Upload
biblioteca-virtual
View
341
Download
3
Embed Size (px)
Citation preview
745
�1
Original Research Communications-method
Clinical and field studies of human lactation:methodological considerations13Kenneth H. Brown, M. D. , Robert E. Black, M. D. , M. P.H. , Alastair D. Robertson, M. S.,
Naheed A hmed Akhtar, M. B. B. S. , Md. Giashuddin Ahmed, M. Sc., and Stan Becker, Ph.D.
ABSTRACT A variety of studies has been completed to assess selected methods that are
presently being used or might potentially be used to evaluate lactation performance. During 64
test-weighings of infants before and after the consumption of a known amount of milk. the mean
± SD “recovery” ofmilk ingestion was 94.9 ± 13.2%. The weight ofmilk extracted by a mechanical
pump was approximately 7% greater than the amount measured during test-weighings of infants of
the same women within periods of 1 wk. To evaluate the possibility of performing abbreviated
studies in field settings, the proportion of24-h milk consumption received during 12 daytime hours
was measured by test-weighings. Daytime consumption ranged from 46 to 58% of24-h consumption
(� ± SD = 52 ± 3%). Attempts to predict the amount of milk consumption during 341 daytime
studies from the age of infants and their frequency and duration of feedings met with limited
success. Although each independent variable was significantly correlated with the amount con-
sumed (multiple r = 0.69, p <0.001). the SE of the estimate (Sy.x) was relatively large. The effects
of time of day and side of extraction on the volume and composition of extracted milk were
determined during 24-h studies of seven women. There were significant changes in the concentration
of fat (p < 0.001) and nitrogen (p = 0.003) during the day and significant differences in the
concentrations of fat (p = 0.04) and lactose (p = 0.04) and in the volume (p < 0.00 1) of milk
produced by each breast. The importance of these findings for the planning and interpretation of
studies of human lactation is discussed. Am J Clin Nutr 1982:35:745-756.
KEY WORDS Human lactation, breast-feeding, human milk, infant feeding
Introduction
Among mothers and scientists of the tech-nologically advanced countries there hasbeen a recent resurgence of interest in breast
feeding. Simultaneously, the World HealthOrganization, recognizing the central impor-tance of breast-feeding for the health of in-fants in the developing world, has launcheda number of studies of present day breast-feeding practices and has stressed the needfor additional research in the physiology of
lactation (1). Basic to the correct design andinterpretation of such studies is the applica-tion of appropriate methodologies to investi-
gate lactation performance in clinical andfield settings. The current studies were com-pleted in an effort to assess some of the
From the Division ofGeographic Medicine (K.H.B.,
A. DR.), Department of Medicine, Gastroenterology and
Nutrition Unit, Department of Pediatrics, School of
Medicine, Division of Human Nutrition, Department of
International Health, Johns Hopkins University, Balti-
more, MD; the International Centre for Diarrhoeal Dis-
ease Research (Formerly The Cholera Research Labo-
ratories) (K.H.B., R.E.B., SB.), Dacca, Bangladesh; the
Center for Vaccine Development (R.E.B.), University of
Maryland School of Medicine, Baltimore, MD: Save the
Children Fund (N.A.A., Md. GA.), U.K., Children’s
Nutrition Unit, Dacca, Bangladesh.
2Supported in part by USAID Grant DAN-l406-G-
SS-l031-00. Supported by an International Center for
Medical Research NIH Grant 5R07AI10048-17, by the
International Centre for Diarrhoeal Diseases Research,
Bangladesh (ICDDR,B) and by the Center for Vaccine
Development.Address correspondence to: Dr. Kenneth Brown.
615 North Wolfe Street, Baltimore, MD 21205.
Received February 23, 1981.
Accepted for publication September 22, 1981.
The American Journal of Clinical Nutrition 35: APRIL 1982. pp. 745-756. Printed in U.S.A.© 1982 American Society for Clinical Nutrition
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
746 BROWN ET AL.
procedures commonly used to evaluate lac-
tation.The difference in an infant’s weight before
and after nursing has been assumed to equalthe weight of milk consumed, but few studieshave attempted to validate this “test-weigh-
ing” technique or to compare it with othermethods of estimating the production or con-
sumption of human milk. Although simple toperform in clinical wards, the test-weighing
technique itself becomes impractical for 24-hstudies in field settings. Thus, more simplified
or abbreviated techniques of investigationmust be developed and evaluated. Further-more, the complete assessment of lactationperformance must include some indication ofthe composition of nutrients in the milk, ne-cessitating the extraction of a representative
sample of the day’s production. Knowledgeof whether a given sample of milk is repre-sentative of the full day’s production requiresthe analysis and comparison of levels of nu-trients in the milk at different times of theday. Specific studies were therefore under-
taken to address these issues in preparationfor subsequent studies of lactation perform-ance among relatively undernourishedwomen of two communities in Bangladesh.
