Archives of Disease in Childhood 1996;75:474-481

ORIGINAL ARTICLES

Nutrient intakes and impact of fortified breakfastcereals in schoolchildren

Helene McNulty, Jill Eaton-Evans, Gordon Cran, Gregory Woulahan,Colin Boreham,J Maurice Savage, Reginald Fletcher, John J Strain

Human NutritionResearch Group,University of Ulster atColeraineH McNultyJ Eaton-EvansG WoulahanJ J Strain

The Queen'sUniversity of Belfast:Department ofStatistics andOperational ResearchG Cran

Department ofPhysical and HealthEducationC Boreham

The NuffieldDepartment of ChildHealth, Institute ofClinical ScienceJM Savage

Kellogg Company ofGreat Britain,ManchesterR Fletcher

Correspondence to:Dr H McNulty, HumanNutrition Research Group,University of Ulster,Coleraine BT52 ISA.

Accepted 20 August 1996

AbstractObjective-To report micronutrient in-takes in Northern Ireland schoolchildren,and to establish the contribution of forti-fied breakfast cereal to overall nutrientintakes and achievement of current di-etary recommendations.Design-Analysis of dietary intakes andphysical characteristics of participants ina randomly selected 2% population sam-ple of 1015 schoolchildren aged 12 and 15years in Northern Ireland during the1990/1 school year.Main outcome measures-Dietary in-takes, physical characteristics, and theirassociation with consumption of fortifiedbreakfast cereal.Results-Mean micronutrient intakeswere generally adequate with the excep-tion oflow intakes offolate (boys and girls)and iron (girls). Fortified breakfast cere-als, consumed by a high proportion (94%boys; 83% girls) of the sample, were asso-ciated with higher daily intakes of mostmicronutrients and fibre and with amacronutrient profile consistent with cur-rent nutritional recommendations. Ap-preciable proportions of subjects who didnot consume fortified breakfast cerealshad daily intakes that fell below the lowerreference nutrient intake for riboflavin,niacin, folate, vitamin B-12, and iron(girls).Conclusions-The results demonstratethe potential of fortification in contribut-ing to micronutrient intakes of schoolchil-dren, particularly where requlirements arehigh, or for those on marginal diets oflownutritional quality.(Arch Dis Child 1996;75:474-481)

Keywords: fortification, micronutrients, dietary intakes,Northern Ireland schoolchildren.

Diet in childhood is considered to be animportant factor in the development of diseasein later life. Not only are eating patternsbelieved to be established in childhood, butmany diet related disease processes may start inchildhood. A number of epidemiological stud-

ies in different populations have demonstrated'tracking' of risk factors for chronic disease, thephenomenon whereby those individuals withplasma lipids, blood pressure, and body massindex in the upper end of the distribution inchildhood appear to remain so duringadulthood.' 2 Thus nutritional targets such asthose set out in the government strategy paperThe Health of the Nation have directed attentionto the diets of children and adolescentsacknowledging that many of the effects of dietare long term.3 More recent reports havere-emphasised that nutritional recommenda-tions for the prevention of chronic disease inadulthood are applicable in full from the age of5 years onwards throughout childhood.4 Ado-lescence is generally considered a nutritionallyvulnerable period, as it is a time of significantchange in lifestyle, food habits, and physiology.Of particular concern is the maintenance ofhealthy body weight, avoidance of excess fatintakes, and achievement of adequate intake ofmicronutrients.

Recent surveys of schoolchildren in the UKhave confirmed a trend away from conven-tional family meal times towards dietary habitsthat are characterised by informal eating andfrequent snacking,"6 with snack foods contrib-uting up to one third of total energy intakes.7Breakfast cereals feature prominently in thediets of children and adolescents, not only atbreakfast but also as snack foods throughoutthe day.8 The majority of breakfast cereals havebeen fortified with a range of vitamins and ironon a voluntary basis for many years, but theeffectiveness of such wide ranging fortificationin the diets of schoolchildren in theUK has notbeen confirmed in a population study. Fortifi-cation may be particularly important in thediets of at risk subgroups, such as female ado-lescents consuming 'slimming diets' to loseweight (estimated to be up to 60% of girls aged11-18 years),9 or where evidence for increasedrequirement of a specific nutrient (for examplefolic acid in the prevention of neural tubedefects) has resulted in the generation ofofficial recommendationsl' at levels far inexcess of those provided by the national diet.The aim of the present study was to report

micronutrient intakes in a random sample ofNorthern Ireland schoolchildren aged 12 and

474

on May 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.75.6.474 on 1 D

ecember 1996. D

ownloaded from

Nutrient intakes andfortified breakfast cereals

15, and to establish the extent to which theconsumption of fortified breakfast cereal bythis population contributes to overall nutrientintakes and achievement of current dietaryrecommendations.

MethodsSUBJECTSThe dietary and other data reported here werecollected as part of a larger study of coronaryrisk factors in Northern Ireland schoolchil-dren. A detailed description of the samplingprocedure used including reasons for non-response, along with information on the natureand extent of factors associated with coronaryrisk in schoolchildren in Northern Ireland, hasbeen published previously." The samplingprocedure of the original study was designed toselect a sample of approximately 250 childrenfrom each of the following age-sex groups: 12year old boys, 12 year old girls, 15 year oldboys, 15 year old girls. Account was taken ofgeographical spread and the different catego-ries of schools in Northern Ireland. Thenumber amounted to a 2% random sample ofeach of the two age populations in NorthernIreland. The procedure allowed for non-response and drop-outs by increasing the targetnumbers by approximately 20%. An overallresponse rate of78% resulted in a total of 1015subjects being tested. Prior ethical approval forthe main study was secured from the Queen'sUniversity of Belfast ethical committee, andwritten consent was obtained from the parentor guardian of each participating subject."The test protocol included a medical

examination during which height and weightwere determined from which body mass indexwas calculated. Skinfold thicknesses wereobtained from four sites for the estimation ofbody composition.'2 Blood pressure was meas-ured twice from the right arm as previouslydescribed." Cardiorespiratory fitness was de-termined by the 20 metre endurance shuttlerun,"3 and converted to predicted maximaloxygen uptake score using linear regression.'4Estimations of total cholesterol concentration(enzymatic technique: CHOD-PAP, Boeh-ringer Mannheim) were carried out on non-fasting venous blood samples obtained underlocal anaesthesia and separated within fourhours.

