34 4.9 12 1.8 Missing 0 35 5.0 32 4.8 _________________________________________________________________________________________________ @ based on children aged 0-4 in the 1996 Census
Haematocrit levels and anaemia in Australian children aged 1-4 years 332
Results It is estimated that 4112 children aged 1-4 years lived in the selected areas and 3542 were located.9 Blood samples were taken from 1575 children but only 1371 had useable results. These children came from 1106 households; 947, 193, 10 and two households had one, two, three and four children respectively participating in the survey. Of the non-usable samples, 52 contained clots, 133 were partly or fully haemolysed and 19 samples did not have hae-matocrit determined. Overall, the age of the children with useable results was similar to the 1996 Census distri-bution11 but with a small deficit of 1-year old children (Table 1). The regional distribution was also similar, though with a small deficit of children from New South Wales.11 Probably owing to the small numbers, the distribution of the 46 Aboriginal children did not reflect the national distribution because only 17% lived in NSW, 10% in Qld, 32% in WA and 15% in the NT compared to the national proportions of 30%, 28%, 14% and 13% respectively.12 The prevalence of anaemia using the WHO criteria was 2.0% (95% CI: 1.3-3.1%) overall (Table 2). Using the US criteria, the classification of 18 children changed the overall prevalence of anaemia was 3.3% (95% CI: 2.4-4.5%). The mean haematocrit level was 38.8% and this varied among the jurisdictions (Table 3). In addition, age, child’s lead level and use of supplements and presence of a vegetable garden, had P values near or below 0.05 in the univariate analysis. In the multivariate model, only jurisdiction of residence, age and use of supplements were significant predictors of haematocrit but together they explained only 4% of the variation in haematocrit levels in the population (Table 4). This result did not change if the two territories with small populations were excluded from the analysis.
Discussion This appears to be the first description of the prevalence of anaemia from a national Australian survey of children aged 1-4 years. The prevalence is very low. Including all the haemolysed and clotted samples in the analysis made little difference to the mean haematocrit (for all 783 boys, mean haematocrit was 38.8% and for all 772 girls, the
mean was 38.9%). Therefore we do not believe that ex-cluding these samples has biased our findings. Our results cannot be compared directly to previous Australian studies because the definition of anaemia is based on haematocrit rather than haemoglobin and all anaemias are included in the current analysis. Although haematocrit is used for anaemia screening in other countries, documents that publish cutoff values for both characteristics do not describe how the haematocrit-based prevalence of anaemia relates to the haemoglobin-based prevalence of anaemia.1,10 Graciter et al.,13 used lower cutoffs of 31% haematocrit and 10g/dL haemoglobin to define anaemia to examine this question for children aged 12-23 months seen in a number of surveys. They found that the haematocrit-based definition generally yielded a lower prevalence of anaemia than the haemoglobin-based defi-nition. However the difference was small when the pre-valence was low, e.g 3.3% anaemic by haemoglobin and 3.0% anaemic by haematocrit.13 Therefore our haematocrit-based prevalence of 2.0-3.3% may be a close estimate of the unknown prevalence of anaemia based on haemoglobin levels. We could not exclude hemoglobinopathies or non-iron deficiency anaemia and so, theoretically, our study should have yielded a higher prevalence than previous Australian studies2-4 but we found a lower prevalence. The US hae-matocrit cut-offs are based on the 5th centile of the haematocrit distribution from their Third National Health and Nutrition Examination Survey calculated after ex-cluding “persons who had a high likelihood of iron de-ficiency”.10 Hence the expected prevalence using the US definition is 5% provided that persons who are likely to be iron deficient are excluded from the analysis. We conclude that the prevalence of anaemia is lower in Australia than the US because, firstly, the confidence interval around the prevalence calculated using the US definition excluded the value of 5% and secondly, because we were unable to exclude children with a high likelihood of iron deficiency. The WHO haematocrit cut-off is based on a conversion factor from haemoglobin levels, but the basis of the haemoglobin level chosen is unclear1 and so the expected prevalence is unclear.
Table 2. Proportion of children who are anaemic according to two criteria by demographic characteristics, weighted to the age- and jurisdiction population distribution in the 1996 Census and corrected for clustering within households _________________________________________________________________________________________________ Characteristic WHO criteria US criteria* % anaemic 95% CI % anaemic 95% CI ___________________________________________________________________________________________________________ Total 2.0 1.3 – 3.1 3.3 2.4 – 4.5 Sex Boys 2.5 1.4 – 4.5 4.0 2.6 – 6.0 Girls 1.6 0.8 – 2.9 2.6 1.6 – 4.1 Age (years) 1 1.8 0.7 – 4.3 1.8 0.7 – 4.3 2 2.5 1.3 – 4.7 4.9 3.1 – 7.6 3 1.1 0.4 – 2.7 3.1 1.7 – 5.4 4 2.8 1.4 – 5.4 3.5 1.9 – 6.2 _________________________________________________________________________________________________ * US criterion is the 5th centile of the haematocrit distribution in children with no hemoglobinopathies or iron deficiency
333 DEM Mackerras, SI Hutton and PR Anderson
Intra-person variation affects biochemical parameters and this has not been corrected for either in the references, our work or previous Australian reports. This means that a number of children who had anaemia or iron depletion on one occasion would not have this on a second occasion, even in the absence of any treatment, and is due to a phenomenon called regression to the mean.14 Looker et al.,
report that an initial 10% prevalence of impaired iron status (based on mean corpuscular volume, transferrin saturation and erythrocyte proto-porphyrin) dropped to 4% when corrected for within-person variability.15 There-fore population-based surveys using a single measure-ment occasion for anaemia, iron depletion or deficiency overestimate the true prevalence of these conditions.
Table 3. Mean haematocrit levels according to demographic and lifestyle characteristics, weighted to the age- and jurisdiction population distribution in the 1996 Census and corrected for clustering within households ___________________________________________________________________________________________________________Characteristic N Haematocrit Mean 95% CI P Total 1371 38.8 38.6-39.1 - Sex Boys 699 38.7 38.4-39.1 0.4 Girls 672 38.9 38.6-39.2 Age (years) 1 286 38.5 38.0-39.0 0.02 2 376 38.8 38.4-39.2 3 375 38.7 38.3-39.0 4 334 39.4 38.9-39.9 Residence New South Wales 377 38.6 38.1-39.1 <0.0001 Victoria 313 39.3 38.9-39.7 Queensland 239 38.2 37.6-38.8 South Australia 120 38.3 37.6-38.9 Western Australia 136 39.6 38.6-40.6 Tasmania 107 39.8 38.2-41.4 Northern Territory 20 38.1 36.8-39.3 ACT 59 41.0 39.8-42.2 Identification Non-Indigenous 1260 38.8 38.6-39.1 0.8 Aboriginal 46 39.1 37.6-40.6 Missing 68 38.6 37.8-39.3 Income <$20,000 345 38.7 38.2-39.2 0.7 $20-30,000 305 38.9 38.4-39.4 $30-40,000 308 39.0 38.6-39.4 >$40,000 338 38.9 38.5-39.2 Missing 75 38.4 37.4-39.5 Education Bachelor or higher 310 38.8 38.4-39.2 0.8 Trade 371 38.9 38.4-39.3 Other 290 39.0 38.5-39.5 None/missing 400 38.7 38.3-39.1 Marital status Married/de facto 1138 38.9 38.6-39.1 0.8 Single parent 145 38.8 38.0-39.6 Other 88 38.6 37.8-39.4 Maternal birth Australian born 1045 38.8 38.5-39.1 0.9 English-speaking countries 132 38.8 38.0-39.6 Non-English-speaking 147 38.9 38.4-39.4 Missing/not in household 47 39.0 37.9-40.0 Smoker present no 948 38.9 38.6-39.2 0.5 in household yes 422 38.7 38.3-39.1 Child’s lead level ≤0.49 μmol/L 1275 38.9 38.6-39.2 0.09 >0.49 μmol/L 96 38.3 37.5-39.0 Takes vitamin/mineral no 1179 38.7 38.5-39.0 0.02 supplements yes 188 39.4 38.8-39.9 Eats beef, lamb, no 88 38.5 37.6-39.4 0.4 pork yes 1282 38.9 38.6-39.1 Eats fish, chicken no 36 39.1 37.9-40.4 0.6 yes 1335 38.8 38.6-39.1 Vegetable garden no 910 38.7 38.4-39.0 0.06 yes 461 39.1 38.8-39.5
Haematocrit levels and anaemia in Australian children aged 1-4 years 334
The single Torres Strait Islander child in the survey did not have a usable sample. There was no difference in haematocrit levels between Aboriginal and non-Indigenous children. Earlier studies reporting high levels of anaemia in Aboriginal children have been done in rural and remote areas.5-7 However most Aboriginal people live in urban areas and this was reflected in the NSLIC (J Donovan, personal communication). Although the sample size is small, our findings results suggest that results from remote areas should not be extrapolated to the urban Aboriginal population. We found that haematocrit levels were the same regardless of maternal birthplace although the numbers were too small to allow examination in great detail. Unlike some other surveys2,8 we did not find that red meat intake was related to anaemia. This may be due to the crudeness of the questions that were asked. However, trials of dietary intervention have had inconsistent results16,17 and so this finding may also indicate that a single food item is not pre-eminent for preventing anaemia. It was not possible to examine a number of associations described previously. The National Health Survey and the National Nutrition Survey were also in the field in 1995 and ascertained duration of breastfeeding, age of introduction of cow milk, years since immigration, and collected a 24-hour dietary recall from those aged 2 years and older.18 It is unfortunate that, although all three surveys collected data in the same year, there was no relationship between them that would have permitted better use of the biochemical data. One interesting relationship would have been a comparison of iron status and dietary intakes. Population intakes should be com-pared to the estimated average requirement (EAR), not the recommended dietary intake.19,20 As the Australian dietary reference data do not describe the EAR, we used the US value of 3.4mg/day for this age group.19 After correction for within-person variation in dietary intake, the 10th centile of iron intake in Australian children aged
2-3 years was 5.3 mg/day,21 indicating that less than 10% of this age group have inadequate iron intakes. This is consistent with the low prevalence of anaemia in the NSLIC. In conclusion, the prevalence of anaemia is very low, 2-3.3%, in children aged 1-4 years in Australia. Although mean haematocrit levels increased with age and supple-ment use and varied among the jurisdictions, these factors explain little of the variation in haematocrit within the population. To examine risks associated with dietary intake and other factors, future national surveys would need to examine iron status, rather than anaemia, to exclude haemoglobinopathies and inflammatory diseases such as respiratory infections22 and to over-sample subgroups where the prevalence is thought to be higher. Our results justify the extra expense involved in a more detailed assessment of iron status rather than collecting only haemoglobin or haematocrit. In addition, to deter-mine the true prevalence, agreement is needed on the definition of iron deficiency and depletion and, ideally, a sub-sample should be re-tested to reduce the effects of intra-individual variability on the prevalence results. References 1. United Nations Children’s Fund, United Nations
University, World Health Organisation. Iron deficiency anaemia. Assessment, prevention and control. A guide for programme managers. World Health Organisation, 2001.
2. Karr M, Alperstein G, Causer J, Mira M, Lammi A, Fett MJ. Iron status and anaemia in preschool children in Sydney. Aust NZ J Public Health 1996;20:618-22.
3. Karr MA, Mira M, Alperstein G, Labib S, Webster BH, Lammi AT, Beal P. Iron deficiency in Australian-born children of Arabic background in central Sydney. Med J Aust 2001; 174:165-68
4. Oti-Boateng P, Seshadri R, Petrick S, Gibson RA, Simmer K. Iron status and dietary iron intake of 6-24-month-old children in Adelaide. J Paediar Child Health 1998; 34: 250-3.
Table 4. Final model for predictors of haematocrit levels, weighted to the age- and jurisdiction population distribution in the 1996 Census and corrected for clustering within households ________________________________________________________________________________________________
Variable Group Haematocrit 95% CI P (%) Constant - 38.2 37.7; 38.8 <0.0001 Age# 2 years old 0.3 -0.3; 0.8 0.02 3 years old 0.1 -0.4; 0.7 4 years old 0.8 0.2; 1.5 Jurisdiction# Victoria 0.6 -0.04; 1.3 <0.0001 Queensland -0.5 -1.3; 0.3 South Australia -0.4 -1.1; 0.4 Western Australia 1.0 -0.1; 2.1 Tasmania 1.2 -0.5; 2.9 Northern Territory -0.5 -2.0; 1.0 Australian Capital Territory 2.4 1.2; 3.6 Vitamin & mineral supplements Takes v does not take 0.6 0.07; 1.1 0.03 _________________________________________________________________________________________________________________# referent group for age is 1 year old, for jurisdiction is New South Wales; overall r2 for model: 0.04
335 DEM Mackerras, SI Hutton and PR Anderson
5. Harris MF, Cameron B, Florin S. Iron deficiency in Bourke children. Aust Paediatr 1988;24: 45-7.
6. Hopkins RM, Gracey MS, Hobbs RP, Spargo RM, Yates M, Thompson RCA. The prevalence of hookworm infection, iron deficiency and anaemia in an Aboriginal community in north-west Australia. Med J Aust 1997; 66: 41-44
7. Fraser J. Evaluation of a child health program to prevent and treat anaemia in Arnhemland. Aust J Rural Health 1996;4:11-7.
8. Mira M, Aperstein G, Karr M, Ranmuthugala G, Causer J, Niec A, Lilburne A. Haem iron intake in 12-36 month old children depleted in iron: case-control study. Brit Med J 1996; 312: 881-3.
9. Donovan J. Lead in Australian children: report on the National Survey of Lead in Children. Australian Institute of Health and Welfare, Canberra 1996.
10. Centers for Disease Control and Prevention. Recommendations to prevent and control iron deficiency in the United States. MMWR 1998; 47 (No RR-3):[inclusive page numbers].
11. ABS. Population by age and sex. Australian states and territories. June 1992 to June 1997. ABS Catalogue No 3201.0, Canberra, 1997
12. ABS. Experimental projections of the Aboriginal and Torres Strait Islander population. 30 June 1996 to 30 June 2006. ABS Catalogue No 3231.0. Canberra, 1998
13. Graciter PL, Goldsby JB, Nichaman MZ. Hemoglobins and hematocrits: are they equally sensitive in detecting anemias? Am J Clin Nutr 1981;34: 61-64.
14. Newell D, Simpson J. Regression to the mean. Med J Aust 1990; 153: 166-8.
15. Looker AC, Sempos CT, Liu K, Johnson CL, Gunter EW. Within-person variance in biochemical indicators of iron status: effects on prevalence estimates. Am J Clin Nutr 1990; 52: 541-7.
16. Engelmann MD, Sandstrom B, Michaelsen KF. Meat intake and iron status in later infancy: an intervention study. J Pediatr Gastroenterol Nutr 1998; 26: 26-33.
17. Makrides M, Leeson R, Gibson RA, Simmer K. A randomized controlled clinical trial of increased dietary iron in breast-fed intakes. J Pediatr 1998; 133: 559-62.
18. ABS and HEALTH. National Nutrition Survey. Nutrient intakes and physical measurements Australia 1995. ABS Cat No 4805.0, Canberra 1998.
19. Food and Nutrition Board, Institute of Medicine. Dietary Reference Intakes for Vitamin A, Vitamin K, Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manganese, Molybdenum, Nickel, Silicon, Vanadium, and Zinc. National Academy Press, Washington, 2002. http:// books. nap.edu/books/0309072794/html/index.html
20. Beaton G H. Recommended dietary intakes: individuals and populations. From Shils M, Olson JA, Shike M, Ross AC (eds). Modern nutrition in health and disease. 9th Edn. Williams and Wilkins, Baltimore, 1999.
21. Australian Bureau Statistics. National Nutrition Survey Nutrient Intakes and Physical Measurements Australia 1995. Cat No 4805.0. ABS, Canberra 1998.
22. Yip R, Dallman PR. The roles of inflammation and iron deficiency as caused of anemia. Am J Clin Nutr 1998; 48: 1295-1300.
Asia Pac J Clin Nutr 2004;13 (4): 336-340 336
Original Article Carotenoid status among preschool children with vitamin A deficiency in the Republic of the Marshall Islands Mary V Gamble PhD1, Neal A Palafox MD2, Barbara Dancheck BS3, Michelle O Ricks MS3, Kennar Briand MD4 and Richard D Semba MD3 1Department of Environmental Health Sciences, Mailman School of Public Health, Columbia University, New York, NY 2Department of Family Practice and Community Health, John A Burns School of Medicine, University of Hawaii, Honolulu, HI 3Department of Ophthalmology, Johns Hopkins University School of Medicine, Baltimore, MD and 4Ministry of Health and Environment, Republic of the Marshall Islands
Although carotenoids are known to be important dietary sources of vitamin A, there have been few epidemi-ological studies that have characterized the serum concentrations of major dietary carotenoids among preschool children with vitamin A deficiency. We conducted a population-based, cross-sectional study of serum pro-vitamin A carotenoids (α-carotene, β-carotene, β-cryptoxanthin), non-provitamin A carotenoids (lutein/ zeaxanthin, and lycopene), and retinol among 278 children, aged 1-5 y, in the Republic of the Marshall Islands. Vitamin A deficiency was defined as serum retinol <0.70 μmol/L. Geometric mean serum concentrations of carotenoids among children with and without vitamin A deficiency were 0.003 vs 0.006 μmol/L for α-carotene (P = 0.0017), 0.011 vs 0.023 μmol/L for β-carotene (P <0.0001), 0.023 vs 0.034 μmol/L for β-cryptoxanthin (P = 0.0075), 0.007 vs 0.012 μmol/L for lycopene (P = 0.037), 0.044 vs 0.052 μmol/L for lutein/zeaxanthin (P = 0.2), and 0.045 vs 0.074 μmol/L for total provitamin A carotenoids (P <0.0001) respectively. In a multivariate analysis adjusting for sex, age (Odds Ratio [O.R.] 1.44, 95% confidence interval [C.I.] 1.16-1.78), and serum provitamin A carotenoids (O.R. 0.49, 95% C.I. 0.34-0.71) were associated with vitamin A deficiency, but serum non-provitamin A carotenoids were not associated with vitamin A deficiency (O.R. 0.93, 95% C.I. 0.67-1.28). Preschool children with vitamin A deficiency in the Republic of the Marshall Islands have extremely low serum concentrations of provitamin A carotenoids and interventions are needed to improve the dietary intake of provitamin A carotenoids among Marshallese children.
Key words: α-carotene, β-carotene, β-cryptoxanthin, retinol, children, vitamin A deficiency, Republic of Marshall Islands Introduction Vitamin A deficiency is a leading cause of growth failure, morbidity, mortality, and blindness among preschool chil-dren in developing countries.1 Epidemiological surveys have shown that some islands in the South and Western Pacific regions have the highest prevalence rates of clinical vitamin A deficiency that have been described in the last twelve years.2 The rates of clinical vitamin A deficiency, i.e. nightblindness and Bitot spots, exceeded 15% in some islands.2 Factors that have been implicated in the recent emergence of vitamin A deficiency in the Pacific region include rapid demographic change, poverty, lack of home-stead food production, and replacement of traditional foods such as breadfruit, banana, taro, yam, sweet potato, pan-danus, coconut, and fish with rice and sweet refined foods of low nutritional quality.2,3 Some traditional cultivars used in the region such as yellow banana, giant swamp taro, and pandanus are carotenoid-rich.4 The Republic of the Mar-shall Islands is among the nations in the Pacific region
that have an extremely high rate of vitamin A deficiency.5
Although foods rich in carotenoids are the dominant source of vitamin A for children in developing countries, there are little epidemiological data on the relationship between serum retinol concentrations and the major serum carotenoids, α-carotene, β-carotene, β-cryptoxanthin, lu-tein, zeaxanthin, and lycopene among children with vita-min A deficiency. In developing countries, children derive an estimated 70-90% of their total vitamin A intake from provitamin A carotenoids (α-carotene, β-carotene, and β-cryptoxanthin) from fruits and vegetables.6 Serum caro-tenoid concentrations reflect the consumption of fruits and vegetables and are widely considered to be the best bio-logical markers for dietary carotenoid intake.7 In this
Correspondence address: Dr. Richard D. Semba, 550 North Broadway, Suite 700, Baltimore, MD 21205, U.S.A. Tel: (410) 955-3572; Fax (410) 955-0629. Email: [email protected]
Accepted 25 June 2004
337 MV Gamble, NA Palafox, B Dancheck, MO Ricks, K Briand and RD Semba
Table 1. Serum carotenoids and retinol among children, by age
study, we describe the distribution of serum carotenoid concentrations and their relationships to vitamin A defi-ciency among children in the Republic of the Marshall Islands. Methods A community-based survey, the Republic of the Marshall Islands Vitamin A Deficiency Study, was conducted be-tween November 1994 and March 1995. The total survey included 919 Marshallese children, ages 1-5, from ten atolls, who represented approximately 20% of the entire population of 1-5 year old children living the Republic of the Marshall Islands.5 The sampling strategy for the study was based on the 1988 census of the Republic of the Marshall Islands. This census provided data on the average number of children of the target age group within each household, determined by dividing the number of children in a locality by the number of households in the same location. This number was then divided into the number of children to be sampled to obtain the number of households to be visited. Households to be visited were chosen at random. When available, the date of birth of children was ascertained from the childrens' health cards. The survey team consisted of at least one Marshallese-speaking health care worker, a phlebotomist, and a medi-cal doctor. Oral consent was obtained from a parent or guardian prior to participation in the survey, and subject anonimity was preserved by removing personal identifiers from the data set. The Ministry of Health and Envion-ment of the Republic of the Marshall Islands supported the project and assisted with the planning and deve-lopment of this evaluation. Blood samples were obtained by venipuncture. Ve-nous blood samples were immediately wrapped in alu-minum foil and stored at 4°C until centrifugation (200 x g, 10 minutes, room temperature) in a local laboratory. Aliquots of serum were made in cryovials, and samples were placed immediately in liquid nitrogen. Serum samples were kept in liquid nitrogen or at -70°C until the time of laboratory analysis, which occurred within two years of the survey. Serum retinol and carotenoids are stable in samples that have been stored at -70°C for more than fifteen years.8 Serum retinol and carotenoids were
measured using reversed-phase high performance liquid chromatography.5,9 Quality control reference standards (Standard Reference Material 968C, National Institute of Standards and Technology, Gaithersburg, MD) were eva-luated to ensure the quality of our measurements. Pooled human standards were used to measure intra- and inter-assay coefficients of variation in laboratory ana-lyses. For serum retinol, the within-assay and between-assay coefficients of variation were 3% and 8%, respectively. For α-carotene, β-carotene, β-cryptoxanthin, lutein/zea-xanthin, and lycopene, the within-assay coefficients of variation were 7.9% and 6.3%, and 5.9%, 8.8% and 5.9%, respectively. The between assay coefficients of variation were 11.3% for β-carotene and 6.2% for total carotenoids. The study protocol was approved by the institutional review board of the Pacific Health Research Institute of Hawaii and the Ministry of Health and Environment of the Republic of the Marshall Islands. The study protocol conforms to the provisions of the Declaration of Helsinki in 1995 (as revised in Edinburgh 2000). Groups were compared using Student's t-test for continuous variables where appropriate, and categorical variables were compared using chi-square or exact tests. Vitamin A deficiency was considered consistent with serum retinol <0.70 μmol/L.10 Non-parametric tests were used to examine trend across age categories. Provitamin A carotenoids were defined as the sum of α-carotene, β-carotene, and β-cryptoxanthin in μmol/L, and non-provitamin A carotenoids were defined as the sum of lutein/zeaxanthin and lycopene in μmol/L. Univariate and multivariate logistic regression analyses were used to estimate the relative risks of factors associated with vita-min A deficiency. Regression coefficients were converted to odds ratios (OR), and the confidence intervals (CI) for the odds ratios were derived from the standard error estimates of the regression coefficients. Results Out of the 919 children in the survey, 278 children had serum carotenoid concentrations measured. The mean age (± SD) of the 278 children was 3.1 ± 1.3 years, and there were 137 boys and 141 girls. The subsample of children involved in the present study was compared by age and
Analyte1 1 to < 2 years (N = 43)
2 to < 3 years (N = 50)
3 to <4 years (N = 73)
4 to < 5 years (N = 54)
5 to < 6 years (N = 57)
P
α-carotene 0.004 (0.002, 0.007) 0.004 (0.002, 0.007) 0.004 (0.002, 0.006) 0.004 (0.002, 0.005) 0.004 (0.003, 0.006 0.94
β-carotene 0.015 (0.010, 0.022) 0.015 (0.011, 0.019) 0.015 (0.011, 0.019) 0.013 (0.008, 0.018) 0.015 (0.011, 0.019) 0.92
β-cryptoxanthin 0.014 (0.010, 0.026) 0.027 (0.019, 0.036) 0.032 (0.024, 0.041) 0.034 (0.027, 0.043) 0.027 (0.022, 0.033) 0.0002
lycopene 0.004 (0.002, 0.007) 0.007 (0.005, 0.012) 0.010 (0.008, 0.013) 0.009 (0.005, 0.014) 0.017 (0.012, 0.024) 0.0003
lutein/zeaxanthin 0.035 (0.024, 0.052) 0.037 (0.029, 0.048) 0.052 (0.041, 0.066) 0.053 (0.038, 0.071) 0.052 (0.042, 0.065) 0.11 provitamin A carotenoids2
0.041 (0.030, 0.055) 0.052 (0.040, 0.069) 0.059 (0.047, 0.073) 0.058 (0.046, 0.073) 0.052 (0.043, 0.061) 0.26
non-provitamin A carotenoids3
0.045 (0.031, 0.066) 0.053 (0.042, 0.066) 0.069 (0.056, 0.086) 0.070 (0.052, 0.094) 0.079 (0.066, 0.097) 0.02
retinol 0.683 (0.604, 0.777) 0.590 (0.516, 0.675) 0.560 (0.511, 0.615) 0.547 (0.498, 0.601) 0.504 (0.464, 0.549) 0.0036 1 Geometric mean (lower and upper 95% confidence limits; 2 Provitamin A carotenoids = α-carotene + β-carotene + β-cryptoxanthin; 3 Non provitamin A carotenoids = lycopene + lutein/zeaxanthin
Carotenoid status among preschool children with vitamin A deficiency in the Republic of the Marshall Islands 338
sex with the children who were not selected for the study. There were no significant differences by age, sex, or retinol concentrations (data not shown). The serum con-centrations of α-carotene, β-carotene, β-cryptoxanthin, lutein/zeaxanthin, and lycopene among children by age in years is shown in Table 1. Serum β-cryptoxanthin and lycopene increased significantly with increasing age, whereas serum retinol concentrations decreased with increasing age. There were no significant differences in α-carotene, β-carotene, lutein/zeaxanthin, or provitamin A carotenoid concentrations with age. Serum carotenoid concentrations were compared between children with and without vitamin A deficiency (Table 2). Children with vitamin A deficiency had significantly lower geometric mean serum concentrations of α-carotene, β-carotene, β-cryptoxanthin, lycopene, and provitamin A carotenoids compared to children without vitamin A deficiency. There were no significant differences in geometric mean serum concentrations of lutein/zeaxanthin or non-provitamin A carotenoids between children with and without vitamin A deficiency. Serum carotenoid concen-trations were compared between boys and girls, and there were no significant differences in geometric mean serum concentrations of any of the carotenoids by sex (data not shown). Geometric mean serum retinol concentrations (SD) were 0.541 (0.507, 0.578) and 0.596 (0.557, 0.636) μmol/L among boys and girls, respectively (P = 0.04). Univariate and multivariate analyses were conducted in order to examine risk factors associated with vitamin A deficiency (Table 3). In univariate analyses, lower con-centrations of serum pro-vitamin A carotenoids and in-creasing age were significantly associated with an in-creased risk of vitamin A deficiency. In multivariate ana-lyses adjusting for sex and age, serum provitamin A
carotenoids were associated with vitamin A deficiency (P <0.0001) whereas non-provitamin A carotenoids were not associated with vitamin A deficiency (P = 0.65).
Discussion To our knowledge, this is the first study to describe the major serum carotenoids among preschool children with and without vitamin A deficiency in a developing coun-try. These data suggest that serum carotenoid concen-trations are profoundly low among preschool children in the Republic of the Marshall Islands. There has been a paucity of studies of serum carotenoids among children in the scientific literature.11 It is notable that the geometric mean carotenoid concentrations for α-carotene, β-carotene, β-cryptoxanthin, lutein/zeaxanthin, and lyco-pene among Marshallese children was below the 5th per-centile for all of these major blood carotenoids in a slightly older group of children in the US. Among 839 children aged 6-7 years who participated in the third National Health and Nutrition Examination Survey in the US, 1988-1994, the 5th percentiles in μmol/L were 0.012 for α-carotene, 0.124 for β-carotene, 0.083 for β-cryptoxanthin, 0.177 for lutein/zeaxanthin, and 0.167 for lycopene.12 The geometric mean concentrations of serum carotenoids of Marshallese children were also below the 25th percentile for the respective serum carotenoids among children in Belize. Among 493 children in Belize, aged 3-9 years, the 25th percentiles in μmol/L were 0.056 for α-carotene, 0.130 for β-carotene, 0.090 for β-cryptoxanthin, 0.176 for lutein/zeaxanthin, and 0.056 for lycopene.11 The geometric mean concentrations of serum carotenoids of Marshallese children were also below that described among children with acute, uncomplicated malaria in India13 and Uganda.14 The serum carotenoid
Table 2. Serum carotenoid concentrations among preschool children with and without vitamin A deficiencyCarotenoid1 (μmol/L) Serum retinol (μmol/L) P < 0.70 (N = 189) >0.70 (N = 89)
α-carotene 0.003 (0.002, 0.004) 0.006 (0.005, 0.008) 0.0017
β-carotene 0.011 (0.009, 0.013) 0.023 (0.019, 0.029) 0.0001
β-cryptoxanthin 0.023 (0.021, 0.027) 0.034 (0.027, 0.043) 0.0075 lycopene 0.007 (0.006, 0.010) 0.012 (0.009, 0.016) 0.037 lutein/zeaxanthin 0.044 (0.038, 0.052) 0.052 (0.043, 0.088) 0.2 provitamin A carotenoids2 0.045 (0.040, 0.052) 0.074 (0.062, 0.089) 0.0001
non provitamin A carotenoids3 0.059 (0.052, 0.069) 0.074 (0.062, 0.089) 0.082
1 Geometric mean (lower and upper 95% confidence limits); 2 Provitamin A carotenoids = α-carotene + β-carotene + β-cryptoxanthin; 3 Non provitamin A carotenoids = lycopene + lutein/zeaxanthin
Table 3. Univariate and multivariate models of factors associated with vitamin A deficiency (retinol <0.70 μmol/L) __________________________________________________________________________________________________Characteristic Univariate Multivariate OR (95% CI) P OR (95% CI) P __________________________________________________________________________________________________
Age (per year) 1.37 (1.13-1.66) 0.0014 1.44 (1.16-1.78) 0.0007 Male gender 1.47 (0.88-2.45) 0.13 1.28 (0.74-2.21) 0.37 Provitamin A 0.50 (0.36-0.70) 0.0001 0.49 (0.34-0.71) 0.0001 carotenoids1 Non provitamin A 0.78 (0.59-1.03) 0.08 0.93 (0.67-1.28) 0.65 carotenoids2 1 Provitamin A carotenoids = α-carotene + β-carotene + β-cryptoxanthin; per loge unit increase; 2 Non provitamin A carotenoids = lycopene + lutein/zeaxanthin; per loge unit increase.
339 MV Gamble, NA Palafox, B Dancheck, MO Ricks, K Briand and RD Semba
and retinol concentrations of Marshallese children were also relatively lower than those of a similar age group of children from urban slums of northeastern Brazil (M. Gamble, unpublished data). In infants from semiurban villages in Pakistan, β-carotene concentrations were undetectable, suggesting that extremely low consumption of carotenoids occurs in other high risk populations.15 The low serum concentrations of carotenoids in Marshallese preschool children appear to reflect the dietary and demographic changes that have occurred over the last few decades in the Republic of the Marshall Islands. Traditional foods such as breadfruit, banana, taro, yam, sweet potato, pandanus, coconut, and fish have been replaced with rice, fatty foods, and refined foods of low nutritional quality, such as canned processed meat, snack foods and chips, and ramen noodles. Similar dietary changes have taken place in the region, such as in the Federated States of Micronesia.4 A limitation of this study is that the survey and the analyses were conducted in 1994-1995, and the vitamin A and carotenoid status may have changed since the survey. Following the initial reports from this survey, a vitamin A capsule distribution program was implemented in the Republic of the Marshall Islands, but little has been done to address the intake of traditional foods among Marshallese children. Serum retinol concentrations appeared to decrease with age whereas serum provitamin A carotenoids were similar across age categories. The decline of serum retinol with increasing age is consistent with the general observation that the risk of xerophthalmia increases with increasing age among preschool children.16-18 The intake of pro-vitamin A carotenoids may not be sufficient to meet the needs of growing preschool children in this population with an extremely high prevalence of vitamin A defi-ciency. Nevertheless, serum concentrations of provitamin A carotenoids were significantly lower among children with vitamin A deficiency compared to children without vitamin A deficiency, consistent with the notion that bioavailable carotenoids reduce the risk of vitamin A deficiency. One factor that may affect the absorption of carotenoids is the amount of dietary fat that is consumed, and it is possible that a relative lack of fat may also be affecting the absorption of carotenoids among preschool children. Serum β-cryptoxanthin and lycopene con-centrations increased with increasing age, and this in-crease might reflect an increased consumption of citrus fruit containing β-cryptoxanthin and foods containing to-mato sauce or tomatoes with increasing age. A limitation of this study is that data on infectious disease morbidity was not collected for all children in the survey. The acute phase response, which can occur during infections, may depress serum retinol concentrations.19 Infections such as malaria are not endemic in the Republic of the Marshall Islands, but diarrheal disease is a common problem among preschool children. Although the measurement of acute phase proteins has been advocated to improve the interpretation of serum retinol concentrations, elevation of acute phase proteins is common among preschool chil-dren who have clinical vitamin A deficiency, i.e night-blindness and Bitot's spots, and elevated acute phase proteins do not provide any criteria to distinguish children with clinical vitamin A deficiency from subclinical
vitamin A deficiency.20 The dietary consumption of dark green leafy vegetables and orange fruits is often advo-cated to combat vitamin A deficiency in developing countries,21,22 and many studies have focused on the dietary consumption of foods rich in β-carotene.23,24 The use of palm fruit and its products may be another effective approach in the alleviation of vitamin A defi-ciency.25 Home gardening projects in the South Pacific appear to be protective against vitamin A deficiency.26 Local measures, such as increasing homestead food production and dietary education are needed to increase the consumption of carotenoid-rich fruits and vegetables among children in the Republic of the Marshall Islands, as the present study shows low concen-trations of dietary carotenoids among these children. Acknowledgements This study was support in part by the Pacific Health Research Institute, UNICEF, the Fergussen Foundation Hawaii, the Hawaii Community Foundation, the National Institute of Child Health and Human Development (HD30042), the National Institutes of Health, and the United States Agency for International Development (Cooperative Agreement HRN A-0097-00015-00). References 1. McLaren DS, Frigg M. Sight and Life manual on vitamin
A deficiency disorders (VADD). Second edition. Basel: Task Force Sight and Life, 2001.
2. Semba RD, Palafox NA. Prevention of nutritional blindness in the South Pacific. Asia-Pacific J Ophthalmol 2002; 14: 6-12.
3. Schaumberg DA, Linehan M, Hawley G, O=Connor J, Dreyfuss M, Semba RD. Vitamin A deficiency in the South Pacific. Public Health 1995; 109: 311-317.
4. Englberger L, Mark GC, Fitzgerald MH. Insights on food and nutrition in the Federated States of Micronesia: a review of the literature. Public Health Nutr 2002; 6: 5-17.
5. Gamble MV, Ramakrishnan R, Palafox NA, Briand K, Berglund L, Blaner WS. Retinol binding protein as a surrogate measure for serum retinol: studies in vitamin A-deficient children from the Republic of the Marshall Islands. Am J Clin Nutr 2001; 73: 594-601.
6. Food & Agricultural Organization/World Health Organi-zation. Requirements of vitamin A, iron, folate and vitamin B12. Report of a joint FAO/WHO expert consultation. FAO Food and Nutrition Series No. 23. Rome: Food and Agricultural Organization of the United Nations, 1988.
7. Food and Nutrition Board, Institute of Medicine. Dietary references intakes for vitamin C, vitamin E, selenium, and carotenoids. Washington DC: National Academy Press, 2000.
8. Comstock GW Alberg AJ, Helzlsouer KJ. Reported effects of long-term freezer storage on concentrations of retinol, beta-carotene, and alpha-tocopherol in serum or plasma summarized. Clin Chem 1993; 39: 1075-1078.
9. Redlich CA, Grauer JN, van Bennekum AM, Clever SL, Ponn RB, Blaner WS. Characterization of carotenoid, vitamin A, and alpha tocopherol levels in human lung tissue and pulmonary macrophages. Am J Respir Crit Care Med 1996; 154: 1436-1443.
10. West KP Jr, Darnton-Hill I. Vitamin A deficiency. In: Semba RD, Bloem MW, eds. Nutrition and health in developing countries. Totowa, NJ: Humana Press, 2001; 267-306.
Carotenoid status among preschool children with vitamin A deficiency in the Republic of the Marshall Islands 340
11. Apgar J, Makdani D, Sowell AL, Gunter EW, Hegar A, Potts W, Rao D, Wilcox A, Smith JC. Serum carotenoid concentrations and their reproducibility in children in Belize. Am J Clin Nutr 1996; 64: 726-730.
12. Ford ES, Gillespie C, Ballew C, Sowell A, Mannino DM. Serum carotenoid concentrations in US children and adolescents. Am J Clin Nutr 2002; 76: 818-827.
13. Das BS, Thurnham DI, Das DB. Plasma alpha-tocopherol, retinol, and carotenoids in children with falciparum malaria. Am J Clin Nutr 1996; 64: 94-100.
14. Nussenblatt N, Mukasa G, Metzger A, Ndeezi G, Eisinger W, Semba RD. Relationship between carotenoids and anaemia during acute uncomplicated Plasmodium falciparum malaria in children. J Health Pop Nutr 2002; 20: 205-214.
15. Northrop-Clewes CA, Paracha PI, McLoone UJ, Thurnham DI. Effect of improved vitamin A status on response to iron supplementation in Pakistani infants. Am J Clin Nutr 1996; 64: 694-699.
16. Khatry SK, West KP Jr, Katz J, LeClerq SC, Pradhan EK, Wu LS, Thapa MD, Pokhrel RP. Epidemiology of xerophthalmia in Nepal: a pattern of household poverty, childhood illness, and mortality. Arch Ophthalmol 1995; 113: 425-429.
17. Sukwa T, Mwandu D, Kapui A, Siziya S, Vamoer A, Mukunyandela M, Chelemu V. The prevalence and distribution of xerophthalmia in pre-school age children of the Luapula Valley, Zambia. J Trop Pediatr 1988; 34: 12-15.
18. Sommer A. Nutritional Blindness: Xerophthalmia and Keratomalacia. New York, Oxford University Press, 1982.
19. Thurnham DI, McCabe GP, Northrop-Clewes CA, Nestel P. Effects of subclinical infection on plasma retinol concen-trations and assessment of prevalence of vitamin A deficiency: meta-analysis. Lancet 2003; 362: 2052-2058.
20. Semba RD, Muhilal, West KP Jr, Natadisastra G, Eisinger W, Lan Y, Sommer A. Hyporetinolemia and acute phase proteins in children with and without xerophthalmia. Am J Clin Nutr 2000; 72:146-153.
21. Bloem MW, de Pee S, Darnton-Hill I. New issues in developing effective approaches for the prevention and control of vitamin A deficiency. Food Nutr Bull 1998; 19: 137-148.
22. De Pee S, Bloem MW, Kiess L. Evaluating food-based programmes for their reduction of vitamin A deficiency and its consequences. Food Nutr Bull 2000; 21: 232-238.
23. De Pee S, Bloem MW, Gorstein J, Sari M, Satoto, Yip R, Shrimpton R, Muhilal. Reappraisal of the role of vegetables in the vitamin A status of mothers in Central Java, Indonesia. Am J Clin Nutr 1998; 68: 1068-1074.
24. Faber M, Phungula MAS, Venter SL, Dhansay MA, Benadé AJS. Home gardens focusing on the production of yellow and dark-green leafy vegetables increase the serum retinol concentrations of 2-5-y-old children in South Africa. Am J Clin Nutr 2002; 76:1048-1054.
25. Solomons NW, Orozco M. Alleviation of vitamin A deficiency with palm fruit and its products. Asia Pac J Clin Nutr 2003; 12: 373-384.
26. Schaumberg DA, O'Connor J, Semba RD. Risk factors for xerophthalmia in the Republic of Kiribati. Eur J Clin Nutr 1996; 50: 761-764.
341 Asia Pac J Clin Nutr 2004;13 (4): 341-347
Original Article The effects of a high calcium dairy food on bone health in pre-pubertal children in New Zealand Megan J Gibbons MSc, RD 1,3, Nigel L Gilchrist MBCh1, Christopher Frampton PhD2, Patricia Maguire RN 1, Penelope H Reilly RN 1, Rachel L March RN 1 and Clare R Wall PhD3 1CGM Research Trust, PO Box 731, Christchurch, New Zealand 2Department of Medicine, Christchurch Hospital. PO Box 4710, Christchurch, New Zealand 3Institute of Food Nutrition and Human Health, Massey University, Auckland, New Zealand
Childhood and adolescence is the period of most rapid skeletal growth in an individual’s lifetime. A greater peak bone mass achieved in the first 2-3 decades of life, may protect against the risk of osteoporotic fracture in later life. The aim of this randomized, controlled study was to assess in pre-pubertal boys and girls (aged 8-10 years) the effect of 18 months of a calcium enriched, cocoa flavoured product on bone density, bone growth and bone size in New Zealand children. One hundred and fifty four pre-pubertal boys and girls (aged 8-10 years) were randomized to receive a high calcium dairy drink or a control drink reconstituted with water for 18 months. They were assessed at baseline and then every 6 months for the first 18 months, while they were having the supplement; they were then followed up 12 months after supplementation had finished. Bone mineral density and bone mineral content were assessed at the total body, hip and spine. Indicators of bone size (vertebral width and height) were also measured at the spine. Anthropometric data was collected, medical history questionnaires were administered (including the Tanner or pubertal stage questionaire), dietary calcium intake was assessed with a calcium food frequency questionnaire and calcium supplement compliance was determined. There was no significant difference between the 2 groups for bone mineral density or bone mineral content at any time point. There was no difference in vertebral height or width at any stage of the study, indicating no additional influence on bone size at the lumbar vertebrae. There were no significant differences between height, weight, lean mass or fat mass at any time point. Both groups had higher habitual calcium intakes than recommended for this age group going into the study and throughout the study. In this 2½ year study (18 months supplementation, 1 year follow-up) we did not observe a difference in bone mineral density in pre-pubertal children. This was probably due to their high habitual dietary calcium intake whereby minimal addition of calcium to the diet reached the threshold level where no further benefit was seen. There were no significant differences between the two groups in body composition. Growth and the mean height and weight remained between the 50th and 75th percentile for their age. We have shown calcium supplementation in children with high habitual dietary calcium intake appears not to have additional effects on bone mass. Calcium supplementation needs to be targeted in those children with low habitual dietary calcium intake.
Key Words: BMD, BMC, pre-pubertal, calcium supplementation, bone health, New Zealand. Introduction The amount of calcium a child needs to support optimal growth and to maximize peak bone mass (PBM) has not been conclusively established. Childhood and adolescence is the period of most rapid skeletal growth. Increase in total body bone mass can be in the order of 7-8% per year1,2 with the highest bone mineral density gain in girls between the ages of 10 to 14.3 A greater PBM achieved in the first 2-3 decades of life, may protect against the risk of osteoporotic fracture in later life4. Accordingly, osteo-porosis is sometimes termed a “paediatric disorder”. Osteoporosis is a major New Zealand (NZ) public health issue affecting approximately one in three women and one in eight men of the elderly population and costs the health care system an estimated $200 million a year.5 It is therefore important that recommendations for calcium intake in children correspond with the intake required to
maximize their PBM, and therefore provide some pro-tection against osteoporosis in later life. There are limited data on the calcium intake of children in NZ, and there have been no published surveys of nutrient intakes in this population. A 1997 National Nutrition Survey found 15-18 year old young adults have a daily average calcium intake of 870mg.6 Other studies have suggested calcium intake decreases from childhood to adolescence,7-8 so it seems probable that the intake of NZ children will be higher than this. Correspondence address: Megan J Gibbons, Institute of Food Nutrition and Human Health, Massey University, Auckland , New Zealand . Tel: + 64 9 4140800; Fax: +64 9 4439640 Email: [email protected] Accepted 25 June 2004
MJ Gibbons, N L Gilchrist, C Frampton, P Maguire, PH Reilly, RL March and CR Wall 342
Eight intervention studies9-16 in children aged 7-14 years examining the effects of calcium supplementation on bone mineral density have been published. The studies, with the exception of the study by Cadogan et al., (1997)15, were all of a randomized double-blind placebo-controlled study design. This is the strongest study de-sign to prove causality, and controlling for potentially confounding factors that are known to influence bone mass. Additional factors associated with peak bone mass are genetics, hormones, physical activity, lifestyle factors such as smoking and nutrition.1 One of these studies has shown an increase in bone size11 corresponding with an increase in bone mineral density when additional calcium is given in the form of calcium fortified foods. In addition, the studies9-15 have suggested the appendicular skeleton, particularly the regions of compact bone, appear more sensitive than the axial skeleton to the effects of calcium supplementation. The main areas of increase in bone mineralisation have been the radial and femoral dia-physis,11 the midshaft radius,10, 12-14 total hip9, and leg and pelvis.15 However, some studies have also found signi-ficant differences in the spine10,13,16 and the total body.15-16 The aim of this randomized, controlled study is to assess in pre-pubertal boys and girls (aged 8-10 years) the effect of 18 months of a calcium enriched, cocoa fla-voured product (New Zealand Milk Limited, Wellington, New Zealand) on bone mineral density, bone growth and bone size at the vertebrae in NZ children. This age group was chosen because previous studies have shown this is a period of most rapid development.2 Supplementation with both milk calcium and calcium salts have been shown to increase bone mineral density in this age group.9-16 There are limited data on supplementation with calcium on bone development in male pre-pubertal children. With the increase in osteoporosis in both males and females, it is also important to consider the effect nutrition may play on their bone development. Subjects and Methods Three hundred and ninety children aged 8-10 years were approached at three local primary schools and asked to attend an information evening with their parents. Of the 211 who attended, 159 satisfied the inclusion criteria and agreed to participate. Exclusion criteria included allergy to dairy products, any major disease states including significant psychological problems. If the child was on any medication that influenced bone growth or meta-bolism [e.g steroids (inhaled or oral), anti-convulsants, thiazide diuretics or vitamin D] they were excluded. Of the 159 randomised children, 5 (2 controls and 3 treat-ment) did not complete the baseline questionnaires after randomisation and took no further part in the study. The children at each school were randomly allocated to either the treatment group or the control group using heel ultrasound values at baseline for stratification. The study was approved by the Southern Regional Health Authority Ethics Committee, (Canterbury) New Zealand in 1997. Informed consent was obtained from the children and their parents.
Protocol All the children visited the research centre and met with
the research nurses, at baseline and then every six months for the first 18 months, while they were having the supplement; they were then followed up 12 months after the supplementation finished. Bone mineral density, bone mineral content and bone size were measured at baseline, 6 months, 12 months, 18 months and 30 months; height and weight were also recorded at these times. In addition, the children completed a calcium food frequency que-stionnaire22 to determine dietary calcium intake at each visit. A medical questionnaire was completed at baseline and at 30 months to check medication use, medical history, previous fractures, family history and caffeine intake; for the females there were also questions asked about menarche history. Pubertal stage (breast and genital development) was assessed by a self-administered Tanner questionnaire17 at baseline, 18 and 30 months. Dietary supplement and compliance The children had sachets of the product (Table 1) deli-vered to them either fortnightly at school or monthly at home over the 18 months of supplementation. Each child was required to have one sachet of the product in the morning and afternoon mixed with hot or cold water as a drink. The children were blinded to the study product as both looked and tasted the same. The high calcium milk provided 600mg of calcium per 40g serve or 1200mg per day. The control milk provided 200mg of calcium per 40g serve or 400mg per day - this is equivalent to two glasses of milk. The children completed a tick sheet after they had consumed the drink, and returned it to the study coordinator each month as a measure of compliance. Compliance was assessed as the number of drinks actu-ally consumed expressed as a percentage of those that should have been taken. The study product was initially chocolate flavoured, however other flavours were intro-duced to maintain compliance.
