Previous - Next
The mean metabolisable energy values at 4-6 and 10-12 weeks of 87 and 89 kcal/kg are substantially lower than the FAO/WHO/UNU (1973 and 1985) recommended dietary allowance for energy of 120 kcal/kg per day in 1973, and 116 kcal/kg per day in 1985. If these new data are accepted, then current estimates of requirement for energy and the recommended dietary allowance would need to be reconsidered.
The major reason for the low values of metabolisable energy reported is the low energy content of breast milk. That these values are so substantially different from those reported in numerous studies on expressed breast milk raises questions of major biological and clinical importance and at the same time invites critical appraisal of the new techniques employed.
In this study, values for milk energy content have been obtained without recourse to the difficulties of breast-milk sampling, and without the need for special apparatus or observer interference which might affect the normal biology of breast-feeding.
At 3 months of age, exclusively breast-fed infants had a mean metabolisable energy intake (MEI) of only 58 kcal/100 mL. It was impossible to compare simultaneously breast-milk energy content obtained by our techniques and that derived by milk expression, since the latter would have involved disruption of normal feeding. However, many investigators (DHSS, 1977) have examined expressed breast-milk composition. A recent DHSS study (1977) derived a value of 70 kcal/100 mL, and other key reports have quoted values as follows: 71 kcal/100 mL (MACY and KELLY'S 1953 review of 1500 publications); 75 kcal/100 mL (MACY and KELLY, 1961); 75 kcal/100 mL (FOMON, 1974); 76 kcal and 70 kcal/100 mL (KON and MAWSON, 1950); 64 kcal/100 mL (MORRISON, 1952); and 80 kcal/100 mL (McCANCE and WIDDOWSON, 1967). Given the technical problems of milk sampling, published values would be expected to vary, and very low fat and energy values can be obtained if milk for analysis is not thoroughly mixed immediately prior to sampling. Nevertheless, earlier evidence suggests a mean value for expressed breast milk of at least 70 kcal/100 mL, a figure which lies well above the upper limit of the 95% confidence interval for values obtained here (95% confidence interval: 54-62 kcal/100 mL).
Two factors add confidence to the validity of these low values for breast-milk energy. Firstly, the value for the energy density of the formula using the same method at 3 months of age was close (within one standard error) to the manufacturers' quoted figure based on macronutrient composition. Secondly, it might be expected that the value for the energy content of breast milk, derived from isotope techniques, would be slightly higher than that for the formula, since mean reported energy values for breast milk are around 2 kcal/100 mL more than the formula used here; and fat absorption and hence energy absorption is likely to be greater from human milk in view of its pattern of fatty acids and the presence of milk lipases. Yet, in contrast, our data showed that the energy content of the formula was significantly higher than that in breast milk (66 versus 58 kcal/100 mL).
These data suggest therefore that the traditional model, using expressed breast milk, may not be the most appropriate one for deriving values for milk fat and energy content. As suggested before, expressed milk may contain high amounts of fat and therefore be energy-rich hindmilk which the breast-fed infant would not normally receive.
A potential problem with our findings, however, is that the value obtained for the energy content of the formula in infants at 6 weeks of age was only 60 kcal/100 mL, substantially lower than the manufacturers' figure of 68 kcal/100 mL. The value for breast milk was even lower at this age, 53 ±2 kcal/100 mL. Why should these 6-week values be so low? Methodological error due to the difficulty in studying energy metabolism in young infants must be considered. However, the standard deviation of values for milk energy were similar at 6 weeks and 3 months, and our previous validation in even smaller, faster-growing preterm infants showed that accurate data for energy expenditure and milk intake were obtained, using similar methods and assumptions. A constant bias due to inappropriate choice of values for RQ (0.87) and the proportion of water liable to fractionation (0.13) seems unlikely; indeed we would regard the latter figure as conservative, and a higher assumed value for fractionated water losses would have reduced further the estimated energy content of breast milk. In younger infants, a greater proportion of ingested energy is stored in new tissue; thus, an error in the estimation of this component would be more influential. Nevertheless, we have calculated that large unidirectional errors in the reference values for lean body mass composition used for calculating energy stored would have a trivial effect on the estimated energy content of breast milk.