DIETARY METHODOLOGYFull details of the dietary methodology, includ-ing the quality control measures undertaken tooptimise consistency of procedures, have beenpublished elsewhere.'5 Briefly, dietary datawere obtained by three trained field workersduring an open ended interview (lastingapproximately one hour) using the diet historymethod'6 to record habitual meal and snackconsumption, in conjunction with a food pho-tographic atlas" and published data to estimatefood portion sizes.'8 Recorded food intakeswere converted into energy and nutrientintakes using a computerised dietary analysisprogram as previously described.'5

Reported food intakes by this dietarymethodology have been shown to be more

representative of energy intake than thoseobtained by the weighed dietary record whenboth methods were independently validated,using doubly labelled water as a measure ofenergy expenditure, in a sample of schoolchil-dren of 12 and 15 years."' However, in order toidentify a small number of subjects whosereported intakes were likely to be under-representative of actual energy (and, in turn,micronutrient) intake based on fundamentalprinciples of energy physiology, a statisticallyderived lower cut off value for ratios ofreported energy intake to calculated basalmetabolic rate (EI:BMR, 1.14)20 was applied;any individual with an EI:BMR ratio of lessthan this value may be classified as anunder-reporter of food intake. When this cutoffpoint was employed in a previous analysis ofthis population,'5 under-reporting was identi-fied in 13.4%, 7.8%, 6.0%, and 5.6% of 15year old girls, 12 year old girls, 12 year oldboys, and 15 year old boys respectively. Thesame principle can be applied to identify over-reporters of food intake by using an upper cutoff value of EI:BMR. The use of these cut offvalues based on multiples of BMR to removethose with doubtful food intakes is still a mat-ter of debate and it is difficult to derive anobjective upper cut off value.2' Thus, in thepresentation of micronutrient intake, all datawere retained and analysed by methods robustto values in the tails of the nutrients beingstudied (see under statistical methods) aspreviously used in the publication of macronu-trient intake in this population.'5 However, forcomparison of intakes according to fortifiedbreakfast cereal consumption level, and forexamination of data according to achievementof dietary reference values, it was felt importantto exclude those subjects classified as under-reporters since the correlation of total nutrientintake with total energy intake for most micro-nutrients means that under-reporting overesti-mates the presence of low intakes.

STATISTICAL METHODSMean values for micronutrients reported intable 1 are medians with 25th and 75thcentiles. Statistical comparisons between sub-ject age-sex groups are based on theKolmogorov-Smirnov two sample test at a 5%significance level. Values with respect tofortified breakfast cereal consumption arereported (tables 2,4, and 5) as means with 95%confidence intervals. Fortified breakfast cerealconsumption categories are 0, 1-20, 1-40, and>40 g/day which correspond to non-consumer,consumption of breakfast cereal two to threetimes weekly, once daily, and more than oncedaily respectively. Statistical comparisons be-tween levels of breakfast cereal consumptionare based on one way analysis of variance withleast significant difference test (in some cases asquare root or log transformation of the datawas employed as appropriate). In tables 2, 4,and 5 a large number of hypothesis tests, eachat the 5% level, are reported. Because of theproblem of multiple testing, the assertion of asignificant difference should be treated withcaution; the existence of a statistically signifi-

475

on May 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.75.6.474 on 1 D

ecember 1996. D

ownloaded from

McNulty, Eaton-Evans, Cran, Woulahan, Boreham, Savage, Fletcher, Strain

Table 1 Median daily intake of micronutrients in Northern Ireland schoolchildren. Median values with 25th and 75thcentiles are given in parentheses

Boys Girls

12years (n=251) 15 years (n=252) 12 years (n=258) 15 years (n=254)

Thiamin (mg) 1.38 (1.08, 1.65) 1.59 (1.27, 1.87) 1.09 (0.92, 1.37) 1.04 (0.86, 1.27)Riboflavin (mg) 1.93 (1.44, 2.41) 2.24 (1.71, 2.86) 1.51 (1.18, 1.83) 1.29 (1.01, 1.71)Niacin (mg) 33.0 (27.6, 39.8) 39.8 (33.0, 47.4) 27.0 (22.7, 32.4) 26.7 (22.0, 32.7)Vitamin B-6 (mg) 1.91 (1.48, 2.34) 2.22 (1.82, 2.63) 1.58 (1.31, 1.89) 1.49 (1.23, 1.86)Vitamin B-12 (gig) 3.4 (2.6, 4.3) 3.9 (3.1, 5.3) 2.6 (2.0, 3.3) 2.5 (1.7, 3.2)Folate (jg) 129 (101, 154) 156 (123, 193) 114 (95, 141) 121 (98, 155)Retinol (jig) 386 (260, 513) 465 (337, 624) 295 (225, 400) 304 (219, 407)Carotene (jig) 1589 (732, 2661) 1523 (707, 2664) 1475 (865, 2454) 1416 (806, 2484)Vitamin C (mg) 61.7 (43.8, 90.9) 73.2 (50.9, 102.4) 68.0 (44.0, 97.6) 71.1 (49.3, 114.8)Vitamin D (jig) 1.74 (1.11, 2.80) 2.18 (1.29, 3.41) 1.67 (0.99, 2.32) 1.66 (0.96, 2.73)Vitamin E (jig) 4.9 (3.8, 6.5) 5.9 (4.5, 7.7) 4.9 (3.7, 6.5) 5.5 (4.2, 7.2)Iron (mg) 12.5 (10.6, 15.6) 15.0 (12.7, 17.6) 10.6 (9.0, 13.0) 10.9 (8.6, 13.1)Zinc (mg) 8.6 (7.3, 10.3) 10.3 (8.4, 12.7) 7.3 (6.0, 8.9) 7.4 (5.9, 9.0)Copper (mg) 1.7 (1.4, 2.0) 1.9 (1.7, 2.4) 1.4 (1.2, 1.7) 1.4 (1.2, 1.8)

