THE QUANTITY AND QUALITY OF BREAST MILK Report on the WHO Collaborative Study on Breast-feeding WORLD HEALTH ORGANIZATION GENEVA 1985ISBN 92 4 154201 2 World Health Organization 1985 Publications. of the World Health Organization enjoy copyright protection in accordance with the provisions of Protocol 2 of the Universal Copyright Convention, For rights of reproduction or translation of WHO publications, in patt or in toto, applications should be made to the Office of Publications, World Health Organization, Geneva, Switzerland, The World Health Organization welcomes such applications. ‘The designations employed and the presentation of the material in this publication do ‘ot imply the expression of any opinion whatsoever on the part of the Secretariat of the World Health Organization concerning the legal status of any country, territory, city or area, ot of its authorities, or concerning the delimitation of its frontiers or boundaries, ‘The mention of specific companies or of certain manufactufers’ products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters, TYPESET IN INDIA PRINTED IN SWITZERLAND 83/5742 — Macmilan/1AM — 10,000Contents Acknowledgements... 2... 6.06500 c ese cece eee ee Introduction... .......... A short review of literature on 1 breast-milk quanti y and composition ...... eee Design of the study. Study of breast-milk quantity : Study of breast-milk composition and quality... Summary and conclusions ......... wees Rr ape References . oe ee eee (Annex) le Adaitionall tablesseyrren terre Annex 2, Guidelines for record forms and sample record forms ...... Annex 3, Test-weighing and sample collection procedures used in the study ..... Annex 4. Laboratory methods, auality “control, and reporting of results ...... bee ebeneeees 80 102 129 143Acknowledgements ‘The World Health Organization wishes to express its special thanks to the principal investigators listed below, whose untiring efforts have made this study possible. The Organization wishes also to acknowledge the guidance and support of Mr W. Z. Billewicz who acted as a consultant for the design and statistical analysis of the study and to Dr L, Hambracus who provided the reference samples of breast milk to the collaborating centres and assisted in the standardization of the laboratory procedures. Thanks are also due to the collaborating centres and national teams in the five participating countries and to the more than 3000 mothers who, together with their families, took part in the study. The help of Ms Capri Fillmore in the preparation of the text is also gratefully acknowledged Participants in the WHO Collaborative Study on Breast-feeding: second phase Guatemala Dr_O. Pineda, Institute of Nutrition for Central America and Panama (INCAP), Guatemala Dr J. J. Urrutia, Ms B. Garcia Hangary Principal investigator Dr I. Gry, Head, Department of Mother, Child, and Youth Care, Ministry of Health, Budapest Chief collaborators Dr P. Cholnoky, Head Paediatrician, General Hospital, Szombathely Dr ©. Gaal, Department of Food Chemistry, National Institute of Food and Nutrition, Budapest Clinical work . Dr M. Csaszar, Miss J. Acs, Mrs M. Csiza, Mrs V. Markoczy Philippines Principal investigator Dr Virginia B. Guzman, Department of Community Health, Institute of Public Health, University of the Philippines System, Manila vivii Biochemical studies Manuel P. Macapinlac, Department of Biochemistry, College of Medicine, University of the Philippines System, Manila Field directors Lilia V. del Castillo, Department of Community Health, Institute of Public Health, University of the Philippines System, Manila Theresita R. Lariosa, Department of Community Health, Institute of Public Health, University of the Philippines System, Manila Biostatistician Susie Ignacio Morelos, Chief, Biostatistics Section, Department of Epidemiology and Biostatistics, Research Division, Philippine Heart Center for Asia, Quezon City Sweden Principal investigator Professor Y, Hofvander, Department of Paediatrics, University Hospital, Uppsala Clinical work U. Hagman, The Swedish National Food Administration, Stockholm Laboratory work G. Fransson,! Professor L. Hambraeus, H. Jonsson, B. Larsson,? C. E, Linder,? R. Ras,? S. A. Slarach,? L. Wahlberg" Zaire Principal investigator Professor H. L. Vis, Institute of Scientific Research, Kinshasa; Department of Paediatrics, Free University of Brussels, Brussels, Belgium Clinical work P. Hennart, Field Director, Zaire Institute of Scientific Research, Kinshasa. Ruchababisha-Migabo and Nyampeta Uwaytu Biochemical studies 1. Mandelbaum, Director, Paediatrics Laboratory, St. Peter’s Hospital, Free University of Brussels, Brussels, Belgium Eugenia Colombara, P. Devroede, Agnés Vuye, and Nadine Herremans, Free University of Brussels, Brussels, Belgium Mrs J. Delonge-Desnoeck, Laboratory Technician, Unit of Reproductive Endocrinology, Free University of Brussels, Brussels, Belgium Dr_A. Heyndrick, Director, Laboratory of Toxicology, Ghent University, Ghent, Belgium’ * Institute of Nutrition, Uppsala University, Uppsala 2 The Swedish National Food Administration, Stockholmvi Dr C. Robyn, Head, Unit of Reproductive Endocrinology, Free University of Brussels, Brussels, Belgium N. Sookvanchsilp, Pharmacist, Laboratory of Toxicology, Ghent University, Ghent, Belgium M. Warnez, Pharmacist, Laboratory of Toxicology, Ghent University, Ghent, Belgium Consultants Mr W. Z. Billewicz, Medical Research Council, Reproduction and Growth Unit, Newcastle-upon-Tyne, England Dr_L. Hambraeus, Institute of Nutrition, Uppsala University, Uppsala, Sweden International Atomic Energy Agency Dr R. M. Parr, Department of Research and Isotopes, IAEA, Vienna, Austria WHO Secretariat Dr’M. Béhar, Nutrition, Division of Family Health, Geneva Dr M. Carballo, Maternal and Child Health, Division of Family Health, Geneva Dr E. M. DeMaeyer, Nutrition, Division of Family Health, Geneva, (Study coordinator) Mr P.-C. Kaufmann, Health Statistical Methodology, Division of Health Statistics, Geneva Dr A. Petros-Barvazian, Di sion of Family Health, Geneva1. Introduction The desirability of a multinational study on breast-feeding was advanced by the Thirteenth International Paediatric Congress in Vienna in 1971, and in the following year, a colloquium arranged by the International Children’s Center in Abidjan recommended, inter alia, that WHO should play a leading role in the organization of such a study. It was also recommended that the latter should be carried out in two phases—the first dealing with the prevalence and duration of breast-feeding, and the second with the volume and composition of breast milk. In early 1973, WHO decided to implement these recommendations, the first phase of the study being initiated shortly thereafter. This phase was completed in 1976, and the results were published by WHO in 1981 in a book entitled Contemporary patterns of breast-feeding (21). Thereafter, plans were made to initiate the second phase to investigate more specifically the volume and composition of breast milk. This second study, the results of which are presented in this book, was timely in view of the questions that have been raised in recent years concerning the adequacy of breast-feeding by mothers living in the developing world, in circumstances that are not always conducive to the best nutrit- ional status and the best physiological performance of the mothers. It has been known for some time that mothers in developing countries tend to produce less milk than mothers living in developed countries. On the other hand, most of the information collected so far indicates that any differences in the composition of breast milk that may exist between mothers living in developed countries and those living in developing countries, do not modify significantly the nutritive value of the milk, especially its energy value. Most of this information, however, has been collected using different methods and is not strictly comparable; it is therefore of doubtful value. In order to be able to compare data from different parts of the world, it has been stressed repeatedly that the methods used by different investigators should be rigorously standardized. The same applies to the determination of the various components of breast milk, and is particularly important with regard to the determination of trace elements. It is only by such rigorous standardization of methods that it will become possible to compare data on breast-milk quantity and quality from different races living under different ecological and2 THE QUANTITY AND QUALITY OF BREAST MILK socioeconomic conditions and, presumably, of different nutritional backgrounds. This volume reports the results of the study concerning the quantity of breast milk, its composition in protein and non-protein nitrogen, lactose, fat, lactalbumin and lactoferrin, vitamin A, and vitamin C and the levels of pesticides present in the breast milk. The results of mineral and trace element determinations in breast milk will be reported in a subsequent volume. It is a common observation that children in developing countries have a growth rate similar to that of children in developed countries up to the age of 3-4 months, Thereafter, their growth curves tend to flatten and diverge from those of children in developed countries. This observation has led many people to question the adequacy of breast-feeding after 3 months of age in developing countries—because either the volume of milk might be insufficient or its nutrient composition might be inadequate. The study reported here of lactating mothers living in contrasting socioeconomic conditions and in different geographical areas of the world was intended to answer such questions. If it were found that mothers in developed and developing countries produced equal quantities of milk of similar nutritional value then the reason for the suboptimal growth curve after 3-4 months of age would have to be sought elsewhere. The present study was designed to provide easily. interpretable information on some of the points raised above. The answers should be of great value not only in terms of what the nutrient requirements of an infant are, but also in terms of how and when to start supplementing the diet of a breast-fed infant, Tentative answers to these questions have been provided in the reports of several joint FAO/WHO ‘scientific groups dealing with nutrient requirements, young child feeding, or both, but an answer supported by reasonably accurate facts collected in various environmental conditions is still required. It is hoped that this collaborative study will lead to recommendations on when and how to supplement the diets of breast- fed children, and that the results will form a basis for nutrition education and supplementation feeding programmes.2. Breast-milk quantity and composition: a review 2.1 Quantity or Volume of Breast Milk 2.1.1 Methodology for the measurement of breast-milk quantity or volume The methods generally used for the measurement of breast-milk quantity or volume can be divided into two categories: those that measure breast secretory capacity—ie., the ability of the mother to secrete milk, and those that measure the infant’s intake of milk. Since the infant's intake of breast milk is influenced by both the maternal secretory capacity and the baby's sucking ability and appetite, one expects the estimates of the infant’s intake to be lower than those of the mother’s ability to secrete milk. Furthermore, the determination of breast-milk quantity generally interferes with the normal interac- tion between the mother and the child and with the family life-style. This may have a negative effect on the let-down reflex, causing lower values for breast-milk volume. The degree of effect on the volume depends on the method used, and, in many cases, on the skill of the person making the measurements. For the purposes of this study, breast-milk quantity, volume, output, or yield will refer to infant intake or maternal secretory capacity, or both. Measurement of maternal ability to secrete milk. The most common method of measuring maternal secretion is by collecting milk by manual or mechanical expression, and measuring the volume over a 24-hour period. Hytten (5/) has observed that manual expression does not yield as high a quantity of milk as does the “test-feeding” method discussed below. On the other hand, he found that the expression of milk using the breast-pump yields higher values than the test-feeding method. Macy et al. (79) found manual expression to yield milk volumes similar to those obtained by Hytten with a breast- pump. The conflicting results among studies illustrate that results obtained by different methods are difficult to compare because of the many factors that may influence the outcome. Another method for measuring maternal secretory capacity is first to make a test-feeding (see below) and then collect the remainder of the milk by expression either mechanically or manually; the two values are then added together.4 THE QUANTITY AND QUALITY OF BREAST MILK Measurement of infant breast-milk intake. The “test-feeding” or “test-weighing” method of measuring the infant breast-milk intake is the method most often used. The quantity of breast milk that the infant ingests is calculated by subtracting the weight of the baby before feeding from its weight after feeding. If the infant is breast-fed during the night, care must be taken to carry out weighings during the night also; it is not sufficiently accurate to calculate the daily quantities on the basis of weighings over part of the day only. Another problem is that if the baby is given short, frequent feeds, the weight gains will be so small as to cause unacceptable errors in the estimates of the quantity of ingested breast milk (23), Infant milk intake has also been calculated from the weight of the mother before and after she suckles her child (97). However, the iculty of measuring accurately small changes of maternal weight is greater than that of measuring similar changes when test-weighing infants. A newly developed method for measuring the infant's intake of breast milk is based on the enrichment of saliva with heavy water (deuterium oxide). A precisely measured amount of heavy water is administered by mouth to the infant and the concentration of heavy - water in the saliva is measured before and 24 hours after the administration. The difference in the two concentrations indicates changes in the total body water over 24 hours, and can be used as a measurement of breast-milk intake if appropriate correction is made. for the intake of any other fluid. This method has the advantage of not interfering with the feeding habits of the baby or the life-styles of the mother and baby. It also has the advantage of being able to provide the average daily intake over several days (usually 11~14 days), whereas test-feeding measurements .are usually done over a one-day period (22). The first results obtained with this method appear, however, to give somewhat higher figures for breast-milk intake than those obtained by the test-weighing method. Breast-milk intake can also be measured by means of a flow meter connected to a nipple-shield through which the infant sucks at his mother's breast. In this way, information is obtained about the milk flow and its pattern, making it possible to calculate the quantity of breast milk ingested by the infant. This method may also be used to determine breast-milk composition. at any time during the feeding period because it makes continuous sampling possible (49, 76). Little, however, is known about the comparability of the results by this method with those obtained by other methods. It is likely that the presence of the shield might affect the child’s willingness to suck. 2.1.2 Factors influencing breast-milk quantity Psychological, physiological, and sociological factors may affect the quantity of breast milk. The fact that these factors are frequently2. REVIEW OF QUANTITY AND COMPOSITION 5 interrelated makes it difficult to identify the degree to which each of them is responsible for differences in the output. Psychological factors. Of all the factors influencing breast-milk volume, the impact of psychological factors has probably been known for the longest time. It has long been known in folk experience that the shock of unpleasant news can cause a mother's milk to dry up rapidly, as Jelliffe & Jelliffe (56) have pointed out. Emotional disturbances and anxiety are well known to interfere with the “let- down” reflex, thus causing less milk to be secreted? Physiological factors. Physiological factors influencing breast-milk quantity include the mother’s capacity to produce milk, her ability to release it, and the infant's ability to consume the milk as well as to stimulate the nipple so that more milk is secreted. The frequency, duration, and force of the infant's sucking all affect breast-milk quantity (123). Belavady? and Athavale (6) have pointed out that smaller infants need less milk to maintain an adequate growth rate and therefore they suck less, allowing the breast to produce and secrete less milk; by sucking more, the higher-birth- weight newborns cause more milk to be produced and secreted. This is consistent with the theory that the quantity of milk secreted by the gland can be considerably influenced by the demand (79). Fluctuations in breast-milk yields during one day, from day to day, and at different times post partum have been noted in several studies. Increase in volume over time, especially during the first few months of lactation, may be due to the greater strength of sucking by the growing infant and also to its greater nutritional needs. Hormonal factors may also