ProcedurBone micompositptiometryWisconsiwere meacontent; measuredlysed usi
Energy
Protein CarbohyFat VitaminVitaminVitaminCalciumPhospho*= µg re
Treatment Placebo
1529 kJ 364 kcal
1650 kJ 393 kcal
10.2 g 10.6 g drate 34.7 g 34.7 g
21.4 g 24.5 g A 656 µg* 656 µg* D 336 IU 320 IU C 52 mg 58 mg 1200 mg 400 mg rus 776 mg 320 mg tinol equivalents
Table 1. Nutritional composition of 2 sachets (80g) of high calcium supplement compared to 2 sachets (80g) of the placebo supplement
es neral density, bone mineral content and body ion were determined by dual-energy x-ray absor- (DPX-IQ; Lunar Radiation Corp., Madison, n). The total body, lumbar spine and total hip sured for bone mineral density and bone mineral lean muscle mass and fat mass were also from the total body scan. The scans were ana-ng the Lunar paediatric database (version 4.79).
343 The effects of a high calcium dairy food on bone health in pre-pubertal children in New Zealand
Statistical analysis Preliminary power calculations indicate that 75 subjects in each group will provide sufficient power (80%) to detect a differential effect of approximately 4% (α = 0.05). Comparisons between the two groups at baseline were made using independent t-tests; ANOVA for re-peated measures was used to compare the changes over time between the two groups. When ANOVA indicated a significant interaction between time and treatment group, comparisons between groups at individual times were made using Fisher’s Least Significant Difference test. All analyses were performed on an intention to treat basis. The statistical package used was SPSS (Statistical Package for Social Sciences) 10.0.1 for Windows (1999) Results Of the children randomized and starting intervention (N=154) 51% were female and 49% were male. There was a mixture of ethnic identities in the study population, however the majority of participants identified themselves as NZ European or Pakeha (81%). Other nationalities were Australian, Dutch, English, NZ Maori, Scottish and Other. The groups were very well matched at baseline on all variables (Table 2). The withdrawal rate was high in this
Table 2. Baseline characteristics of both groups, mean values (+ SEM)
Treatment (N = 74)
Control (N = 80)
P value
Age (years) 9.4 (0.1) 9.4 (0.1) 0.952 Ratio males to females
36:38 39:41
Height (cm) 135.1 (0.8) 135.4 (0.8) 0.767 Weight (kg) 31.5 (0.7) 32.4 (0.9) 0.399 Dietary calcium intake (mg)
934 (44) 985 (53) 0.461
Total Body BMD (g/cm2)
0.881 (0.006) 0.881 (0.007) 0.951
Total Body BMC (g)
1152 (23)
1172 (29) 0.603
L1-L4 Spine BMD (g/cm2)
0.742 (0.009) 0.730 (0.009) 0.329
L1-L4 Spine BMC (g)
23 (1) 23 (1) 0.800
Total Hip BMD (g/cm2)
0.813 (0.014) 0.792 (0.015) 0.305
Total Hip BMC (g)
18 (0.4) 17 (0.5) 0.727
Total Fat (kg) 5.8 (0.4) 6.3 (0.5) 0.428 Total Lean (kg) 23.8 (0.3) 24.3 (0.4) 0.357 Tanner 1 (breast/genital)#
1 (1-2) 1 (1-2)
Tanner 2 1 (1-2) 1 (1-2)
The region of interest was adjusted to match previous scans and anatomical changes due to growth. Quality assurance was carried out three times per week during the study. The coefficient of variation for repeated scans of the lumbar spine was 1%, the femoral neck 2.5% and 0.5% for the total body. The coefficient of variation for repeated scans of lean muscle mass was 1.1% and 1.9% for total fat mass. The hip bone areas were not analysed because of growth, positioning and coefficient of variance factors. Height was measured using a wall mounted stadio-meter (Harpendin); children were measured three times to within 2mm and the average obtained. Weight was mea-sured on electronic calibrated scales (Salter). All mea-surements were undertaken by assessors blinded to the treatment received by each child. Dietary calcium intake was calculated from a calcium food frequency22 question-naire using data from the New Zealand food composition tables.study, 53.9% withdrew at some point in the 18 months from taking the supplement. Of these children 66.3% continued to be monitored at each data collection point. The main reason for the children withdrawing was a dis-like of the study products, both treatment and control, and an inability to have 2 cups a day. We have included the data points for the children who did not finish the sup-plement in the analysis. Both groups were evenly matched in terms of withdrawals from the study (Ncontrol = 39, Ntreatment = 44). Most withdrawals occurred in the first six months of starting the intervention. The mean compliance for those who finished treatment (N=71) was 80.6%. No significant differences were seen in the changes in the anthropometric values between the two groups and the minimum P value (0-30) from repeated measures of ANOVA = 0.531. Throughout the study the mean value for both the males and females in both groups remained between the 50th and 75th percentile for both height and weight [National Child Health Survey (NCHS) growth curves for children].
(pubic hair)#
Table 3. Height, Weight, Total Body Lean Mass and Fat Mass for the Treatment and Control group at baseline, 6, 12,18 and 30 months (±SEM)
Baseline (N txt=74, N con=80)
6 Months (N txt=45, N con=58)
12 Months (N txt=56, N con=60)
18 Months (N txt=60, N con=65)
30 Months (N txt=58, N con=65)
P value (0-30)
Height (cm) Treatment 135.1 (0.8) 138.4(1.2) 141.6 (1.0) 144.3 (1.0) 151.0 (1.0) 0.607 Control 135.4 (0.8) 138.9 (1.0) 142.8 (1.0) 145.3 (1.0) 151.8 (1.1)
Weight (kg) Treatment 31.5 (0.7) 33.5 (1.1) 36.3 (1.0) 38.4 (1.0) 42.9 (1.2) 0.548 Control 32.4 (0.9) 33.8 (1.0) 37.4 (1.2) 39.0 (1.2) 44.0 (1.4)
Lean Mass (g) Treatment 23806 (346) 24995 (548) 26458 (468) 27783 (488) 30978 (577) 0.823 Control 24318 (421) 25337 (501) 26874 (545) 28114 (580) 31178 (672)
Fat Mass (g) Treatment 5775 (411) 6794 (643) 7932 (658) 8391 (650) 9532 (740) 0.531 Control 6317 (534) 6751 (608) 8340 (825) 8628 (790) 10260 (874)
MJ Gibbons, N L Gilchrist, C Frampton, P Maguire, PH Reilly, RL March and CR Wall 344
Table 4. Percentage change from baseline for the Total Body, Lumbar Spine, Total Hip, Trochanter and Femoralneck bone mineral density for the Treatment and Control groups at, 6, 12, 18 and 30 months (±SEM)
There were no significant differences in the bone mineral density changes between the groups, although some trends were evident at the total hip and trochanter. The changes 0-18 months also showed no significant advan-tage to either treatment (minimum P value =0.567). There were no significant differences in the indicators of bone size between the groups. The changes 0-18 months
also showed no significant advantage to either treatment (minimum P value =0.675) Throughout the study habitual calcium intake did not change, indicating that the supplement did not influence food choice e.g substituting calcium rich foods for other foods (mean treatment 30 months = 876, mean control 30 months = 894). There were no significant differences in the percentage change in bone mineral content between the groups (minimum P value =0.268). There was no difference in the change in pubertal stage for the children between groups (Tanner stage 1 at baseline, Tanner stage 2 at 30 months). Fifteen girls started menstruating during the study duration, seven in the treatment group and eight in the control group.
Discussion Our study shows there is no beneficial effect of taking a high calcium dairy supplement when the habitual dietary calcium intake is already high. There was no difference between the groups, although trends in the total hip and trochanter were evident. There may be a limit at which calcium supplementation has a positive effect on bone mineralisation. In the control group the mean intake during supplementation was 1395mg per day, and the
6 Months
(N txt=45, N con=58) 12 Months (N txt=56, N con=60)
18 Months (N txt=60, N con=65)
30 Months (N txt=58, N con=65)
P value (0-30)
Total Body (%) Treatment 1.6(0.9) 4.4(0.9) 5.1(0.9) 9.4(1.0) 0.737 Control 1.6(0.9) 3.3(1.0) 4.3(1.0) 8.9(1.1)
L1-L4 Spine (%) Treatment 2.3(1.6) 5.9(1.5) 8.4(1.5) 16.3(1.9) 0.616 Control 4.1(1.5) 5.8(1.5) 8.6(1.5) 16.8(2.1)
Total Hip (%) Treatment 1.0(2.1) 4.8(1.7) 6.8(1.6) 14.0(1.9) 0.081 Control 2.4(2.0) 3.9(1.8) 5.4(1.6) 12.4(2.0)
Trochanter (%) Treatment 0.9(2.1) 6.2(1.9) 8.6(1.9) 15.8(2.2) 0.088 Control 2.8(2.2) 5.1(2.0) 7.5(2.0) 14.9(2.2)
Femoral Neck (%) Treatment 0.7(2.1) 6.7(1.7) 8.9(1.9) 15.4(1.9) 0.447 Control 3.4(1.9) 7.0(1.8) 8.2(1.7) 15.3(1.7)
Table 5. Percentage change from baseline in parameters of bone size; L1-L4 lumbar spine width, area, height and volumetric density for 6, 12, 18 and 30 months (±SEM)
6 Months (N txt=45, N con=58)
12 Months (N txt=56, N con=60)
18 Months (N txt=60, N con=65)
30 Months (N txt=58, N con=65)
P value (0-30)
L1-L4 Width (%) Treatment 2.3 (1.4) 4.9 (1.3) 7.2 (1.2) 12.1 (1.2) 0.676 Control 2.6 (1.1) 5.5 (1.2) 7.2 (1.1) 12.1 (1.3) L1-L4 Area (%) Treatment 2.9 (2.4) 8.1 (1.7) 13.7 (1.8) 25.0 (2.0) 0.511 Control 4.9 (2.0) 10.4 (1.9) 14.3 (2.0) 25.8 (2.4) L1-L4 Height (%) Treatment 1.9 (0.8) 3.8 (0.7) 6.2 (0.7) 10.1 (1.6) 0.293 Control 2.7 (0.7) 5.0 (0.8) 7.1 (0.8) 12.1 (1.1) L1-L4 Volumetric Density (%) Treatment 6.5 (5.6) 19.6 (4.9) 28.3 (5.0) 54.3 (6.5) 0.603 Control 14.0 (5.3) 20.9 (5.2) 32.6 (5.6) 60.5 (7.3)
800
850
900
950
1000
0 6 12 18 30
Time (months)
Cal
cium
inta
ke (m
g)
TreatmentControl
P =0.459
P=0.841
P =0.655
P=0.464
P =0.810
Figure 1. Comparison between the habitual calcium intake in the treatment and control groups throughout the study.
345 The effects of a high calcium dairy food on bone health in pre-pubertal children in New Zealand
Total Hip Bone Mineral Content
01020304050
6 12 18 30
Time (months)
Perc
ent (
%)
Total Body Bone Mineral Content
0
10
20
30
40
50
6 12 18 30
Time (months)
Perc
ent (
%)
Femoral Neck Bone Mineral Content
0
10
20
30
6 12 18 30
Time (months)
Perc
ent (
%)
TreatmentControl
L1-L4 Spine Bone Mineral Content
0102030405060
6 12 18 30
Time (months)
Perc
ent (
%)
Figure 2. Percentage change from baseline for total body, L1-L4 lumbar spine, total hip and femoral neck bone mineral content at 6, 12, 18 and 30 months.
treatment group was even higher at 2106 mg per day. Weaver et al.,18 and Wastney et al.,19
have shown through calcium balance studies that maximum retention of calcium occurred at a level of 1300mg per day, which has since become the recommended intake in the USA for this age group. This group has estimated that an increase in calcium intake from 918mg to 1300mg per day could increase skeletal mass by 4% per year in pre-pubertal children. This is consistent with the increase in bone mineral density we saw in both the treatment and control groups. The previous studies that have looked at the effect of calcium supplementation on bone health and have reported varying levels of habitual calcium intake
(inm6inwpththscwT
Figure 3 . Differences in the calcium intakes in the interve
0
500
1000
1500
2000
2500
John
son
Lloyd
Lee (
1994
)
Lee (
1995
)
Cadog
an
Bon
Study
Inta
ke o
f Cal
cium
(mg/
d)
calcium threshold (1300
Fig. 3). Three studies had comparable baseline calcium takes to this study,10-11,16 in that they were approxi-ately 900mg/d; three studies9, 14-15 had intakes between
00mg/d and 750mg/d and the other two studies had low takes of around 300mg/d.12-13 In contrast to the studies ith the high intakes of calcium, our study used a placebo roduct containing an additional 400mg/d of calcium - is lifted the control group to around 1300mg. This was e only study to lift both the control group and the
upplemented group over the “threshold” level for dietary alcium. This is an obvious flaw in the study design and ould need to be addressed if this study was repeated. he main reason for the control group receiving calcium
ntion trials that have been carried out in pre-pubertal children
jour
Nowso
nDibb
a
This Stud
y
Supplemented GroupControl Group
mg)
MJ Gibbons, N L Gilchrist, C Frampton, P Maguire, PH Reilly, RL March and CR Wall 346
The main reason for the control group receiving calcium in the supplement was that the ethics committee at the time felt there was sufficient scientific evidence that calcium has an effect on bone health and that it is unethical to not provide some calcium to the growing skeleton. In addition, the researchers did not know how high baseline calcium intake was. In contrast, other studies22 into calcium intake in NZ children and ado-lescents have reported much lower levels than this study. This may be one of the main reasons no differences were seen in indicators of bone health. We used a dairy calcium food product to supplement our subjects. Two previous studies11,15 looking at this age group have also used a calcium-containing food product to supplement their experimental group. Fluid milk pro-vides an ideal source of calcium because it is inexpensive; the lactose in milk aids in calcium absorption, and it provides other nutrients that support growth and development in children. Our baseline calcium intake was in line with the recommended intake of 800-900mg per day. However, the method used to determine habitual intake may not have been sufficient as we only used a single food fre-quency questionnaire. The most accurate method of die-tary assessment is a weighed food collection and then assays to determine calcium mineral content. Studies have shown that even if food record diaries are collected for 9 days for girls and 18 days for boys the “true” calcium intake may be inaccurate.20,21 Thus, the values we obtained for habitual calcium intake may potentially be slightly over or under estimated. Bone mineral content may be a better indicator of accretion in bone mineralisation than bone density, especially in this age group where the bones are still growing. We found no difference in bone mineral content between the two groups. Differences between the two groups were not observed in bone size during the supple-mentation period or during follow-up. This is different to results found by Bonjour et al.,11 where they found an increase in bone size between the supplemented and control groups. In their study the results of the children in the lowest quartile for baseline dietary calcium intake were compared to those in the highest quartile. It was only in the children with a low dietary calcium intake that an effect on bone size and height was evident. There is a common perception that dairy products are high in fat and consequently should be limited in the diet, or excluded. They are often thought to be the major cause of weight gain and undesirable changes in anthropometric measures. We found no adverse changes in the anthro-pometric measures, lean mass, fat mass, height and weight between the groups. This is consistent with pre-vious findings in teenage girls.22 Pubertal stage is a known determinant of bone mass. Thus using subject groups closely matched in this respect reduced the likelihood that differences in pubertal pro-gression would obscure interpretation of bone mea-surements. The two groups in our study were matched at baseline and remained matched. There were a large number of the children who stopped taking the sup-plement during the supplementation period due to a dislike of the product. Product acceptability is important
in an intervention study and a prolonged compliance run-in period in this study may have reduced the high drop-out. Compliance in the children that liked the supplement was high. This may be due to supervision from either the parent or the teacher in the classroom. A variety in the flavours was important for maintaining compliance and reducing the rate of withdrawals. Results from previous studies7,8 have suggested there is a decrease in dietary calcium intake from childhood to adolescence. By deve-loping dietary habits to include the frequent intake of milk during childhood and adolescence it is likely that calcium intakes will be higher in later years. Studies that look at compliance are important to determine whether children can achieve the recommended daily intake from diet. In this 2½ year study (18 months supplementation, 1 year follow-up) we did not observe a difference in bone mineral density in pre-pubertal children. We feel this is due to a high habitual dietary calcium intake, that even with minimal addition of calcium to the diet a threshold level was reached where no further benefit was seen. There were no adverse effects on body composition from the children taking the high calcium dairy product as throughout the study the children grew normally, and the mean height and weight remained between the 50th and 75th percentile for their age. Calcium supplementation in children needs to be targeted in those children with low habitual dietary calcium intake. Acknowledgement We would like to thank the children and their parents who participated in the study. The three schools and their staff that participated in the study: Redcliffs Primary School, St Martins School and Cashmere Primary School, and New Zealand Milk Limited, Wellington, New Zealand for there assistance with funding. References 1. Bonjour JP, Theintz G, Law F, Slosman D, Rizzoli R.
Peak Bone Mass. Osteoporosis Int 1994; Suppl 1: S7-S13. 2. Nelson DA, Simpson PM, Johnson CC, Barondess DA,
Kleerekoper M. The accumulation of whole body skeletal mass in third- and fourth-grade children: effect of age, gender, ethnicity and body composition. Bone 1997; 20: 73-78.
3. Bonjour JP, Theintz G, Buchs B, Slosman D, Rizzoli R. Critical years and stages of puberty for spinal and femoral bone mass accumulation during adolescence. J Clin Endocrinol Metab 1991; 73:555-563.
4. Matkovic V, Fontana D, Tominac C, Goel P, Chesnut CH. Factors that influence peak bone mass formation: a study of calcium balance and the inheritance of bone mass in ado-lescent females. Am J Clin Nutr 1990; 52:878-888.
5. Osteoporosis New Zealand Incorporated 2002, Website www.osteoporosis.org.nz
6. Russell DG, Parnell WR, Wilson NC, Faed J, Ferguson E, Herbison P, Horwarth C, Nye T, Reid P, Walker R, Wilson B, Tukuitonga C. NZ Food: NZ People. Key results of the 1997 National Nutrition Survey. Ministry of Health, Wellington, New Zealand, 2000.
7. Harel Z, Riggs S, Vaz R, White L, Menzies G. Adolescents and calcium: what they do and do not know and how much they consume. J Adol Health 1998; 22 (3): 225-228.
347 The effects of a high calcium dairy food on bone health in pre-pubertal children in New Zealand
8. Wosje KS, Specker BL. Role of calcium in bone health during childhood. Nutr Reviews 2000; 58 (9): 253-268.
9. Nowson CA, Green RM, Hopper JL, Sherwin AJ, Young D, Kaymakci B, Guest CS, Smid M, Larkins RG, Wark JD. A co-twin study of the effect of calcium supplementation on bone density during adolescence. Osteoporosis Int 1997; 7:219-225.
10. Johnston CC, Miller JZ, Slemenda PH Reister TK, Hui S, Christian JC, Peacock M. Calcium supplementation and increases in bone mineral density in children. N Eng J Med 1992; 327 (2): 82-87.
11. Bonjour J-P, Carrie A-L, Ferrari S, Clavien H, Slosman D, Theintz G, Rizzoli R. Calcium enriched foods and bone mass growth in prepubertal girls: a randomised, double-blind, placebo-controlled trial. J Clin Invest 1997; 99 (6): 1287-1294.
12. Lee WTK, Leung SSF, Wang S-H, Xu YC, Zeng WP, Lau J, Oppenheimer SJ, Cheng JC. Double-blind, controlled calcium supplementation and bone mineral accretion in children accustomed to a low calcium diet. J Am Diet Assoc 1994; 60: 744-750.
13. Lee WTK, Leung SSF, Leung DMY Tsang HS, Lau J, Cheng JC. A randomised double-blind controlled calcium supplementation trial, and bone and height acquisition in children. Brit J Nutr 1995; 74: 125-139.
14. Dibba B, Prentice A, Ceesay M Stirling DM, Cole TJ, Poskitt EM. Effect of calcium supplementation on bone mineral accretion in Gambian children accustomed to a low-calcium diet. Am J Clin Nutr 2000; 71: 544-549.
15. Cadogan J, Eastell R, Jones N Barker ME. Milk intake and bone mineral acquisition in adolescent girls: randomised, controlled intervention trial. Br Med J 1997; 315: 1255-1259.
16. Lloyd T, Andon M, Rollings N Martell JK, Landis JR, Demes CM, Eggli DF, Kieselhorst K, Kulin HE. Calcium supplementation and bone mineral density in adolescent girls. J Am Med Assoc 1993; 270 (7): 841-844.
17. Duke PM, Litt IF, Gross RT. Adolescents’ self-assessment of sexual maturation. Pediatrics 1980; 66:918-920.
18. Weaver CM, Martin BR, Plawecki KL, Peacock M, Wood OB, Smith DL, Wastney ME. Differences in calcium metabolism between adolescent and adult females. Am J Clin Nutr 1995; 61:577-581.
19. Wastney ME, Martin BR, Peacock M, Smith D, Jiang XY, Jackman LA, Weaver CM. Changes in calcium kinetics in adolescent girls induced by high calcium intake. J Clin Endo Metab 2000; 85 (12): 4470-4475.
20. Miller JZ, Kimes T, Hui S, Andon MB, Johnston CC. Nutrient intake variability in a paediatric population: implications for a study design. J Nutr 1991; 121: 265-274.
21. Gibson RS. Principles of nutritional assessment. New York: Oxford University Press, 1990.
22. Merrilees MJ, Smart EJ, Gilchrist NJ Frampton C, Turner JG, Hooke E, March RL, Maguire P. Effects of dairy food supplements on bone mineral density in teenage girls. Eur J Nutr 2000; 6: 256-261.
Asia Pac J Clin Nutr 2004; 13 (4): 348-352 348
Original Article Comparison of serum levels of iron, zinc and copper in anaemic and non-anaemic pregnant women in China Ai-Guo Ma MD1, Xue-Cun Chen MD2, Rong-Xian Xu MD3, Ming-Ci Zheng PhD4 Yu Wang MD5 and Jue-Sheng Li MD1 1 Institute of Human Nutrition, Medical College of Qingdao University, Qingdao, China; 2 Institute of Nutrition and Food Hygiene, Chinese Academy of Preventive Medicine, Beijing, China; 3 Fujian Medical University, Fuzhou, China; 4 Guilin Medical College, Guilin, China; 5 Lanzhou Medical College, Lanzhou, China.
Zinc and copper deficiency is associated with anaemia or iron deficiency and affects fetus growth and pregnant women during pregnancy. To examine iron, zinc and copper status of Chinese pregnant women with and without anaemia in the third trimester, 1185 subjects were enrolled for measurements of Hb, ferritin, transferrin, soluble transferrin receptor (sTfR), and serum iron, zinc and copper. The results showed that there were lower levels of ferritin (14.1 μg/L) and transferrin (3.33 g/L) in subjects with Hb≤100g/L as compared with subjects with Hb≥101g/L. sTfR levels in subjects with Hb≤100g/L were significantly higher than those in subjects with Hb≥120 g/L (38.5 nmol/L vs. 25.04 nmol/L, P<0.001). Serum iron was lower in subjects with Hb≤100g/L than those with Hb≥120 g/L (871µg/L vs. 990 µg/L, P<0.01). Lower levels of serum iron and zinc were also found in anaemic (Hb<110g/L) as compared with non-anaemic women (Hb≥110g/L). Frequencies of marginal deficiencies in serum iron and zinc were 41.58% and 51.05% respectively higher in anaemic than in non-anaemic subjects. Distribution of serum zinc and iron showed a deceasing trend as Hb decreased. Few anaemic as well as non-anaemic subjects had copper deficiency although copper and Hb levels were found inversely correlated and the ratio of copper/iron was higher in anaemic than in non-anaemic group. In conclusion, a lower level of serum zinc in anaemic pregnant women might be related to anaemia and iron deficiency during pregnancy. Therefore, combined zinc and iron supplementation should be recommended to Chinese pregnant women, especially those with anaemia.
Keyword: anaemia, pregnancy, iron, zinc, copper, Qingdao, China Introduction Minerals have important influences on the health of pregnant women and the growing fetus. Iron deficiency results in anaemia, which may increase the risk of death from hemorrhage during delivery.1 Zinc deficiency is common in developing countries.2 It has been estimated that 82% of all pregnant women worldwide suffer from zinc deficiency.3 Maternal zinc deficiency is a public health problem because zinc has an important role in the expression of genetic potential, nuclear acid metabolism, and protein synthesis.4,5 Zinc is therefore critical for fetal growth. Deficiencies in other minerals such as copper and calcium have also been associated with complications of pregnancy and abnormal fetal development.6 Multi-mineral interactions of iron-zinc-copper have been re-ported. A high concentration of iron can interfere with zinc uptake when no dietary ligands are present. High iron intakes may also interfere with copper absorption when both are taken simultaneously. Modest increase in zinc intake may have positive effect on immune functions but higher amounts can interfere with copper and iron absorption, which in turn can adversely affect immune functions.7 A recent study examining serum copper, iron, and zinc levels in Chinese subjects has been reported.8 However, the study involved only 70 gravidas and 45 unpregnant
women. To date, no large scale studies have been per-formed to investigate the mineral status of pregnant women in China. The present study was therefore carried out to determine serum iron, zinc, and copper levels in women with and without anaemia during the third trimester of their pregnancy.
Subjects and methods Subjects 1185 clinically normal pregnant women, aged 20-35 y, were examined during the third trimester of their preg-nancy. All subjects had no abnormal bleeding, no iron and other mineral supplements, and attended the clinics re-gularly for prenatal care. They were enrolled in a random manner in each clinic and can therefore be regarded as a random sample of all pregnant women attending the clinic. Subjects were from two rural areas and two developing cities in China. Pregnant women in the cities were mainly Correspondence address: Professor Aiguo Ma, Institute of Human Nutrition, Medical College of Qingdao University, 38# Dengzhou Road, 266021, Qingdao P.R.china. Fax: +86 532 3812434 E-mail: [email protected] Supported by a grant from the Nestlé Foundation Accepted 25 June 2004
349 A-G Ma, X-C Chen, R-X Xu, M-C Zheng , Y Wang and J-S Li
Hb≤100 101 – 119 Hb≥120 P
N Mean ± SD1 N Mean ± SD N Mean ± SD Haematocrit (l/L) 342 29.8 ± 4.82 a 334 33.0 ± 3.76b 385 37.8 ± 4.93c Serum iron (µg/L ) 301 871 ± 0.48a 382 1080 ± 0.75b 347 990 ± 0.54c 0.000 Ferritin (μg/L) 316 14.1±10.53a 402 18.8 ± 13.52b 407 19.34 ± 18.63bc 0.000 Transferrin (g/L) 351 3.33 ± 0.44a 406 3.37 ± 0.42ab 428 3.46 ± 0.46c 0.000 sTfR ( nmol/L) 43 38.5 ± 22.60a 90 25.19 ± 9.99b 23 25.04 ± 7.87bc 0.000
1Values (mean±SD) within rows with different superscript letters are statistically different at P<0.05 using one-way ANOVA.
Table 1. Iron status of pregnant women with anaemia and nonanaemia in China1
from the middle socio-economic class whereas subjects in the rural areas were from the lower class. The study was carried out in accordance with the ethical standards of the local authorities. Informed consent was obtained from all subjects.
Sample collection and methods Fasting venous blood samples from 1185 subjects (5 ml/person) were taken at antenatal examination during 7 to 9 o’clock in the morning.9 Samples were stored on ice for transport to the laboratory. Haematocrit and haemo-globin concentrations were measured. Serum was sepa-rated from blood by centrifugation at 2000 × g for 15 min at room temperature upon arrival. Serum samples were then stored separately at -200C in the dark and transported to the laboratory for analyses of ferritin, transferrin, serum iron, zinc, copper and sTfR. Serum sTfRs were measured by enzyme-linked immu-nosorbent assay using a commercial kit (R&D Systems, Minneapolis, USA). The concentrations of test samples were determined in a Bio Rad Microplate Reader (model 3550: Richmond. CA) set at 450 nm. To test the reli-ability of sTfR kits, three control serum samples were assayed in each experiment. Haemoglobin concentrations were measured with the cyanomethaemoglobin method and haematocrit with the micromethod. A standard haemoglobincyanide solution was used for quality control of haemoglobin measure-ments. Measurement of serum ferritin was performed by radioimmunoassay,10 as described by the manufacturer (The North Biol.Tec. Institute, Beijing, China). Trans-ferrin (TRF) was determined by using a commercially available kit (Yadu Biotech Co. Shanghai, China). Levels of serum zinc, iron and copper were determined by atomic absorption spectrophotometry (PE 3100, USA). Iron status was assessed by using multiple criteria. Abnormal values are classified as: Hb < 110 g/L, HCT
(haemotacrit) <33%, SF<12 µg/L and TRF<2.1 g/L. Marginal deficiency of serum nutrients was defined as serum iron <700µg/L, serum zinc <700µg/L, and serum copper <700µg/L. Statistical analysis Data were cleaned by visual and logical checks and analyzed by using the SPSS program (10.0 version). Ratios of serum zinc/iron or zinc/copper were obtained by dividing individual serum zinc with iron or copper. Differences in biochemical parameters including haema-tocrit, serum iron, ferritin, transferring and sTfR among three groups of subjects (Hb≤100g/L, Hb 101-119g/L, Hb≥120 g/L) were analyzed by one-way ANOVA (F-test); serum zinc, copper and iron in two groups of subjects (Hb<110g/L, Hb ≥110 g/L) were examined by Independent Samples t-test. Statistical significance was accepted at a probability level of 0.05. Data were expressed as mean ± SD. Results 1185 subjects were divided into three groups based on their haemoglobin concentrations. Iron status of pregnant women was evaluated using five parameters including haematocrit, serum iron, ferritin, transferrin and sTfR. Comparisons of these parameters among three groups of subjects (Hb ≤ 100g/L, 101-119g/L, and ≥120 g/L) are shown in Table 1. Levels of serum iron in Hb ≤ 100g/L and Hb ≥ 120 g/L were 871µg/L and 990µg/L (P<0.05) respectively. There were low levels of ferritin (14.1 μg/L) and transferrin (3.33 g/L) in subjects with Hb ≤ 100g/L as compared with other groups. sTfR levels of 156 subjects were examined. A higher value of sTfR with 38.5 nmol/L was found in Hb ≤ 100g/L than 101-119g/L and Hb ≥ 120 g/L groups (P <0.001), while there was no difference between 101-119g/L and Hb ≥ 120 g/L (P>0.05).
Table 2. Comparison of three serum minerals between Hb <110 and ≥ 110 g/L
Hb<110 Hb≥110
P
N Mean ± SD1 Mean±SD zinc (μg/L) 523 701.8 ± 220.6 719.0 ± 252.0 0.000 iron (μg/L) 509 888.6 ± 498.2 1087.3 ± 700.1 0.000 copper (μg/L) 520 1733.3 ± 573.8 1788.3 ± 539.8 0.107 Cu/Zn 508 2.67 ± 1.16 2.57 ± 1.11 0.146 Cu/Fe 488 2.44 ± 1.51
N 555 521 550 545 510 2.15 ± 1.306 0.001
Serum levels of iron, zinc and copper in anaemic and non-anaemic pregnant women in China 350
Table 3. Frequencies of marginal deficiency of iron, zinc and copper in anaemia1 and nonanaemia
suTwloTbraan seBsu
Anaemia Non-anaemia Abnormal range P
N % % Haematocrit 536 77.99 19.62 <33% 0.000 Ferritin 531 51.04 38.18 <12µg/L 0.000 Serum iron 505 41.58 30.49 <700µg/L 0.000 Serum zinc 523 51.05 45.05 <700µg/L 0.048 Serum copper 520 4.23
N 555 521 550 545 510 1.64 <700µg/L 0.011
1H
b<110g/L as anaemia, Hb≥110g/L as nonanaemia
zinc
0.50
0.60
0.70
0.80
50 70 90 110 130 150
haemoglobin concentration (g/L)
zinc
(ug/
L)
iron
0.20
0.60
1.00
1.40
50 70 90 110 130 150
haemoglobin concentration (g/L)
seru
m ir
on (u
g/L)
(a) (b)
copper
1.20
1.60
2.00
2.40
2.80
50 70 90 110 130 150
haemoglobin concentration (g/L)
copp
er (u
g/L)
Cu/Zn
2.00
2.50
3.00
3.50
4.00
4.50
5.00
50 70 90 110 130 150
haemoglobin concentration (g/L)
Cu/
Zn
(c) (d)
Figure 1. (a) distribution of serum zinc to haemoglobin concentrations; (b) distribution of serum iron to haemoglobin concentrations;
(c) distribution of serum copper to haemoglobin concentrations; (d) distribution of serum Cu/Zn to haemoglobin concentrations.
Serum levels of iron in 1030 subjects, zinc in 1078 bjects, and copper in 1070 subjects were shown in
able 2. Mean serum zinc was 701.8μg/L and serum iron as 888.6μg/L in the anaemic group, which were much wer than those in the non-anaemic group (P <0.001). here was no difference in mean serum copper levels etween anaemic and non-anaemic groups (P >0.05). The tio of copper/iron was higher in anaemic than in non-aemic subjects (P <0.001).
Frequency of abnormal haematologic results and rum mineral concentrations are shown in Table 3. ased on ferritin levels, iron depletion in anaemic bjects was more common (51.04%) than those without
anaemia (38.18%). Frequencies of marginal deficiencies of serum iron and zinc were higher (41.58% and 51.05%) in anaemic than non-anaemic (30.49% and 45.05%) women. Notably, few anaemic as well as non-anaemic subjects had copper deficiency. Figure 1 showed the changeable trends of serum zinc, iron and copper with the distribution of haemoglobin concentrations. Serum zinc and iron levels were increased as Hb increased (Fig. 1a & 1b). The marginal deficiencies of iron and zinc kept below the level of Hb <90g/L. Nevertheless, serum copper levels and the copper/zinc ratio were decreased as Hb increased (Fig. 1c and 1d). There was a steep decrease of copper at the lower level of
351 A-G Ma, X-C Chen, R-X Xu, M-C Zheng , Y Wang and J-S Li
Hb value from 70-90 g/L, while a steep increase was observed with zinc and iron. Discussion The present study, which involved a large population of Chinese pregnant women, showed that there were iron and zinc deficiencies in anaemic women in the third trimester of their pregnancy. Consistent with reported findings11,12 some biochemical indices for trace elements fall in parallel with red blood cell volume, and serum con-centrations of zinc and iron were positively correlated with maternal haemoglobins. The positive correlation of haemoglobin concentrations with zinc and iron indicates that deficiencies of the two minerals were common and more severe in anaemic pregnant women. Iron and zinc deficiencies in pregnant women may result from an expanded blood volume, an increased need for zinc and iron, and also likely, poor intake and poor bio-absorption of iron and zinc. Although the average iron intake (23.52 mg/d) by anaemic women was close to the Chinese RDA (28 mg/d) for pregnant women, traditional Chinese vegetable diets may result in low iron bio-absorption.13 The daily intake of zinc by anaemic subjects reached 19.83 mg/d. However, serum zinc levels in these people were very low. The high frequencies of marginal zinc deficiency (<700µg/L) in both anaemic (51.05%) and non-anaemic (45.05%) groups indicate that zinc defi-ciency is common in Chinese pregnant women. Several dietary factors are known to affect zinc absorption as a result of physico-chemical interactions in the intestine. Phytate, a component in plants with the highest concen-tration found in seeds (cereal grains/legumes/nuts), inhibits zinc absorption, as does calcium. Physiological states, such pregnancy and lactation, increase the demand for absorbed zinc.14 On the contrary, dietary proteins enhance zinc absorption. The molar ratio of phytate to zinc (P:Z) in diets has been used to estimate the absorption efficiency of zinc, with P:Z<5 being associated with relatively high absorption of zinc, P:Z 5-10 with moderate absorption, and P:Z >15 with low absorption.15 Serum copper concentrations were in inverse relation-ship with maternal haemoglobin concentrations. Although frequencies of marginal copper deficiency were low in both anaemic (4.23%) and nonanaemic (1.64%) groups, copper deficiency has deleterious effects on pregnancy and therefore should not be neglected. On the other hand, high copper intake can decrease zinc absorption by com-peting with zinc at absorption sites.16 Iron deficiency during pregnancy is common and has serious short and long term consequences such as fetal growth retardation and cardiovascular problems in the adult offsprings. In our study, we noted a higher ratio of copper/iron (2.44) in anaemic (Hb <110g/L) than in non-anaemic subjects (Hb ≥110 g/L). The higher copper/iron ratio in anaemic subjects may be attributed to the fact that iron deficiency increases copper levels in the maternal liver, serum and placenta, although it has much less effect in the fetal serum and liver. Apart from maternal ceruloplasmin, mRNA levels of copper-regulated proteins are not changed. Copper oxidase, which is believed to fulfil the function of hephaestin in placenta, is regulated by copper as well as by iron.17 Our findings are consistent with those
of Huang et al.,18 that hair concentrations of copper, zinc, iron and calcium in the gravidas in all three trimesters were significantly lower than those in non-pregnant women in Tianjin, China.
During pregnancy, nutrients are transferred from mother to fetus across the placenta. To compensate for iron deficiency, proteins involved in iron transfer, such as the transferrin receptor, are upregulated. Serum transferrin receptor (sTfR) has thus been used as an indicator for early tissue iron deficiency. Unlike ferritin, sTfR con-centrations are not affected by conditions such as in-flammation and are therefore considered a better para-meter for determining the iron status.19 sTfR concen-trations have been shown to be associated with serum iron, serum ferritin and haemoglobin.20 In our study, 50.5% of pregnant women were found anaemic (Hb <110 g/L), among whom 51.0% had depleted iron stores reflected by low ferritin concentration (<12 μg/L). sTfR values in severe pregnant women (Hb <100 g/L) were 1.53 times higher than other subjects (Hb 101-119g/L and ≥120 g/L). There was no difference in sTfR levels within the latter group. The findings suggest that severe anaemia may induce sTfR, whereas moderate anaemia (Hb 101-110g/L) had no such obvious effect. Elevation of sTfRs resulting from severe iron deficiency has also been reported in pregnant women in the later part of preg-nancy.21 Hence, sTfR assay may be a better test for ill and hospitalized patients in whom, despite having iron deficiency, ferritin levels are normal or elevated.22
Prophylactic iron supplementation is recommended in developing countries for all pregnant women in the second and third trimesters of pregnancy. In other countries, such as Great Britain, iron supplementation is recommended only to anaemic women with diagnosed iron deficiency.23 Safety issues related to iron supplemen-tation have been raised.24 Zinc supplementation during pregnancy is recommended by the Institute of Medicine (IOM) in USA although IOM concluded that there was insufficient evidence to support the supplementation.25 The high frequencies of marginal zinc deficiency (<700µg/L) in both anaemic (51.05%) and non-anaemic (45.05%) groups observed in our study indicate that zinc deficiency is common in Chinese pregnant women. Our findings thus provide strong evidence that necessitates zinc supplementation for pregnant Chinese women. Pro-phylactic doses of 20-25 mg elemental zinc/d have generally been used in pregnant women in developing countries.26 The latest survey in China showed that the daily intake of zinc by anaemic pregnant women (19.83 ± 10.02 mg/d) and non-anaemic pregnant women (18.39 ± 8.49 mg/d) was lower than the generally recommended dosage. So zinc or combined iron and zinc supplemen-tation for pregnant women should be considered in China.13 Copper deficiency has not been observed, sup-plementation in pregnancy is therefore not recommen-ded.25 On the other hand, when zinc supplements are given to individuals with low copper intakes, a copper supplement should be given to compensate for the zinc-copper interaction.27
In conclusion, the present study revealed the unbalanced nutritional status of iron, zinc and copper in Chinese pregnant women. To decrease the adverse
Serum levels of iron, zinc and copper in anaemic and non-anaemic pregnant women in China 352
outcomes of mineral malnutrition and to improve fetal growth and the health of pregnant women, further efforts should be made to develop and implement an appropriate mineral supplementation regime for Chinese pregnant women. Acknowledgements We sincerely thank Dr Evert Shouten, Dr Joseph Hautvast, Dr Beat Schürch and Dr Zhang Huihong for their help. We also thank Zhang Xiuzhen Liang Hui and Li Yong for coordinating the study; Du wei, Wang Zhixu, Xu Hongwei and Zhang Shehua for measuring sTfR, transferrin, ferritin. We are grateful to Liu Hui and Xin Qianyi for their technical assistance with serum zinc, iron and copper analyses;Han Xiuxia and Song Xuxia for data analysis. References 1. Black RE. Micronutrients in pregnancy. Br J Nutr 2001; 85
(suppl 2): 193-197. 2. Jameson S. Zinc status in pregnancy: the effect of zinc
therapy on perinatal mortality, prematurity, and placental ablation. Ann N Y Acad Sci 1993;678:178-192.
3. Caulfield LE, Zavaleta N, Shankar AH, Merialdi M. Potential contribution of maternal zinc supplementation during pregnancy to maternal and child survival. Am J Clin Nutr 1998; 68 (suppl): 449s-508s.
4. Saskia JMO, Clive EW, Robert EB. The need for maternal zinc supplementation in developing countries: an unresolved issue. J Nutr 2003; 133: 817s-827s.
5. Villar BL, Falchuck KH. The biochemical basis of zinc physiology. Physiol Rev 1993;73:79-118.
6. Villar J, Merialdi M, Gülmezoglu AM, Abalos E, Carroli G, Kulier R, de Oni M. Nutritional interventions during pregnancy for the prevention or treatment of maternal morbidity and preterm delivery:An overview of randomized controlled trial. J Nutr 2003; 133: 1606s-1625s.
7. Lonnerdal B. iron-zinc-copper interactions. In micro-nutrient interactions: impact on child health and nutrition. Washington, DC: ILSI Press,1996: 3-10.
8. Zhou CL, Zhang XG, Liu CM. Determination of serum Cu, Fe and Zn in pregnant women. Guizhou Medical J 2000;24: 333-334.
9. Hotz C, Peerson JM, Brown KH. Suggested lower cutoffs of serum zinc concentrations for assessing zinc status: reanalysis of the second national health and nutrition examination survey data (1976-1980). Am J Clin Nutr 2003; 78: 756-764.
10. Chu SL, Chen YZ, Xue JZ, Ni QG, Wang YL. Measurement of clinical application of ferritin by radioimmunoassay. Chin J Hemato 1983; 4: 296-298.
11. Bothwell TH. Iron requirements in pregnancy and stra-tegies to meet them. Am J Clin Nutr 2000;72 (suppl): 257S-264S.
12. Ladipo OA. Nutrition in pregnancy: mineral and vitamin supplement. Am J Clin Nutr 2000; 72 (suppl): 280s-290s.
13. Ma AG, Chen XC, Zheng MC, Wang Y, Xu RX, Li JS. Iron status and dietary intake of Chinese pregnant women with anaemia in the third trimester. Asia Pac J Clin Nutr 2002; 11 (3):171-175.
14. Krebs NF. Overview of zinc absorption and excretion in the human gastrointestinal tract. J Nutr. 2000;130 (5S Suppl): 1374S - 1377S.
15. Hotz C, Lowe NM, Araya M, Brown KH. Assessment of the trace element status of individuals and population: The example of zinc and copper. J Nutr 2003; 133: 1563s-1568s.
16. Sheldon WL, Aspillaga MO, Smith PA, Lind T. The effect of oral iron supplementation on zinc and magnesium levels during pregnancy. Br J Obstet Gynaecol 1985; 92: 892-898.
17. Gambling L, Danzeisen R, Fosset C, Andersen HS, Dunford S, Srai SK, MCArdle HJ. Iron and copper interactions in development and the effect on pregnancy outcome. J Nutr 2003; 133 (5 Suppl 1): 1554S-1556S.
18. Huang HM, Leung PL, Sun DZ, Zhu MG. Hair and serum calcium, iron. Copper, and zinc levels during normal pregnancy at three trimester. Biol Trace Elem Res 1999; 69 (2): 111-120
19. Ferguson BJ, Skikne BS, Simpson KM, Bayne RD, Cook JD. Serum transferrin receptor for the detection of iron deficiency in pregnancy. Am J Clin Nutr 1992; 54: 1077-1081.
20. de Azevedo Paiva A, Rondo PH, Guerra-Shinohara EM, Silva CS. The influence of iron, vitamin B(12), and folate levels on soluble transferrin receptor concentration in pregnant women. Clin Chim Acta 2003;334(1-2):197-203.
21. Simmons WK, Cook JD, Bingham KC, Thomas M, Jackson J, Jackson M, Ahluwalia N, Kahn SG, Patterson AW. Evaluation of a gastric delivery system for iron supplementation in pregnancy. Am J Clim Nutr 1993; 58: 622-626.
22. Akinsooto PJ, Ojwang T, Govender J, Moodley CA, Connolly V. Soluble transferrin receptors in anaemia of pregnancy. J Obstet Gynaecol 2001; 21(3):250-252.
23. Hibbard BM. Iron and folate supplemnts during pregnancy: supplementation is valuable only in selected patients. BJM 1988; 297:1324-1326.
24. Yip R. Significance of an abnormally low or high haemoglobin concentration during pregnancy: special consideration of iron nutrition. Am J Clin Nutr 2000; 72 (suppl): 272s-279s.
25. Institute of Medicine. Nutrition during pregnancy. Washington, DC; National Academy Press, 1990
26. Gibson RS, Ferguson EL. Nutrition intervention strategies to combat zinc deficiency in developing countries. Nutr Res Rev 1998; 11: 115-131.
27. National Research Council. Recommended dietary allowances. 10th ed. Washington, DC: National Academy Press, 1989.
353 Asia Pac J Clin Nutr 2004;13 (4):353-358
Original Article Effects of 4 weeks iron supplementation on haema-tological and immunological status in elite female soccer players Hyung-Sook Kang PhD1 and Tatsuhiro Matsuo PhD2 1 Korea Sports Medical Nutrition Institute, 48-19 Song Pa-Dong,Song Pa-Gu, Seoul, Korea
2 Faculty of Agriculture, Kagawa University, Ikenobe, Miki-cho, Kita-gun, Kagawa 761-0795, Japan
The effects of 4 weeks iron supplementation on haematological and immunological status were studied in 25 elite female soccer players aged 20-28 years. The subjects were randomized and assigned to one of the following two groups; subjects given 40 mg/day iron supplementation (S group) or those given placebo (C group). The oral iron supplementation (40 mg elemental iron) was taken in 15 ml solution once a day by the S group, and the C group took a placebo for 4 weeks. Daily energy and protein intakes met the Korean Recommended Dietary Allowances. Blood haemoglobin concentration did not change in the S group, but decreased significantly (P<0.05) in the C group over the 4-week experimental period. Haematocrit, mean cell volume, mean cell haemoglobin and total iron binding capacity decreased significantly, and mean cell haemoglobin concentration increased significantly (P<0.05) in both the S and C groups. Plasma ferritin concentration increased significantly (P<0.05) in the S group, but did not change in the C group. The change of plasma immunolgical parameters and erythrocyte anti-oxidative enzyme activities were almost the same between the S and C groups. These results suggest that 4 weeks of iron supplementation by elite female soccer players significantly increased body iron stores and inhibited decrease of haemoglobin concentration induced by soccer training.
Key words: iron supplementation, haematological parameter, immune function, elite soccer player, Korean Introduction Iron deficiency is one of the leading nutritional problems in the world.1,2 Its most common clinical manifestation is anaemia, and the work impairment caused by iron defi-ciency anaemia has been thoroughly documented.3-6 Iron deficiency reduces physical performance4,6-8 probably through combined effects on oxygen consumption8-10 and muscle metabolism.10-12 Many studies suggest that elite female athletes may be at increased risk of iron defi-ciency.13-16 Clement and Asmundson13 reported that 82% of female Canadian distance runners were iron deficient, as estimated by serum ferritin levels, which are believed to accurately reflect the size of the body iron stores.16 Another report found that despite normal haemoglobin (Hb) and serum iron values, bone marrow showed either an absence or only traces of iron.14 Several other investigators have confirmed this surprisingly high incidence of iron de-ficiency in active persons.9, 15 As female athletes are already at increased risk due to the superimposed requirements related to menstruation, the possibility of increased iron demand associated with exer-cise is of particular concern to those engaged in physical activity. As a result, a variety of supplementation regimens are recommended to ensure adequate iron status.17 These intervention programs are predicated on the assumption that nutritional iron deficiency does indeed lead to signi-ficant disability. Reduction of iron deficiency is also aimed at reducing the risk of developing anaemia and perhaps other performance-related problems.17 Rowland et al.,18
noted that 4-week oral iron treatment improved serum ferritin levels (8.7 to 26.6 μg/L) in nonanaemic iron de-ficient runners. Schoene et al.,10 studied the effect of 2 weeks of iron therapy on exercise performance in trained, mildly iron deficient female athletes. They reported that performance was unchanged after therapy.19 On the other hand, the immune system seems par-ticularly sensitive to the availability of iron.19 Iron is needed for DNA synthesis, and for the activity of the iron-dependent enzymes that are involved in the killing of microorganisms. An iron deficiency can thus cause an overall atrophy of immune tissues.20 Alterations in immune responses can occur early in the course of reduction of iron stores.21 Various immune responses can be suppressed by vigorous and intensive athletic training.22-24 Thus, exercise-induced immunological impairment may be related to iron deficiency. However, evaluation of the relationship between iron status and immune function has not been adequately investigated in elite female athletes. Thus, the purpose of this investigation was to determine whether 4 weeks of iron treatment would improve haematological Correspondence address: Dr Tatsuhiro Matsuo, Faculty of Agriculture, Kagawa University, 2393 Ikenobe, Miki-cho, Kita-gun, Kagawa 761-0795, Japan. Tel: 81-87-891-3082; Fax: 81-87-891-3021 Email: [email protected] Accepted 23 July 2004
Hyung-Sook Kang and Tatsuhiro Matsuo 354
jects S C
N = 11 N = 14
22.6 ± 2.0 23.8 ± 2.8 165.2 ± 6.0 163.9 ± 5.7 57.9 ± 4.5 57.0 ± 4.9 21.3 ± 1.7 21.2 ± 1.2 24.1 ± 3.1 23.7 ± 3.1 14.0 ± 2.5 13.5 ± 2.1 43.9 ± 3.1 43.5 ± 3.9
lement group; C, Control group.
