Breast milk proteins provide both nutritional and physiological benefits to infants, and longitudinal, circadian, and individual variations have been reported to influence the breast milk protein concentration (1–6).
Although there are several studies on longitudinal changes in total protein concentration of breast milk, information on the circadian variations, in particular variation in the 2 main classes of breast milk proteins whey and casein, is limited. Furthermore, to our knowledge, there have been no short-term studies (within the day) investigating the variability in both whey and casein composition.
It is important to study the short-term variations in individual protein concentrations because this information has implications for appropriate sample collection of these constituents for future nutritional studies. This information may also help in understanding the control of the milk intake mechanism in infants because protein composition has been reported as an essential factor contributing to gastric emptying. Casein and whey proteins behave differently during digestion in the stomach and have different emptying rates (7), and therefore could contribute to variations in infant feeding patterns.
Previous studies have shown that not only there are changes in the total protein concentration but also the proportion of casein and whey proteins vary during the course of lactation in both term (8) and preterm (9) human milk. Although total protein concentration decreases as lactation progresses, casein concentration, which is almost undetectable during the first days of lactation, increases, with a simultaneous decrease in the concentration of total whey proteins. As a result of these changes, the ratio of casein to whey proteins in breast milk has been found to vary throughout lactation (ie, from initiation to up to 12 months of lactation) and reported to be low in early lactation (10:90), to increase in mature milk (40:60), and to level out at 50:50 in late lactation (8). Because it is known that there are longitudinal changes in the relative concentration of casein and whey proteins during lactation, further investigation is required to understand whether the casein:whey ratio varies over the short term, that is, throughout the day.
The present study was designed to investigate the variation in protein concentration of breast milk casein and whey fractions in mothers of term infants. More specifically, we aimed to investigate whether the protein concentrations differed between fore and hind milk and between breasts of mothers during 4 time points within a 24-hour period, as well as investigate the extent of variation between mothers.
Breast milk samples were obtained from 25 mothers of healthy, term singleton infants recruited through the Western Australian branch of the Australian Breastfeeding Association, and through Child and Adolescent Community Health Nurses in the Oceanic Health Region. The infants (n = 25) had a mean age of 3.6 ± 1.9 months (range 1.0–8.0 months) and were mostly boys (n = 18). All of the infants (n = 25) were exclusively breast-fed on demand except for 2 infants older than 6 months who were receiving complementary solid foods. The infants were growing appropriately for age. Briefly, mothers were between 25 and 39 years old (mean 33 ± 4.0 years, n = 4, age unknown), and the majority of them were primiparous (n = 16). Seventeen mothers (mean age 33 ± 3 years) agreed to provide the samples for the analysis of protein concentration in fore and hind milk samples (study component A, mean age of infants 3.7 ± 2.1, range 1–8 months, percentage of male infants 82%). Fifteen mothers (mean age 31 ± 4 years), including a subgroup of 7 mothers from study component A, provided the samples with sufficient volume for the analysis of circadian variation in protein concentration during a 24-hour breast-feeding period (study component B, mean age of infants 3.2 ± 1.7, range 1–6 months, percentage of male infants 67%). Mothers were also asked to measure their 24-hour milk productions.
All of the mothers provided written informed consent to participate in the study, which was approved by the University of Western Australia human research ethics committee.
Study Component A
Mothers (n = 17) performed a 15-minute simultaneous expression of left and right breasts using 2 separate electric breast pumps (Symphony, Medela AG, Baar, Switzerland). This was done under supervision and milk samples (1–2 mL) were collected at the beginning (fore) and end (hind) of the expression session by the supervising researcher. Total volume of milk expressed during the session was recorded. Subjects who participated in this part of the study attended our experimental rooms at the Breastfeeding Centre at King Edward Memorial Hospital, Subiaco, Western Australia.
Study Component B
Mothers (n = 15) collected the milk samples (1–2 mL) at their homes before and after each feed from each breast during a 24-hour period by either hand expression or manual hand pump into 5-mL polypropylene vials (Disposable Products Pty Ltd, Adelaide, Australia). For the present study, the day was divided into 4 intervals of 6 hours and named as morning (4:01 AM–10:01 AM), day (10:01 AM–4:00 PM), evening (4:01 PM–10:00 PM), and night (10:01 PM–4:00 AM) as described previously by Kent et al (10).
