Colostrum and Mature Human Milk of Women from London, Moscow, and Verona: Determinants of Immune Composition

Abstract

Cytokines and growth factors in colostrum and mature milk may play an important role in infant immune maturation, and may vary significantly between populations. We aimed to examine associations between environmental and maternal factors, and human milk (HM) cytokine and growth factor levels. We recruited 398 pregnant/lactating women in the United Kingdom, Russia, and Italy. Participants underwent skin prick testing, questionnaire interview, and colostrum and mature milk sampling. HM cytokine and growth factor levels were quantified by electro-chemiluminescence. We found significant geographical variation in growth factor levels, but no evidence of variation between sites in cytokine detectability. There was an inverse correlation between time of milk sampling and growth factor levels in colostrum for Hepatocyte Growth Factor (HGF) and TGFβ1 and TGFβ3, but not TGFβ2, and levels were significantly higher in colostrum than mature milk for all growth factors. The kinetics of decline were different for each growth factor. Cytokines were present at much lower levels than growth factors, and the decline over time was less consistent. HM growth factors and cytokine levels vary between populations for unknown reasons. Levels of HM mediators decline at different rates postpartum, and these findings suggest specific biological roles for HM growth factors and cytokines in early postnatal development.

Keywords: colostrum; cytokines; environmental influence; growth factors; human milk; immune modulators; immunologically active molecules.

Figure 1

Unadjusted growth factors concentration (pg/mL) in colostrum and breast milk across all sites (overall) and at each site of collection (London, Moscow, and Verona). ** p value < 0.01.

Figure 2

Association between site of collection and growth factor concentrations in colostrum and breast milk. Data shown are mean (bold line) and 95% CI (shaded area) for log transformed concentrations of HGF (a); TGFβ1 (b); TGFβ2 (c); and TGFβ3 (d) in the UK (red), Russia (blue), and Italy (green). A multilevel mixed-effect regression model was used for all analyses which were adjusted to the following factors: Parity, Maternal Atopy, Maternal age, Site (Country) of collection, Mode of delivery, Mould presence at home, Pets at home or regular contact; Exposure to tobacco smoke, at recruitment; At least one self-reported maternal infection during pregnancy; Maternal diet.

Figure 3

Effect of correcting growth factor concentrations for total protein level on the relationship between concentration and time. Data show nonlinear curves for unadjusted concentrations (pg/mL) of HGF (blue), TGFβ1 (orange), TGFβ2 (green), and TGFβ3 (pink) in colostrum in relation to time of sampling. Correlations were explored using raw data (a); and the ratio of growth factor concentration to protein concentration in the same sample (b). Trends in absolute and relative growth factors concentration in colostrum change over time. All correlation coefficients were statistically significant apart from the TGFβ3/Protein ratio in Figure 3b.