Table 2. Characteristics of sub
Age years Height cm Weight kg Body mass index kg/m2 Percentage body fat % Fat mass kg Fat free mass kg
Values are means±SD. S, Iron supp
Table 1. Example of daily schedule for the subjects1
Time of day (h) Items
06:00 - 07:30 Climbing and walking
08:00 - 10:00 Breakfast and rest 10:00 - 12:00 Soccer training 12:30 - 14:30 Lunch and rest 14:30 - 18:00 Soccer training 18:30 - 20:00 Dinner and rest 20:00 - 21:30 Weight training 1One month before Universiade Game.
parameters in elite female athletes with nonanaemic iron deficiency. In addition, we investigated the effects of iron supplementation on immunological status in female iron deficient athletes. Methods Subjects Twenty-five elite female soccer players (20-28 years old) were recruited from the Korean national team for this study. The subjects severely trained for 7-9 hours every-day. An example of a daily training schedule is shown in Table 1. All procedures were approved in advance by the Ethics Committee of the Korean Sports Medical Nutrition Institute and were in accordance with the Helsinki Declaration of 1964, as revised in 2000. After a detailed explanation of this study, each subject gave her informed written consent. The subjects were determined to be free of disease by a medical examination before the study. No subjects were using illegal drugs or taking medications that affect body weight. The day of the menstrual cycle when they began and ended the study was noted because fluctuations in metabolic parameters can occur during the cycle.25 Subjects started the iron supplementation with their training season immediately after biochemical pre-tests. Because the experimental period was 4 weeks, most of the women were at about the same point of their cycle (mid-follicular phase) when blood haematological and immunological parameters were remeasured at the end. Subjects were randomized and assigned to one of the following two groups: (1) subjects given 40 mg/day iron supplementation (S group) or (2) those given placebo (C group). The characteristics of subjects belonging to the S and C groups are shown in Table 2.
Supplementation Iron supplement and placebo were purchased com-mercially (Daewoong Pharmaceutical Ltd., Seoul, Korea). The experimental treatment consisted of oral iron supple-mentation (40 mg elemental iron) taken in 15 ml solution once a day, as tolerated. The C group took a placebo, which appeared identical to the active agent and was taken in 15 ml solution once a day, as tolerated.
Dietary intake The daily food intake of the subjects was not controlled, but energy and protein intake met the Korean Recom-mended Dietary Allowances (RDA). A dietary assessment
was performed using a 24-h recall method. The subjects were asked to record their complete food intake during the 3 days of the study. The daily intake of nutrients was calculated from these records using a nutritional analysis program (CAN-pro, Korean Nutrition Society).
Measurement procedures Subjects underwent several measurements before starting the experiment and again after the 4 weeks while still in training. End measurements were conducted >24 h after the previous exercise. The procedures were performed in the following order: blood and plasma biochemical tests (haematological and iron-related measurements26, white blood cell counts27, leucocyte differential27, plasma immunoglobulin28, erythrocyte antioxidative enzyme acti-vities29-31). Evaluations of biochemical parameters of blood and plasma were requested from Green Cross Reference Laboratory Co., (Seoul, Korea).
Body composition The subjects’ height, weight and measurements were taken by conventional methods. Skinfold thickness was determined by caliper. Percentage of body fat, fat mass and FFM were calculated from skinfold thickness (subscapular and triceps) as described previously.32,33
Percentage of body fat (%BF) is calculated with the following formula: BS = W0.425 X H0.725 X 71.84 /10000 BS, Body surface area (m2); W, body weight (kg); H, height (cm) BD = 1.0923-0.000514 x x= (SFt + SFs) X BS/W X 100 BD, Body density (kg/m3); SFt, triceps skinfold thickness (mm); SFs, subscapular skinfold thickness (mm) %BF = (4.570 / BD - 4.142)×100 Statistical analysis The mean and standard deviation (SD) were reported for all measurements. Data were analysed using repeated measures ANOVA followed by Student’s paired t-tests to show differences in variables from baseline to 4 weeks and using Student’s unpaired t-tests to show differences in variables between the S and C groups. A value of P <0.05 was considered to be significant.
355 Iron supplementation and haematological and immunological status in elite female soccer players
Results Dietary intake Daily nutrient intakes during the experiment and per-centages of RDA are shown in Table 2. Daily intake of energy and nutrients were not different between the S and C groups. Energy intake was about 100% of RDA and protein intake was over 100% of RDA for both the S and C groups (Table 3). Calcium, iron, and vitamin A were insufficient compared to RDA (Table 3). Percentages of energy as protein, fat and carbohydrate were 14.8, 28.2 and 57% for the S group and, 15.1, 28.9, and 56% for the C group. The sources of the iron from daily meals included a combination of meats, fish, eggs, beans, grains, and vegetables. Mean percentages of iron sources were 32.4% for animals, 67.6% for plants in S group and 29.4% for animals, 70.6% for plants in C group.
Haematological parameters All pre-and post-experiment blood and plasma haema-tological test results were within the standard values for adult Korean women. Blood haemoglobin concentration did not change in the S group, but decreased significantly (P<0.05) in the C group over the 4-week experimental period (Table 4). Red blood cells and plasma iron con-centration did not change in either the S or C group (Table 4). Haematocrit, mean cell volume (MCV), mean cell haemoglobin (MCH) and total iron binding capacity (TIBC) decreased significantly, and mean cell haemo-globin concentration (MCHC) increased significantly (P<0.05) in both the S and C groups (Table 4). The change in MCH was significantly (P<0.05) greater in the C group than in the S group (Table 4). Plasma ferritin concentration increased significantly (P<0.05) in the S group, but did not change in the C group (Table 3).
Immunological parameters White blood cells and plasma immunoglobulin were not different between pre- and post-experiment results in either the S or C group (Table 5). Percentages of neutro-phils and basophils did not change in either the S or C group over the 4-week experimental period (Table 5). The percentage of lymphocytes decreased, and percen-tages of monocytes and esoinophils increased in both the
S and C groups, but the change of esoinophils in the C group was not significant (Table 5).
Antioxidative enzyme activities Erythrocyte glutathione peroxidase (GPx) and Catalase activities were not different between pre- and post-experiment results in either the S or C group (Table 6). Superoxide dismutase (SOD) activity increased signifi-cantly (P<0.05) in both the S and C groups over the 4-week experimental period (Table 6). Discussion The results show that 4 weeks iron supplementation (a daily dose of 40 mg elemental iron) to elite female soccer players significantly increased plasma ferritin concen-tration and inhibited decrease of Hb concentration in-duced by soccer training. The change in Hb values pre- to post-was not different in the S group. This is in agree-ment with Pate et al.,34 who found that oral iron sup-plementations, when administered to nonanaemic female athletes, have no statistically significant effects on Hb levels. However, they did note a modest improvement from 14.4 to 15.0 g/dl over their treatment period (5-9 weeks with 50 mg elemental iron per day). There was thus some question prior to the present study as to whether changing Hb levels would influence our results. Newhouse et al.,35 reported that 8-week iron supple-mentation (100 mg elemental iron per day) did not in-fluence Hb concentration (13.4 to 13.5 g/dl) in prelatent or latent iron deficient females. Rowland et al.,18 demon-strated that 4-week iron supplementation (975mg ferrous sulfate per day) has no effect on Hb concentration (13.1 to 12.9 g/dl) in female endurance runners. These results suggest that oral iron supplementation did not increase blood Hb level in female athletes. However, the results of the present study may indicate that iron supplementation inhibits the decrease in Hb level induced by heavy training. In this study, most of the haematological para-meters (Hb, Ht, MCV, MCH, MCHC, and TIBC) decreased over the 4-week experimental period (training season) in both the S and C groups. Newhouse and Clement36 suggested that iron deficiency induced by
Table 3. Dietary intake S C N =11 N = 14 Energy kcal 2154 ± 377 (107)* 2083 ± 343 (103) Protein g 78 ± 10 (133) 66 ± 11 (129) Fat g 68 ± 11 66 ± 13 Carbohydrate g 312 ± 69 292 ± 61 Fibre g 4.5 ± 0.6 4.1 ± 1.2 Calcium mg 538 ± 121 (75) 571 ± 154 (82) Phosphorus mg 1103 ± 214 (156) 1118 ± 141 (160) Iron mg 13.4 ± 2.8 (84) 13.3 ± 2.1 (82) Sodium g 4.1 ± 0.9 3.9 ± 1.0 Potassium g 2.7 ± 0.5 2.5 ± 0.5 Vitamin A mgRE 676 ± 141 (97) 614 ± 146 (88) Vitamin B1 mg 1.2 ± 0.2 (118) 1.1 ± 0.2 (112) Vitamin B2 mg 1.3 ± 0.3 (105) 1.2 ± 0.2 (103) Niacin mgNE 15.1 ± 2.4 (117) 14.7 ± 2.1 (114) Vitamin C mg 152 ± 79 (277) 105 ± 36 (192) Cholesterol mg 521 ± 119 490 ± 112 Values are means±SD. S, Iron supplemet group; C, Cotrol group. RE, equivalent retinol weight; NE, equivalent niacin weight. *Percentage of recommended dietary allowance.
Hyung-Sook Kang and Tatsuhiro Matsuo 356
Table 4. Pre- and post-experiment haematological measurements
S N =11
C N = 14
Before After Change Before After Change Haemoglobin g/dl 12.8 ± 1.4 12.2 ± 0.9 -0.6 ± 0.9 13.2 ± 1.3 12.3 ± 1.2* -0.9 ± 0.6
Red blood cells x106/mm3 4.24 ± 0.17 4.19 ± 0.27 0.03 ± 0.25 4.31 ± 0.32 4.20 ± 0.30 -0.04 ± 0.27
Haematocrit % 40.6 ± 3.2 37.8 ± 2.4* -2.8 ± 3.1 41.9 ± 3.3 38.3 ± 3.1* -3.6 ± 1.7
MCV fl 95.8 ± 6.4 90.3 ± 5.6* -5.5 ± 1.8 97.4 ± 2.2 91.1 ± 3.0* -6.3 ± 2.0 MCH pg 30.1 ± 2.8 29.5 ± 2.5* -0.6 ± 0.5 30.6 ± 1.2 29.3 ± 1.4* -1.4 ± 0.7#
MCHC g/l 31.3 ± 1.4 32.3 ± 1.3* 1.0 ± 0.8 31.6 ± 0.8 32.1 ± 0.7* 0.6 ± 0.6
Plasma iron μg/dl 77 ± 34 78 ± 31 1 ± 26 71 ± 18 85 ± 46 14 ± 45 TIBC μg/dl 479 ± 67 388 ± 75* -91 ± 36 456 ± 39 392 ± 50* -64 ± 11
Plasma ferritin μg/l 21.5 ± 28.3 33.3 ± 33.4* 11.2 ± 8.3 16.6 ± 9.4 24.1 ± 15.8 7.6 ± 15.3 Values are means ±SD. S, Iron supplement group; C, Control group. MCV, mean cell volume; MCH, mean cell haemoglobin; MCHC, mean cell haemoglobin concentration; TIBC, total iron binding capacity. P < 0.05 vs. pre-experiment values (repeated measures ANOVA and Student's paired t-test). # P<0.05 vs. S group (Student's t-test).
Table 5. Pre- and post-experiment immunological measurements
S N = 11
C N = 14
Before After Change Before After Change White blood cell x103/mm3 3.88 ± 1.67 4.52 ± 1.00 0.64 ± 0.91 4.10 ± 1.60 4.60 ± 0.22 0.50 ± 0.88 Neutrophil % 51.1 ± 4.8 49.6 ± 10.6 -1.5 ± 4.2 49.6 ± 5.9 54.7 ± 9.8 5.1 ± 4.9 Lymphocyte % 42.4 ± 4.2 34.6 ± 8.2* -7.8 ± 3.1 43.6 ± 7.1 34.6 ± 8.2* -9.0 ± 4.7 Monocyte % 4.0 ± 1.3 7.2 ± 1.8* 3.2 ± 1.8 4.4 ± 1.9 7.1 ± 2.4* 2.7 ± 2.0 Esoinophil % 2.1 ± 1.5 4.0 ± 3.3* 1.9 ± 0.5 2.1 ± 1.9 2.9 ± 1.5 0.8 ± 0.9 Basophil % 0.5 ± 0.5 0.6 ± 0.5 0.1 ± 0.5 0.5 ± 0.5 0.6 ± 0.5 0.1 ± 0.5 IgG mg/dl 1207 ± 171 1258 ± 225 51 ± 26 1211 ± 197 1210 ± 188 -1 ± 45 IgA mg/dl 170 ± 68 172 ± 67 2 ± 30 213 ± 84 214 ± 74 1 ± 51 IgM mg/dl 134 ± 46 156 ± 53 22 ± 39 135 ± 30 146 ± 29 11 ± 22 Values are means±SD. S, Iron supplement group; C, Control group. *P<0.05 vs. pre-experiment values (Repeated measures ANOVA and Student's
sports training is not a true anaemia in that iron is not limiting red blood cell production. An increase in plasma volume is presumed to account for most of the initial drop in Hb, although red blood cell destruction also contributes to the decrease.37,38 Evidence in favor of the latter con-comitant change includes: (1) the degree of change in Hb concentration is greater than that accountable to increased plasma volume31 (2) there is an increase in mean red blood cell size38,39 (3) red blood cell osmotic fragility decreases and (4) serum haptoglobin decreases.38,40 Short-term haematological change induced by sports training is thus an early adaptation to endurance exercise.38
Inadequate dietary iron intake appears to be a major contributing factor to the prevalence of iron deficiency. In this study, iron intakes of the subjects who completed the 4-week experimental period averaged 13.3 mg/day for the C group. It should be noted that no condition except iron deficiency has been reported to produce a low serum ferritin concentration.41 Many female athletes had intakes below the Korean recommended intake of 16 mg/day, which reinforced the common finding that it is difficult for the menstruating female to meet her iron demands when consuming the typical Western diet.42
The difference between the S and C groups was the change in plasma ferritin levels. The S group’s mean level of plasma ferritin rose 54.9%. Although statistically signi-ficant, this rise in plasma ferritin is still modest when one considers that the normal range extends to 160 μg/l, and that the mean level of a large screening (n=1104) of U.S. female nonathletes was 69.6 μg/l.9 Schoene et al.,10 supplemented with a similar dosage (300 mg/day), but for only two weeks, and found an increase in ferritin levels from 10.0 to 22.1 μg/l. In the present study, it was hoped that 4 weeks of supplementation would be sufficient to bring the mean ferritin levels to greater than 60 μg/l. Ferritin levels below 64 μg/l may still indicate an iron-deficient state.43 Heinrich et al.,43 correlated iron absorp-tion with serum ferritin concentration. Diagnostic 59Fe2+ absorption appeared to be a more sensitive indicator of depleted iron stores. It was concluded that serum ferritin values up to 64 μg/l could still be representing prelatent iron deficiency. Exhausted iron stores cannot be definitely excluded as a possibility until serum ferritin concentration rises above this level. Newhouse et al.,35 suggested that supplementation of prelatent/latent iron deficient female athletes should ideally be continued for perhaps 16 weeks to ensure that mean levels reach the 60-70 μg/l range.
paired t-test).
357 Iron supplementation and haematological and immunological status in elite female soccer players
The immune system itself appears to be particularly sensitive to the availability of iron.19 Iron deficiency de-presses various aspects of immune function, including the lymphocyte proliferative response to mitogen stimu-lation,44 macrophage interleukin-1 production,45 and na-tural killer cell cytotoxic activity46; the latter possibly owing to the reduced production of interferon associated with iron deficiency. Phagocyte function is impaired by low iron availability, as evidenced by decreased bacte-ricide, lowered myeloperoxidase activity and a decrease in the oxidative burst.47 In contrast, high concentrations of ferric irons inhibit phagocytosis of human neutrophils in vitro.48 In this study, the change in immunological parameters and antioxidative enzyme activities were nearly the same for both the S and C groups. The percentage of lymphocytes and monocytes, and SOD activity changed during the 4-week experimental period in both the S and C groups. These results may show an adaptation to heavy training. In conclusion, the results demonstrate that 4 weeks of iron supplementation (a daily dose of 40 mg elemental iron) given to elite female soccer players significantly increased body iron stores and inhibited decrease of Hb concentration induced by soccer training, but did not influence immune functions and antioxidative enzyme activities. Iron supplementation would appear to be necessary for elite female athletes, but a more detailed study is required to clarify the effects of iron supple-mentation. References 1. Beaton GH. Epidemiology in iron deficiency. In: Jacobs A,
Worwood M, eds. Iron on Biochemistry and Medicine. London: Academic Press, 1974; 477-490.
2. Siimes .A, Refino C, Dallman PR. Manifestations of iron deficiency at various levels of dietary iron intake. Am J Clin Nutr 1980; 33: 570-574.
3. Edgerton VR, Bryant SL, Gillespie CA, Gardner GW. Iron deficiency anaemia and physical performance and activity of rats. J Nutr 1972; 102: 381-400.
4. Davies CTM, Chukweumeka AC, Van Haaren JPM. Iron deficiency anaemia: its effect on maximum aerobic power and response to exercise in African males aged 17-40 years. Clin Sci 1973; 44: 555-562.
5. Gardner GW, Edgerton VR, Barnard RJ, Bernauer EM. Cardiorespiratory, haematological and physical perfor-mance responses of anemic subjects to iron treatment. Am J Clin Nutr 1975; 28: 982-988.
6. Viteri FE, Torun B. Anaemia and physical work capacity. Clin Haematol 1974; 3: 609-626.
7. Finch CA, Gollnick PD, Hlastala MP, Miller LR, Dillmann E, Mackler B. Lactic acidosis as a result of iron deficiency. J Clin Invest 1979; 64: 129-137.
8. Perkkio MV, Jansson LT, Brooks GA, Refino CJ, Dallman PR. Work performance in iron deficiency of increasing severity. J Appl Physiol 1985; 58: 1477-1480.
9. Davies KJA. Maguire JJ, Brooks GA, Dallman PR, Packer L. Muscle mitochondrial bioenergetics, oxygen supply, and work capacity during dietary iron deficiency and repletion. Am J Physiol 1982; 242: E418-E427.
10. Schoene B, Escourrou P, Robertson HT, Nilson KL, Parsons JR, Smith NJ. Iron repletion decreases maximal exercise lactate concentrations in female athletes with minimal iron deficiency anaemia. J Lab Clin Med 1983; 102: 306-312.
11. Ohira Y, Edgerton VR, Gardner GW, Senewiratne B, Barnard RJ, Simpson DR. Work capacity, heart rate and blood lactate responses to iron treatment. Br J Haematol 1979; 41: 365-372.
12. McLane JA, Fell RD, McKay RH, Winder WW, Brown EBL, Holloszy JO. Physiological and biochemical effects of iron deficiency on rat skeletal muscle. Am J Physiol 1981; 241: C47-C54.
13. Clement DB, Asmundson RC. Nutritional intake and haematological parameters in endurance runners. Physician Sports Med 1982; 10: 37-73.
14. Ehn L, Carmack B, Hoglund S. Iron status in athletes involved in intense physical activity. Med Sci Sports Exerc 1980; 12: 61-64.
15. Wishnitzer R, Vorst E, Berrebi A. Bone marrow iron depression in competitive distance runners. Int J Sports Med 1983; 4: 27-30.
16. Jacobs A, Miller E, Worwood M, Beamish MR, Wardrop CA. Ferritin in the serum of normal subjects and patients with iron deficiency and iron overload. Br J Nutr 1972; 4: 206-208.
17. Bothwell TH, Charlton RW, Cook JD, Finch CA. Iron metabolism in man. Oxford: Blackwell Scientific Publications, 1979; 7-43.
18. Rowland TW, Desiroth MB, Green GM, Kelleher JF. The effect of iron therapy on the exercise capacity of non-anaemic iron-deficient adolescent runners. Am J Dis Child 1988; 142: 165-169.
19. Galan P, Thibault H, Preziosi P, Hercberg S. Interluikin 2 production in iron deficient children. Biol Trace Elem Res 1992; 32: 421-425.
20. Chandra RK. Immunocompetence in undernutriton. J Pediatr 1972; 81: 1194-1200.
21. Chandra RK. Nutrition as a critical determinant in susceptibility to infection. World Rev. Nutr Dietet 1976; 25: 166-188.
22. Shephard RJ, Shek PN. Immunological hazards from nutritional imbalance in athletes. Exer Immunol Rev 1998; 4: 22-48.
Table 6. Pre- and post-experiment antioxidative enzyme activities
S N =11 C
N =14 Before After Before After GPx U/g Hb 0.90 ± 0.59 0.70 ± 0.68 0.73 ± 0.38 0.84 ± 0.47 SOD U/g Hb 5854 ± 1666 10020 ± 2431* 6756 ± 1600 9736 ± 2534*
Catalase U/g Hb 768 ± 850 566 ± 939 668 ± 710 534 ± 428 Values are means±SD. S, Iron supplement group; C, Control group. GPx, glutathione peroxidase; SOD, superoxide dismutase; Hb, haemoglobin.
*P<0.05 vs. pre-experiment values (Repeated measures ANOVA and Student's paired t-test).
Hyung-Sook Kang and Tatsuhiro Matsuo 358
23. Shephard RJ, Shek PN. Heavy exercise, nutrition and immune function: Is there a connection? Int J Sports Med 1995; 16: 491-497.
24. Bishop NC, Blannin AK, Walsh NP, Robson PJ, Gleeson M. Nutritional aspects of immunosuppression in athletes. Sports Med 1999; 28: 151-156.
25. Bisdee JT, James WPT, Shaw MA. Changes in energy expenditure during the menstrual cycle. J Nutr 1989; 61: 287-199.
26. International Nutritional Anaemia Consultative Groug. Measurements of Iron Status. Washington, DC: The Nutrition Foundation Inc., 1985; 4-54.
27. Ortoft G, Gronbaek H, Oxlund H. Growth hormone admin-istration can improve growth in glucocorticoid-injected rats without affecting the lymphocytopenic effect of the gluco-corticoid. Growth Horm IGF Res 1998; 8: 251-264.
28. Sternberg JC. A rate nephelometer for measuring specific proteins by immunoprecipitate reactions. Clin Chem 1977; 23: 1456-1460.
29. Winterbourn CC, Hawkins RE, Brian M, Carrell RW. The estimation of red cell superoxide dismutase activity. J Lab Clin Med 1975; 35: 337-341.
30. Paglia DE, Valentine WN. Studies on the quantitative and quantitative characterization of erythrocyte glutathione peroxidase. J Lab Clin Med 1967; 70: 158-160.
31. Johansson LH, Hakan LA. A spectrophotometric method for determination of catalase activity in small tissue sample. Anal Biochem 1988; 174: 331-336.
32. Nagamine S. Valuation of body fatness by skinfold, In: Asahina K, Shigiya R, eds. Physiological, adaptability and nutrition status of the Japanese: growth, work capacity and nutrition of Japanese. Tokyo: University of Tokyo Press, 1975; 16-20.
33. Brozek J, Grande F, Anderson JT. Densitometric analysis of body composition: revision of some quantitative assum-ptions. Ann NY Acad Sci 1963; 110: 113-140.
34. Pate RB, Maguire M, Wyk JV. Dietary iron supple-mentation in women athletes. Physician Sports Med 1979; 7: 81-88.
35. Newhous IJ, Clement DB, Taunton JE, McKenzie DC. The effects of prelatent/latent iron deficiency on physical work capacity. Med Sci Sports Exer 1989; 21: 263-268.
36. Newhouse IJ, Clement DB. Iron status in athletes an update. Sports Med 10988; 5: 337-352.
37. Frederickson LA, Puhl JL, Runyan WS. Effects of training on indices of iron status of young female cross-country runners. Med Sci Sports Exer 1983; 15: 271-276.
38. Eichner ER. Runners macrocytosis: a clue to footstrike hemolysis. Am J Med 1985; 78: 321-325.
39. Davidson RJL. Exertional haemoglobinuria: a report on three cases with studies on the haemolytic mechanism. J Clin Pathol 1969; 17: 536-540.
40. Williamson MR. Anaemia in runners and other athletes. Physician Sportsmed 1981; 9: 73-79.
41. Valberg LB. Plasma ferritin concentration: their clinical significance and relevance to patient care. Can Med Assoc J 1980; 122: 1240-1248.
42. Clement DB, Sawchuk LL. Iron status and sports perfor-mance. Sports Med 1984; 1: 65-74.
43. Heinrich HC, Bruggemann J, Gabbe EE, Glaser M. Correlation between diagnostic 59Fe2+ absorption and serum ferritin concentration in man. Z Naturforsch 1977; 32: 1023-1025.
44. Chandra RK. Nutrition and immunity: lessons from the past and new insights into the future. Am J Clin Nutr 1991; 53: 1087-1101.
45. Helyar L, Sherman AR. Iron dificiency and interleukin 1 production by rat leukocytes. Am J Clin Nutr 1987; 46: 346-352.
46. Sherman AR. Zinc, copper and iron nutriture and immunity. J Nutr 1992; 122: 604-609.
47. Dallman PR. Iron deficiency and the immune response. Am J Clin Nutr 1987; 46: 329-334.
48. Van Asbeck S, Marx JJM, Struyvenberg A. Effect of iron (III) in the presence of various ligands on the phagocytic and metabolic activity of human polymorphonuclear leukocytes. J Immunol 1994; 132: 851-856.
359 Asia Pac J Clin Nutr 2004;13 (4):359-365
Original Article High prevalence of low dietary calcium and low vitamin D status in healthy south Indians CV Harinarayan MD, DM, T Ramalakshmi MSc and U Venkataprasad MSc
Department of Endocrinology & Metabolism, Sri Venkateswara Institute of Medical Sciences, Tirupati, India
Calcium and vitamin D under nutrition can adversely affect the bone mineral metabolism. There is no population-based study from India documenting dietary habits, serum calcium and vitamin D levels. Our study investigated the dietary habits of rural and urban societies in and around Tirupati and their relationship with serum calcium, phosphorous and vitamin D [25(OH)D] levels. Four hundred and seven subjects from 5 villages around Tirupati, (rural population) and 125 asymptomatic staff of our hospital (urban population) were studied. Dietary intakes of calcium, phosphorous and phytates were documented by diet history. Serum calcium, phosphorus and 25 (OH) D levels were estimated in 191 rural subjects and 125 urban subjects. Compared to urban subjects, rural subjects had a significantly lower intake of dietary calcium (P <0.0001) and a significantly higher dietary phytate/calcium ratio and serum calcium and 25 (OH) D levels (P <0.0001). Dietary calcium intake was inadequate in both rural and urban subjects compared to the recommended daily allowances (RDA) for our country. About 31% of the population had normal vitamin D levels, 54% had vitamin D insufficiency and 15% vitamin D deficiency. About two-thirds of the population had low levels of vitamin D. Inadequate dietary calcium intake associated with high phytate/calcium ratio reduces the bioavailable calcium in the gut. Hence, there is a need to fortify food with calcium and to propose new guidelines for 25 (OH) D in Indian subjects. Multicentric studies with large sample populations are required to generate normal standards and nationally relevant guidelines.
Key Words: diet, calcium, serum 25 (OH) D, vitamin D, fortification of foods, RDA, ICMR, rural, urban, South India Introduction Nutritional factors play a vital role in bone homeostasis during adulthood. During infancy, childhood and ado-lescence, increasing dietary calcium intake favours bone mineral accrual. Adequate calcium intake along with vitamin D helps to maintain bone mineral mass attained at the end of growth period i.e. peak bone mass. Serum 25-hydroxyvitamin D [25(OH)D] is the most reliable indicator of vitamin D levels of an individual. Vitamin D in-sufficiency [25(OH)D levels between 10 – 20 ng/ml] is associated with secondary hyperparathyroidism (SHPT). Low dietary calcium intake further amplifies the para-thyroid response to vitamin D insufficiency. The SHPT, which ensues, mobilizes mineral and matrix from skeleton leads to a high risk of fracture.1-5 Vitamin D deficiency and/or poor dietary calcium intake can lead to a defect in mineralization of bone (Rickets in children; Osteomalacia in adults). Rickets and osteomalacia are known to develop in immigrant Indians who migrate away from the equator.6-9 This was attributed to the poor cutaneous synthesis of vitamin D due to pig-mentation and inadequate sunlight exposure along with an inadequate dietary calcium intake. Vitamin D deficiency [25 (OH) D levels <10 ng/ml] was presumed to be rare in tropical countries like India. Previously, we reported the prevalence of low vitamin D levels in India in a group of normal subjects and patients with primary hyperpara-thyroidism.10 Subsequently other reports ensued.11-14 It is
indeed surprising to find low vitamin D in healthy subjects in India, a country with abundant sunshine. So far, there is no large study documenting the dietary habits, serum calcium and vitamin D levels of an Indian population. We studied the dietary habits, and its relationship with serum calcium, and vitamin D [25-hydroxy cholecalciferol 25 (OH) D] in patients residing in Tirupati and surrounding villages.
Materials and methods Between January 2000 and July 2003, 407 apparently healthy, asymptomatic subjects from Sathyavedu [SY] (Latitude [Lat] 13.260N and Longitude [Long] 79.570E), Peddathippasamudram [PTM] (Lat.13.430N,Long. 78.13 0E), Sandramakula palli [S. Palli] (Lat.13.400N, Long. 78.140E), Adharam [A] (Lat.13.370N, Long. 79.470 E) and Kandluru [K] (Lat.13.360N, Long. 79.470E) villages around Tirupati, belonging to Chitoor district, Andhra Pradesh were included in the study. They constituted the Tirupati rural population. Students and staff of the Sri Venka-teswara Institute of Medical Sciences (SVIMS), Tirupati Correspondence address: Dr CV Harinarayan, Additional Professor & Head, Department of Endocrinology & Metabolism, Sri Venkateswara Institute of Medical Sciences, Tirupati – 517 507, AP, India. Tel: 91-877-2287152; Fax: 91 -877-2286803 Email: [email protected] Accepted 28 May 2004
CV Harinarayan, T Ramalakshmi and U Venkataprasad 360
fpsssv ttspsdrrsmctn
–
catalogue No. 68100E). The minimal detectable limit of 25 (OH) D assay is 1.5 ng/ml [reference range 9 to 37.6 ng/ml]. The kit manufacturer to monitor assay perfor-mance provided two quality control sera (control A and B). The control A (range: 11.6 to 24.4 ng/ml) and the control B (range: 34.7 to 73.5 ng/ml) values for 25 (OH)D assay were 17.8 ng/ml and 57.3 ng/ml respectively. The intra- assay (at 12.75 ng/ml) and inter-assay (at 11.0 ng/ml) variations for 25 (OH) D were 0.9% and 3.95% respectively.
Statistical methods A statistical analysis was performed using SPSS package (version10). Descriptive results are presented as mean + standard error of mean (SEM). One-way analysis of variance (ANOVA) was used to estimate the differences
Parameter Tpt - Rural Tpt – Urban
N 191 125
Age (Yrs) 44 + 1.03 45.5 + 0.95
S. Calcium (mg/dl) 10 + 0.05* 9.71 + 0.06
S. Phosphorous (mg/dl) 3 + 0.04* 3.28 + 0.53
S. 25(OH)D (ng/ml) 21 + 0.46* 13.52 + 0.59
Dietary Calcium (mg/day) 264 + 1.94* 356 + 5.0
Dietary Phosphorous (mg/day) 490 + 4.98* 721 + 10.2
Dietary Phytates (mg/day) 200 + 1.9 207 + 4.7
Phytates/Calcium Ratio 1 + 0.01* 0.58 + 0.01
Table 1. Dietary Pattern, serum calcium and 25 (OH) D statuses of Tirupati Urban and Rural population
between the study groups. If a significant difference was found, a multiple comparison test was performed using
Values represent mean+ SEM; *P <0.0001; 25(OH)D – forconversion from ng/ml to nmol/l – multiply by 2.5[TPT] (Lat.13.40 0 N, Long. 77.2 0 E), Chitoor district, Andhra Pradesh and their asymptomatic relatives (N=125) were studied. They constituted Tirupati urban population. In all the above locations, the average duration of sunlight is around 8 to10 hours per day throughout the year. Winters are short with minimum and maximum temperatures ranging from 170C to 300C with poor rain-all. Most often, there is little seasonal variation of the eak sunlight. The visual skin complexion of the subjects tudied is wheatish to dark in color. Most of the rural ubjects are agricultural workers who are exposed to unlight for a period of 8 to10 hours a day. Cloths or eils did not restrict the exposure to sunlight. In all the villages, a prior visit was undertaken to study he pattern of living and dietary. The subjects were asked o remain fasting on the day of collection of blood ample. The dietary intake of calcium, phosphorous and hytates were documented by recalling the diet con-umed in the previous 5 to 7 days. The documentation of ietary pattern was by a single observer. The validity and epeatability of the documentation was rechecked at andom by one of us (authors) over the period of the tudy. There was no significant error in the docu-entation of dietary history. From the raw weights, the
alcium and phosphorous intakes were calculated using he published food composition table detailing the utritive value of Indian foods.15
For all patients fasting venous blood samples were collected from the most accessible peripheral vein between 0800 to 0900 hours in the fasting state without applying tourniquet for the estimation of serum calcium, phosphorus and on ice for 25(OH)D. The serum was separated in refrigerated centrifuge at 40C and stored at
200C until the analysis for the estimation of 25(OH)D. The blood samples collected from village populations were transported under cool packs until they were separated and stored for further analysis. The 25 (OH)D levels were estimated in 191 rural subjects and 125 urban subjects. The serum calcium and phosphorus levels were estimated by titrimetric method16 and by Fiske Subba Row method17 respectively. The 25 (OH)D concentrations were measured by competitive radioimmunoassay after acetonitril extraction (DiaSorin, Stillwater, MN, USA,
LSD post hoc test to analyze the differences between the study groups. Probability value of P<0.05 was considered significant.
Results The diet in rural subjects consisted of 1700 KJ/day approximately. Carbohydrates provided 75% of the total energy intake, proteins 10%, fat 5%, vegetables 5%, and milk and milk products 5%. The carbohydrate source was from cereals [Rice – 60% and Ragi (Eleusine Coracana) – 40%]. Vegetable sources included drumstick leaves, brinjals and tomatoes. Animal sources of protein were consumed once fortnightly. The diet in urban subjects consisted of 2200 KJ/day approximately. Carbohydrates provided 55% of the total energy intake, proteins 10% and fat 10%. Vegetables contributed 10% to total energy intake and milk and milk products contributed 15% (Fig. 1). The carbohydrate source was primarily from cereals with rice providing 50% of total carbohydrates, wheat 25% and ragi 25%. Vegetable sources included amaranth leaves, cauliflower, carrots, ladies fingers, other seasonal vegetables and tubers. Animal sources of protein were consumed once a week. There was no other source of calcium or any other mineral in both groups. Milk is not fortified with calcium or vitamin D in India. The mean + SEM of dietary calcium, phosphorous, phytates and dietary phytate/calcium ratio, serum cal-cium, phosphorous and 25 (OH) D levels of the Tirupati rural and urban population is described in Table 1. The age groups of Tirupati rural and urban populations were comparably similar. The daily dietary calcium intake by both Tirupati rural and urban populations were low (mean + SEM: rural 264 + 1.94; urban 354 + 5 mg/day) compared to that of Recommended Daily/Dietary Allow-ance (RDA) issued by the Indian Council of Medical Research (ICMR) for the Indian population. Dietary calcium, phosphorous and serum phosphorous were significantly lower (P<0.0001) in the rural subjects compared to the urban subjects. Though the dietary phytates were comparable in both the rural and urban groups, the dietary phytate/calcium ratio was significantly (P <0.0001) higher in rural subjects (Table 1). The serum calcium and 25 (OH) D levels were significantly higher P <0.0001) in the rural subjects compared to the urban subjects.
361 High prevalence of low dietary calcium and low vitamin D status in healthy south Indians
Parameter 25(OH)D < 10 ng/ml (Group – 1)
Vitamin D deficiency
25(OH)D 10 – 20 ng/ml (Group – 2)
Vitamin D insufficiency
25(OH)D > 20 ng/ml (Group – 3)
Normal vitamin D levels
Rural Urban Whole Group
Rural Urban Whole Group
Rural Urban Whole Group
N (%) 5 44 49 (15%) 107 63 170 (54%) 79 18 97 (31%) Age (yrs) 40+7.4 46+1.8 46 +1.7 42+1.4 44+1.28 43+1.0 47+1.5 48+2 47+1.3 S.Ca mg/dl
10+0.25* 9.76+0.08 9.79+0.07 10+0.07* 9.61+0.1 9.87+0.06 10+0.07* 9.9+0.18 10+0.07
25 (OH) D ng/ml
9.04+0.7@ 7.36+0.3 7.53+0.28 16.7+ 0.26@ 14.3+0..34 15.83+0.23 27+0.5@ 25.8+0.95 26.92+0.44
Dietary Ca mg/day
227+12* 355+9.1 342+10 264+3* 360+7.4 299+5 266+2.15* 344+10 281+4.15
Dietary Phos mg/day
416+21 730+16 699+20 481+7 724+16 481+7 506+7 694+23 544+10
Phyt./ Calc. Ratio
0.87+0.03* 0.6+0.01 0.63+0.01 0.79+0.01* 0.58+ 0.01 0.71+ 0.01 0.72+0.01* 0.54+0.02 0.68+0.01
Table 2. Dietary pattern, serum calcium and 25(OH)D status of Tirupati urban and rural population – categories based on 25 (OH) D levels
alues represent mean + SEM; *P=<0.0001 compared to urban group; @ P <0.001 compared to urban group
V The 25(OH)D levels of the sample (rural and urban groups) were classified into: group 1 vitamin D defi-ciency [25(OH)D levels <10ng/ml]; group 2 vitamin D insufficiency or marginal intake [25(OH)D levels 10-20 ng/ml] and group 3 normal vitamin D [25(OH)D levels >20ng/ ml].18 Based on this classification only 31% (N=97) of the sample population had normal vitamin D levels. About 50% (N =170) had vitamin D insufficiency and 15% (N=49) had vitamin D deficiency (Table 2). Severe vitamin D deficiency (25(OH)D levels <5ng/ml) was found in three subjects (1% of the whole population).The 25(OH)D levels ranged from zero to 4.05 ng/ml in the severe vitamin D deficiency group. They did not have any other secondary cause attributable to vitamin D deficiencies. All of them were urban subjects. The village and urban subjects were sub-classified based on 25 (OH) D levels into three groups (Table 2). The rural population had significantly (P <0.001) higher 25(OH)D levels compared to the urban group in all the three sub-categories. One way ANOVA amongst the three groups (between rural and urban subjects) revealed significantly (P <0.0001) lower dietary calcium, higher
0
10
20
30
40
50
60
70
80
Carbohydrate % Protein % Fat % Vegetables % Milk & Products % Rural Urban
Figure 1. Dietary pattern (percentage of total energy intake) of rural and urban subjects
Percentage of total energy intake
CV Harinarayan, T Ramalakshmi and U Venkataprasad 362
phytate/calcium ratio, and higher serum calcium in the vitamin D deficiency group compared to the vitamin D insufficient group and the group with normal vitamin D levels in rural subjects (Table 2). Discussion The dietary intake of calcium in first generation normal Asian Indian immigrants in USA19 was found to be less than two-thirds of the dietary reference intake recommen-ded for a normal person as per the guidelines of the USA. Recently the RDA has been revised and redefined as the Dietary Reference Intake (DRI), which is a collaborative effort between USA and Canada.20 The RDA for calcium in India recommended by the Indian Council of Medical Research (ICMR) is lower than the recently revised recommendations by the USA and Canada (Table 3).21–23 There is neither a recommendation for dietary intake of vitamin D nor a monitored food fortification program for the intake of calcium or vitamin D by ICMR. The dietary intake in the urban group was high in calories, milk, milk products and vegetables. The major cereal consumed was rice, rather than ragi and wheat, which has lower phytate levels. Even the carbohydrate portion was occasionally replaced by sweets containing milk and its products. The dietary calcium intake by the Tirupati rural population is less than that of the urban population. Intake of Ragi (rich in phytates) by the rural population retards the absorption of calcium from the gut. The daily consumption of milk and milk products was only 5% of their total energy intake. The other source of calcium was from leafy vegetables (especially drumstick leaves). There is no other source of vitamin D in the diets of the sample population. Nevertheless, the dietary calcium intakes by both the rural and the urban samples were much lower than the RDA for calcium as per the ICMR guidelines (Table 3). These data highlight the high prevalence of inadequate dietary calcium intake across the population compared to the RDA. To the best of our knowledge, there are no population-based studies from India comparing rural and urban populations with their dietary habits and 25 (OH) D levels. There are reports of very low dietary intakes of calcium (<300 mg/day) in patients with osteomalacia.24,25 Besides this, it has been shown in the studies by Panwar et al.,26 that the calculated values for all nutrients are significantly higher than the analytical values. Hence, a patient with a calculated low intake of calcium with a background diet containing foods high in phytates, as in our study, may be more calcium deficient than calculated from dietary intake data. The inadequate dietary calcium intake is significant when viewed in the background of high phytate/calcium ratio associated with low 25 (OH)D levels. Phytate in the diet retards the absorption of calcium in the gut. Though the 25(OH)D levels were high in rural subjects in all the three groups, the dietary calcium intake was inadequate with high phytate/calcium ratio compared to the urban subjects (Table 2). The high phytate/calcium ratio in the rural subjects retards calcium absorption. In the present study, all the subjects had adequate sunlight exposure and the dress code did not affect the exposure to sunlight. About two-thirds (69%) of the population
Category India14,15 USA16 Units mg/day mg/day Infants
Infants 0–6 months 500 500 Infants 6– 12 months 500 750
Children Boys & Girls 1 – 9 yrs 400 800 10 – 15 yrs 500 1200 -1300 16 – 18 yrs 500 1200 -1300
Men 400 800 -1000 Women 400 800 - 1000 Pregnant & Lactating 1000 1200 - 1300
Table 3. Recommended Dietary Allowances of calcium in India and USA
have low levels of vitamin D. About 15% of the popu-lation had vitamin D deficiency and 54% had vitamin D insufficiency. In our study, all patients with severe 25 (OH)D deficiency (<5ng/ml) were from the urban sample. The significantly higher levels of 25 (OH) D in the rural population compared to the urban population can be partly explained by the former group having greater ex-posure to sunlight as a result of their agricultural occu-pation. Vitamin D insufficiency is associated with secondary hyperparathyroidism (SHPT), which is further amplified by inadequate calcium intake. Thus, in the background of low vitamin D levels and inadequate dietary calcium intakes, when an individual is exposed to the additional insult of an environmental toxin like fluoride, the clinical expression of the disease is altered. Various studies have shown that the effect of an environmental toxin like fluoride on bone mineral metabolism is severe and more complex in children with poor dietary calcium intake when compared to the children with adequate dietary calcium intake.27–29 It has also been shown that calcium absorptive performance of the gut is a function of 25(OH)D status of an individual.30,31 When there are low 25 (OH) D concen-trations, the effective calcium absorption from the gut is reduced.31,32 This is further amplified by the low dietary calcium intake. The SHPT consequent to inade-quate dietary calcium intake and low 25(OH)D concentrations mobilizes mineral and matrix from the skeleton. This increases he risk of fractures, especially in post-menopausal women and elderly patients. These are further amplified by age related changes with calcium supplementation.33 High phytate/calcium ratio amplifies the inadequate dietary calcium intake. There are several studies documenting that vitamin D and calcium supplements have synergistic effects in preventing proximal femoral fractures in postmenopausal and older patients.34-36 The shortcomings of the inade-quate dietary calcium intake associated with the reduced bioavailable calcium in the gut due to phytates and age related calcium conservation in the gut can be overcome by up-revising the RDA for calcium, restricting phytates in food and recommending new guidelines for vitamin D. Monitored food fortification programs are to be imple-mented at national level so that the existing diets are
mothers
363 High prevalence of low dietary calcium and low vitamin D status in healthy south Indians
supplemented with food that are rich (or have been enriched) with calcium. In summary, two-thirds of the study population (rural and urban groups) had low vitamin D levels. There is no RDA proposed by ICMR for vitamin D for the Indian population. The revised DRI of the USA and Canada20 recommends 400 IU for vitamin D for those aged under 50 and 800IU for those aged 50 plus. The dietary calcium intake by the urban and rural populations did not meet the existing RDA by ICMR. The present study has certain methodological limi-tations. The urban sample was a sample of convenience because of logistic and operational reasons. However, even after taking these limitations into account, our obser-vations argue strongly for the revision of the RDA for calcium and new recommendations for the 25(OH)D for the Indian population. Multicentric studies with a large sample size are required to generate normal standards for the purpose of nationally relevant guidelines.
Acknowledgements (contributions of each author). There is no conflict of interest by any of the authors.
References 1. Parfitt AM, Gallagher JC, Heaney RP, Johnston CC, Neer
R, Whedon GD. Vitamin D, and bone health in the elderly. Am J Clin Nutr 1982; 35: 1014-31.
2. Villareal DT, Civitelli R, Chines A, Avioli LV. Sub-clinical vitamin D deficiency in postmenopausal women with low vertebral bone mass. J Clin Endocrinol Metab 1991; 72: 628-34.
3. Holick MF. Vitamin D: the underappreciated D-lightful hormone that is important for skeletal and cellular health. Curr Opin Endocrinol Diabetes 2000; 9: 87-98.
4. Lips P. Vitamin D deficiency and secondary hyperpara-thyroidism in the Elderly: Consequences for bone loss and fractures and Therapeutic implications. Endocrine Reviews 2001; 22: 477-501.
5. Harris SS, Soteriades E, Stina Coolidge JA, Mudgal S and Hughes BD. Vitamin D insufficiency and hyper-parathyroidism in a low income, multiracial, elderly population. J Clin Endocrinol Metab 2001; 85: 4125-30.
6. Preece MA, Ford JA, MacIntosh WB, Dunnigan MG, Tomlinson S, O’Riordan JLH. Vitamin D deficiency among Asian immigrants to Britain. Lancet 1973;1:907–10.
7. Dent CE, Rowe DJF, Round JM, Stamp TCR. Effect of chapattis and ultraviolet irradiation on nutritional rickets in Indian immigrants. Lancet 1973; 1: 1282–84.
8. Awumey EMK, Mitra DA, Hollis BW, Rajiv kumar and Bell NH. Vitamin D metabolism is altered in Asian Indians in the southern United Status: A clinical research center study. J Clin Endocrinol Met 1998;83:169–73.
9. Gibson RS, Bindra GS, Nizan P, Draper HH. The vitamin D status of East Indian Punjabi immigrants to Canada. Br Jr Nutr 1987; 58 (1): 23-9.
10. Harinarayan CV, Gupta N, Kochupillai N. Vitamin D status in primary hyperparathyroidism in India. Clin Endocrinol (Oxf). 1995; 43: 351-8.
11. Goswami R, Gupta N, Goswami D, Marwaha RK, Tandon N, Kochupillai N. Prevalence and significance of low 25-hydroxy D concentrations in healthy subjects in Delhi. Am J Clin Nutr 2000; 72: 472-5.
12. Agarwal KS, Mughal MZ, Upadhyay P, Berry JL, Marwer EB, Puliyel JM. The impact of atmospheric pollution on vitamin D status of infants and toddlers in Delhi, India. Arch Dis Child 2002; 87 (2):111-3.
13. Arya V, Bhambari R, Godbole MM, Mithal A. Vitamin D status and its relationship with bone mineral density in healthy Asian Indians. Osteoporosis Int 2004; 1: 56-61.
14. Tandon N, Marwaha RK, Kalra S, Gupta N, Dudha A, Kochupillai N. Bone mineral parameters in healthy young Indian adults with optimal vitamin D availability. Nat Med J India 2003; 16: 298-302.
15. Gopalan C, Ramasastri BB, Balasubramanyam SC. Nutritive value of Indian Food. Indian Council of Medical research, National Institute of Nutrition, Hyderabad 1998.
16. Raghuramulu N, Madhavan Nair K and Kalyanasundaram S. A Manual of Laboratory Techniques. Indian council of Medical Research, National Institute of Nutrition, Hyderabad 1983;136-7.
17. Raghuramulu N, Madhavan Nair K and Kalyanasundaram S. A Manual of Laboratory Techniques, Indian Council of Medical Research, National Institute of Nutrition, Hyderabad 1983; 140-1.
18. Lips P. Vitamin D deficiency and secondary hyper-parathyroidism in the Elderly: Consequences for bone loss and fractures and Therapeutic implications. Endocrine Reviews 2001; 22: 477-501.
19. Jonnalagadda SS, Diwan S. Nutrient intake of first generation Gujarati Asian Indian immigrants in the U.S. J Am Coll Nutr 2002; 21: 372-80.
20. RDA – Recommended Dietary Allowance of nutritional elements. Available at URL: http://www.anyvitamins.com/ rda.htm.
21. Food composition table. In: Gopalan C, Sastri BVR, Balasubramanyam SC, eds. Nutritive value of Indian foods. Hyderabad, India: National Institute of Nutrition ICMR 1996; Appendix 1: 92– 4.