The samples collected before and after the feed were pooled. Then milk samples from the left and right breasts that were closest to the centre of these time points were selected for the analysis of the concentration of proteins. Samples were initially stored in the mother's home freezer and then transported on ice to the laboratory, where they were kept at −20° C until further analysis.
24-hour Milk Production Measurements
Mothers measured their 24-hour milk productions at their homes by test weighing the infants before and after each feed, for each breast, using electronic scales (Baby Weigh Scale, Medela AG) and recorded the volume of milk intake of each feed from each breast (11).
Isolation of Casein and Whey
The milk samples were thawed and, after mixing, defatted by centrifugation at 10,000 g for 10 minutes at 4° C for the analysis of concentration of proteins. Casein and whey proteins were separated by the method described by Kunz and Lonnerdal (12,13). In short, casein and whey were separated from the defatted (skim) milk by the addition of calcium chloride to the skim milk; pH was adjusted to 4.3 with 1 mol/L HCl followed by ultracentrifugation at 189,000 g for 90 minutes at 4° C. The clear supernatant containing whey was separated, and the casein pellet was washed with double deionised water and centrifuged again at 189,000 g for 60 minutes. The supernatant wash was collected, and the casein pellet was dissolved in surfactant solution of 6 mol/L urea and 4% 3-((3-cholamidopropyl) dimethylammonio)-1-propanesulfonic acid. The samples were stored in −80°C freezer until analysed.
The separated fractions were subjected to electrophoresis in 15% sodium dodecyl sulfate-polyacrylamide gel (14). In the electrophoresis, purified human milk β-casein, lactoferrin, and serum albumin (Sigma-Aldrich, Castle Hill, Australia) were included to control for sample purity. Protein bands were identified by comparing with those of purified human milk proteins and with protein standards of known molecular weights varying between 10 and 250 kDa (Precision Plus Protein Standards; Bio-Rad Laboratories, Richmond, CA).
The protein concentrations of skim, whey, casein, and casein wash fractions of the samples were determined by the Bradford protein assay using a commercial protein reagent (Bio-Rad). Human milk protein standards were prepared by determining the concentration of an aliquot of mature breast milk as described by Atwood and Hartmann (15). The protein assays were carried out by the procedure described by Mitoulas et al (2). Skim milk and whey samples were diluted 1 in 30; for casein samples, a 1 in 10 dilution was used. The protein concentration of casein samples was measured by comparing with protein standards that were prepared in the same surfactant solution used to dissolve the casein pellet to avoid any background interference. For all of the assays, a quality control sample was analysed, along with all of the samples. The recovery of a known amount of the protein added to the milk samples was 99.8% (standard error 1.4, n = 12). The detection limit of the assay was 0.045 g/L (standard error 0.002, n = 30), and the interassay coefficient of variation was 6.4% (n = 50).
Statistical analysis was performed using the R program, version 2.7.2 (R Foundation, Vienna, Austria). The package NLME was used for linear mixed-effects modelling. For the linear mixed models, different intercepts for each individual were used as the random effect in all of the models, unless otherwise specified. Descriptive statistics are reported as mean ± standard deviation, unless otherwise stated. P < 0.05 was considered statistically significant.
Fore and Hind Milk
Protein concentrations of fore and hind milk samples were compared using univariate linear mixed-effect models that tested the calculated difference against zero (intercept-only model). Expression volume was tested as univariate predictor.
Linear mixed-effects models were used to compare the circadian variation in protein concentrations by considering time of day as a factor. Age of infant and feed volume were tested as univariate predictors, and where significant relations were seen, they were included as covariates in the time point model for that protein fraction. Models presented include only significant covariates. Infant age was either included as linear predictor, or stratified into younger (2 months or younger) and older (3 months or older).
Comparison of Breasts
Comparison of average feed volumes and 24-hour milk intakes between breasts used the paired Student t test. Individual differences in baseline protein concentrations between the breasts were tested by including breast as a random effect in circadian variation models.
The influence of mechanical expression on protein composition was assessed in a subgroup of 7 mothers who participated in both components of the study, using coefficient of variance as a measure of magnitude.
Mothers (n = 25) participating in all of the components of the present study completed the 24-hour milk production to measure the feeding patterns of the infants and had no concerns about their lactation.
The mean 24-hour breast milk intake was 758 ± 162 g (range 467–1113 g), and there was no significant difference between the 24-hour milk intake from the left (368 ± 133 g, range 133–665 g) and right breasts (372 ± 128 g, range 85–598 g). Breast-feeding frequency (a continuous sucking session from 1 breast (10)) ranged from 6 to 21 feeds/24 hours (mean 14 ± 3). The average breast-feed volume was not significantly different between the left (62 ± 22 g) and right (60 ± 22 g) breasts. The milk intake patterns were within the range reported previously for these measurements (10).
