Moderators, mediators, and components of a standalone smartphone application for postpartum depression: secondary analysis of a randomized controlled trial

PURPOSE: The purpose of this study was to identify potential moderators, mediators, and efficacious components of a standalone app intervention on postpartum depression symptoms and functional impairment. METHODS: A total of 264 women were randomized to intervention (Motherly app) or to an active control group. We analyzed depressive symptoms assessed by the Edinburgh Postnatal Depression Scale and functional impairment by the Clinical Global Impression at post-treatment as outcomes. Various potential moderators were assessed at baseline. Potential mediation variables were the level of behavioral activation (BA) measured by the Behavioral Activation for Depression Scale - Short Form and the Response-contingent positive reinforcement (RCPR) by the Reward Probability Index, at post-treatment. Intervention component analysis was evaluated using app engagement data. RESULTS: The results indicate that intervention effects on depressive symptoms were moderated by the number of children and pediatric consultations, with stronger effects observed among mothers with fewer children and higher consultation frequency. Functional impairment was moderated by time spent with the child, with greater improvements among mothers with lower caregiving demands. Mediation analyses showed BA and RCPR mediated the effect of our intervention on depressive symptoms and functional impairment. Total and BA achievements were significantly associated with reduction in depressive symptoms. CONCLUSION: Our findings suggest that our app intervention efficacy is influenced by maternal characteristics and childcare responsibilities. BA and RCPR are key mechanisms, and higher engagement with app components produce greater symptom reduction.