A most likely major contributing explanation for the observed low metabolisable energy content of formula or breast milk at 6 weeks is the inappropriateness of the standard Atwater conversion factors. These factors are used to estimate metabolisable (available) energy from macronutrient composition; they are significantly smaller numerically than those which would be used to calculate gross energy. Accurate definition of these factors depends on knowledge of nutrient absorption from the gut. It is possible that energy absorption at 4-6 weeks is less than at 10-12 weeks; and it is absorbed energy which would be derived from the doubly-labelled water method.
The low metabolisable energy and
energy content of human milk reported here suggests there is indeed a need for reappraisal
of recommendations for energy intake in early infancy. We are currently exploring the
energy intake of older infants and children. Nevertheless, more validation studies are
required.
If we choose to base estimated energy requirements in early life on the 'factorial' or 'food intake' approaches, then it is encouraging that the new methodologies available are likely to improve our estimates. The question is whether we should be satisfied with the philosophy behind the current approaches. My own view is that we should not, though without intended criticism of various international bodies who have had to make recommendations on the best available data.
A number of debatable assumptions have had to be made. For instance, estimation of energy requirement has depended on an internationally-agreed assumption that NCHS growth data should be taken as a standard. Yet, optimal growth has never been examined in relation to long-term health. Moreover, growth performance has apparently been subject to secular trends. For example, studies on children's skinfold thicknesses show that in Britain (WHITEHEAD, PAUL and COLE, 1989), Australia (BOULTON, 1981), Canada (YEUNG, 1983), and Germany (SCHLUTER et al., 1976) modern feeding practices appear to result in thinner children than those studied 20 years ago. Nowadays, breast-feeding is actively promoted and the breast-fed infant, beyond 4 months is lighter than its bottle-fed counterpart. This difference is maintained throughout the rest of infancy and on to the preschool period. In both these examples medical advice has played a major role in changing growth performance; but are these changes good or bad? Clearly, we can set the energy requirement at a level which depends on our current judgement on how babies should grow, but without satisfactory outcome information we are left with judgements and not data.
To explore another avenue, the successful validation of the doubly-labelled water method will undoubtedly encourage investigators to use total energy expenditure measurements to derive energy requirements throughout childhood. Beyond infancy, total energy expenditure becomes a reasonable proxy for metabolisable energy intake since energy retention or loss during the study period is likely to be small and can be adjusted for without introducing significant error. However, can requirement be defined as the energy intake of modern children? Effectively the child is being asked to advise the adviser on how to advise the child.
We have recently identified a
further confounding factor; namely that high energy intakes may increase energy
expenditure without a corresponding large increase in energy storage. Thus, our
preliminary, unpublished data show that 30 formula-fed infants aged 4-12 weeks had a mean
energy intake 16% higher than 20 breast-fed infants, yet, they expended 80% of this excess
and laid down only 20% in new tissue. Our findings indicate that formula-fed infants
consume more energy than breast-fed babies, but do not gain weight (i.e., stored energy)
appreciably faster; thus, weight gain may not always be a good indicator of energy intake.
Moreover, there is some circularity in the argument that energy expenditure should be used
to define the needs for energy intake, if energy intake can determine energy expenditure.
The confusion in this field is part of a more general problem in infant nutrition research. It might be asked why, after intensive research throughout this century, there is still uncertainty over almost every aspect of nutritional practice. When such a large body of scientific data exists, and yet the answers remain in doubt, it seems reasonable to reappraise the questions. It is relevant in this context to consider the evolution of other areas of clinical investigation. Normally, research in clinical science is a three-phase process. Phase 1 documents anecdotal observation and clinical experience. Phase 2 includes broader-based epidemiological observation or detailed physiological experiment. Finally, in phase 3, outcome studies are employed to test the hypothesis that specific interventions result in desirable clinical outcome responses. Successful studies in this last category place clinical management on a secure scientific footing. To take an analogy: it would be of limited value to conduct a phase-2 study on the physiologic effects of a range of new blood pressure lowering agents if it had not been shown in a phase-3 study that it was actually worthwhile treating hypertension, in terms of improved outcome. It seems that infant nutrition in general, and indeed the field of energy requirements, has become largely stuck at phase 2 in its evolutionary progress. Investigators of energy metabolism have been preoccupied by phase-2 physiological research and have not addressed the critical, long-term phase-3 question: does it matter what energy intake we give infants and young children in terms of their performance and morbidity in later life?