cant difference does not necessarily infer apractically important difference. The reportingof such differences was, however, considered tobe useful to the reader. The least significantdifference test was carried out only when thevariance ratio in the analysis of variance wassignificant at p<0.05.When the EI:BMR cut off ratio of 1.14 was

employed to exclude the subjects who wereclassified as under-reporters of actual foodintake, an edited sample comprising boys aged12 years (n=237) and 15 years (n=238) andgirls aged 12 years (n=238) and 15 years(n=221) was obtained. This edited sample wasused in all analyses except for the median (25thand 75th centile) values reported in table 1 (seeunder dietary methodology).

ResultsMedian (25th and 75th centiles) daily intakesof micronutrients for each age-sex group arereported in table 1. The 15 year old boys hadsignificantly (p<0.001) higher intakes than the12 year old boys who, in turn, had significantly(p<0.001) higher intakes than girls of thiamin,riboflavin, niacin, vitamin B-6, vitamin B-12,retinol, iron, zinc, and copper. The 15 year oldboys also had significantly higher intakes thanthe 12 year old boys of folate and vitamin E.The latter, in turn, had a significantly(p<0.001) higher intake of folate than the 12year old girls. Intakes of folate, vitamin C, andvitamin E by the older girls were alsosignificantly (p<0.05) higher than the youngergirls. The intake ofvitamin D in the 15 year oldboys was significantly higher than in the 12year old boys (p<0.01) and also higher than inthe 12 year old girls (p<0.001).A high proportion of Northern Ireland

schoolchildren reported eating fortified break-fast cereal (95 and 92% of males aged 12 and15 years; 87 and 78% of females aged 12 and15 years respectively). Of those consumingbreakfast cereal, a substantial proportion (84%of male consumers, 60% of female consumers)consumed more than 20 g/day-that is, at leastone serving daily.

Estimates of daily intake of micronutrientstraditionally added to breakfast cereals, alongwith calcium and zinc at four levels ofconsumption of fortified breakfast cereal, aregiven in table 2. Intakes of thiamin, riboflavin,

and niacin increased with increasing consump-tion of breakfast cereal in all sex-age groups.Vitamin B-6 intake was significantly differentbetween the breakfast cereal consumptiongroups only in 12 year old girls who consumedmore than 40 g/day fortified breakfast cereal. Ingeneral, higher breakfast cereal consumptionwas associated with higher vitamin B-12intake, except for boys aged 12 years. Folateintake significantly increased with increasingconsumption of fortified breakfast cereal in theyounger adolescents and was greater in girlsaged 15 who consumed more than 20 g/daybreakfast cereal and in 15 year old boys whoconsumed more than 40 g/day breakfast cereal,compared with non-consumers or those con-suming lower levels. In general, higher break-fast cereal consumption was associated withhigher iron intake in all age-sex groups except15 year old boys. Calcium intake increasedwith increasing breakfast cereal consumptionin all age-sex groups. Zinc intake was higher atbreakfast cereal consumption levels greaterthan 20 g/day in 15 year old boys and girls, andgreater than 40 g/day in 12 year old boys, whilezinc intake increased with increasing breakfastcereal consumption in 12 year old girls. Therewas no consistent association between intakesof micronutrients not typically added to break-fast cereal, including retinol, carotene, vitaminE, vitamin C, and copper, and breakfast cerealconsumption in age-sex group (values notreported).The percentages ofNorthern Ireland school-

children who did not meet the lower referencenutrient intake (LRNI) for micronutrientstraditionally added to breakfast cereal and cal-cium and zinc at four levels of consumption offortified breakfast cereal are shown in table 3.In all cases, a decreasing percentage of subjectsnot meeting the LRNI value was found withincreasing level of consumption of fortifiedbreakfast cereal. Vitamin D data are not givenowing to the uncertainty concerning referencevalues for this micronutrient." Neither thiaminnor vitamin B-6 data are included in the tablesince no subject had a daily intake below theLRNI for either of these micronutrients.

Daily intakes of energy, energy yieldingnutrients, and fibre at four levels of consump-tion of fortified breakfast cereal in NorthernIreland schoolchildren are given in table 4.

476

on May 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.75.6.474 on 1 D

ecember 1996. D

ownloaded from

Nutrient intakes andfortified breakfast cereals

00 I 01 _00°'0'-

rA 0'.. . .

N 00 00 0a V

D Onn00 -4 000-00 0

O

'

uo de O -tn000 00

IA0 -00

cest- Le

0 IA N Ci00 '00- V

'40 0 01

0

te ONO

01NN C0

0

0-LI)

IA N IA

N IA 0 ~04-- -

O _ _00

_ _ O

00 It 0~

en

'00 01d

4-- --04-4

00IA 0's

0' N 0'sr

-e c t V-

,_ t_ t_ _

Nd IA 0

--I NA 0

'00 IA 00

4-1 +1 4.1 4 t

%O un O d

cS 'g 0 0 o

cI 01~CI 01

00 01 0'0

roo

cIV

xo o1 -4 CO

IA IA IA 01

tC) 00 00~c

csR sr t

Lo) t- as Le)

ci- - -(:

