Infant sex predicts breast milk energy content

AMERICAN JOURNAL OF HUMAN BIOLOGY 22:50–54 (2010) Infant Sex Predicts Breast Milk Energy Content CAMILLE E. POWE,1* CHERYL D. KNOTT,2 AND NANCY CONKLIN-BRITTAIN11Department of Anthropology, Harvard University, Cambridge, Massachusetts 02138 2Department of Anthropology, Boston University, Boston, Massachusetts 02115 During human evolutionary history, and for many around the world, breast milk is the primary source of nutritional energy for infants. Variation in breast milk quality might logically have important effects on infanthealth, growth, and development, yet the sources of this variation remain largely unelucidated. We quantified nutrientand energy content of breast milk from 25 healthy, well-nourished Massachusetts mothers with infants aged2–5 months. We examined several potential sources of variation in milk quality, particularly feeding patterns, infantsex, and maternal breast growth during pregnancy. After controlling for time since last feeding, a known correlateof milk composition, we found that mothers of male infants produced milk that had 25% greater energy contentthan mothers of female infants (P < 0.001). Change in maternal bra cup size during pregnancy was associated with16.17 kcal/100 ml greater energy content of milk (P 5 0.009), but was not significant after taking infant sex intoaccount. Greater nutritional investment in sons may account for the greater observed growth rates in male com-pared to female infants. Am. J. Hum. Biol. 22:50–54, 2010.
Milk caloric content is known to vary greatly between characteristics, breast development, and infant character- individual human mother–infant pairs, but the source of istics reflective of energy requirement.
much of this variation is unexplained (Allen et al., 1990;Michaelsen et al., 1994; Mitoulas et al., 2002; Prentice et al., 1981a,b). During pregnancy, under the influence ofestrogen, prolactin, progesterone, and other hormones, We conducted a cross-sectional study of well-nourished breast glandular tissue undergoes a period of glandular proliferation and differentiation (Gardner and Shoback, Eligible mothers were exclusively breastfeeding their 2-to 5-month-old infants. We considered mothers to be exclu- Breast milk energy content may differ between individ- sively breastfeeding if their infants relied on breast milk ual mothers based on genetic differences or because of for all of their caloric needs, but we did not exclude moth- differences in the proliferation and differentiation of milk ers whose infants had received formula on a rare occasion epithelial cells during pregnancy under a specific hormo- (1–3 times in life). Both stay-at-home mothers who nursed nal milieu. This may be associated with maternal their infants at the breast, and working mothers who nutritional status, innate characteristics, or fetal cues.
pumped while away from their infants and fed their Prior research has shown that milk caloric content fails to infants at the breast while at home, could participate. We respond to maternal dietary intake even among malnour- recruited mothers and their infants from Harvard Univer- ished women (Prentice et al., 1983). Exercise similarly sity Health Services OB/GYN Department in Cambridge, does not affect the caloric content of breast milk (Dewey Massachusetts, The Center for Breastfeeding in Sand- et al., 1994). There is some evidence that the size of mater- wich, Massachusetts, and mothers groups in Watertown, nal fat stores is associated with milk caloric content in Arlington, and Boston, Massachusetts through flyers, poorly nourished populations, but it is unclear whether emails, or through clinic staff. All procedures involving this relationship holds up in well nourished populations.
human subjects were approved by the Committee on the Differences in fat stores alone do not explain most of the Use of Human Subjects at Harvard University, Cam- variation in milk caloric content between human mothers bridge, Massachusetts and we obtained written informed (Barbosa et al., 1997; Brown et al., 1986; Nommsen et al., 1991; Perez-Escamilla et al., 1995; Prentice et al., 1981a).
We met with each subject pair once at either their home Despite the fact that changes in the breast during preg- or one of two clinical settings by the subject’s choice.
nancy are responsible for breast milk production capacity Mothers were asked to pick a time at which their infant postpartum, breast size and change in breast size during would not typically be feeding if possible. We collected pregnancy has seldom been examined in studies of milk data on maternal characteristics, infant characteristics, and feeding behaviors by self-report. Participants com- Differences in milk energy content between mothers pleted a survey which requested information about their may alternatively be responsive to infant demand, giventhat prolactin, the hormone which promotes the produc- Contract grant sponsor: Harvard College Research Program, Cambridge, tion of milk is exquisitely sensitive to the suckling stimu- lus. It is well established that milk fat content changes *Correspondence to: Camille E. Powe, 159 Longwood Avenue, No. 1 over a feeding and is related to the fullness of the breast, Brookline, MA 02446. E-mail: [email protected] but it has not been shown that an infant’s level of demand Received 14 December 2008; Revision received 8 February 2009; is associated with changes in milk composition or caloric content. We designed a study to determine whether differ- Published online 16 June 2009 in Wiley InterScience (www.interscience.