play a role, Although Habicht and coworkers (42) and Deem (23) have suggested that breast-milk secretion does not vary during the day, most studies have shown a diurnal fluctuation (25, 51, 55, 78). Denis & Talbot (25) found the maximum breast-milk yield at around 09h00; Macy (78) found it to be in both morning and late evening, Jaso (55) found maximum yields at 06h00 and 23h00. Almost all these studies showed wide fluctuations in milk yield, as well as differences between mothers in the time of the peak milk yield; on average, there seemed to be a trend among the data towards higher yields in the morning. It is possible that some of the observed diurnal variation may not be only a manifestation of the circadian rhythm, but may also be related to the intervals between feeds. Variations in breast-milk volume during the day, whatever their cause, make it necessary to measure the volume over a 24-hour period in order to achieve an accurate measurement. The amount of milk secreted varies with the period of lactation. It * Infant and young child feeding—current issues. Unpublished WHO document FHE/ICF/19.3, 1979. * Belavady, B. Personal communication (1974).6 THE QUANTITY AND QUALITY OF BREAST MILK increases sharply. during the first few days post partum (59), reaches a plateau at about 4-6 weeks post partum, remains at this level, sometimes for several months, and then starts to decline. Belavady' observed in Indian women that the highest values of milk volume were attained around 4-6 weeks post partum and were maintained for about one year. Similar observations have been made by other investigators in India and elsewhere (39, 97, 115). Hytten (52) observed that the yield from the two breasts differed in most women studied, the average difference being about 50ml per day. This difference could result from the mother suckling her infant more on one breast than on the other. ‘An increase in plasma concentration of prolactin appears to play a key role in the initiation (16) and maintenance of lactation (1/8). A study in Edinburgh, Scotland, (50) showed that a mother’s prolactin concentration remained above normal until there was a decrease in the daily frequency and duration of breast-feeding. This study found that the introduction of supplementary foods to the infant coincided with a decline in the prolactin concentration in the mother’s plasma. Hennart et al. (48) suggest that it is the decline in sucking frequency that causes the prolactin concentration to drop. Bunner et al. (/6), however, observed continued lactation in mothers whose basal normal concentrations of prolactin were not increased, Some drugs and hormones are known to influence breast-milk volume. Combined oral contraceptives containing 30yg of estrogen appear to decrease milk volume (74). Alcohol is known to suppress oxytocin release, and may therefore interfere with the let-down reflex. Thompson (104) found a correlation between heavy smoking and decreased milk production, although he pointed out that heavy smokers are often nervous and that this may lead to reduced milk production. Most cultures use galactogogues to help stimulate milk production. In several countries beer is recommended as a galactogogue. Jelliffe & Jelliffe (56) suggest that part of the effectiveness of many galactogogues comes from the soothing belief that they will work, thus allowing the mother to relax and facilitating the stimulation of the let-down refiex. Social factors. Social factors may have an influence on breast-milk quantity, though their influence is. usually through psychological or physiological mechanisms, or a combination of these. For example, a society may believe that breast-feeding is distasteful and this may make the mother feel uncomfortable about doing so and thereby cause inhibition of the let-down reflex. In a society in which the mother must work away from home, there may be a negative influence on her capacity to secrete breast milk, on the amount of breast milk the infant ingests, or on both, unless special facilities for breast-feeding * Belavady, B. Personal communication (1974).2. REVIEW OF QUANTITY AND COMPOSITION 7 are provided at the place of work. The breast-feeding working mother is often forced to give the baby breast-milk substitutes or supplements while she is away from home. This use of breast-milk substitutes or the early introduction of food supplements may also influence the quantity of breast milk ingested by the infant. When the baby’s nutritional needs are partially met by foods other than breast milk, the baby will suck the breast less frequently and less vigorously, and, consequently, less breast milk will be produced (19, 87). 2.1.3 Influence of maternal nutritional status and diet on breast-milk quantity The relationship between breast-milk output and the maternal nutritional status is complicated by the fact that many aspects of the latter defy a clear and generally applicable definition; at the same time psycho- logical and sociological factors may also be important. While it is safe to assume that maternal nutrition may, to some extent, influence the output of breast milk, output may also be influenced by maternal nutrition indirectly by way of a psychological effect. For instance, when a mother fears that she has lost too much weight or has not eaten well enough to produce “good” or “enough” milk for her baby, this may inhibit the let-down reflex. On the other hand, the mother’s nutritional status and diet are often determined by the social norms of the society to which she belongs—e.g., there are often prescribed quantities and types of food for mothers. The mother’s ability to acquire food and what is considered to be the desirable body-weight for a woman in her particular society also influence milk output. Maternal nutrition may thus influence the breast-milk volume either directly or indirectly, but it is difficult to determine whether maternal nutrition is primarily responsible for changes in breast-milk output. To date, most of our knowledge concerning the impact of maternal nutrition on breast-milk output comes from maternal diet supplemen- tation studies. Their interpretation is not always easy for the following reasons: (a) the food supplement may replace part of the diet instead of complementing it; (6) the length of diet supplementation may be too short; (c) the timing of supplementation may not be right—instead of starting supplementation during lactation, it may be advisable to start earlier, for example, during pregnancy in order to build up the fat stores; (d) the quantity and the composition of the supplement may not be correct in relation to the daily food intake of the mother before and during dietary supplementation; and (e) the diet supplementation of a mother may not be correctly related to the degree of malnutrition. From the above it is easy to understand why studies of the effect of maternal diet supplementation on breast milk have produced mixed results. Chavez & Martinez (18) were able to increase milk volume8 THE QUANTITY AND QUALITY OF BREAST MILK with supplements of 1260 kJ/day (300 keal/day) given from the 45th day of gestation until weaning; although there was a measurable positive impact on child growth, the milk became more dilute. Postpartum protein supplementation of 25-50 g of protein per day in Nigeria produced increases in infant intake and maternal secretion capacity (30). In this study, the mothers were able to secrete more milk than the infants consumed even before diet supplementation, and after the commencement of diet supplementation the babies’ milk consumption increased for some unexplained reason. The babies showed a significant improvement in growth (both in height and in weight) compared with the growth of babies in a control group. In the Gambia, nursing mothers’ diets were supplemented with over 2940 kJ /day (700 kcal/day) for 12 months so that the mean maternal intake was 9622kJ/day (2291 kcal/day) (92); although the Gambian mothers showed an initial improvement in maternal body weight and subcutaneous fat stores, there was no increase in the intake of breast milk by the infant. Studies by Belavady (/2) and Gopalan (46) also showed no increase in breast-milk yield with diet supplementation of the nursing mother. ‘A few studies have looked at maternal nutrition and breast-milk volume, and have tried to draw inferences about the influence of maternal nutritional status on the volume of breast milk. Khin Maung Naing and coworkers (61) found that while well-nourished mothers were able to secrete significantly more milk than infants were able to ingest, this was not the case for malnourished women, Hanafy et al. (47) found that malnourished women produced 22% less milk than the well-nourished ones. Bailey (8) was unable to find a correlation between the maternal weight:height ratio or skinfold thickness and the quantity of breast milk, except for severely malnourished women. As for the relationship between the mother’s dietary intake and milk output, a recently conducted study in the Gambia showed a significant drop in breast-milk quantity during the rainy season when the already low daily food intakes of the mothers dropped by an additional 1050-1260kJ/day (250-300 kcal/day) (92); however, the rainy season is also the period with the highest infant morbidity. It was pointed out in this study that for infants born during the dry season the mother's milk nourished the infant for the first three months of life so well that the infant grew faster than usual despite the fact that maternal dietary intake was well below that recommended for even non-pregnant, non-lactating women. 2.1.4 Comparison of data on breast-milk intake in developed and developing countries. Several investigators have reported that the average daily intake of breast milk by infants in developed countries is in the range of 600-800 ml/day rather than 850 ml/day, which is the figure usually2. REVIEW OF QUANTITY AND COMPOSITION 2 used in calculations of the dietary requirements of lactating women and of infants below 6 months of age (104, 123). The highest quantities reported for a population were by Rattigan and coworkers (97), who used the method of test-weighing the mothers; the quantities were 1187 ¢/24h, 1238 g/24h, 1128 gj24h, and 880 g/24h at 1, 3, 6 and 12 months of lactation, respectively. There have been reports of wet-nurses having similarly high quantities of breast milk (78). Schanler & Oh (99), studying the breast milk of mothers of premature babies in the USA found that the mean volumes of expressed milk during the first couple of weeks post partum ranged between 1098 and 1673 ml/day. The volume of milk the infants ingested was, of course, consider- ably less than the mother's supply. A comparison of breast-milk intake in developed and developing countries indicates that the intake is smaller in the latter, as illustrated in Table 1. 2.2 Composition of Breast Milk Breast-milk quantity alone does not give a clear picture of its nutritional adequacy. To get a better picture, one must also look at the composition of the milk. The method of breast-milk collection may influence the composition, and this must be considered when comparing data from different studies, Other factors that need to be taken into account when comparing breast-milk composition data are listed in section 2.2.1 below. Table 2 gives data from some recent studies on breast-milk composition conducted mainly in developing countries. 2.2.1 Factors affecting breast-milk composition The factors that may influence breast-milk composition and which may need to be considered in assessing data on the composition of breast-milk are as follows: (a) changes in the composition of milk over time (according to length of time post partum); (b) changes in milk composition during a single feed; (c) diurnal fluctuations in composition; (d) effect on composition of the interval between feeds; (e) differences in milk composition between the two breasts; (f) changes in composition attributed to mother’s menstrual cycle or pregnancy; (g) effect on composition of the degree of pressure exerted if the milk is expressed; (H) effect of the length of gestation prior to the beginning of lactation; and () effect of the interval of time between pregnancies.THE QUANTITY AND QUALITY OF BREAST MILK 10 eseuunod snouier ut (Aep/6) IW rsee1q jo AnUEND bergen 696-976 se 09g ey os (ezot-tes) (901-423) Lyk 1ze 188-998 £08 vis-08g els oy (696-6) (6z01-2z9) eek 86L 86L-79L s08 Les 967 { ve 619 cae ely ez (cig-cep) (468-62) 199 ‘989 £99-0r9 us our we (cece) (18¢-L98) BLL { es i) zee, ety vay sojewes sorew (2) epeueg (0Z4) puer6ua (z6) equeD (ZB) OOK (6) EIU! (OL) eAUoY a Jo (s)uuow (214) uepemsMN 2. REVIEW OF QUANTITY AND COMPOSITION 11 ha ova 904-72 #840 OP 8AM SEDI IY gis ve-ee 2-2 909 wee toe 1z-0z var oz-6t rel (ose els Brut 9ES-1 Lg t LoL ies 9st ov sir e8s-95 sis viel yy ez uy Zit 89-19 Hob ely Olé zov eer { 68 cve-si8 98 bto-pes et (oc6-018) (ez -445) ove. tie z18-6e8 y98 i)THE QUANTITY AND QUALITY OF BREAST MILK 12 (s) 99% = (s) 9s4 - smuow 6'¢-'1 (uowom ($1) o9'9 = (Si) eve = syquow ¢'1-s°0 peBainud) (s) 99 = (8) 997 - ‘SyIvOU! 0-0 a) edomna (sezze (sez) ve (ER) ZL - sunuow 24 (gez)ec —(gez) ge (Bez) EL = ‘suiuou LL (Bez) &% (gz) €€ (g€z) £40 a syvoW OL (sez) 2 = (Bez) VE (EZ) PLO = syluow & (sez) $2 (ez) 8Z_~— (EZ) SL'O syiuow g (sez) 992 = (Bez) CZ _=— (EZ) 6L'O = syiuow Z (ez) 92 (ez) 9Z_~— (EZ) ZB'O = syluoW g (sez) 92 = (ez) SZ_—«(EZ) L610 = syiuow g (sez) £2 (ez) 9Z_~—«(BEZ) BB'O oe syuow y (sez) 72 = (ez) EZ ~—(BEZ) 6B; - syjuow ¢ (sez) Zs (ez) Sz (Bez) OVE - syiuow Z (wez) oe (Bez). (BEZ) EOL = wuuow 1 (4) 1483 (ez)ore (ez) ss'@ =~ (6z) 90b (62) (144 00L/Ie24 v9) 692, syiuow 21-2 (9z) S02 = (dz) eve = (92) SVE (92) (1m OOL/Ie>4 G9) Ezz, suuow! 2-7 (ez) 62 (BZ) BLE (OE) BLL_—_eL 82) (us OOL/IeDY Z9) 09 sqquous 7-1 (4g) euing (1m ooV/8) (1001/8) (144 001/68) (Ww oot/ra) voxel sem ojduses 1s 801087 we4 ureioid A610U3 tueym ‘winued 150d eu), ‘Anunog YIU 3see1q Jo susuoduiod 10few! UO saIpms Jo suNsey —Z 1GeL13 2. REVIEW OF QUANTITY AND COMPOSITION (LL) yore (24) ce¢ (ZL) zoe (GL) 669 (ce) 6e't (ve) eve (av) e6'9 (99) £69 (6c) 969 (gt) isc (€2) vL (62) 99: (ge) ore ©) Wwe (6) ove (sr) exe (92) er (yh) eve (e) ere sass yo quo lepuyeasaqusie u sind e payods assouto SUN (ce) evz (v2) Lez (38) oe (a0) o6'y (99) 68’ (62) es (sh) we (v2) 6L€ (ez) eve (ge) vee (e) avy (6) ery (uy) (zu) (zi) 1) (ee) (vz) (8) (sv) (99) (62) (oe) (se) (se) (sr) (s) (we) (1) (sr) (sz) (wi) (e) ove ot as ze $60 601 tor ove we we eve SOL Lob 90° leL lez 968 we get wh Loz (1) (14 owieox 99) Lez (ez) (sz) (ge) (€) (Iw OoL/e94 G9) ELz (6) (iw ooLieos 92) ozé sujuow ¢'9-G'¢, sujuow ¢'E-s'| sywuow ¢'1-$'°0 ‘syuow $'0-0 suuow 9 < syuow 9 > sywuow €2-0 sue0A €-% sue0A 2-1 1884 1-0 syiuow gt < sywuow gt-eL suuow 21-2 syuow 9-Z skep OF LL ‘shep OLE skep 2-1 syuow g'9< syiuow g'9-G°e syjuow ¢'e-G'L sywuow ${-'0 (#2) vapams, (LL) eouns mon (041) ehuey (G4) eseuopyy (€1) erpuy14 THE QUANTITY AND QUALITY OF BREAST MILK The above list is by no means exhaustive, and other factors may also influence breast-milk composition. (a) Fluctuations in breast-milk composition over the duration of breast-feeding. Changes in the composition of breast milk over time are so great that breast milk has been classified into three types. Colostrum is the first; it is generally considered to last from birth until about 5 days post partum. The next type is transitional milk, which lasts from the 6th to around the 15th day post partum. The final type of milk is called mature milk; it starts around the 15th day post partum and lasts until the complete weaning of the child. The periods of secretion of these different types of milk may vary from one woman to another, and different researchers have defined the periods differently. The composition of breast milk changes relatively rapidly during the first few days post partum (54). Many of the components of milk are found in higher concentration in colostrum than in mature milk. Protein concentration is higher, the peak being reached in the first few days post partum (46, 66), droping sharply thereafter (13). This high concentration appears to be due, inter alia, to a higher content of antibody proteins, especially immunoglobulin A and lactoferrin (46). Some vitamins, minerals, and elements have also been found in higher concentrations in the colostrum than in mature milk. Among these minerals are calcium (54), iron (54), mercury (57), molybdenum (63), phosphorus (54), potassium (54), possibly selenium (54, 72, 73), sodium chloride (66), and sulfur and zinc (54). Kulski & Hartmann (66), however, found lower values for calcium and phosphorus in colostrum than in mature milk. The higher concentrations of iron, nitrogen, and sulfur in colostrum have been linked with its higher concentration of protein (46,54). Among the vitamins, vitamin A (13,81), and vitamin E (13) have been found in higher concentrations in colostrum than in mature milk, The concentrations of water- soluble vitamins appear to increase rapidly during the first few days post partum (13). In transitional milk, the concentrations of most constituents shift gradually to. the concentrations found in mature milk. Some components, however, drop rapidly in concentration during the first few days post partum. For example, the concentrations of iodine and magnesium drop very rapidly at the onset of transitional milk and remain low throughout lactation (20,63). Thus, the time boundaries for transitional milk may not necessarily be the same for all breast- milk components. Various studies comparing mature milk with colostrum have found mature milk to be richer in