C5 al DA I

'0N0. N

de0Ci 0I

IA I N N00 01/ 00

'0IA0a 9

0' CO I 00

00 IA N N- Oi - -

0A N 0' d1 Ii 00 .1 * -----0N

00000 en

-i -O

Cl 0N

.0Ot - in 0

O d1 On0 0.00

-0 IA IA+---, ncr-t-0000 00NNo 00

_ a4 - -4 O

I-N--, -I-

DCit 9

000

Ci 00 Ci! le

Ci Ci 0 In

m tn

N N 0)--ci

0 0 r 0

00 00

beo 00Ci OAI N -

_--_0

0 0 0 0

WIAW'00

--VCi N O d

IA N I 0-

cri a' '00

C) en d, ~o-

In t O O

d0-0 0

00(~000

'0'ONC4N

----0

t-, ON

0 o 0 a0

- I -)

Nr vc

en C14e d

- 0 0' _

N0000.I N -'C

cq - - C)

0000CIA0' 0.0.__~-0O

_- Cie0000Z

IA°w- Ci '08

0oxIAAI'IA R N_ Ci 0

I -

0' 00'0

IA Ci 000 0C.s-c_-o

.0 0 _009.°06

IA 0

r

_o t- '00°NNN0'V

00

NC01

0 0'4- 4- 4- cq

00 N IO -ao N- 00 -0IA 00 '0 'o0 01 N -'00000 V

I Q

0 0r- 1co 1

0 0''00a, C0

0-I 01 0'

C t_ _e_

0 O O O

000C 1 0

t0'0

0 0 0 00 I5'D00 6

- Ci Ci IA----V

C0 t 00

e c I OV

0 IA 0'I. 41.

_ ON cn _

-CiCi41e

N C14 ei enV

---- IAO a, _ cn

0 0 0 0

IA1 41 &I 000

cq0000t

oC NC 6

0 0 0 0

-- -1 ci-

en t en UE

00N00O

'0I '0 N

0 V0

IA. A

0- NS 0

i 00 I 0

00- 7,00

00

Nh 000t

- - Co

-S N 0' 00V

0' 00 IA 41IA IA '0000----0

rIA O' O

N 00' 0'

0 00-0 _

N-CiosCCi 000'-0

0 0 4'

b 0o

+1 41 "t Cq0000 00\000 00

0 Ln 04Le) 00

o

O O ~0 0' 0

I00

'O -

0ncn00-0 A 00

4---++ -0'00e No

N 0'. Ci

N A 0'

ci4 Ci4 enA0

IA 0 Ci -

4- 4-1 4- 4-00 NU00

001 IA- -CiI

CON _iN0

ci ci IAc00 '00

0 '

CC0 IA -0004-1 44-1I 4-14

----0

0 0 N Ci

( (S ((SC_4 V

t-CS C

0' 4-10 IA

_% tr c d

Ia 0- I 00

0 0 C0

o1 4U _ 00

a N " al

c 0' I I

000

'r oq o rll

-'IA -4 9

0 IA 0' 00

-N 0Se0D IA 00

0000cs

'o 41 '00

0' 0000.0--_/

ciFdeiici ._2222N

0' IA0 41'O'000 N N----0

0d 0 0cCa 0100 010010 01

01 0? 0~No A c-i A la, A 0o. cicI A

Energy intakes were not significantly differentbetween the breakfast cereal consumptiongroups in 15 year olds of either sex, but 12 yearold boys and girls who consumed more than 40g/day breakfast cereal-that is, more than oneserving daily-had higher energy intakes thannon-consumers or those consuming lower lev-els. Boys (both age groups) who consumedmore than 40 g/day breakfast cereal andyounger girls who consumed more than 20g/day breakfast cereal had significantly lowerpercentage contribution of fat to total energyintakes than non-consumers or those consum-ing lower levels of breakfast cereal. Converselyhigher carbohydrate (as a percentage of totalenergy) intakes were generally associated withhigher breakfast cereal consumption in boysand younger girls. Intake of total sugars washigher in 15 year old boys consuming morethan 40 g/day and in 12 year old girls consum-ing more than 20 g/day breakfast cerealcompared with non-consumers or those con-suming lower levels. In the case of 12 year oldboys and 15 year old girls, there was no signifi-cant difference in total sugars according tobreakfast cereal consumption level, but starchintake was higher in both groups at levels ofconsumption above 40 g/day. Fibre intake gen-erally increased with increasing consumptionof breakfast cereal in girls and was higher inboys (both age groups) who consumed morethan 40 g/day.

Subject and physical characteristics at fourlevels of consumption of fortified breakfastcereal were also assessed (table 5). A trendtowards lower total cholesterol concentrationswith increasing level of consumption of break-fast cereal was apparent in all groups, butsignificantly (p<0.001) so only in girls aged 12years. Cardiorespiratory fitness as determinedby maximal oxygen intake score was alsogreater (p<0.04) in 12 year old girls who con-sumed breakfast cereal compared with non-consumers. There were non-significant trendstowards lower body mass index and percentagebody fat with increasing level of consumptionof breakfast cereal in all age-sex groups.Neither systolic nor diastolic blood pressurewas significantly different between the variousbreakfast cereal consumption levels in any age-sex group.

041

.2

01

co

0

0

41

c0

o.

Cs

a0

V.u

01A

a4-

a010i

c1w0 Discussion

The availability of data from a survey con-

ducted in 1990/1 of a randomly selected popu-lation sample (n=1015; 2%) of 12 and 15 yearolds in Northern Ireland" provided an

opportunity to examine the intakes of a rangeof micronutrients and the importance ofbreakfast cereal fortification to such intakes.The macronutrient intake of this populationhas been previously reported'5 but data on

micronutrient intake are not currently avail-able. Breakfast cereals are the only foods eatenby a substantial proportion of NorthernIreland schoolchildren which are fortified witha wide range of micronutrients. The only othercommonly eaten fortified foods are those forti-fied with specific micronutrients only (for

477

5-.S

.t

MS

015

5.2gt;3

C.)