Since this may seem, at first sight, a most difficult objective, some justification is needed for its pursuit. The possibility that dietary experience during early life has lasting consequences invokes the concept of 'programming', meaning here the general process whereby a stimulus (or insult), applied during a 'critical' or 'sensitive' period of development, has a long-term or permanent effect in the organism. Many examples (LUCAS, 1987) of programming, resulting either from internal triggers or external agents, have been described in animals and man.
There is increasing evidence that early diet may act as a programming agent in animals. It is not the purpose of this article to review these data, but a few examples will be given to illustrate the point.
A number of investigators have explored whether early dietary manipulation programs alter metabolism in ways that might be relevant to the development of obesity, abnormalities in lipid metabolism, and atherosclerosis. In HAHN'S (1984) studies on rats, litter size was manipulated. He examined outcome, for instance, when litter size was 14 or reduced to 4; in small litters the pups would be overfed during the short breast-feeding period. In adulthood, rats from small litters had significantly higher plasma cholesterol and insulin concentrations. Hahn also showed that weaning animals on to a high-carbohydrate diet in the neonatal period resulted in similar long-term effects on plasma cholesterol and insulin, and in addition influenced in adulthood the activities of key enzymes concerned with fatty acid and cholesterol metabolism (fatty acid synthetase and HMG-CoA reductase activities). Such biochemical changes in man could be important in relation to adult morbidity.
Rats are born immature, and it might be argued that they are not a good model for man, but long-term programming has also been demonstrated in primates. In a study by LEWIS and coworkers (1986), infant baboons were randomly assigned to one of three formulas for the first 4 months. The formulas provided low, normal and high energy intakes. After the 4-month period all the animals were fed in the same way. The excess weight gained during infancy in the animals with high intakes was soon lost. In males the loss was permanent; adult males overfed in infancy had normal adult weight. In female baboons, however, early overfeeding resulted in a dramatic rise in body weight and fat mass during adolescence and early adult life. In this instance, the effects of the initial 'programming' event were not manifested until much later in life, raising the important question as to how such a 'memory' could tee 'stored', passed through many cell generations, and then 'expressed'.
The influence of breast- versus formula-feeding on later lipid metabolism and vascular disease has also been explored in a series of studies by MOTT (1986), MOTT et al. (1989) and LEWIS et al. (1985) using the baboon model. In man, clearly it is not possible to randomly assign infants to breast- or formula-feeding; consequently, all outcome studies comparing the two are confounded by the considerable socioeconomic and educational differences between breast- and bottle-feeders. In baboons such randomisation is possible. In the studies of Lewis and Mott, these assignments were only applied during infancy (first 4 months); beyond that the animals were fed in the same way. Compared with the formula-fed group, those who were breast-fed in infancy had increased cholesterol absorption, reduced cholesterol turnover, higher plasma levels of LDL and VLDL cholesterol and, when placed on a high saturated fat (Western style) diet, they developed lower levels of potentially protective HDL cholesterol in adult life. These lipid abnormalities would be expected to result in an increased risk of atherosclerosis, and indeed the animals showed, at post mortem, a significantly greater area of atherosclerotic plaque if they had been breast-fed in infancy (LEWIS et al., 1985). By using formulas with different cholesterol content, the investigators found that cholesterol intake itself was not the factor which accounted for their findings, which remain unexplained. The significance of these data for man are unknown, and indeed it could be argued that, in humans, morbidity from vascular disease may be more related to thrombotic events than to atherosclerosis per se. Nevertheless, these findings indicate that early life may be a critical period for nutrition in primates and emphasise the importance of long-term, phase-3 studies on human nutrition.
In our own unit, we have strictly randomly assigned nearly 1000 preterm infants to the diet they received in the early weeks and we are now following them up indefinitely. The cohort is approaching 78 years. Data analysed so far, after blind evaluation at 18 months, show that the brief period of dietary manipulation after birth results in major differences between diet groups in growth and neurodevelopment (LUCAS et al., 1989a; LUCAS et al., 1989b). This study also serves to demonstrate the feasibility of long-term outcome studies in man.
There are a number of opportunities
to examine formally the effects of early energy intake on later outcome in healthy babies,
born at full term. Currently we are conducting a prospective randomised trial of energy
intake in formula-fed babies with planned long-term follow-up. Interestingly, our
unpublished pilot data indicate that energy intake at 3 months in normal-term infants is
not related to skinfold thickness at that age, but significantly related to skinfold
thickness at 2 years. Weaning is another time period when randomised intervention would be
feasible. It would certainly be disappointing if, in another 20 years time, the same
issues were being debated when, in the meantime, critical outcome data could have been
collected on the effects of early energy intake on later growth, obesity, neurodevelopment
and markers for vascular disease.