ences in milk composition were associated with maternal INFANT SEX PREDICTS BREAST MILK ENERGY CONTENT age, race, parity, and breastfeeding history with previous feeding per day was calculated as: Minutes feeding per infants, pre-pregnancy weight, pregnancy weight gain, day 5 (Number of daytime feedings) 3 (Average length of pre-pregnancy and current bra size, age at menarche, and daytime feedings) 1 (Number of nighttime feedings) 3 medical conditions. Participants were also asked to com- (Average length of nighttime feedings). Infant weight gain plete a 24 h dietary recall. Mothers also reported their was calculated as the birth weight subtracted from the infant’s birth weight, birth date, due date, typical feeding current weight. We used pounds and ounces for the units schedule, medical conditions, and whether they fed on one of infant weight rather than grams because mothers were breast more than the other. They were asked to complete better able to recall their infants birth weights in pounds a breastfeeding/pumping log from the past 12 h.
Infant weight was measured in pounds using the We conducted an exploratory statistical analysis to Health-O-Meter Grow with Me Teddy Bear Baby to Tod- determine which of the maternal and infant characteris- dler Scale (Jarden Corporation, Rye, NY). Infant length in tics that we measured were correlated with milk energy inches was measured using the Seca 210 Infant Measure density. We calculated Pearson’s correlations between con- Mat for Infants and Small Children (Seca, Hamburg). We tinuous maternal and infant characteristics. Student’s measured maternal weight with the Tanita HD317 Digital t-tests (two-tailed) were used to compare the characteris- Scale with Memory Function (Tanita Corporation, Arling- tics and feeding behaviors of male infants with those of ton Heights, Illinois). Maternal height was measured female infants, the characteristics of the mothers of male using a 6-foot long ruler. Maternal body fat was measured infants to those of the mothers of female infants, the com- using the Omron HPF 306 Body Fat Tracker (Omron position of the milk of mothers with male infants to the Healthcare, Kyoto, Japan), which uses bioelectrical composition of the milk of the mothers with female infants.
impedance to determine body fat percentage.
Linear regression procedures were used to create mod- A milk sample was collected from each participant at els for milk energy density (kcal/100 ml) which included the study meeting using an electric breast pump. Mothers maternal and infant variables associated with milk were permitted to use their own pumps, but if they chose energy density. Goodness of fit was checked by examining to or if they did not own a pump, they used the Ameda residuals. All analyses were performed using Statistical Elite (Hollister, Liberty, IL) hospital grade breast pump Package for the Social Sciences (SPSS).
with a sterile collection kit. Mothers were instructed topump one breast until it was empty. Milk samples were given a random number prior to analysis.
Samples were placed on ice before freezing and freeze We collected data from 25 mother–infant pairs between drying. Total nitrogen was determined using the Kjeldahl September 2005 and February 2006. The characteristics of procedure (Pierce and Haenisch, 1947). Crude protein the mothers, the infants, and their feeding behaviors are (CP) was estimated as (Total Nitrogen) 3 6.25. The con- summarized in Table 1. The liquid energy density of milk centration of free simple sugars (FSS) was determined was highly variable between mothers. Kilocalories per 100 using the sulfuric acid-phenol method (Dubois et al., 1956; ml ranged from 32.9 to 116.1 with a mean of 68.48 6 18.87.