.Os

'0o

S

.2

r

on May 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.75.6.474 on 1 D

ecember 1996. D

ownloaded from

McNulty, Eaton-Evans, Cran, Woulahan, Boreham, Savage, Fletcher, Strain

Table 3 Percentage ofsubjects who did not achieve the LRNIfor micronutrients traditionally added to breakfast cereal and calcium and zinc by breakfastcereal consumption

Breakfast cereal consumption (glday)

0 1-20 21-40 > 40

Boys Girls Boys Girls Boys Girls Boys Girls

Age (years) 12 15 12 15 12 15 12 15 12 15 12 15 12 15 12 15Sample size 11 20 31 49 42 29 67 85 95 95 106 60 89 94 34 27

Riboflavin 18 15 35 20 0 0 3 6 1 0 0 0 0 0 0 0Niacin 18 20 13 8 5 0 3 2 0 1 1 0 0 0 3 4Vitamin B-12 0 10 10 8 0 0 0 7 0 1 0 2 0 0 0 0Folate 100 75 94 96 90 86 93 95 89 82 93 80 85 57 76 70Iron 0 0 35 20 0 0 22 17 1 0 13 8 0 0 11 4Calcium 18 10 32 4 0 0 9 3 0 0 4 0 0 0 3 0Zinc 0 0 23 2 7 0 7 0 2 0 8 2 0 1 0 0

No subject had a daily intake of thiamin or vitamin B-6 that was below the LRNI. Vitamin D data are not shown because of the uncertainty concerning referencevalues.22

example wheat flour with calcium, iron,thiamin, niacin; margarine with vitamin A andvitamin D).Mean micronutrient intakes of this popula-

tion were generally well above referencevalues,22 with the exception of folate intakewhich fell well below the estimated averagerequirement (EAR) in all age-sex groups apartfrom boys aged 15 years, and iron intake (wellaccepted to be a problem in girls),8 which fellbelow the EAR for females of both age groups.The proportion of Northern Ireland adoles-

cents habitually consuming fortified breakfastcereal in the present study (94% boys, 83%girls) was higher than that previously reportedin 14-16 year old adolescents in Nottingham-shire (71% boys, 46% girls)23 and in a sample of5000 adolescents aged 16-17 years throughoutGreat Britain (79% boys, 63% girls).24 Thismay be due to the younger age group ofsubjects in the present study and an apparenttrend towards decreased consumption ofbreakfast cereal in older adolescents,8 but mayalso reflect regional differences in dietary prac-tices. Comparison of daily intakes of micronu-trients traditionally added to breakfast cerealsin non-consumers compared with consumersof fortified breakfast cereal shows significantlyhigher values among the latter. In many casesmicronutrient intakes increased significantlywith increasing level of consumption of forti-fied breakfast cereal. This finding cannot beexplained simply in terms of overall higherenergy intakes with increased breakfast cerealconsumption since the finding was observed in15 year old boys and girls whose overall energyintakes were not significantly different whenconsumers and non-consumers were com-pared, or when energy values at different levelsof breakfast cereal consumption were com-pared.Although intakes of most micronutrients in

this population appear adequate when ex-pressed as mean values, appreciable propor-tions of those Northern Ireland schoolchildrenwho did not consume fortified breakfast cerealhad daily intakes below the LRNI for indi-vidual micronutrients (riboflavin, 23%; niacin,13%; vitamin B-12, 8%; folate, 92%; iron(girls), 26%; calcium, 14%; zinc (girls), 10%).The LRNI of any nutrient is the amountestimated to meet the needs of the bottom 3%of the population; if individuals are habitually

eating less than the LRNI they will almost cer-tainly be deficient.22 For most micronutrientsthe percentage of subjects with overall intakesbelow the LRNI decreased with increasinglevel of consumption of fortified breakfastcereal; among those consuming 20 g/day ormore-that is, at least one serving of breakfastcereal daily-few subjects with such low dailyintakes were found. This clearly demonstratesthe potential of fortification to benefit thosesubjects consuming diets of low/marginalnutritional quality. The exceptions were folate(in all age-sex groups) and to a lesser extentiron (in younger girls in particular), for whichhigh proportions of subjects had intakes fallingbelow the LRNI despite consuming high levelsof fortified breakfast cereal, as reflected in thelow mean intakes of these nutrients.

Evidence that these observations reflect realeffects of fortification and are not simply theresult of an overall dietary pattern for whichbreakfast cereal consumption is a marker issupported by data on the intakes ofmicronutri-ents such as retinol, carotene, vitamin E, andvitamin C, not usually added to breakfastcereal during fortification, which showed nosignificant differences when examined accord-ing to breakfast cereal consumption level in anyage-sex group (not shown). Since fortifiedbreakfast cereals are consumed with milk it isnot possible to separate the dietary effects ofthe two foods by this type of analysis; high con-sumers of breakfast cereal almost certainly alsoconsume high milk levels. Thus the significantincreases in intakes of nutrients such ascalcium and zinc according to breakfast cerealconsumption level are most likely due to themilk consumed with them, since breakfastcereals are not typically fortified with, or natu-rally rich in, these nutrients. Likewise theassociation demonstrated between fortifiedbreakfast cereal consumption and riboflavinintake probably reflects the dietary composi-tion of both foods.The public health rationale for fortification

(the addition of nutrients to a food to levelsabove those normally present in that food) isthat additional intakes of nutrients may bebeneficial in population groups.25 The questionof folic acid fortification has received muchattention recently following the publication ofconclusive evidence of the effectiveness of thevitamin in the prevention ofneural tube defects

