Cindy-Lee Dennis, PhD


An RCT to Evaluate the Effect of Home-Based Peer Support on Maternal-Infant Interaction, Infant Health Outcomes, and Postpartum Depression

Hypotheses: The primary hypothesis of this study was that home-based peer support would improve maternal-infant interactions. Secondary hypotheses included that home-based peer support would: improve infants’ cognitive development; improve infants’ social development; decrease average daily salivary cortisol levels in infants; reduce maternal depressive symptomatology; and improve maternal perceptions of social support.

Methods: This randomized controlled trial of a treatment approach to PPD attempted to enrol 104 mothers and their infants (determined by power analysis) in Alberta and New Brunswick who screen above 12 on the Edinburgh Postnatal Depression Scale. Mother-infant dyads were randomly assigned to either the intervention or control group. Each intervention group mother was matched with an experienced peer who has recovered from PPD. Peers provided 12 weekly home visits of approximately one hour duration to provide informational, affirmational, emotional, and instrumental support and promote optimal maternal-infant interactions. Paid professionals (e.g. nurses, social workers) were resources for the peers. Control group mothers received an abbreviated form of the intervention at the end of a 12-week waiting period.

Measures: Measures assessed for their reliability, validity, and feasibility in previous studies by the investigators were administered at study enrolment, 6 weeks post-randomization, and 12 weeks post-randomization. Outcome measures assessed maternal-infant interaction, infant cognitive development, infant social development, infant salivary cortisol levels, and maternal depressive symptomatology. In addition, peers documented activities in logs and field notes were gathered from professionals.

Analysis: In addition to basic descriptive statistics, three multivariate analysis of variance (MANOVA) tests were used to examine the primary and secondary hypotheses. With the independent variables of group and time, and groupings of the dependent variables related to maternal-infant interaction (teaching and feeding interactions, parenting style), infants (cognitive and social development, salivary cortisol) and mothers (maternal depressive symptomatology, support satisfaction) were used to test for the study’s overall intervention, time, or interaction effect. Thematic content analysis was done on the textual data.

Results: Contrary to the hypothesized direction of relationships, results favored the control group. A significant difference between groups was observed on one of the two measures of maternal-infant interactions. Several other measures favored the control group including mothers’ depressive symptoms and social support scores. No significant treatment effects were observed in infant IQ scores or diurnal salivary cortisol levels. Discussion: Findings suggest that a base level of professional support for postpartum depression (e.g. three home visits and at least three phone calls from professional nurses) may be optimal in reducing depressive symptoms and promoting maternal-child interactions. Moreover, maternal-infant interaction teaching may be more optimally delivered by professional nurses.

< Return to Randomized Controlled Trial Research Activities