Methods and results
Study sites
All ward studies were accomplished in aneight-bed metabolic ward situated at the Chil-dren’s Nutrition Unit (CNU) in urban Dacca,Bangladesh. The ward was separated fromthe rest of the hospital, and the study roomwas furnished simply to appear more like avillage home than a hospital ward. The studysubjects nursed their infants openly, as iscustomary in Bangladesh; no special nursingareas were provided for the women. Mealswere prepared for the mothers according tofixed dietary protocols (to be described in
detail in subsequent communications); butthe women were responsible for basichousekeeping chores and for the care of theirinfants while staying on the ward. Themother-infant pairs stayed in the ward for aslittle as 2 days for “outpatient studies” and
for as long as 6 wk for “inpatient studies.”The village studies were completed in sev-
eral villages of Matlab Thana, Comilla Dis-
trict, Bangladesh, the rural setting ofa variety
of health-related activities of the Interna-tional Centre for Diarrhoeal Disease Re-search, Bangladesh (formerly the CholeraResearch Laboratory). The geographic, so-cial, and demographic characteristics of thearea have been described previously (2, 3).The individual studies were performed in thehome of each study subject.
Study subjects
Participants in the ward studies were re-
cruited from a large group ofwomen enrolledin a social service program located in a“bustee” area (semiurban slum) in outlyingDacca. After learning of the purposes anddesign of the study, interested women weretransported to the ward to visualize the studysite; consenting subjects were then acceptedfor participation in future studies.4 The 61women ranged in age from 14 to 39 yr (me-
dian 20 yr), were all primiparous except threewomen, and were in good, general health.None was using hormonal contraceptives atthe time of the study. Their infants’ ages
ranged from 1 to 9 months. In return for theirparticipation the subjects and their infantsreceived routine preventive and curativehealth services as well as remuneration in lieuof potential earnings. The 13 infants takingpart in studies designed to validate the test-weighing technique were selected from inpa-tients recovering from severe protein-caloriemalnutrition at the Children’s Nutrition Unit.
The village study subjects were selected
from a group of mothers who were alreadyparticipating in a separate study of infants’dietary intakes, and who were found to beparticularly interested in and cooperativewith the earlier investigations. The womenwere 17 to 42 yr of age (median 25 yr), ingood health and had from one to eight livingchildren (median three), including the nurs-ing infants who ranged from 6 to 29 monthsof age. Several additional women (usuallyrelatives of field workers) were enrolled inthe studies comparing 12- and 24-h milk con-sumption, since it was particularly difficult to
The research protocols were approved by the re-
spective Committees on Human Volunteers of the School
of Hygiene, Johns Hopkins University, of the Cholera
Research Laboratory and of the Bangladesh Medical
Research Council.
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
60
28
12
52
STUDIES OF HUMAN LACTATION 747
arrange 24-h, home-based studies in the con-
servative village society.The methods and results of the individual
parts of the study will be presented togetherto facilitate comprehension. A list of each ofthe individual studies, as well as their numberand location is found in Table I.
Validation of test-weighing technique
Quantification of infants’ consumption ofmother’s milk has traditionally been accom-pushed by the test-weighing technique. In thepresent study, diapered infants were weighedbefore and after feedings by a clinical assist-ant using a balance calibrated daily againstknown weights. The diapers were covered byplastic pants and were not changed until the
second weighing was completed. The wardbalance (Detecto Infant Scale) was sensitive
to 5 g; the balance used in field studies wasa modified solution balance (Heavy DutySolution Balance, Ohaus) sensitive to I g.During the validation studies the infants were
fed by spoon from cups of milk that had beenpreweighed by a second study assistant on abalance (Dial-O-Gram Balance, Ohaus) sen-
sitive to 0.1 g. After the infants had been fedby the mother or by a study assistant otherthan the individual responsible for weighingthe infants, the cup of milk was reweighed.The infant’s weight change was then ex-
pressed as a percentage of the change in theweight of milk in the cup. Three studies wereeliminated from consideration because either
a sizeable amount of milk was spilled or thebaby regurgitated.
TABLE 1
A minimum of nine validation studies wascompleted by each ofthe six clinical assistantsresponsible for the test-weighing of infantsduring longitudinal, clinical studies of lacta-tion performance. The accuracy of the“weighed-feeding” procedure, as determined
by the mean percentage of a known weight
of ingested milk that could be measured bytest-weighings ranged from 89.4 to 98.6% forthe individual clinical assistants (Table 2).The SDS ofthe recoveries ranged from 4.8 to24.4% for the different assistants. The mean± SD overall recovery of milk ingestion dur-
ing all 64 validation studies was 94.9 ± 13.2%.The variation in the recoveries tended to begreater when smaller amounts of milk wereconsumed.