ASHWORTH, A.: Metabolic rates during recovery from protein-calorie malnutrition. The need for a new concept of specific dynamic action. Nature, 223. 407-409 (1969).
BLAND, J.M., ALTMAN, D.G.: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet, i, 307-310 (1986).
BOULTON, J.: Nutrition in childhood and its relationships to early somatic growth, body fat, blood pressure and physical fitness. Acta Paediatr. Scand., Suppl. 284 (1981).
BROOKE, O.G., ASHWORTH, A.: The influence of malnutrition on the post-prandial metabolic rate and respiratory quotients. Br. J. Nutr., 27, 407-487 (1972).
COWARD, W.A.: Measuring milk intake in breast-fed babies. J. Pediatr. Gastroenterol. Nutr., 3, 275-279 (1984).
COWARD, W.A., WHITEHEAD, R.G., SAWYER, M.B., PRENTICE, A.M., EVANS, J.: New method for measuring milk intakes in breast-fed babies. Lancet, ii, 13-14 (1979).
COWARD, W.A., COLE, T.J., SAWYER, M.N., PRENTICE, A.M., ORR-EWING, A.K.: Breast milk intake measurement in mixed-fed infants by administration of deuterium oxide to their mothers. Hum. Nutr. Clin. Nutr., 36C, 141-148 (1982).
COWARD, W.A., PRENTICE A.M.: MURGATROYD, P.R. et al.: Measurement of CO2 and water production fates in man using 2H18O-labeled H2O - comparisons between calorimeter and isotope values. In: Human Energy Metabolism Physical Activity and Energy Expenditure Measurements in Epidemiological Research Based upon Direct and Indirect Calorimetry. Euronut. Report, 5, 126-128 (1984).
COWARD, W.A., ROBERTS, S.B., PRENTICE, A.M., LUCAS, A.: The 2H218O method for energy expenditure measurements - clinical possibilities, necessary assumptions and limitations, pp. 169-177. European Society for Parenteral and Enteral Nutrition, 7th Congress, Munich, 1986.
COWARD, W.A., ROBERTS, S.B., COLE, T.J.: Theoretical and practical considerations in the doubly-labelled water (2H218O) method for the measurement of carbon dioxide production rate in man. Eur. J. Clin. Nutr., 42, 207-212 (1988).
DAVIES, P.S.W., LUCAS, A.: Quetelet's index as a measure of body fatness in infants. Early Hum. Dev., 1990 (in press).
DHSS Report on health and social subjects, No. 12. The composition of mature human milk. HMSO, London, 1977.
DURNIN, J.V.G.A.: Some problems in assessing the role of physical activity in the maintenance of energy balance. In: Energy Intake and Activity (Current Topics in Nutrition and Disease Vol ii), pp. 101-103, E. POLLITT, P. AMANTE (Eds.). Alan R. Liss Inc., New York, 1984.
FOMON, S.J.: Infant Nutrition, 2nd ed. WB Saunders, Philadelphia, PA, 1974.
FOMON, S.J., HASCHKE, F., ZIEGLER, E.E., NELSON, S.E.: Body composition of reference children from birth to age 10 years. Am. J. Clin. Nutr., 35, 1169-1175 (1982).
FAO: Energy and protein requirements. Report of a Joint FAO/WHO Ad Hoc Expert Committee. FAO Nutrition Meetings Rep. Ser. 52, 1973.
FAO/WHO/UNU: Energy and protein requirements. WHO Tech. Rep. Ser. No. 724. WHO, Geneva, 1985.
GARROW, J.S.: Energy balance and obesity in man, pp. 17-47. 2nd ed. Biomedical Press, Elsevier/North Holland, 1978.
GUDINCHET, F., SCHUTZ, Y., MICHELIN, J-L., STETTLER, E., JÉQUIER, E.: Metabolic cost of growth in very low birthweight infants. Pediatr. Res., 16, 1025-1030 (1982).
HAHN, P.: Effect of litter size on plasma cholesterol and insulin and some liver and adipose tissue enzymes in adult rodents. J. Nutr., 114, 1231 (1984).