fat, lactose, glucose, and urea (13,66). The concentration of fat, lactose, magnesium and manganese appears to remain stable after the onset of mature milk. However, the concentration of some components has been reported to decrease2, REVIEW OF QUANTITY AND COMPOSITION 1s between the onset of mature milk and weaning, For example, the concentrations of sodium, phosphorus, copper, iron, calcium, and zinc have been shown to decrease slowly over the duration of lactation (10, 11, 54, 112). There are many inconsistencies between studies on changes in the concentration of various constituents in mature breast milk during the progress of lactation; these changes, however, are partly due to the differences in the methods of sampling and the difficulty of obtaining representative samples of breast milk. In Burma, Khin Maung Naing and coworkers (61) found that the protein content was significantly higher at 1-4 months post partum than at 7-12 months post partum. In Indonesia, however, Boediman and coworkers (/5) found the mean concentration of protein in breast milk to be the same for each of 3 years post partum. Fomon (32) noted that the protein content decreased during the first 6 months of lactation, and then remained constant from 6-24 months post partum. Abdel Kader et al. (/) found the fat concentration in breast milk to increase with the duration of lactation rather than remaining the same, as other studies seem to indicate. The unexplained differences in trends reported in various studies suggest that more attention should be paid to physiological, nutritional, or other factors which may influence breast-milk composition, as well as to the possible influence of the volume and the concentration of the nutrients. (6) Diurnal fluctuations in composition. Hytten (51) found wide diurnal fluctuations in fat content. He observed that the fat concentration in milk rises between 05h00 and 10h00 and then gradually declines. Picciano (88) found that the iron concentration increases during the daytime. Not all breast-milk components have been tested for diurnal fluctuations, but copper, zinc, and magnesium appear to remain constant throughout the day (94). (0) Changes in composition during a single feed. It is known that fat concentration increases during suckling (44,78, 102). Protein has also been shown to increase slightly over the duration of the feed (25, 44,78). Hall (44) found that the lipid concentration was about 4-5 times greater in hind-milk than in fore-milk and that the protein concentration increased by 50% during one feed, although the extent of change varied between women. (4) Changes in composition attributed to the interval between ‘feeds. A study by Kon & Mawson (65) showed that the longer the interval between feeds, the lower the concentration of fat. (e) Differences in the composition of breast milk between the mother's two breasts. Hytten (52) observed the differences in milk composition between the woman's two breasts: 92% of the women16 THE QUANTITY AND QUALITY OF BREAST MILK studied had differences in fat concentration, most of which were statistically significant, averaging approximately 14%. The differences between the breasts in lactose and nitrogen concentrations were small, but if one breast produced milk with a higher concentration of fat, that milk was also likely to have a higher concentration of protein. There was no apparent pattern to the differences in composition. (A) Fluctuations in composition attributed to a mother's menstrual cycle or pregnancy. Prosser & Hartman (93) have found increases in sodium and decreases in potassium and lactose concentrations in breast milk during the mid-follicular and mid-luteal phases, and have suggested that the baby's refusal of the breast at certain times during the menstrual cycle may be related to this. Belavady (/0) reported a difference in breast-milk composition between pregnant and non- pregnant lactating women. She found pregnant women’s milk to have twice as much protein and an increase in vitamin B,, and total ash compared with non-pregnant women’s milk. (g) Fluctuations in composition attributed to degree of pressure exerted in expressing breast milk. Widdows & Lowenfeld (125) reported that the degree of pressure used to express milk influences the fat content of the milk; this may be linked to the already- mentioned compositional changes during the feed. (h) Differences in ‘composition according to length of gestation. Gross and colleagues (41), comparing the milk of mothers of premature babies with that of mothers of full-term babies, found that the milk of the former had significantly higher concentrations of protein, sodium, and chloride but a lower lactose concentration. The energy content was found to be similar, as were the concentrations of potassium, calcium, phosphorus, and magnesium. Schanler & Oh (99) found similar results when they compared the breast milk of mothers of premature babies with milk from a donor pool; in this study, the increased concentration of protein, sodium, and chloride appeared to last for about 3 weeks after birth. ( Interval between pregnancies. In a study of the effects of age, parity, and birth-interval on milk composition, Abdel Kader and coworkers (1) found birth-interval to be the most important factor; the concentrations of protein and lactose appeared to increase with longer intervals between pregnancies, 2.2.2 Influence of maternal dietary intake and nutritional status on breast-milk composition. Most studies have found the energy, protein, fat, and lactose contents of mature breast milk to be little influenced by variations,2. REVIEW OF QUANTITY AND COMPOSITION 7 within certain limits, in the nutritional status or dietary intake of the mother (8, 13,61, 74). The energy content of breast milk in the Gambia, where during the best months of the year lactating women were consuming only 7.1 MS day (1700 keal/day), was 302 kJ/100 mi (72 keal/100ml). This is very similar to the energy content of 290 kJ /100 ml (69 keal/100 ml) found in the milk of British women. Furthermore, it was found that when the lactating mothers in the Gambia had an intake of only 4.6-5 MJ/day (1100-1200 kcal/day), the energy value of their milk decreased by only 10%, (123). In studies of the breast milk of Gambian women, economically advantaged and disadvantaged Ethiopian women, and Swedish women, all samples showed similar concentrations of protein despite clear differences in dietary intake and nutritional status (74, 123). Even during the time of the year when Gambian women had the lowest dietary intake, the total nitrogen content decreased by only 10%, Hanafy et al. (47), however, observed a lower concentration of protein in the milk of malnourished mothers. Lindblad & Rahimtoola (70) found the protein content of the milk of very poor Pakistani women to be about the same as that of women from more affluent countries. The casein to whey-protein ratio has been reported to be higher in the milk of malnourished women (77). The effect of protein intake on the protein concentration of breast milk has been studied by manipulating the dietary protein intakes of mothers. Forsum & Lonnerdal (34) studied the effect of high protein (20% of total energy intake) and low protein (8% of total energy intake) diets on the milk of three Swedish women; the high-protein diet was associated with a higher concentration of total nitrogen, true protein, and non-protein nitrogen in the breast milk, but no difference was observed in the concentrations of lactoferrin, lactalbumin, or serum albumin. Edozien and coworkers (30), on the other hand, found that the protein concentration in breast milk remained unchanged after diet supplementation with protein. It should be noted that studies on the effects of dict supplementation on breast-milk composition are subject to the same methodological problems as those found in studies on the effects of diet supplementation on breast-milk quantity. Lonnerdal and coworkers (74) found the same lactose concentration in the breast milk of economically advantaged and disadvantaged Ethiopian women and of Swedish women. Khin Muang Niang et al. (60) reached the same conclusion when studying malnourished and well-nourished mothers. Edozien et al. (30) found the lactose concentration in breast milk unchanged after diet supplementation with protein. It appears from these studies that lactose concentration is not influenced by the diet and nutritional status of the mother. The concentration of fat in breast milk also does not appear to be18 THE QUANTITY AND QUALITY OF BREAST MILK influenced much by the mother’s diet. For example, British women get about 40°% of their dietary food energy from fats, whereas the Gambian women get no more than 10%; yet, one study (123) found the fat concentration in the breast milk of the two groups to be essentially the same. On the other hand, the fatty acid pattern appears to be influenced by the diet (45,53). Insull and colleagues (53) found that women who were-on a diet rich in polyunsaturated fats produced milk with a high content of polyunsaturated fats, while the total fat content remained the same. A low intake of water-soluble vitamins by the mother results in a decreased concentration of such vitamins in breast milk (/3, 26, 27). In more recent studies, the concentrations of riboflavin (90), vitamin C (9), thiamin, folic acid, vitamin B,, and pantothenic acid (/23) in breast milk have been found to be closely correlated with the mother’s dietary intake and/or the concentrations of these vitamins in her plasma.’ Supplementing the mother’s diet with vitamins, particularly water-soluble vitamins, has shown in a number of studies to increase the vitamin concentration in her milk (14, 58,65, 95, 114). It appears, therefore, that the concentrations of water-soluble vitamins in breast milk are influenced by the mother’s diet and nutritional status. As to fat-soluble vitamins in human milk, Lonnerdal et al. (74) found a significantly lower concentration of vitamin A and B- carotene in the breast milk of economically disadvantaged Ethiopian mothers compared with the milk of well-nourished mothers. In Guatemala, Arroyave (5) has shown that the fortification of sugar with vitamin A resulted in an increased concentration of vitamin A in the breast milk. As far as minerals and trace elements are concerned, the influence of the dietary intake varies widely. Potassium concentrations have been found to be independent of dietary intake (54). DeFilippi and coworkers (24) compared the breast milk of women on a low salt diet with the milk of women on a normal diet and found no significant differences in sodium or potassium concentrations. Supplementation of-maternal diet with iron and manganese failed to show an increase in the iron or manganese content of breast milk in several studies (14, 84, 108). Of all the minerals, the relation between the calcium concentration in breast milk and the nutritional status and diet of the mother has been the most variable. On the one hand, there have been reports claiming that calcium concentration in human milk is uninfluenced by the nutritional status or the dietary intake of the mother (/3,84). In another study, poorly fed Chinese mothers with histories of severe calcium depletion were able to. produce milk with normal calcium concentrations (7/). On the other hand, Drummond and coworkers (28) found the calcium concentration of breast milk to be dependent on the calcium intake of the mother, and Toverud & Toverud (1/06)2. REVIEW OF QUANTITY AND COMPOSITION 19 observed that with the administration of calcium and vitamin D the concentration of calcium in breast milk did increase. 2.3 Adequacy of Breast Milk in Supporting Infant Growth ‘The amount of breast milk required by an infant at a specific age and the most beneficial composition are still being studied. At present one of the best ways of assessing the adequacy of breast milk is to determine how well it sustains the infant’s rate of growth along standardized growth curves. However, this method has some limitations, First, the infant’s growth is influenced not only by the maternal ability to secrete milk and the infant's capacity to suck milk, but also by the infant’s ability to utilize the nutrients in the milk, which may be adversely influenced by disease. Second, it has been pointed out (/07) that the standardized WHO growth reference values come from a population with a large proportion of non-breast-fed and mixed-fed infants, and it may not be correct to assume that these infants grow at the same rate as infants who are exclusively breast- fed. In fact, some studies have recorded accelerated growth spurts in exclusively breast-fed infants in the early post-partum months (7). However, it must be remembered, particularly when dealing with data for infants of various races for whom specific standards may not be available, that one is more concerned with whether the infants are growing at the rate appropriate for their age than with their actual centile position on the chart that happens to be available. Table 3 lists studies on the duration of breast-milk adequacy. The Table 3. Age at which the growth curves of exclusively breast-fed infants start to diverge from reference curves in different countries 3-4 months 5-6 months Country Reference Country Reference Bangladesh Khan (60) China, Peoples Shanghai Child Egypt (urban) Shukry et al. (707) Republic of Health Coordi- ration Ethiopia Eksymr (37) Group (100) Gambia Whitehead ot al. Egypt (urban) —_Lebshtein & El Bahay (721) (69) Jordan Kimmance (62) India Ghosh (38) India Gopalan (39) Iran (urban) Geissler et al. (36) Ivory Coast Lauber & Reinhardt (68) Mexico Martinez & Malaysia Dugdale (29) Chavez (82) Sierra Leone USAID (109) New Guinea Bailey (8) Sweden Waligren (177) Tanzania, United Poeplau et al. (89) Uganda Welbourn (119) Republic of United States Ahn & Maclean (2) of America20 THE QUANTITY AND QUALITY OF BREAST MILK Table 4. Some organochlorine contaminants of breast milk by country Organochlorine compound —_—_Year of study Sample size Median (mg/kg) and country DDT complex Austria® 1978 182 253 Canada 1975 100 0.034 Federal Republic of Germany? 1979 374 1.26 Guatemala 1974/1974 81/290 1.40/0.4 Hungary 1976 10-86 0.53-0.097 Japan 1979 365 0.037 Sweden 1979 41 0.046 ‘Switzerland 1974 6 0.063 USA 1975 1436 0.070 Hexachlorocyclohexane (HCH) isomers Canada 1975 100 0.001 Guatemala 1971 46 0.006 Hungary 1978 2 0.011 Japan 1979 353 0.035 Sweden 1979 41 0.0032 Switzerland 1975 6 0.0051 USA 1975 1436 0.003 Heptachlor and its epoxide Austria 1978 182 0.013 Guatemala 1974 a1 0.001 Japan 1979 33 0.0005 Switzerland 1974 6 0.003 Usa 1978 1436 0.001 Aldrin and dieldrin Austria 1978 182 0.009 Guatemala 1974 81 0.001 Japan 1979 53 0.0013 Switzerland 1974 6 0.0031 Hexechlorobenzene (HCB) 1978 182 1.74 1975 100 0.001 Federal Republic of Germany 1979 374 070 Hungary 1978 2 0.0040 Japan 1979 38 0.0025 1976 620 0.00003 1972 202 0.0300 ‘Sweden 1979 41 0.0041 Switzerland 1978 51 0.010 usa 1975 1436 0.001 * Sours: Joint FAQ|WHO Food and Animal Feed Contamination Monitoring Programme: summary of dte receive ram colaborating canis -1877 to 1960, Unpublahed WHO document Ne FAO. ESNIMON OCC(O1 2B, WHO ere 198, 198" The anahis of orgenochlaine compounds and heavy mata in breast mk ae in some cate # representative survey of the whale county and in otbereasae only of pat ofthe county. * Datacn oranochionne compounds am Ausra the Federal Repubc of Garmeny, and Mexico are expressed ona {at base: ta fom othr counties are ons whole milk ba.2. REVIEW OF QUANTITY AND COMPOSITION 21 studies have been divided into two groups: those that found infant growth faltering at about 5-6 months of age, and those that found infant growth faltering at about 3-4 months of age; all the infants in those studies were exclusively breast-fed. One of the difficulties of dividing such studies in this way is that the term “faltering” is imprecise, The meaning attributed to it by different authors varies and is not always clearly stated. It should be noted that breast-milk volumes in both developing and developed countries fail to meet the generally accepted energy requirements of the child (126) at about 3 months post partum (/22); however, despite this apparent “inadequacy” in volume or energy, in many studies infants have grown adequately far beyond 3 months of age. As a result, Whitehead and coworkers (J2/) have suggested that the recommended daily requirements are too high for the first 6 months of life 2.4 Contaminants in Breast Milk The amount and type of contaminants in breast milk are influenced by their solubility in fat or water, their concentration in maternal plasma, the degree of ionization, and the mechanism by which they are transported to the mammary glands. Since human milk contains a higher concentration of lipids than human plasma, liposoluble compounds, including liposoluble contaminants, will tend to con- centrate in’ the milk (64,86). Contaminants may include various medications, such as sulfa drugs, bromides, caffeine, and amphet- amines (33,43,67), as well as alcohol, nicotine, oral contraceptives, mycotoxins, and heavy metals (4, 54, 64, 85, 86) Pesticides, especially the organochlorine compounds, and heavy metals have received particular attention. Among the latter, mercury, lead, and cadmium have been found in human milk (35,57,64, 83). Contaminated food and water and atmospheric pollution are the principal sources of these contaminants (3, 35,37). Knowles (64) has noted that the concentration of some contaminants in human milk may be higher than that in the milk of other mammals; more DDT, for instance, can be found in human milk than in an