Fourteen similar studies of single feedingswere completed in the village subjects’ homes
to define the accuracy of the test-weighing
technique under field conditions. The recov-ery of a known amount of ingested milkranged from 56.2 to 117.8% in the field settingwith a mean ± SD of 88.6 ± 14.4%.
Comparison of the weights of milk consumed
and the weights of milk extracted
In order to determine whether the twodifferent methods of estimating the amountof milk production provide similar data, val-
ues obtained from the same woman by test-weighing and by mechanical extraction were
compared over short intervals of time (lessthan 8 days). During the period of extractionall milk was extracted from both breasts by
Summary of methodological evaluations, locations of studies,
and numbers of studies and subjects
. � . No. Of mothersPurpt�.c of �.tud� Ltcati�n ?So. of .tudie�
. or ntant�
Validation oftest-weighing . Urban ward 64 13
Rural homes 14 14Comparison of test-weighing and mechanical extrac- Urban ward 167 72*
tion
Determination of density of milk Urban ward 511
Comparison of amount of milk consumed in 12- and Urban ward 250
24-h periods Rural homes 12
Prediction of human milk consumption from age of Rural homes 341
infant, frequency and duration of feedings
Determination of changes in volume and composition Urban ward 7
of extracted milk by side of extraction and time of
day
* Mothers participating as both inpatients and outpatients are considered twice.
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
748 BROWN ET AL.
TABLE 2
Comparison of the change in infant weight before and after feedings with the weight of milk ingested
Clinical Assistant No of studiesAverage amount of milk
ingested
Average wt change of
infant
‘Recovery of ingested milk by
weighed-feeding technique
g g ‘:�
1 10 160 147 91.1 ± 8.4*
2 9 113 109 89.4± 24.4
3 9 153 148 96.1±8.2
4 9 154 148 96.4 ±4.8
5 9 146 140 94.4 ± 9.8
6 18 142 141 98.6 ± 13.8
All studies 64 144 139 94.9 ± 13.2
* Mean ± SD.
a mechanical breast pump (Egnell) at 3-hintervals and pooled for the 24-h period. Two
groups of women were studied during eitherinpatient or outpatient periods in the clinicalward. During the 1 18 inpatient studies of 23mother-infant pairs, the total amount of each
day’s test-weighings measured during someor all of the 6 days either before or just after
a day of mechanical extraction was comparedto the weight of that extracted milk. Duringthe 3-day outpatient studies of 49 mothersand infants, the data from the 2 days of test-weighings were compared to those for theintervening day’s extracted milk. Paired datawere analyzed by substracting the test-weigh-
ing data for a given day’s study from therespective extraction data for the same
woman during the same study period. Aspresented in Table 3, those mean differenceswere always positive and usually statistically
significant during the inpatient studies. Inother words, mechanical extraction yieldedslightly higher estimates of milk productionthan did test-weighings. The 3-day outpatientstudies yielded similar results, but the differ-ences were smaller than those measured dur-ing the inpatient studies. Whereas the mean(± 1 SE) weight of extracted milk was 717(± 21) g, the weight of milk measured by test-weighings averaged 7 14 (± 18) and 678 (± I 8)g, respectively, on the day before and the dayafter extraction. Only the difference on theday after extraction was statistically signifi-cant (p < 0.0 1 ). If the test-weighings from allof the paired studies were corrected for theamount of underestimation as determined inthe earlier recovery studies, the amount ofmilk production as estimated by either test-weighing or extraction would be similar.
Density of extracted milk
To estimate the amount of milk consumed
by a single infant on multiple days, the leastdisruptive measurement procedure should be
utilized. Although test-weighings are a moresuitable technique than extraction of milk,extraction would be necessary periodically todetermine the concentration of nutrients inthe milk. To estimate the infant’s consump-tion of nutrients, one could then multiply theconcentration of nutrients in extracted milkby the amount of milk consumed on thosedays when nursings were weighed. This cal-culation assumes little day to day variation inthe concentration of nutrients.
Whereas test-weighings measure theweight of milk consumed, the concentrations
of nutrients are generally expressed per unitsof volume. The amount of milk measured
during test-weighings must be divided by itsdensity to calculate the volume consumed.To learn whether the density of milk is con-stant for all women the weights and volumes
of extracted milk were compared during 5 1 1separate 24-h ward studies of 60 women. Themean ± SD density for all studies was 1.036
± 0.017 g/ml, the grand mean for individualwomen was 1 .038 ± 0.01 1 g/ml. There wasno statistically significant correlation between
maternal age or infant age and the density of
extracted milk.