478

on May 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.75.6.474 on 1 D

ecember 1996. D

ownloaded from

Nutrient intakes andfortified breakfast cereals

-

e N 0N CC-

o0 0 0 r °000 0 N

I 0

_ _ N -

004%00I'llO en t)C

a- 60al 9

C _ O 0%

----4 0

00 -NoCO C6ef It) 00V%N V

o) N co No

D 0n0D0 U0 0 0\_._ _

___cO

-R 0---0Co rl Co t o

0% C-C I -

00 00C

un tnao

0% CC-C oad ad

CC In tt4N

ad0 6

CC4. 4)1

%000

O0 -

---,4. -0,

CC-

It) - 0%o

(o tt-,

-0o00000

tq OR _o 0

tn v 00N CC N

I0 00

t- a, N to

CS!llS CS N

00 0000

(R -O~a

0-8

qE 00 ON %Oe 0

(C4-C--00 °

.n v

00 4 00(q -4 O -4

o 0%00000%CO

4--° 4-U14D"0N000 O

On Ul 00_

00 00

Le0 0 -_00_00004 IA -4 -1

0% 000 0

1- .- .4 -

0000C) t(0

_0 0 00 a

u--(C)su

0 C O 0

00 0

I 0

00Ch00a

0 0 4'1

IO O - N 0

0% - - O

If- 0C I 0

00 (C)OIIn0-(C CUCC----0CC)Vl() 0% C 0

----0eWde m

00a

00 0% 0

00)~o00

C_ t"qx-. a00_

IAC C0

00 0%i t_il_ a, 00 0O

0 000

ioaoCnV

000t 00CE- CC) .900

s0

N N %0

(C4 a, ON

0 0%

N00 0

00.-l(C) CC)IA

0000N 00 C

C') 00

VIt ) (C)

Io- - C)t _000_

lpCSi

_N IA 0%00666

0-00

6 w0 4-

w tqOUOR 0 O O

00b 00

(C)CC

0 . 0 -

00 00C 0

-NC LO 00 a, le - a 1

_ e o o00 eC c40% a,% en

C') (CC C' C')000 0

N N - -

0000%0141"

* 0

"! OR Ci -

a,6 CC).9"t

d4C')Co)V

OR) .Ci.C) V

eno-

00tli00

00 _ -

CC CC) CC) CC 9

- -1 N tSO

00 V- C') a,1

60 00

oo %o aA 00olt 4000 0

0 CC 000-1 -1 --0

00 sNCC 0

00I 00

NoV, 00 C

CC)

0N NV

41 .91 4*

000CC 0% V V

a, a, ad

OR O~6 \

0aaC -

VCs) t-

" 000~c

C' -C N 0----Vq

0 0'

0%CA 0- N910 V (C)

C' > 8' C, e>c t

4)o C)oY 0 o o Y X o

0 I01

m 0' A "M O _ cs A 0%s _ A 4 cO- A

(NTDs)."6 Although the mechanism(s) in-volved are still under investigation, and alsoappear to involve an independent role for vita-min B-1 2,27 such findings have led to thegeneration of new folic acid recommendationsby the UK Department of Health.'0 These rec-ommendations are almost identical with thosepublished by other national authorities in theUSA and Australia.28 29 The recommendationfor the prevention of first occurrence ofNTDsin women with no previous history (represent-ing 95% of all cases) is 400 pg/day folic acid,over and above current intakes of the vitamin.'0Folic acid fortification offers the most promis-ing means of ensuring that the recommendedlevels are reached by the target population-that is, all women of child bearing age-sinceincreasing natural food folate sources alonewould require a threefold increase in currentfolate intakes.'0 Moreover, even when a signifi-cant increase in food folate intake is achievedexperimentally, it has recently been shown tobe relatively ineffective (compared with supple-ments and fortified food) at increasing erythro-cyte folate status, probably because of thepoorer bioavailability of the vitamin whenpresent in the polyglutamate (food folate)form." Supplementation with folic acid, al-

0) though effective, is unlikely to reach a major.2 proportion of those who need it." In the

present study the folate intakes of female ado-v lescents (1 14 and 121 jg/day in 12 and 15 year

olds respectively) fell far short of recom-mended levels, failing even to achieve the refer-ence value of 200 pg/day which was set before

5% the publication of the new folic acid recom-a mendations specific to the prevention of

NTDs.10 There is cause for concern in thisregard since most teenage pregnancies areunplanned (therefore folic acid supplementa-tion before conception is unlikely to be in

C-' place), and the incidence of NTDs, which isparticularly high in Northern Ireland," appearsto be higher in younger (and older) mothers

a than the general population of women in theirreproductive years.'0 The fact that folateintakes increased significantly with increasingconsumption of fortified breakfast cereal dem-onstrates the potential of fortification to make avaluable contribution to overall folate intakes,but clearly current levels of fortification are nothigh enough to ensure that the recommendedincreased intakes are reached by this vulnerablegroup. The present study therefore supportsrecent arguments'4 for a public health initiativeto include a more aggressive approach to folicacid fortification. New evidence on the role of

g. folic acid'5 (and to a lesser extent vitamins B-6o and B-12)'6 in lowering plasma homocysteine

level, an independent risk factor for coronaryheart disease in the general population, is likelyto strengthen the arguments for fortificationwith these micronutrients.

a Any potential benefit of consumption of for-tified breakfast cereal on micronutrient intakewould be irrelevant if it resulted in a dietary

.2 pattern that was in conflict with currentrecommendations for the prevention of chronicdisease.' 4 Thus the extent to which breakfastcereal consumption patterns of Northern

479

C-Z

CO

020

CC15%0.

A

I0A)0-00l

C-

.,

'-

0t

0,

on May 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.75.6.474 on 1 D

ecember 1996. D

ownloaded from

480 McNulty, Eaton-Evans, Cran, Woulahan, Boreham, Savage, Fletcher, Strain

\Z.- 0 0 00 0 0 00 0 0 00 0 00 0

4C '-0000-141 41 41 414- -1 414?

