Effects of a postpartum depression intervention: subgroup analyses from a cluster randomized trial

Limited research has examined participant subgroups enrolled in postpartum depression interventions. Via a cluster-randomized trial of the Mothers and Babies (MB) postpartum depression preventive intervention, we examined changes in depressive symptoms across pre-specified participant subgroups: race/ethnicity, first-time mother, language of intervention receipt, and client education. We also examined interaction effects between subgroups and study arm (usual care vs. intervention) on symptom trajectories.

A total of 629 pregnant women (mean weeks' gestation = 22.0 weeks, mean age 26.6 years) completed self-report assessments at baseline, post-intervention, and 12 and 24 weeks postpartum. The primary outcome was the Quick Inventory of Depressive Symptomatology (QIDS). Analyses involved linear mixed models fitted with baseline QIDS score, patient characteristics, study arm, and three-way interaction for QIDS scores at follow-up time points.

Minority participants exhibited larger mean QIDS decreases than non-minorities, with varying trajectories across study arm and minority subgroup (three-way interaction term p = 0.048). First-time mothers receiving MB also exhibited lower mean scores at 24 weeks postpartum than their multiparous counterparts over time (p=0.028). Trends in QIDS scores were comparable, although not significant, over time among women in the language (p = 0.929) and education subgroups (p = 0.635).

MB may have greater benefits when delivered to racial and ethnic minorities and first-time mothers, providing added guidance to practitioners planning to implement the intervention. Future research should also consider additional organizational and system-level variables that may be associated with varied intervention effects.