Abstract
Changes to the maternal inflammatory milieu may be a mechanism through which maternal psychosocial stress is transmitted to the fetus. Research investigating a limited number of immune markers may miss important signals. We take a proteomics approach to investigate maternal lifetime stress and 92 biomarkers of immune system status. Participants were enrolled in an urban, dual-site (Boston, n = 301 and New York City, n = 110) pregnancy cohort. We measured maternal lifetime history of stress and trauma using the validated Life Stressor Checklist-Revised (LSC-R). We measured a panel of 92 immune-related proteins in mid-pregnancy serum using proximity extension assay technology. We leveraged the dual-site study design to perform variable selection and inference within the cohort. First, we used LASSO to select immune markers related to maternal stress among Boston mothers. Then, we performed OLS regression to examine associations between maternal stress and LASSO-selected proteins among New York City mothers. LASSO regression selected 19 immune proteins with non-null coefficients (CCL11, CCL23, CD244, CST5, CXCL1, CXCL5, CXCL10, CX3CL1, FGF-23, IL-5, IL-7, IL-10, IL-17C, MCP-2, MMP-1, SLAMF1, ST1A1, TNF-β, and TWEAK). Of these, only the chemotactic cytokine CX3CL1 (i.e. fractalkine) was significantly associated with maternal stress among the validation sample (percent change in LSC-R score per 1% increase in relative fractalkine expression: 0.74, 95% confidence interval: 0.19, 1.28). Expanding research suggests fractalkine plays an important role in many aspects of pregnancy and fetal development and is stress-sensitive. We found that maternal lifetime history of stress and trauma was significantly associated with elevated serum fractalkine levels during pregnancy.