Comparison of the amounts of milk consumedduring 12 and 24 h
Ideally, quantitative studies of milk con-sumption should take place in the subjects’homes with as little disruption as possible ofthe mothers’ routine activities and feeding
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
STUDIES OF HUMAN LACTATION 749
TABLE 3
Comparison of the amount of milk produced by individual women in 24 h, as
estimated either by extraction of milk or by test-weighings, by day
of test-weighing before or after day of
“paired” extraction
Study
Day before or after extraction
I 2 3 4 5 6
One day of extraction followed
by 6 days of test-weighings
Amount ofrnilk measured 715.8 695.8 694.2 716.0 708.9 704.4
during test-weighings (g/
day)
Mean difference* 96.3t 70.3t 8 1 .9f 60. It 58.7t 68.8tSE ofdifference 26.0 14.6 16.2 17.8 16.8 15.8
No. of paired studies 26 50 55 55 49 51
Six days of test-weighings fol-
lowed by I day of extrac-
tion
Amount of milk extracted in 694.5 7 12.3 7 10. 1 69 1 .4 689.3 729.7
paired studies
Mean difference* 48.6t 29.5 33.Ot 5l.7t 46.8t 48.2
SEofdifference 12.1 15.1 14.4 13.2 12.6 28.5No. of paired studies 63 60 63 63 57 20
a Mean difference between extracted amount less test-weighing amount during paired studies of individual women
(g/day).
t Paired t tests (p < 0.01).:1:Paired t tests (p < 0.001).
patterns, and without potentially anxiety-pro-yoking changes in surroundings. However, itis not logistically feasible nor socially accept-able to accomplish large numbers of 24-hstudies of milk intake in subjects’ homes, soinvestigations taking place in the field must
be abbreviated. Therefore, preliminary stud-ies comparing 12-h intakes to actual 24-h
intakes were undertaken to determine whatproportion of the day’s total intake was con-sumed between 6 AM and 6 PM These 24-hstudies were completed in I 2 village homes
and the data obtained were compared to datafrom 250 studies of 28 women enrolled in theward studies.
The proportion of the entire day’s intakethat was ingested between 6 AM and 6 PMranged from 37 to 72% (mean ± SD, 53 ±
9%) in the 12 studies of village subjects. Sincethe number of villagers who consented to the24-h studies in their homes was limited, theproportions were also calculated for a sam-
pling of outpatients who were enrolled in theward studies on at least 5 different days. Thewoman-specific mean proportions of the full
day’s intake that were consumed during the
daytime ranged from 46 to 58% during mul-tiple studies of 28 individual women (mean
±SD, 52 ± 3%). All but two of the SDs of theproportions for studies within individualwomen ranged between 3 and 10%.
Prediction of the amount of milk consumptionf rom age of infant andfrequency andduration offeedings
Since the test-weighing technique is cum-bersome for field studies, an attempt wasmade to identify a more easily obtainable
estimate of the quantity of milk consumed ina 12-h period. During an 8-month period,three hundred forty-one 12-h weighed feed-
ing studies were completed for 52 infants inthe field. The field assistants timed each com-plete feeding (both breasts) to the nearestwhole minute. The relationship between thenumber and total duration of feeding, as well
as infant age (independent variables) and theamount ofmilk consumed in 12 h (dependentvariable) was analyzed by stepwise multiplelinear regression (4) with the hope of iden-
tifying independent variables that accuratelypredict the amount of milk consumed.
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
Results of multiple regression analysis relating age of infant, total amount of time suckling, and frequency of
nursing (independent variables) to amount ofmilk consumed (dependent variable) during 12-h studies* (n = 341)
Partial SD of partial PartialIndependent Mean ± SD regression regres.ston F p
variablescoefficient coefficient )df= 1.337)
0.92
0.32
26.5
2 I.2
3.80 20.7
<0.001
<0.001
<0.001
750 BROWN ET AL.
The infant’s age, the frequency of feedings,and the total amount of time at the breastwere significantly correlated with the amount
of milk consumed (Tables 4 and 5). The ageof the infant was negatively correlated with
the frequency of feeding (r = -0.2 1), the
amount of time nursing (r = -0.2 1), and theamount of milk consumed (r = -0.34). Thenumber of feedings and time at the breastwere both positively correlated with theamount of milk consumed (r = 0.63 for both)and with each other (r = 0.82). Although the
ability ofthe independent variables to predictthe amount of milk consumed was enhanced
when all three were considered simultane-
ously (multiple r = 0.69), they were togetherable to explain only about 50% of the varia-tion in milk consumption. The SE (Sy.x) ofthe estimate of the amount of milk consumedby an individual was 81.8 g. Therefore the95% confidence limits of the estimate was ±
54.8% ofthe mean amount ofmilk consumed.