equal volume of cow’s milk. Table 4 gives some data on the concentration of organochlorine compounds and heavy metals in breast milk.3. Design of the WHO Collaborative study 3.1 Population Groups Studied Since “nutritional status” is impossible to define directly, it was decided to compare women in strongly contrasted socioeconomic groups and to use anthropometric, biochemical, and clinical observations in an attempt to identify the possible differences in nutritional status within these groups. Following the scheme used in the first phase of the study (2/), it was decided to study three groups of mothers: —economically advantaged, educated families in an urban area; —poor and usually poorly educated families in an urban area; and —families in rural areas, usually following a traditional way of life and often dependent on subsistence agriculture and local marketing. The centres collaborating in the study were located in Guatemala, Hungary, the Philippines, Sweden, and Zaire. The communities from which the mothers were selected and the characteristics of the study samples have been described in the report on the first phase of the study—Contemporary patterns of breast-feeding (21). It was realized that in some of the collaborating centres the groupings indicated above would have little meaning, and modifications of this general scheme were agreed upon in advance. The groups studied in each of the collaborating centres together with the number of mother-child pairs are shown in Table 5. Annex 1, Table 1 gives the number of cases studied by study group and children’s age. 3.2 Methods Subjects were recruited for the study on the basis of information collected on each person using a set of record forms; Annex 2 contains sample forms and guidelines for their completion. Breast-milk quantity and composition were studied in mothers with infants of different ages. Sampling at 1, 3, 6, 9, and 18 months was 2324 ‘THE QUANTITY AND QUALITY OF BREAST MILK Table 5. Study groups and number of mother-child pairs Country Characteristics Group Number Hungary? Mothers with university education A 143 ‘Mothers with primary education 8 141 Mothers living in a rural environment c 141 Well-to-do mothers A 90 Urban-poor mothers B 114 Rural mothers c 160 Philippines Well-to-do mothers A 135 Urban-poor mothers B 114 Rural mothers c 237 ‘Sweden ‘Only one group of urban mothers was studied 168 Zaire Urban mothers? B ™ Rural mothers. c 722 + Attough initly analysed sepsiately, the thee groups were combined becbuse oftheir homogeneity in ai important charetwiis. "The aatincton betwebn very poor and not so poor ubae mothers was thought tobe impracticable mandatory, and additional studies at 2, 4, 12, and 24 months were optional. The acceptable limit of deviation was + 1 week up to 4 months of age, + 2 weeks at 6 months of age, and + 1 month thereafter. The highest age investigated in any particular centre or study group was dictated by the duration of breast-feeding. The minimum number of subjects at each age was stipulated as 28, except for the two highest age groups, where 16 cases were acceptable, The intake of subjects was to be spread evenly over the year to allow a check on possible seasonal fluctuations. Each mother-child pair was investigated on one occasion only. Mothers with multiple births or those using hormonal contraceptives (except in Hungary) were excluded, and efforts were made to limit as much as possible the number of primigravidae. The mothers selected for the study were free from recent or current debilitating diseases, and the children had no important malformations or disabilities that could impair sucking. Malnutrition, unless it was related to other diseases or to common intestinal parasites, was not regarded as a reason for exclusion. A clinical examination of the mother and child took place on the day of breast-milk quantity measurement to ascertain whether there were any complaints that could invalidate the measurement. The quantity of breast milk was estimated by test-weighing over a period of 24 hours, and was recorded in grams without ‘any adjustment for the interval between the first and the last feed. The records of mothers for whom the full test schedule was not completed were rejected at source. The details of the test-weighing procedure and of the collection of samples for laboratory analysis are given in3. DESIGN OF THE COLLABORATIVE STUDY 25 Annex 3. The methods used for the determination of the compo: of breast milk are described in Annex 4. Pilot studies were carried out in all centres in order to familiarize the research assistants with these procedures. In order to assess the validity of the test-weighing procedure, all centres were advised to test-weigh bottle-fed babies given feeds of 50, 100, 150, and 200 ml. To give an idea of the order of magnitude of the error involved in measuring a single feed, the results of such experiments carried out in the Philippines are reported in detail in Annex 3. With the exception of weight index and the classification of breast- fed babies, all other variables used in this study are self-explanatory and are defined in Annex 2. The weight index defined as weight (kg)/(height (m))*, correlated very little with maternal height, but correlated well with maternal arm circumference. However, in Hungary and Zaire (Group B) smaller women tended to be overweight, the correlation coefficients of —0.18 and —0.21 being significant at P<0.01. Although the weight index is an indicator of weight in relation to height, it is known from other work to be fairly highly correlated with adiposity (r~ 0.8) whether measured by the sum of skinfold measurements or by body-density methods. In order to avoid any misunderstanding, only the terms given below will be used in the report to describe groups of breast-fed babies. Breast milk only: babies receiving breast milk and nothing else. Fully breast-fed: babies receiving breast milk only and babies receiving breast milk and drinks or fruit juice of insignificant energy content. Breast-fed with supplements: babies receiving breast milk and any other foods of significant energy content. As already mentioned, mothers taking hormonal contraceptives were to be excluded from the study. However, this recommendation could not be followed in Hungary where nearly 20°% of mothers were found to be using hormonal contraceptives with a daily dose of 0,5 mg of etynodiol. Therefore, before proceeding with the analyses it was necessary to check whether this had any effect on the observed breast-milk quantities. Table 6 gives means and standard deviations of breast-milk quantities for Hungarian women using and not using etynodiol, and these figures show that there was no reason to exclude women taking this particular hormonal preparation since it had little or no effect on the observed breast-milk quantity. It was impossible to devise a common, or indeed a strictly random, method of selecting mothers. Therefore, a short description of the selection procedure, possible biases, and other relevant characteristics of the samples is given for each study centre.26 THE QUANTITY AND QUALITY OF BREAST MILK Table 6. Means and standard deviations. of breast-milk quantities (g) for women using and not using 2 hormonal contraceptive conti etynodiol diacetate (Hungary) Child's age (months) Use of hormonal —————~ contraceptives 1 2 3 4 6 No 6064123 6774148 6722146 6354161 5464161 3)" (72) (65) (65) (65) Yes 730 6494127 7094161 6201194 518152 a (ia) (20) (20) (20) + Figures in parentheses incicee number of subjects, 3.3 Hungary Three groups of mothers were investigated: ——mothers with a university degree or equivalent qualification; —mothers with primary education; and —mothers living in a rural environment, The first two groups were recruited from among mothers living in seven districts of the capital, Budapest, and the third group from among those living in the county of Pest. The task of recruiting cooperative breast-feeding mothers with children of the required age was undertaken by health visitors. Health visitors tended not to enlist mothers who breast-fed once or twice a’ day at the time of the recruiting interview for fear that they would not continue to breast- feed until the test-weighing session. Therefore, the group of mothers with infants aged 6 months or 4 months, ages at which the daily number of feeds tends to decrease, is likely to contain a somewhat inflated proportion of mothers who breast-fed more frequently. The guidelines to the record forms were strictly observed but two concessions had to be made: (a) some primiparous mothers were enlisted; and (b) the users of oral contraceptives could not be excluded, as explained above. The oral contraceptive in use contained a daily dose of 0.5 mg of etynodiol. All tests. were supervised by a health worker, and were carried out in the mothers’ homes. When a mother was judged to be intelligent and reliable, she was allowed to do the late-evening or night test- weighing by herself. 3.4 Sweden The mothers who delivered at the University Hospital, Uppsala, were enlisted for this study,soon after delivery. The selection did not3. DESIGN OF THE COLLABORATIVE STUDY 27 take into consideration the socioeconomic status of the mother. Very few of those invited refused to cooperate. Practically all mothers were given training in test-weighing while in hospital, In addition, one day before the test-weighing, a research assistant instructed the mother again and performed yet another trial run. The principles of selection specified in the guidelines were adhered to except for the inclusion of primigravidae. None of the mothers in the study was using an oral contraceptive. 3.5 Guatemala The urban well-to-do group (Group A) was recruited by obtaining names and addresses of women in the last trimester of pregnancy as well as of mothers with babies under 3 months of age from well- known doctors in Guatemala City. Of those contacted, 12% refused to take part in the study. The urban poor group (Group B) was recruited by means of a house-to-house search for mothers with 0-9 months old babies in a specified area of Guatemala City. People were advised of the study in advance through the health committee of the area, There were no open refusals, but 37% of mothers did not keep their appoint- ments. The selection of the rural group (Group C) was facilitated by the existence of birth records maintained by an on-going population research study by the Institute of Nutrition for Central America and Panama in the village of Santa Maria Cauqué. All families with children aged 0-18 months, as well as expectant mothers, were visited and invited to take part in the study; 11% refused to cooper- ate. In Group A the test-weighing sessions took place in the mothers’ homes. The field worker stayed from 07h00 to 19h00 and taught the mothers to weigh their babies. During the night the mothers were allowed to do the test-weighing themselves. Accommodation was provided for the mothers and infants involved in the Group B study. Women left their babies in the field workers’ care during the day and returned to breast-feed as many times as was usual for them. Mothers slept with their babies in the place of the study and returned home next morning after completing the test- weighing schedule. In the rural group (Group C) most test-weighing sessions took place in the mothers’ homes during the day and at the health centre during the night. This arrangement was preferred by most mothers. The guidelines for the study were followed rigorously, except for the inclusion of primigravidae. None of the women in the study was using a contraceptive of any type.28 THE QUANTITY AND QUALITY OF BREAST MILK 3.6 Philippines The urban well-to-do group (Group A) was drawn from mothers living in the town of Paranaque, part of metropolitan Manila. This group does not include the really wealthy families who were “unapproachable”; however, in terms of socioeconomic indicators, it is greatly superior to the group of urban poor (Group B) drawn from Pasay City, which is oue of the cities within the metropolitan area of Manila. The rural sample (Group C) was drawn from San Rafael, a town on the island of Luzon. The recruitment was done by means of a house-to-house inquiry and all eligible mothers were invited to take part in the study. The percentages of refusals were 19%, 1%, and 9% for Groups A, B, and C, respectively. The offer of ‘monetary compensation, free medicines (if needed), and free board and lodging during the test-weighing sessions influenced without doubt the rate of cooperation in Groups B and C. The test-weighing sessions in the urban well-to-do group were carried out in the subject's house under the supervision of research assistants between 0800 and 20h00. The overnight test-weighings were done by the mother herself (with the exception of 12 mothers who allowed the research assistant to stay with them for the full 24 hours), The research assistant returned before 08h00 the next day to perform the final test-weighing. In the urban-poor and the rural groups, all test-weighing procedures were carried out centrally in accommodation specially secured for the purpose. The study guidelines were followed in all respects except for the inclusion of primigravidae. Oral contraceptives were not used by any of the mothers included in. the study. 3.7 Zaii Two groups of mothers were investigated. The first, an urban group of mothers (Group B), does not strictly correspond to the definition of “urban-poor” given in the guidelines, but it does not include the economically advantaged mothers. The subjects were recruited from Bagira and Kadutu, which are two of the three districts of the city of Bukavu (east Zaire). The second group comprises rural mothers (Group C) who were recruited from the region of Kabare, which is to the west of Bukavu between Lake Kivu and the mountains of Mitumba. It should be noted that the two groups have different proportions of Tutsi and Bantu women, who have different physical charac- teristics, as will be shown later on. Bantu women predominated in both groups, but the proportion of Tutsi women was greater in Group C (rural) than in Group B (urban). Medical and administrative authorities in the two areas helped to publicize and explain the aims of the study. The recruitment was3. DESIGN OF THE COLLABORATIVE STUDY 29 carried out by research workers with the help of public health personnel. All lactating mothers were included in the study irrespective of the age of the child; refusals were very rare. Special accommodation for carrying out the tests was secured in both study areas. The tests were carried out under constant supervision. The guidelines in Annex 2 were strictly followed, but since all lactating mothers were included in the study, a proportion of primiparous mothers were inevitably enlisted. For practical reasons the analysis of breast-milk constituents was not performed on individual samples, but was done on pooled samples of breast milk taken at the same stage of lactation. Hormonal contraceptives were not used by any of the mothers included in the study. Lactating but pregnant mothers were excluded.4. Study of breast-milk quantity 4.1 Data Collected in Different Countries Annex 1 gives all the basic data collected in the study; tables of immediate relevance are included in the text, It is important to remember that the data are cross-sectional, and that any trends over time do not represent changes in any individual mother. Table 7 shows the mean breast-milk quantities, by the child’s age, country, and mother's socioeconomic group. The mean quantities for boys were slightly greater than those for girls, but the differences were small and erratic. 4.1.1 Hungary Milk quantity increased during the first 3 months and then started to decrease. The number of test-weighings (feeds) decreased with increasing child’s age (Annex 1, Table 7). A breast-milk substitute was being given to 10% of the children at 4 months and to 20% at 6 months (Annex 1, Table 12). By 3 months of age, about 50% of children were receiving other foods, mainly vegetables, and at 6 months 100% were receiving food supplements (Annex 1, Table 7) 4.1.2 Sweden Milk quantity increased up to 4 months and dropped sharply afterwards. The number of test-weighings decreased with increasing child’s age, more particularly at 6 months and later (Annex 1, Table8). The proportion of children receiving “energy” supplements—ie., supplements other than watery drinks or fruit juice—was very low for the first 3 months, reaching about 50% by the fourth month and nearly 100% by the sixth month of lactation (Annex 1, Table 8). Very few children received breast-milk substitutes (See Annex 1, Table 12); but, from the fourth month onwards some children were given a pap composed of diluted cow's milk (2/3 milk, 1/3 water) mixed with cereals; this was coded as “other supplements”. aL(suwow) 96¢ spn 7) (x9) (oa) (se) (ae) (6s) (99) (us) (os) (za) GELFELZ PLIFOSE ZL FREE PLIFLOv EGLFALE OLIFLGE Loi Fe9e ELLFOSE ZeLFo9e eGLteee 9 dnoID (8) (a2) (ev) (09) eo (su) (ga) (ee) (991) (se1) lezFzas G7zF1es B6LFZ/S ZGLFEGS E6LFEL9 BELT IPO SBLFLO9 ZozTEBS gszFoG9 pye+E09 a dnoID euez (oz) (zz) (92) (ze) (62) (ez) (sz) (oe) (2) BGLFELE GyLFloy GBLFZOS OLLF YES geLtERG LOZFElo IzzF7Z9 giztEs9 L8LF1ZG 9. dnoIn : (yy cr) (ez) (ie) (oe) (zz) (ze) (ez) (ee) - weLFyez SELF ley esters eglFess viztleg LolFoRs LiLFeeg partsg g4itzog =a dno1p (on) (on (on (oz) (sz) (ve) - a sit ize LULFeLe vezFeye ooztOze zrZT MOY LeLT9EE — ¥ dhol - (2) = (gz) (sz) (ee) = (2) 7 e127 26s a “81% 709 tL 88S = 1917 989 7 lerFers 9 dnoip - a a (gz) (sz) el (oe) 5 (gz) oglF 199 81 F 98s = eLLF 89. . egiF6is a dno = 2 = a = = a (gz) (oe) (ze) = = = = = A gezFes9 zzzt19s oyeFbza =v dnoIp ejewareng = = = = (ez) (sz) (sz) (sz) (gz) (gz) LOEFZay BOT F095 ELF LGL © SBF OLL «BLT SPL GL TZ¥O vopoms = a s (a8) (ge) (se) co) (v8) 2 fs 7 = fs OStF6ES goltleo crLFigo pyLFeL9 ez1FLO9 Avebuni cd 8h st zu 6 9 v € z b dnos6 pue Anunog “knunoo Aq saseo Jo raquinu pue u a6e s,piiyo pue ‘dois Apms In@p puepuers yum (6) Atnuend yw-Isee1q UEEW “2 819eL4, STUDY OF BREAST-MILK QUANTITY 33 4.1.3 Guatemala In Group A there was an increase in milk quantity up to 3 months of age; no data are available beyond this stage. “Energy” supplements were received by 60% of the children in the first month and by 75% at 3 months (see Annex 1, Table 9). About 50% of children received breast-milk substitutes from the first month onwards (see Annex 1, Table 12) In Groups B and C there was an initial increase in milk quantity up to about 6 months and very little change thereafter. The proportions of children receiving “energy” supplements rose in Group B from 35% in the first month to 45% at 6 months and to 60°% at 9 months. The corresponding proportions for Group C were 10%, 40%, and 70% (see Annex 1, Table 9). In Group B about 30% of children were receiving breast-milk substitutes at | and 3 months, while in Group C breast-milk substitutes were hardly used (Annex 1, Table 12). The number of test-weighings (or feeds) changed very little with the age of the child (Annex 1, Table 9). 