22. Swaminathan M. Recommended Dietary Intake of Nutrients. Indian Council of Medical Research, 1981.
23. Recommended Dietary Allowances. In: Essentials of Food & Nutrition. Fundamental aspects. Place of publication: Bappco Pub 1997; 1:508–21.
24. Standing Committee on Scientific evaluation of Dietary Reference Intakes. Calcium. In: Dietary Reference Intakes for calcium, Food and Nutrition Board, National Academy Press 1999; 71–145.
25. Rajeswari J, Balasubramanian K, Bhatia V, Sharma VP, Agarwal AK. Aetiology and clinical profile of osteomalacia in adolescent girls in northern India. Natl Med J India 2003; 16: 139-42.
26. Mathew JT, Seshadri MS, Thomas K, Krishnaswami H, Cherian AM. Osteomalacia - Fifty-five patients seen in a teaching institution over a 4-year period. J Assoc Physicians India 1994; 42: 692-4.
27. Panwar B, Punia D Analysis of composite diets of rural pregnant women and comparison with calculated values. Nutr Health 2000; 14: 217-23.
28. Krishnaramachari KAVR. Skeletal fluorosis in humans. A review of recent progress in understanding of the disease. Prog Food Nutr Sci 1986; 10: 279-314.
29. Teotia M, Teotia SP, Singh KP. Endemic chronic fluoride toxicity and dietary calcium deficiency interaction syn-dromes of metabolic bone disease and deformities in India: year 2000. Indian J Pediatr 1998; 65: 371-81.
30. Mithal A, Trivedi N, Gupta SK, Kumar S, Gupta RK. Radiological spectrum of endemic fluorosis: relationship with calcium intake. Skeletal Radiol 1993; 22: 257-61.
31. Bischoff HA, Stahelin HB, Dick W, Akos R, Knecht M, Salis C, Nebiker M, Theiler R, Pfeifer M, Begerow B, Lew RA Conzelmann M. Effects of vitamin D and calcium supplementation on falls: A randomized controlled Trial. J Bone Min Res 2003; 18: 343 –51.
CV Harinarayan, T Ramalakshmi and U Venkataprasad 364
32. Heaney RP. Vitamin D depletion and effective calcium absorption. A letter to the editor. J Bone Min Res 2003; 18: 1342.
33. Heaney RP, Dowell MS, Hale CA, Bendich A. Calcium absorption varies within the reference range for serum 25-hydroxyvitamin D. J Am Coll Nutr 2003; 22: 142-6.
34. Food and Nutrition Board, Institute of Medicine. Dietary reference intakes of calcium, magnesium, phosphorous, vitamin D and fluorides. Washington DC: National Academy Press, 1997.
35. Szulc P, Meunier PJ. Synergistic effect of vitamin D and calcium in preventing proximal femoral fractures in older patients. Joint Bone Spine 2003; 70: 157-60.
36. Brazier M, Kamel S, Maamer M, Agbomson F, Elesper I, Garabedian M, Desmet G, Sebert JL The markers of bone remodeling in the elderly subjects: effects of vitamin D insufficiency and its correction. J Bone Miner Res 1995; 10: 1753-61.
37. Sahota O, Gaynor K, Harwood RH, Hosking DJ. Hypovitaminosis D and 'functional hypoparathyroidism'-the NoNoF (Nottingham Neck of Femur) study. Age Ageing 2001; 30: 467-72.
365 Asia Pac J Clin Nutr 2004;13 (4):365-371
Original Article Health characteristics of older Australian dietary supplement users compared to non-supplement users Sonya Brownie ND, PGDipSci1 and Margaret Rolfe BSc, MStat2 1School of Natural and Complementary Medicine, Southern Cross University, NSW, Australia 2Australian Centre for Complementary Medicine Education and Research, a joint venture of the University of Queensland and Southern Cross University, Australia.
The aim of this study was to measure the prevalence of dietary and health supplement use among Australians aged 65 years and over, and to contrast the health differences between supplement users and non-supplement users. Data was obtained from 1,263 randomly selected older Australians, who provided general demographic data, in addition to information related to their health, symptoms experienced and uses of medication, including dietary supplements. Supplement use was reported by 43% of the sample (52% of females and 35% of males). This investigation has revealed distinct differences in the health profile of older supplement users compared to non-users. Although there was no difference in the number of visits to medical doctors or self-rated health status between supplement users and non-supplement users, supplement users were more likely to report arthritis and osteoporosis, and experience more symptoms and consume more medication than non-supplement users. In contrast, there was a reduced likelihood of taking a supplement for those with hypertension and by those using blood pressure medication and heart tablets. These results suggest that older dietary supplement users may benefit from education and professional advice to assist them make appropriate and informed choices, particularly if they expect these preparations to attenuate their health concerns.
Key Words: ageing, dietary supplementation, elderly, Australians Introduction It is well established that among the general adult popu-lation supplement users can be characterised on the basis of several demographic and lifestyle features, including gen-der, ethnicity, level of education, income, self-rated health status, weight status, nutrient intake, cigarette and alcohol consumption.1-13 Compared to non-supplement users, supplement users also appear to hold contrasting beliefs about the nutritional adequacy of the food supply; the extent of chemical conta-mination of foods; the role of supplemental intakes of nutrients (e.g vitamin and mineral preparations) and their perception of their nutritional status.14,15 Worsley14,15 found that in a sample of females (mean age 42 years) living in Adelaide, supplement users ex-pressed a higher level of concern about the quality and pu-rity of the food supply than non-supplement users. Sup-plement users also reported higher levels of medication usage, alcohol intake, and work-related stress and were more inclined to practice some form of relaxation activity, such as meditation. In comparison to what is known about the demographic and attitudinal profile of general adult Australian supplement users, the patterns and prevalence of supplement use by older individuals in this country is not as well understood. Since the 1980s, there have been several attempts in the United States to assess rates of supplement usage by individuals aged 65 years and over, and to profile the
characteristics of supplement users. In older individuals supplement use tends to be a marker for a number of positive health related behaviours. These individuals report higher levels of education and healthier lifestyle practices than older non-supplement users.11,16-18 Gender is the most predictive determinant of dietary supplement use across all age groups, including older individuals.1,6-9,16,17,19-26 The use of dietary supplements in this group is commonplace. Depending on the population studied and the method of data collection, rates of usage for individuals aged 65 years and over range from 33% to 46%.1, 6, 8, 9, 11, 16, 19-22 There is a lack of information about the health profile of older supplement users and what data does exist is conflicting. Some studies have found that although there was no significant relationship between supplementation and self-rated health status16,22 supplement use was higher among those with poor self-rated health status.22 In contrast, data from a large sample (n=3,939) of Americans aged 65 years and over suggested that supplement use was significantly less likely for those with poor self-rated health status.18 Correspondence address: Sonya Brownie, School of Natural and Complementary Medicine PO Box 157 Southern Cross University Lismore, NSW Tel: + 02 6626 9131 ; Fax: +02 6626 9135 Email: [email protected] Accepted 23 July 2004
S Brownie and M Rolfe 366
One study found that the relationship was gender specific; according to Wallstrom et al., (1996) female supplement users had worse perceived health than female non-users, while no such association was observed for males.27 Compared to the United States, surprisingly little is known about the dietary supplement practices of older Australians. The most reliable estimates of supplement use by older Australians are based on a small number of national and regional investigations, namely the National Health Survey28, the National Nutrition Survey29, the Blue Mountain Eye Study12 and an earlier investigation con-ducted by Horwath and Worsley (1989).30 According to these studies, 20%-40% of Australians aged 60 years and over report the use of some form of dietary supplement. Supplement use was positively associated with gender (female),12,28,29 higher levels of education,12,30 socio-economic status12,30 and weight status.12 Only one study, a regional investigation of South Australian residents (N = 2,195), has specifically explored the relationship between health status and dietary supple-ment use among individuals aged 65 years and over.30 In this study, dietary supplement users and non-users did not suffer from different medical conditions, nor did they vary significantly in the number of bouts of illness or visits to their general medical practitioner. The current investigation was designed to address the paucity of data on this topic in Australia. The aims of the study were threefold; firstly, to quantify the extent of dietary supplement use in a representative sample of older Australians; secondly, to examine the association between supplement use and self-rated health status and lastly, to determine if older supplement users differ from older non supplement users in regard to the total number and types of health conditions reported, symptoms experienced and medication used. Methods Survey implementation In January 2001 a letter outlining the nature of the investigation and inviting participation was posted to a proportionally random selection of 2,457 older Austra-lians stratified by State and Territory from the 2000 Australian Electoral Commission roll. Two weeks later the 12-page questionnaire was posted. A reminder to complete the survey card was posted to all participants 4 weeks after the original letter had been mailed. Each respondent received the initial letter, questionnaire and follow-up postcard and a pre-paid envelope in which to return the completed survey. Completed surveys were received from 1,263 elderly Australians. This represents a response rate of 62% after allowing for confirmed non-deliveries. The survey instrument was designed to obtain information about health and lifestyle practices of older Australians, including their use of dietary supplements. The development of the survey involved two in-depth focus groups and survey pre-testing by residents in an aged care facility and individuals aged 65 years and over living independently. In total, 70 participants contributed to the construction and modification of the instrument.
Southern Cross University Human Research Ethics Committee approved the project. A copy of the survey is available upon request. Definition of supplement user In this study a supplement user was defined as someone who ticked ‘yes’ to the following question. “Do you take any of the following types of supplements; vitamins, minerals, herbal preparations or other health products?” Examples of each type of supplement were included in the questionnaire. ‘Other health products’ referred to such preparations as garlic tablets, fish oils, evening primrose oil, phytoestrogen formulas etc. Data analysis The Statistical Package for the Social Sciences for Windows (SPSS Inc, Chicago, version 10.0, 1999) was used to conduct the data analysis. Chi-square tests were conducted to determine the relationships between the supplement use and primary health variables. Two-way analysis of variance was used to examine the impact of gender and supplement use on the total number of con-ditions reported, symptoms experienced and medication used. Breslow Day Homogeneity of Odds Ratios for gender and Mantel-Haenszel Common Odds Ratio were used to estimate the probability of supplement use according to conditions, symptoms and medication used after controlling for gender. For all statistical tests, a significance level of P <0.05 was used. Results General features The demographic profile of the sample – age, gender, living arrangements, education, present income and ethni-city (as determined by place of birth) – has been reported elsewhere.31 Gender divided the sample into almost two equal groups i.e. 51% males (N=641) and 49% females (N=622). Respondents were aged between 65 and 98 years. For males, ages ranged from 65-98 years with a mean age of 73 years (SD 6.21). For females, ages ranged from 65-95 years with a mean age of 74 years (SD 6.76). The majority of males (64%) and females (63%) were aged between 65-74 years. Twenty-nine percent of males and 27% of females were aged between 75-84, 7% of males and 10% of females were aged between 85 and over. Health features: according to gender and supplement use Supplement use was unrelated to the number of visits to medical doctors (GP’s and specialists) in the six months prior to survey for males (χ2 (df:3) = 0.208, P = 0.996) and females (χ2 (df:3) = 0.957, P = 0.812); the number of days ill enough to restrict activity in the six months prior to survey for males (χ2 (df:3) = 5.272, P = 0.153) and females (χ2 (df:3) = 3.863, P = 0.277) and self-rated health status for males (χ2 (df:2) = 2.299, P = 0.317) and females (χ2 (df:2) = 0.379, P = 0.827). Table 1 shows the relationship between gender, supplement use and these health variables.
367 Health characteristics of older Australian dietary supplement users compared to non-supplement users
Table 1. Health features of the sample, gender and supplement useTable 2. Types of supplements utilised by survey sample Use of health and dietary supplements Forty-three percent of the sample (N =548, 224 males and 324 females) reported using at least one dietary or health supplement at the time of survey. More than half (52%, N = 324) of all females reported the use of at least one health or dietary supplement compared to just over one-third (35%, N = 224) of men. For all types of supple-ments, females were heavier users than males. The usage of health and dietary supplements by this sample is shown in Table 2. The supplements used most often were vitamin C and bioflavonoids (26%), multi-vitamin/mineral preparations (17%), fish oils (17%), vitamin E (16%), calcium (+/- vitamin D) (13%), garlic capsules or oil (11%), vitamin B (single or mixed) (9%), single vitamin or single mineral – other than those listed (7%), zinc (6%) and Gingko biloba (5%). Not shown on this table are those preparations taken by fewer than 5% of the subjects.
ANOVA results - total numbers of health conditions, symptoms experienced and medication used Square root transformations of a) total number of health conditions reported, b) total number of symptoms experienced and c) total amount of medication used was undertaken to improve the normality requirements for two-way ANOVA. Two-way ANOVA were conducted to explore the impact of gender and supplement utilisation on the transformed total number of health conditions, total number of symptoms experienced and total amount of medication used. For the total number of health conditions supplement use was not significant (P = 0.454) and gender was close to significant (P = 0.089) with the mean number of total health conditions for females (mean 2.21) greater than males (mean 2.00). For the total number of symptoms there was no significant gender effect (P = 0.116) but a significant supplement utilisation effect (P = 0.000) with mean number of symptoms for supplement users (mean 6.19) significantly higher than for non-supplement users (mean 5.22). For purposes of data analysis the frequency of medication use was dichotomised (no or yes which included daily, regularly and occasionally). For total number of medications used both factors supplement use and gender were statistically significant, P = 0.021 and P = 0.001 respectively. The total number of medications used by supplement users (mean 7.03) was significantly higher compared to non-supplement users (mean 6.19). Overall, females (mean 7.25) were higher users of medication than males (mean 5.98). Odds ratios results - health conditions, symptoms and medication used Mantel-Haenszel Odds Ratio was conducted to provide estimates of the probability of using supplements accor-ding to the type of health conditions reported, symptoms
Variable Total Any Supplement No Supplement
M
N =641
F
N =622
M
N =221
F
N =324
M
N =417
F
N =297 N % % % % % % % Visits to doctor (during the past 6 months) No visits 124 10 12 8 10 9 11 7 1-4 visits 730 58 53 59 58 58 57 60 5-9 visits 292 23 25 22 22 22 24 22 10+ visits 115 9 10 11 10 11 8 11 P value 0.121 Days ill enough to restrict activity (during the past 6 months) No days 991 78 79 79 80 76 77 81 1-9 days 200 16 16 15 17 16 16 14 10-49 days ill 58 5 4 5 3 6 5 5 50+ days ill 13 1 1 1 0 2 2 1 P value 0.613 Health status Fair to poor 286 23 24 22 25 22 23 22 Good 487 39 38 40 34 39 40 41 Very good to excellent 474 38 38 38 41 39 37 37 P value 0.807 * P value significant at P=<0.05, ** P value significant at P=<0.005
Supplement
N %
of supplement users
N=548 Vitamin C + bioflavonoid 141 26 Multivitamin/mineral 97 17 Fish oil 95 17 Vitamin E 89 16 Calcium (+/- vitamin D) 73 13 Garlic (capsules or oil) 60 11 Vitamin B (single or mixed) 53 9 Single vitamin or single mineral* 43 7 Zinc 34 6 Gingko biloba 28 5 * = other than those listed
S Brownie and M Rolfe 368
Common Odds Ratio
Confidence Intervals P
% who take supplements
% who take supplements
Health conditions Yes No Arthritis 50 38 1.48 1.18-1.87 0.000** Hypertension 40 46 0.77 0.61-0.96 0.024* Osteoporosis 64 41 2.16 1.47-3.17 0.000**
Symptoms experienced Yes No Anxiety 51 40 1.44 1.12-1.84 0.004** Constipation/diarrhoea 50 41 1.44 1.11-1.85 0.006* Fatigue 49 39 1.43 1.14-1.80 0.002** Lack of motivation 48 42 1.32 1.02-1.69 0.031* Sore throat 54 41 1.73 1.25-2.40 0.001** Trouble with neck, back or spine
49 38 1.58 1.26-1.98 0.000**
Urinary problems 49 41 1.42 1.09-1.85 0.009* Medication used Yes No
Allergy or hay fever 57 41 1.86 1.11-2.60 0.000** Antacids 50 42 1.46 1.11-1.95 0.009* Antibiotics 53 40 1.66 1.28-2.17 0.000** Blood pressure tablets 41 45 0.78 0.62-0.98 0.034* Cold and flu tablets 51 42 1.47 1.05-2.08 0.027* Cough medication 52 42 1.52 1.08-2.13 0.016* Eye drops or eye preparations
50 41 1.39 1.08-1.80 0.011*
Haemorrhoids or piles ointments
54 42 1.61 1.08-2.39 0.019*
Heart tablets 37 45 0.74 0.55-1.00 0.050* Pain relievers 48 40 1.34 1.07-1.69 0.012* Rheumatism/arthritis preparations
50 41 1.44 1.12-1.86 0.005*
*P value significant at P=<0.05; ** p value significant at P=<0.005 All odds ratios for supplement use by health conditions, symptoms experienced and medication used were homogenous over gender.
Table 3. Odds of taking a supplement according to health conditions, symptoms experienced and medication used
symptoms experienced and use of medication after controlling for gender. The odds of taking a supplement were significantly higher for those with arthritis (OR 1.48, CI 1.18-1.87) and osteoporosis (OR 2.16, CI 1.47-3.17). The odds of taking a supplement were significantly lower for those with hypertension (OR 0.77, CI 0.61-0.96). The odds of taking a supplement were significantly higher for those experiencing anxiety (OR 1.22, CI 0.97-1.53), constipation/diarrhoea (OR 1.44, CI 1.11-1.85), fatigue (OR 1.43, CI 1.14-1.80), lack of motivation (OR 1.32, CI 1.02-1.69), sore throat (OR 1.73, CI 1.25-2.40), trouble with the back, neck or spine (OR 1.58, CI 1.26-1.98), urinary problems (OR 1.42, CI 1.09-1.85). The odds of taking a supplement were significantly higher for those using allergy or hay fever medication (OR 1.86, CI 1.33-2.60), antacids (OR 1.46, CI 1.11-1.95), antibiotics (OR 1.66, CI 1.28-2.17), cold and flu tablets (OR 1.47, CI 1.05-2.08), cough medicines (OR 1.52, CI 1.08-2.13), eye drops or eye preparations (OR 1.39, CI 1.08-1.80), ointments or suppositories for hae-morrhoids (OR 1.61, CI 1.08-2.39), pain relievers (OR 1.34, CI 1.07-1.69) and rheumatism/arthritis preparations (OR 1.44, CI 1.12-1.86). The odds of taking a supplement were significantly lower for those individuals taking blood pressure medication (OR 0.78, CI 0.62-0.98) or heart tablets (OR 0.72, CI 0.53-0.95). Table 3 details the odds of taking a supplement according to health conditions, symptoms experienced and medication used (only significant findings reported).
Discussion Using a self-administered mail questionnaire data was obtained from 1,263 Australians aged 65 years and over proportionally selected (across all States and Territories) from the 2000 Australian Electoral Commission roll. This study has two unique features. Firstly, it is one of the largest surveys ever conducted in Australia aimed at examining the supplement practices of predominantly independently-living older individuals. Secondly, this study represents the most recent attempt in Australia to comprehensively profile the health differences between older supplement users and older non-supplement users. Supplement use was reported by 43% of the sample. This is consistent with international prevalence studies which show that supplement use in the range of 33% - 46% is typical among individuals of this age group.1, 6, 8, 9,
11, 16, 19-22 Results from our study conflict with an earlier finding,30 which reported that supplement users and non-users did not suffer from different medical conditions. Overall, 42% (n=534) and 11% (n=135) respectively, of our sample suffered from arthritis and osteoporosis. Individuals with these conditions were significantly more likely to use dietary supplements, than those without these conditions. However the types of supplements used by these individuals does not accord with current evidence. For example, recent studies have confirmed that gluco-samine sulphate is effective in relieving joint pain and may retard the progression of arthritis.32-35 Interestingly, half of those who experienced arthritis reported the use of
369 Health characteristics of older Australian dietary supplement users compared to non-supplement users
supplements, yet less than 5% specifically used gluco-samine preparations. In regard to osteoporosis, a similar observation was noted. Whereas almost two-thirds (64%) of those with osteoporosis used supplements, fewer than 15% specifically used calcium (with or without vitamin D) preparations. Houston et al.,21 and Yu et al.,12 revealed that persons with hypertension were less likely to take supplements than those without hypertension. This is consistent with the results of our study which also found that conditions such as heart disease, stroke, diabetes, and cancer appear to act as impediments to supplement use. A number of factors might explain this finding. For example, indi-viduals with these conditions may consider supplements ineffective to mitigate against these potentially life-threatening illnesses. They might be concerned about the cost, inconvenience and safety of taking dietary supple-ments concurrently with prescription medication. Although supplement users experienced more sym-ptoms (such as anxiety, constipation or diarrhoea, fatigue, lack of motivation, sore throats, trouble with their back, neck or spine and urinary problems), than non-supplement users, there was no significant difference in the way they perceived their health – 77% of both supplement users and non users considered their health as good to excellent. The total number of symptoms experi-enced was unrelated to gender. Physicians are often cited as the most influential source of information used by the elderly in making their decision to use nutritional supplements22,36,37 yet there was no difference in the number of visits to doctors or medical specialists between supplement users and non-users. Two other studies have measured the relationship between supplement and medication use and both found that there was no significant difference in the total number of medicines used by supplement users compared to non-supplement users.24,38 Neither study however examined whether the types of medicines taken varied between supplement users and non-supplement users. In our study, higher levels of medication use were recorded for supplement users compared to non-supplement users, and females compared to males. Overall, women reported a significantly higher usage of prescription and over-the-counter medication, including health and dietary supplements (52%). Supplement users were more likely to report the use of antacids, antibiotics, cold and flu tablets, cough medicines, eye preparations, ointments for haemorrhoids, pain relievers, medication for rheumatism/ arthritis and sleeping tablets compared to non-supplement users. In contrast, there was a reduced likelihood of taking a supplement for those with hypertension and by those using blood pressure medication and heart tablets. Because elderly people are at increased risk of adverse drug reactions38 it is important that studies designed to measure the use of dietary supplements in the elderly also obtain information about the use of prescription and over-the-counter medication. There is consensus within the international literature, that among older individuals poor dietary intake coupled with physiological changes associated with ageing, leads to inadequate levels of several nutrients including protein, calcium, zinc and B-group vitamins (including vitamin
B12, B6 and folate) and vitamin D.39-47 Current patterns of supplement usage reveal that although older indi-viduals consume a range of products, they favour preparations that are of limited value in addressing these known nutrient inadequacies. Our study, and many others,6,11,12,16,17,19-24,28,30,38,48,49 have shown that among older individuals, the most popular supplements are pre-parations which contain vitamin C, multivitamin/ minerals, fish oils, vitamin E, calcium, B group vitamins, zinc and iron. Older individuals frequently cite ‘the desire for more energy’ and ‘the prevention of coughs and colds’ as reasons for using supplements containing vita-mins B12, C, D and E.17,22,23,48 In the absence of defi-ciencies, the impact on energy and immunity of supple-mental amounts of these nutrients is probably of little benefit. Older supplement users may have unrealistic expectations of what can be achieved by taking additional amounts of certain nutrients. The popularity of some particular nutrients raises concern about the safety of their use by older individuals. Single nutrient preparations, such as vitamin C and vitamin E, often contain the highest multiples of the recommended dietary intake, compared to multi nutrient preparations. Therefore users of single nutrient supple-ments are at greatest risk of developing potentially adverse effects of taking excess doses of nutrients. Care needs to be taken since vitamin C supplements may inter-fere with diagnostic tests such as occult blood stool tests and urine glucose tests. Examination of the faeces for occult blood is used to screen for gastrointestinal patho-logy, including neoplasms. The presence of vitamin C in the faeces can inhibit occult blood tests, producing false negative results, which mask the true prevalence of blood in the stool.50 At supplementation above 400 IU/day, vitamin E may decrease platelet adhesion and increase clotting time. Individuals consuming vitamin E whilst on anticoagulants are well advised to have periodic moni-toring of their clotting times.51 Numerous investigations have evaluated the relation-ship between dental status,26 marital status,2,24,26,30,37 alcohol intake,8,16,27 dietary adequacy,30,54,53 seasonal vari-ation,27 membership to health insurance18 and supple-ment utilisation among predominantly older individuals. These variables were unaccounted for in this study, limiting the extent to which our findings can be compared to the existing literature and the general population of older supplement users. In conclusion, these results extend the current understanding of the contrast between supplement users and non-supplement users. We have been able to show distinct differences in the health profile of older supple-ment users compared to non-supplement users. Possibly supplement users view these products as a means to mini-mise their suffering, arrest deterioration and prolong functional independence. If so, older supplement users may need assistance to make informed, appropriate choices. Future research efforts need to establish the motives that maintain the widespread use of supplements by the elderly population; identify the sources of information, which guide the choices made by older supplement users and measure the contribution dietary supplements make to the nutritional status of older
S Brownie and M Rolfe 370
individuals. Until such data is available, the potential for benefit or harm associated with supplement use by older individuals cannot be established.
Acknowledgement This study received funding from Blackmore’s Ltd. Pty and Southern Cross University, Lismore, NSW. We are grateful for their financial contribution to this project. We would also like to thank Dr. Lyndon Brooks for his assistance in the construction of the survey and subsequent data analysis, Ian Howden and Sue Evans for providing critical feedback on the manuscript and Dr. Stephen Myers and Dr. John Stevens for their contribution to the study design. Our appreciation is extended to all those older Australians who provided such valuable information. References 1. Ervin RB, Wright JD, Kennedy-Stephenson J. Use of
dietary supplements in the United States, 1988-94. Vital & Health Statistics - Series 11: Data From the National Health Survey 1999; 244 (i-iii): 1-14
2. Koplan JP, Annest JL, Layde PM, Rubin GL. Nutrient Intake and Supplementation in the United States (NHANES II). Am J Public Health 1986; 76 (3): 287-289.
3. Kim I, Williamson DF, Byers T, Koplan JP. Vitamin and mineral supplement use and mortality in a US cohort [comment]. Am J Public Health 1993; 83 (4): 546-50.
4. MacLennan AH, Wilson DH, Taylor AW. Prevalence and cost of alternative medicine in Australia.[comment]. Lancet 1996; 347 (9001): 569-73.
5. MacLennan AH, Wilson DH, Taylor AW. The escalating cost and prevalence of alternative medicine. Preventive Med 2002; 35 (2): 166-73.
6. Stewart ML, McDonald JT, Levy AS, Schucker RE, Henderson DP. Vitamin/mineral supplement use: a telephone survey of adults in the United States. J Am Diet Assoc 1985; 85 (12): 1585-1590.
7. Block G, Cox C, Madans J, Schreiber GB, Licitra L, Melia N. Vitamin supplement use, by demographic charac-teristics. Am J Epidemiology 1988; 127 (2): 297-309.
8. Subar AF, Block G. Use of vitamin and mineral supplements: demographics and amounts of nutrients consumed. Am J Epidemiology 1990; 132: 1091-1101.
9. Bender MM, Levy AS, Schucker RE, Yetley EA. Trends in prevalence and magnitude of vitamin and mineral supplement usage and correlation with health status. J Am Diet Assoc 1992; 92 (9): 1096-1101.
10. Dorant E, Van den Brandt P, Hamstra A, Feenstra MH, Goldbohm RA, Hermus RJJ, Sturmans F. The use of vitamins, minerals and other dietary supplements in the Netherlands. Int J Vitamin & Nutrition Research 1993; 63 Supp: 4-10.
11. Slesinski MJ, Subar AF, Kahle LL. Trends in use of vitamin and mineral supplements in the United States: the 1987 and 1992 National Health Interview Surveys. J Am Diet Assoc 1995; 95(8): 921-3.
12. Yu XQ, Smith W, Webb K, Mitchell P, Leeder SR. Prevalence and predictors of dietary supplement use in an older Australian population. Aust J Nutr & Dietetics 1999; 56 (2): 69-75.
13. Kermode S, Myers S, Ramsay L. Natural and comple-mentary therapy utilisation on the north coast of NSW. Aust J Holistic Nursing 1998; 5 (2):7-13.
14. Worsley A. Dietary supplementation, naturalistic values and attitudes to the food supply among Australian women. Ecology of Food and Nutrition 1992; 28: 211-217.
15. Worsley A. Australian dietary supplementation practices: implications for health professionals. In: Wahlqvist ML, King R, McNeill J, Sewell R, eds. Food and Health: issues and directions. London: John Libbey, 1987; 11-17.
16. Lyle BJ, Mares-Perlman JA, Klein BE, Klein R, Greger JL. Supplement users differ from nonusers in demographic, lifestyle, dietary and health characteristics. J Nutr 1998; 128 (12): 2355-62.
17. Sheehan ET, Delett A, Read MH, Bendel R, Bhalla VS, Bock MA, Harrill I, Mitchell M, Schutz HG, Standal BR. Vitamin and mineral supplement practices and nutrition beliefs of the elderly in seven western states. Nutr Research 1989; 9: 251-258.
18. Gray SL, Hanlon JT, Fillenbaum GG, Wall WE, Jr., Bales C. Predictors of nutritional supplement use by the elderly. Pharmacotherapy 1996; 16 (4): 715-20.
19. Hartz SC, Otradovec CL, McGandy RB, Russell RM, Jacob RA, Sahyoun N, Peters H, Abrams D, Scura LA, Whinston-Perry RA. Nutrient supplement use by healthy elderly. J Am College of Nutr 1988; 7 (2): 119-128.
20. Schneider CL, Nordlund DJ. Prevalence of vitamin and mineral supplement use in the elderly. The Journal of Family Practice 1983; 17 (2): 243-247.
21. Houston DK, Daniel TD, Johnson MA, Poon LW. Demographic characteristics of supplement users in an elderly population. The Journal of Applied Gerontology 1998; 17 (1): 79-96.
22. Read MH, Graney AS. Food supplement usage by the elderly. J Am Diet Assoc 1982; 80: 250-253.
23. Hale WE, Stewart RB, Cerda JJ, Marks RG, May FE. Use of nutritional supplements in an ambulatory elderly population. J Am Geriatrics Soc 1982; 30 (6): 401-3.
24. Vitolins MZ, Quandt SA, Case LD, Bell RA, Arcury TA, McDonald J. Vitamin and mineral supplement use by older rural adults. J Gerontology Series A - Biological Sciences & Medical Sciences 2000; 55 (10): M613-7.
25. Looker A, Sempos CT, Johnson C, Yetley EA. Vitamin-mineral supplement use: association with dietary intake and iron status of adults. J Am Diet Assoc 1988; 88: 808-814.
26. Messerer M, Johansson SE, Wolk A. Sociodemographic and health behaviour factors among dietary supplement and natural remedy users. European J Clin Nutr 2001; 55: 1104-1110.
27. Wallstrom P, Elmstahl S, Hanson BS, Ostergren P, Johansson U, Janzon L, Larsson SA Demographic and psychosocial characteristics of middle-aged women and men who use dietary supplements. European J Public Health 1996; 6 (3): 188-95.
28. ABS. National Health Survey. Australian Bureau of Statistics 1995, Canberra; 4377.0
29. Worsley A, Crawford D. Australian dietary supplementation practices: Health and Dietary Supple-ments. Med J Aust 1984; 140: 579-583.
30. Horwath CC, Worsley A. Dietary supplement use in a randomly selected group of elderly Australians. Results from a large nutrition and health survey. J Am Geriatrics Soc 1989; 37: 689-696.
31. Brownie S, Myers SP. Dietary and health supplement use among older Australians: results from a national survey. Australasian J Ageing 2003; 22(4): 171-178.
32. Richy F, Bruyere O, Ethgen O, Cucherat M, Henrotin Y, Reginster JY. Structural and symptomatic efficacy of glucosamine and chondroitin in knee osteoarthritis: a comprehensive meta-analysis. Archives Internal Med 2003; 163 (13): 1514-1522.
371 Health characteristics of older Australian dietary supplement users compared to non-supplement users
33. Bruyere O, Honore A, Ethgen O, Rovati L, Giacovelli G, Henrotin Y, Seidel L, Reginster JY Correlation between radiographic severity of knee osteoarthritis and future disease progression. Results from a 3-year prospective, placebo-controlled study evaluating the effect of glucosamine sulphate. Osteoarthritis and Cartilage 2003; 11(1): 1-5.
34. Blakeley J, Ribeiro V. A survey of self-medication practices and perceived effectiveness of glucosamine products among older adults. Complementary Therapies in Med 2002; 10 (3):154-160.
35. Pavelka K, Gatterova J, Olejarova M, Machacek S, Giacovelli G, Rovati LC. Glucosamine sulphate use and delay of progression of knee osteoarthritis: a 3-year, randomized, placebo-controlled, double-blind study. Archives Internal Med 2002; 162 (18): 2113-23.
36. Cotugna N. Predictors of supplement use in the elderly. Part 1: a review of the literature. J Nutr for the Elderly 1989; 8 (3/4): 3-33.
37. Daly MP, Sobal J. Vitamin/mineral supplement use by geriatric outpatients in the United Kingdom. J Nutrition for the Elderly 1990; 10 (1): 55-64.
38. Brzozowska A, Enzi G, Amorin Cruz J. Medicine use and supplementation practice among participants of SENECA Study. J Nutrition, Health & Aging 2002; 6 (1):34-8.
39. Curran J. Overview of geriatric nutrition. Dysphagia 1990; 5: 72-76.
40. Drewnowski A, Warren-Mears VA. Does aging change nutrition requirements? J Nutrition, Health & Aging 2001; 5 (2): 70-4.
41. Chernoff R. President's page: nutrition and health for older Americans. J Am Diet Assoc 1996; 96 (10): 1053.
42. Blumberg JB. Nutritional needs of seniors. J Am Coll Nutr 1997; 16 (6): 517-523.
43. Young VR. Macronutrient needs in the elderly. Nutr Rev 1992; 50 (12): 454-462.
44. Gariballa SE, Sinclair AJ. Nutrition, ageing and ill health. Br J Nutr 1998; 80: 7-23.
45. Finkelstein JA, Schiffman SS. Workshop on taste and smell in the elderly: an overview. Physiology and Behaviour 1999; 66 (2): 173-176.
46. Blechman MB, Gelb AM. Aging and gastrointestinal physiology. Gastroenterology 1999; 15(3): 429-438.
47. Chen CC, Schilling LS, Lyder CH. A concept analysis of malnutrition in the elderly. J Advanced Nursing 2001; 36 (1): 131-42
48. Freeman MS, Sargent RG, Sharpe RG, Waller JL, Powell FM, Drane W. Cognitive, behavioral and environmental correlates of nutrient supplement use among independently living older adults. J Nutr Elderly 1998; 17 (3): 19-40.
49. Margarey AM, Tiddy JA, Wilson PC. The diets of elderly men and the use of dietary supplements: Australian non-veterans compared with war veterans. Aust J Nutr & Diet 1993; 50 (1): 25-28.
50. Jaffe R, Kasten B, DS Y. False negative stool occult blood tests caused by ingestion of ascorbic acid. Annals of Internal Med 1975; 83: 824-826.
51. Pryor WA. Vitamin E and heart disease: basic science to clinical intervention trials. Free Radical Bio & Med 2000; 28 (1): 141-64.
52. Kirk SF, Cade JE, Barrett JH, Conner M. Diet and lifestyle characteristics associated with dietary supplement use in women. Public Health Nutr 1999; 2 (1): 69-73.
53. de Jong N, Ocke MC, Branderhorst HAC, Friele R. Demographic and lifestyle characteristics of functional food consumers and dietary supplement users. Br J Nutr 2003; 89: 273-281.
Asia Pac J Clin Nutr 2004;13 (4): 372-376 372
Original Article Nutritional status of Saudi males living in the Riyadh nursing home Adel A Alhamdan PhD
Community Health Sciences Department, College of Applied Medical Sciences, King Saud University, Riyadh, Saudi Arabia.
This study evaluated the nutritional status of residents in the Riyadh nursing home, using anthropometric and haematological measurements. All male residents (N=84; age range 24-80 years) in the Riyadh nursing home were included in the study. Weight, height, body mass index, triceps skin fold thickness, and mid-arm muscle circumference were measured. Furthermore, serum concentrations of albumin, haemoglobin and haematocrit were measured. About 13% of adult residents and 11% of elderly residents were considered to be underweight (body mass index <18.5 kg/m2). From estimations of fat mass in the periphery, using triceps skin fold thickness, it appears that the elderly residents had significantly lower fat mass compared to the adult residents (P <0.05). The results showed that more than 40% of residents had low mid-arm muscle circumference (<22.3 cm). Serum albumin concentration was significantly lower in the elderly group than in the adult group (P <0.01). No significant difference was found in haematocrit level between the adult and elderly residents. Within the adult group, about 38% of residents had low haemoglobin level (<12 mg/dl), and this proportion was even higher, about 55%, in elderly residents. Based on body mass index or albumin to determine the prevalence of malnutrition among residents, the results have shown that the prevalence of undernourished residents was not higher than the prevalence of undernourished nursing-home residents reported in other studies. The percentage of elderly residents with anaemia was appreciable. Thus, undernourished and anemic residents should have special dietary and medical attention. Early detection of malnutrition upon admission would lead to early intervention and thus to reduced complications and medical-treatment costs. Staff working in nursing homes should be aware of the nutritional guidelines for health and disease.
Key Words: Nutritional status, anthropometric measurements, haematological measurements, nursing home, elderly, Saudi males, Saudi Arabia, Gulf Introduction Malnutrition is common in nursing homes. It has been reported that the prevalence of malnutrition in nursing-home residents ranges from 23-85%.1-3 The proportion of people who are 60 years of age and older in the Saudi pop-ulation is 4%, which is similar to that for other Gulf countries: Bahrain 5.2%, Kuwait 3.9%, Oman 4.5%, Qatar 6%4. With the attention that the government of Saudi Arabia is paying to improve health services, the proportion of elderly people is expected to increase in the future. Poor nutritional status is one of the major factors associated with mortality in older persons.5 It is well documented that poor nourishment impairs immunity, decreases resistance to infection and reduces the antioxidant-defence mechanisms of the body.6 The elderly are at increased risk of being undernourished because of inadequate food intake (e.g anorexia, dysphagia, reduced physical ability to eat as in arthritis), reduced desire to eat (e.g. depression, chronic pain), decline in food digestion or absorption, compro-mised metabolic pathways, poor dental hygiene or because of chronic diseases that are commonly found in the elderly.7,8 Mentally retarded persons in nursing homes are also at risk of becoming malnourished. In mentally retarded subjects, food choices may be restricted as a result of
physical abnormalities and poor control over feeding.9 Therefore, during a stay in a long-term care facility home, nutrition should become a major component of the reha-bilitative process in the elderly as well as in mentally-retarded subjects. The male nursing home in Riyadh, Kingdom of Saudi Arabia, accommodates and provides health and rehabi-litative care for elderly Saudi men aged 60 years and over, as well as Saudi men below 60 years of age, with stable conditions of neurospychiatric diseases (e.g. schi-zophrenia, dementia, hemiparalysis and Down syn-drome). Almost all of the people living in the home are without family or financial support. The present study was conducted in the Riyadh nursing home to assess the nutritional status of residents using anthropometric and haematological measurements.
Correspondence address: Adel Abdulwahab Alhamdan, Community Health Sciences Department, College of Applied Medical Sciences, King Saud University, P. O. Box 6838, Riyadh 11452, Saudi Arabia. Tel: +966 (1) 4355392; Fax: + 966 (1) 4355883 Email: [email protected] Accepted 28 May 2004
373 AA Alhamdan
Subjects and methods All Saudi males living in the Riyadh nursing home were included in the study (N=84). The mean age of the residents was 55.6 ± 16.9 years (range 24-80 years). Because of the large variation in the age of residents, and because of the effect of aging on physiological, functional and health status10,11, it was decided that the subjects should be divided into two age groups. The first group was the adult group (less than 60 years old, N = 39). The mean age of adult residents was 39.7 ± 1.63 years (range 24-58 years). The second group was the elderly group (≥ 60 years old, N =45). The mean age of elderly residents was 69.4 ± 0.77 years (range 60-80 years). The study was approved by the Nursing home ethics committee. Anthropometric and haematological measurements All anthropometric measurements were collected by a single well-trained physician. Measurements of weight (to the nearest 0.1 kg) and height (to the nearest 0.1 cm) were made using a portable scale and a portable stadiometer, respectively. Knee height was used to estimate the stature of a person who could not stand, or for a person with an obvious spinal curvature. The following equation was used to estimate the stature from knee height. Stature for men = [(2.02 × knee height) – (0.24 × age)].12 Body mass index (BMI) was calculated by dividing the weight in kilograms by the square of the stature in meters (kg/m2). Triceps skin fold thickness (TSFT) was taken on the back of the left upper arm by measuring the distance from the acromion to the olecranon process and marking the half point. TSFT was measured to the nearest 0.2 mm with a calibrated caliper, and it provides an estimate of body fat. To measure mid-arm muscle circumference (MAMC), which reflects muscle mass, mid-arm circumference (MAC) was taken on the front of the left-upper arm by measuring the halfway distance between the inferior aspect of the acromion and the olecranon. MAC was measured to the nearest 0.1 cm using a flexible non-elastic tape. MAMC was determined by measuring MAC and TSFT using the following equation: MAMC (cm) = MAC (cm) – [0.314 × TSFT (mm)].13 BMI (kg/m2) was classified into underweight [< 18.5 kg/m2], normal weight [18.5-24.9 kg/m2], overweight [25-29.9 kg/m2], and obese [> 30 kg/m2] (14). The following desirable ranges, 4-25 (mm) and 22.3-30.6 (cm), were used for the TSFT and MAMC measurements, respectively.15 For haematological analysis, a fasting venous blood sample was taken. Analysis included serum albumin, haemoglobin (Hb) and hematocrit (Hct). A serum albumin level < 35 g/l was considered to be low.16 A serum Hb level between 12 and 13.9 mg/dl was considered to be bordering on low, while a serum Hb level of <12 mg/dl was considered to be low. A serum Hct level between 37% and 43% was considered to be bordering on low, while a serum Hct level < 37% was considered to be low.17
Statistical analysis Results were expressed as mean values ± standard error of the mean (SEM). When comparing the mean values be-tween adult and elderly residents, statistical analysis was
performed using unpaired Student’s t test, with a signi-ficance level of P <0.05. The number and the percentage of adult or elderly residents with normal and abnormal anthropometric and haematological measurements were also estimated. Results The mean values of the BMI and MAMC showed no significant difference between the adult and the elderly residents. However, the results of the TSFT showed that the elderly residents had significantly lower mean values than the adult residents (Table 1). Table 1. Comparison of the anthropometric measurements between adult and elderly residents Table 2. Comparison of the haematological measurements between adult and elderly residents. Bagpgo5rvrbrt iracecitw
Measurement Adult residents
Elderly residents
P-value
BMI (kg/m2) 25.7 ± 0.93 24.2 ± 0.85 0.23
MAMC (cm) 27.6 ± 0.74 25.8 ± 0.67 0.083
TSFT (mm) 16.0 ± 1.46 11.5 ± 1.07 0.015
Anthropometric measurements. BMI: Body mass index. MAMC: Mid-arm muscle circumference. TSFT: Triceps skin fold thickness. Values are means ± SEM. N = 39 for the adult group, and N = 45 for the elderly group. Measurements with P<0.05 indicate a signi-ficant different between the groups.
Vtd
Measurement Adult residents
Elderly residents
P value
Albumin (g/l) 42.0 ± 0.68 38.5 ± 0.60 0.002
Hg (mg/dl) 14.1 ± 0.20 13.6 ± 0.22 0.053
Hct (%) 41.1 ± 0.60 40.6 ± 0.70 0.59 alues are means ± SEM. N=39 for the adult group, and n=45 for
he elderly group. Measurements with P<0.05 indicate a significant ifferent between the groups.
ased on BMI measurement, Table 3 shows that 12.8% nd 11.1% of the residents in the adult group and elderly roup, respectively, were considered underweight. The ercentage of obese residents was 25.6% in the adult roup, but only 13.3% in the elderly group. Measurement f the MAMC indicated that 41% of adult residents and 5.6% of the elderly residents were below the desirable ange (Table 3). Despite the fact that the TSFT-mean alues were lower in the elderly residents than in the adult esidents, no elderly resident was found to have a TSFT elow the desirable range (Table 3). Three (6.7%) elderly esidents and 7 (17.9%) adult residents had TSFT above he desirable range. Serum albumin concentration was significantly lower n the elderly group than in the adult group (Table 2). The esults show that the percentage of adult residents with lbumin concentration below 35 g/l was 2.6%, and the orresponding percentage was much higher (18%) in the lderly residents (Table 4). Table 2 shows that Hg oncentration was slightly lower in the elderly group than n the adult group, but the lower value was not sta-istically significant (P = 0,053). No significant difference as found in Hct levels between the adult and elderly
Nutritional status of Saudi males living in the Riyadh nursing home 374
Table 3. The number and percentage1 of adult and elderly residents with normal and abnormal anthropometric mea-surements. residents (Table 2). Within the adult group, 38.4% of residents had Hg level below normal, and the corresponding percentage was higher for elderly residents (55.5%) (Table 4). A large proportion of adult (82%) and elderly (75.6%) residents had an Ht level below normal (Table 4). Discussion Hypoalbuminemia and being underweight has been reported to affect 30-50% of nursing-home residents.18 In the present study, when BMI was used as an indicator of malnutrition (underweight and obese subjects), the percentage of malnourished residents was 38.4% and 24.4% for the adult and elderly respectively. When albumin was used as an indicator of malnutrition, the percentage of undernourished residents was 2.6% and 17.8% for the adult and elderly residents, respectively. Based on the results of the BMI and albumin, we can conclude that the prevalence of malnutrition in the Riyadh nursing home was not higher than the prevalence reported for other nursing homes.1,2,3,18 Using the World Health Organization (WHO) criteria of anaemia (< 13 g/l in men), the prevalence of anaemia in the elderly has been reported to be in the range of 8% to 44%.19,20 In the present study, also when using the WHO criteria of anaemia, the percentage of anemic residents in the nursing home was 17.9% (n=7) in the adult residents and 40% (n=18) in the elderly residents. Thus, the pre-valence of anaemia in the elderly residents was appre-ciable. It has been reported that the most common causes of anaemia in the elderly are chronic diseases (30% to 45% of cases) and iron deficiency (15% to 30% of cases), while 15% to 25% of cases have no identifiable cause.21 The dietary habits of old people and their previous dietary histories (what they ate and drank during childhood and adulthood), in addition to genetic factors, affect the degree of the response to aging,
Table 4. The number and percentage1 of adult and elderly residents with normal and abnormal haematological mea-surements. so that they age at different rates. Because of the large variability in health status, physiologic function, physical activity, and nutritional status among older adults, there is no set of reference data for a representative sample of the elderly. Hence, it is important to combine anthropometric measurements with other biochemical and/or dietary measurements to get a more reliable indication of their nutritional status. When dietary inadequacy is defined as the dietary intake of four or more of the following nutrients (protein, thiamine, vitamin A, riboflavin, vitamin C, niacin, cal-cium and iron) falling below two-thirds of the recom-mended dietary allowance, Shahar and his colleges have shown that in elderly Malaysian subjects, consuming less than three meals per day is one of the determinants of dietary inadequacy.22 In the present study, we have found that all elderly residents (N=45) ate at least two full meals per day, 86.6% (N=39) of whom ate three full meals per day. When the dietary intake of dairy products, beans, eggs, fish, meat and poultry was recorded as markers for protein intake, we found that most of the residents, 97.3% of adult residents and 84.4% of elderly residents, con-sumed at least one serving of dairy products/day, two or more servings of beans or eggs per week, and ate fish, meat or poultry every day. Several studies have shown a significant decrease in lean body mass and a significant increase in body fat in elderly subjects.23-25 The present study has shown a decrease in lean body mass in elderly subjects, as measured by MAMC, when compared to adult residents, but this was not statistically significant (P <0.083). However, when using TSFT to estimate body fat, the results showed a significant reduction in the fat mass in the elderly subjects compared to the adult subjects. In old age, a redistribution of fat occurs, in which more fat is deposited in the trunk than in the extremities. Thus, adipose tissue thickness decreases in the limbs and increases in the abdominal area and around the internal organs.26-28 Minten et al., (1991) have found a relatively
Measurements Adult residents (N=39)
Elderly residents(N=45)
BMI (kg/m2) Underweight Normal weight Overweight Obese
5 (12.8%)
12 (30.8%) 12 (30.8%) 10 (25.6%)
5 (11.1%)
24 (53.3%) 10 (22.2%) 6 (13.3%)
MAMC (cm) Below desirable range Desirable range
Above normal range
16 (41.0%) 23 (59.0%)
0 (0%)
25 (55.6%) 19 (42.2%)
1 (2.2%)
TSFT (mm) Below desirable range Desirable range
Above normal range
1 (2.6%) 31 (79.5%) 7 (17.9%)
0 (0%) 42 (93.3%)
3 (6.7%)
Anthropometric measurements. BMI: Body mass index; MAMC: Mid-arm muscle circumference; TSFT: Triceps skin fold thickness.1Sum percentage may not equal 100% due to rounding.