Composition and volume of milk extractedfrom both breasts during 24 h
In order to determine an optimal samplingprocedure for measuring the concentration ofnutrients in and the volume of extracted hu-
TABLE 4
Amount of mother’s milk consumed (g) during 12-h studies by age of infant,
total amount of time s uckling, and frequen cy of nursing
.Ageoftnfant
Total time at
breast
Fre quency of feedi ngs (n/ I 2-h study)
n
O�5
Mean
amountn
6-7
Mean
amountn
8+
Mean
amountn
All�-
Mean
amount
6-Il 0-59
60-89
90+Total
11
2
013
224*
402
252
1
24
4
29
274
333
402
340
0
10
8
18
410
453
430
12
36
12
60
228
358
436
348
12-17 0-59
60-89
90+
Total
47
9
2
58
228
318
261
243
5
35
3
43
287
343
370
337
2
16
19
37
328
328
414
372
54
60
24
138
237
334
396
307
18+ 0-59
60-89
90+
Total
61
11
0
72
193
243
201
9
29
2
40
244
294
296
282
0
18
13
31
348
352
350
70
58
15
143
200
301
344
256
All 0-59
60-89
90+
Total
119
22
2
143
210
288
261
223
15
88
9
112
260
324
368
319
2
44
40
86
328
355
402
376
136
154
51
341
217
327
390
293
a Amount of milk consumed (g).
TABLE S
Age of infant (mo) 16.3 ± 4.9 -4.74
Amount oftime 66.0 ± 24.4 1.48
at breast (mm)
No. offeedings 6.1 ± 2.1 17.30
DependentVariable
Amount of milk 292.7 ± I 12.3
consumed (g) ____________ ______________
a Multiple r = 0.668, SE (Sy.x) = 81.84, p < 0.00 I.
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
STUDIES OF HUMAN LACTATION 751
>
1t
�1
8 2� a a ‘�‘
0 � 0 �, 0 1,p.. #{149} �O *
z
0 #{149} 0 #{149} 0 �p.. ,p � *
a
CuS
t...
�1 0U�
o2; �
.�
8#{149} t_
- 0U
E8 ��, 0-
aU
� E- .�
a
Ea
E
a
§a
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
fat, and lactose in, and volume of extracted breast milk
Woman
Nitrogen concentration (mg/
dlFat concentratton
.(g/dl) Lactose
.concentration (g!dl) volume (ml)
- Left breast Right breast Left breast Right breast Left breast Right breast Left breast Right breast
1 148
±1 1*
156
±6
1.79
±0.67
2.14
±0.60
7.31
±0.33
7.50
±0.40
57
±1567
±20
2 167
±6
174
± I 1
2.22
±0.88
2.69
±0.74
8.10
±0.3 1
8.04
±0.27
48
± 10
60
±7
3 165
±1 1
147
±15
2.23
±0.60
2.15
±0.42
8.03
±0.20
8.06
±0.3835
±1354
±9
4 158
±15
169
±22
2.20
±0.62
2.31
±0.58
7.94
±0.46
8.16
±0.34
58
±8
58
±15
5 137
±5
138
±5
1.74
±0.67
1.77
±0.458.28
±0.57
8.25
±0.3629
±635
±9
6 140
±20
158
±15
1.53
±0.46
1.64
±0.67
7.25
±0.33
7.38
±0.29
76
±12
31
±9
7 171
±8
139
±6
2.71
±1.02
3.18
±0.71
7.85
±0.62
8.12
±0.23
35
±13
91
±22
All 155
±17
155
±18
2.05
±0.782.27
±0.76
7.82
±0.54
7.93
±0.45
48
±19
56
±23
a Mean ± SD.
752 BROWN ET AL.
man milk, milk was collected separately from
each breast every 3 h during the course ofsingle 24-h studies of seven women. The vol-
umes and weights ofthe milks were measuredand well mixed aliquots were then stored at-20#{176}C before analysis for nitrogen, lactose,and fat. Nitrogen concentration was deter-mined by a semimicro-Kjeldahl procedure
(5) with an accuracy of 94.2 ± 8.4%, lactosewas measured by the method of Folin (6)with an accuracy of 99. 1 ± 5.4% and total fatwas quantified gravimetrically after extrac-tion by the Rose-Gottlieb procedure (5) withan accuracy of 100.5 ± 1.9%.
The data for the concentration of nutrients
and the volume of milk were summarized todetermine changes related to the time of sam-pling (Fig. 1) and side of extraction (Table
6). Analyses of variance (4) were then com-pleted as shown in Table 7. The time-relateddifferences were statistically significant in thecase of nitrogen (F = 3.82, p = 0.003) and fat(F = 5.30, p < 0.001). Whereas the minimumnitrogen concentration was slightly more than90% of its maximum, the minimum fat con-centration was less than 65% of its maximum.
TABLE 6
Although the time-related changes for lactose
concentrations and volume were not statisti-cally significant for the group as a whole,there were significant interactions betweentime and individuals for these two variables.