- -t~c' 1tt 00(q00)O Ut Cf-oo 4 0enCCt000s00 00C0 0

0 00 0 0 00 0 0 0 00 0 00 0

C. 'IOtux in % nr t- 00o ~ ~ o0)

0 -0'0'44 4.400 4I"0 t-'Ir 4 N o'oooo 0

000I00 0 0 00 0 00 0 00-

C. 00 C- 00C tnO 00 00fto 00'00)C

0)~~~~~~~~~~~~~~~~f

NO 0 0 00 0 00 0 0 0 00 0 00 0

0 rot~~~~~~~~~~~~t----n 0%~oOc-cC0- C'-N 0N 00'IO000t-

~~~ ~ ~ ~ ~ e~~~~~ie1C-~~~~~~~~~00 ---- 01 _11_

0) CC~~~~~~~~~~~nU)l0U)tntncn Cn CC.. C~~~~~~~~~~~~~~~u

0000cr(N 00O 0 r I fo000000 00000 Cu

-0 0 00 000- 00

b 4 -400' --4Ct-o0'If 0 '0 00C14 1C(7.0'0'0'0.-0Ct0~~~~~~~~~~~~~~C~~Ir0__ ''0'0000001 -1 -11-

c* o 6 (io6o66v rn\6~t;o6 o -: -

0)~~~~~~~0~~~~~)C'~~~~~~ -0- 4( (4N 0 N NNN 0~~-~~~-~~~-~~~-OoOoO~~~~~~Cu

0 C0000 000C-0 0000 C0000

e~00 o0r00 ON N0 ON 00 00 000t C- O 2\

C 000 00004 - 7, N 000 00 -00\00C0C-C-06 c! 66''0 00'oIfoC-oo6 0C C,.Cq-4- 00'-C C-oor-1 _

focor

C04

0) 0 Cu0 0Cu 0 Cu 0 Cu 0 00) 0 0) 0~~~~~~~~~~~~~~~~~~~1)0 0 0) 00O0)00)0

o "D ,0-t r--~\0t -N o 0c qI

0)uC) 0))n IOC 0) 0C~~~~~~~~Cu~~~~~~~~~~~~~~~~~~~;C

0 0 00 Cu 00Cu 0.00

0 N t0 - t' 0 (1)c " Cu~ocf 0NCEu~~~~~~~~~~~v:o ~ ; ciC V \; C C ; (F ,fiA .c

Ireland schoolchildren relate to macronutrientintake and health indices was also examined inthe present study. Fat as a percentage of dietaryenergy has previously been reported for thispopulation to be high at 39%,15 and consider-ably above current Department of Health3 4recommendations of 35%. Although none ofthe breakfast cereal consumption subgroupsmet the recommendation, the finding that fatintake of this population decreased withincreasing consumption of breakfast cereal isconsistent with other studies."4 37 38 Corre-sponding increases in carbohydrate (as apercentage of dietary energy), starch, and fibrewere shown to be associated with increasedbreakfast cereal consumption, generally con-forming with the recommendation of increas-ing intake of complex carbohydrates as a meansof achieving a reduction in fat as a percentageof total energy.4 The consumption of fortifiedbreakfast cereal in the present study was notconsistently associated with increased intake oftotal sugars; thus the increase in carbohydrateenergy associated with breakfast cereal con-sumption cannot be explained by increasedsugars alone. The trends in subject and physi-cal characteristics such as body mass index,percentage body fat, and cholesterol level,which tended to decrease with increasedconsumption of breakfast cereal, may be theindirect result of the inverse relationshipbetween fat energy and breakfast cereal level.Alternatively these trends may simply be areflection of a general dietary and lifestyle pat-tern for which breakfast cereal consumption isa marker. In any case, with the exception ofcholesterol levels in 12 year old girls, none ofthese trends was significant. Support forconcerns that the sodium content of somemanufactured food may contribute to elevatedblood pressure4 was not evident in the case ofbreakfast cereals in the present investigation inwhich there were no significant differences insystolic or diastolic blood pressure associatedwith the consumption of breakfast cereal.

In conclusion, although mean micronutrientintakes of Northern Ireland adolescents gener-ally appear to be adequate, there is cause forconcern in the case of low folate intake (in bothboys and girls) and low iron intake in girls.Fortified breakfast cereals, which are con-sumed by a high percentage of NorthernIreland schoolchildren, appear to make asignificant contribution to micronutrient in-takes, particularly in subjects consuming dietsofpoor nutritional quality. Effects on macronu-trient intakes associated with breakfast cerealconsumption are consistent with current di-etary recommendations.

This study was supported by the Northern Ireland Chest, Heartand Stroke Association, the Department of Health and SocialServices (Northern Ireland), and Kellogg Company of GreatBritain Ltd. In addition the contribution of the following peopleto the data collection in the original survey is gratefullyacknowledged: Paula Robson, Elizabeth Archer, and HilaryMorrison (dietary data); Craig Mahoney (fimness testing);Daphne Primrose (medical aspects).

1 Lauer RM, Lee J, Clarke WR. Factors affecting the relation-ship between childhood and adult cholesterol levels: theMuscatine Study. Pediatrics 1988;82:3 10-28.

on May 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.75.6.474 on 1 D

ecember 1996. D

ownloaded from

Nutrient intakes andfortified breakfast cereals 481

2 Webber LS, Srinivasan SR, Wattingey WA, Berenson GS.Tracking of serum lipids and lipoproteins from childhoodto adulthood: the Bogalusa heart study. Am J7 Epidemiol1991;133:884-99.

3 Department of Health. Health of the nation: a'strategy forhealth in England. London: HMSO, 1992.

4 Department of Health. Report on Health and SocialSubjects No 46. Nutritional aspects of cardiovascular disease.London: HMSO, 1994.

5 Adamson A, Rugg-Gunn A, Butler NR, et al. Nutritionalintake, height and weight of 11 and 12 year old Northum-brian children in 1990 compared with informationobtained in 1980. BrJNutr 1992;68:543-63.