Measurements Adult residents (N = 39)
Elderly residents (N = 45)
Albumin (g/l) Normal (≥ 35) Low (< 35)
38 (97.4%) 1 (2.6%)
37 (82.2%) 8 (17.8%)
Hb (mg/dl) Normal (≥ 14)
Borderline (12-13.9) Low (< 12)
24 (61.5%) 13 (33.3%)
2 (5.1%)
20 (44.4%) 19 (42.2%) 6 (13.3%)
HCT (%) Normal (≥ 44)
Borderline (37-43) Low (< 37)
7 (17.9%)
27 (69.2%) 5 (12.8%)
11 (24.4%) 25 (55.6%)
9 (20%)
Hb: haemoglobin; HCT: haematocrit. 1Sum percentage may not equal 100% due to rounding.
375 AA Alhamdan
low correlation between body mass index and body fat (assessed by biceps and triceps skin-fold thickness) in old age.29 Thus, using TSFT may not be a good indicator for the estimation of body fat in elderly subjects. Due to this alteration in fat distribution, measuring waist-to-hip ratio may be a better predictor of overall fatness in old age.29,30 Despite the fact that there are some difficulties in interpretation of anthropometric measurements in old age, due to changes in body composition, these are essential for providing basic descriptive information.26 Early detection of malnourished residents upon admission to the nursing home would lead to early intervention and thus to a reduction in medical-treatment costs and also reduced complications. The staff working in the nursing home should be aware of the nutritional guidelines for health and disease. Improving the know-ledge of nutrition amongst staff and residents would improve the nutritional status of all individuals in the nursing home. Kim et al., (1981) have observed a sig-nificant improvement in dietary intake during the period of a program related to nutrition and healthy food habits, but this improvement was not sustained after the program ended, indicating that individuals in nursing homes need great attention and a continuous nutritional education program.31 Good eating habits based on moderation and variety and following general guidelines for reducing the risk of chronic diseases, for example by maintaining healthy weight, reducing the intake of dietary fat, particularly saturated fat, increasing the intake of fruit and vegetables and whole-grain products, and using salt and sugar in moderation, is the best advice that we can give the residents of nursing homes. Acknowledgement We would like to thank Dr. Yasser Alquathy for his valuable comments, and for his valuable assistant in taking the anthropommetric measurements and for collecting the blood samples. References 1. Seiler WO. Clinical pictures of malnutrition in ill elderly
subjects. Nutrition 2001; 17(6): 496-498. 2. Silver KJ, Morley JE, Strome LS, Jones D, Vickers L.
Nutritional status in an academic nursing home. Journal of the American Geriatric Society 1988; 36: 487-491.
3. Shaver HJ, Loper JA, Lutes RA. Nutritional status of nursing home patients. Journal of Parenteral and Enteral Nutrition 1980; 4: 367-370.
4-. Hafez G, Bagchi K, Mahaini R. Caring for the elderly: a report on the status of care for the elderly in the eastern Mediterranean region. Eastern Mediterranean Health Journal 2000; 6(4): 633-643.
5. Morley JE. Management of nutritional problems in sub-acute care. Clinics in Geriatric Medicine 2000; 16: 817-834.
6. Chandra RC. Nutrition and the immune system. Proceedings of the Nutrition Society 1993; 52: 77-84.
7. Johnson RM, Kaiser FE, Kerstetter JE, Reuben DB. Maintaining good nutrition in the elderly. Patient Care 1995; 29(18): 46-55.
8. Cederholm T, Jägr´en C, Hellström K. Nutritional status and performance capacity in internal medical patients. Clinical Nutrition 1993; 12: 8-14.
9. Lindeman, AK. Resident managers’ nutrition concerns for staff and residents of group homes for mentally retarded adults. Journal of the American Dietetic Association 1991; 91(5): 602-604.
10. Morley JE. Anorexia of aging: physiologic and pathologic. American Journal of Clinical Nutrition 1997; 66(4): 760-773.
11. Marcus EL, Berry EM. Refusal to eat in elderly. Nutrition Reviews 1998; 56(6): 163-171.
12. Chumlea, W.C., Roche, A.F., Mukherjee, D., eds. Nutritional assessment of the elderly through anthropometry. Colombus, Ohio: Ross Laboratories, 1987.
13. Bishop CW, Bowen PE, Ritchey SJ. Norms for nutritional assessment of American adults by upper arm anthropometry. American Journal of Clinical Nutrition 1981; 34: 2530-2539.
14-. National Heart, Lung and Blood Institute Expert Panel. Executive summary of the clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. Journal of the American Dietetic Association 1998; 98: 1178-1191.
15. Nelson KJ, Coulston AM, Sucher KP, Tseny RY. Prevalence of malnutrition in the elderly admitted to long-term-care facilities. Journal of the American Dietetic Association 1993; 93 (4): 459-461.
16. Heymsfield SB, Williams PJ. Nutritional assessment by clinical and biochemical methods. In: Modern Nutrition in Health and Disease. Eds. ME Shils & VR Young, pp. 817-860. Philadelphia, Pa: Lea and Febiger, 1988.
17. King, J.W.; Fauker, W.R., eds. Critical resources in clinical laboratory sciences, Cleveland, Ohio: CR Press, 1973.
18. Rudman D. Nutrition and fitness in elderly people. American Journal of Clinical Nutrition 1989; 49: 1090-1098.
19. Ania BJ, Suman VJ, Fairbanks VF, Melton LJ. Prevalence of anaemia in medical practice: community versus referral patients. Mayo Clinic Proceedings 1994; 69(8): 730-735.
20. Daly MP. Anaemia in the elderly. American Family Physician 1989; 39: 129-136.
21. Joosten E, Pelemans W, Hiele M, Noyen J, Verhaeghe R, Boogaerts MA. Prevalence and causes of anaemia in a geriatric hospitalized population. Gerontology 1992; 38: 111-117.
22. Shahar S, Dixon RA, Earland J. Development of a screening tool for detecting undernutrition and dietary inadequacy among rural elderly in Malaysia: Simple indices to identify individuals at high risk. International Journal of Food Sciences and Nutrition 1999; 50 (6): 435-444.
23. Noppa H, Andersson M, Bengtsson C, Bruce A, Isaksson B. The population study of women in Göteborg, Sweden. American Journal of Clinical Nutrition 1980; 33: 155-162.
24. Steen GB, Isaksson B, Svanberg A. Body composition at 70 and 75 years of age: a longitudinal population study. Journal of Clinical and Experimental Gerontology 1979; 27: 185-200.
25. Noppa H, Andersson M, Bengtsson C, Bruce A, Isaksson B. Body composition in middle-aged women with special reference to the correlation between body fat mass and anthropometric data. American Journal of Clinical Nutrition 1979; 32: 1388-1395.
26. De Groot LC, Sette S, Zajkas G, Carbajal A, Amorim JA. Nutritional status: anthropometery. European Journal of Clinical Nutrition 1991; 45 (supp. 3): 31-42.
Nutritional status of Saudi males living in the Riyadh nursing home 376
27. Euzi G, Gasparo M, Biondetti PR, Fiore D, Senisa M, Zurlo F. Subcutaneous and visceral fat distribution according to sex, age and overweight evaluated by computed tomography. American Journal of Clinical Nutrition 1987; 45: 7-13.
28. Baumgartuer RN, Heymsfield SB, Roche AF, Bernardino M. Quantification of abdominal composition by computed tomography. American Journal of Clinical Nutrition 1989; 50: 221-226.
29. Minten VA, Lowik MRH, Deurenberg P, Kok FJ. Inconsistent associations among anthropometric measurements in elderly Dutch men and women. Journal of the American Dietetic Association 1991; 91 (11):1408-1412.
30. Kuczmarski, RJ. Need for body composition information in elderly subjects. American Journal of Clinical Nutrition 1989; 50: 1150-1157.
31. Kim S, Schriver JE, Campbell KM. Nutrition education for nursing home residents. Journal of the American Dietetic Association 1981; 78(4): 362-366.
377 Asia Pac J Clin Nutr 2004;13 (4):377-384
Original Article Dietary intake and the risk of coronary heart disease among the coconut-consuming Minangkabau in West Sumatra, Indonesia Nur I Lipoeto MD MMedSci PhD 1, Zulkarnain Agus, MD MPH 1, Fadil Oenzil MD, PhD 1, Mark L Wahlqvist BMedSc, MBBS, MD FRACP, FACN 2 and Naiyana Wattanapenpaiboon BSc (Pharm), MSc (Pharm), PhD 2 1Faculty of Medicine, Andalas University, Padang,West Sumatra,Indonesia 2Asia Pacific Health & Nutrition Centre, Monash Asia Institute, Monash University, Melbourne, Victoria, Australia
Several nutrition and non-nutritional pathways are recognised in the development and occurrence of cardiovascular disease. In many populations, high intakes of saturated fat are associated with elevated serum cholesterol concentrations and increased coronary heart disease (CHD) mortality. However, several studies report that hyperlipidaemia and heart diseases are not common among populations who consume coconut, a source of saturated fat. A case-control study was conducted among the Minangkabau known to be high coconut consumers to examine the difference in food patterns and risk of coronary heart disease (CHD) between the coronary cases and their gender- and age-matched apparently healthy counterparts serving as controls. Eligible subjects with CHD were identified through the co-operation of five participating hospitals located in Padang and Bukittinggi in West Sumatra, Indonesia. A total of 93 eligible cases (62 men and 31 women) in the Case group and 189 subjects (113 men and 76 women) in the Control group were recruited. Information on the intakes of individual foods and dishes over the preceding 12 months was obtained using a semi-quantitative food frequency questionnaire. The Case group had significantly higher intakes of meats, eggs, sugar, tea, coffee and fruits, but lower intakes of soy products, rice and cereals compared to the controls. Coconut consumption as flesh or milk was not different between cases and controls. The cases had significantly higher intakes of protein and cholesterol, but lower intake of carbohydrate. Similar intakes of saturated and unsaturated fatty acids between the cases and controls indicated that the consumption of total fat or saturated fat, including that from coconut, was not a predictor for CHD in this food culture. However, the intakes of animal foods, total protein, dietary cholesterol and less plant derived carbohydrates were predictors of CHD.
Key Words: dietary intake, coconut consumption, coronary heart disease, case-control study, Minangkabau, saturated fat, Padang, Bukittinggi, West Sumatra, Indonesia. Introduction Epidemiological studies suggest a strong association between coronary heart disease (CHD) and several dietary factors.1-4 High intakes of saturated fat in different pop-ulations are associated with the elevation of serum cho-lesterol concentrations and the mortality in CHD.5-7 Expe-rimental and metabolic studies suggest that coconut con-sumption can cause hyperlipidaemia and atherosclerosis. However, several studies report that hyperlipidaemia and heart diseases are uncommon among high coconut con-suming populations.8-9 Furthermore, most studies to date have assessed Caucasian populations consuming high intakes of total fat and low intakes of fish with the risk of developing CHD.10 Only few studies are available ex-ploring associations between coronary disease events in Asian populations with low total dietary fat and high fish and coconut intakes.11 Minangkabau food culture therefore provides a unique opportunity to investigate whether the CHD risk could be predicted by the food patterns. This
case-control study was conducted to examine the difference in food patterns and CHD risk between the CHD cases and their gender- and age-matched healthy individuals serving as the controls.
Subjects and methods The study was conducted from February to August 1999. This research project was approved by the Monash University Standing Committee on Ethics in Research Involving Humans in November 1998 (Project Number 98/458) and acquired a written permission from the local government in West Sumatra, Indonesia. All subjects gave Correspondence address: Dr N Tikky Wattanapenpaiboon, Monash Asia Institute, Building 11A, Monash University VIC 3800, Australia. Tel: (+61 3) 9905 8145; Fax: (+61 3) 9905 8146 Email: [email protected] Accepted 16 April 2004
NI Lipoeto, Z Agus, F Oenzil, ML Wahlqvist and N Wattanapenpaiboon 378
written informed consent prior to their participation in the study. All patients diagnosed with CHD less than 6 months were eligible for this study. All patients enrolled in the study had to have received the diagnosis from cardio-logists. The diagnosis followed the World Health Organi-zation criteria which was based on information on typical symptoms, typical changes in electrocardiograph (ECG) or enzymes.12 Eligible patients were identified through the co-operation of five participating hospitals located in two cities in West Sumatra. The two cities were Padang, which is the capital of West Sumatra province, and Bukittinggi, a more rural county situated on the mountainous area 88 km north of Padang. Subjects in the case groups were recruited from the outpatient clinic of the Cardiovascular Unit in the five hospitals in Padang and Bukittinggi. Subjects in the Control group were recruited from the outpatient Ear, Nose and Throat and the Eyes Clinics from the same hospitals and came from the same areas as the cases. The controls were randomly selected from people matched to the Case subjects on the basis of age and gender. Subjects in the Control group who had health problems related to cardiovascular diseases, such as hypertension and diabetes mellitus, were not included. Pregnant women were also excluded. A total of 93 eligible cases (62 men and 31 women) and 189 subjects (113 men and 76 women) in the Control group were recruited. A questionnaire on demography, health status, lifestyle and general food habits and practices was developed and administered. Information on the intakes of individual foods and dishes over the past 12 months was obtained using a semi-quantitative food frequency questionnaire. Where an ingredient in a mixed dish or recipe was used in small amounts to add flavours, it was difficult to quantify. For example, coconut milk and grated coconut intakes were assessed in dishes where they were major ingre-dients. Where small amounts of either coconut milk or grated coconut were used, such as chicken satay or beef rendang, the dishes were included in the estimation of chicken or beef intake, but not in the estimation of coconut milk intake, even though they contained coconut milk (Table 1). When food dishes were categorised as animal or plant foods, this was also judged from the major ingredient(s) of those dishes (Table 2). Nutrient intakes
included carbohydrates, proteins, total fat, monoun-saturated fat, polyunsaturated fat (n-6 and n-3) and cho-lesterol. Information on the nutrient content of food items was obtained from the Nutrient Composition of Indo-nesian Foods and the Nutrient Composition of Malaysian foods.13 Data on the fatty acid composition of various foods was taken from a USDA Nutrient Database for Standard Reference, www.nal.usda.gov/fnic/cgi-bin/ list_ nut.pl., retrieved on the 27th of February 2001 and used to calculate individual fatty acid intakes (United State Department of Agriculture, 2001). Subjects whose daily energy intake was implausibly low or high, for example, less than 2,094 kJ (500 kcal) or more than 14,650 kJ (3,500 kcal), were not included in the further data analyses of food or nutrient intakes.14 Nutrient intakes were presented in actual grams/day and as a percentage of total energy intake. Statistical analysis The Statistical Analysis System (SAS software version 6.12 for Windows, SAS Institute Inc., NC, USA, 1996) was used for all data analysis. All data analysis proce-dures were performed under the SAS/ASSIST. De-scriptive statistics were used to report sample distri-butions and attributes for confounders and antecedent factors. Mean, standard deviation and percentiles were used for continuous variables, whereas for discrete va-riables, frequency and percentage were derived. To examine the associations of CHD events with food and nutrient variables, all subjects were divided into four equal groups according to the quartile values of food and nutrient intakes. The odds ratio (OR) was computed as the rate in a specific quartile divided by the value in the group with lowest intake. In multivariate analysis, energy intake, the percentage of energy derived from fat intake, and other potentially confounding variables were simul-taneously included into the models.14 Results Food intake Table 1 shows the mean intakes of each food group for Cases and Controls The cases had significantly higher intakes of meats, eggs, sugar, tea, coffee and fruits, but lower intakes of soy products, rice and cereals compared
Table 1. Descriptive statistics of average daily food consumption (g/day) by food groups for cases and controlsCases Controls Food groups Mean ± SD Mean ± SD Fish, seafood and products 67.5 ± 33.8 58.4 ± 30.8 Meats and meat products 47.1 ± 40.2** 34.5 ± 28.5 Eggs 37.7 ± 31.7** 26.8 ± 19.3 Milk and milk products 20.0 ± 39.0 17.0 ± 30.4 Soy products 90.1 ± 82.5 100.7 ± 79.5 Legumes and nuts 81.1 ± 43.1 77.0 ± 54.1 Coconut milk and grated coconut 42.0 ± 21.7 38.2 ± 18.2 Rice and cereals 382.3 ± 108.9*** 407.4 ± 116.0 Vegetables 338.0 ± 146.6 303.4 ± 108.4 Fruits 129.9 ± 73.7** 114.9 ± 69.8 Non alcoholic beverages 131.1 ± 105.6** 126.4 ± 78.6 Sugar 34.6 ± 28.8* 31.6 ± 21.0 Palm oil 25.3 ± 14.2 23.3 ± 10.5 Significantly different from the Control group: *, P<0.05; **, P<0.01; ***, P<0.001.
379 Dietary intake and the risk of coronary heart disease among the Minangkabau in West Sumatra
to the controls. The use of animal fat spreads was not common amongst the Minangkabau. For margarine, only 22% and 7% of the cases and controls respectively used margarine (1.5 teaspoons/d consumed by the cases compared to 0.7 by the controls) (P<0.001).
Animal foods The animal food group included fish, eggs, beef, chicken, and dairy foods. The Case group had significantly higher intakes of total animal foods, compared to the Control group (247 g/d vs 187 g/d, P<0.0001). This was mainly due to the difference in the intakes of meat and eggs. Table 2 shows distribution of the Cases and Controls according to animal foods consumed. The odds ratio for subjects who consumed animal foods in the highest quartile (above 210 g) to those in the lowest quartile (below 108 grams) was 4.8 (95% CI 2.25-10.30, P<0.0001).
Plant foods The plant food group included rice and cereal, tempeh and tofu, legumes and nuts, vegetables, fruits and coconut milk and grated coconut. There was no difference in plant food intake between the two groups. Average intake of plant food was 1,061 g/d for the Case group and 1,028 g/d in the Control group.
Macronutrients Table 3 shows the descriptive statistics of the macro-nutrients for the Case and Control groups. In total, the
Table 2. Distribution of cases and controls according to their animal food consumption
Total Mean ± SD Cases n = 93
Total energy (kcal) 1765 ± 534 Total energy (kJ) 7502 ± 2436 Protein (g) 92.0 ± 33.5 ** % total energy 20.6 ± 3.1 *** Nutrient density (g/MJ) 12.3 ± 1.9 Carbohydrates (g) 204.1 ± 55.9 ** % total energy 55.8 ± 7.5 *** Nutrient density (g/MJ) 28.0 ± 4.6 Total fats (g) 47.2 ± 20.9 % total energy 23.6 ± 5.6 Nutrient density (g/MJ) 6.4 ± 1.4 Dietary cholesterol (mg) 296 ± 205 ***Fibre (g) 10.1 ± 4.0
Controls n = 189 Total energy (kcal) 1657 ± 487 Total energy (kJ) 6934 ± 2038 Protein (g) 79.3 ± 28.2 % total energy 19.0 ± 2.7 Nutrient density (g/MJ) 11.4 ± 1.6 Carbohydrates (g) 204.7 ± 52.9 % total energy 57.6 ± 6.5 Nutrient density (g/MJ) 30.1 ± 3.9 Total fats (g) 44.0 ± 18.6 % total energy 23.4 ± 4.9 Nutrient density (g/MJ) 6.2 ± 1.3 Dietary cholesterol (mg) 187 ± 109 Fibre (g) 9.3 ± 3.6
Significantly different from the Control group: *, P<0.05; **, P<0.01; **
Cases Controls Animal food intake (g/d) n % n % <108.1 14 15.1 57 30.1 108.1 – 149.8 18 19.4 52 27.5 149.8 – 210.4 27 29.0 43 22.8 >210.4 34 36.6 37 19.6
Case group had significantly higher intakes of protein and cholesterol, but a lower intake of carbohydrate (P<0.0001 for all cases). Between the men, the cases had signi-ficantly higher intake of total energy, protein and cholesterol, but a lower intake of carbohydrate (P<0.05, P<0.01, P<0.0001 and P<0.001 respectively). Between the women, the cases had a significantly lower intake of carbohydrate (P<0.05). As expected, rice was the major food source of total energy intake in this population. Rice and cereals contributed as much as 32% and 36% of total energy for the Case and Control groups, respectively. Other major food sources of total energy were fish, vegetables and soy dishes for both groups. Amongst the macronutrients included in the univariate logistic regression analysis, intakes of protein and cholesterol were found to be a risk factor for CHD. More subjects in the Case group were in the highest quartiles of protein and total cholesterol (Table 4). The odds ratio for those with intakes of protein and cholesterol in the highest quartile were 1.01 (95% CI 1.01–1.02) and 1.00 (95% CI
Table 3. Macronutrient intakes of the case and control groups
Men Women Mean ± SD Mean ± SD
n = 62 n = 31 1823 ± 548 * 1647 ± 491 7658 ± 2303 6893 ± 2054
93.1 ± 34.2 ** 88.5 ± 32.5 **** * 20.3 ± 3.3 ** 21.1 ± 2.7 ****
12.1 ± 1.9 12.6 ± 1.6 * 213.2 ± 58.3 ** 186.0 ± 46.1 * 56.6 ± 7.5 ** 55.3 ± 7.5 *
28.4 ± 4.5 27.8 ± 4.5 47.7 ± 21.2 46.4 ± 20.6 23.1 ± 5.5 24.6 ± 5.8
6.1 ± 1.5 6.5 ± 1.5 * 327 ± 207 **** 235 ± 192
10.1 ± 4.1 10.0 ± 3.6 n = 113 n = 76
1666 ± 475 1643 ± 507 6971 ± 1987 6877 ± 2122 80.4 ± 28.9 77.7 ± 27.1 19.1 ± 2.8 18.9 ± 2.5 11.4 ± 1.7 11.3 ± 1.5
205.7 ± 50.0 203.1 ± 57.4 57.7 ± 6.7 57.5 ± 6.2 30.1 ± 40.5 29.9 ± 3.7 43.9 ± 18.1 44.2 ± 19.3 23.2 ± 4.8 23.6 ± 5.1
6.2 ± 1.3 6.3 ± 1.4 190 ± 113 182 ± 103 9.2 ± 3.4 9.4 ± 3.9
*, P<0.001; ****, P<0.0001.
NI Lipoeto, Z Agus, F Oenzil, ML Wahlqvist and N Wattanapenpaiboon 380
Table 4. Distribution of cases and controls in the highest and lowest quartiles, the odds ratio (95% confidence intervals) according to their protein and cholesterol intakes
Cases Controls
n (%) n (%) Odds ratio
95% CI
Protein (g/day)
1st quartile (<61.7 g/day) 16 (17.2) 54 (28.6) 1.01 1.01 – 1.02
4th quartile (>100.2 g/d) 35 (37.6) 36 (19.0)
Cholesterol (mg/day)
1st quartile (<125.2 mg/d) 16 (17.2) 55 (29.1) 1.005 1.003 – 1.007
4th quartile (>271 mg/d) 40 (43.0) 30 (15.9)
1.00–1.01), respectively, compared to those in the lowest quartile. Dietary fat and fatty acids Table 5 shows percent energy contribution of dietary fat for the Case and Control groups. The table shows that there were no significant differences in the intakes of saturated and unsaturated fatty acids between the cases and controls, except for arachidonic acid (C20:4). The intakes of individual saturated fatty acids (SFAs) were similar in both groups. Lauric (C12:0), palmitic (C16:0) and myristic (C14:0) acids accounted for 43%, 25% and 17% of total fat intake for both the Case and Control groups. The average intake of marine and plant n-3 fatty acids was 1.9 g/d for the Case group and 1.7 g/d for the Control group. Table 6 shows the percentage of fat from food groups containing fat. Fish dishes were the major source of total fat in both groups followed by soy dishes, rice and cereal dishes. The amounts of coconut milk and grated coconut were estimated in various dishes with coconut as the major ingredient. Fish and soy dishes were the major food sources for monounsaturated fatty acids (MUFAs). The Case and Control groups were not different in terms of food sources of MUFAs, PUFAs and n-3/n-6 fatty acid ratio. As expected, fish was almost the only source of long chain n-3 in the present study - it accounted for more than 90% of n-3 fatty acid intake in both groups. The cases had a significantly higher intake of arachidonic acid (C20:4) due to the higher consumption of meat and eggs. The proportion of food sources for n-6 fatty acids was similar in both groups. Egg, fish and beef were the most important sources of food for arachidonic acids, while soy, eggs and beef were the most important food sources for linoleic acid. Multivariate analysis To determine the predictive power of food and nutrient variables for coronary events, regression of food and nutrient variables was performed. Results are presented in Table 7. Included in this model were food variables, such as animal food intake, and nutrient variables such as total energy, total protein, carbohydrate and cholesterol intake. In addition, other aspects of diet, especially saturated fat intake, and lifestyle factors such as physical activity and stress level, are known to be related to CHD events. These variables were included in the regression analysis.Presented in this table are the partial correlation coefficients for the variables included in the models. Table 7 shows the odds ratio and 95% CI of CHD by food and nutrient intakes. For the total population, higher intake of carbohydrate, higher physical activity, lower animal intake and stress level, were protective against CHD. For men, carbohydrate intake was no longer protective, but lower animal intake and stress, and higher physical activity, were protective. For women, total carbohydrate intake, animal intake and stress levels were predictors for CHD. Discussion One major concern in estimating food intake in a case control study is the strong influence of current diet on recall of previous diet. In a case control study by Willet et al.,14 cases tended to over or underreport past dietary practices. However, the author suggested that diet may be recalled with acceptable levels up to approximately 10 years; beyond this period greater uncertainty exists. The incidence of CHD in this population could not be explained by the introduction of Western foods, because the influence of Western food in this study population was nearly nil. Although the cases had significantly higher intakes of margarine (1.5 teaspoon/d) compared to the controls (0.7 teaspoon/d), the intake of dairy foods (other than milk) was nil and fast food was 0.7 g/d. In contrast, a study in Japan suggested an association between the increase of Western style fat-rich foods such as butter and margarine, cheese, bread and ham & sausage with an increase of mortality from degenerative diseases.15
Total fat intake The present study did not find any significant differences in dietary fat intake between the Case and Control groups, and there was no relationship with CHD. Although other studies have found positive correlation with CHD,1 total fat intake was not always associated with CHD risk or mortality. Reviewing a wide range of studies, there is little evidence that a high intake of dietary fat predisposes to CHD.16-19 However, most of such studies were done in populations with high fat intakes. Populations with low fat intakes tended to be at low risk of CHD,20 and have lower plasma lipid concentrations.21 The increased intake of saturated fat in different populations has been confirmed to be associated with the
381 Dietary intake and the risk of coronary heart disease among the Minangkabau in West Sumatra
elevation of serum cholesterol concentrations and CHD mortality.22-23 It has been suggested that a high intake of coconut oil may contribute to this relationship.24-25 In contrast, when studies were conducted within popu-lations,26 it was not always possible to demonstrate signi- ficant relationships between intake of saturated fat and the incidence of CHD, as it was shown in the present study. Likewise, it has been difficult to demonstrate significant relationships between saturated fat intake and serum lipid levels in an observational study.27
There are several possible explanations for this, such as insufficient precision of the methods used for dietary surveys, large intra-individual variations in food intake, genetic variation with low and high responders to changes in dietary fat, or a small and heterogenous sample with regard to age and gender. Another possibility is that relationships between dietary fat intake and serum lipid levels are more complex than that has been realised hitherto. For example, the source of saturated fatty acids e.g from plant or animal may be important and have
Table 5. % Energy contribution of dietary fat for the Case and Control groups
Total Men Women % Energy
Mean ± SD Mean ± SD Mean ± SD Cases
Total dietary fat 23.6 ± 5.6 23.1 ± 5.5 24.6 ± 4.8 Saturated fatty acids 14.3 ± 3.6 14.0 ± 2.3 14.5 ± 3.3 Short chain fatty acids 1.0 ± 0.3 0.9 ± 0.3 1.0 ± 0.3 Capric acid (C10:0) 0.8 ± 0.2 0.7 ± 0.2 0.8 ± 0.2 Lauric acid (C12:0) 6.1 ± 1.7 5.8 ± 1.5 6.3 ± 1.6 Myristic acid (C14:0) 2.5 ± 0.6 2.4 ± 0.6 2.5 ± 0.6 Palmitic acid (C16:0) 3.5 ± 0.9 3.5 ± 0.9 3.5 ± 0.9 Stearic acid (C18:0) 1.5 ± 0.4 1.5 ± 0.4 1.4 ± 0.4 Monounsaturated fatty acids 5.3 ± 1.5 5.2 ± 1.5 5.3 ± 1.5 Palmitoleic (C 16:1) 0.3 ± 0.1 0.3 ± 0.1 0.3 ± 0.1 Oleic acid (C 18:1) 5.1 ± 1.5 5.1 ± 1.5 5.2 ± 1.6 Polyunsaturated fatty acids 4.4 ± 1.7 4.0 ± 1.4 4.8 ± 1.5 n-6 fatty acids 3.4 ± 1.5 3.1 ± 1.3 3.8 ± 1.4 Linoleic acid (C18:2) 3.3 ± 1.5 3.0 ± 1.3 3.7 ± 1.4 Arachidonic acid (C20:4) 0.06 ± 0.03**** 0.08 ± 0.03**** 0.07 ± 0.03* n-3 fatty acids 0.7 ± 0.3 1.0 ± 0.3 1.1 ± 0.3 α-linolenic (C18:3) 0.3 ± 0.0 0.3 ± 0.1 0.4 ± 0.2 Eicosapentaenoic acid (C20:5, EPA) 0.2 ± 0.1 0.2 ± 0.1 0.2 ± 0.1 Docosahexaenoic acid (C22:6, DHA) 0.5 ± 0.2 0.5 ± 0.2 0.5 ± 0.2 n-3/n-6 fatty acid ratio 0.3 ± 0.2 0.3 ± 0.2 0.3 ± 0.1 P:S ratio 0.3 ± 0.1 0.3 ± 0.1 0.3 ± 0.1
Controls Total dietary fat 23.4. ± 4.9 23.2 ± 4.8 23.6 ± 5.1 Saturated fatty acids 13.9 ± 3.0 13.8 ± 2.9 14.1 ± 3.1 Short chain fatty acids 1.0 ± 0.3 1.0 ± 0.2 1.0 ± 0.3 Capric acid (C10:0) 0.8 ± 0.2 0.8 ± 0.2 0.8 ± 0.2 Lauric acid (C12:0) 6.0 ± 1.4 6.0 ± 1.4 6.0 ± 1.4 Myristic acid (C14:0) 2.4 ± 0.5 2.4 ± 0.5 2.5 ± 0.6 Palmitic acid (C16:0) 3.5 ± 0.8 3.4 ± 0.7 3.6 ± 0.8 Stearic acid (C18:0) 1.4 ± 0.3 1.4 ± 0.3 1.4 ± 0.4 Monounsaturated fatty acids 5.1 ± 1.1 5.0 ± 1.1 5.1 ± 1.1 Palmitoleic (C16:1) 0.3 ± 0.1 0.3 ± 0.1 0.3 ± 0.1 Oleic (C18:1) 5.0 ± 1.2 5.0 ± 1.2 5.1 ± 1.2 Polyunsaturated fatty acids 4.4 ± 1.4 4.3 ± 1.3 4.4 ± 1.4 n-6 fatty acids 3.5 ± 1.3 3.4 ± 1.3 3.6 ± 1.3 Linoleic acid (C18:2) 3.4 ± 1.3 3.4 ± 1.3 3.5 ± 1.3 Arachidonic acid (C20:4) 0.06 ± 0.02 0.06 ± 0.02 0.06 ± 0.02 n-3 fatty acids 0.9 ± 0.2 0.9 ± 0.3 0.9 ± 0.3 α-linolenic (C18:3) 0.4 ± 0.2 0.3 ± 0.1 0.4 ± 0.1 Eicosapentaenoic acid (C20:5, EPA) 0.1 ± 0.1 0.1 ± 0.1 0.1 ± 0.0 Docosahexaenoic acid (C22:6, DHA) 0.4 ± 0.2 0.4 ± 0.2 0.4 ± 0.1 n-3/n-6 fatty acid ratio 0.3 ± 0.1 0.3 ± 0.2 0.3 ± 0.1 P:S ratio 0.3 ± 0.1 0.3 ± 0.1 0.3 ± 0.1
Significantly different from the Control group: *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001.
NI Lipoeto, Z Agus, F Oenzil, ML Wahlqvist and N Wattanapenpaiboon 382
differing effects on CHD risk. Subjects in the Case group of the present study were recruited from a well diagnosed group of CHD patients from several hospitals in West Sumatra, and those in the Control group were recruited carefully to match the age and gender of the Cases. The use of validated FFQ may have helped identify similarities between the Case and Control groups in terms of total fat intake, contribution to total energy intake by total fat and individual fatty acids.
Fatty acid intakes Only limited information is available regarding the association between individual SFAs and the risk of CHD. Experimental studies have found that different classes of SFAs have different effects on plasma lipid and lipoprotein concentrations.28-29 The differential effects of specific saturated fats on plasma lipids and lipoproteins imply that these fats may have different effects on CHD risk.3 In the present study, both the cases and the controls had a similar total SFA intake of about 27 g/d, equivalent to 129 g coconut milk or 31.5 g coconut oil. The results do not support an association between total SFA intake and CHD events. Moreover, this finding further suggests that the consumption of coconut products, especially coconut milk or coconut oil, which were the main source of SFAs in the Minangkabau cuisine, does not increase CHD risk, at least in this study population. Results from several studies indicate that diets high in monounsaturated fatty acids have a more favourable effect on serum lipo-proteins,30 and are cardio-protective.31 However, in this study, no association was found between monounsa-turated and polyunsaturated fat and CHD events, although
the cases had a significantly higher intake of arachidonic acid than the controls. A similar study was reported with coronary angiographic findings.32
Total energy, protein and dietary cholesterol intakes The present study provides no evidence that total energy intake has implications for CHD, although other studies have found a positive relationship between total energy intake and CHD.1 A few case-control studies examined the association of CHD with protein intake. Smit et al., (1999) found a significant positive association between CHD with protein intake.4 Analyses of the association between protein intake and CHD risk are difficult to inter-pret because they involved simple comparisons of means between cases and non-cases without adjustment for intakes of specific types of dietary fatty acids. In the present study, in univariate analysis, the odds ratio of subjects with total protein intake in the top 25% compared to the lowest 25% was 3.2 (95% CI 1.6–6.6). But after adjusting for other risk factors, total protein failed to enter the model, suggesting that it was not a risk factor for CHD. Dietary cholesterol intake was associated with an increased risk of CHD in some studies but not others.33-34
In the present study, the intake of cholesterol was positively associated with CHD in the total population, especially in men, but not in women. In univariate analysis, the odds ratio of subjects with cholesterol intake in the highest quartile, compared to the lowest quartile, was 4.7 (95% CI 2.3–9.7). Dietary cholesterol has been reported to increase liver cholesterol synthesis, resulting in down-regulation of LDL receptor concentrations.35 The
Table 6. Percentage of total fat from food groups
Cases Controls Food groups Mean ± SD Mean ± SD
Fish, seafood and products 26.9 ± 10.7 24.0 ± 11.6 Soy products 18.1 ± 11.9 ** 22.1 ± 12.5 Cereal and cereal products 11.8 ± 8.3 12.9 ± 7.5 Meats and meat products 11.5 ± 7.2 ** 9.2 ± 6.4 Vegetables 9.8 ± 4.5 10.3 ± 4.3 Eggs 10.2 ± 7.3 9.0 ± 5.3 Nuts and seeds 4.4 ± 5.1 4.4 ± 4.4 Milk and milk products 4.1 ± 8.7 4.5 ± 9.3 Fruits 2.8 ± 2.4 2.8 ± 1.9 Confectionary 0.2 ± 0.7 0.2 ± 0.7 Non alcoholic beverages 0.4 ± 1.2 0.7 ± 1.7 Significantly different from the Control group: **, P<0.01.
Table 7. Odds Ratio (95% confidence interval) of coronary events by food and nutrient variables
Total Men Women Total Carbohydrate (highest vs lowest quartile)
0.7 (0.36 – 1.47)
NA 0.98 ** (0.97 – 0.99)
Animal food intake (highest vs lowest quartile)
4.8 **** (2.25 – 10.30)
5.6 *** (1.99 – 16.89)
4.7 * (1.28 – 16.98)
Physical activity (highest vs lowest quartile)
0.4 ** (0.2 – 0.8)
0.3 * (0.1 – 0.7)
NA
Stress level (highest vs lowest quartile)
2.9 ** (1.6–6.5)
2.8 * (1.2 – 6.3)
3.6 ** (1.3 – 10.3)
Smoke (highest vs lowest quartile)
NA NA 0.2 ** (0.04 – 0.7)
Variables entered into model include the intakes of animal food, carbohydrate, protein, cholesterol, saturated fat, physical activity, stress level and smoking status; NA: data not available. The variable was removed from the model (the significance level did not reach 0.15). Significantly different from the odds ratio 1.0: *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001.
383 Dietary intake and the risk of coronary heart disease among the Minangkabau in West Sumatra
background quality of dietary intake, bound in Minang-kabau people, as with other population, effects serum cholesterol and lipoprotein status. Food intakes In general, the intakes of several food groups were different between the cases and the controls. Only total animal intake was found to be independently correlated with CHD events. A higher fish intake (which was also associated with higher n-3 fatty acids), has been recognised to reduce very-low-density lipoproteins, inhibit thromboxane production, increase prostacyclin synthesis, reduce the likelihood of thrombosis, risk of cardiac arrhythmias, and blood viscosity.10,36 Other populations with higher fish intake, such as the Eskimos and the Japanese, have long been recognised to have a low rate of CHD.37-38 However, in the present study, it was observed that both the Case and Control groups had a high intake of n-3 fatty acids. Some studies conducted in North America and European countries, where the fish consumption was lower than in the present study, had also found no correlation between fish intake and CHD incidence.26,39 In this population, coconut (a saturated fat source) is used in the cooking of fish and vegetables. Hence, any potentially adverse effects of coconut-derived saturated fat may be offset by the cardio-protective role of the coconut-associated fish and vegetable intakes in the Minangkabau food culture. Conclusions Two conclusions can be drawn from this study. Firstly, intakes of Western foods in this study population were minimal, so that the incidence of CHD cannot possibly be explained by the introduction of Western foods. Secondly, the results from this study showed that the intakes of total fat and saturated fat were not associated with CHD events. Thus, CHD could not be predicted by saturated fat intake in this population. In contrast, intakes of animal food, total protein, dietary cholesterol and total carbohydrate were found to be predictors of CAD. References 1. Tzonou A, Kalandidi A, Trichopoulou A, Hsieh CC,
Toupadaki N, Willett W, Trichopoulou D. Diet and coronary heart disease: a case-control study in Athens, Greece. Epidemiology 1993; 4: 511-516.
2. Shekelle RB, Shryock AM, Paul O, Lepper M, Stamler J, Liu S, Raynor WJ. Diet, serum cholesterol and death from coronary heart disease. The Western Electric study. N Engl J Med 1981; 304: 65-70.
3. Hu FB, Stampfer MJ, Manson JE, Ascherio A, Colditz GA, Speizer FE, Hennekens CH, Willett WC. Dietary saturated fats and their food sources in relation to the risk of coronary heart disease in women. Am J Clin Nutr 1999; 70: 1001-1008.
4. Smit E, Nieto FJ, Crespo CJ. Blood cholesterol and apolipoprotein B levels in relation to intakes of animal and plant proteins in US adults. Br J Nutr 1999; 82: 193-201.
5. Keys A. Serum cholesterol response to dietary cholesterol. Am J Clin Nutr 1986; 44: 309.
6. Barr SL, Ramakrishnan R, Johnson C, Holleran S, Dell RB, Ginsberg HN. Reducing total dietary fat without reducing saturated fatty acids does not significantly lower total plasma cholesterol concentrations in normal males. Am J Clin Nutr 1992; 55: 675-681.
7. Kromhout D, Bloemberg B, Feskens E, Menotti A, Nissinen A. Saturated fat, vitamin C and smoking predict long-term population all-cause mortality rates in the Seven Countries Study. Int J Epidemiol 2000; 29: 260-265.
8. Lindeberg S, Lundh B. Apparent absence of stroke and ischaemic heart disease in a traditional Melanesian island: a clinical study in Kitava. J Intern Med 1993; 233:269-275.
9. Kumar PD. The role of coconut and coconut oil in coronary heart disease in Kerala, south India. Tropical Doctor 1997; 27: 215-217.
10. Krauss RM, Eckel RH, Howard B, Appel LJ, Daniels SR, Deckelbaum RJ, Erdman JW Jr, Kris-Etherton P, Goldberg IJ, Kotchen TA, Lichtenstein AH, Mitch WE, Mullis R, Robinson K, Wylie-Rosett J, St Jeor S, Suttie J, Tribble DL, Bazzarre TL AHA Dietary Guidelines: revision 2000: A statement for healthcare professionals from the Nutrition Committee of the American Heart Association. Circulation 2000; 102: 2284-2299.
11. Singh RB, Niaz MA, Ghosh S, Beegom R, Agarwal P, Nangia S, Moshiri M, Janus ED. Low fat intake and coronary artery disease in a population with higher prevalence of coronary artery disease: the Indian paradox. J Am Coll Nutr 1998; 17: 342-350.
12. Rose GA, Blackburn H, Gillum RF, Prineas RJ. Cardiovascular Survey Methods, 2nd Edition, Monograph Series No. 56, Geneva, World Health Organization, 1982
13. Mukrie NA, Chatidjah S, Masoara S, Alhabsji S. Indonesian Food and Nutrinet Composition. Jakarta, Ministry of Health, 1995.
14. Willet WC. Diet and Coronary Heart Disease. In: Willett WC, ed. Nutritional Epidemiology. 2nd Edition, New York: Oxford University Press, 1998: 148-156.
15. Kato I, Tominaga S, Kuroishi T. Relationship between westernization of dietary habits and mortality from breast and ovarian cancers in Japan. Japanese J Cancer Res 1987; 78: 349-357.
16. Grundy SM, Grundy S.M. The optimal ratio of fat-to- carbohydrate in the diet. Annu Rev Nutr 1999; 19: 325-341.
17. Hooper L, Summerbell CD, Higgins JP, Thompson RL, Capps NE, Smith GD, Riemersma RA, Ebrahim S. Dietary fat intake and prevention of cardiovascular disease: systematic review. BMJ 2001; 7289: 757-763.
18. Hodgson JM, Wahlqvist ML, Hsu-Hage B. Diet hyper-lipidaemia and cardiovascular disease. Asia Pac J Clin Nutr 1995; 4: 304-313.
19. Wahlqvist ML, Yamori Y. The Okinawan Round-table on Nutritional Cardiovascular Disease. Summary and recommendations. Asia Pac J Clin Nutr 2001; 10: 172.
20. Dreon DM, Fernstrom HA, Williams PT, Krauss RM. A very low-fat diet is not associated with improved lipoprotein profiles in men with a predominance of large, low-density lipoproteins. Am J Clin Nutr 1999; 69: 411-418.
21. Raeini-Sarjaz,M, Vanstone CA, Papamandjaris AA, Wykes LJ, Jones PJ. Comparison of the effect of dietary fat restriction with that of energy restriction on human lipid metabolism. Am J Clin Nutr 2001; 73:262-267.
22. Barr SL, Ramakrishnan R, Johnson C, Holleran S, Dell RB, Ginsberg HN. Reducing total dietary fat without reducing saturated fatty acids does not significantly lower total plasma cholesterol concentrations in normal males. Am J Clin Nutr 1992; 55: 675-681.
23. Kromhout D, Bloemberg B, Feskens E, Menotti A, Nissinen A. Saturated fat, vitamin C and smoking predict long-term population all-cause mortality rates in the Seven Countries Study. Int J Epidemiol 2000; 29: 260-265.
NI Lipoeto, Z Agus, F Oenzil, ML Wahlqvist and N Wattanapenpaiboon 384
24. Pronczuk A, Patton G.M, Stephan ZF, Hayes KC. Species variation in the atherogenic profile of monkeys: relationship between dietary fats, lipoproteins, and platelet aggregation. Lipids 1991; 26: 213-222.
25. Anderson JT, Grande F, Keys A. Independence of the effects of cholesterol and degree of saturation of the fat in the diet on serum cholesterol in man. Am J Clin Nutr 1976; 29: 1184-1189.
26. Ascherio A, Rimm EB, Stampfer MJ, Giovannucci EL, Willett WC. Dietary intake of marine n-3 fatty acids, fish intake, and the risk of coronary disease among men. N Engl J Med 1995; 332: 977-982.
27. Samuelson G, Bratteby LE, Mohsen R, Vessby B. Dietary fat intake in healthy adolescents: inverse relationships between the estimated intake of saturated fatty acids and serum cholesterol. Br J Nutr 2001; 85: 333-341.
28. Nicolosi RJ. Dietary fat saturation effects on low-density-lipoprotein concentrations and metabolism in various animal models. Am J Clin Nutr 1997; 65 (5 Supl): 1617S-1627S.
29. Kelly FD, Sinclair AJ, Mann NJ, Turner AH, Abedin L, Li D. A stearic acid-rich diet improves thrombogenic and atherogenic risk factor profiles in healthy males. Eur J Clin Nutr 2001; 55: 88-96.
30. Mensink RP, Temme EH, Hornstra G. Dietary saturated and trans fatty acids and lipoprotein metabolism. Ann Med 1994; 26: 461-464.
31. de Lorgeril M, Salen P, Martin JL, Monjaud I, Boucher P, Mamelle N. Mediterranean dietary pattern in a randomized trial: prolonged survival and possible reduced cancer rate. Arch Internal Med 1998; 158:1181-1187.
32. Hodgson JM, Wahlqvist ML, Boxall JA, Balazs NDH. Can linoleic acid contribute to coronary artery disease? Am J Clin Nutr 1993; 58:228-234.
33. Simkin-Silverman L, Wing RR, Hansen DH, Klem ML, Pasagian-Macaulay AP, Meilahn EN, Kuller LH. Prevention of cardiovascular risk factor elevations in healthy premenopausal women. Prev Med 1995; 24:509-517.
34. Addis PB, Carr TP, Hassel CA, Huang ZZ, Warner GJ. Atherogenic and anti-atherogenic factors in the human diet. Biochemical Society Symposia 1995; 61: 259-271.
35. Schaefer EJ. Effects of dietary fatty acids on lipoproteins and cardiovascular disease risk. Am J Clin Nutr 1997; 65 (5 Suppl): 1655S-1666S.
36. Agren JJ, Vaisanen S, Hanninen O, Muller AD, Hornstra G. Hemostatic factors and platelet aggregation after a fish-enriched diet or fish oil or docosahexaenoic acid supple-mentation. Prostaglandins Leukot Essent Fatty Acids 1997; 57: 419-421.
37. Kromann N, Green A. Epidemiological studies in the Upernavik district, Greenland. Incidence of some chronic diseases 1950-1974. Acta Medica Scand 1980; 208: 401-406.
38. Hirai A, Hamazaki T, Terano T, Nishikawa T, Tamura Y, Kamugai A, Jajiki J. Eicosapentaenoic acid and platelet function in Japanese. Lancet 1980; 2 (8204): 1132-1133.
39. Kromhout D, Bosschieter EB, de LezenneCoulander C. The inverse relation between fish consumption and 20-year mortality from coronary heart disease. N Engl J Med 1995; 312: 1205-1209.
385 Asia Pac J Clin Nutr 2004;13 (4): 385-390
Original Article Nutritional analysis of blenderized enteral diets in the Philippines Mary M Sullivan MPH, RD1, Pearl Sorreda-Esguerra RND2, Maria Bernadette Platon RND3, Cynthia G Castro RND4, Nancy R Chou, MS, RD5, Susan Shott PhD6, Gail M Comer MD5 and Pedro Alarcon MD, PhD5 1 Takeda Pharmaceuticals North America, Lincolnshire, IL, USA 2 Philippine Heart Center, Manila, Philippines 3 University of Santo Tomas Hospital, Manila, Philippines 4 Manila Sanitarium, Manila, Philippines 5 Abbott Laboratories, Abbott Park, IL, USA 6 Rush University, Chicago, IL, USA
The objective of this study was to analyze the nutritional quality and viscosity of blenderized enteral tube feedings (BTFs) from four hospitals in the Philippines. Samples of two different BTFs (one standard and one modified) were collected from each hospital on three separate occasions and analyzed for macronutrients, micronutrients, and viscosity. There was considerable variation among the BTFs for the concentrations of most nutrients measured. For standard BTF samples, the caloric density ranged from 66-123 kcal/100g and the percentages of total weight for protein, carbohydrate, and fat ranged from 1.5-4.0%, 8.6-21.4%, and 0.27-3.40%, respectively. Levels of specific vitamins were undetectable in 10 standard and 15 modified BTF samples. In samples where vitamin levels were detectable, results were: vitamin A, 625-8850 IU/kg; riboflavin, 0.40-5.00 mg/kg; and pyridoxine, 0.14-3.00 mg/kg. Mineral concentrations also varied greatly (eg calcium, 64-524 mg/kg; sodium, 148-886 mg/kg; iron, 3.0-13.7 mg/kg; and zinc, 1.8- 11.5 mg/kg). Correlation coefficients were statis-tically significant only for carbohydrate (r = 0.48, P = 0.017). Measured values tended to be lower than expected values for all nutrients, although the difference was statistically significant only for calories (P = 0.023). The viscosity of BTF samples ranged from 2.3-45,060 centipoise, excluding three samples that were too viscous for analysis. This study demonstrates that hospital prepared blenderized enteral tube feedings render unpredictable levels of micronutrients and macronutrients and appear likely to deliver less than the desired amounts of nutrients. Additionally, the viscosity of these feedings may be unsuitable for infusion through feeding tubes.