In other words, there were statistically sig-nificant diurnal changes in the concentration
of lactose (F = 3.38, p < 0.001) and volume(F = 1.80, p = 0.03) within individual subjectsbut the times of those changes were not con-sistent for all individuals in the study group.The magnitude of change in lactose concen-tration was not great as the minimums weregreater than 90% of the maximums for all butone subject. However, some individuals hadas much as a 2-fold difference in the volumeof milk produced during different intervals ofthe 24-h period.
The differences between the volume ofmilk produced by the right and left breastwere of statistical and biological significance(Table 6). As a group the differences werenot great, but for some women the less pro-
ductive breast yielded only 65% ofthe volumeof the more productive one. Although the
concentrations oflactose and fat in milk from
Effects of side of extraction on concentration of nitrogen,
from seven women
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
STUDIES OF HUMAN LACTATION 753
TABLE 7
Results of analysis of variance for volume and three components of extracted breast milk with factors, side of
extraction, time of day, and interactions
Factor Source ofvariation OF Mean Square F Significance
Nitrogen
Main effects Individual
Breast
Time
6
I
7
1812
4
498
13.88
0.03
3.82
<0.001
0.863
0.0032-Way interactions Individual x breast
Individual x time
Breast X time
6
42
7
1286
120
132
9.86
0.92
1.01
<0.001
0.605
0.438
Lactose
Main effects Individual
Breast
Time
6
1
7
2.1 16
0.323
0.666
29.91
4.56
0.94
<0.001
0.038
0.4862-Way interactions Individual x breast
Individual x time
Breast X time
642
7
0.065
0.239
0.102
0.923.38
1.44
0.493<0.001
0.214
Fat
Main effects Individual
BreastTime
6
17
3.308
1 .2031.430
12.26
4.46
5.30
<0.001
0.041
<0.001
2-Way interactions Individual x breast
Individual x time
Breast X time
6
42
7
0. 195
0.505
0.162
0.72
1.87
0.60
0.635
0.023
0.753
Volume
Main effects Individual
Breast
Time
6
1
7
1880
1930
106
15.91
16.34
0.90
<0.001
<0.001
0.5 192-Way interactions Individual x breast
Individual X time
Breast X time
6
42
7
35 1 1
213
256
29.72
1.80
2.17
<0.001
0.030
0.057
each breast were also statistically different,the magnitude of those differences was rela-tively unimportant when grouped data were
analyzed. Even when individual women wereconsidered, the differences between milk col-lected from each breast were not great, exceptfor two women whose milk from one breastcontained about 80% of the concentration offat as compared to milk from the other breast.
Discussion
The evaluation of lactation performance isof fundamental importance to nutritionistsconcerned with the health of infants and theirmothers, particularly in the lesser developedcountries. The techniques for such evaluation
are limited by the central dilemma that theimpact of the study procedures themselves onlactation remains uncertain. At best, one can
compare a variety of techniques and assumethat those procedures that are least invasive
and that require the fewest changes from the
natural setting are most likely to estimate theusual lactation performance with accuracy.The issues of cost, comfort of study subjects,
and the level of technical sophistication re-quired must also be considered in planningsuch studies.
Test-weighing is the procedure used mostcommonly to estimate the amount of milkconsumed by the nursing infant. The proce-dure is technically simple, requires minimalequipment and training of personnel, andinterferes relatively little with the usual proc-ess of nursing. However, since the amount ofmilk consumed at a single feeding may besmall, the measurement error can introducevariability that is a sizeable proportion of theamount consumed. Nevertheless, in the pres-ent studies the mean recovery of a knownamount ofingested milk attained a fair degreeof accuracy, even under field conditions, ifmultiple studies were undertaken. The test-weighings consistently underestimated the
amount of milk consumed, probably because
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
754 BROWN ET AL.
small amounts of milk were spilled or regur-gitated. Insensible water loss of the infant
between weighings undoubtedly also ac-counts for some of the underestimation. Fi-nally, any urine or stool passed after theinitial weighing and inadvertantly not in-cluded in the fmal weighing would also re-duce the recoveries. The range of variabilityof the technique also depends on the carewith which the clinical assistants perform
their tasks. In the present studies some assist-ants were clearly superior to others in thatregard.
The test-weighing procedure has been
compared to other methods of estimatingmilk production in previous studies. Cowardet al. (7) found that their technique of isotopedilution by breast milk ingestion yieldedhigher estimates of milk consumption than
weighed-feedings in six infants. In the presentstudy we found that extraction of milk by themechanical pump yielded an estimate of milkproduction that was about 6 to 7% higherthan that measured by test-weighings withinthe same week. Hytten (8) also found thatmechanical extraction yielded 6% more milk
than the sum of weighed feedings plus man-ual extraction of residual milk in studies thattook place on consecutive days (8). It is pos-sible that mechanical extraction recoversmore milk than is actually consumed by theinfant or that test-weighings systematically
underestimate the amount of milk produced.Since the difference in the estimates obtainedby the two techniques was similar to the
difference between test-weighings and actualconsumption observed during the “recoverystudies,” the latter hypothesis seems morelikely to be correct.
