Maternal characteristics are associated with human milk anti-inflammatory proteins in two populations.

Milk anti-inflammatory compounds are ubiquitous in milk but vary greatly within and between populations. The causes of this variation and how this variation impacts infant phenotype is not well-characterized. The goal of this study was to explain how maternal characteristics across two disparate populations impact the levels of TGF-β2 and IL-1ra in human milk. Two populations of mothers, one from rural Kenya and the other from urban U.S., were queried about months since birth, age, sex of infant, height, BMI, triceps skinfold, parity, post-birth resumption of menstrual period, and exclusive breastfeeding. Mothers' foremilk was assayed for TGF-β2 and IL-1ra as well as % milk fat. Mixed models were used to measure the relationships between maternal characteristics and milk biomarkers, adjusting for population. Statistically significant maternal characteristics were then used to develop path models incorporating infant phenotype. Path results indicated that maternal height and months postpartum significantly predicted milk TGF-β2, which then significantly predicted infant height-for-age. Exclusive breastfeeding and milk fat percent predicted IL-1ra, which was not related to infant weight-for-age. These results have implications for understanding the intergenerational effect of maternal context on infant phenotype via biomarkers in human milk.