Clinical phenotypes of epigenetic biomarker positive versus biomarker negative postpartum depression

Postpartum depression (PPD) can be predicted by biomarkers of 3rd trimester methylation at two estrogen-responsive genes (TTC9B, HP1BP3), but the mechanism(s) by which these epigenetic changes lead to PPD are unknown. We investigated whether clinical characteristics (psychiatric histories, depression onset, stress/anxiety/trauma) differ between those with (BM+) and without (BM-) these biomarkers of PPD risk. Participants (N = 206) completed clinical diagnostic interviews and stress, anxiety, and trauma questionnaires at up to 7 visits during pregnancy and postpartum. Biomarker status was determined using 3rd trimester blood samples. Linear mixed effects models compared outcomes between biomarker groups (positive/negative). Mediation models assessed the average causal mediation effect of anxiety on the relationship between biomarker status and PPD. BM+ individuals were more likely to 1) have a prior history of generalized anxiety (X² = 10.06, p = .003) and bipolar II disorder (X² = 7.68, p = .01), 2) develop PPD (X² = 3.97, p = .04), and 3) be depressed during both pregnancy and postpartum (X² = 6.52, p = .02). BM+ individuals scored higher on the perinatal anxiety screening scale (PASS, p = .006), perceived stress scale (PSS, p = .001), and state-trait anxiety inventory (STAI)-trait subscale (p = .02). Anxiety outcomes on the PASS (95% CI: 0.01-0.10, p = .004), PSS (95% CI: 0.007-0.10, p = .03), and STAI-State (95% CI: -0.001-0.06, p = .06) mediated the relationship between biomarker status and PPD development. Findings suggest distinct clinical phenotypes between BM+ and BM- PPD, with BM+ individuals displaying significantly higher anxiety/stress, which may contribute to their higher risk of developing PPD. This presents an opportunity to investigate mechanisms in more homogenous groups based on biomarker status.