Key Words: blenderized enteral feedings, caloric density, vitamin content, mineral content, viscosity, Philippines. Introduction Enteral tube feedings are commonly used in hospitals to provide nutritional support. While commercial, ready-to-use formulas have been available for over 20 years, many institutions prefer the use of “homemade”, blenderized tube feedings (BTF). This preference may result from believing them to be more “natural” (physiologic) or more eco-nomical. Blenderized tube feedings typically contain common foodstuffs such as milk, eggs, meat, soft fruits, and vegetables that are pureed in a food blender or mixer. Other BTFs are made from a base of a commercial nutri-tional powder, which is reconstituted with water or other liquid. To this base, other foods may be added to modify the consistency or nutritional composition. While BTFs appear to permit flexibility with regard to the selection of ingredients, and therefore nutritional content, problems with their use have been reported. Gallagher-Allred analyzed prepared BTFs for nutritional content, osmolality, and bacterial contamination.1 An insti-tutionally prepared "high calorie" formula expected to deli-ver 1.5 kcal/mL yielded only 1.0 kcal/mL on analysis. In
addition, this "high calorie" formula did not meet the US Recommended Dietary Allowances (US RDA) for vitamin B12, biotin, iron, and copper in 3,000 kcal. By contrast, commercial feedings designated to provide 1.0 kcal/mL and 1.5 kcal/mL met all nutrient standards and provided the expected caloric density. A similar study was conducted in the Philippines in which 17 hospitals were randomly se-lected to provide blenderized diets for analysis.2 Measured calories were consistently much lower than expected values. In practice, this would result in patients receiving a lower intake of energy and micro-nutrients than prescribed. In Gallagher-Allred's study, the propensity for tube occlu-sion was determined by allowing tube-feeding formula to drip unrestricted through a size 8 or 12 French nasogastric feeding tube. Milk-based and commercial feedings flowed unaided through a size 12 French feeding tube and Corresponding author: Dr Pedro Alarcon, AP30-2, D-06NR, 200 Abbott Park Rd, Abbott Park, IL 60064-6149, USA Tel: (847) 938-5577; Fax: 847-938-8355 Email: [email protected] Accepted 11 May 2004
MM Sullivan, P Sorreda-Esguerra, MB Platon, CG Castro, NR Chou, S Shott, GM Comer and P Alarcon 386
feeding tube and delivered 240 mL in less than 30 minutes. The institutionally prepared "high calorie" for-mula did not flow through the size 12 tube. Milk-based and commercial formulas were also able to flow unaided through a size 8 French nasogastric tube, but the blen-derized formula could not. There are no published standards for tube feeding viscosity; however, tube feeding formulas should be expected to flow through a small bore feeding tube at the desired infusion rate without causing tube occlusion. Commercial fiber-containing feedings, which are known to flow unaided through small bore (8 French) feeding tubes, typically have a viscosity of less than 60 centipoise (cps). Tube occlusion may result from inadequate flushing, addition of medication to the feeding, or co-agulation of the protein moiety by gastric acid.3 Tube occlusion has been cited as a reason why enteral tube feedings may not meet patient's energy requirements.4 The purpose of this study was to evaluate the nutritional content and viscosity of hospital-prepared blenderized enteral tube feedings in the Philippines. Materials and Methods Four hospitals in Manila, Philippines were selected for participation in the study. The participating hospitals all used BTFs as a standard of practice for their enteral tube fed patients. Each hospital submitted two different tube feeding recipes; one representing a "standard" or general diet and the other a “modified” or therapeutic (eg dia-betic, sodium-restricted, anti-diarrheal) diet of the hospi-tal's choice. The providers of the recipes believed them to be nutritionally complete, providing all essential nutri-ents. Supplies used for the collection of the tube feeding samples (sterile containers, dry ice, cooler) were provided by SGS Philippines, Inc. (Makati City, Philippines). Each hospital prepared at least one liter of both BTF recipes on three separate occasions with an SGS technician present. The BTFs were prepared by the same personnel and with the same procedures used in the preparation of tube feedings for patients. Immediately following the pre-paration of all tube feedings, the SGS technician took a 600 mL aliquot of each feeding and divided it into three sterile, Nalgene bottles using clean technique. An addi-tional 10 mL was poured into a sterile, Nalgene bottle for microbial analysis. Results of microbial analysis have been reported elsewhere.5 Samples were sealed, labelled and immediately placed in a cooler with dry ice. Samples
were transported from the hospital to the laboratory on dry ice and transferred to a -70oC freezer within ten hours of collection. SGS Philippines analyzed the samples for viscosity and the following nutritional components: vitamins (A, riboflavin, and pyridoxine), minerals (cal-cium, magnesium, phosphorus, sodium, iron, potassium, and zinc), cholesterol, saturated fatty acids, caloric density, and the percentage (by weight) of carbohydrate, fat, and protein. The laboratory used for this study lacked methods sensitive enough to reliably detect low levels of the following vitamins and minerals: vitamins B12, C, D, E, and K, thiamin, niacin, folic acid, selenium, and iodine. Therefore, these vitamins and minerals were not con-sidered for analysis. Methods of analysis are shown in Table 1. The dietitians from each institution determined the “expected” nutrient content of each recipe, which was derived from the recipe ingredients. If this was not available, the recipe was analyzed using nutritional ana-lysis software (Nutritionist V, First Data Bank, San Bruno, CA, 1998) and Philippine Food Composition Tables.6 The “expected” nutritional content was com-pared with the measured nutritional content as determined from the laboratory analyses. The recipes were generally prepared from blended foods such as meat, fruit, and vegetables. However, two hospitals used a commercial powder formula as a base to which water alone or water and fruit were added: Hospital A, standard and modified feedings; Hospital B, standard feeding. Recipes for all feedings are shown in Table 2. Statistical methods All statistical tests were two-sided with a 0.05 signi-ficance level. Statistical analyses were performed sepa-rately for each hospital and for the combined data from all hospitals. For each BTF, descriptive statistics were cal-culated for nutritional parameters and viscosity. Paired statistical tests were used to determine whether the measured nutrient levels differed from the expected levels. Paired t-tests were used to compare results when the differences showed a normal or approximately normal distribution, and the nonparametric paired sign test was used when the differences showed an extremely non-normal distribution. Correlation coefficients were obtained for measured versus expected nutrient content to determine how close these values were. Pearson correlation coefficients were
Table 1. Methods of analysis*
Laboratory Test Method Water Soluble Vitamins:
Pyridoxine, riboflavin High performance liquid chromatography
Minerals Sodium, potassium, magnesium, calcium, iron, zinc
Atomic absorption spectrophotometry
Phosphorus Spectrophotometry Macronutrients
Total fat Sohxlet method
Saturated fat, cholesterol Gas chromatography Protein Kjeltic automatic protein analyzer Carbohydrate, calories Computation
Viscosity Viscosimeter * References for analysis of vitamins and minerals: pyridoxine, riboflavin-HPLC (In-house method); minerals-AOAC 985.35; saturated fat, cholesterol-AOAC 1995
387 Nutritional analysis of blenderized enteral diets in the Philippines
used to test the hypothesis of zero correlation when the measured nutrient levels had a normal or approximately normal distribution. Spearman correlation coefficients were used when the measured nutrient levels had an extremely non-normal distribution. SPSS for Windows (version 8) was used for data management and statistical analyses. Results There was a high degree of variability in the concen-trations of most nutrients measured both among samples from a single hospital and among samples from different hospitals. In addition, vitamin levels were not detectable in all samples. Descriptive statistics for all nutrients include only those samples with detectable levels. The measured nutrient concentrations of the standard and modified feedings from each hospital are presented in Tables 3 and 4, respectively. At each hospital, the values for caloric density and the percentages (by weight) of carbohydrate, fat, and protein were generally within ± 20%-30% of the mean value at that hospital. However, variability for cholesterol concentrations was much greater, particularly for standard feedings prepared at Hospital D (range 14-172 mg/kg) and modified feedings prepared at Hospital B (range 4-198 mg/kg). For standard feedings, the mean caloric density between hospitals ranged from 80.9 - 106.5 kcal/100g and the mean per-centages (by weight) of protein, carbohydrate, and fat ranged from 2.13% - 3.63%, 12.0% - 18.7%, and 1.63% - 2.57%, respectively. Vitamin levels were undetectable in many samples. For standard BTFs, vitamin A and riboflavin were de-tectable in 9 of the 12 samples, and pyridoxine was
detectable in 8 of the 12 samples. For modified BTFs, pyridoxine was detectable in 9 of 12 samples, vitamin A in 7 of 12 samples, and riboflavin in 5 of 12 samples. Concentrations of all vitamins measured varied widely, particularly for feedings prepared at Hospital B (vitamin A, 2250 - 8075 IU/kg; riboflavin, 1.10 - 5.00 mg/kg; pyridoxine, 0.30 - 3.00 mg/kg). Variability for mineral concentrations was generally within ± 20% of the mean value at each hospital. How-ever, for standard feedings, sodium and calcium concen-trations ranged from 148 - 389 mg/kg and 64 - 204 mg/kg, respectively, at Hospital D, and zinc concen-trations ranged from 3.6 - 11.5 mg/kg at Hospital A. For modified feedings, sodium and potassium concentrations ranged from 144 - 404 mg/kg and 423 - 1242 mg/kg, respectively, at Hospital D, and zinc concentrations ranged from 2.0 - 9.3 mg/kg at Hospital A. Mean concen-trations of minerals also varied considerably between hospitals for standard feedings: calcium, 139 - 467 mg/kg; phosphorus, 293 - 499 mg/kg; iron, 3.4 - 11.2 mg/kg; sodium, 280 - 679 mg/kg; potassium, 757 - 1095 mg/kg; and zinc, 2.8 - 8.6 mg/kg. Mean concentrations of most minerals measured in standard and modified feedings were generally higher when prepared from powdered formula; i.e standard feeding Hospital A and B, modified feeding Hospital A. There were marked discrepancies between the measured and expected values for calories and percentage of carbohydrate, fat, and protein in both the standard and modified feedings (Tables 5 and 6). There was no overall correlation between measured and expected nutrient concentrations. When all feedings were analyzed to-gether, the Pearson correlation coefficients were almost
Table 2. Recipes for blenderized enteral tube feedings
Hospital A Hospital B Hospital C Hospital D Standard Feedings
Ensure ® powder†, 333 g Ensure ® powder†, 265 g Squash, 135 g Banana, 4 whole peeled Tap water, 1350 mL Tap water, 960 mL Banana, 80 g White bread, 5 slices Nonfat dry milk, 17 g Lugao*, 240 mL White bread, 150 g Egg, cooked, 1 Corn oil, 26 mL Corn oil, 7.5 mL Chicken breast, 67.5 g White sugar, 4.2 g Lugao*, 360 mL 1500 mL Total 1200 mL Total 1000 mL Total Total Volume NA
Modified Feedings (Constipating Diet) (Natural Formula Diet) (High Fibre Low Cholesterol Diet) (Diabetic Diet) Ensure® powder†, 289 g Squash, 245 g Squash, 180 g Bananas 4.5 whole peeled Banana, 2.5 whole peeled Banana, 5 whole peeled Banana, 120 g White bread, 5 slices Tap water, 1275 mL Egg cooked, 272 g Pineapple juice, 120 mL Egg, cooked, 1 Corn oil, 60 mL Mung beans, 62 g Corn oil, 7.5 mL White sugar, 12.6 g Nonfat milk, 8.5 g Lugao*, 240 mL Egg, cooked 12.5 g White sugar, 16.8 g Oatmeal, 227 g White bread, 110 g Corn oil, 10 mL Olive oil, 12.5 mL 1500 mL Total 1000 mL Total 1000 mL Total Total Volume NA †Ensure®, Abbott Laboratories, Zwolle, Netherlands; * Lugao = rice gruel; NA = not available
MM Sullivan, P Sorreda-Esguerra, MB Platon, CG Castro, NR Chou, S Shott, GM Comer and P Alarcon 388
Table 3. Measured nutrient concentrations of BTF samples: standard feedings
Hospital A Mean + SD (Range)
Hospital B Mean + SD (Range)
Hospital C Mean + SD (Range)
Hospital D Mean + SD (Range)
Macronutrients Calories (kcal/100g) 80.9+9.1 (73.1-90.9) 85.9+14.8 (70.4-99.8) 90.2+21.7 (66.0-108.0) 106.5+19.8 (84.5-123.0) Protein (%)# 2.83+0.60 (2.20-3.40) 3.00+0.44 (2.70-3.50) 3.63+0.55 (3.00-4.00) 2.13+0.65 (1.50-2.80) Carbohydrate (%)# 13.4+2.7 (10.9-16.2) 12.0+2.9 (8.6-13.8) 15.2+3.9 (10.8-18.4) 18.7+3.5 (14.7-21.4) Fat (%)# 1.77+0.47 (1.40-2.30) 2.16+1.66 (0.27-3.40) 1.63+0.40 (1.20-2.00) 2.57+0.58 (1.90-2.90) Saturated fat (%)# 1.13+0.31 (0.80-1.40) 2.08+1.57 (0.34-3.40) 1.13+0.23 (1.00-1.40) 2.53+0.70 (1.80-3.20) Cholesterol (mg/kg) 16.8+8.2 (7.4-21.7) 12.1+7.9 (3.0-17.0) 27.2+22.0 (7.6-51.0) 114+87 (14-172)
Vitamins Vitamin A (IU/kg) % with detectable level$
6967+2581 (4025-8850) 100%
4977+2930 (2250-8075) 100%
1197+602 (625-1825) 100%
ND 0%
Riboflavin (mg/kg) %with detectable level$
3.05+1.20 (2.20-3.90) 67%
3.05+2.76 (1.10-5.00) 67%
2.80+0.28 (2.60-3.00) 67%
0.55+0.21 (0.40-0.70) 100%
Pyridoxine (mg/kg) %with detectable level$
2.30+0.28 (2.10-2.50) 67%
1.90+1.42 (0.30-3.00) 100%
0.14 33%
0.65+0.64 (0.20-1.10) 67%
Minerals Calcium (mg/kg) 467+51 (426-524) 406+18 (388-424) 313+60 (256-376) 139+71 (64-204) Magnesium (mg/kg) 175+23 (154-200) 155+24 (135-181) 120+22 (102-145) 130+38 (104-173) Phosphorus (mg/kg) 454+42 (410-493) 396+22 (372-413) 499+60 (430-539) 293+82 (235-387) Iron (mg/kg) 11.2+2.3 (9.3-13.7) 8.5+1.8 (6.4-9.7) 3.4+0.6 (3.0-4.1) 4.5+1.4 (3.5-6.1) Sodium (mg/kg) 679+204 (478-886) 672+102 (590-786) 405+78 (350-494) 280+122 (148-389) Potassium (mg/kg) 998+141 (862-1143) 1095+183 (884-1210) 757+90(704-860) 822+218 (608-1043) Zinc (mg/kg) % with detectable level$
8.6+4.4 (3.6-11.5) 100%
8.1+0.1 (8.0-8.2) 67%
2.9+1.1 (1.8-4.0) 100%
2.8+0.4 (2.3-3.0) 100%
SD = standard deviation; # Percentages based on weight (g/100g). N = 3 for all hospitals; ND = not detectable in any sample; $ % of samples with detectable level of nutrient
zero for measured versus expected caloric density (correlation coefficient r = -0.056, P =0.80) and measured versus expected percent protein (r = 0.045, P = 0.83). For the measured versus expected percent fat, the Pearson correlation coefficient was negative, but not statistically significant (r =–0.27, P=0.20). A statistically significant, positive Pearson correlation coefficient was observed for measured versus expected percent carbohydrate, but this was only a moderate correlation (r = 0.48, P = 0.017). The nonparametric sign test was used to investigate the tendency for the measured values to be higher or lower than the expected values. When all samples were ana-lyzed together, the measured values tended to be lower than the expected values for all nutrients, but the difference was statistically significant only for calories (P=0.023). P values for the differences between measured and expected values for carbohydrate, fat, and protein were P = 0.064, P = 0.093, and P = 0.093, respectively. Viscosity measurements were obtained for 21 of the 24 samples; three samples were too viscous for analysis. The mean viscosity for the 21 samples was 2,617 cps (median 21.6 cps; range 2.3 - 45,060 cps). The wide range of viscosity values reflected the range of ingredients used for preparation of the feedings. Viscosity was uniformly much lower and more consistent for feedings prepared from powdered formulas and water (usually <10 cps) than for feedings prepared from blenderized whole food ingredients (20 - 45,060 cps). Discussion This study demonstrates that hospital prepared blen-derized enteral tube feedings render unpredictable and inconsistent levels of micronutrients and macronutrients. These feedings were likely to deliver less than the ex-pected amount of nutrients based on the actual recipes.
There may be clinical implications of under-delivering nutrients in at-risk patient groups, such as paediatric patients, and hypermetabolic patients who have higher nutrient needs compared to normal metabolic ones. For example, a burned patient may require 3000 kcal per day. If a formula were selected that unknowingly provided only 2000 kcal/day or only 66% of his caloric require-ments, this deficit could have significant adverse clinical outcomes such as accelerated loss of lean body mass. The variability in nutrient levels observed over three days of BTF preparation was significant. At Hospital B, preparation of the same recipe (modified diet) on three separate days provided a nearly 50-fold range in mea-sured cholesterol content (4 - 198 mg/kg). Variations in the nutrient compositions of hospital prepared enteral feedings have been observed in other studies.1,2 The nutrient composition of feedings prepared from normal foodstuffs depends on the nutrient compositions of the foods used. These compositions can vary according to the geographical source of the food, the season and stage of maturity when the food was harvested, food processing methods, storage conditions, and cooking methods.2 These factors could explain some of the variability in the nutrient composition of the feedings. It is of concern that 11 standard and 17 modified BTF samples did not have detectable levels of specific nutrients: pyridoxine, riboflavin, vitamin A or zinc. The expected amount of each nutrient was derived from the BTF recipes. The result of undetectable levels of specific nutrients in the prepared samples could have been due to inadequate preparation of the recipes (ie substitution or omission of food from the recipe) or loss of the nutrient during storage of the raw foods prior to BTF preparation. While short-term lack of vitamins may be without consequence, long-term deficiency could negatively impact nutritional status.
389 Nutritional analysis of blenderized enteral diets in the Philippines
The mean calcium content for two hospitals was inade-quate. The mean calcium content derived for 1000 kcal was only 90 mg for Hospital B's modified feeding and 131mg for Hospital D’s standard feeding. This is clearly insufficient for all patient groups. While the US Dietary Reference Intakes recommend an intake of 1000 mg calcium per day8 for adults 19-50 years old, needs for some individuals, such as those at risk for osteoporosis, may be as high as 1500 mg/day.9 The mean viscosity (2,617 cps) for the 21 samples analyzed was more than 43 times higher than typical commercial formulas (60 cps). It is likely that some of these samples would not flow easily through nasogastric
or nasoenteric feeding tubes and could occlude these tubes. To prevent tube occlusion from a high viscosity formula, rapid feeding by bolus method or the use of large bore feeding tubes may be required. In general, these methods of feeding are poorly tolerated compared to con-tinuous feeding through a small bore feeding tube. Coben et al., compared lower oesophageal sphincter pressure (LES) in response to a rapid feeding bolus versus con-tinuous feeding in tube fed adults.10 The LES was signi-ficantly lower following the bolus feeding than after the continuous feeding. Relaxation of LES is associated with gastroesophageal reflux.11,12 Gastroesophageal reflux has also been associated with the use of large bore feeding tubes.11
Table 4. Measured nutrient concentrations of BTF samples: modified feedings
Hospital A (Constipating Diet)
Mean + SD (Range)
Hospital B (Natural Formula Diet) Mean + SD (Range)
Hospital C (High Fibre Low Cholesterol Diet)
Mean + SD (Range)
Hospital D (Diabetic Diet)
Mean + SD (Range)
Macronutrients Calories (kcal/100g) 97.8+28.3 (73.9-129.0) 64.8+14.1 (50.7-78.9) 82.2+4.0 (77.7-85.0) 98.5+3.7 (94.6-102.0) Protein (%)# 2.77+0.59(2.10-3.20) 1.17+0.45(0.71-1.60) 2.57+0.21(2.40-2.80) 1.87+0.40(1.40-2.10) Carbohydrate (%)# 16.2+5.7 (11.2-22.4) 11.1+2.6 (8.1-13.2) 14.4+1.1 (13.2-15.4) 16.3+1.6 (15.0-18.1) Fat (%)# 2.43+0.51(2.00-3.00) 1.93+0.67(1.50-2.70) 1.63+0.06(1.60-1.70) 2.87+0.55(2.30-3.40) Saturated fat (%)# 1.37+0.38(1.10-1.80) 1.47+0.23(1.20-1.60) 1.20+0.36(0.90-1.60) 2.80+0.66(2.10-3.40) Cholesterol (mg/kg) 19.9+12.2 (6.0-28.9) 110+98 (4-198) 15.5+5.5 (10.8-21.5) 106+79 (16.2-164)
Vitamins Vitamin A (IU/kg) % with detectable level$
3882+322 (3655-4250) 100%
1832 33%
1372+490 (870-1850) 100%
ND 0%
Riboflavin (mg/kg) % with detectable level$
3.4 33%
ND 0%
5.10+2.12(3.60-6.60) 67%
0.68+0.59(0.27-1.10) 67%
Pyridoxine (mg/kg) % with detectable level$
2.15+0.50(1.80-2.50) 67%
0.30+0.28(0.11-0.50) 67%
0.40+0.14(0.30-0.50) 67%
0.36+0.21(0.20-0.60) 100%
Minerals Calcium (mg/kg) 454+75 (374-524) 58.0+20.8 (34-70) 299+50 (250-350) 141+53 (80-174) Magnesium (mg/kg) 171+22 (148-192) 44.0+3.5 (40-46) 155+32 (124-188) 136+47 (100-190) Phosphorus (mg/kg) 432+81 (348-510) 156+50.1 (102-201) 502+27 (475-530) 310+15 (294-322) Iron (mg/kg) 9.1+1.6 (7.3-10.4) 2.1+1.2 (1.1-3.4) 5.5+1.5 (3.8-6.6) 5.1+2.1 (3.2-7.4) Sodium (mg/kg) 523.3+62.3 (456-579) 187.3+41.8 (140-219) 369+90.7 (271-450) 261+132 (144-404) Potassium (mg/kg) 1098+222(901-1339) 268+80(176-320) 1203+234(1035-1470) 776+421(423-1242) Zinc (mg/kg)
% with detectable level$ 6.0+3.7 (2.0-9.3)
100% 1.6
33% 3.5+0.6 (2.8-4.0)
100% 2.6+1.0 (1.6-3.5)
100% SD = standard deviation; # Percentages based on weight (g/100g). N = 3 for all hospitals; $ % of samples with detectable level of nutrient; ND = not detectable in any sample
Table 5. Comparison of measured and expected nutrient content: standard feedings
Calories (kcal/100g)
Carbohydrate (%)
Fat (%)
Protein (%)
Hospital A Measured: Mean + SD
(Range) Expected:
80.9 + 9.1 (73.1-90.9)
88.0
13.4 + 2.7 (10.9-16.2)
11.9
1.77 + 0.47 (1.40-2.30)
3.10
2.83 + 0.60 (2.20-3.40)
3.10 Hospital B Measured: Mean + SD (Range) Expected:
85.9 + 14.8 (70.4-99.8)
53.6
12.0 + 2.9 (8.6-13.8)
7.2
2.16 + 1.66 (0.27-3.40)
1.90
3.00 + 0.44 (2.70-3.50)
1.90 Hospital C Measured: Mean + SD (Range) Expected:
90.2 + 21.7 (66.0-108.0)
152.0
15.2 + 3.9 (10.8-18.4)
20.5
1.63 + 0.40 (1.20-2.00)
5.30
3.63 + 0.55 (3.00-4.00)
5.50 Hospital D Measured: Mean + SD (Range) Expected:
106.5 + 19.8 (84.5-123.0)
129.0
18.7 + 3.5 (14.7-21.4)
23.9
2.57 + 0.58 (1.90-2.90)
2.45
2.13 + 0.65 (1.50-2.80
2.80 SD = standard deviation; N = 3 for all hospitals
MM Sullivan, P Sorreda-Esguerra, MB Platon, CG Castro, NR Chou, S Shott, GM Comer and P Alarcon 390
Table 6. Comparison of measured and expected nutrient content: modified feedings
Calories (kcal/100g) Carbohydrate (%)
Fat (%)
Protein (%)
Hospital A (Anti-diarrheal Diet) Measured: Mean + SD (Range) Expected:
97.8 + 28.3 (73.9-129.0)
84.9
16.2 + 5.7 (11.2-22.4)
13.0
2.43 + 0.51 (2.00-3.00)
2.50
2.77 + 0.59 (2.10-3.20)
2.60 Hospital B (Natural Formula Diet) Measured: Mean + SD (Range) Expected:
64.8 + 14.1 (50.7-78.9)
148.0
11.1 + 2.6 (8.1-13.2)
14.2
1.93 + 0.67 (1.50-2.70)
8.10
1.17 + 0.45 (0.71-1.60)
4.60 Hospital C (High Fibre Low Cholesterol Diet) Measured: Mean + SD (Range) Expected:
82.2 + 4.0 (77.7-85.0)
112.0
14.4 + 1.1 (13.2-15.4)
18.1
1.63 + 0.06 (1.60-1.70)
3.20
2.57 + 0.21 (2.40-2.80)
2.70 Hospital D (Diabetic Diet) Measured: Mean + SD (Range) Expected:
98.5 + 3.7 (94.6-102.0)
126.0
16.3 + 1.6 (15.0-18.1)
23.5
2.87 + 0.55 (2.30-3.40)
2.30
1.87 + 0.40 (1.40-2.10)
2.70 SD = standard deviation; N = 3 for all hospitals
Situations that prevent the patient’s energy require-ments from being met through enteral nutrition have been identified. The reasons stated in the literature for caloric goals not being met in the hospitalized tube fed patient are: mechanical complications with the tube, gastro-intestinal intolerance, and cessation of feeding due to diagnostic procedures.4,13 This study has shown that unreliable composition of blenderized enteral formulas may pose an additional risk for inadequate nutritional intake among tube fed patients. Recent studies by De Jonghe et al., and McClave et al., have investigated the adequacy of energy delivery through tube feeding using available commercial formulas.13,14 In these studies, the investigators did not question the reliability of the nutri-tional content of the feeding formula since the composition of commercial feedings is assumed. It was unfortunate that we were not able to measure all nutrients. However, we did have a representative sample of micronutrients, including: vitamins, minerals, and electrolytes. Whenever the nutrient content of an enteral feeding does not correspond with the expected nutrient levels, adverse outcomes may result. The results of this study demonstrate that despite standardized recipes, hospital prepared enteral tube feedings render unpredictable levels of micro-nutrients and macronutrients. These feedings were more likely to under-deliver rather than over-deliver nutrients, which may result in clinical and nutritional implications for patients at risk of malnutrition.
Acknowledgements This study was supported by Abbott Laboratories. References 1. Gallagher-Allred CR. Comparison of institutionally and
commercially prepared formulas. Nutritional Support Services 1983; 3: 32-34.
2. Tanchoco CC, Florentino RF, Flores EG, Castro Ma CA, Portugal TR. Survey of blenderized diets prepared by some hospitals in Metro Manila: Phase II. Nutrient composition of blenderized diets. Hospital Journal 1990; 22: 17-26.
3. Belknap DC, Seifert CF, Petermann M. Administration of
medications through enteral feeding catheters. Am J Crit Care 1997; 6: 382-392.
4. Abernathy GB, Heizer WD, Holcombe BJ, Raasch RH, Schlegel KE, Hak LJ. Efficacy of tube feeding in supplying energy requirements of hospitalized patients. JPEN 1989; 13: 387-391.
5. Sullivan MM, Sorreda-Esguerra P, Santos EE, Platon BG, Castro CG, Idrisalman ER, Chen NR, Shott S, Comer GM. Bacterial contamination of blenderized whole food and commercial enteral tube feedings in the Philippines. J Hosp Infect 2001; 49: 268-273.
6. The Philippine Food Composition Tables 1997. Food and Nutrition Research Institute, Dept of Science and Techno-logy, Manila, Philippines.
7. Brody T. Nutritional Biochemistry. San Diego, California: Academic Press; 1994.
8. Food and Nutrition Board, Institute of Medicine: Dietary Reference Intakes for Calcium, Phosphorus, Magnesium, Vitamin D, and Fluoride. Washington, DC: National Academies Press; 1997.
9. Levenson DI, Bockman RS. A review of calcium preparations. Nutr Rev 1994; 52: 221-232.
10. Coben RM, Weintraub A, DiMarino AJ Jr, Cohen S. Gastroesophageal reflux during gastrostomy feeding. Gastroenterology 1994; 106: 13-18.
11. Cabre E, Gassull MA. Complications of enteral feeding. Nutrition 1992; 8: 1-9. 12. Kirby DF, DeLegge MH, Fleming CR. American
gastroenterological association technical review on tube feeding for enteral nutrition. Gastroenterology 1995; 108: 1282-1301.
13. De Jonghe B, Appere-De-Vechi C, Fournier M, Tran B, Merrer J, Melchior JC, Outin H. A prospective survey of nutritional support practices in intensive care unit patients: what is prescribed? What is delivered? Crit Care Med 2001; 29: 8-12.
14. McClave SA, Sexton LK, Spain DA, Adams JL, Owens NA, Sullins MB, et al. Enteral tube feeding in the intensive care unit: factors impeding adequate delivery. Crit Care Med 1999; 27: 1252-1256.
391 Asia Pac J Clin Nutr 2004;13 (4): 391-395
Original Article
Lipid peroxidation and antioxidants status in patients with papillary thyroid carcinoma in India Namasivayam Senthil M Phil and Shanmugam Manoharan PhD Department of Biochemistry, Faculty of Science, Annamalai University, Annamalai Nagar, Tamil Nadu, India
The levels of lipid peroxidation products (TBARS), non-enzymatic antioxidants and enzymatic antioxidants activity were investigated in plasma and erythrocytes of twenty clinically diagnosed stage II papillary thyroid cancer patients and an equal number of age and sex matched healthy subjects. An increase in the levels of lipid peroxidation products, decrease in non-enzymatic antioxidants levels and enzymatic antioxidant activities in plasma and erythrocytes were detected in papillary thyroid cancer patients as compared to healthy subjects. Impairment in antioxidant defence mechanisms are responsible for enhanced lipid peroxidation observed in plasma and erythrocytes of papillary thyroid cancer patients.
Keywords: papillary thyroid cancer, lipid peroxidation, antioxidants, India
Introduction Thyroid cancer, the most common endocrine cancer, is a cancerous tumour or growth located within the thyroid gland. The annual incidence rate of thyroid carcinoma varies from 0.5 to 10 cases per 100,000 in different parts of the world.
1 Thyroid cancer accounts for 64% of deaths attributable to malignant endocrine neoplasms, more than all other endocrine cancers combined.
2 Papillary carci-noma and follicular carcinoma are the most common types respectively accounting for about 70% and 15% of cases.
3 In the USA, about 18,000 new cases of thyroid cancer and 1,200 deaths are reported to occur every year.
4 The pattern of thyroid cancer in India is different from that seen in Western countries. In Bombay, North India, the thyroid cancer incidence was found to be at the lowest level in both sexes and it is about three times more frequent among women than men.
5 In Chennai, South India, thyroid carcinoma constitutes about 1-2% of all cancers.
6 Thyroid cancer occurs two to three times more frequently in women than in men.
Free radicals are highly reactive species generated in vivo as by-products of normal metabolism. Although free radicals are utilized by the immune system to kill microbes, free radicals are toxic when generated in excess. Free radicals can damage proteins, lipids, carbo-hydrates, and nucleic acids. Plasma membranes are critical targets of free radical reactions. Lipid pero-xidation is a complex process occurring in aerobic cells and reflects the interaction between molecular oxygen and polyunsaturated fatty acids. Lipid peroxidation has been implicated in the pathogenesis of a variety of diseases including cancer.
7,8 The continuous production of oxidants, are however, balanced by equivalent synthesis
of antioxidants. Antioxidants act as radical scavengers, hydrogen donors, electron donors, peroxide decomposers, singlet oxygen quenchers, enzyme inhibitors and metal-chelating agents. The antioxidant defence system includes antioxidant enzymes (superoxide dismutase, catalase, glu-tathione peroxidase) and small molecule antioxidants (vitamin E, vitamin C and reduced glutathione).
9,10 Studies have demonstrated altered levels of lipid pero-xides and antioxidants in tumour tissues of thyroid cancer patients.
11,12 However, the altered pattern of lipid pero-xidation products and antioxidants have not been well do-cumented in plasma and erythrocytes of papillary thyroid cancer patients. Hence, the present study was undertaken to analyse the levels of lipid peroxidation products and antioxidants in plasma and erythrocytes of papillary thyroid cancer patients.
Methods and Patients Twenty clinically diagnosed stage II papillary thyroid carcinoma (tumour size >1cm) patients from Raja Muthiah Medical College and Hospital, Annamalai University, Annamalai Nagar, India, who had not undergone any previous treatment for their tumours were chosen for the study. An equal number of age and gender matched healthy subjects were also investigated. The patients and healthy subjects were from both genders ranging in age from 40-60 years.
Correspondence address: Dr S Manoharan, Dept. Biochemistry, Faculty of Science, Annamalai University, Annamalai Nagar-608 002, Tamilnadu. India Tel: +91-4144-238343; Fax: +91-4144-238145 Email: [email protected] Accepted 25 June 2004
N Senthil and S Manoharan 392
Table 1. Blood picture of healthy subjects and papillary thyroid cancer patients
The healthy subjects were free from any systemic di-seases. The red and white blood cells and blood haemo-globin levels were significantly reduced in papillary thyroid cancer patients as compared to healthy subjects (Table 1). Blood samples were collected by venous arm puncture into heparinised tubes and the plasma was separated by centrifugation at 3000 rpm for 15 minutes. After plasma separation, the buffy coat was removed and the packed cells were washed thrice with physiological saline. Known volumes of erythrocytes were lysed with hypotonic phosphate buffer at pH 7.4. The hemolysate was separated by centrifugation at 10,000 rpm for 15 minutes at 20o C. The erythrocyte membranes were iso-lated according to the procedure of Dodge et. al.,13 with a change in buffer according to Quist.14 The erythrocytes remaining after the removal of plasma were washed three times with 310mM isotonic tris- HCl buffer (pH 7.4). Haemolysis was performed by pipetting out the washed erythrocytes suspension into polypropylene centrifuge tubes which contained 20mM tris- HCl buffer (pH 7.2). The erythrocyte membranes were sedimented in a high-speed centrifuge at 10,000 rpm for 40 minutes. The supernatant was decanted and erythrocyte membrane pellet was made up to known volume by using 0.2 M isotonic tris- HCl buffer (pH 7.4). Aliquots from these preparations were used for the estimation of thiobarbituric acid reactive substances (TBARS) and vitamin E. Lipid peroxidation was estimated as evidenced by the formation of TBARS. Lipid peroxides in plasma were assayed by the method of Yagi.
15 Plasma was depro-teinised with phosphotungstic acid and the precipitate was treated with thiobarbituric acid at 90oC for 1 hour. Absorbance of the pink coloured complex formed was measured at 535nm. TBARS in erythrocytes and erythro-cyte membranes was estimated by the method of Donnan.
16 Absorbance of the pink chromogen formed by the reaction of thiobarbituric acid with breakdown pro-ducts of lipid peroxides was read at 535 nm. Vitamin E was estimated by the method of Desai.
17 The method involves the reduction of ferric ions to ferrous ions by the tocopherol and the formation of a pink coloured complex with bathophenanthroline orthophos-phoric acid. Absorbance of the stable chromophore was measured at 536nm. Vitamin C level was estimated by the method of Omaye et al.
18 The dehydroascorbic acid
forma colhydra525nbancKelleyello(DTNgroup Taccoamouwith specitathioaccooxidemethinhibtetrazassayof H620n StatiThe comphypobetwerythlation ResuTablthrocand pwerecompdietaeryththyrosignidecrepatiethe lthion
Parameters Healthy subjects
Papillary thyroid cancer patients
RBC count (106 cells/µL)
5.35 ± 0.53 4.83 ± 0.48*
WBC count (103 cells/µL)
11.3 ± 1.3 10.08 ± 1.35*
Haemoglobin (g/dL)
14.8 ± 1.35 11.01 ± 1.12*
Values are expressed as mean ± S.D; N = 20; * Significantly different from healthy subjects, P <0.001
Plasm(nmoEryth(pmoErythmem
Parameter Normal subjects
Papillary thyroid cancer patients
a TBARS l/ml)
2.89 ± 0.30 6.57 ± 0.7*
rocytes TBARS l/mg Hb)
3.20 ± 0.32 6.63 ± 0.6*
rocyte 0.37 ± 0.03 1.07 ± 0.06*
Table 2. Plasma, erythrocytes and erythrocyte membranes TBARS levels in healthy subjects and papillary thyroid cancer patients
branes TBARS
(nmol/mg protein)Values are expressed as mean ± SD; N = 20; *Significantlydifferent from healthy subjects P <0.001
ed by the oxidation of ascorbic acid by copper, forms oured product on treatment with 2,4 dinitro phenyl zene, which was measured at an absorbance of
m. Reduced glutathione was measured at an absor-e of 412nm according to the method of Beutler and y.
19 The method was based on the development of w colour, when 5,5’-dithio-bis-2-nitrobenzoic acid B) was added to compound containing sulphydryl s.
he activity of glutathione peroxidase was estimated rding to the method of Rotruck et al.
20 A known nt of hemolysate preparation was allowed to react H2O2 in the presence of reduced glutathione for a fied time period, and the remaining reduced glu-ne content was measured at an absorbance of 412nm
rding to the method of Beutler and Kelley.19 Super- dismutase activity was assayed at 520nm by the
od of Kakkar et al.21 The assay was based on the
ition of NADH – Phenazine methosulphate nitroblue olium formation. The activity of catalase was ed by the method of Sinha22 based on the utilization
2O2 by the enzyme. The colour developed was read at m.
stical analysis values are expressed as mean ± SD. Statistical arisons were done by Student’s t-test. The null thesis was rejected for P<0.05. The relationship een lipid peroxidation and antioxidants in plasma and rocytes were determined using Karl Pearson’s corre- coefficient.
lts e 2 shows the levels of TBARS in plasma, ery-ytes and erythrocyte membranes of healthy subjects apillary thyroid cancer patients. The TBARS levels
significantly increased in thyroid cancer patients as ared to healthy subjects. Table 3 shows the levels of
ry antioxidants (vitamin C, and E) in plasma and rocyte membranes of healthy subjects and papillary id cancer patients. The vitamin C and E levels were ficantly decreased in plasma and vitamin E was ased in erythrocyte mem-branes of thyroid cancer nts as compared to healthy subjects. Table 4 shows evel of reduced glutathione and activity of gluta-e peroxidase, in plasma and erythrocytes of healthy
393 Oxidants and antioxidants in thyroid cancer
subjects and papillary thyroid cancer patients. The level of glutathione and glutathione peroxidase activity was decreased in thyroid cancer patients as compared to healthy subjects. Table 5 shows the activities of super-oxide dismutase and catalase in erythrocyte lysate of healthy subjects and papillary thyroid cancer patients. The superoxide dismutase and catalase activities were significantly decreased in thyroid cancer patients as compared to healthy subjects. Table 6 shows Karl Pearson’s correlation coefficient for lipid peroxidation and antioxidants in plasma, erythrocyte lysate and ery-throcyte membranes in papillary thyroid cancer patients. Statistically significant negative correlation was observed between lipid peroxidation against vitamin E, SOD, and CAT in patients with papillary thyroid cancer. Although negative correlation was observed between lipid pero-xidation against vitamin C, reduced glutathione and glu-tathione peroxidase in papillary thyroid cancer patients, the values were statistically insignificant.
Discussion In the present study, we observed an increase in lipid peroxidation products (TBARS) and decline in anti-oxidants status in plasma, erythrocytes and erythrocyte membranes of papillary thyroid cancer patients as com-pared to healthy subjects. Durak et al.,
23 have reported an increase in malondialdehyde (MDA) levels and decrease in the activities of enzymatic antioxidants as compared to non-cancerous tissues. Sadani and Nadkarni12 have reported that the higher production of reactive oxygen species is responsible for the elevated lipid peroxidation in adenomas and carcinoma of thyroid tissues. Mano et al.,
11 have reported an increase in lipid peroxidation and disturbed antioxidant enzymes in thyroid tumour tissues of patients with papillary carcinoma. They suggested that the lipid peroxides are not completely scavenged in papillary carcinoma tissues and therefore these substances affect some role in cell function of thyroid tissues. The reactive oxygen species play a crucial role in the pathogenesis of tissue damage in many pathological conditions via a peroxidation of membrane phospholipids. Gutteridge24 has focused lipid peroxidation and anti-oxidants as biomarkers of tissue damage. Lipid peroxides that are generated at the site of tissue injury could be transferred through circulation to other organs and tissues
to provoke damage by propagating lipid peroxidation. Elevated levels of lipid peroxides have been reported in the erythrocytes and erythrocyte membranes of various cancer patients.
25,26 Our results lend credence to these observations. Hence, the observed increase in plasma TBARS in papillary thyroid cancer patients may be related to over production of lipid peroxides in erythro-cytes, erythrocyte membranes or tumour tissues itself with consequent leakage into plasma. Antioxidants have been reported to protect cell against cancer. Antioxidants play an important role in scavenging lipid peroxides that are generated at the site of tumours. The antioxidant capacity is determined by dynamic interactions between nonenzymatic antioxidants and antioxidative enzymes.
27 α-Tocopherol is known to have the greatest biological activity of the various stereo-isomers of the vitamin E. In vivo α tocopherol is the most abundant lipid soluble antioxidant and acts as an impor-tant inhibitor of membrane lipid peroxidation.
28 Mano et al.,29 reported an increase in vitamin E concentration in thyroid tumor tissues as compared to normal tissues. Vitamin C, the most important antioxidant in the plasma, scavenges a variety of oxidants. Dumitrescu et al.,30 have reported a decrease in plasma MDA level after one month of vitamin C administration in patients with differentiated thyroid cancer who have undergone surgery and are being treated with Iodine.
131 Hence, the decreased levels of vitamin E and C observed in the present study can be correlated to elevated plasma lipid peroxidation or utilization of these antioxidants by tumour tissues to scavenge excess lipid peroxides that are generated in the tumour tissues. Glutathione peroxidase, superoxide dismutase and cata-lase are most important enzymes of the cell antioxidant defense system. Glutathione peroxidase, catalyses the decomposition of H2O2 to H2O and reduces organic
Plasm vit vitEryth vi (µ
Parameter Normal subjects
Papillary thyroid cancer
patients Plasma GSH (mg/dl) 48.7 ± 3.76 38.2 ± 3.02* Erythrocytes GSH (mg/ dl)
52.6 ± 4.1 39.3 ± 2.6*
Plasma GPx (U/L) 222.68 ± 28.8 186.13 ± 8.44* Erythrocyte lysate GPx 34.06 ± 3.5 26.6 ± 2.48*
Table 4. Levels of reduced glutathione and activity of glutathione peroxidase, in plasma and erythrocyte lysate of healthy subjects and thyroid papillary cancer patients
Table 5. Activities of superoxide dismutase and catalase in erythrocyte lysate of healthy subjects and papillary thyroid cancer patients
Ery
Parameter Normal subjects
Papillary thyroid cancer patients
a amin C (mg/dl)
1.56 ± 0.15
0.64 ± 0.06*
amin E (mg/dl) 1.41 ± 0.13 0.85 ± 0.08* rocyte membranes
tamin E g/mg protein)
2.08 ± 0.19 1.58 ± 0.14* SOD CAT
Parameter throcyte lysate
Normal subjects
Papillary thyroid cancer patients
(U/mg Hb) 2.14 ± 0.24 1.64 ± 0.16* (U/mg Hb) 5.86 ± 0.58 4.42 ± 0.38*
(U/g Hb) Values are expressed as mean ± SD; N = 20; *Significantly different from healthy subjects P <0.001
Table 3. Plasma and erythrocyte membranes dietary antioxidant levels in healthy subjects and papillary thyroid cancer patients
Values are expressed as mean ± SD; N = 20; *Significantly different
from healthy subjects P <0.001
Values are expressed as mean ± SD; N = 20l; *Significantly different from healthy subjects P <0.001
N Senthil and S Manoharan 394
peroxide to corresponding alcohols. The decrease in glu-tathione peroxidase activity has been reported in thyroid cancer tissues as compared to normal tissues.
31 Reduced glutathione, an important intracellular antioxidant, acts both as cofactor for glutathione peroxidase and as a direct active scavenger to remove reactive species such as hy-droxyl radical, peroxynitrite, carbon centered radicals and singlet oxygen. Reduced glutathione also play a very effective role in the protection of erythrocyte from oxi-dative damage.
32 Lower levels of reduced glutathione in plasma and erythrocytes have been reported in various pathological conditions including cancer.
33 Hence the de-creased activities of plasma and erythrocytes glutathione peroxidase in papillary thyroid cancer patients may be due to lowered level of reduced glutathione observed in plasma and erythrocytes. Decreased activities of catalase and superoxide dis-mutase have been demonstrated in various pathological diseases including cancer.
34 Decreased activity of super-oxide dismutase and unaltered activity of catalase has been reported in thyroid tumour tissues.
35 Lowered activities of these enzymes observed in erythrocytes of thyroid cancer patients can be related to enhanced lipid peroxidation observed in the erythrocytes of papillary thyroid cancer patients. Hence, we feel that significantly impaired enzymatic and non enzymatic antioxidant de-fence systems are responsible for the elevated lipid pe-roxidation products in plasma and erythrocytes of papi-llary thyroid cancer patients. References 1. Dal Maso L, La Vecchia C, Franceschi S, Preston-Martin
S, Ron E, Levi F, Mack W, Mark SD, McTiernan A., Kolonel L, Mabuchi K, Jin F, Wingren G, Galanti MR, Hallquist A, Glattre E, Lund E, Linos D, Negri E. A pooled analysis of thyroid cancer studies. Cancer Causes Control 2000; 11: 137-144.
2. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics. CA. Cancer J Clin 1999; 49: 8-31.
3. Ortapamuk H, Arican P, Alp A. Parapharyngeal lymph node involvement in papillary thyroid carcinoma. Clin Nucl Med 2003; 28: 947-948.
4. Wu XC, Chen VW, Steele B, Roffers S, Klotz JB, Correa CN Carozza SE. Cancer incidence in adolescents and young adults in the United States, 1992-1997. J Adolesc Health 2003; 32: 405-415.
5. Rao RS, Parikh DM, Mistry RC, Rao SR. Evidence-based protocols for the management of well-differentiated carcinomas of the thyroid. Asian J Surg 2002; 25: 319-324.
6. Dorairajan N, Pandiarajan R, Yuvaraja S. A descriptive study of papillary thyroid carcinoma in a teaching hospital in Chennai, India. Asian J Surg 2002; 25:300-303.
7. Halliwell B, Gutteridge JMC. Lipid peroxidation, oxygen radicals, cell damage and antioxidant theraphy. Lancet 1994; 1: 1396-1397.
8. Ray G, Akthar Hussain S. Oxidants, antioxidants and carcinogenesis. Indian J Exp Biol 2002; 40: 1213-1232.
9. Bray TM. Antioxidant and oxidative stress in health and disease. Soc Exp Biol Med 1999; 23: 195-198.
10. Gey KF. Vitamin E plus C and interacting co-nutrients required for optimal health. Biofactors 1998; 7: 113-174.
11. Mano T, Shinohara R, Iwase K. Changes in free radical scanvengers and lipid peroxide in thyroid glands of various thyroid disorders. Horm Metab Res 1997; 29: 351-354.
12. Sadani GR, Nadkarni GD. Radiation Medicine Centre (BARC), Tata Memorial Centre Annexe, Parel, Bombay, India. Role of tissue antioxidant defence in thyroid cancers. Cancer Lett 1996; 109: 231-215.
13. Dodge JF, Mitchell G, Hanhan DJ. The preparation and chemical characterization of haemoglobin free ghosts of human red blood cells. Arch Biochem Biophys 1968; 110: 119-130.
14. Quist EH. Regulation of erythrocyte membrane shape by calcium ion. Biochem Biophys Res Commun 1980; 92: 631-637.
15. Yagi K. Lipid peroxides and human disease. Chem Phys Lipids 1978; 45: 337-351
16. Donan SK. Thiobarbituric acid test applied to tissues from rats treated in various ways. J Biol Chem 1950; 182: 415-419.
17. Desai FD. Vitamin E analysis and methods for human tissue. In: Feischer S, Packer L. eds. Methods Enzymol. New York: Academic Press, 1984; 105: 138-145.
18. Omaye ST, Turbull TD, Sauberlich HE. Selected method for the determination of ascorbic acid in animal cells, tissues and fluids. In: McCromic DB, Wright DL, eds. Methods Enzymol. Academic Press, 1979; 62: 3-11.