6 McNeill G, Davidson L, Morrison DC, et al. Nutrientintake in schoolchildren: some practical considerations.Proc Nutr Soc 1991;50:37-43.

7 Robson P, Strain JJ, Cran GW, et al. Snack energy andnutrient intakes ofNorthern Ireland adolescents. Proc NutrSoc 1991;50:180A.

8 Department of Health. Report on Health and SocialSubjects No 36. The diets of British schoolchildren. London:HMSO, 1989.

9 Hill A. Causes and consequences of dieting and anorexia.Proc Nutr Soc 1993;52:211-8.

10 Department of Health. Report from an expert advisorygroup. Folic acid and the prevention of neural tube defects.London: Department of Health, 1992.

11 Boreham C, Savage JM, Primrose D, et al. Coronary riskfactors in schoolchildren. Arch Dis Child 1993;68: 182-6.

12 Durnin JVGA, Rahaman MM. The assessment of theamount of fat in the human body from measurements ofskinfold thickness. BrJ Nutr 1967;21:681-9.

13 Boreham CAG, Paliczka VJ, Nicols AK. A comparison ofthe PWC170 and 20-MST tests of aerobic fimess inadolescent schoolchildren. J Sports Med Phys Fitness 1990;30:19-23.

14 Riddoch C. Northern Ireland health and fitness survey. Areport by the Division of Physical and Health Education,The Queen's University of Belfast, 1990.

15 Strain JJ, Robson PJ, Livingstone MBE, et al. Estimates offood and macronutrient intake in a random sample ofNorthern Ireland adolescents. BrJNutr 1994;72:343-52.

16 van Stavern WA, de Boer JD, Burema J. Validity andreproducibility of a dietary history method estimating theusual food intake during one month. Am Jf Clin Nutr 1985;42:554-9.

17 Lee P, Cunningham K Irish national nutrition survey.Dublin: The Irish Nutrition and Dietetic Institute, 1990.

18 Crawley H. Food portion sizes. London: HMSO, 1988.19 Livingstone MBE, Prentice AM, Coward WA, et aL Valida-

tion of estimates of energy intake by weighed dietary recordand diet history in children and adolescents. AmJ Clin Nutr1992;56:29-35.

20 Goldberg GR, Black AE, Jebb SA, et al. Critical evaluationof energy intake data using fundamental principles ofenergy physiology. 1. Derivation of cut-off limits to identifyunder-recording. Eur J Clin Nutr 1991;45;569-81.

21 Black AE, Goldberg GR, Jebb SA, et al. Critical evaluationof energy intake data using fundamental principles of

energy physiology. 2. Evaluating the results of publishedsurveys. EurJ Clin Nutr 1991;45:583-99.

22 Department of Health. Report on Health and SocialSubjects No 41. Dietary reference values forfood energy andnutrients for the United Kingdom. London: HMSO, 1991.

23 Lund BK, Gregsun K, Neale RJ, Tilston CH. The breakfastarrangements ofschookhildren. Food Marketing Group, Uni-versity of Nottingham, 1990.

24 Crawley HF. The role of breakfast cereals in the diets of16-17 year old teenagers in Britain. Journal of HumanNutrition and Diet 1993;6:205-16.

25 British Nutrition Foundation. Briefing paper. Food fortifica-tion. London: The British Nutrition Foundation, 1994.

26 MRC Vitamin Study Research Group. Prevention of neuraltube defects: results of the Medical Research Council vita-min study. Lancet 1991;338:131-7.

27 Kirke PN, Molloy AM, Daly LE, et al. Maternal plasmafolate and vitamin B-12 are independent risk factors forneiral tube defects. QJ Med 1993;86:703-8.

28 US Department ofHealth and Human Services, Centers forDisease Control and Prevention. Recommendations for theuse of folic acid to reduce the number of cases of spinabifida and other neural tube defects. MMWR MorbidMortal Wkly Rep 1992;41:1-7.

29 National Health and Medical Research Council. Revisedstatement on the relationship between dietaryfolic acid and neu-ral tube defects such as spina bifida. 115th session. Australia:NHMRC, 1993.

30 Gregory J, Foster K, Tyler H, Wiseman M. The dietary andnutritional survey ofBritish adults. London: HMSO, 1990.

31 Cuskelly GJ, McNulty H, Scott JM. Effect of increasingdietary folate on red-cell folate: implications for theprevention of neural tube defects. Lancet 1996;347:657-9.

32 Clark MA, Fisk NM. Minimal compliance with DepartmentofHealth recommendation for routine folate prophylaxis toprevent neural tube defects. BrJ Obstet Gynaecol 1994;101:709-10.

33 Eurocat Working Group. Prevalence of neural tube defectsin 20 regions ofEurope and the impact of prenatal diagno-sis, 1980-1986. J Epidemiol Community Health 199 1;45:52-8.

34 Wald NJ, Bower C. Folic acid and the prevention of neuraltube defects. BMJ 1995;310:1019-20.

35 Boushey CJ, Beresford SAA, Omenn GS, et al. A quantita-tive assessment of plasma homocysteine as a risk factor forvascular disease. JAMA 1995;274:1049-57.

36 Ubbink JB, Vermaak WJH, van der Merwe A, Becker PJ.Vitamin B-12, vitamin-B-6, and folate nutritional status inmen with hyperhomocysteinaemia. Am J Clin Nutr1993;57:47-53.

37 Jenkins DJA, Wolever TMS, Vuksan V, et al. Nibbling versusgorging: metabolic advantages of increased meal frequency.NEngl3tMed 1989;321:929-34.

38 Gibson SA, O'Sullivan KR. Breakfast cereal consumptionpatterns and nutrient intakes of British schoolchildren. JRSoc Health 1995;115:366-70.

on May 23, 2021 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.75.6.474 on 1 D

ecember 1996. D

ownloaded from