4.1.4 Philippines In Group A, the mean breast-milk quantity was very small and did not change with the age of the child. Fewer test-weighings were done (Annex 1, Table 10) in this group compared with Groups B and. C. From the first month onwards, about 80% of children received “energy” supplements (Annex 1, Table 12) and about 70% received breast-milk substitutes (Annex 1, Table 12); this may explain the low quantities of breast milk recorded in Group A. In Groups B and C there was an increase in milk quantity at 2-3 months, then a gradual decrease up to about 15 months, and a fairly sharp fall thereafter. “Energy” supplements were given to about 30— 35% of children up to 4 months of age, to 70% at 6 months, and to 90% or more at 9 months and older (Annex 1, Table 10). Breast-milk substitutes were given to 20-30% of children at all ages (Annex 1, Table 12). The average number of test-weighings showed little decrease with increasing child’s age up to about 12 months of age (Annex 1, Table 10). 4.1.5 Zaire There was a large difference in the mean breast-milk quantity between Groups B and C. (It should be noted that the nutritional status of Group C mothers was much poorer than that of mothers in Group B). For both groups the intake remained nearly constant throughout the period of lactation. Breast-milk substitutes were rarely used in either group. In the initial stages of lactation there was a difference in the amount of food supplement given to the children in Groups B and C. While in the34 THE QUANTITY AND QUALITY OF BREAST MILK former the proportion of children fed with supplementary foods rose from about 10% to 40% during the first 3 months, in the latter group the corresponding figures were 50% and 70%. This may be related to the differences in the quantity of milk observed in the two groups (Table 6). The two groups were very similar after 4 months, and the proportion of children receiving supplementary foods rose from 70% to 100% at the age of 12 months. The average number of test-weighings decreased only slightly with increasing child’s age. 4.1.6 Discussion of the findings The number of mother-child pairs in individual age groups of children is far too small to make inferences about the shape of the breast- milk quantity distributions. It may be noted, however, that 41% of differences between the mean and the median were below 10g, 67% below 20, and that the number of occasions on which the median was smaller than the mean did not differ much from that on which the opposite was true (see Table 6). The mean breast-milk quantities recorded for Sweden were the highest among all countries in the first 4 months (significantly higher than any other group at months 2, 3, and 4). The mean quantities recorded for the Philippines Group A, and for Zaire Group were significantly lower than those for any other group for reasons indicated above, Otherwise there were no consistent differences over the range of ages studied. Breast-milk quantity is determined by the interplay of maternal supply and child’s demand, both of which are affected by a number of factors, including the breast-feeding routine. Before test-weighing the mothers were asked how many times they had breast-fed the child during the preceding 24 hours. It can be seen from Annex 1, Tables 7-11 that the reported number of feeds during the preceding day tended to be slightly less than the number of feeds during the test. It may be that having no other duties some mothers fed the child more frequently than was usual. The discrepancies may also be due to day- to-day variation in the number of feeds given by mothers who did not follow a rigid schedule, or simply to inaccuracies of recall by mothers, On average, however, there was reasonable agreement between the reported number of feeds during the day preceding the test and the number recorded during the test. Thus, although test-weighing may affect the usual breast-feeding routine to some extent, the aim of disrupting it as little as possible seems to have been at least partially achieved, Table 8 shows that there was a tendency for the mean breast-milk quantity to increase with the number of suckling sessions (ice., number of test-weighings). There was also a small correlation between the duration of suckling in minutes per 24 hours and the observed breast-milk quantity.4, STUDY OF BREAST-MILK QUANTITY 35 Table 8. Average breast-milk quantity (g) by number of test-weighings Child's age (months) County and number of 1 12 test-weighings Hungary <4 = — 470 (44) — 46 640 (20) 676 (75) 614 (41) = >6 597 (64) 714(10) = Sweden c : ae. = 429 (11) 324 (18) _ 46 654(15) 772 (20) 644 (17) 723. (9) >6 628 (13) 787 (8) — = Guatemala <8 438 (33) 584(31) 483 (6) 469(10) 89 549 (26) 628 (28) 576 (30) 619 (24) — >10 B14 (29) 674 (26) 695 (20) 625 (22) — Phitippines (Groups B and C) ra 318 (7) 481 (8) 389 (9) 317 (6) 398 (12) 89 4484 (27) 631 (31) 645 (28) 664 (36) 854 (40) >10 646 (25) 689 (21) 609 (21) 573(15) 677 (3) Zaire 7 (Group 8) <8 367 (14) 320 (4) 472 (6) 476 (5) 427 (2) B9 439 (24) 480 (34) 519.(22) 507 (9) 445 (12) >10 686 (97) 666 (61) 683 (67) 642 (58) 639 (46) (Group C) <8 201 (6) 273(10) 178 (12) 27411) 297 (8) 69 294 (14) 305 (16) 326 (36) 347 (25) 411 (32) >10 384 (32) 409 (31) 412 (52) 414 (55) 425 (45) + Flours in arenthene indicate numbers of subjects. Table 9 shows the duration of suckling per 24 hours and the estimated breast-milk quantity per 15 minutes of suckling. The mean duration of suckling sessions did not vary greatly between study groups and was about 15-20 minutes. The duration of suckling per 24 hours decreased with child’s age in parallel with the number of feeds, the decrease being much more pronounced in Hungary and Sweden. The high quantities per 15 minutes in Hungary and Sweden and the very low quantities in Zaire Group C should be noted. It may also be noted that the average quantities per 15 minutes are‘THE QUANTITY AND QUALITY OF BREAST MILK 36 Sus 809 86s 8 es 829 80s uw gi Jed Ainuen osFerL = ke Fert S Opes: = WWF zg1 og F eg, = pa FzaL voneing 9 dnoip Bus Bex 829 8 ap tu g1 sed Aunuen - a . eeFzer — ceFozL «= oy FESL «=z F EOL vowing a dnoip 89s 86e uw g1 40d Aunueng, . - : : : voFest 89 F E0z eng vy dnoig eyewerend Buy 89g Ben 899 iw gL Jed Anuen . 2 seF or StF ys ovFeor ws F Gy uoneing wopems Leo Bau 828 tu gy 460 Arnuen 5 a : . wes 9zF18 fF LLL uoneing ‘AeBunyy et st zt 6 9 € 1 {syiuow) Be s py 188019 UeaW pue (Gg F siNoY pz Jed UW) BuI}Y9NS Jo UONeINp UeEW “6 eIeL37 4. STUDY OF BREAST-MILK QUANTITY Bye 9sF at Bz ev izh 8 vy le LoL 8ze er asi 859 eve lel 819 687 vol oy a7 ¥6 sie zaF vol 889 ovF 1eL 569 ze¥60L Bee SF LL 869 Ww vel 5 04 oes vit Bue ee 60 8.99 6z>98 5 6z to s@L 6 ox ev eb 89 ee 6z1 8 ze aye 2b Sus oz 66 Boe esF tat 879 eet zp 819 ov opt Soe 66> 08L 8 sy ge 901 Buz 997 t6L 609 ly zs 5o9 ear apt 59g LeF eth Boy 9S OLL 29igne yo mEWON uw g4 40d Anuenp uoneing, ‘9 dno ui g1 49d Anueng, voneing a dnoip, ‘aNeZ uw gy 2d Anueno, uoneing ‘9 dno1p ww g4 10d Arnueng, voneing g dnop, uw gy sd Arnueng, voneing, vy dnoig souddyy38 THE QUANTITY AND QUALITY OF BREAST MILK remarkably stable from 6 months of age onwards. The results for Hungary and Sweden indicate probably good milk flow and/or vigorous sucking while those for Zaire Group C poor milk flow and perhaps less aggressive sucking, 4.2 Maternal Characteristics and Breast-milk Quantity There were no correlations between breast-milk quantity and maternal characteristics, including weight, weight index, serum concentrations of total protein and albumin, erythrocyte volume fraction, and haemoglobin concentration, except for values obtained in Zaire There, breast-milk quantity and maternal weight in Group B (urban) showed a low correlation within children’s age groups (r= 0.10, P<0.01, 7504.£.) and maternal serum protein and albumin concentra- tions were correlated with breast-milk quantity in Group C (r = 0.15 and r = 0.12, respectively, P<0.01, 678 d.f.). However, in spite of the lack of significant linear correlations, it is possible that breast-milk quantity is noticeably depressed only below some low “threshold” values of maternal characteristics. This possibility was explored with generally negative results, except, perhaps, in Zaire. In Zaire, breast-milk quantity was little affected by the age of the child. However, in order to avoid any confusion in the data arising from the high milk yields usually seen during the first 2-3 months post partum, data in Table 10 are restricted to breast-milk quantities recorded between 3 and 18 months, A closer examination of the data indicates that for serum albumin there may be a threshold value’ below which the quantity of milk produced declines. Table 10 shows that for various higher concen- trations of albumin the maternal milk quantity does not change significantly. But below 30g/l the milk yield suddenly drops, suggesting the presence of a threshold effect. However, this effect was not seen in other countries, probably because concentrations of albumin as low as those seen in Zaire were never encountered. The average breast-milk quantities were much lower in Group C than those in Group B. This was true for all the variables studied mother’s weight, her weight index, and her serum albumin concentra- tion. Also, in both groups, lower values for each variable were associated with lower average breast-milk quantities. Thus, when two groups were plotted together, significant correlations between average breast-milk quantities and the three variables were found. But these correlations appear to be artefacts because when the two groups were plotted separately no such correlations could be seen. This point is emphasized in order to caution. research workers that if the population sample under study is not homogeneous in terms of socioeconomic characteristics, there is a possibility of spurious conclusions.39 4. STUDY OF BREAST-MILK QUANTITY (a) (ee) 7 (ve) (s) (ez) OCLFsee OBLFEI9 = Ob< e1z+se9 az< geLFeoz — 60Z +004 or< (en) (zp) (an) (oe) (1) (ue) ELLF8Ep BBLFOIO «= BEBE = LLL OGE GET F ONO eat vsLtzze © 9sz+es9 69-39 (se) (er) (gz) (rou) (ez) ae) LLL+Z0v © GBL¥ez9 = LE-OE = GHLFISE gst F O19 9e-¥z “siFsee 6614909 9-09 aay (eu) (ea) (eet) 3) (sou) WyLFIse PLLFGI9 © SEE = GLLF ELE 6 BBS eee, istFsee ler FoLo 6s-ss (ge) (ve) (ost) (sz) (sv) (ez) eeitzee olztove eee = pL Q6e © OBL FLAS leer “olFese vel Fels vs-08 (ee) (a9) (wen) (oe) (ri) (se) LLLF6LE eBLFES lef zat FO9e zal + 9E9 6reL eOLF7Se — L6LF Lo erst (zzz) (6) (ga) w (ou) (ze) sO9LF BEE gliz+s9s OE = ezGEF GLE — Oz F 929 aL> OLLF eve —-ZOL-FOvS. oy> gdnaig = g dnoig we) gdnoig = g dnoig gdnoig = g dnoiy (o) unary —— xepur 148!en, — 2UBlom jewarey ‘Aunuenb > ‘Anuend 511 ‘Anuend 3H. (suiuow g4-€) au127 ‘sonsueioereyo jeuie1eu UeAIB 10} (seseynuaied ul) saseo yo Jaquinu pue uoneiAap piepuers yum (6) sonnueNd yIIui-Iseo1q UEEW “OL 142140 THE QUANTITY AND QUALITY OF BREAST MILK A considerable proportion of mothers in Guatemala and the Philippines were taking supplementary proteins, vitamins, minerals, etc. with their normal diet. In Guatemala, 53% of Group A mothers were taking supplements “to increase milk”, 28°% were taking vitamins, and 67% protein supplements. In Groups B and C taken together the corresponding percentages were 46%, 4%, and 48%. In the Philippines Group A, supplements “to increase milk” were taken by 59% of mothers, vitamins by 54%, and protein supplements by 43%. The corresponding percentages in Group B were 64%, 55%, and’ 43%, respectively, and in Group C 31%, 23%, and 14%, respectively. The observed breast-milk quantities were not associated with either the type or presence of supplements taken by the mother. This does not necessarily mean that supplements are generally ineffective since a proper test of their efficacy would require a longitudinal study on the same mother-child pairs. 4.3 Children’s Characteristics and Breast-milk Quantity No relationship was found between breast-milk quantity and parity. At the time of test-weighing, the nutritional status of the children was clinically assessed, and the children were divided into three categories: above average, average, and below average. The groups differed in mean arm circumference in the order implied by the clinical assessment. To compare breast-milk intake of those classified as above average with the intake of those assessed as below average, the intakes were expressed in. terms of standardized deviations from the mean for a given age group of children in a study group in a country ie. {d = (observed volume —mean volume)/standard deviation}. For Hungary, children judged above average for nutritional status yielded a mean standardized deviation (d) of +0.22 (111)! and those judged below average gave —0.35 (13). In the same order, the results for Guatemala were +0.44 (14) and —0.95 (4), for the Philippines + 0.36 (39) and —0.55 (19), for Zaire Group B, +0.02 (137) and = 0.08 (25) and Group C+0.04 (50) and —0.29 (105). However, in Zaire, 30% of children were not classified for nutritional status and for the rest the classification was based on the children’s serum albumin concentration rather than on clinical assessment. In the Swe: series, all children were classified as average for nutritional status. In all other countries; breast-milk intakes of those judged above average were significantly higher than intakes of those judged to be below average for nutritional status. Taking the standard deviation of milk quantity as 150-200 g the difference in daily average intake between the two groups may be estimated as approximately 70-95 g * Figures in parentheses indicate the total number of children.4, STUDY OF BREAST-MILK QUANTITY 41 Table 11 shows the within-study-group correlations between breast- milk quantity and the child’s present weight and the weight gained since birth. It should be remembered that birth weights were mostly reported by the mothers, except in Hungary and in Sweden where this information was obtained from birth certificates; the figures for weight gain may therefore be inaccurate in some cases. It must also be remembered that although all children were breast-fed, milk was not their only source of energy. The number of children fully breast- fed was unfortunately too smail to consider this group separately. For the first 3 months, most correlations of present weight with breast- milk quantity are statistically significant, and so are most of those between weight gain and breast-milk quantity. However, owing to the small number of subjects in the study, statistical significance is less important than the fact that the correlations are of similar order in all five countries. At 6 months and later, the correlations are small and erratic with the exception of Zaire, where there is hardly any change in the degree of correlation up to 15 months of age. These results indicate that at least during the first months of life, there is a correlation between child's growth and milk intake, and that there must be many other factors (such as genetic constitution and infectious diseases) that influence the child's growth. Table 12 shows the mean present weights of children whose diets were supplemented and of those whose diets were not supplemented, classified according to the quantities of breast milk they received. Similar results to those shown in Table 12 were obtained for Hungary and Guatemala, Groups B and C, and for Zaire for weight gained since birth. Children whose diets were supplemented received other foods in addition to breast milk; watery drinks and fruit juice were of Table 11. Correlation coefficients within study groups between breast- milk quantity and child's weight (r,) and between breast-milk quantity and weight gain since birth (r,) Child's age Hungary Sweden Guatemala Philippines Zal Tmonth —r, 0.46 (78)* 0.65 (26) 0.32 (80) 0.44 (87) 0.29 (180) 7, 0.37 (78) 0.58 (26) 0.24 (74) 0.23 (60) 0.22 (128) 2months 1, 0.22 (80) 0.51 (26) 0.28 (28) 0.37 (72) 0.23 (202) 7, 0.20 (80) 0.53 (26) 0.22 (28) 0.21 (38) 0.21 (142) Smonths 1, 0.35 (79) 0.24 (26) 0.39 (77) 0.42 (73) 0.24 (181) nr, 027 (79) 0.04 (26) 0.31 (77) 0.36 (44) 0.36 (95) 4 months 7, 0.08 (78) 0.10 (26) — 0.50 (54) 0.21 (119) ry 0.96 (78) -013 (268) — 0.19 (22) 021 (69) months 1, —0.01 (79) 0.33 (26) 0.01 (61) 0.07 (69) 0.24 (210) ry 0.08 (78) 0.17 (26) 0.02 (45) -0.03 (32) 0.12 (138) “Figures in parenthesas indicate the dogrees of raedomTHE QUANTITY AND QUALITY OF BREAST MILK (enge (ce) be (ze)e9 (ez) 9 (ze) eg (nes (er) ee (Zi) 6 ve nee () v9 oc lies oo er = 800 < a9 (be. @Nz9 (es (irs @) ro (ice ) zy 8669-009 @ ye (nro @) £9 (1) vo (€) ey (8) er (OnOE (5) OF 866r-008 (co) 99 = a) 9s a @ev wee ©) ve Boos > {9 pue g sdnoip) eewaeng, = (eee @) Be —az)ee OD —) 99 ez) 7 wes @ee (es oo Lo (Her 8004 < = @ ce = no. es ro (ner = 8669-005 @) v6 a woe a a () 6 = 8 66y-008 = anes = ) 7% = = = () ee 8008 > uepems a a (Yeo sex ~— (ov) 9g. (SS eB)OP = CL) OF 7 = — (ever (sz)6S = (oL)99 Lew = Boo. < = a @ co ec zs z)9S eo = 8669-005 = (eee () 9g (Loe = (L) OF Se6v-008 i @ ve = = = = 8008 > ‘ebuny, ‘S004 10410 $PO0} 18410 S00} JOO $00} 18410 _SPOO} 19410 $09} 104 SPO} 1240 SOO, Ja40 Buynis001 ON BuInleoey BuIAleDe ION GuInIe9ey —GuieDe1 10N GuInla0ay GuInIee1 ON BuIAIoDey 42 6 9 & L ‘Anuenb > -seeiq pu AnUNOD (sinuow) 96e 5, Mo eAiquenb s1u!