19. Beutler E, Kelley BM. The effect of sodium nitrate on RBC glutathione. Experienta 1963: 96-97.
20. Rotruck JT, Pope AL, Ganther HE, Swanson AB, Hafeman DG, Hockstra WG. Selenium biochemical role as a component of glutathione peroxidase. Science 1973; 179: 588-590.
21. Kakkar P, Das B, Visvanathan PN. A modified spectrophotometric assay of superoxide dismutase. Ind J Biochem Biophys 1984; 21: 130-132.
22. Sinha KA. Colorimetric assay of catalase. Anal Biochem 1972; 17: 389-394.
23. Durak I, Bayram F, Kavutcu M, Canbolat O, Ozturk HS. Impaired enzymatic antioxidant defense mechanism in cancerous human thyroid tissues. J Endocrinol Invest 1996; 19: 312-315.
24. Gutteridge JMC. Lipid peroxidation and antioxidants as biomarkers of tissue damage. Clin Chem 1995; 41: 1819-1828.
25. Manoharan S, Shreeram S, Nagini S. Life style can induce lipid peroxidation and compromise of antioxidant defense mechanisms in the erythrocytes of oral cancer patients. Med Sci Res 1996; 24: 397-400.
Parameters Papillary thyroid cancer patients
Plasma TBARS-Vitamin E
- 0.546*
TBARS-Vitamin C - 0.339NS TBARS-Glutathione - 0.238 NS TBARS-Glutathione peroxidase - 0.120 NS Erythrocyte lysate TBARS-Glutathione
- 0.226 NS
TBARS- Glutathione peroxidase - 0.172 NS TBARS-Superoxide dismutase - 0.537* TBARS-Catalase - 0.454* Erythrocyte membranes TBARS-Vitamin E
- 0.465*
Table 6. Karl Pearson’s correlation coefficient for lipid peroxidation and antioxidants in papillary thyroid cancer patients
*Values are statistically significant at P < 0.05; NSNot significant
395 Oxidants and antioxidants in thyroid cancer
26. Ray G, Batra S, Shukla NK, Des S, Raina V, Ashok S, Husain SA. Lipid peroxidation and antioxidant status in breast cancer. Breast Cancer Res Treat 2000; 59: 163-170.
27. Mates JM. Effects of antioxidant enzymes in the molecular control of reactive oxygen species toxigology. Toxicology 2000; 153: 83-104.
28. McCay PB. Vitamin E interactions with free radicals and ascorbate. Annu Rev Nutr 1985; 5: 323.
29. Mano T, Iwase K, Hayashi R, Hayakawa N, Uchimura K, Makino M, Nagata M, Sawai Y, Oda N, Hamada M, Aono T, Nakai A, Nagasaka A, Itoh M. Vitamin E and coenzyme Q concentrations in the thyroid tissues of patients with various thyroid disorders. Am J Med Sci 1998; 315: 230-232.
30. Dumitrescu C, Belgun M, Olinescu R, Lianu L, Bartoc C. Effect of vitamin C administration on the ratio between the pro- and antioxidative factors. Rom J Endocrinol 1993; 31: 81-84.
31. Hasegawa Y, Takano T, Miyauchi A, Matsuzuka F, Yoshida H, Kuma K, Amino N. Decreased expression of glutathione peroxidase mRNA in thyroid anaplastic car-cinoma. Cancer Lett 2002; 182: 69-74.
32. Banerjee KK, Marimuthu P, Chaudhri RN. Glutathione and lipid peroxidation status. Indian J Public Health 1998; 42: 20-23.
33. Gerad MD, Chaudrie J. Metabolism and antioxidant functions of glutathione. Pathol Biol 1996; 44: 77-83.
34. Oberly LW, Oberley TD. The role of superoxide dismutase and the gene amplification in carcino-genesis. J Ther Biol 1984; 106: 403.
35. Felt V, Cerny K, Klimova H, Kremenova J, Rehak F. Enzyme changes in thyroid tissue with reference to thyroid cancer. Acta Univ Carol Med Praha 1976; 22: 435-442.
Asia Pac J Clin Nutr 2004;13 (4):396-400 396
Original Article Potential anticancer effect of red spinach (Amaranthus gangeticus) extract Huzaimah Abdullah Sani MSc1, Asmah Rahmat PhD1, Maznah Ismail PhD1, Rozita Rosli PhD2 and Susi Endrini PhD1 1Department of Nutrition and Health Sciences 2Department of Human Growth and Development, Faculty of Medicine and Health Sciences,Universiti Putra Malaysia, 43400, Serdang, Selangor Darul Ehsan, Malaysia.
The objective of this study was to determine the anti cancer effects of red spinach (Amaranthus gangeticus Linn) in vitro and in vivo. For in vitro study, microtitration cytotoxic assay was done using 3-(4,5-dimethylthiazol-2-il)-2,5-diphenil tetrazolium bromide (MTT) kit assay. Results showed that aqueous extract of A gangeticus inhibited the proliferation of liver cancer cell line (HepG2) and breast cancer cell line (MCF-7). The IC50 values were 93.8 µg/ml and 98.8 µg/ml for HepG2 and MCF-7, respectively. The inhibitory effect was also observed in colon cancer cell line (Caco-2), but a lower percentage compared to HepG2 and MCF-7. For normal cell line (Chang Liver), there was no inhibitory effect. In the in vivo study, hepatocarcinogenesis was monitored in rats according to Solt and Farber (1976) without partial hepatectomy. Assay of tumour marker enzymes such as glutathione S-transferase (GST), gamma-glutamyl transpeptidase (GGT), uridyl diphosphoglucuronyl transferase (UDPGT) and alkaline phosphatase (ALP) were carried out to determine the severity of hepatocarcinogenesis. The result found that supplementation of 5%, 7.5% and 10% of A. gangeticus aqueous extract to normal rats did not show any significant difference towards normal control (P <0.05). The exposure of the rats to chemical carcinogens diethylnitrosamine (DEN) and 2-acetylaminofluorene (AAF) showed a significant increase in specific enzyme activity of GGT, GST, UDPGT and ALP compared to normal control (P <0.05). However, it was found that the supplementation of A. gangeticus aqueous extract in 5%, 7.5% and 10% to cancer-induced rats could inhibit the activity of all tumour marker enzymes especially at 10% (P <0.05). Supplementation of anti cancer drug glycyrrhizin at suggested dose (0.005%) did not show any suppressive effect towards cancer control (P <0.05). In conclusion, A. gangeticus showed anticancer potential in in vitro and in vivo studies.
Key Words: red spinach, anticancer effect, in vitro and in vivo studies. Introduction Cancer is believed to be the result of external factors combined with a hereditary disposition for cancer. It is a neoplasm characterized by the uncontrolled growth of the anaplastic cells that tends to invade surrounding tissue and to metastasize to distance body sites.1 In Malaysia, cancer is one of the leading causes for morbidity and mortality. It was estimated to be 30,000 cases annually.2 Cancers are complicated diseases. Although epidemiological data on populations may help identify exogenous agents, the probability of identifying the agent is not enough unless there are good dose-response data for humans or animal models. A group of vegetables with considerable anti-carcinogenic properties are the cruciferous vegetables. In epidemiological studies, it was shown that intake of cruciferous plants is inversely associated with kidney, prostate, bladder, colon, rectum and lung cancer risk.3-8 Malaysia has a variety of natural resources. Previous studies showed that low consumption of vegetables is found to be associated with the increased risk of cancer.9 Antioxidant activity present in these vegetables perhaps may have some benefits in cancer. Epidemiological studies
suggest that vitamin E and other antioxidants may reduce cancer incidence. It has been observed that people who eat diets rich in fruits and vegetables, which are rich in antioxidants, have lower incidences of cancer.10 Amaranthus tender (Red spinach: Amaranthus gange-ticus) is a carotene-rich food available in Malaysia that has potential as a dietary source of chemopreventive phyto-chemicals. Consumption pattern of beta-carotene rich foods from 500 household of Coimbatrore District in India was studied.11 Results indicated that greens mainly were purchased from market and were consumed 2 –3 times per week. In-vitro and in-vivo experiments have been analysed from red spinach for possible anticancer agents. The cytotoxic effect of Amaranthus gangeticus aqueous and ethanolic
Correspondence address Assoc. Prof. Dr Asmah Rahmat, Department of Nutrition and Health Sciences, Universiti Putra, Malaysia 43400, Serdang, Darul Ehsan Malaysia., Tel: + 603-89468443; Fax: + 603-89426769 Email: [email protected] Accepted 30 January 2004
397 H A Sani, A Rahmat , M Ismail, R Rosli and S Endrini
extracts were used to determine the IC50-value against several cancer cell lines such as non-estrogen dependent breast cancer cell lines (MDA-MB-231), estrogen-dependent breast cancer cell lines (MCF-7), liver cancer cell lines (HepG2), colon cancer cell lines (Caco-2) and transformed liver cell lines (Chang Liver). In-vivo studies were focused primarily for identifying the crude extract against hepatocarcinogenesis in rats induced by Diethyl-nitrosamine (DEN) and 2-acetylaminofluorene (AAF). We report the effect of red spinach on the enzyme tumour markers activity including Gluthathione S-Transferase (GST, EC 2.5.1.18), γ-Glutamyl Transpeptidase (GGT, EC 2.3.2.2), Alkaline Phosphatase (ALP, EC 3.1.3.1) and Uridyl Diphospho Glucoronyl Transpeptidase (UDPGT, EC 2.4.1.18) from liver of rats. Materials and Methods In-vitro studies The in-vitro studies were designed to determine cytotoxic effect of the leaves aqueous and ethanolic extracts against several cancer cell lines such as MDA-MB-231, MCF-7, HepG2, Caco-2 and Chang Liver. The leaves of Amaranthus gangeticus were obtained from a supplier at Seri Serdang, Selangor. Ethanol ex-traction method was used according to Ali et al (1996).12 One hundred grams of fresh leaves were ground and soaked in water or 90% ethanol at room temperature overnight. The extracts were then filtered and evaporated with a rotary evaporator. After that, the dried residue was stored at -80°C and freeze-dried. The extract was ready for the treatment (in vitro). MTT Assay (Boehringer Mannheim) HepG2, Caco2, MDA-MB-231, MCF-7 and Chang liver cell lines culture were obtained from American Type Culture Collection (ATCC). HepG2 cells were cultured in Earl’s Minimum Essential Medium; MCF-7, MDA-MB-231 and Caco2 cells were cultured in Dullbecco’s Modified Eagle’s Medium; and Chang liver cells were cultured in Roswell Park Memorial Institute 1640 supplemented with 10% of fetal bovine serum (FBS), 100 IU/ml of penicillin and 100 μg/ml of Streptomycin using 25-cm2 flasks, in 5% CO2 incubator at 37˚C. The viability of cells was determined with trypan blue reagent. Expo-nentially growing cells were harvested, counted with hae-mocytometer and diluted with a particular medium. Cell culture with the concentration of 1 x 105 cells/ml was prepared and was plated (100 μl/well) onto 96-well plates (NUNCTM, Denmark). The diluted ranges of extracts were added to each well and the final concentrations of the test extracts were 5, 10, 20, 40, 60, 80, and 100 μg/ml. The proliferative activity was determined using the MTT assay (3- [4, 5 - dimethylthiazol - 2-yl]-2,5-diphenyl tetrazolium bromide). The incubation period used was 72 hours. After solubilization of the purple formazan crystals were completed, the spectrophotometrical absorbance of the plants extract was measured using an ELISA reader at a wavelength of 550 nm. The cytotoxicity was recorded as the drug concentration causing 50% growth inhibition of the tumour cells (IC50 value). % cell viability = OD sample (mean) x100% OD control (mean)
After the determination of the cytotoxicity percentage, graphs were plotted with the percentage of cytotoxicity against its respective concentrations.. In-vivo studies The in-vivo studies were conducted to determine the effect of three different doses of red spinach juice against hepatocarcinogenesis. A total of 64 male Sprague-dawley rats, each initially weighing between 120 – 150 g were housed individually at 27oC and were maintained on nor-mal or treated rat chow. The rats were divided into nine groups i.e. group I: control (basal diet) (N), group II-IV: AG-supplemented diet (5%, 7.5% and 10%) in drinking water, group V: cancer (DEN/AAF) with basal diet after week 4 (C), group VI- IX: cancer (DEN/AAF)-AG supplemented diet (C 5, C 7.5 and C 10), group XI cancer (DEN/AAF)-treated with Glycyrrhizin). The crude extract was prepared from the modification of a previous method.13 A 100 g of A.gangeticus leaves were ground in 1000 ml of distilled water (10%) and filtered. The filtrate was diluted with distilled water to obtain the concentration that was used (5, 7 and 10%). The extract was stored at 4oC. Hepatocarcinogenesis was induced according to the method of Solt and Farber (1976),14 but without partial hepatectomy. Animals in the groups 2, 7-14 were intra-peritoneally given a single injection of DEN (200 mg/kg body weight) dissolved in corn oil at the beginning of the experiment to initiate hepatocarcinogenesis. After 2 weeks of feeding with standard basal diet, promotion of hepa-tocarcinogenesis was done with administration of AAF (0.02% in basal diet) for 2 weeks without partial hepa-tectomy. Treatment with AG (at different concentration) was given as a substitute to distilled water in Groups II - IV and glycyrrhizin in group XI. A summary of the protocol is presented in (Fig. 1).
Groups Treatments NC Basal diet + water N 5.0 Basal diet + AG 5.0 N 7.5 Basal diet + AG 7.5 N 10 Basal diet + AG 10 CC Basal diet + water AAF + water Basal diet + water C 5.0 Basal diet + AG 5.0 AAF + AG 5.0 Basal diet + AG 5.0 C 7.5 Basal diet + AG 7.5 AAF + AG 7.5 Basal diet + AG 7.5 C 10 Basal diet + AG 10 AAF + AG 10 Basal diet + AG 10 C 5.0 Basal diet + GL AAF + GL Basal diet Weeks DEN 0 2 4 14
Figure 1. Study protocol for in-vivo studies. DEN, 200 mg/kg diethylnitrosamine (ip); AAF, 0.02% 2-acetylaminofluorene; NC=control; C=DEN/AAF; AG 5=Amaranthus gangeticus extract 5%; N=Normal, AG 7.5=Amaranthus gangeticus 7.5%; AG 10=Amaranthus gangeticus extract 10%; GL=Glycyrrhizin; CC=Cancer control
Potential anticancer effect of red spinach extract 398
Table 1. IC50 values of aqueous and ethanolic extracts from Amaranthaus gangeticus extracts Determination of glutathione, γγγγ-glutamyl transpeptidase, GST, UDPGT and alkaline phosphatase. Blood was taken immediately from the orbital sinus vein and plasma was separated by centrifugation at 3000 rpm 4oC and used for GGT and ALP assays. The rats were sacrificed by cervical dislocation at 14 weeks from the DEN injection. The livers were weighed and stored at –70oC before use. The microsomal fraction of the liver was prepared according to the method of Speir and Wattenberg (1975).15 GGT assay was determined in the microsomal fraction while ALP activity and the level of GSH were determined in the homogenate of the liver. Plasma and liver GGT activities were assayed following the method of Jacobs (1971)16 and the activities were expressed as units per litre and units per gram protein, respectively. The microsomal pellet was first resuspended in 5 volume of 0.1 M Tris-HCl buffer pH 8.2 containing 1 mM MgCl2. Alkaline phosphatase activity was assayed by the method of Jahan and Butterworth (1986).17 One unit of activity was expressed as the amount of enzyme required to catalyse the release of 1 µmol p-nitrophenol/min under the condition stated. Liver ALP and GGT were expressed as units per litre per
milligram protein, respectively. Protein concentration was determined by using the method of Bradford (1976).18 The activity of GST in the liver cytosol was assayed according to the method of Habig et al., (1974)19 using CDNB and DCNB as the substrates. Specific activity was defined as µmol/min/mg protein in the cytosol. UDPGT activity in the liver microsome was assayed by the method of Vassey and Zakim (1972)20 using p-nitrophenol as substrate and uridyl diphosphoglucuronyl acid (UDPGA) as glucuronic acid source. Specific activity of UDPGT was expressed as µmol/min/mg protein. Statistical analysis Statistical comparisons were carried out using student’s t-test. Probability level of P<0.05 was chosen as the criterion of statistical significance. Values reported were mean ± SD.
Results In vitro studies The IC50 from aqueous and ethanolic extracts of A.gangeticus are shown in (Table 1). The ethanolic extract of A.gangeticus was observed to inhibit the proliferative of HepG2 cells (IC50 27.75µg/ml), MDA-MB 231 cells (IC50 27.75µg/ml) and MCF-7 cells (IC50 12.50 µg/ml) whereas the aqueous extract inhibited the proliferation of HepG2 and MCF-7 cells (IC50 93.8 and 98.8µg/ml respectively). In vivo studies The result showed that supplementation of 5%, 7.5% and 10% of A.gangeticus aqueous extract to normal rats did not show any significant difference towards normal control (P <0.05). The exposure of the rats to chemical carcinogens DEN/AAF showed a significant increase in specific enzyme activities of GGT, UDPGT, GST and ALP compared to normal control (P <0.05). However, it was found that the supplementation of A.gangeticus aqueous extract to the DEN/AAF-treated rats decreased all tumour marker enzymes, especially at 10% (P<0.05).
IC50 (µg/ml) No.
Types of cell lines
Ethanolic extract
Aqueous extract
1. HepG2 27.75 93.8
2. MCF-7 12.5 98.8
3. MDA-MB-231 27.75 >110
4. Caco-2 >100 >100
5. Chang Liver >100 >100
0 0 .5 1
1 .5 2
2 .5 3
3 .5 4
4 .5 5
N N 5 N 7 .5 N 1 0 C C 5 C 7 .5 C 1 0 C G T r e a t m e n ts
Sp
ec
ific
ac
tiv
ity
of
mic
ros
GG
T(um
ol/
min
/mg
pro
tein
d ,e f , i
e , f , i e , f , i
a ,e ,h , i
a ,b , c d ,g ,h
a ,b , c g ,h e , f , i
d ,e , f , i
a ,b , c d ,g ,h
Figure 2. Effect on specific activity of microsom GGT in 3 months. Mean ± S.D. (N = 8). N - Normal, N5 – Normal + 5% dose of A. gangeticus N7.5 – Normal + 7.5% dose of A. gangeticus, N10 – Normal + 10% dose of A. gangeticus, C – Cancer induced, C5 – Cancer induced + 5% dose of A. gangeticus, C7.5 – Cancer induced + 7.5% dose of A. gangeticus, C10 – Cancer induced + 10% dose of A. gangeticus, CG – Cancer induced + Glycirrhizin. a: P <0.05 compared to normal; b: P <0.05 compared to normal + 5% dose of A. gangeticus; c: P <0.05 compared to normal + 7.5% dose of A. gangeticus; d: P <0.05 compared to normal + 10% dose of A. gangeticus; e: P < 0.05 compared to cancer induced; f: P <0.05 compared to cancer induced + 5% dose of A. gangeticus; g: P <0.05 compared to cancer induced + 7.5% dose of A. gangeticus; h: P<0.05 compared to cancer induced + 10% dose of A. gangeticus ii:: PP<<00..0055 ccoommppaarreedd ttoo ccaanncceerr iinndduucceedd ++ GGllyycciirrrrhhiizziinn
99 H A Sani, A Rahmat , M Ismail, R Rosli and S Endrini
30
0.2
0.4
0.6
0.8
1
1.2
N N5 N7.5 N10 C C5 C7.5 C10 CG
Treatments
Spec
ific
activ
ity o
f GST
(um
ol/m
in/m
gpr
otei
n) g,h,i
e,f,g,i
d,g,h,id,g,h,i
d,e,fe,f
d,e,f
g,h,id,e,f d,e,f
g,h,id,e,f
a,b,c
a,b,ca,b,c
a,b,ca,b,c
a,b,c
FFiigguurree 33.. EEffffeecctt oonn ssppeecciiffiicc aaccttiivviittyy ooff GGSSTT iinn 33 mmoonntthhss.. MMeeaann ±± SS..DD.. ((NN == 66--88)).. N - Normal, N5 – Normal + 5% dose of A. gangeticus N7.5 – Normal + 7.5% dose of A. gangeticus, N10 – Normal + 10% dose of A. gangeticus, C – Cancer induced, C5 – Cancer induced + 5% dose of A. gangeticus, C7.5 – Cancer induced + 7.5% dose of A. gangeticus, C10 – Cancer induced + 10% dose of A. gangeticus, CG – Cancer induced + Glycirrhizin. a: P <0.05 compared to normal; b: P <0.05 compared to normal + 5% dose of A. gangeticus; c: P <0.05 compared to normal + 7.5% dose of A. gangeticus; d: P <0.05 compared to normal + 10% dose of A. gangeticus; e: P <0.05 compared to cancer induced; f: P <0.05 compared to cancer induced + 5% dose of A. gangeticus; gg:: PP <<00..0055 ccoommppaarreedd ttoo ccaanncceerr iinndduucceedd ++ 77..55%% ddoossee ooff AA.. ggaannggeettiiccuuss;; hh:: PP <<00..0055 ccoommppaarreedd ttoo ccaanncceerr iinndduucceedd ++ 1100%% ddoossee ooff AA.. ggaannggeettiiccuuss;; ii:: PP <<00..0055 ccoommppaarreedd ttoo ccaanncceerr iinndduucceedd ++ GGllyycciirrrrhhiizziinn
FiNofinnof: h:
FiNgaGdotoca
0
1
2
3
4
5
6
N N5 N7.5 N10 C C5 C7.5 C10 CG
Treatments
Spec
ific
Act
ivity
of U
DPG
T(m
mol
/min
/mg
prot
ein)
g,h,ig,h,i g,h,i
e,fg,h,i
e,fe,f e,f
d,g,ha,b,c d,g,ha,b,c
a,b,c,d,ha,b,cd,e,f,h a,b,c,de,f,g,i
gure 4. Effect on specific activity of UDPGT in 3 months. Mean ± S.D. (N = 6-8). N - Normal, N5 – Normal + 5% dose of A. gangeticus 7.5 – Normal + 7.5% dose of A. gangeticus, N10 – Normal + 10% dose of A. gangeticus, C – Cancer induced, C5 – Cancer induced + 5% dose A. gangeticus, C7.5 – Cancer induced + 7.5% dose of A. gangeticus, C10 – Cancer induced + 10% dose of A. gangeticus, CG – Cancer duced + Glycirrhizin. a: PP <0.05 compared to normal; b: PP <0.05 compared to normal + 5% dose of A. gangeticus; c: PP <0.05 compared to rmal + 7.5% dose of A. gangeticus d: PP <0.05 compared to normal + 10% dose of A. gangeticus; e: PP <0.05 compared to cancer induced; PP <0.05 compared to cancer induced + 5% dose of A. gangeticus; g: PP <0.05 compared to cancer induced + 7.5% dose of A. gangeticus
PP <0.05 compared to cancer induced + 10% dose of A. gangeticus; i: PP <0.05 compared to cancer induced + Glycirrhizin
gureormalngeti
lycirrse of cancncer
0
1
2
3
4
5
6
7
N N5 N7.5 N10 C C5 C7.5 C10 CG
Treatments
Spec
ific
Act
ivity
of A
LP (u
mol
/min
/mg
prot
ein)
f,g,i
a,b,e f,g,ic,d,h c,d,h
a,b,e f,g,i
c,d,h c,d,h c,d,h a,b,e c,d,h
5. Effect on specific activity of ALP in 3 months. Mean ± S.D. (N =6-8). N - Normal, N5 – Normal + 5% dose of A. gangeticus N7.5 – + 7.5% dose of A. gangeticus, N10 – Normal + 10% dose of A. gangeticus, C – Cancer induced, C5 – Cancer induced + 5% dose of A. cus, C7.5 – Cancer induced + 7.5% dose of A. gangeticus, C10 – Cancer induced + 10% dose of A. gangeticus, CG – Cancer induced + hizin. a: PP <0.05 compared to normal; b: PP <0.05 compared to normal + 5% dose of A. gangeticus; c:PP <0.05 compared to normal + 7.5% A. gangeticus d: PP <0.05 compared to normal + 10% dose of A. gangeticus; e: PP <0.05 compared to cancer induced; f: PP <0.05 compared er induced + 5% dose of A. gangeticus; g: PP <0.05 compared to cancer induced + 7.5% dose of A. gangeticus h: PP <0.05 compared to induced + 10% dose of A. gangeticus; i: PP <0.05 compared to cancer induced + Glycirrhizin
Potential anticancer effect of red spinach extract 400
Discussion GGT, GST and ALP have been recognized as a positive marker for hepatocytes, which have undergone malignant transformation.21 Our previous studies on tocotrienol supplementation, administered over the short or long-term, attenuated the impact of carcinogens in the rats.21-23 In rats treated with DEN/AAF, vitamin E supplementation atte-nuated GGT and ALP activities and blood GSH levels. The optimum dose required for highest attenuation of the tumour marker enzyme activities was 34mg/kg diet for α-tocopherol and 30mg/kg diet for γ-tocotrienol. Higher doses of the vitamin did not show further attenu-ation in the level of the tumour marker enzyme activities. The IC50–values of ethanolic extracts of A.gangeticus from in-vitro studies were very much lower when compared to the aqueous extracts. This may be due to the high antioxidant activities that offer protection against damage due to free radicals.24 Carotenoids, vitamin E and fibres in plants have been implicated as anticarcinogenic agents.25-26 Based on the low IC50 values, ethanolic extract was used for iso-lation of cytotoxic (bioactive) compounds. The mecha-nism of the cytotoxic effect of these bioactive compounds obtained from this plant is being studied.. Further study is needed since they have pronounced effect on some of the tumour biomarkers, which en-courages the utility in human studies. Moreover, there was no evidence suggesting side effects of the extracts towards normal cells, indicating a potent preventive agent for cancer. Conclusion These results indicate that A.gangeticus possess hepato-protective properties against chemical carcinogenesis. Acknowledgement The author would like to thank to IRPA Grant 06-02-04-0050 and UPM Short-term Grant, 2000. References 1. Anderson KN, Anderson LE. Glanze WD. Mosby’s
Medical Dictionary. Mosby Year Book. 1998. 2. Department of Public Health. Malaysia’s Health 1999.
Department of Public Health; Ministry of Health, 1999. 3. Yuan JM, Gago-Dominguez M, Castelao JE, Hankin JH,
Ross RK, Yu MC. Cruciferous vegetables in relation to renal cell carcinoma. Int J Cancer 1998; 77: 211-216.
4. Jain MG, Hislop GT, Howe GR, Ghadirian P. Plant foods, antioxidants, and prostate cancer from case control studies in Canada. Nutr Cancer 1999; 34: 173-184.
5. Cohen JH, Kristal AR, Stanford JL. Fruit and vegetable intakes and prostate cancer risk. J Natl Cancer Inst 2000; 92: 61-68.
6. Michaud DS, Spielgelman D, Clinton SK, Rimm EB, Willet WC, Giovannucci EL. Fruit and vegetable intake of bladder cancer in a male prospective cohort. J Natl Cancer Inst 1999; 91: 605-613.
7. Graham S, Dayal H, Swanson M, Mittelman A, Wilkinson G. Diet in the epidemiology of cancer of the colon and rectum. J Natl Cancer Inst 1978; 61: 709-714.
8. Agudo A, Esteve MG, Pallarés C, Martinez-Ballarin I, Fabregat X, Malats N, Machengs I, Badia A, Gonzalez CA. Vegetable and fruit intake and the risk of lung cancer in women in Barcelona Spain. Eur J Cancer 1997; 33: 1256-1261.
99.. Tavani A, Vecchia CL. Fruit and vegetable consumption and cancer risk in a Mediterranean population. Am J Clin Nutr 1995; 61 (Suppl): 1374S-1377S.
1100.. Packer L. Protective role of vitamin E in biological system. Am J Clin Nutr 1991; 65: 1050S-1055S.
1111.. Rajamnal D, Chandrasekhar U, Premakumari S, Saishree R. Biomedical Environment Science BES 1996; 9: 213-222.
12. Ali AM, Macken M, Hamid M, Lajis NH, El-sharkawy SH, Murakoshi. Antitumour promoting and antitumour activities of the crude extract from leaves of Juniperus chinensis. J Ethnopharmacology 1996; 53 : 165-169.
13. Conney AH, Wang ZY, Huang MT, Ho CT, Yang CS. Inhibitory effect of oral administration of green tea on tumourigenesis by ultraviolet light, 12-0-tetradecanoyl-phorbol-13-acetate and N-nitrosodiethyl-amine in mice. In: Lee W, Martin L, Charles BW, Gary K, eds. Cancer Chemoprevention. New York: CRC Press Inc, 1992; 362.
14. Solt D, Farber E. New principle for the analysis of chemical carcinogenesis. Nature 1976; 263: 701-703.
15. Speier CL, Wattenberg LW. Alterations in microsomal metabolism of benzo (a)-pyrene in mice fed butylate hydroxyanisole. J Natl Cancer Inst 1975; 55: 469 – 472.
16. Jacobs WLW. A colorimetric assay for gamma-glutamyl-transpeptidase. Clinica Chimica Acta. 1971 31: 175-179.
17. Jahan M, Buterworth PJ. Alkaline phosphatase of chick kidney. Enzyme. 1986; 35: 61 – 69.
18. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 1976; 72: 248 – 254.
19. Habig WH, Pabst MJ, Jacoby WB. Glutathione S-transferases. The first enzymatic step in mercapturic acid formation. J Biol Chem 1974; 249: 7130 – 7139.
20. Vessey DA, Zakim D. Regulation of microsomal enzymes by phospholipids. J Biol Chem 1972; 247: 3023 – 3028.
21. Asmah R, Wan Zurinah WN, Nor Aripin S, Abd. Gapor MT, Khalid AK. Long-Term Administration of Toco-trietnols and Tumour-Marker Enzyme Activities During Hepatocarcinogenesis in Rats. Nutrition 1993; 3: 229-232.
22. Asmah R, Wan Zurinah WN, Alini M, Zanariah J, Rosnani I, Khalid AK, Nor Aripin S, Effect of γ-Tocotrienol and α-Tocopherol on Blood Glutathione and Tumour Marker Enzymes during Chemical Hepatocarcinogenesis in the Rat. J Clin Biochem Nutr 1993; 15: 195-202.
23. Asmah R, Wan Zurinah WN, Abdul Gapor MT, Khalid BAK. Long-term tocotrienol supplementation and glutathione-dependent enzymes during hepatocarcino-genesis in the rat. Asia Pac J Clin Nutr 1993; 2: 129-134.
24. Huzaimah AS, Asmah R, Maznah I, Rozita R. Antioxidative Activities of Organic Extracts of Coleus blumei, Amaranthus gangeticus and Coleus amboinicus Compared to α-Tocopherol and Butylated Hydroxytoluene (BHT). Mal J Bio Mol Biol 2000; 5: 47-51.
25. Committee on Diet Nutrition and Cancer. Diet Nutrition and Cancer. Washington D.C: National Academy Press, 1982.
26. Hayatsu H, Arimoto S, Negishi T. Dietary inhibitors of mutagenesis and carcinogenesis. Mutat Res 1988; 202: 429-446.
401 Asia Pac J Clin Nutr 2004;13 (4): 401-408
Original Article Inhibitory effect of clonal oregano extracts against porcine pancreatic amylase in vitro Patrick McCue PhD1, Dhiraj Vattem PhD2 and Kalidas Shetty PhD2 1Program in Molecular and Cellular Biology, University of Massachusetts, Amherst, MA 01003 2Department of Food Science, University of Massachusetts, Amherst, MA 01003
Oregano (Origanum vulgare) is a rich source of natural phenolic antioxidants and has potential to be a source of nutritional ingredients for functional foods. Herbs such as oregano have long been used in food preservation and in traditional medicine in the treatment of common ailments and have potential for positive modulation of oxidation-linked diseases such as diabetes. One of the potentially important components of anti-diabetic activity by oregano extract is mild amylase inhibition by phenolic antioxidants to help contribute towards management of hyperglycemia. Previously, we reported the ability of rosmarinic acid, one of the principal phenolic components of oregano, to inhibit porcine pancreatic amylase (PPA) activity. Here, we investigated the effect of 50% ethanol extracts of eleven phenolic antioxidant-rich oregano clonal lines on the activity of PPA in vitro. To this end, we analyzed extract total soluble phenolic content by the Folin-Ciocalteu reagent method, rosmarinic acid (RA), protochatechuic acid (PA), quercetin, and p-coumaric acid (pCA) contents by HPLC, antioxidant activity as 1,1-diphenyl-2-picryl-hydrazyl (DPPH) radical scavenging, and PPA-inhibitory activity by incubation of the enzyme with clonal oregano extracts and characterization of the activity of the phenolic-bound enzyme. Clonal oregano extracts inhibited the activity of PPA in vitro by 9-57%. Amylase inhibition by oregano extract was associated with extract total phenolic content and RA, quercetin, PA, and pCA content, as well as extract antioxidant activity and protein content. Our finding that clonal oregano extracts can inhibit PPA supports a potential new functionality for oregano as a n anti-hyperglycemic agent. This provides an opportunity for a food-based strategy for modulation of starch breakdown to glucose, which could contribute to the management of hyperglycemia and diabetes complications in the long term.
Keywords: amylase inhibitors, oregano, herbal extracts, rosmarinic acid, hyperglycemia, diabetes mellitus, obesity Introduction Hyperglycemia is one of the major problematic symptoms associated with Type 2 diabetes mellitus, as well as pre-diabetes impaired glucose tolerance.1 Elevated postprandial glucose levels and persistent hyperglycemia can lead to cellular damage and is associated with the development of retinal, renal, neurological, and cardiovascular disease.2, 3 As oral anti-diabetic therapies that act on the liver risk hepatic dysfunction, much recent attention has been placed on the development of agents to combat hyperglycemia as an anti-diabetic therapy.4,5 Many anti-hyperglycemia agents currently in use, such as acarbose or metformin, act to inhibit or retard various reactions of glucose metabolism, are synthetic, and may have negative side-effects at high doses.2,6 A current goal in anti-diabetic research is to identify anti-hyperglycemic agents that are safe and that lack any negative side-effects. Traditional medicines from various cultures have given us potential prospects to explore the benefits of natural dietary remedies of symptoms associated with Type 2 diabetes mellitus, such as hyperglycemia. Traditional Indian and Chinese medicines have long used plant and herbal extracts as anti-diabetic agents.7,8 These plants are typically rich in phenolic compounds. Herbs used in traditional Indian medicine, such as Holy Basil (Ocimum sanctum) and oregano (Origanum vulgare), are high in rosmarinic acid (RA) content.9
Recently, we reported the strong inhibitory activity of RA against porcine pancreatic amylase, the enzyme responsible for the breakdown of starch into glucose.10 In light of this finding, we hypothesized that part of the anti-diabetic effect of RA-containing herb extracts may be due to the anti-amylase activity of RA. In the current study, we screened extracts of 11 clonal lines of oregano for anti-amylase activity. These clonal lines are of single seed origin and were isolated using tissue culture tech-niques from individual heterozygous seed.11 Further, the results were compared to measurements of oregano extract, total phenolic and RA content, as well as anti-oxidant (free-radical scavenging) activity. Methods Plant materials Shoots from specific clonal lines of oregano (Origanum vulgare) were generated previously from individual heterozygous seedlings following germination of a heterogeneous seed population.11, 12
Correspondence address: Dr Kalidas Shetty, Department of Food Science, Chenoweth Laboratory, University of Massachusetts, Amherst, MA 01003, USA Tel: (413)- 545- 1022; Fax:: (413)-545-1262 Email: [email protected] Accepted 21 May 2004
P McCue, D Vattem and K Shetty 402
Shoot culture of oregano Under aseptic conditions in a laminar flow hood, each existing clonal line was subcultured at room temperature on Petri plates containing Murashige & Skoog basal salt medium supplemented with 1 mg/ mL BAP, 3% sucrose, 1X Nitsch & Nitsch vitamins, and 0.3% gellan gum.13
Extraction procedure One gram of fresh shoot tissue (per clone) was submerged in 5mL of 50% EtOH and incubated for 3d at -20˚C. Next, each sample was diluted with 5mL of 50% EtOH, homogenized, and centrifuged 20 min at 10,000 rpm at 4˚C. The supernatant was used as the crude oregano extract. Extractions using 95% EtOH were performed for HPLC analyses of RA content in each clonal line.
Treatment of α-amylase with clonal oregano extracts Treatment of α-amylase was performed as previously described, with some modifications.10 Fifty mg of powdered porcine pancreatic α -amylase (Sigma) was added to 27 mL of dH2O and adjusted to pH 6.9. For each oregano extract, a volume equivalent to 400 μg total phenolic content was diluted to 3 mL with 50% EtOH, added to the above solution, and readjusted to pH 6.9. Amylase-oregano extract mixtures were incubated 24h at 4˚C with stirring. For comparison studies, 400 μg of synthetic antioxidants BHT and Trolox were in 50% EtOH were used. The control was 50% EtOH.
Characterization of α-amylase activity �-Amylase activity was determined by the method of McCue and Shetty10, using starch as a substrate in a colo-rimetric reaction using 3,5-dinitrosalicylic acid. A stan-dard curve was generated for the splitting products (reducing groups) using D-(+)-maltose monohydrate.10 Activity was calculated as units/ mg protein, where 1 unit was defined as the amount of enzyme required to liberate 1 μmol of maltose under assay conditions. Protein content was determined using the Bio-Rad protein assay kit. Data was reported as amylase inhibition (AI) index values,
defined herein as the ratio of the amylase activity of the control (enzyme alone) to that of the enzyme/ clonal extract mixture.14 Values greater than 1 indicate α-amylase inhibition.
Total soluble phenolic content assay The total soluble phenolics content in each extract was determined using a previously described method.13 A phenolic standard curve was established at 725nm with (+)-catechin (25-200 μg/ mL) of in 95% EtOH.
Phenolic content determination by HPLC One mL of each 95% EtOH clonal oregano extract adjusted to 200 μg total phenolic content per mL was passed through a 0.45 μm filter. HPLC analysis was performed using an Agilent 1100 series system equipped with an autosampler, a variable wavelength diode array detector (set at 333nm for RA; 306nm for all other phenolics), and a Zorbax SB-C18 column. The injection volume was 5 μL and flow rate 1mL/min. The solvents used were 10 mM phosphoric acid, pH 2.5, and MeOH. The MeOH concentration was increased to 60% for 8 min, to 100% for 7 min, then to 0% for 3 min, and maintained for another 7 min. Pure RA, PA, quercetin, and pCA was used for a standard curve and for sample spiking.
Antioxidant activity Antioxidant activity was determined as percent sca-venging of 1,1-diphenyl-2-picryl-hydrazyl (DPPH) ra-dicals.15 A volume equivalent to 200 μg total phenolic content for each extract was diluted to 1mL with 50% EtOH and mixed with 1mL of 0.1 mM ethanolic DPPH solution. After 30 min at RT, absorbance at λ = 517nm was measured.
Protein concentration The protein concentration of each 50% ethanol extract of clonal oregano was determined using the Bio-Rad protein assay kit. Bovine serum albumin (BSA) was used for the standard.
Figuvalu
0
0.5
1
1.5
2
2.5
3
Control O-1 O-9 O-11Y O-11M O-12 O-17 OK-17 O-23 O-24 O-26 GO-19-1 Oregano Clonal Line
AI IIndex
a
b, h
c, f, h
d
c, h
e, h b, h
e, f, h
b, g, h g, h
c
h
re 1. Inhibition of porcine pancreatic amylase activity in vitro by extracts of clonal oregano. Amylase inhibition (AI) index es above 1 indicate inhibitory activity. Data with different letters are significantly different by ANOVA at P <0.05.
403 Inhibitory effect of clonal oregano extracts against porcine pancreatic amylase in vitro
Descriptive Correlation Excel XP) fof possibly p
Protein struRasMol (Worg/) Molecvisualize thamylase (Pro Results andEffect of cloIn this studyto have stro(Fig.1). Thefor extract O0.28 and cactivity. Eihad AI indexexperimentsapproximate Relationshipand α -amylPrevious rephenolic com
Figure 2
Oregano0-0-
0-10-10-10-1
OK0-20-20-2
GO-a Data are µgc '_' symbol =
Table 1. Co
statistics coefficients were determined (using Microsoft or activities/ compounds that were suspected ossessing synergistic action.
cture analysis indows Version 2.7.2.1; http://openrasmol. ular Graphics Visualization Tool was used to e published structure for porcine pancreatic tein Data Bank code: 1DHK).16-18
Discussion nal oregano extracts on α-amylase activity , extracts of clonal oregano lines were found ng inhibitory activity against PPA in vitro
strongest anti-amylase activity was observed -24 which had an AI index value of 2.32 ±
orresponded to 57% inhibition of enzyme ght of the 11 clonal oregano extracts tested values greater than or equal to 1.5. For these
, an AI index value of 1.5 corresponded to ly 33% α-amylase enzyme inhibition.
between clonal extract phenolic content ase inhibition search has noted the ability of various pounds to inhibit α-amylase activity.19, 20
Therefore, to further define the nature of the amylase inhibition mechanism in response to clonal oregano extracts, the extract anti-amylase activities were com-pared to extract total soluble phenolic contents (Fig.2). Anti-amylase activity appeared to be related to phenolic content in some extracts but not in others. In Figure 2, anti-amylase activity of extracts O-1 to O-17 was strongly linked to total soluble phenolic content (correlation coefficient = 0.85). However, anti-amylase activity of the remaining extracts (OK-17 to GO-19-1) was largely un-related to total soluble phenolic content (correlation coefficient = -0.51). As some of the extracts in this latter group still possess strong anti-amylase activity, it is possible that other factors, such as phenolic synergies or synergies to other extract components (i.e. proteins), may be involved. Relationships between α- amylase inhibition and specific phenolic contents in clonal oregano extracts Oregano is known to be a rich source of RA, a biphenolic compound previously reported to possess strong anti-amylase activity.10 To explore potential phenolic com-pounds involved in the anti-amylase activity of clonal oregano extracts, we determined the concentration of selected phenolic compounds known to occur in oregano and related herb species. Table 1 shows the contents of
0
100
200
300
400
500
600
700
800
O-1 O-9 O-11Y O-11M O-12 O-17 OK-17 O-23 O-24 O-26 GO-19-1
Clonal Oregano Line
Phen
olic
con
tent
( µg
cate
chin
/mL)
0
0.5
1
1.5
2
2.5
AI In
dex
Total Phenolics
AI Index
. Comparison of clonal extract anti-amylase activity to total phenolic content and AI, amylase inhibition.
Clone RA PA Quercetin pCA 1 64.5 ± 0.5 -c 12.0 ± 0.1 9 42.6 ± 0.4 - 8.8 ± 0.04 0.15 ± 0.01 1Y 114.0 ± 1.4 - 9.9 ± 2.3 0.53 ± 0.09 1M 116 ± 1.4 - 9.5 ± 0.1 - 2 101 ± 0.2 - 16.5 ± 0.2 1.10 ± 0.02 7 184 ± 2.3 - 12.6 ± 0.1 0.99 ± 0.02
-17 136 ± 1.4 - 18.4 ± 0.4 0.61 ± 0.04 3 76.0 ± 1.1 - 15.3 ± 1.3 - 4 53.1 ± 0.1 - 5.5 ± 0.3 - 6 31.5 ± 0.3 - 8.9 ± 0.4 0.24 ± 0.01
19-1 72.6 ± 0.2 8.5 ± 0.1 15.2 ± 0.01 - phenolic per 200µg of total phenolic content per extract. b Data are mean ± SD of three replicates. not detected, or of negligible amount. RA, rosmarinic acid; PA, protocatechuic acid; pCA, para-Coumaric acid.
ncentrationa of select phenolic compounds in extractsb of clonal oregano by HPLC
P McCue, D Vattem and K Shetty 404
RA, PA, quein samples (extract. Thewas RA. Tquercetin. pCPA was dete The anti-extracts werand pCA, boanti-amylasenot correlateas shown in1, O-9, O-strongly corThere was between antall but twoextracts (Figextracts O-1
0
0.5
1
1.5
2
2.5
Control O-1 O-9 O-11Y O-11M O-12 O-17 OK-17 O-23 O-24 O-26 GO-19-1
Oregano Clonal Extract
AI In
dex
0
20
40
60
80
100
120
140
160
180
200
Ros
mar
inc
Acid
Con
tent
, µg/
mL
AI Index
Rosmarinic Acid
Figure 3. Comparison of clonal oregano extract anti-amylase activity to rosmarinic acid content and AI amylase inhibition.
0
2
4
6
8
10
12
14
16
18
20
Control O-1 O-9 O-11Y O-11M O-12 O-17 OK-17 O-23 O-24 O-26 GO-19-1
Clonal Oregano Extract
Que
rcet
in, µ
g/m
L
0
0.5
1
1.5
2
2.5
AI In
dex
Quercetin
AI Index
Figure 4. Comparison of clonal oregano extract anti-amylase activity to quercetin content and AI amylase inhibition.
rcetin, and pCA that were detected by HPLC 200μg total phenolics) of each oregano clonal major phenolic detected in all of the samples he second highest phenolic detected was A was detected in 6 of the 11 extracts, while
cted in only one (GO-19-1). amylase activities of the clonal oregano e compared to the contents of RA, quercetin, th individually and in sum with PA. Overall, activity of the clonal oregano extracts did well with RA content (coefficient = - 0.01),
Figure 3. However, for 4 of the extracts (O-17, and OK-17) anti-amylase activity was related to RA content (coefficient = 0.62). a strong correlation (coefficient = 0.75)
i-amylase activity and quercetin content for (OK-17 and O-24) of the clonal oregano . 4). Anti-amylase activity in clonal oregano 7 and OK-17 were positively correlated to
pCA content (coefficient = 1.0), however, none of the other extracts were associated with pCA content (Fig. 5). For 7 of the 11 clonal oregano extracts (O-1, O-9, O-12, O-17, OK-17, O-26, and GO-19-1), anti-amylase activity was strongly correlated (coefficient = 0.81) to the sum total phenolic content of RA, quercetin, pCA, and PA, as detected by HPLC (Fig. 6). As the correlations between clonal extract anti-amylase activity and phenolic contents are not absolute, we are led to believe that other extract components may be involved. Relationship between clonal extract antioxidant activity and α-amylase inhibition Three of the best known mammalian α-amylase inhibitors are acarbose, a carbohydrate inhibitor, Tendamistat, a proteinaceous inhibitor from Streptomyces, and α-amylase inhibitor 1, a lectin-like proteinaceous inhibitor from the common bean Phaseolus vulgaris.18 Structural analysis indicates that the proteinaceous inhibitors interact with
405 Inhibitory effect of clonal oregano extracts against porcine pancreatic amylase in vitro
with α-amythe carbohyresidues in to inhibit published sthe occurrethe outer suthat reductcould causaffect the a Many p
0
0.2
0.4
0.6
0.8
1
1.2
Control O-1 O-9 O-11Y O-11M O-12 O-17 OK-17 O-23 O-24 O-26 GO-19-1
Oregano Clonal Extract
pCA,
µg/
mL
0
0.5
1
1.5
2
2.5
AI In
dex
pCA
AI Index
Figure 5. Comparison of clonal oregano extract anti-amylase activity to pCA content. AI, amylase inhibition; pCA, para-coumaric acid.
0
50
100
150
200
250
Control O-1 O-9 O-11Y O-11M O-12 O-17 OK-17 O-23 O-24 O-26 GO-19-1
Oregano Clonal Extract
Phen
olic
Con
tent
, µg/
mL
0
0.5
1
1.5
2
2.5
AI In
dex
HPLC Phenolic Sum
AI Index
Figure 6. Comparison of clonal oregano extract anti-amylase activity to the sum total content of the phenolics rosmarinic acid,
quercetin, para-coumaric acid, and protocatechuic acid, as detected by HPLC. AI, amylase inhibition.
lase and block access to the active site, while drate inhibitor acarbose directly interacts with the active site (Glu233, Asp300, and Asp197) the enzyme.18 In this work, analysis of the tructure of PPA (PDB code: 1DHK), revealed nce of 5 sets of disulfide bridges that occur on rface of the enzyme (Fig. 7). We hypothesize
ion of these cysteine residues by antioxidants e structural alterations that may negatively ctivity of the enzyme. henolic compounds are known to possess
antioxidant activity in vitro and in vivo.21-23 Since oregano is a rich source of the phenolic antioxidants RA and quer-cetin, we sought to explore whether antioxidant activity of the clonal oregano extracts might be involved in the anti-amylase activity. Antioxidant activity of the clonal ore-gano extracts was determined as percentage DPPH free-radical scavenging. Additionally, two synthetic anti-oxidants, BHT and Trolox, were also tested for anti-amylase activity. Anti-amylase activity of 6 of the 11 clonal oregano extracts and the 2 synthetic antioxidants was found to strongly correlate (coefficient = 0.87) to
P McCue, D Vattem and K Shetty 406
Figure 7. Structural representation of porcine pancreatic alpha-amylase. Cysteine residues are shown in green (chain B) and in blue (chain A). Three important residues of the active site (Glu233A, Asp300A, Asp197A) are shown in red (center of molecule). Arrows indicate each of the 5 disulfide bridges located on the outer surface of the enzyme.