Separation of Casein and Whey
Figure 1 depicts the skim milk, separated whey, and casein fractions. Although it is known to be difficult to completely separate the whey and casein proteins (9), it can be seen that the casein band was not predominantly present in the separated whey and vice versa. Therefore, the acid-soluble fraction was defined as whey and acid-insoluble fraction as casein.
Protein Concentration in Fore and Hind Milk
Study component A examined the protein concentrations in fore and hind milk from a single breast expression for 17 mothers. These data are presented in Figure 2. Overall, the mean protein concentrations were 12.85 ± 2.2 for skim milk, 7.06 ± 1.85 for whey, and 2.99 ± 0.73 g/L for casein. The concentration of protein was not significantly different between the fore and hind milk samples in the skim (mean difference −0.12 g/L, P = 0.78), whey (mean difference −0.06 g/L, P = 0.89), and casein (mean difference −0.09 g/L, P = 0.23) fractions.
A negligible amount of protein (0.52 ± 0.19 g/L) was present in the casein wash, which was found to be predominately casein based on gel analysis (data not shown). The difference between the fore and hind milk casein wash values (mean difference 0.01 g/L, P = 0.82) was also not significantly different. Therefore, for analysis purposes, the casein and casein wash values were summed together. In addition, the total volume of milk expressed was not found to affect the protein concentration of any fraction.
Circadian and Interindividual Variation in Protein Concentration
Study component B analysed the protein concentration at the 4 time periods (morning, day, evening, night) within 24 hours for 15 mothers. These data are presented in Figure 3. The average protein concentrations in this dataset for all of the mothers were found to be 13.55 ± 2.1, 7.6 ± 1.5, and 3.4 ± 0.97 g/L for skim milk, whey, and casein, respectively.
The total protein concentration in skim milk samples was similar at all time points in the day (P = 0.14). There was a nonsignificant (P = 0.055) trend toward a higher protein concentration in the evening samples, with 0.67 g/L (5%) more protein measured when compared with morning. Neither whey protein concentration (P = 0.55) nor the casein protein concentration (P = 0.18) was found to be significantly affected by the time of day.
Although the average protein concentration for all of the mothers in each of the fractions remained constant throughout the day, there were marked variations observed between individual mothers. The protein concentration in skim milk, whey, and casein fraction ranged from 7 to 18.1 (coefficient of variance [CV] 15.5%), 4.45 to 11.35 (CV 19.8%), and 1.66 to 5.26 g/L (CV 28.4%), respectively, between mothers.
When compared with skim milk, the variation in whey and casein fractions was more pronounced both within and between mothers. Analysis of the CV of the protein concentration for samples from individual mothers showed that for the whey fraction, 8 mothers, and for the casein fraction, 6 mothers, had a CV >10% during the day; however, although variation >10% existed (11.7%–18.5% whey; 13.1%–26.8% casein), the pattern was not consistent with a circadian variation.
No relation was seen between the volume of milk consumed by the infant during the feed (feed volume) and the protein level in the skim (P = 0.83), whey (0.08), and casein fractions (0.09).
Protein concentration of the whey fraction was found to decrease significantly with the age of the baby (P = 0.03), whereas the effect was not significant for the skim milk and casein fractions. In addition, there was no relation between infant age (either in months, or stratified into younger and older infants) and protein concentration of any fraction at any time point of the day.
There was a significant relation between the protein content of skim milk to whey and casein fraction (P < 0.05). Higher protein concentration of skim milk was associated with higher concentrations of both whey and casein fraction.
Protein Concentration Between Breasts
There was no overall difference in the protein concentration of skim (P = 0.09), whey (P = 0.054), and casein fractions (P = 0.14) between breasts (Fig. 4). Similarly, no consistent differences were detected in the protein levels between breasts across selected time points of the day; however, individual variations were found in 3 mothers who had consistently different whey protein concentrations between their breasts.
When mean protein composition of samples obtained from breast expression was compared with the hand-expressed breast-feed samples in a subgroup of 7 mothers, the variability was <10%.
The present study measured the short-term changes in protein composition of term human milk. Within mothers, the protein concentration of skim, whey, and casein fractions during a 24-hour period did not vary between fore and hind milk samples or the left and right breasts (Figs. 2 and 4). There was, however, considerable variation between mothers.