-Iseeiq Aq pue (in pue syuup BuIpnioxe) siuauieiddns poo} perisaes Aay 10U 40 eyteym Aq (Bx) 1yBlem wee S,uEIPIIYD “ZL e1eL43 4, STUDY OF BREAST-MILK QUANTITY nes vorpyye jo sequin pur sosoqunied ey sont, @ ro lia)oa (igivs (Dor ees = twee (wz) ve u = @ v9 2 ) zs ae Qiy MW ce WW) 6 So0L < () t2 aoe =e) OL ELOQ =) VS OS (e) GE) Ge 5669 005, ) vo eos =) os (oles =) ok (OL) OE Zu) ve Seep 00€ (ee (ezza Se eziey a) te SHY CE 6) So0e > (9 dnoig) anez (oe (idee De one zayes = (ue eee (ge lw (ee (ee = (9) be (ELL = oz) "9 (8) LG (ve) = (9) So0e< (oe ez)ee =) L9G z)es = (GL)9G = (e)BE —() Ge 8669 003 (oe Se =) ee (BBD Sy «= eS = Gee) ve Fesr oe = a =o = (ee (@) 6v =) be) Ae S008 * {@ dnoig) auez (te (9992 (LOL — (iw) £9 (ov) Sweet iw (ce (6) ees) €£ —) 69 eS =e) v9 (6) OW 8) Boor < (y) be (ez)ee e's) Ga)g9 izes OWS = (L)@E KE) OW 5669-005 GQ) ce (yee = ek a) eS) OS =~) ESL PE AB) Be6r-008 Ors 9) ze = () v9 a () sy (i) Ze) GE Sooe > (9 pue g sdnoip) souiddiyug44 ‘THE QUANTITY AND QUALITY OF BREAST MILK little energy value and therefore not regarded as supplements for the purposes of this tabulation. The “All” rows show that there was no difference between the weights of the babies who received diet supplements and those who did not. These data were also analysed in a different way and it appears that infants not receiving diet supplements had somewhat larger milk intakes. The difference between the mean breast-milk intakes of infants who received supplements and those who did not was estimated for some of the groups shown in Table 12. At 3 months, the differences were 150 g for Hungary, 195 g for Guatemala (Groups B and ©), and 110g for the Philippines (Groups B and C) in favour of infants who did not receive supplements. The corresponding differences at 6 months were 85g for Guatemala (Groups B and C) and 70g for the Philippines (Groups B and C). In Zaire, there were no consistent differences between the two groups of children; if anything, the mean breast-milk intakes of the children who received diet supplements were greater. Table 12 also shows that, whatever the breast-milk intake, infants who received food supplements were not heavier than those who did not. In addition, the weights of the former tended to increase with milk intake, implying, perhaps, that some of the food supplements contained insignificant amounts of energy compared to the ingested breast milk. It should be remembered that: (a) the information on supplementation was qualitative; (b) the 24-hour milk quantity estimates are not necessarily representative of the “usual” intake of milk; and (c) the health histories of the children prior to the investigation were not known. These factors would tend to attenuate possible differences related to supplementation which might in reality exist. The distribution of breast-milk quantities indicates that very few mothers could supply enough milk in order to meet the energy needs of a 6-month-old child without diet supplementation, 4.4 Season and Breast-milk Quantity An effort was made to spread the intake of new subjects more or less evenly throughout the year. There was no indication of any clear seasonal variation in breast-milk quantity except in Zaire. In Zaire Group B, the mean milk quantities in the first 6 months were 566g in November-January, 585 g in February~April, 801 g in May-July, and 568g in August-October; the corresponding mean quantities in Group C were 267, 394, 425, and 313. respectively. In both groups the seasonal effect was ‘significant at P<0.01. Fig. 1 illustrates seasonal variations in Zaire in greater detail and gives 95% confidence limits for quarterly means. These variations are probably a result of the striking changes in food availability at different seasons. Seasonal variation in breast-milk quantity has also been reported from the Gambia, low quantities being recorded during the rainy season (124).4, STUDY OF BREAST-MILK QUANTITY 45 4.5 Maternal Characteristics During Lactation Annex 1, Tables 2-6 give maternal characteristics by children’s age The weight index decreased slightly as lactation progressed in all study groups, except in Zaire. Other maternal characteristics showed no trends with time. It must be remembered that the data are cross- sectional and that any time-related trends are not followed necessarily by a majority of mothers. Note also that there were very large differences in all maternal characteristics between Groups B and C in Zaire (see Annex 1, Table 6); this is due to the different proportion of Bantu and Tutsi women in the two groups and also to the difference in the nutritional status of the two groups. Although corresponding variation in breast-milk quality was observed in Zaire, there was no seasonal variation in maternal weight, weight index, or erythrocyte volume fraction, except that in Group C maternal albumin was clearly higher in the February—April and May-July than in August October and November—January quarters. 4.6 Variations in Children’s Characteristics Within Study Groups A summary of children’s characteristics is presented in Annex 1, Tables 7-11. It is interesting to note that, with the exception of Zaire, there were hardly any differences between study groups in the infant’s present weight, length, or arm circumference. Annex 1, Table 12 gives the percentages of children receiving a given type of food or drink in addition to breast milk. The category “other”, which is very important, requires some elaboration. In Hungary “other” represented vegetables prepared with milk and, usually, cheese. In Sweden “other” was predominantly a pap of two- thirds milk mixed with cereals. In the remaining countries “other” represented a variety of foods with large carbohydrate content: rice, beans, potatoes, noodles, tortillas, biscuits, etc. Thus, in all countries foods classified as “other” represented a significant contribution to the energy content of the diet 4.7 Post-partum Amenorrhoea and Milk Quantity At the time of recruitment, mothers were asked whether or not they had started menstruating since their last delivery. If the answer was positive they were asked about the time interval between the delivery and first menstruation, At the time of test-weighing, mothers were again asked if they were currently menstruating. The statistical analysis shows that children from mothers who were still not menstruating had marginally greater quantities of breast milk, though the difference was not statistically significant. Among mothers46 THE QUANTITY AND QUALITY OF BREAST MILK who had started menstruating, there was no difference in observed breast-milk quantity between those who were bleeding and those who were not at the time of test-weighing 4.8 Food Intake (Other than Breast Milk) During the Study Some information about foods given to the child during the 24 hours of test-weighing was collected in Guatemala and Hungary. Although this is not equivalent to a weighed-diet survey and not necessarily representative of what the child usually received, the data can be used to give an approximation of the energy value of additional foods. The mean energy values of supplementary foods (except fruit and drinks) are given in Table 13, At 3 months, supplementary foods accounted for about 420 kJ (100 keal) and 840 kJ (200 kcal) daily in Hungary and Guatemala (Groups B and C), respectively, and at 6 months supplementary foods accounted for about 798 kJ (190 kcal) in Guatemala (Groups B and C), and 1260 kJ (300 kcal) in Hungary. Table 14 shows that children with lower breast-milk intakes tended to receive more supplements; it is obviously not possible to decide from the data whether the supplementation was causing a lowering of the milk intake or whether the opposite was true, This association was statistically significant in Hungary (P<0.01) but not in Guatemala, where children with breast-milk intakes below 500g also tended to receive more supplements, Diarrhoea (as reported by mothers) in the week preceding the test was not associated with supplementation either in Guatemala or in the Philippines, the only countries for which reports on diarrhoea were sufficiently frequent to warrant analysis. 4.9 Serum Prolactin and Milk Quantity Prolactin was estimated for some Swedish mothers and nearly all Group B and C mothers in the Philippines and in Zaire. Table 15 shows that prolactin concentrations decreased with increasing child’s age; a similar tendency was observed for breast-milk quantity in some, though not all, study groups (Table 7). The estimations for Sweden and Zaire were carried out in a laboratory different from the one in which the estimations for the Philippines were done. Therefore, even though the differences in prolactin concentrations between Sweden and Zaire on the one hand, and the Philippines on the other, are statistically significant, they should be treated » with caution. Comparison of the two sets of means for Zaire shows that the concentrations of prolactin in Group B were higher than those in Group C (P<0.001). The correlations between serum prolactin and breast-milk quantity at given ages of children were very erratic; this47 4. STUDY OF BREAST-MILK QUANTITY aL aw at on z £ ——_-Selqns jo Jequiny (coz¥ zee) (GBLF vor) (9s1Fs0z) (veF6on) = — (Foy = (os 011) 69ST vEEL pELFLézL So0F LOB e97F 8Sy B67 F PLL Oz Z9y 2 dnowg, = i st zt - eh 6 si99Igns jo soquuny a (ziFce1) (ezztssz) = — (een F oz) — (26 601) ZS+S/5 LO 1ZOL (zLF oy8 gee +9z9 @ dno1p oz aL el sioeiqns jo sequin - — er Fzv) (orLFssi) (ori Fev) yssFacs BSF LLL B97 L19 v dnoi sjeworen9 cad ae a +8 uw sp 8 1 sigeigns yo soquiny = = = (orF ive) (aFear) (99F66) (aT IvL) ——(ozz) = = = BOSFOEOL ESET HEL LLZFOLY wre FTES vee Atwuny e z t gi st 6 9 ’ (syruow) 96e s,py4D, dnou8 pue Anunog ‘S{UUP pue iny, UeY JaINO siuawia|ddns BuIAJeoe! UeIPIIYD Aq BuIyBlom-1501 Jo sinoy pz up BuLinp panjeaai spo, Jo (seseyiuased Ul {29) 7-4 UL an|eA ABIPUD Btn Jo SUOHE!Aep piepuels pue sUEEW “EL e1qeL‘THE QUANTITY AND QUALITY OF BREAST MILK 48 ‘® (1e0¥ sez) L641 (3) (leo ze) vat (9) @ ® @ = (y) eox €21) Lig que saapuysosaquaie vent poner OL, (2) (1e94 gov) pLLL squow gL an ( sqnuow $1 (en) syuow & w syuou 9 (Ot) (1224 Lez) 046 syiuow € (g) (1294 891) Zz9 wquow 1 (9 +8 sdnoip) ejewaren5 (e1) (ie0¥ 02) es (ge) (1294 p6z) gez4 (98) (1294 98) Lest syuow 9 (6b) (leo zeL) vs9 (98) (i294 661) 968 (91) (1294 9€2) L66 sywuaw 7 (ol) (ieox €9) saz. (62) (124 001) ozy (9) (194 281) 8¢9 suuow € (2) (eax S24) S75 (y) (eX SLL) €ay (2) (199 012) 788, syiuow Z = o(1)- (W204 022).926 yquow | ‘AoBunH, ‘a1 10 BO0L 8669-008 96e s,uaipiyo pue yoru syrus-1se03g 6003 sepun on = ‘dnoi6 ‘Anunog ayeqU! ¥I:UN-Isea1q 01 BuIpiod9e pay! 219 vaipliyo Aq U9ye) SxUUP pu IIN4y UeY J8YI0 SiUAWIe;ddns jo (Fy) Sonjen ABOU LEEW “PL 9192L49 4, STUDY OF BREAST-MILK QUANTITY ioe ~ wn oss ure Ag a4 ow pavanvos eam osu pur /syun ov pena} AyBuBHO SION sma '8HE7 UE YOpEM 04 "UOIPIYDJO equ 24812} vesm\RUEN LB! Mg ord go> s00 6987 (82) cos + seee (€8) Lev Lose & szoz uve (2) 99 1glz (Ze) 6Lv= S917 (ve) 699+ 9967 (2) Lue F o9ez (iz) Loe> wre (ze) 062 ¥ s9ez (gz) 982 + g608 a(€8) ble F 997 t (synuows) AroAujop saye ow, 9 dnoig a dnog 2ulez 9 dnoig a dnog vy dnos5, sauldaiiyg 9 dnoig @ dnog v dnoig ejewareng, uapamg, Are6unH dnoiB pue Anunoy jw 1SBaIq JO Sanyea ABiouB UeAW “LL SIGeLTHE QUANTITY AND QUALITY OF BREAST MILK 6) 0707 5) 230328) cor 0s, foci ca ete Ophetesiog\Ve 0a x99 oz . 2 | | a loc tee oo Jor jos | 6 uuew 9 aiuow : 5 oz t oe t ov 08 © unvow . 1 nuow oo 96e s,pilyo pue dnoiB Apmis Aq “yiIu 1seeiq u! suonenuesuOD (V/8) esoroe| ueaw rsulebe (1/6) rey ueow jo yde1B seneos y “E “B145. COMPOSITION AND QUALITY 61 5.7 Pesticides in Breast Milk Table 18 shows the concentrations of various pesticides found in breast milk in $ countries. The breast milk of the Hungarian and Guatemalan women had higher pesticide concentrations than those found for the Swedish, Filipino, and Zairian mothers. In all countries except Sweden DDE was present in the highest concentration; in Sweden the concentration of polychlorobiphenyls were higher. In general, the pesticide with the second highest concentration was DDT. In Guatemala and the Philippines, the comparison of study groups showed some statistically significant differences. In Guatemala, 2,4-DDT concentrations were higher in Group B than in Groups A and C (P<0.01). For DDE, the rural Group C had lower (P< 0.01) concentrations than either Group A or Group B in both countries. For DDT, Group B (urban-poor) had significantly higher concentra- tions (P<0.01) than Groups A and C in Guatemala, while in the Philippines Group A had a higher concentration (P<0.05) than Groups B and C. For beta-HCH and dieldrin, Group C in Guatemala had a lower concentration than Groups A and B (P<0.01). It appears that in many instances the concentration of pesticides in the breast milk of urban mothers was higher than that found for rural mothers (Table 19). The reasons for such a difference are clear as far as Guatemala is concerned. The rural mothers were from an area in the highlands where no pesticides were being used and where people mostly consumed locally produced foods. On the other hand, the urban mothers consumed foods originating from the lowlands, where pesticides were being widely used for the production of crops such as cotton and coffee In Hungary, the concentrations of several pesticides in breast milk at 1 month were significantly higher (P<0.01) than those found at 6months. The concentrations of the different pesticides at 1 and 6 months, respectively, in each case were: DDT—335 y1g/l and 189 nei; HCH—6yg1 and 4yg); and beta-HCH—18ygil and 7yg/l. This trend was not noticeable in other countries, but usually the number of samples at different stages of lactation was very small. It may be useful to put matters in perspective. Taking DDE as an example, it is seen that a breast-fed child is likely to absorb during 6 months 29mg of the pesticide in Hungary, 14mg in Guatemala, 4 mg in the Philippines, 4 mg in Sweden, and 3 mg in Zaire. All of the compounds considered in this study are difficult to excrete and tend to accumulate, Whether the “doses” quoted above are detrimental to the child’s progress in terms of physical or possibly mental development is not known, But any outcry about contaminated breast milk is not justified because the problem is not one of breast-feeding. Any child taken off breast-feeding will continue to ingest the pesticides in his daily diet just as does his mother who now passes some of it to her° ‘nonlne o sequmu steps sosaqionied of SNBLL soidwies oqnueaiod uunipoyy 40 0N «ww onez 06 ° ° 8 8 moo ° ° 80H ° ° ¥6 ° ° ° ° ° 0 situeyaig -o10}4oAiod ° ei ° ° ° z ° oe ° ° ° ° ° ° ° ‘ ° ° ° ° ° ¢ z zs ° ° L ° e ° ° ° ° s a SOLS HOH BW ° ° 8 ° oie ° 6% HOH-muwed ° ° ° ° ° e ° fz HOH-udie et a ss 8 wiLS kk we aq ° ° ° ° ° z 1 er aaa 9 for woe ez LOB 30a b ° ° ° ° L z oe Lane, ° ° ° ° ° z ° of 300-97 ones egenseous aa sed eoied sojguies ojnusoied sojaues uipoy Wis UEIPOW JOON MOG «UOIPEW 0-ON OG UEIPEWN 40°ON, - — ES eprnseg (e) i) ow e(i8) (ox/6") opoms seb eyewsieng soursdid (V/6r/) ui rseaig u! suonenusoucs apionseg “gL 91qe,5, COMPOSITION AND QUALITY 63 Table 19. Median concentrations of selected pesticides in breast milk, Guatemala and Philippines by study group Guatemala (149/!) Philippines (j.a/ka) Pesticide ~ a SESE aa Group A Group B GroupC GroupA Group B Group C (ane 8) (28) (1) (28) (21) 2.4-DOT 15 43 22 24 42 13 DDE 192 175 18 96 32 2 oT 23 49 20 19 12 "1 Beta-HCH® 24 15 5 = = = Dieldrin 32 23 13 34 ° 0 * Figures in 9 indicate numbers of subeets 2 Inthe cate ofthe Philippines. only four semples wete found tobe contaminated. child in the breast milk. There are only two ways of combating the problem: (a) by developing compounds that are easily excreted; (b) by identifying the more harmful compounds and strictly controlling their use. 5.8 Quantity and Quality of Breast Milk Although the breast-milk samples for chemical analyses were collected 3-4 hours after the test-weighing procedure was completed, their composition will be regarded here as representative of the whole 24-hour period. 