In order to determine the density of humanmilk, the volumes and weights of multipleextracted samples were measured. The spe-
cific gravity of human milk has been reported(9) as 1.032, which is similar to the presentfindings. Since the densities did not changewith infant or maternal age for our sample,most of the variability appears to be ex-plained by individual differences and mea-surement error. Therefore, when correctingmilk weight to volume or vice versa it wouldbe preferable to use the conversion factor asdetermined for the woman in question.
To facilitate the collection of data on lac-tation performance in field studies, feedings
can be weighed during a 12-h period rather
than during a complete 24-h period. On the
average, approximately 52% of the 24-h in-take was consumed during the daytime pe-
nod by infants of mothers from our popula-tions, both in field and ward studies. Al-though there was considerable within and
between subject variability in the proportionof milk consumed during daytime hours, thatvariability could be reduced by performingmultiple studies. Whether such abbreviatedstudies are acceptable depends on the natureand the design of research proposed. Fieldstudies of lactation performance that were
completed in rural West Africa used a similartechnique of 12-h test-weighing to estimate
milk consumption (10). In those studies,which took place between 7 AM and 7 PM,less than 50% of the full day’s intake was
consumed during the 12-h studies. The dif-ferences in the proportions consumed duringdaytime hours may relate to the fact that adifferent 12-h interval was studied or mayindicate culture-specific differences in breast-feeding practices. It would seem necessary toestimate the conversion factor directly for anypopulation to be studied by an abbreviated
technique.The attempt to estimate milk consumption
by indirect methods was generally unreward-
ing in that the majority ofvariability in intakedata was unexplained by the independentvariables chosen. Unless better explanatory
variables can be identified we believe that theestimation of milk ingestion of an individualmust continue to rely on a more direct tech-
nique of assessment, such as those techniquesmentioned above. However, similar indirecttechniques may be of value in estimating the
amount of milk consumption by a group ofinfants from a given population. Again, theusefulness of the technique will depend on
the purpose of the data collection. The mu!-tiple regression analysis was of additionalinterest in explaining some of the factorsaffecting lactation performance in Banglade-shi mothers. For example, milk intake tendedto decrease as the infants became older. In-creased total duration of suckling and fre-
quency of feedings were both independentlyassociated with higher levels of milk con-sumption in children of all ages. This latterfinding is at odds with the results of the WestAfrican studies reported by Whitehead et a!.
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
STUD1ES OF HUMAN LACTATION 755
1. Anonymous. WHO/UNICEF Meeting. Lancet1979:2:841-3.
(10). Although they stated that the variationin total milk intake between individuals atany given age could not be explained by the
frequency of feedings, they did report thatthe amount ofmilk intake fell with infant ageand was accompanied by a gradual reduction
in the frequency of feedings. We cannot ex-plain the differences in these two sets of data.
The primary concern of nutritionists inves-
tigating lactation performance is to determinethe total amount of nutrients ingested by theinfant and simultaneously lost by the mother.Both the volume of milk consumed and theconcentration of nutrients in that milk mustbe measured. Since it is the entire day’s milk
that is of interest, the milk that is collectedfor laboratory analysis must be representativeof the whole day’s production. The technical
difficulties that interfere with the accuratemeasurement of milk nutrients have beendiscussed by Hytten and Thomson (I 1) who
concluded that “there is no escape from thenecessity to obtain complete 24 hour samples
if misleading measurements are to beavoided.” Hytten (12) also claimed that milkmust be collected from both breasts sincesubstantial differences may exist in the vol-
ume and composition of milk from eachbreast.