Strickland and Parsons, 1972). Lipids were extracted from Figure 1 shows the distribution of milk energy densities in dried breast milk using a method modified from the Asso- our sample. Milk energy density was not significantly cor- ciation of Official Analytical Chemists (Association of Offi- related with maternal age, maternal parity, maternal pre- cial Analytical Chemists, 1984). Ash was measured by pregnancy weight, maternal pregnancy weight gain, heating sample to 5008C overnight in a muffle furnace, maternal current weight, maternal body fat percentage, and the weight of the residue was measured (Association maternal bra size (pre or post pregnancy), lactation his- of Official Analytical Chemists, 1984). Total nonstructural tory, maternal weight, or household income. Milk energy carbohydrates (TNC) were estimated as the percent fat, density was not significantly correlated with infant age, percent ash, and percent crude protein subtracted from age corrected for gestational age, infant (birth or current) 100%. Milk caloric density was calculated by using the weight, infant (birth or current) length, infant weight fractions of TNC, lipid, and CP. Lipid was assumed to gain, infant feeding frequency, or minutes feeding per day.
have the physiological fuel value of 9 kcal/g, TNC and Among 25 mothers, the liquid energy density of milk CP were assumed to have 4 kcal/g (National Resource produced by mothers of male infants was 75.56 kcal/ 100 ml versus 60.811 kcal/100 ml of milk produced by Maternal and infant ages were calculated from the birth mothers of female infants (t-test, P 5 0.049). To control for dates reported on the questionnaire using an age calcula- the fullness of the breast, which can affect the milk fat tor website (Pon, 2000). Corrected infant age (corrected content, and thus caloric content of the milk, we controlled for gestational age) was calculated using the same calcula- for time since last feeding (Daly et al., 1993; Jackson tor, but by plugging in the due date of the infant instead of et al., 1988; Prentice et al., 1981a). The effect of infant sex the birth date (Pon, 2000). Self reported prepregnancy was statistically significant after adjustment for time and current bra sizes were scored by cup size. Size A was since last feeding (multiple regression, P < 0.001) (Table given a score of 1 and each ascending size was given the 2). After this adjustment in multiple regression analysis, next ascending number: B 5 2, C 5 3, etc. Change in cup mothers of male infants in our sample had a milk energy size was taken as pre-pregnancy size subtracted from cur- concentration that was 24.68 kcal/100 ml greater than rent size. Lactation history was calculated for each milk energy of mothers of females. Based on a milk con- mother as the total number of months spent breastfeeding sumption volume of 788 ml per 24 h (Daly et al., 1993), the children in her lifetime. Time spent nursing two infants difference in composition between milk consumed by at the same time was counted twice. The average number infants of different sexes would result in a difference of of feedings per 24 h was calculated as: day feedings 1 194.5 kcal per day in infant intake. The sex difference in night feedings. The average number of minutes spent milk composition is sufficient to account for differences in TABLE 1. Characteristics of mothers and infants Milk energy density among 25 Massachusetts mothers.
of prior animal studies in rhesus macaques and red deer (Hinde, 2007; Landete-Castillejo et al., 2004).
Increased demand for milk from male infants could fea- sibly result in an increase in the caloric density of their mothers’ breast milk. It has previously been suggested that male infants may consume 8–10% more milk than females (Michaelsen et al., 1994). Consistent with this, male infants (mean 5 9.27 times per day) in our study on average fed more times per day than female infants (mean 5 7.95 times per day), although this result did not reach statistical significance. Fat content is inversely correlated to the fullness of the breast (Kent et al., 2006). Therefore, at a given feeding frequency, an infant who extracts more volume at each feeding will drive the mean fat content of his mother’s milk up. This happens because the infant consumes more of the high fat hind milk at the current feeding and because the breast will not be as full at the subsequent feeding. It has previously been shown that increased feeding from a single breast, initiated by the mother, can increase the mean fat content per feeding, but this maternal behavior change could not increase totalinfant caloric intake, most likely because total intake is growth between male and female infants, with male controlled by infant demand (Woolridge et al., 1990).
infants gaining an average of 3.72 kg from birth to the Taken alone, the increased milk consumption of male time of study and female infants gaining 2.96 kg from infants and milk composition dependence on the fullness birth to the time of study (P 5 0.49), despite female of the breast may not explain a difference in average milk infants being slightly older than males at the time of caloric content, because prolactin levels are responsive to study (Table 3). There was not a statistically significant suckling stimulus and will serve to match milk supply to correlation between milk energy density and absolute However, faster growing infants logically demand ever Each unit increase in maternal bra cup size over preg- increasing milk production from their mothers by empty- nancy up until the time of study was associated with a ing a progressively increasing fraction of the milk 16.17 kcal/100 ml greater milk energy density (P 5 0.009).
contained in their mother’s breast at each feeding. While The effect size increased after adjustment for time since the mother’s production would increase to match demand, last feeding and the association remained statistically sig- the infant demand would remain ahead of the milk pro- nificant (P 5 0.029). The association between change in duction. Thus, the increased caloric content of milk pro- maternal cup size and milk energy density was not signifi- duced by mothers of male infants may be caused by a cant after taking infant sex into account. This was likely greater rate of increase of energy requirement in male related to mothers of male infants having a larger average infants. This is consistent with the greater weight gain increase in cup size (1.35 vs. 1.00), although this result from birth observed among male infants in this study.
did not reach statistical significance (P 5 0.246).