5.9 Breast-milk Quantity and Total Nitrogen, Fat, and Lactose Concentrations Table 20 shows average correlations of breast-milk quantity with total nitrogen, fat, and lactose for all children of 1-9 months of age within each study group and for all children at 1, 3, 6, and 9 months of age in all study groups. It was not possible to obtain correlations for each age group within each study group because of the small number of subjects in the study groups; therefore average correlations within study groups or age groups were calculated. It was observed that the total nitrogen concentration tended to decrease as the 24-hour breast-milk quantity increased. The average correlations between milk quantity and fat within the study groups were not statistically significant and were erratic with respect to sign. The correlations between milk quantity and lactose indicated that lactose tended to increase with quantity; these correlations were mostly statistically significant. Details of volume and chemical analysis of seventh-day breast milk of 150 women were published by Hytten (5/). From these data,THE QUANTITY AND QUALITY OF BREAST MILK s90>d 9 1054 q ‘wopaey 0 snidep ey rerpu semiwuand vain eu. alert) 8z0 (eri) ov0— lyon) veO— syuow 6 2(922) LV0 (vez) L00 alvez) e€°0— sujuow 9 (492) Leo (192) toro- (292) SE°0— sujuow € alvlz) Sz0 (zz) Lo0- (S22) Le0- iuow | (Sdnou6 Apmis jje ut saBe 1uareyp re UsupiEYD [Ie 404) (est) s0°0 (eau) 910 a(cau) 6zo~ 9 dno (zat) 600 2(Z91) 8L0— a(Z91) 970~ @ dno @(601) L€°0 (801) 910- (601) L70— yianoip Souldditt4 (001) 94.0 (001) 90°0- (00}) Loo- gdnoip a(LOl) 6770 (304) 20°0— a(€01) 6z'0— a dnop, a(v@) 960 (v8) Zo0- (v8) 8¥0- ‘v dnoig ejewarenp, 2(ZOL) 170 (zou) ovo a(Zou) ev'0- uepems, o(€¥2) LO (eve) to alove) St'0— AvebunH {dnoub Apms yoee uiyum aBe Jo syauoU! G-| JO eIpIIYO IIe 104) es0122] pue Anueng, 22} pue Arnueng ebony je101 pue AUEAD ~ —— - — dno16 pue Anunog jo2miog suone}ou03 eSdnouB Apmis 12 u! eBe 4o syluow 6 pUE ’g ’E ‘| 2e VOpIIYO {Ie 10} pue dno1B APMIS Yoee UIYIA 982 Jo syiuOU G-1 Jo UAIPIIYO ITe 403 aso10e] pue ‘rey ‘UaBOIIIU je101 yaIM AiUEND ¥IILI-1S2019 Jo SUONE|e109 Beery “OZ 914215, COMPOSITION AND QUALITY 65 correlations were calculated between milk volume and total nitrogen, fat, and lactose and were found to be —0.43, 0.02, and 0.45, respectively. This pattern of correlations is the same as that found for average correlations in this study. 5.10 Energy Value of Ingested Breast Milk Table 21 gives estimated energy values (kJ/24 hours) of ingested breast milk. The energy value of the breast milk of Swedish mothers was the greatest. The lowest values were seen in Philippines Group A and in Zaire Group C. In the case of Philippines Group A, one should note that artificial feeding was started very early, and mothers had little motivation to breast-feed The estimations given in Table 21 are, however, subject to large errors. The first source of error concerns the estimates provided by Guatemala and Hungary of the additional foods given to the children during the 24 hours in which the test-weighings were done; the energy Table 21. Mean energy values (kJ/24h) of ingested breast milk* Child's age (months) Country and group a 1 3 6 9 Guatemala Group A 12182605 1516 +655 - = (32) (28) Group 8 12222458 13442475 14204508 1386 +424 (27) (30) (28) (24) Group C 12684336 1671+403 1369340 1390+475 (27) (28) (28) (27) Hungary 1508+365 16464445 1423-4437 (@3) (83) (78) = Philippines Group A 7774407 8194575 8314403 844 +529 (34) (21) (16) (18) Group 8 13154462 16634455 13944588 12944416 (32) (31) (30) (i) Group C 15624638 1528+605 14284496 1378 +609 (27) (28) (28) (32) ‘Sweden 1936 +655 2528+823 18064672 1846 +1000 (28) (28) (28) (26) Zeiree Group B 1457 1373 1600 1558 Group C 861 16 879 958 “To conve these nergy values nto kes 24 divide by 42. ‘m parenteeee indicate numbers of subjects © timated trom poole sample the abimated sand B and C,rxpectivety ddaviatons ware 192/24 h and 169/24 for Groups66 THE QUANTITY AND QUALITY OF BREAST MILK values of different foods had to be calculated and subtracted from the child’s total energy intake. Also, the dietary information provided by Guatemala and Hungary was not necessarily representative of the foods the child normally received at home. Secondly, the calculation of the energy content of breast milk was based on a sample of breast milk taken 3-4 hours after the test-weighing, and was assumed to be representative of the milk ingested by the child during the whole of a period of 24 hours. Finally, the total volume of milk recorded during the test-weighing was taken to represent the child’s usual intake of Fig. 4. Mean body weights and energy intakes of children by age and country 2520). £ £2100) 1680] HUNGARY GUATEMALA, Group & = GUATEMALA, Group 8 GUATEMALA, Group C >— 6 ‘Age (months)5. COMPOSITION AND QUALITY 67 breast milk. Nevertheless, it was thought worth while to gather the various evidence together and form a picture, however inaccurate, of the energy content of breast milk of women in different countries and in different socioeconomic groups. Fig. 4 shows the mean weight of children at given ages, mean energy intake per 24hours (breast milk + supplements), and mean energy intake per kg of body weight. It must be emphasized that the data are cross-sectional and the means are joined by straight lines only to facilitate the recognition of points belonging to a given study group. The weight of the Guatemalan children started diverging from that of the Hungarian children somewhere between 3 and 6 months of age. At the same time, the total energy intake (kJ/24 hours) became considerably different, and so did the daily intake per kg of body weight. ‘A number of implications follow from Fig. 4. First. the Hungarian children gained 4 kg in weight from birth to 6 months (which was as much as the Swedish children gained) with an average daily energy intake of, say, 358 kJ (85 kcal) per kg of body weight. Assuming that the energy value of breast milk is 2604kJ/I (620kcal) (which is approximately the Hungarian figure), this would imply an intake of about 1 litre of breast milk daily. Since, as was seen before, only a very small proportion of the Hungarian children achieved such intakes, supplementation is clearly required somewhere between 3 and 6 months of age. Finally, Fig. 4 suggests that the usual observation that the weight curves of children from developing countries flatten out and start diverging from the weight curves of children in developed countries at about 6 months of age, may be explained by inadequate diet supplementation and repeated infections rather than by an inadequate quantity of breast milk, since as can be seen from Table 21 the average energy ingested per 24 hours in breast milk at 6 months differed little between countries, if one excludes Sweden, Group A of the Philippines, and Group C of Zaire. It has already been indicated that the two latter groups represent special situations. 5.11 Energy Content of Breast Milk and Maternal and Child Characteristics In general, the maternal and child characteristics do not appear to be correlated with either the major constituents (protein, lactose, and fat) of breast milk or the milk’s energy content. In Zaire, the maternal and child characteristics could not be correlated with the milk’s energy content because the milk was pooled for analysis; this precluded the investigation of the possible threshold effect seen in the correlations of milk quantity and maternal characteristics The correlations between total energy derived from breast milk and child's weight or weight gain from birth were positive, and were very68 ‘THE QUANTITY AND QUALITY OF BREAST MILK similar to the correlations between the latter two variables and breast- milk quantity; therefore, these correlations will not be presented separately. The implication of this finding is that information on breast-milk quantity, which is comparatively simple to obtain, is in practice as valuable as the much more difficult analysis of its composition.6. Summary and conclusions Study groups in Guatemala, the Philippines, and Zaire were deliberately selected to represent groups of mothers with different traditions, ways of life, socioeconomic status and, it was hoped, nutritional background. Hungary and Sweden represent two countries in which prolonged breast feeding is not frequent but in which the mothers currently enjoy relatively high economic and nutritional standards. It is worth examining whether these selection criteria have indeed produced real differences in maternal characteristics. Differences in weight and height between groups were large but difficult to interpret because of ethnic differences. There were no real differences between study groups in weight-for-height (weight index), except in Zaire, where the rural group had clearly lower weight-for-height than the urban group. This could be due to a real difference in the nutritional status of the two groups and/or to ethnic differences in body frame, the proportion of Bantu and Tutsi women being different in the two groups. Women in the rural group were much taller (157.1em) and lighter (50.8kg) than women in the urban group (153.5cm and 55.5kg). There were also differences in maternal arm circumference. In Guatemala, Groups A and B had significantly (P<0.01) greater values than the rural Group C; in the Philippines, Group A had significantly greater values (P<0.01) than Group C; and in Zaire, the urban group had larger mean arm circumference than the rural Group C (P<0.01). Maternal serum albumin was significantly (P<0.01) higher in the urban than in the rural groups in Guatemala, the Philippines, and Zaire. In Zaire, in Groups B and C, the mothers had significantly lower albumin values than groups from any other country. There was a great overlap, from the nutritional point of view, between the urban and rural groups in the three developing countries covered by the study. However, the above comparison indicates that the group selection resulted in some group differences in nutritional status, the rural groups being the most disadvantaged. This does not imply that the difference between urban and rural areas always in the same direction, because there are no doubt situations in which the urban immigrants are worse off than the people in rural areas. 6970 THE QUANTITY AND QUALITY OF BREAST MILK The maternal characteristics changed very little with the stage of lactation. However, as has repeatedly been stressed, it must be remembered that the data are cross-sectional. The weight index decreased in most study groups as would be expected, since the production of breast milk uses up maternal fat stores accumulated during pregnancy. In Zaire, however, the weight index did not fall with the stage of lactation, which may imply that maternal fat stores were low and breast-milk production depended on current food consumption. Other maternal characteristics did not change. The quantity of breast milk recorded for any mother-child pair depends on many factors, including the mother’s ability to supply milk and the feeding routine adopted (the two may be related), the child's demand, health, and the availability of supplementary foods. In the first 3-4 months of lactation the mean volumes recorded for Swedish mothers were significantly greater than those for any other study group, but in later months the comparisons between study groups were confused, possibly by changes in feeding patterns and increasing use of food supplementation. The lowest volumes were recorded in the Philippines Group A (well-to-do) and Zaire Group C (rural). The low mean volumes in the Philippines may, possibly, be attributed to the very extensive use of breast-milk substitutes, The low mean values in rural Zaire were possibly due to the generally poor state of nutrition of the population, In all study groups breast-milk volume was found to increase with the number of feeds. This may be interpreted as evidence that more frequent stimulation leads to increased secretion. However, an alternative explanation that mothers who have less milk feed less frequently is equally plausible, and is supported by the quite large volume differences between the intakes of babies with different numbers of feeds, seen, for example, in the urban-poor and the rural groups in Guatemala, the Philippines, and Zaire, where about 90°% of mothers breast-feed on demand. It is possible that both interpretations are correct depending on the circumstances. The large number of feeds per day reported by the mothers, and actually observed during test- weighing, suggests that for the urban-poor and rural groups the observed mean breast-milk volumes may be at or near the maximum the mothers could supply. The differences in the flow of milk, as measured by the quantity sucked by the child in 15 minutes, between Hungary and Sweden as compared with Guatemala, the Philippines, and Zaire Group B, may mean simply that a child put to the breast less frequently is more eager to suck. But the low results for Zaire Group C may indicate a genuine deficiency in milk production. Seasonal variation in breast-milk volume was noted in Zaire only; in the region studied, the seasonal pattern of food availability coincides with that observed for breast-milk quantity. The quantity of breast milk was not correlated with any of the maternal characteristics, and the search for a threshold value below6. SUMMARY AND CONCLUSIONS a which the flow of milk could be regarded as impaired was unsuccessful except in Zaire, where mothers with very low serum albumin had significantly lower breast-milk volumes than those with higher albumin values. However, in other study groups, there were hardly any mothers with comparable serum albumin concentrations; perhaps this made the search for a threshold effect ineffective. ‘A considerable proportion of mothers in Guatemala and the Philippines were taking diet supplements. The breast-milk volumes recorded for mothers taking diet supplements and for those who did not take them did not differ. This result does not mean that the supplements had no effect on breast-milk supply because the data are cross-sectional; a proper test of the efficacy of a given supplement requires a longitudinal study over a sufficiently long period for the supplement to show its effect. There was little relation between the milk quantity and the recommencement of menstruation, though mothers who had not yet started menstruating again tended to have marginally larger quan- tities. Serum prolactin concentrations were estimated in Sweden, the Philippines, and Zaire, and the concentrations in Zaire were found to remain high for as long as 24 months. The average correlations between prolactin concentrations and breast-milk volumes were low and statistically not significant. The results suggest that while elevated prolactin concentrations are necessary for the maintenance of lactation, the actual concentration in an individual mother bears little, if any, relation to her milk output. In all countries (except Sweden, where all children were assessed as average), the children assessed as above average for nutritional status had breast-milk intakes significantly greater than the volumes recorded for children of below-average nutritional status. The difference in breast-milk intake between these two groups was estimated as 70-95g of breast milk per day. This should not be interpreted as meaning that the difference in nutritional status is due to the larger quantities of breast milk ingested by those with above- average nutritional status, since many of the children investigated were also receiving food supplements, Although it has been shown for Guatemala and Hungary that children receiving more milk tend to receive less food supplemen- tation, it is not known whether this relationship is true for other study groups. During the first 3 months of lactation, there was a statistically significant correlation between breast-milk volume and child’s weight or the weight gained since birth. At 6 months and later, the correlations became erratic and very low, presumably as a result of the increasing importance of supplementary foods. The only exception was in Zaire, where correlations of the same order were observed from the Ist to the 15th month of lactation, in spite of the fact that nearly all children were given supplementary foods after the age of 6 months.2 ‘THE QUANTITY AND QUALITY OF BREAST MILK Up to the age of 6 months, supplements other than drinks were given least frequently in Sweden and in Guatemala Group C (rural). It must be made clear that food supplementation does not imply the use of breast-milk substitutes. In the rural group in Zaire, 46% of the children received food supplements (gruel of bananas or manioc) in the first months of life, whereas breast-milk substitutes were given to fewer than 2% of children of any age. In fact, except for the Group A (well-to-do) children in Guatemala and the Philippines, no more than a quarter of children receiving supplements were given breast-milk substitutes. The comparison of weights of children receiving food supplements and those not receiving supplements showed practically no difference between the two groups at any age. There was also no difference in weight between the two groups when the children were classified according to breast-milk intake. The results of the semi-quantitative dietary survey of foods given during the test-weighing session help to explain this apparent paradox. In Guatemala and Hungary, among the children receiving supplements, those who had high intakes of breast milk received less food as supplements than those with low breast-milk intakes; for example, in the two countries, at 6 months of age, children with a breast-milk intake of 700g or more received 670kI (160kcal)/day in the form of supplements which was S45kJ (130kcal)/day less than that received by children with intakes below 500 g of breast milk. In addition, it was found that with the exception of Zaire, the average breast-milk volume taken by children not receiving food supplements was greater than that taken by those receiving supplements. Chemical analyses showed that, in general, the concentration