HAMILL, P.V.V., DRIZD, T.A., JOHNSON, C.L., REED, R.B., ROCHE, A.F.: NCHS growth curves for children, birth to 18 years. US Department of Health, Education and Welfare Publications No. PHD 78-1650. National Centre for Health Statistics, Hyattsville, MD, 1977.
HOW, T.J., ASHMORE, M.P., ROLFE, P., LUCAS, A., LUCAS, P.J., BAUM, J.D.: A Doppler ultra-sound technique for measuring human milk flow. J. Med. Eng. Technol., 33, 66-71 (1979).
KLEIN, P.D., JAMES, W.P.T., WONG, W.W. et al.: Calorimetric validation of the doubly labeled water method for determination of energy expenditure in man. Hum. Nutr. Clin. Nutr., 38C, 95-106 (1984).
KON, S.K. MAWSON, E.H.: Human milk. Special Report Series of the Medical Research Council No. 269. HMSO, London, 1950.
LEWIS, D.S., MOTT, G.E., McMAHAN, C.A., MASORO, J., CAREY, K.D., McGILL, H.C.: Deferred effects of preweaning diet on atherosclerosis in adolescent baboons. Atherosclerosis, 5, 347-354 (1985).
LEWIS, D.S., BERTAUT, H.A., McMAHON, A., McGILL, H.C., CAREY, K.D., MASORO, E.J.: Preweaning food intake influences the adiposity of young adult baboons. J. Clin. Invest., 78, 899-905 (1986).
LIFSON, B., GORDON, G.B., McCLINTOCK, R.: Measurement of total carbon dioxide production by means of D218O. J. Appl. Physiol., 7, 704-710 (1955).
LIFSON, N., McCLINTOCK, R.: Theory of use of the turnover rates of body water for measuring energy and material balance. J. Theoret. Biol., 12, 46-74 (1966).
LUCAS, A.: Diet in early life: evidence for its later effects. In: Scientific Report No. 8. Infant Nutrition and Cardiovascular Disease, pp. 11-16, D. BARKER (Ed.). MRC Environmental Epidemiology Unit, Southampton, U.K., 1987.
LUCAS, A., LUCAS, P.J., BAUM, J.D.: The nipple-shield sampling system: a device for measuring the dietary intake of breast-fed infants. Early Hum. Dev., 4, 365-372 (1980).
LUCAS, A., EWING, E., ROBERTS, S.B., COWARD, W.A.: Measurements of milk intake by deuterium dilution. Arch. Dis. Child., 62, 796-800 (1987a).
LUCAS, A., EWING, G., ROBERTS, S.G., COWARD, W.A.: How much energy does the breast-fed infant consume and expend? Br. Med. J., 295, 75-77 (1987b).
LUCAS, A., MORLEY, R., COLE, T.J., GORE, S.M., DAVIS, J.A., BAMFORD, M.F.M., DOSSETOR, J.F.B.: Early diet in preterm infants and developmental status in infancy. Arch. Dis. Child., 64,1570-1578 (1989a).
LUCAS, A., BROOKE, O.G., BAKER, B.A., BISHOP, N., MORLEY, R.: High alkaline phosphatase activity and growth in preterm neonates. Arch. Dis. Child., 64, 902-909 (1989b).
MACY, I.G., KELLY, H.J., SLAON, R.E.: The composition of milks. National Research Council Publication No. 254. National Academy of Sciences, Washington, DC, 1953.
MACY, E.G., KELLY, H.J.: Human milk and cows' milk in infant nutrition. In: Milk: The Mammary Gland and Its Secretion. Vol. II, pp. 265-304, S.K. KON, A.T. COWIE (Eds.). Academic Press, New York, 1961.
McCANCE, R.A., WIDDOWSON, E.M.: The composition of foods. HMSO, London, 1967.
MORRISON, S.D.: Human Milk. Commonwealth Bureau of Animal Nutrition Technical Communication No. 18. University Press, Aberdeen, U.K., 1952.
MOTT, G.E.: Deferred effects of breast-feeding versus formula feeding on serum lipoprotein concentrations and cholesterol metabolism in baboons. In: The Breast-fed Infant: A Model for Performance, pp. 144-149, L.J. FILER Jr., S.J. FOMON (Eds.). Report of the Ninety-First Ross Conference on Paediatric Research. Ross Laboratories, Columbus, OH, 1986.