Our results of no significant change in fore and hind milk total protein concentration (Fig. 2) are in agreement with other previous studies carried out both in women (2,18) and in other species (15). Furthermore, no circadian change was found in protein concentration of any fraction, with no difference between morning, day, evening, and night feed samples (Fig. 3). This finding supports the earlier work of Mitoulas et al (2); however, Lammi-Keefe et al (5) found a significant time-of-day effect on protein nitrogen. The change observed in the study by Lammi-Keefe et al could be associated with the change in fat concentration during the day because it appears that a whole-milk sample was used for protein analysis.
During lactation, protein composition changes to meet the growing requirements of the infant (19). A decrease in whey protein concentration is balanced by a subsequent increase in casein concentration (8). Changes in the concentration of specific whey proteins have also been reported. The protective components of whey proteins (lactoferrin or secretory immunoglobulin A) are higher in early lactation and subsequently decrease in concentration, whereas the nutritional α-lactalbumin whey protein increases during the course of lactation (20); however, we did not find any short-term variations in the relative proportion of casein and whey proteins.
Our finding of no consistent change in protein concentration between the breasts is supported by previous studies (2,21) (Fig. 4); however, we did observe consistent differences in protein concentration, particularly in the whey fraction of some individuals. These mothers (n = 3) had a consistently higher protein concentration (approximately 1.5 g/L) in one breast over the other for all of the samples collected during a 24-hour period. Of these individuals, 2 had asymmetrical productions, with 1 breast producing approximately 40% of the milk production of the other. These were older mothers (ages 33 and 37 years). The third individual did not show a difference in production between breasts, and her age was unknown. Neville et al (21) has observed the inconsistent differences in the milk composition of micronutrients between left and right breasts and suggested that mastitis may contribute to these differences. The mothers participating in the present study were healthy and without symptoms of mastitis. Therefore, the factors behind these differences are not clear; however, there is a possibility that the variations could be caused by differences in milk production between breasts. The variations have previously been reported in the protein composition of mammary lobes in the breast, and it was suggested that the feedback inhibitor of lactation individually regulates the protein synthesis in each mammary lobe (22). This could result in variation of protein content. Thus, the possibility exists that variation observed in the present study in addition to milk production may also be caused by this proposed mechanism, but this requires further investigation.
The variations across the day in some of the individuals were not found to be indicative of circadian variation. In addition, individual variation in protein concentrations across the day was less than variation between mothers. These results highlighted the marked variability present in milk composition between mothers, which is not associated with maternal age, and are consistent with other studies (4,23,24).
Individual proteins have greater variability than the total protein, which supports earlier reports of greater variability in individual proteins both within and between mothers (25). Although we did not measure individual subunits of casein, the greater variability in the casein fraction could be related to the presence of these subunits. The individual variations in whey protein could be the result of active secretion of certain immunoglobulins by the mammary gland in some individuals (26); however, they were also associated with infant age. Our findings of casein and whey having large variation between mothers support the idea that each woman has a characteristic breast milk protein profile.
The mean casein and whey protein concentrations in the present study is similar to those reported by others (8,27) during lactation; however, the mean total protein concentration in the present study is slightly higher than the reported values of 10 to 12 g/L (18). Furthermore, the variability of <10% between the 2 methods of milk removal suggests that the method of milk expression did not alter the protein composition.
Because breast milk composition may be affected by several factors, including environmental conditions, it is important to investigate the variability in different protein fractions of breast milk during a feed, between breasts and during the course of the day to determine the appropriate sampling protocol for these components. The limitation of the present study includes the lack of determination of the effect of longitudinal changes on protein concentration; however, this area has already been studied (2,8). Moreover, further studies are required to understand the effect of infant-related factors (eg, infant body weight) on milk composition.
In conclusion, no short-term changes in the concentration of whey and casein were found. This establishes that it will not be necessary to account for these short-term factors in nutritional studies. Therefore, for sampling purposes, a single breast milk sample (fore or hind milk at any time point of the day) can be used to estimate a protein concentration of skim, whey, and casein fractions for a mother for that day. Although there was no significant difference between breasts, consistent differences between breasts for 3 individuals indicate that milk sample from 1 breast cannot be reliably considered to be representative of both breasts. It is therefore recommended that samples should still be taken from both breasts of an individual mother. In addition, longitudinal changes in protein concentration should also be taken into account.