of nitrogen in breast milk was highest at 1 month of age, fell up to 3-6 months, and fluctuated with no particular pattern at later ages. Fat and lactose concentrations showed no clear relation to the stage of lactation (from 1 month onwards). The correlations between the concentrations of these three constituents were low but preponder- antly negative between lactose and fat, and between lactose and total nitrogen; the correlations were mostly positive between fat and total nitrogen. For protein nitrogen, Group C (rural) showed lowest concentrations in Guatemala and the Philippines, but in Zaire the rural group had slightly higher concentrations than Group B (urban). For fat and lactose there were no differences between study groups in Guatemala and the Philippines, but Zaire Group C (rural) had higher concentrations of fat and lower concentrations of lactose than Group B. There were some notable differences between countries, Though interlaboratory differences make the comparisons difficult, some of the differences are too large to discount. The high nitrogen concentrations in Zaire (Group C) and the low concentrations in the Philippines (Groups B and C) stand out among other results. The low fat and high lactose concentrations in Guatemala deserve further investigation.6. SUMMARY AND CONCLUSIONS B This fat-lactose relation is of importance in the explanation of the aforementioned “compensation” effect. The lactalbumin and lactoferrin concentrations in the breast milk of Swedish mothers were highest at 1 month and then fell to a fairly stable level from 3 to 9 months of age. The vitamin C concentrations estimated in Hungary and Sweden were 50-60 mg/l. The concen- trations of vitamin A tended to decrease with the increase in the child’s age, and the rural groups in Guatemala and the Philippines had the lowest concentrations. The presence of various pesticides in breast milk was examined in Guatemala, Hungary, the Philippines, and Sweden, and large differences were observed between countries. It is interesting that the concentration of pesticides in breast milk of urban mothers tended to be higher than that observed in the milk of rural mothers. Whether the concentrations of pesticides found in breast milk are detrimental to the child’s development is not known. The problem, however, is not specific to breast-feeding but to dietary intake in general. A child taken off breast milk will still continue to ingest pesticides with the ordinary family diet, if the latter is contaminated. The correlations between the quantity of breast milk ingested during the test-weighing session and the concentrations of the main breast- milk constituents in the sample taken 3-4 hours later are of interest, though why they occur is not clear. Breast-milk quantity and fat showed little correlation, but volume and nitrogen and breast-milk quantity and lactose showed negative and positive correlations, respectively. Similar results may be deduced from the data of Hytten (S51) for 7th-day breast milk. When the energy content of the principal breast-milk constituents was calculated it was found that the milk of the Swedish mothers had the highest energy content; thus Swedish babies had the advantage of not only the large quantity of breast milk noted above, but also milk of high energy content. There were only small differences between all other countries or groups in this respect. Of particular interest was the lack of difference between Groups A, B, and C in Guatemala and the Philippines where the standard of living and the economic and nutritional background of Group A mothers were undoubtedly superior to those of mothers in Groups B and C. Comparing the mean concentrations of fat and lactose in the study groups it was found that the groups with high fat concentrations tended to have low concentrations of lactose and vice versa. This “compensation” effect leading to comparative equality of breast-milk energy content was also noted by Bailey (8). In terms of the proportion of total energy content of breast milk attributable to fat, the countries can be arranged into three groups: Sweden and Hungary (62% and 59%), the Philippines and Zaire (54% and 52%), and Guatemala (37%). There were no correlations between the mothers’ or the children’s characteristics and the concentrations of the main breast-milk14 THE QUANTITY AND QUALITY OF BREAST MILK constituents or its energy content. Present weight and the weight gained since birth were correlated with the total energy content derived from breast milk, but the correlations were no better than those obtained with breast-milk quantity. These findings are also in agreement with those of Bailey (8) For two countries (Guatemala and Hungary) it was possible to estimate the total energy intake per day (derived from breast milk and supplementary foods, if any). From these somewhat hypothetical calculations it appears that only a small proportion of mothers could satisfy the needs of a 7-kg baby. This implies the advisability of supplementation from about 4-6 months of age, on average. Inadequate food supplementation and the load of infectious diseases rather than lack or “poor quality” of milk may be responsible for the usual downward divergence of the growth curves of babies from the developing countries from those of babies from the so-called developed countries. The arguments in favour of encouraging breast-feeding are undoubtedly sound. Breast milk is the most economical food for young babies, and it also has definite immunological advantages. The advocates of prolonged breast-feeding, however, seem to under- estimate the fact recognized by all mothers that sooner rather than later the maternal supply of breast milk ceases to satisfy the growing nutritional needs of a healthy child and supplementation becomes unavoidable, However, supplementation with breast-milk substitutes has its own problems. There are always the dangers of infection resulting from poor hygiene, over-dilution of feeds resulting from economic pressures, and poor education of mothers. These fears, though real, may be somewhat misleading. For example, apart from the well-to-do groups of mothers in Guatemala and the Philippines, less than 25% of all babies receiving food supplements were receiving breast-milk substitutes in any country at any age. Furthermore, it is worth remembering that the “‘infection” argument applies as much to other food supplements as it does to breast-milk substitutes (98). It seems, therefore, that in addition to the maintenance and promotion of breast-feeding, some resources should be devoted to teaching mothers when and how to supplement the diet of breast-fed babies with foods prepared from locally available products. The demonstration that babies fed in this way can grow well and healthy would constitute an effective and economically sound contribution to the control of malnutrition in young children.26, 21, 28, 29, 30, References AppEL KADER, M. M. ET AL. Journal of biosocial science, 4: 403 (1972). AN, C. HL, & MACLEAN, W. C. American journal of clinical nutrition, 33: 183 (1980), AMINZakl, L. eT AL Journal of pediatrics, 88: 91 (1974), ARENA, J. M, Nutrition today, 5 (4): 2 (1970) | ARROYAVE, G. ET AL Evaluation of sugar fortification with vitamin A at the national level, Washington, DC, Pan American Health Organization, 1979 (PAHO Scientific Publication No. 384). AtHavate, V. B. Indian journal of child health, 9: 175 (1960), Bay P. VA. er at. Tropical and geographical medicine, 32: 158 (1980) Baitey, K. V. Journal of tropical pediatries, 11: 35 (1965). Bares, C.J. EF aL Transactions of the Royal Society of Tropical Medicine and Hygiene, 76: 253 (1982) BrLavapy, B. Indian jownal of medical research, 47: 217 (1959) BELAVADY, B. Acta paediatrica scandinavica, 67; 566 (1978). BeLavay, B. In: H. Aebi & R. Whitehead, ed. Maternal nutrition during pregnancy and lactation. Bern, Verlag Hans Huber, 1980. Briavapy, B. & Goratax, C. Indian journal of medical research, 47: 234 (1959). Brravapy, B. & Goratan, C. Indian journal of medical research, $0: 518 (1960). BoEDIMAN, D. et AL. Journal of tropical pediatrics and environmental child health 2544): 107 (1979). Bunner, D. L. er AL American journal of obstetrics and gynecology, 1 (1978) Cuanpna, R. K. Nutrition research, 11 25 (1981). GuAvez, A. & Martinez, C. In: H, Aebi & R. Whitehead, ed. Maternal nutrition diuring pregnancy and lactation, Bern, Verlag Hans Huber, 1980, p. 274 Cuavez, A. Er AL In: Proceedings of the Ninth International Congress of Nutrition, Mexico, 1972, Vol. 2, Basle, Karger, 1975, p. 9. CustLean TopiNe EDucaTiONal BUREAU. ladine content of foods. London, Stone House, 1952, Contemporary patterns of breast-feeding: report on the WHO Collaborative Study con Breastfeeding. Geneva, World Health Organization, 1981 Cowarn, W. A. ETAL Lancet, 2: 13 (1979) Dem, H. E. Archives of disease in childhood, 18: 93 (1931). DeFiuient, J. P. er at. dcia paediatrica seandinavica, 70: 417 (198)). Denis W. & Tatnot, FB. American journal of diseases of children, 18: 93 250 919), Deopiar, A. D. & RaMAakRisHNAN, C. V. Indian journal of medical research, 47: 352 (1959), Deopuar, A. D. & RAMAKRISHNA, C. V. Journal of tropical pediatrics, 6: 44 (1960), DRUMMOND, J.C. Er AL. British medical journal, 2: 757 (1939). Ducpate, A. E. British journal of nutrition, 26: 423 (1971), Epoziex, J.C. et AL. Journal of nutrition, 106: 312 (1976) Exsmyr, R.A trial 10 change infant feeding practices in an Ethiopian village. Addis Ababa, Ethiopian Nutrition Institute, 1969 1516 THE QUANTITY AND QUALITY OF BREAST MILK 32. Fomon, S. J. Infant nutrition, 2nd ed. Philadelphia, W.B. Saunders, 1974 33. FowLer M, Journal of tropical pediatrics and environmental child health, 22: 34 (1976). 34, Forsum, E. & LONNERDAL B. American journal of clinical nutrition, 33: 1809 (1980), 35. GatsteR, W. A. Environmental health perspectives, 1S: 135 (1976). 36. Gerster, C. er AL. American journal of clinical nutrition, 31: 160 (1978), 37. GHELBERG, N. W. ET AL Igiena (Bucharest), 21: 17 (1972) 38. Gnosn, 8. Lancer, 1: 281 (1981), 39. Goratan, C. Journal of tropical pediatrics, 4: 87 (1958). 49. Goratan, C. Indian journal of medical research, 46: 317 (1958). 41. Gross, S.J. er AL Journal of pediatrics, 96: 641 (1980). 42. Hapicut, J. P. er at. In: Proceedings of the Ninth International Congress of Nutrition,” Mexico, 1972 Vol. 2, Basel, Karger, 1975, p. 106. 43. HaGvann, S. B. et AL. Bulletin of environmental contamination and toxicology. 9: 169 (1973) 44, HALL, B. Lancet, 1: 779 (1975). 45. Hawpracus, L. In: H. Acbi & R. G. Whitehead, ed. Maternal nutrition during pregnancy and lactation. Bern, Verlag Hans Huber, 1980, p. 233. 46. Hampracus, L. er aL Acta’paediatrica scandinavica, 67: 561 (1978). 47. HaNAFY, M._M. ET AL. Journal of tropical pediatrics and environmental child health, 18: 187 (1972). 48. Henwanr, P. er At. Clinical endocrinology, 14: 349 (1981), 49. How, T. V. Er AL. Journal of medical engineering and technology, 3: 66 (1979). 50. Howie, P. W. EYAL British medical journal, 283: 757 (1981). 51. Hytten, F. E. British medical journal, 1: 249 (1954), 52. HytteN, F. E, Proceedings of the Nutrition Society, 1S: vi (1956). 53. INsutt, W. ET AL Journal of clinical investigation, 38: 443 (1959) 54. IyENGAR, G. V. Elemental composition of human and animal milk. Vienna, International Atomic Energy Agency, 1982 (IAEA-TECDOC-269), 55. Faso, E. Revista espanola de pediatria, 5: 364 (1949). 56. Jecuirre, D. B. & Je.uirre, E. F. P. Human milk in the modem world, Oxford, ‘Oxford University Press, 1978. 57. Juszkiewicz, T. ET AL. Polski (ygodnik lekarski, 30: 365 (1975). 58. KaRMARKAR, M. G. ET aL. Acta paediatrica, $2: 473 (1963). 59. Kaucuer, M. Et AL American journal of diseases of children, 7 60. KHAN, M. American journal of clinical nutrition, 33: 2356 (1980). 61, Kaw Mauno Naina, Et AL. American journal of clinical mutrition, 33: 2665 (1980), 62. Kiumance, K. J. Journal of tropical pediatrics and environmental child health 18:313 (1972) 63. KIRSCHGESSNER, M. Schriftenreihe tiber Mangetkrankheiten, 6: 49 (1956). 64 Knowies, J. A. Clinical toxicology, 7: 69 (1974) 65. Kon, S. K. & Mawson, E. H. British Medical Research Council Special Report No. 269 (1950), 66, Kuusk1, J. K. & HartMann, P. E, Australian journal of experimental biology and ‘medical science, 59: 405 (1981). 67. Lan, E. P. ETAL Archives of industrial hygiene and occupational medicine, 3: 245 cagsp. 68. Launer, E. & REINHARDT, M. American journal of clinical nutrition, 32: 1159 (1979), 69. Lepsutein, A. K. & EL BaHay, A. M. Journal of the Egyptian Public Health Association, 51: 246 (1976) 70. LixppLap, B. S. & RawIMTOOLA, R. J. Acta paediatrica scandinavica, 63: 125 (i974), 71. Liu, 8. H. er AL: Journal of clinical investigation, 19: 327 (1940). 72. Lompeck, I. Er aL. European journal of pediatrics, 125: 81 (1977). 42 (1945).REFERENCES oH 73, LoMBECK, I. eT AL. European journal of pediatrics, 129: 139 (1978. 74, LONNERDAL B. ET AL American journal of clinical nutrition, 29: 1134 (1976) 75. LONNERDAL, B. Et AL American fournal of clinical nutrition, 33: 816 (1980). 16. Lucas, A. Er AL. Early human development, 4: 365 (1980). 77. Lurz, P. & Part, B. S. Proceedings of the Nutrition Society, 17: iii (1958). 78. Macy, I. G. er aL American journal of diseases of children, 39: 1186 (1930). 79. Macy, IG. Er aL American journal of diseases of children, 70: 135 (1985) 80. Macy, LG. ex AL The composition of milks. A compilation of the comparative composition and properties of human, cow and goat milk colostrum and transitional ‘ilk. Washington, DC, National Academy of Science, National Research Council, 1953 81. Macy, L. G. & Keutey, H. J. In: S. K. Kon & A. T. Cowie, ed. Milk: the ‘mammary gland and its secretion, Vol. Il, New York, Academic Press, 1961, p. 26S. 82. Manrinez, C. & CHAvE2, A. Nutrition reports international, 4: 139 (1971) 83. Muaray, G. K. & REA, U. Journal of dairy science, $4: 1001 (1911). 84. Neuwisten S. Schweizerische medizinische Wochenschrift, 18: 396 (1952), Abstracted in: American journal of diseases of children, 8§: 337 (1953), 85. OiszyNaMarzys, A. E. er aL Boletin de la Oficina Sanitaria Panamericana, 74:93 (1973) 86. OtszyNaManzys, A. E, deta paediatrica scandinavica, 67: $71 (1978). 87. PasRica, S, Indian journal of nutrition and dietetics, 10: 282 (1973). 88. PicctaNo, M. F. Nutrition reports international, 18: 5 (1978), 89. PorpLau, W. er at In: H. Krant & H. D. Cremer, ed. Investigations into health and nutrition in east Africa. Munich, Weltforum Verlag, 1968. 90. PRENTICE, A.M. In: H. Aebi & R.'G. Whitehead, ed. Maternal nutrition during pregnancy and lactation. Bern, Verlag Hans Huber, 1980. 91. PRENTICE, A. M. ET AL Lancet, 2: 886 (1980) 92. PRENTICE, A.M. EF AL American journal of clinical nutrition, 34: 2790 (1981). 93, ProseR, C. G. & Hartmann, P. E, Proceedings of the Australian Society of Medical Research, 13: 14 (1980) 94, RAIALAKSHML K. & SRIKANTIA, S. G. American journal of clinical nutrition, 33: 664 (1980), 95. RAMASASTRY, B. V. & INDRAVATL D, Indian journal of medical research, (1957), 96. Rao, K. S. ET AL Bulletin of the World Health Organization, 20: 328 (1959), 97. RaTriGAN, 8. ETAL. British journal of mucrtion, 48: 243 (1981), 98. RowLano, M. G. M. er aL Lancet, 2: 136 (1978), 99. ScHANtER, R. J. & OH, W, Journal of pediatrics, 96: 679 (1980) 100. SHANGHAI CHILD HEALTH CooRDINATION Grour. Journal of tropical pediatrics and environmental child health, 21: 284 (1976). 10. SHuKRy, AS. ET AL. Gazette of the Egyptian Paediatric Association, 21: 47 (1973) 102. Sypow, G. V. Acta Paediatrica, 32: 756 (1944), 103. THomrsox, W. B. American journal of obstetrics and gynecology, 20: 662 (1933). 104, THomsow, A.M. & Brack, A. E. Bulletin of the World Health Organization, $52; 163 (1975), 105. Tin Tin 00 & Kui MAUNG-NAING. Food and nutrition bulletin, 4 (8): 66 (1982). 106. Toverup, K. U, & Toverun, G. Acta paediatrica, 12: 1 (1931). 107. UxDERwoon, B. A. & HorvaNpeR, Y. Acta paediatrica scandinavica, Suppl. 294 (1982). 108. UNDERWOOD, E. J. Trace elements in relation to human and animal nutrition, 4th ed., New York, Academic Press, 1977, 109. USAID. Sierra’ Leone National Nutrition Survey. Washington, DC, 1978. 110. VAN STEENBERGEN, W. L. er AL, Journal of tropical pediaircs, 27: 155 (1981), IIL. VauGHaN, L.A. er aL In: M, Kirshgessner, ed, Proceedings of the Third Symposium on Trace Element Metabolism in Man and Animals, Munich, 1977 112. VAUGHAN, L.A. Er AL American journal of clinical nutrition, 32: 2301 (1979) 2 4478 THE QUANTITY AND QUALITY OF BREAST MILK 113, VENKATACHALAM, P. S. A study of the diet, nutrition and health of the people of the Chimbu area (New Guinea highlands). Moresby, New Guinea, Department of Public Health, 1962, (Monograph 4) 114, VENKATACHALAM, B.S. eT AL Journal of pediatrics, 61: 262 (1962) 115. VENKATACHALAM, P. S. et aL Journal of tropical pediatrics, 13: 70 (1967). 116. Voosr, J. L. Clinical obstetrics and gynecology, 8: 435 (1978), 117, WaLLoREN, A. Acta paediatrica, 32: 778 (1944), 118. WeicHert, C. E, Advances in pediatrics, 27: 391 (1980), 119. WetsourX, H. F. Journal of tropical pediatrics, 1: 34 (1955). 120. WitTenEAD, R. G. & Paul, A. A. Lancet, 2: 161 (1981), 121, Wurrentan, RG. er AL Food and mutrtion bulletin, Suppl. 5, p. 242 (1981) 122. Wuiteneap, R. G. ev at. In: B. Wharton, ed. Topics in pediatrics 2: nutrition in childhood, Tunbridge Wells, Pitman Medical, 1980, p. 20 123. WatTEHEAD, R. G., ED. Food and mutrtion bulletin, Suppl. 6 (1982) 124. Wurteneap, RG. er at, Lancet, 2: 178 (1978) 125. Wippows, S. T, & LoweNreLD, M. F. Biochemical journal, 27: 1400 (1933) 126. WHO Technical’ Report Series, No, $22, 1973.