0
0.5
1
1.5
2
2.5
Con
trol
O-1
O-9
O-1
1Y
O-1
1M
O-1
2
O-1
7
OK-
17
O-2
3
O-2
4
O-2
6
GO
-19-
1
BHT
Trol
ox
Clonal Oregano Extract
AI In
dex
0
10
20
30
40
50
60
70
80
90
100
% D
PPH
Sca
veng
ing
AI IndexAntioxidant Activity
Figure 8. Comparison of clonal oregano extract anti-amylase activity to antioxidant activity given as percentage DPPH free-radical
scavenging. AI, Amylase inhibition.
antioxidanFig. 8). Hextracts (Omechanismt activity (O-1 to O-11M, O-26, and GO-19-1; owever, as the activities of the remaining 5 -12 to O-24; Fig. 8) did not correlate, other s including direct interactions remain possible.
Relationship between clonal extract protein content and α-amylase inhibition Phenolic compounds are known to bind proteins, including enzymes.24 As the inhibition profile for clonal
407 Inhibitory effect of clonal oregano extracts against porcine pancreatic amylase in vitro
oregano extralated to phenothe involvemobserved PPAextract proteiOverall, thereprotein conteFor the first 6the correlatioWhen the datadded, the costrong (coeffprotein may mediated by protein-pheno Conclusion In this study, to have stronPPA inhibitio11Y) to 57%related to excontent of thquercetin, PAspecific phenoPPA inhibiticorrelated webetter with quextract was gsynthetic antiinhibit PPA, to the oregaoregano extraprotein contephenolic syne
The findextracts inhibhow these co
0
0.5
1
1.5
2
2.5
Control O-1 O-9 O-11Y O-11M O-12 O-17 OK-17 O-23 O-24 O-26 GO-19-1
Clonal Oregano Extract
AI In
dex
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
BSA
Equi
vale
nts,
mg/
mL
AI Index
Protein
Figure 9. Comparison of clonal oregano extract anti-amylase activity to extract protein content. AI, Amylase inhibition; BSA, bovine serum albumin.
cts against PPA was not absolutely corre-lic content, there remains a possibility for ent of protein-phenolic synergies in the inhibition. Therefore, we compared the
n content to extract PPA inhibition (Fig. 9). was a moderate correlation between extract nt and PPA inhibition (coefficient = 0.36). extracts listed in Figure 9 (O-1 to O-17),
n was much stronger (coefficient = 0.74). a for 2 additional extracts (O-23, O-24) was rrelation decreased somewhat but remained icient = 0.65). These results suggest that be involved in the PPA inhibitory activity clonal oregano extracts, perhaps through lic synergies.
extracts of clonal oregano lines were found g inhibitory activity against PPA in vitro. n varied by extract and ranged from 9% (O- (O-24). Generally, PPA inhibition was
tract total phenolic content and the sum e four major phenolics in oregano, RA, , and pCA. We were unable to identify a lic as solely responsible for oregano extract
on, as some extract anti-PPA activities ll with RA content, while others correlated ercetin content. PPA inhibition by oregano
enerally related to antioxidant activity. The oxidants BHT and trolox were found to although at intermediate efficacy compared no extracts. Finally, PPA inhibition by ct was also found to correlate with extract nt, suggesting a possible role for protein-rgies. ing that phenolic antioxidant-rich oregano it α-amylase activity may help to explain nstituents of traditional Asian anti-diabetes
medicines could confer their therapeutic benefits. Further, the results of our study suggest the potential for oregano as a part of functional foods for the dietary control of hyperglycemia and support its inclusion as a natural, safe, anti-diabetic therapy for modulation of Type 2 diabetes mellitus. References 1. Brownlee M. Biochemistry and molecular cell biology of
diabetic complications. Nature 2001; 414: 813-820. 2. Haffner SM. The importance of hyperglycemia in the non-
fasting state to the development of cardiovascular disease. Endocrine Rev 1998; 19 (5): 583-592.
3. Haller H. The clinical importance of postprandial glucose. Diabetes Res Clin Pract 1998; 40: S43-S49.
4. Mudaliar S, Henry RR. New oral therapies for type 2 diabetes mellitus: the glitazones or insulin sensitizers. Annu Rev Med 2001; 52: 239-257.
5. Costacou T, Mayer-Davis EJ. Nutrition and prevention of type 2 diabetes. Annu Rev Nutr 2003; 23: 147-170.
6. Ohmura C, Tanaka Y, Mitsuhashi N, Atsum Y, Matsuoka K, Onuma T, Kawamori R. Efficacy of low-dose metfor-min in Japanese patients with type 2 diabetes mellitus. Curr Ther Res 1998; 59 (12): 889-895.
7. Chen H, Feng R, Guo Y, Sun L, Jiang J. Hypoglycemic effects of aqueous extract of Rhizoma Polygonati Odorati in mice and rats. J Ethnopharm 2001; 74: 225-229.
8. Grover JK, Yadav S, Vats V. Medicinal plants of India with anti-diabetic potential. J Ethnopharm 2002; 81: 81-100.
9. Yang R, Shetty K. Stimulation of rosmarinic acid in shoot cultures of oregano (Origanum vulgare) clonal line in response to proline, proline analogue, and proline precursors. J Agric Food Chem 1998; 46 (7): 2888-2893.
10. McCue P, Shetty K. Inhibitory effects of rosmarinic acid extracts on porcine pancreatic amylase in vitro. Asia Pac Clin Nutr 2004; 13(1): 101-106.
11. Eguchi Y, Curtis OF, Shetty K. Interaction of hyperhydricity-preventing Pseudomonas sp. with oregano (Origanum vulgare) and selection of high phenolics and rosmarinic acid-producing clonal lines. Food Biotechnol 1996; 10 (3): 191-202.
P McCue, D Vattem and K Shetty 408
12. Shetty K, Curtis OF, Levin RE, Witkowski R, Ang W. Prevention of vitrification associated with in vitro shoot culture or oregano (Origanum vulgare) by Pseudomonas spp. J Plant Physiol 1995; 147: 447-451.
13. McCue P, Shetty K. Clonal herbal extracts as elicitors of phenolic synthesis in dark-germinated mungbean for improving nutritional value with implications for food safety. J Food Biochem 2002; 26: 209-232.
14. Correia RTP, McCue P, Vattem DA, Magalhães MMA, Macêdo GR, Shetty K. Amylase and Helicobacter pylori inhibition by phenolic extracts of pineapple wastes bioprocessed by Rhizopus oligosporus. J Food Biochem 2004; In press.
15. McCue P, Horii A, Shetty, K. Solid-state bioconversion of phenolic antioxidants from defatted soybean powders by Rhizopus oligosporus: role of carbohydrate-cleaving enzymes. J Food Biochem 2003; 27 (6): 501-514.
16. Sayle RA, Milner-White EJ. RASMOL: biomolecular graphics for all. Trends Biochem Sci 1995; 20(9): 374-376.
17. Bernstein HJ. Recent changes to RasMol, recombining the variants. Trends Biochem Sci 2000; 25 (9): 453-455.
18. Bompard-Gilles C, Rousseau P, Rougé P, Payan F. Substrate mimicry in the active center of a mammalian �-amylase: structural analysis of an enzyme-inhibitor complex. Structure 1996; 4: 1441-1452.
409 News & Views
News & Views
Symposium and workshop on healthy lifestyle programs for weight management Rodolfo F. Florentino, MD, PhD Nutrition Foundation of the Philippines, Quezon City, Philippines Introduction Overweight and obesity are becoming public health concerns in many countries in Asia. With the change in lifestyle now occurring in the population, not only is the problem growing among adults, but the problem now seems to be creeping into the children population. The shift toward a more sedentary lifestyle and the con-sumption of calorie-dense diets, coupled with improving economic base and environmental conditions, is appa-rently contributing to the transition from the problem of undernutrition to overnutrition in many segments of the population. While public health strategies for the control of this problem are being put in place in some countries of the region, the issues of impact and sustainability of these strategies remain to be elucidated. School-based inter-vention programs are an attractive strategy to prevent the problem from spreading, but again their impact and sus-tainability remain at issue. The Second Asia-Oceania Conference on Obesity organized by the Malaysian Association for the Study of Obesity (MASO) on September 7-9, 2003, in Kuala Lumpur, Malaysia, provided an opportunity for the Inter-national Life Sciences Institute Southeast Asia Region (ILSI SEAR) to pursue its program on promoting healthy lifestyle and physical activity for weight management. ILSI SEAR sponsored a symposium within the MASO conference focusing on evaluating the impact and sus-tainability of public health strategies for the control of obesity, followed by a post-conference workshop on school-based intervention programs directed to the pre-vention and control of obesity in children.
Symposium: evaluating impact and sustainability of healthy lifestyle programs for weight management The Symposium on Healthy Lifestyle Programs for Weight Management was chaired by Dato’ Dr. M. Jegathesan, Deputy President of the Olympic Council of Malaysia, and co-chaired by Dr. Richard Winsley, Uni-versity of Exeter of the United Kingdom. In his paper, Dr M Jegathesan declared that while the principles of health-enhancing activities including exercise and physical fitness are fairly well understood, there are barriers to their application, both at the individual and community levels. At the individual level, behaviour change does not occur because of lack of conviction and motivation, inertia, lack of resources, and presence of
competing interests. It requires the personal decision of the individual to make the change – demonstrated by the commitment to see it through and the motivation to sustain it. In fact, health-enhancing activities can be readily integrated into most people’s lives by making relatively minor adjustments to their daily routines. At the community level, barriers to promotion of physical activity include lack of political will, contrary interests, inadequate resources, and inappropriate criteria for prio-rity setting. Together with promotive policies, programs and facilities; local governments and public authorities must also build the awareness, provide the know-how, and promote an enabling and facilitating environment. Promotion of a healthy lifestyle should be broad-based and integrated into national policies involving many sectors such as health, development, education, housing, town planning and transportation. This requires inter-sectoral and even ministerial collaboration and cooper-ation. With modest investment, Dr. Jegathesan concluded that great returns could be achieved in terms of improved health and less health care costs. Ms. Debra L. Kibbe, Acting Executive Director, ILSI Center for Health Promotion, USA, emphasized the important role of schools in promoting healthy nutrition and physical activity habits in children and adolescents. Ms. Kibbe reviewed several school-based programs that integrate physical activity and nutrition education and have produced successful results. She cited the Healthy Start and the Animal Trackers programs for preschool children focusing on development of healthful nutrition habits and gross motor skills; CATCHTM (Coordinated Approach To Children's Health) for K – 8th grade children which showed positive differences in self-reported daily energy intake and vigorous activity; the TAKE 10! program for K – 5th grade students which demonstrated reduced off-task and fidgeting behaviors among children; SPARK (Sports, Play and Active Recreation in Kids) for pre-K to middle school with its focus on quality physical education and self-monitoring; TEENS (Teens Eating for Energy and Nutrition in Schools) for 7th graders with its peer-led nutrition education strategy; and GEMS (Girls Health Enrichment Multi-site Program) for 8-10 year old African American females, which showed positive outcomes in terms of BMI, after school physical activity, and reduced tele-vision viewing. Finally, Ms. Kibbe described the School Health Index (SHI) developed by the Centers for Disease Control and Prevention in the US. The eight SHI modules (http://www.cdc.gov/nccdphp/dash/SHI/) address: Health Policies and Environment, Health Education, Physical Education and Other Physical Activity Programs, Nutrition Services, School Health Services; School Counseling, Psychological, and Social Services; Family and Community Involvement, and Staff Health Promotion. In summary, school-based programs in the US have demonstrated reduced hours of television watching, increased frequency and duration of physical activity, decreased time off task in the classroom, decreased intake of total and saturated fats, increased consumption of fruits and vegetables, slower rate of increase in BMI percentile, and improved blood lipid levels. Ms Kibbe concluded that school interventions
News & Views 410 should be culturally and linguistically sensitive, grade- and age-appropriate, comprehensive in coverage of health issues, convenient, and low in cost. Dr Andrew P Hills of the Queensland University of Technology, Australia, addressed the importance of a multisectoral approach in promoting sustainable physical activity programs. He described three levels of physical activity prevention programs in the form of a pyramid: Level 3: Public health approach to physical activity promotion, which is the area of health professionals; Level 2: Limited advice such as generalized activity pre-scription; and Level 3: Individualized approach through specific exercise prescription, which is the province of exercise physiologists. Dr Hills pointed out salient barriers to physical activity promotion according to socio-economic level. For those in the higher socio-economic levels, an ‘unpredictable lifestyle’ and long and erratic work hours prevent them from doing physical exercise; for those in the middle level, ‘lack of interest’ is the main factor; for those in lower classes, ‘inconvenient access’, geographical location, and physical and psychological health are the principal barriers. For all socio-economic levels, inclement weather or lack of time is often to blame. Thus, physical activity intervention programs should be tailored to the needs of specific sub-groups, such as the different socio-economic classes, men vs. women, older vs. young adults. Dr Hills described a simple evaluation framework for assessing the impact of public health programs of this nature: determining Reach, assessing Effectiveness, determining the extent of Adoption and Implementation, and assessing sustainability and Maintenance (REAIM). Dr Hills concluded that the key question is how to promote physical activity and prescribe exercise to maximize enjoyment, adherence, and tolerance. Dr Teh Kong Chuan, Director of the Sports Medicine & Sports Science Division, Singapore Sports Council, described the Singapore experience in promoting physical activity at the national level. In Singapore, physical activity is promoted by many organizations, with the Singapore Sports Council as the main player. The Council has been promoting sports in the country since its inception in 1973, through the provision of sports facilities, the implementation of ‘Sports for All’ programs and activities, and marketing of sports to the general population. The provision of sports facilities follows the Master Plan of Sports Facilities first formulated in 1975. In 1985, the Master Plan called for providing sports facilities within 3 km radius of the homes of most Singaporeans. In 1996, the Sports for Life program was launched, targeting the less physically active population including senior citizens, housewives and working adults. A nationwide survey in 2001 showed that 38% of the population played sports one or more times a week (16% three or more times a week), compared to 34% in 1997, and daily physical activity averaged 65 minutes per day (compared with 88 minutes per day in 1997), with household chores accounting for 24.6 minutes. Dr Teh concluded that with the activities of daily living becoming less physical in nature, participation in physical activity needs to be further improved. The
medical personnel should play a more prominent role in promoting regular physical activity. Discussion The discussion that followed the symposium papers dealt with the issue of getting the involvement of teachers and parents. Getting the cooperation of the schools and teachers should involve a top-down approach where both physical fitness and academic achievement are empha-sized starting with the principal, and finding a champion in the school for physical education. Getting the teachers to be physically active themselves is also important. Involving the parents is more difficult. Encouraging parents to limit television viewing time and encouraging children to reduce hours of total screen time (computer, video, and TV) are simple, specific, and proven messages. Policy makers need convincing evidence that physical inactivity is in fact increasing and that physical activity is effective in promoting health. Several indicators were mentioned such as monitoring increase in steps taken per day, weight maintenance in children, and examining specific health outcomes including mental and psy-chosocial concerns among children. Food provided in school canteens should be examined and modified, for example, portion size, nutrient content, and pricing of foods that are being recommended to improve health. The issue of vending machines is controversial in some schools in the US because some of the proceeds actually go to supporting physical education, sports or other extracurricular activities. The challenge of intervention sustainability is still one area that requires further research. Strategies that have proven effective in some programs were mentioned, including convincing the principals and teachers on the link between health and learning; adequate training of the teachers; having a champion at the local level; and utilizing curriculum tools that are easy to integrate with the existing curriculum. The Post-MASO conference satellite workshop on school-based intervention programs for healthy weight management The Satellite Workshop that immediately followed the Second Asia-Oceania Conference on Obesity provided a review of successful strategies and programs in schools that aim to prevent and manage overweight and obesity in children and a discussion of the lessons and challenges learned from such programs. The objective of the Work-shop was to encourage nutrition planners and health edu-cators to include balanced and science-based pro-grams that teach sound nutrition and encourage physical activity among school children. As in the above Symposium on Evaluating the Impact of Healthy Weight Management, the Workshop was chaired by Dato’ Dr. M. Jegathesan. Mrs.Yeong Boon Yee, Executive Director, ILSI SEAR, welcomed the 24 participants from Brunei, Malaysia, Philippines, Singapore, and Thailand, together with invited experts from USA and Australia. Debra Kibbe observed that in the US, physical
411 News & Views
education and physical activity opportunities in schools are on the decline. For example, in order to increase time for academic instruction, many US schools have reduced or eliminated recess apparently with the assumption that sedentary environment will foster learning. On the contrary, Ms. Kibbe presented studies that showed the link between physical activity and academic achievement. Ms. Kibbe cited the Planet Health program for 6th – 8th graders, with its goal of being physically active, watching TV for less than two hours and eating fruits and vegetables everyday, together with eating fat in moderation; a study of the California Department of Education, which showed that as fitness scores increased, increase in reading and math scores followed; the Brain Gym program which demonstrated the link between movement and academic achievement; the TAKE 10! program, which showed reduction in off-task time and fidgeting; as well as other studies that showed improved classroom mood and improved memory with increased physical activity. Ms Kibbe concluded that a positive relationship does exist between physical activity and nutrition, and children’s ability to perform academically. Areas that need further research include the effect of micronutrient deficiency and physical performance and gross motor skill development, the influence of exercise on specific academic objectives, standardized test scores in very fit children vs those that are unfit, and how the effect of physical activity pro-motion in childhood tracks into adulthood. Dr Andrew P Hills expressed the view that the school is an under-used setting for providing and facilitating innovative physical activity strategies in children. Dr. Hills emphasized the need for favorable school policies, improved teacher training and support, improved curri-culum, and provision of physical facilities that promote physical activities. However, there should be a greater interaction among parents, teachers, and children, such that parents should be actively involved in any school program promoting physical activity in children. Among the strategies that engage parents in promoting physical activity of their children, Dr Hills cited the “Walk to/Walk from School” program, aside from actually playing with their children, providing encouragement, serving as active role models, and providing transport to the children’s activity settings. In addition to the schools and parents, Dr Hills pointed to the Early Childhood Care and Education Centers working together with the whole community, in promoting physical activity in young children. Dr Hills suggested the application of the Natural Learning Theory Framework for Learning to the promotion of physical activity in children. Dr Hills recommended that the physical activity setting should provide an environment that is relaxed and friendly, supportive and caring, fun and enjoyable. Other strategies include focusing on health instead of weight per se, fostering enjoyment, encouraging “can do” mentality, and downplaying shame and blame associated with body fatness. During the discussion, Dr Hills emphasized that the key for involving the schools is the active leadership of a local champion to promote physical activity in schools.
Case studies The next session in the workshop dealt with case studies of on-going school-based programs including their strategies and methods, an assessment of their impact, and the problems and challenges facing them. Ms Kibbe described the TAKE 10!® program in greater detail. TAKE 10! is a classroom-based program designed for elementary school children in kindergarten through fifth grade. The curriculum tool integrates 10-minute periods of physical activity with academic con-tent and learning objectives in math, science, language arts, and social studies. The program is now distributed to over 4500 schools in 46 states in the US. An eva-luation of the program indicate that teachers use the program on average of 3-5 times per week, students demonstrate less fidgeting and time off-task, and that each activity takes 35-50 kcal of energy expenditure. The TAKE 10!® Middle School program for students Grades 11 to 14 (6th through 8th in the US), encourages students to accumulate at least 10,000 steps daily while taking an imaginary trek in the United States, at the same time addressing important learning objectives. Using an electronic step counter, each student monitors his/her physical activity throughout the day. Results of quali-tative evaluations suggest an average increase of 21% of the reported daily step counts from baseline. Focus group discussions also indicated favorable reactions from teachers and students. Ms Kibbe also described the Animal Trackers Pre-School program, which is a gross motor skill development program for 3 to 5 year old children. The program activities integrate various motor skills with preschool content and learning areas. The implementation results demonstrated an average increase of structured physical activity by 50 minutes per student per week. Dr Andrew Hills pointed to other strategies arising from their experience in Australia, such as prescribing activities in and out of school, e.g weekend activities; providing teachers with adequate tools for implemen-tation; training teachers as well as parents; and providing teachers and parents with their main goal. The objective is to fill the gap that is not being provided by physical education in school. Dr Sangsom Sinawat, Ministry of Health, Thailand, described the overweight and obesity control program among school children in the country following the survey in 12 big cities in 2001. The strategies being employed include creation of awareness among policy makers, administrators, health care personnel and the public sector; meeting with school administrators; deve-lopment of support tools such as manuals; putting up weight reduction camp prototype; and setting up moni-toring and evaluation mechanism. Lunch menus have been developed for the school canteens in project schools; overweight children are asked to submit food record everyday; and physical exercises in school and after school are extensively promoted. Parents are involved as partners for the project through the PTAs. After the program has been set up, there has been a slight improvement in prevalence of overweight in school
News & Views 412 children, from 13.6% to 12.3% (N = 1.2,984). The plan is to expand the program to cover all schools in the country and integrate the program with health promotion scheme to make it sustainable. Dr Kallaya Kijboonchoo, Institute of Nutrition, Mahidol University, Thailand, described the Nutrifit pro-gram, a nutrition education and physical fitness training package developed by INMU. The program was tested in two private schools in Bangkok Metropolitan Area and two government provincial schools. A fitness corner with physical fitness equipment and a healthy fitness zone chart display were set up in each school, except that in the intervention schools, INMU staff facilitated the nutrition and physical activity training components every other week for 7 months. No differences in nutritional status and in physical fitness test using Fitnessgram (developed by Cooper Institute for Aerobic Research in the US) were found between the control and intervention groups. Among the lessons learned from the project are the importance of involving the school authorities, peer pressure, and access to a physical fitness corner and equipment. Dr Ruzita Abdul Talib, Department of Nutrition and Dietetics, Universiti Kabangsaan Malaysia, discussed the impact of a school-based nutrition education program on nutrition knowledge and food habits of Malaysian school children, which is part of the Healthy Lifestyle in Malaysian Children (HELIC) study. The education package consists of five topics focused on healthy food pyramid, fiber and health, healthy menu, food and disease, and healthy lifestyle. The tools used included compact disk, flipchart, food picture card, pamphlet and poster, together with a teacher’s manual and student activity book. The project was tested in four secondary schools serving as intervention schools, and four other schools serving as control. In the intervention schools, three to four classes were randomly selected to receive the education lessons for five weeks. Evaluation of the program showed significantly higher scores in nutrition knowledge and eating habits after 6 months of inter-vention among the students in the intervention schools compared to the control, while no significant differences were found in the students’ average weight, height and food intake. Mrs Anna Jacob of ILSI Southeast Asia Region, Singapore, described the Power Kids Eat Smart & Power Kids On the Go package that has now been made available to all 210 primary schools in Singapore as a tool for the Trim and Fit Club (TAF) being run by the Singapore schools. The Power Kids Eat Smart which is the nutrition module in the package, uses a story format meant to encourage children to adopt healthy eating habits by relaying messages in a fun way, accompanied by activity sheets and goal cards. The Power Kids On the Go package provides teachers with an instruction manual and activity cards for warm up and cool down exercises as well as aerobic and game activities. In 2001 when the package was made available to all Singapore schools, the teachers received it very enthusiastically. The post-program survey conducted by ILSI SEAR in 2002 showed mostly positive responses from the teachers and children. At the same time, the survey also showed areas
where the program could be improved. Among the challenges facing the program are the high turnover of school staff, the need to retrain teachers and to revitalize the program for ease in implementation, and the need for greater parental involvement. Dr Visal Kantaranakul of the Board of Rehabilitation Medicine, Ramathibodi Hospital, Thailand, described the results of the pilot implementation of the Powerkids Program in a convent school in Thailand after the trans-lation of the package by Dr Suttilak and further adapted to fit Thai culture and the curriculum with the help of an advisory committee. According to the teachers, the parents have no time to complete the food diaries and physical activity logs, and the program competes for time with other school activities. Before the program is expanded to other schools, there is a need for more support from the Ministry of Education and to further revise the Powerkids package for clearer understanding. The sustainability of the program needs to be planned. The last of the case studies presented in the workshop was described by Dr Rodolfo F. Florentino of the Philippine Association for the Study of Overweight and Obesity (PASOO). The Whiz Kids Through Fitness project of PASOO patterned after the Powerkids program of Singapore, is a program directed to schoolchildren to promote healthy lifestyle, with emphasis on physical activity and proper diet. The project is being piloted among Grades I to III pupils in one private school in Manila. The program consists of a physical activity component and a nutrition education component that are integrated in the curriculum. The nutrition education component includes ten simple nutrition messages spread throughout the school year, while the physical activity component consists of physical activities and exercises in and outside the classroom. The preliminary evaluation among a random sample of the children showed a decrease in the prevalence of overweight over less than a year’s time, from 16.7% to 14.3% using NCHS stan-dards. The greatest challenge of the program is in improving the extent and manner of integrating the Whiz Kids lessons and activities into the busy curriculum of the children. Physical education sessions are irregular, and so are classroom physical exercises. The plan is to expand the program in the pilot school to the higher grades, while promoting the program in other schools, both public and private.
Round Table Discussion The round table discussion that followed was chaired by Dr. Tee E-Siong, President of the Nutrition Society of Malaysia. The discussion focused on program design, implementation challenges, evaluation, and other issues including research. In designing school-based programs directed towards healthy weight management, the involvement of parents, and even the whole family, should be considered. Aside from parents, teachers are important role models, so that the education of the teachers themselves, not to mention their own health, should also be considered. The program should have clear and specific objectives with defined behavioral goals, taking into account the need for a balance between science and practicality. In the Asian situation
413 News & Views
consideration should be given to the presence of both undernourished and overnourished children in the same class. Since school canteens exert a strong influence on children’s dietary practice, their policies, content, and procedures should be examined and their operators co-opted into the program. In fact, the whole school environ-ment should be conducive to including wholesome and safe physical activities for the children. Apart from strategies to promote physical activity, strategies to decrease inactivity are also important. As a whole, the program design should be culturally specific. Thus the program should be adaptable to differences in ethnic groups and environmental change opportunities. Likewise the program should be age-specific (and potentially gender-specific), so that the design of the program would accommodate the needs of the target group. In this regard, concern should also be directed to preschool children as targets of health pro-motion programs as a means to prevent pediatric over-weight. Among the challenges to program imple-mentation, getting the involvement of parents appeared to be the most difficult. Several suggestions were given such as focus group discussions with parents and giving incentives such as free lectures with giveaway materials or behavior tools of interest to them. The parents should see what is in it for them as well as their children, such as improving their own habits, preventing family conflicts around food or activity, and reducing their child’s social isolation or teasing experiences related to weight or size. Raising the over-all awareness of the importance of proper food and nutrition and physical activity in families would promote parent involvement. Integrating the program into the busy school curri-culum is another challenge. It was pointed out that having a separate program outside of the school curriculum will not work. This will add extra load to the task of the teacher who is already overburdened with academic tasks. Convincing school authorities to adopt a program in the school to promote the integration on nutrition and physical activity is another challenge. The program will have to be adapted to the school environment. On a higher level, how to put forward a national policy, or even local level policy, on a school-based program for weight management is another big challenge, as it competes with national and subnational priorities. Involving top-level school autho-rities in orientation programs on the benefits of physical activity and nutrition education in schools could be a start. The issue of sustainability was discussed at length. Again, a national policy expressing the political will to pursue school-based programs for weight management should be sought. Awareness of political and program leaders of the growing problem of obesity in children and its short- and long-term effects including its burden on health care while convincing them on the benefits of a preventive program, has to be promoted. In this regard, the important role of professional organizations concerned with the control and prevention of obesity was pointed out. Continuous moni-toring of the problem by such organizations would help influence decision making. On the other hand, revita-lization of programs that are already in operation in order to achieve greater effectiveness should be continually pursued. To achieve greater sustainability, involving other
health professionals and child care providers in the program should be sought. Appropriate educational aids are important in bringing about greater effectiveness. The tools should be simple, clear, easily implementable, and appropriate to the school and environment. Involvement of the school and the teachers themselves in laying out the content and design of the aids was suggested. The workshop emphasized the importance of process and impact evaluation in attaining success of school-based programs. Several simple techniques and tools for evaluation were discussed. Qualitative and quantitative evaluation may start from a simple starting point, pro-gressing into more complex matching of goals and objectives with accomplishments related to the inter-vention. The method of evaluation should be defined ahead of time. The use of REAIM (Reach, Effectiveness, Adoption, Implementation, and Maintenance) evaluation model was suggested. In evaluating the program, it is important to seek a balance between what we want to know and what the school wants to know. For example, an evaluation of behavioral change achieved by the program could be an important criterion for success. The workshop suggested important research areas that need to be explored. The appropriate indices and cut-off points for the assessment of overweight and obesity in Asian children still need to be defined. The effect of physical activity on very fit vs unfit children need further investi-ation, so does the effect of physical activity in toddlers on physical performance and academic achievement in the long term. As a final note, the workshop suggested closer networking among the countries in the region as a means of knowing more about what is going on in the other countries, such as through an inventory of programs, and sharing of experiences, strategies and approaches that work through follow-up workshops. The workshop ended with a synthesis of the round table discussion by Dr Rodolfo Florentino and the closing remarks by Mrs Yeong Boon Yee.
News & Views 414
Future Events
March 10-13, 2005
8th National Rural Health Conference, Alice Springs, NT. Contact: http://www.ruralhealth.org.au, [email protected]
September 19-24, 2005 IUNS 18th International Congress of Nutrition: Nutrition Safari for Innovative Solution. Durban, South Africa. The Congress will focus on innovative solutions for global nutrition problems and will aim to build capacity among "young" nutritionists. Contact: http://www.puk.ac.za/iuns, [email protected] October 4-9 2009 19th International Congress of Nutrition, Bangkok, Thailand Global Nutrition Initiative. Contact: [email protected]; tel/fax: 662-590-4333
Asia Pacific Journal of Clinical Nutrition
Instructions for Authors AIMS AND SCOPE Asia Pacific Journal of Clinical Nutrition is the official journal of the Asia Pacific Clinical Nutrition Society (APCNS). The aims of the APCNS are to promote the education and training of clinical nutritionists in the region and to enhance the practice of human nutrition and related disciplines in their application to health and the prevention of disease. Asia Pacific Journal of Clinical Nutrition will publish original research reports, reviews, short communication, and letters to the editors. Letters to the editors may take several forms: they may be either very short articles (500 words) containing new material, case reports, or comments on previous papers or other topics of current interest. News, book reviews and other items will also be included. The acceptance criteria for all papers are the quality and originality of the research and its significance to our readership. Except where otherwise stated, manuscripts are peer-reviewed by two anonymous reviewers and the Editor. The Editorial Board reserves the right to refuse any material for publication and advises that authors should retain copies of submitted manuscripts and correspondence as material cannot be returned. Final acceptance or rejection rests with the Editorial Board. SUBMISSION OF MANUSCRIPTS All articles submitted to the journal must comply with these instructions. Failure to do so will result in return of the manu-script and possible delay in publication. Manuscripts should be written so that they are intelligible to the professional reader who is not a specialist in the particular field. Where contributions are judged as acceptable for publication on the basis of scientific content, the Editor or the Publisher reserve the right to modify typescripts to eliminate ambiguity and repetition and improve communication between author and reader. If extensive alterations are required, the manuscript will be returned to the author for revision. Covering letter Papers are accepted for publication in the journal on the under-standing that the content has not been published or submitted for publication elsewhere. This must be stated in the covering letter. Authors must also state that the protocol for the research project has been approved by a suitably constituted Ethics Committee of the institution within which the work was undertaken and that it conforms to the provisions of the Declaration of Helsinki in 1995 (as revised in Edinburgh 2000). All investigations on human subjects must include a statement that the subject gave informed consent and patient anonymity should be preserved. Any experiments involving animals must be demonstrated to be ethically acceptable and where relevant conform to National Guidelines for animal usage in research. Authors should declare any financial support or relationships that may pose conflict of interest.
Submission The original manuscript and three copies, should be submitted
to: Professor Mark L Wahlqvist Editor-in-Chief: Asia Pacific Journal of Clinical Nutrition, Asia Pacific Health and Nutrition Centre, Monash Asia
Institute, 8th Floor Menzies Building, Monash University, Wellington Road, Clayton, Victoria 3800, Australia.
Email: [email protected] Tel: +61 3 9905 8145; fax: +61 3 9905 8146. (As a rule, registered mail should be used. If courier mail is used, please send manuscripts to the above address.)
Copyright Papers accepted for publication become copyright of HEC PRESS and authors will be asked to sign a transfer of copyright form. In
signing the transfer of copyright it is assumed that authors have obtained permission to use any copyrighted or previously published material. All authors must read and agree to the conditions outlined in the Copyright Assignment Form, and must sign the Form or agree that the corresponding author can sign on their behalf. Articles cannot be published until a signed Copyright Assignment Form has been received. PREPARATION OF THE MANUSCRIPT Submissions should be printed, doubled-spaced, on one side only of A4 paper. The top, bottom and side margins should be 30 mm. Laser or near-letter quality print is essential. All pages should be numbered consecutively in the top right-hand corner, beginning with the title page. Indent new paragraphs. Turn the hyphenation option off, including only those hyphens that are essential to the meaning.
Style Manuscripts should follow the style of the Vancouver agreement detailed in the ‘Uniform Requirements for Manuscripts Submitted to Biomedical Journals’, as presented in JAMA 1997;277:927–34 hhttp://www.acponline.org/ journals/ annals/01jan97/unifreqr.htm). The journal uses US/UK spelling and authors should therefore follow the latest edition of the Merriam–Webster’s Collegiate Dictionary/Concise Oxford Dictionary. Please indicate your pre-ference and use one or the other exclusively. If you do not specify, by default UK spelling will be used. All measurements must be given in SI units as outlined in the latest edition of Units, Symbols and Abbreviations: A Guide for Medical and Scientific Editors and Authors (Royal Society of Medicine Press, London). Abbre-viations should be used sparingly and only where they ease the reader’s task by reducing repetition of long, technical terms. Initially use the word in full, followed by the abbreviation in parentheses. Thereafter use the abbreviation. At the first mention of a chemical substance, give the generic name only. Trade names should not used. Drugs should be referred to by their generic names, rather than brand names.
Parts of the manuscript Manuscripts should be presented in the following order: (i) title page, (ii) abstract and keywords, (iii) text, (iv) acknowledge-ments, (v) references, (vi) figure legends, (vii) tables (each table complete with title and footnotes) and (viii) figures. Footnotes to the text are not allowed and any such material should be incorporated into the text as parenthetical matter.
Title page The title page should contain (i) the title of the paper, (ii) the full names of the authors and (iii) the addresses of the institutions at which the work was carried out together with (iv) the full postal and email address, plus facsimile and telephone numbers, of the author to whom correspondence about the manuscript, proofs and requests for offprints should be sent. In keeping with the latest guidelines of the International Committee of Medical Journal Editors, each author’s contri-bution to the paper is to be quantified. The title should be short, informative and contain the major key words. A short running title (less than 40 characters, including spaces) should also be provided.
Abstract and key words Articles must have an abstract that states in 250 words or less the purpose, basic procedures, main findings and principal conclusions of the study. The abstract should not contain abbre-viations or references. Five key words should be supplied below the abstract and should be taken from those recommended by the Index Medicus Medical Subject Headings (MeSH) browser list (http://www.nlm.nih.gov/mesh/meshhome.html).
Asia Pacific Journal of Clinical Nutrition
Text Authors should use subheadings to divide the sections of their manuscript: Introduction, Materials and methods, Results, Discussion, Acknowledgments, References.
Acknowledgments The source of financial grants and other funding should be acknowledged, including a frank declaration of the authors’ industrial links and affiliations. The contribution of colleagues or institution should also be acknowledged. Thanks to anonymous reviewers are not allowed.
References The Vancouver system of referencing should be used. In the text, references should be cited using superscript Arabic numerals in the order in which they appear. If cited only in tables or figure legends, number them according to the first identification of the table or figure in the text. In the reference list, the references should be numbered and listed in order of appearance in the text. Cite the names of all authors. Names of journals should be abbreviated in the style used in Index Medicus. Reference to unpublished data and personal communi-cations should appear in the text only. References should be listed in the following form: Journal article
1. Tomaino RM, Decker EA. High fat meals and endothelial function. Nutr Rev 1998; 56: 334–343.
Book 2. Fildes VA. Breasts, bottles and babies. A history of infant
feeding. Edinburgh: Edinburgh University Press, 1986. Chapter in a Book
3. Willet WC. The use of biomarkers in nutritional epidemiolgy. In: Kok FJ, Veer PV, eds. Biomarkers of dietary exposure. London: Smith-Gordon, 1991; 9–14.
Tables Tables should be self-contained and complement, but not duplicate, information contained in the text. Tables should be numbered consecutively in Arabic numerals. Each table should be presented on a separate sheet of A4 paper with a comprehensive but concise legend above the table. Tables should be double-spaced and vertical lines should not be used to separate columns. Column headings should be brief, with units of measurement in parentheses; all abbreviations should be defined in footnotes. Footnote symbols: †, ‡, §, ¶, should be used (in that order) and *, **, *** should be reserved for P-values. The table and its legend/ footnotes should be understandable without reference to the text. All lettering/numbers used in tables should be font style 'times new roman' and font size 8.5 or 9. Figures All illustrations (line drawings, bar charts and photographs) are classified as figures. Figures should be cited in consecutive order in the text. Figures should be sized to fit within the column (85 mm), intermediate (114 mm) or the full text width (177mm). Line figures or bar chart figures should be drawn in a computer graphics package (e.g EXCEL). All lettering used in figures should be font style 'times new roman' and font size 9. Important: All figures must be electonically inserted at the end of the manuscript (that has been typed in the word processing package WORD - see section 'Manuscripts on disk) or in a separate file if a different computer program is used. Photographs should be supplied as sharp, glossy, black and white photographic prints and must be unmounted. Individual photographs forming a composite figure should be of equal contrast, to facilitate printing, and should be accurately squared. Photographs need to be cropped sufficiently to prevent the subject being recognized, or an eye bar used; otherwise, written permission to publish must be obtained. Magnifications should be indicated using a scale bar on the illustration. Colour photographs should be submitted as good
quality, glossy colour prints. A charge of $A1,100/US$660 for the first three colour figures and $550/US$330 for each extra colour figure thereafter will be charged to the author. Figure legends Legends should be self-explanatory and typed on a separate sheet. The legend should incorporate definitions of any symbols used and all abbreviations and units of measurement should be explained so that the figure and its legend is understandable without reference to the text. (Provide a letter stating copyright authorisation if figures have been reproduced from another source.) MANUSCRIPTS ON DISK Authors are required to provide their manuscripts on disk; however, disks should not be sent until the manuscript has been accepted. Use a new disk rather than a reformatted disk; the disk must contain the relevant file(s) only. It is essential that the hardware and the word processing package version are specified on the disk (e.g. IBM, WORD 7), as well as the first author’s surname, the journal title and the manuscript number. IMPORTANT: Authors should supply their accepted paper formatted in the computer software WORD. The entire article (including tables and figures) should be supplied as a single file. If the figures cannot be inserted into the manuscript, provide as separate files. The following instructions should be adhered to: • It is essential that the final, revised version of the accepted
manuscript and the file saved on disk are identical. • Do not use the carriage return (enter) at the end of lines within a
paragraph. • Turn the hyphenation option off. • Specify any special characters used to represent non-keyboard
characters. • Take care not to use l (ell) for 1 (one), O (capital o) for 0 (zero) or
ß (German esszett) for b (Greek beta). • Use a tab, not spaces, to separate data points in tables. • If you use a table editor function, ensure that each data point is
contained within a unique cell, i.e. do not use carriage returns within cells.
• Digital figures must be supplied as .tif or .eps files at a resolution of at least 300 d.p.i. (high-resolution print-outs are also required).
PROOFS, OFFPRINTS AND PAGE CHARGES Proofs Proofs will be sent via email as a WORD file formatted in APJCN style with 'tracking' activated so that all changes made to the manuscript by the author will be visible. The edited proofs must be emailed within 5 days of receipt. Alterations to the text and figures (other than the essential correction of errors) are unacceptable at proof stage and authors may be charged for excessive alterations. Authors should therefore supply an email address to which proofs can be emailed. Proofs will be faxed if no email address is available. If absent, authors should arrange for a colleague to access their email, retrieve the emailed manuscript and check and return them to the publisher on their behalf. Offprints & Pdf file Free offprints are not provided to authors. Authors who are subscribers to the journal will receive a free copy of their paper as a pdf file. A minimum of 50 offprints will be provided upon request, at the author’s expense (AU$150 for short papers 1-4 pages, $210 for longer papers >4 pages double sided) .
APJCN online & CD-ROM
Prior to 2004: Full papers are available free-of-charge; Journal abstracts and full papers will continue to be available on-line free-of-charge via the website http://www.healthyeatingclub.org/APJCN/ Volumes 1-11 can also be purchased on CD-ROM for AU$170 (see back cover of journal).
Issues after 2003: subscribers have on-line access to full papers on the website www.healthyeatingclub.org (password access).
Asia Pacific Journal of Clinical Nutrition �
healthyeatingclub.org
HEC PRESS
Editor: Professor Mark Wahlqvist Asia Pacific Journal of Clinical Nutrition www.healthyeatingclub.org
Managing Editor: Dr Antigone Kouris-Blazos Asia Pacific Journal of Clinical Nutrition www.healthyeatingclub.org
In 2001, a nutrition website was launched called healthyeatingclub.org. This site has at its helm Professor Mark Wahlqvist. The content is science-based and up-to-date and canvasses visionary and cutting edge ideas in the field of healthy eating. The HEC website currently offers: 1. an on-line course in nutrition and healthy eating for the general public 2. an on-line wellness program (WOLP) 3. an expanding library of past and present PhDs in nutrition from around the world (this service also provides a great opportunity for PhD students to publish their thesis on the internet and make it internationally accessible via a reputable site) 4. on-line fact sheets and nutrition books for doctors, nutritionists and the general public 5. Asia Pacific Journal of Clinical Nutrition or APJCN (HEC PRESS is the publishing arm of HEC and is responsible for publishing APJCN)
AIM of HEC:
Provide up-to-date reliablewith resources, products andevelop healthy eating patt
What is free? - Authoritative Fact Sheets - Monthly newsletter - Internet TV channel "Healthy Moments" - Downloadable dietary assessment tools - On-line book "Food Questions and Answers" - Recipes - Games
food and health information d feedback to help members erns.
Members 1. Subscribers (AU$29.95 per year): access to the on-line book "Food Facts" and on-line course module on "Healthy Eating Tips" 2. Megabyte (AU$195): benefits of subscribers plus completion of the on-line nutrition course What can you buy?
- on-line healthy eating course modules - an on-line 6 week wellness program - nutrition books, CDs, exercise video - Food products - Asia Pacific Journal of Clinical Nutrition Personal: Hard copy (including on-line access) AU$135; On-line access only AU$120. Institution: Hard copy (including on-line access) AU$400; On-line access only AU$170.
.
Free on-line Journal issues prior to 2003 Journal abstracts and full papers will continue to be available on-line free-of-charge. From http://www.healthyeatingclub.org/APJCN/ Journal issues prior to 2003 on CD-ROM The consolidated journal vol 1-11 can be purchased on CD-Rom for AU$170, US$100 by using the subscription form below. Current Subscribers Subscribers have on-line access to full papers from ww.healthyeatingclub.org/APJCN Nutrition Society of Australia 28 years of Conference Proceedings (abstracts) Now available FREE on-line at www.healthyeatingclub.org/APJCN/
---------------------------------------------------ORDER FORM---------------------------------------------------
Please tick the boxes below corresponding to the type of subscription and whether you wish to purchase the CD-ROM
Hard copy (including on-line access Online access only Personal Price: AU$135 AU$120 Institution: AU$400 AU$170 Past Issues of APJCN on CD-ROM: Volumes 1-11 AU$170 REPRINTS: Manuscripts 1-4 pages double sided 50 copies AU$150 100 copies AU$210
Manuscripts 5-8 pages double sided 50 copies AU$250 100 copies AU$310 (prices do not include postage and handling) I enclose CHEQUE payment of ...........................(Australian citizens can pay by personal cheque, others must pay by BANK DRAFT in AUSTRALIAN CURRENCY ONLY) payable to HEC PRESS. I wish to pay by Visa MasterCard (Other cards are not accepted) Please debit my card no.
Amount in Australian dollars............................Expiry date on card…....…../.............………………….. …………….
Signature......................................……………………………………………………………………………………………….
Name...............................................................................................................................................................................
Address..............................................................................................................................................................................................
.................................................................................................................................Postcode........................................
Telephone..............................................................Fax...........................................................................................................
Please send me a receipt……Yes No Date this form was faxed/mailed…….…/………/………….
Email address (compulsory if want receipt) ....................................................…….………………………….……………
IMPORTANT: For on-line access please provide preferred Username................................. Password....……....…......…
Return this form to : HEC PRESS, c/o Dr Antigone Kouris-Blazos, Suite 4, 2 Elm Grove, McKinnon, Victoria,
Australia, 3204. Tel +61 (0) 3 95767987 Fax +61 (0) 3 95154544 Mob +61 0408551702 Email: [email protected]
Healthy Eating Club Pty Ltd ABN : 93 078 641 906
Suite 4, 2 Elm Grove, McKinno Victoria 320
Austral
Tel +61 (0)3 9576798 Fax +61 (0)3 9515454
Ed
HEC PRESS
APJCN SUBSCRIPTION FORM
n4 ia
7 4
g g
Asia Pacific Journal of Clinical Nutrition
itors: Mark Wahlqvist MD, Australia; Akira Okada MD, JapanPrint ISSN: 0964-7058; Online ISSN: 1440-6047
Frequency: Quarterly (plus supplements)
Contents: Original Articles Can a food frequency questionnaire be used to capture dietary intake data in a 4 week clinical intervention trial?
PAULINE XIE XINYING, MANNY NOAKES AND JENNIFER KEOGH
318
Body mass status of school children and adolescents in Kuala Lumpur, Malaysia FOONG MING MOY, CHONG YING GAN AND MOHD KASSIM SITI ZALEHA
324
Haematocrit levels and anaemia in Australian children aged 1-4 years DOROTHY EM MACKERRAS, SUSAN I HUTTON AND PHILIP R ANDERSON
330
Carotenoid status among preschool children with vitamin A deficiency in the Republic of the Marshall Islands MARY V GAMBLE, NEAL A PALAFOX, BARBARA DANCHECK, MICHELLE O RICKS, KENNAR BRIAND AND RICHARD D SEMBA
336
The effects of a high calcium dairy food on bone health in pre-pubertal children in New Zealand MEGAN J GIBBONS, NIGEL L GILCHRIST, CHRISTOPHER FRAMPTON, PATRICIA MAGUIRE, PENELOPE H REILLY, RACHEL L MARCH AND CLARE R WALL
341
Comparison of serum levels of iron, zinc and copper in anaemic and non-anaemic pregnant women in China AI-GUO MA, XUE-CUN CHEN, RONG-XIAN XU, MING-CI ZHENG, YU WANG AND JUE-SHENG LI
348
Effects of 4 weeks iron supplementation on haematological and immunological status in elite female soccer players HYUNG-SOOK KANG AND TATSUHIRO MATSUO
353
High prevalence of low dietary calcium and low vitamin D status in healthy south Indians CV HARINARAYAN, T RAMALAKSHMI AND U VENKATAPRASAD 359
Health characteristics of older Australian dietary supplement users compared to non-supplement users SONYA BROWNIE AND MARGARET ROLFE 365
Nutritional status of Saudi males living in the Riyadh nursing home ADEL A ALHAMDAN 372
Dietary intake and the risk of coronary heart disease among the coconut-consuming Minangkabau in West Sumatra, Indonesia NUR I LIPOETO, ZULKARNAIN AGUS, FADIL OENZIL, MARK L WAHLQVIST AND NAIYANA WATTANAPENPAIBOON
377
Nutritional analysis of blenderized enteral diets in the Philippines MARY M SULLIVAN, PEARL SORREDA-ESGUERRA, MARIA BERNADETTE PLATON, CYNTHIA G CASTRO, NANCY R CHOU, SUSAN SHOTT, GAIL M COMER AND PEDRO ALARCON
385
Lipid peroxidation and antioxidants status in patients with papillary thyroid carcinoma in India NAMASIVAYAM SENTHIL AND SHANMUGAM MANOHARAN
391
Potential anticancer effect of red spinach (Amaranthus gangeticus) extract HUZAIMAH ABDULLAH SANI, ASMAH RAHMAT, MAZNAH ISMAIL, ROZITA ROSLI AND SUSI ENDRINI
396
Inhibitory effect of clonal oregano extracts against porcine pancreatic amylase in vitro PATRICK McCUE, DHIRAJ VATTEM AND KALIDAS SHETTY
401
News & Views Symposium and Workshop on Healthy Lifestyle Programs for Weight Management RODOLFO F. FLORENTINO Future Events
409
414
Asia Pacific Journal of Clinical Nutrition. Volume 13, Number 4, 2004
HEC PRESS www.healthyeatingclub.org/APJCN/
![Norsk Varmblod - Kubberød 14. september 2019...For Romance I [OLD] f. 2009 DE433334054009 Fürst Romancier [OLD] f. 2004 DE433331650604 Fürst Heinrich [WESTF] f. 1998 DE341411680498](https://img.pdfslide.us/doc/110x75/607b7fd26146356dcf4ce40c/norsk-varmblod-kubberd-14-september-2019-for-romance-i-old-f-2009-de433334054009.jpg)