The data from the present study are con-
sistent with Hytten’s claims, but some quali-fications are justified. Although the concen-
trations of macronutrients in milk changedsignificantly throughout the day, particularlywhen individuals rather than groups were
considered, only the changes in fat concentra-tions were of relatively great magnitude. Ifone were interested only in lactose concentra-tion, for example, a randomly obtained sam-ple would not be too misleading in that itslactose concentration would probably be less
than 5% different from the daily mean con-centration for that individual. The time-re-lated changes in fat concentration that wereobserved in this study are consistent with thefindings of other investigators (13, 14) and
emphasize the necessity of complete collec-tions to assess the total daily fat (and energy)production accurately. On the other hand,Lauber and Reinhart (15) reported variations
in the protein and lipid concentrations rang-
ing from 5 to 15% and chose to ignore thosedifferences in their study of women in theIvory Coast. The time of day appeared to
influence the concentrations of’ f’at and nitro-
gen in opposite directions. Nims et al. (13)also reported that maximal fat and proteinconcentrations generally occurred at different
times of day, but the actual values were notpresented. The differences in the productive
capacity ofthe two breasts were most impres-sive for volume and fat concentrations. Be-cause the differences in volume were so greatit appears that milk must be extracted from
both breasts if the extraction data are to beused as estimates of the volume of milk pro-
duced. However, the concentrations of nutri-ents in milk from each breast were not verydissimilar, especially if data from a group of
subjects are of interest. However, if the con-centration of milk fat from individual sub-
jects is the object of study, specimens shouldbe obtained from both breasts. Recently, Hall(16) reported that the concentrations of fat in
the fore-milk and hind-milk from the breastsuckled second in the course of a nursing aregreater than in the respective fractions fromthe breast suckled first. Unfortunately, wedid not record which breast was extractedfirst in our own studies. It is possible that the
differences observed between breasts reflectconsistent biases in which breast was pre-
sented first for extraction.Several recent publications have reviewed
the numerous proven and theoretical advan-tages of breast feeding ( 17, 18). Because ofour interest in the nutritional well-being ofchildren and their mothers, particularly in
the lesser developed countries, we have re-cently undertaken studies of the lactationperformance of marginally nourished womenin Bangladesh. The present methodologicalevaluations were initiated in an effort to de-termine the appropriate techniques for those
investigations. The data reported should beof similar use to other individuals planningstudies of human lactation or evaluating datafrom such research.
The authors appreciate the collaboration of the staff
of the Save the Children Fund, U.K., Children’s Nutri-
tion Unit, and the efforts of the field workers of the
Matlab Research Area. We also thank Drs. R. B. Sack,
G. G. Graham, W. B. Greenough, and M. M. Levine for
support and advice during the planning and analysis of
these studies. The encouragement of M. R., M. B., S. B..
and N. B. is also gratefully acknowledged.
References
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from
756 BROWN ET AL.
2. Curlin GT, Chen LC, Houssain SB. Demographic
crisis: the impact of the Bangladesh civil war on
births and deaths in a rural area of Bangladesh.
Popul Stud 1976;30:87-105.
3. Ruzicka LT, Chowdhury AKMA. DemographicSurveillance System-Matlab, Vols I and III. Sci-
entific reports no 9 and 1 1 . Dacca: Cholera ResearchLaboratory, 1978.
4. Snedecor GW, Cochran WC. Statistical methods.
6th ed. Ames: The Iowa State University Press, 1967.
5. Horwith W, ed. Official methods of analysis of the
association of official analytical chemists. 12th ed.Washington, DC: Association of Official Analytical
Chemists, 1975.
6. Folin 0. The determination of sugar in blood and innormal urine. J Biol Chem 1926;67:357-70.
7. Coward WA, Whitehead RG, Sawyer MB, PrenticeAM, Evans J. New method for measuring milkintakes in breast-fed babies. Lancet 1979;2:l3-l4.
8. Hyuen FE. Clinical and chemical studies in humanlactation. 1. Collection of milk samples. Br Med J1954; 1:175-6.
9. Lawrence RA. Breast-feeding: a guide for the med-ical profession. St. Louis: CV Mosby Co, 1980.
10. Whitehead RG, Rowland MGM, Hutton M, et al.Factors influencing lactation performance in rural
Gambian mothers. Lancet 1978;2:l78-8l.
11. Hytten FE, Thomson AM. Nutrition ofthe lactatingwoman. In: Kon SK, Cowie AT, eds. Milk, themammary gland and its secretion. New York: Ace-demic Press, 1961.
12. Hytten FE. Clinical and chemical studies in human
lactation. II. Variation in major constituents during
a feeding. Br Med J l954;l:176-9.13. Nims G, Macy IC, Brown M, Hunscher HA. Human
milk studies. IX. Variations in the composition ofmilk at four hour intervals during the day and night.
Am J Dis Child l932;43:828-44.
14. Hytten FE. Clinical and chemical studies in human
lactation. III. Diurnal variation in major constituents
of milk. Br Med J 1954;l:l79-82.
15. Lauber E, Reinhart M. Studies on the quality ofbreast milk during 23 months of lactation in a ruralcommunity of the Ivory Coast. Am J Clin Nutr1979;32:l 159-73.
16. Hall BH. Uniformity of human milk. Am J Clin
Nutr l979;32:304-l2.17. Jelliffe DB, Jelliffe EFP. The uniqueness of human
milk. Am J Chin Nutr l97I;24:968-1024.
18. Jelhiffe DB, Jelliffe EFP. Human milk in the modern
world. Oxford: Oxford University Press, 1978.
by on August 22, 2009
ww
w.ajcn.org
Dow
nloaded from