The protein concentration of skim milk, whey, and casein varies greatly between mothers. Therefore, it is possible that variation in these protein concentrations between mothers may affect milk intake in infants during a 24-hour period and hence be responsible for the feed variation observed between exclusively breast-fed infants (10). This is an area that requires further investigation.
1. Kunz C, Rodriguez-Palmero M, Koletzko B, et al. Nutritional and biochemical properties of human milk, part I: general aspects, proteins, and carbohydrates. Clin Perinatol
2. Mitoulas LR, Kent JC, Cox DB, et al. Variation in fat, lactose and protein in human milk over 24 h and throughout the first year of lactation. Br J Nutr
3. Allen JC, Keller RP, Archer P, et al. Studies in human lactation: milk composition and daily secretion rates of macronutrients in the first year of lactation. Am J Clin Nutr
4. Clark RM, Ross SA, Hill DW, et al. Within-day variation of taurine and other nitrogen substances in human milk. J Dairy Sci
5. Lammi-Keefe CJ, Ferris AM, Jensen RG. Changes in human milk at 0600, 1000, 1400, 1800, and 2200 h. J Pediatr Gastroenterol Nutr
6. Michaelsen KF, Skafte L, Badsberg JH, et al. Variation in macronutrients in human bank milk: influencing factors and implications for human milk banking. J Pediatr Gastroenterol Nutr
7. Khoshoo V, Brown S. Gastric emptying of two whey-based formulas of different energy density and its clinical implication in children with volume intolerance. Eur J Clin Nutr
8. Kunz C, Lonnerdal B. Re-evaluation of the whey protein/casein ratio of human milk. Acta Paediatr
9. Sanchez-Hidalgo VM, Flores-Huerta S, Matute G, et al. Whey protein/casein ratio and nonprotein nitrogen in preterm human milk during the first 10 days postpartum. J Pediatr Gastroenterol Nutr
10. Kent JC, Mitoulas LR, Cregan MD, et al. Volume and frequency of breastfeedings and fat content of breast milk throughout the day. Pediatrics
11. Arthur PG, Hartmann PE, Smith M. Measurement of the milk intake of breast-fed infants. J Pediatr Gastroenterol Nutr
12. Kunz C, Lonnerdal B. Human milk proteins: separation of whey proteins and their analysis by polyacrylamide gel electrophoresis, fast protein liquid chromatography (FPLC) gel filtration, and anion-exchange chromatography. Am J Clin Nutr
13. Kunz C, Lonnerdal B. Human-milk proteins: analysis of casein and casein subunits by anion-exchange chromatography, gel electrophoresis, and specific staining methods. Am J Clin Nutr
14. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature
15. Atwood CS, Hartmann PE. Collection of fore and hind milk from the sow and the changes in milk composition during suckling. J Dairy Res
16. Deleted in proof.
17. Deleted in proof.
18. Saarela T, Kokkonen J, Koivisto M. Macronutrient and energy contents of human milk fractions during the first six months of lactation. Acta Paediatr
19. Lonnerdal B. Nutritional and physiologic significance of human milk proteins. Am J Clin Nutr
20. Cuilliére ML, Abbadi M, Molé C, et al. Microparticle-enhanced nephelometric immunoassay of alpha-lactalbumin in human milk. J Immunoassay
21. Neville MC, Keller RP, Seacat J, et al. Studies on human lactation. I. Within-feed and between-breast variation in selected components of human milk. Am J Clin Nutr
22. Murase M, Mizuno K, Nishida Y, et al. Comparison of creamatocrit and protein concentration in each mammary lobe of the same breast: does the milk composition of each mammary lobe differ in the same breast? Breastfeed Med
23. Weber A, Loui A, Jochum F, et al. Breast milk from mothers of very low birthweight infants: variability in fat and protein content. Acta Paediatr
24. Bauer J, Gerss J. Longitudinal analysis of macronutrients and minerals in human milk produced by mothers of preterm infants. Clin Nutr
25. Goldfarb MF, Savadove MS, Inman JA. Two-dimensional electrophoretic analysis of human milk proteins. Electrophoresis
26. Peitersen B, Bohn L, Andersen H. Quantitative determination of immunoglobulins, lysozyme, and certain electrolytes in breast milk during the entire period of lactation, during a 24-hour period, and in milk from the individual mammary gland. Acta Paediatr Scand
27. Velonà T. Protein profiles in breast milk from mothers delivering term and preterm babies. Pediatr Res