To our knowledge, this also is the first report to demonstrate a link between breast enlargement duringpregnancy and milk energy content. The growth of breast tissue may more accurately represent the relative mass This is the first report, to our knowledge, of an infant of structures devoted to milk production and storage.
sex-based difference in breast milk energy content in Consistent with this, Neifert et al. (1990) showed that humans. Our findings, however, are consistent with those females with less breast enlargement during pregnancy INFANT SEX PREDICTS BREAST MILK ENERGY CONTENT TABLE 2. Associations between subject characteristics TABLE 3. Infant and milk characteristics by infant sex No adjustment for time since last feeding.
this study include the small sample size and the propor- are more likely to have lactation insufficiency, as meas- tion of data collected by self-report. Despite this, the ured by infant weight gain. Hytten (1995) suggests that results are intriguing and should be followed with further the lower levels of pregnancy breast enlargement in older study. A larger study with prospective data collection mothers are associated with their decreased milk produc- should be conducted to verify our results. Despite limita- tion; however, this relationship was not examined tions, the results of our study raise several other possibil- directly. Our study advances the results of these prior ities for future research. It remains to be seen whether studies by providing evidence for a direct association the association between infant sex and milk energy between breast enlargement during pregnancy and milk content will hold up in less energy rich populations more similar to the environment in which humans evolved. A The objective of this study was not to determine the study conducted in a society with scarce energy resources mechanism by which milk composition varies among may provide insight into whether the sex-based difference women, thus we cannot rule out alternative mechanisms by which the sex based difference in milk compositioncomes about, including fetal sex-based differences in hor-mones secreted by the placenta during pregnancy, which may contribute to breast glandular development. This is For assistance with the statistical analyses we thank supported by our data in that mothers of male infants had Louise Ryan. For comments on the manuscript we thank greater cup size changes than the mothers of female Neil Powe and Melissa Emery Thompson. For support infants in our study, although this did not reach statistical with recruitment of subjects we thank Harriet Torgerson significance. Once infant sex was taken into account, the of HUHS and Anna Blair of the Center for Breastfeeding, association between maternal cup size change and milk energy density was no longer statistically significant, rais-ing the possibility that change in cup size partially medi-ates the relationship between infant sex and milk energy density. Interestingly a study of 244 Boston women foundthat pregnant women carrying male infants consume 10% Allen JC, Keller PR, Archer P, Neville MC. 1990. Studies in human lacta- tion: milk composition and daily secretion rates of macronutrients in the more calories than pregnant women carrying female first year of lactation. Am J Clin Nutr 54:69–80.
infants, raising the additional possibility that male off- AOAC (Association of Official Analytical Chemists). 1984. Official Methods spring place greater demands on their mothers even in of Analysis of the Association of Official Analytical Chemists. Arlington, utero (Tamimi et al., 2003). These greater demands could VA: Association of Official Analytical Chemists. p 159–160.
Barbosa L, Butte NF, Vollalpando S, Wong WW, Smith EO. 1997. Maternal contribute to the greater, though nonstatistically signifi- energy balance and lactation performance of Mesoamericans as a func- cant, cup size change seen in the mothers of male infants.
tion of body mass index. Am J Clin Nutr 66:575–583.
Of note, the sex-based difference in milk energy content Brown KH, Akhtar NA, Robertson AD, Ahmed MG. 1986. Lactational is consistent with the Trivers-Willard hypothesis for sex- capacity of marginally nourished mothers: relationships between mater-nal nutritional status and quantity and proximate composition of milk.
specific parental investment (Trivers and Willard, 1973).
This hypothesis predicts that when male reproductive Daly SE, Rosso AD, Owens RA, Hartmann PE. 1993. Degree of breast emp- success is more dependent on parental investment than tying explains changes in the fat content, but not fatty acid composition, female reproductive success, mothers will tend to invest of human milk. Exp Physiol 78:741–755.
