Abstract
The first postpartum year is a vulnerable period that leads to increased maternal stress. In this study, maternal stress, as measured by perceived stress scale (PSS) scores and human milk (HM) cortisol concentrations, was systematically compared with infant behavior, HM composition of immunomodulators (interleukin [IL]-6, IL-8, IL-10, IL-1β, tumor necrosis factor [TNF]-α), and growth factors (transforming growth factor beta [TGFβ]-2, and epidermal growth factor [EGF]), as well as markers of mammary tight junction and local inflammation within the breast (sodium-to-potassium [Na:K] ratio and somatic cell count [SCC], respectively) over five data collections in 18 days. There were no correlations between PSS and HM cortisol concentrations. Inconsistent correlations were seen with stress markers, infant behavior, inflammation, growth factors, and mammary epithelial function across the timepoints. On collection day five (C5), Mothers with cortisol higher than the median (0.69 ng/ml) reported more total infant crying time (U=54.00, p<0.01), but no other infant behavior variables were correlated with PSS or HM cortisol concentrations at other collections. On collection day 1 (C1), PSS was positively correlated with IL-6 (rs = -0.45), IL-10 (rs = 0.41), and TNF-α (rs=-0.41; p<0.05). Mothers who scored in the moderate PSS range had higher IL-6 (U=49.50, p = 0.035) than mothers in the low range on C1. The subjects with moderate HM cortisol had significantly higher TGFβ2 concentrations than the low HM cortisol group on C5 (U=22.5, p< 0.01). The SCC was positively correlated with IL-1β concentrations on C1 (rs=0.40, p<0.05) and with TGFβ2 on C5 (rs=0.72, p<0.01). Inconsistent correlations between stress markers with infant behavior, inflammatory cytokines, HM growth factors, and mammary epithelial function over different collections highlight the need for larger-scale studies on this subject.