MOTT, G.E., JACKSON, E.M., McMAHON, C.A., McGILL, H.C.: Cholesterol metabolism in adult baboons is influenced by infant diet. J. Nutr., 1989 (in press).
NAGY, K.: CO2 production in animals: analysis of potential error in the doubly labeled water method. Am. J. Physiol., 238, R466-R473 (1980).
NEVILLE, M.C., KELLER, R.P., SEACAT, J., CASEY, C.E., ALLEN, J.C., ARCHER, P.: Studies on human lactation. 1. Within-feed and between-breast variations in selected components of human milk. Am. J. Clin. Nutr., 40, 635-646 (1984).
PRENTICE, A., PRENTICE, A.M., WHITEHEAD, R.G.: Breast-milk fat concentrations of rural African women. 1. Short-term variations within individuals. Br. J. Nutr., 45, 483-494 (1986).
PUTET, G., SENTERRE, J., RIGO, M., SALLE, B.: Nutrient balance, energy utilization and composition of weight gain in very-low-birthweight infants fed pooled human milk or a protein formula. J. Pediatr., 105, 79-85 (1984).
REICHMAN, B., CHESSEX, P., VERELLEN, G. et al.: Dietary composition and macronutrient storage in preterm infants. Pediatrics, 72, 322-328 (1983).
ROBERTS, S.B., COWARD, W.A., SCHLINGENSEIPEN, K-H., NOHRIA, V., LUCAS, A.: Comparison of the doubly labelled water (2H218O) method with indirect calorimetry and a nutrient balance study for simultaneous determination of energy expenditure, water intake and metabolisable energy intake in preterm infants. Am. J. Clin. Nutr., 44, 315-322 (1986).
ROBERTS, S.B., LUCAS, A.: Measurement of urinary constituents and output using disposable napkins. Arch. Dis. Child., 60, 1021-1024 (1985a).
ROBERTS, S.B., LUCAS, A.: Effect of two extremes of dietary intake on protein accretation in preterm infants. Early Hum. Dev., 12, 301-307 (1985b).
SAINT, L., MAGGIORE, S., HARTMAN, P.E.: Yield and nutrient content of milk in eight women breast-feeding twins and one women breast-feeding triplets. Br. J. Nutr., 56, 49-58 (1986).
SCHLUTER, K., FUNFACK, W., PACHALY, J., WEBER, B.: Development of subcutaneous fat in infancy. Standards for tricipital, subscapular and suprailiacal skinfolds in German infants. Eur. J. Pediatr., 123, 255-267 (1976).
SCHOELLER, D.A., VAN SANTEN, E.: Measurement of energy expenditure in humans by doubly labelled water method. J. Appl. Physiol., 53, 955-959 (1982).
SOUTHGATE, D.A.T., BARRETT, I.M.: The intake and excretion of caloric constituents of milk by babies. Br. J. Nutr., 20, 363-372 (1966).
SOUTHGATE, D.A.T., DURNIN, J.V.G.A.: Calorie conversion factors. An experimental reassessment of the factors used in the calculation of the energy value of human diets. Br. J. Nutr., 24, 517-535 (1970).
United Nations University: Nutrient requirements for catch-up growth, pp. 34-48. In: Protein Energy Requirements under Conditions Prevailing in Developing Countries: Current Knowledge and Research Needs. The United Nations University, Tokyo, 1979.
WATERLOW, J.C.: Basic concepts in the determination of nutritional requirements of normal infants. In: Nutrition during Infancy, pp. 1-19, R.C. TSANG, B.L. NICHOLAS (Eds.). Hanley and Belfus Inc., Philadelphia, PA, 1989.
WEIR, J.B. de V.: New method for calculating metabolic rate with special reference to protein metabolism. J. Physiol., 109, 1-9 (1949).
WHITEHEAD, R.G., PAUL, A.A., COLE, T.J.: Diet and growth of healthy infants. J. Hum. Nutr. Dietetics, 2, 73-84 (1989).
WILLIAMSON, S., FINUCAINE, E., ELLIOTT, J., GAMSU, H.R.: Effect of heat treatment of human milk on absorption of nitrogen, fat, sodium, calcium and phosphorus by preterm infants. Arch. Dis. Child., 53, 555-563 (1978).
YEUNG, D.L.: Infant nutrition: a study of feeding practices and growth from birth to 18 months. Canadian Public Health Association, Ontario, 1983.