Dewey KG, Lovelady CA, Nommsen-Rivers LA, McCrory MA, Lonnerdal in males and males will be willing to expend more effort to B. 1994. A randomized study of the effects of aerobic exercise by lactat- obtain parental investment. Well-nourished mothers may ing women on breast milk volume and composition. N Engl J Med have been selected to invest more energy in their male infants than female infants or male infants may be more Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. 1956. Colorimetric method for determination of sugars and related substances. Anal Chem aggressive than females in procuring energy from their Gardner D, Shoback D. 2007. Greenspan’s basic & clinical endocrinology, In summary, this study of 25 exclusively breastfeeding mothers is the first report of an association between milk Hinde K. 2007. First-time macaque mothers bias milk composition in favor energy density and both infant sex and maternal breast Hytten FE. 1995. The clinical physiology of the puerperium. London: Far- size change during pregnancy in humans. Limitations of Jackson DA, Imong SM, Silprasert A, Ruckphaopunt S, Wooridge MW, Perez-Escamilla R, Cohen RJ, Brown KH, Rivera LL, Canahuati J, Dewey Baum JD, Amatayakul A. 1988. Circadian variation in fat concentration KG. 1995. Maternal anthropometric status and lactation performance in of breastmilk in a rural Northern Thai population. Br J Nutr 59:349– a low-income Honduran population: evidence for the role of infants. Am Kent JC, Mitoulas LR, Cregan MD, Ramsay DT, Doherty DA, Hartmann Pierce WC, Haenisch EL. 1947. Quantitative Analysis. New York: Wiley.
PE. 2006. Volume and frequency of breastfeedings and fat content of Pon, 2000. Age Calculator. Available online http://www-users.med.cornell.
breast milk throughout the day. Pediatrics 117:387–395.
edu/spon/picu/calc/agecalc.htm. Accessed March 23rd, 2006.
Landete-Castillejo T, Garcia A, Lopez-Serrano FR, Gallego L. 2004. Mater- Prentice AM, Prentice A, Whitehead RG. 1981a. Breast-milk fat concentra- nal quality and differences in milk production and composition for male tion of rural African women. I. Short-term variations within individuals.
and female Iberian red deer calves. Behav Ecol Sociobiol 57:749–799.
Michaelsen KF, Larsen PS, Thomsen BL, Samuelson G. 1994. The Copen- Prentice A, Prentice AM, Whitehead RG. 1981b. Breast-milk fat concentra- hagen Cohort Study on Infant Nutrition and Growth: breast milk intake, tions of rural African women. II. Long term variations within a commu- human milk macronutrient content, and influencing factors. Am J Clin Prentice AM, Roberts SB, Prentice A, Paul AA, Watkinson M, Watkinson Mitoulas LR, Kent JC, Cox DB, Owens RA, Sherriff JL, Hartmann PE.
AA, Whitehead RG. 1983. Dietary supplementation of lactating Gam- 2002. Variation in fat, lactose, and protein in human milk over 24 h and bian women. I. Effect on breast-milk volume and quality. Hum Nutr Clin throughout the first year of lactation. Br J Nutr 88:29–27.
Neifert M, DeMarzo S, Seacat J, Young D, Leff M, Orleans M. 1990. The Strickland JDH, Parsons TR. 1972. A practical handbook of sea-water influence of breast surgery, breast appearance, and pregnancy-induced analysis. J Fish Res Bd Canada 167:311.
breast changes on lactation sufficiency as measured by infant weight Tamimi RM, Lagiou P, Mucci LA, Hsieh CC, Adami HO, Trichopoulous D.
2003. Average energy intake among pregnant women carrying a boy Nommsen LA, Lovelady CA, Heinig MJ, Lonnerdal B, Dewey KG. 1991.
compared with a girl. Br Med J 326:1245–1246.
Determinants of energy, protein, lipid, and lactose concentrations in human Trivers RL, Willard DE. 1973. Natural selection of parental ability to vary milk during the first 12 mo of lactation: The DARLING study. Am J Clin the sex ratio of offspring. Science 179:90–92.
Woolridge MW, Ingram JC, Baum JD. 1990. Do changes in pattern NRC (National Resource Council). 1980. Recommended Dietary Allow- of breast usage alter the baby’s nutrient intake? Lancet 336: ances. Washington, DC: National Academy of Sciences, 1980.

Source: http://www.stiftungstillen.ch/logicio/client/stillen/file/stillen/Fachpersonal/Powe_2010.pdf