Meta-Analysis of the Effectiveness of Biological and Non-Biological Treatments for Postpartum Depression

• Main Author: Christian, Sarah Jeung soon
• Format: Others
• Published: BYU ScholarsArchive 2013
• Subjects: postpartum depression; treatment outcome; psychotherapy; pharmacology; exercise; meta-analysis; Psychology
• Online Access: https://scholarsarchive.byu.edu/etd/3571; https://scholarsarchive.byu.edu/cgi/viewcontent.cgi?article=4570&context=etd

I provided an updated, comprehensive review of treatments for mothers diagnosed with postpartum depression. Studies included in this meta-analysis were single-group pre-posttest, non-randomized and randomized controlled studies published from 1986 to 2010 that included face-to-face psychotherapy and psychopharmacology as well as non-traditional methods such as exercise and nurse-assisted counseling. 53 published studies were analyzed. The randomized studies showed a moderate to large effects (d= 0.72 to 1.25, k= 9) when postpartum interventions were compared to a control condition, and smaller effects (d= 0.3 to 0.57, k = 13) to treatment as usual. When postpartum interventions were compared to each other there was small to no difference in effect sizes (k = 9). All of the non-randomized comparisons showed no significant difference, except when therapy was compared to treatment as usual (d= 0.55, k = 2). Pre-post studies showed large effect sizes for therapy (d= 0.95, k = 7) and medication treatments (d= 4.30, k = 5). Influence analyses suggest that two studies had a large effect on aggregate effect sizes and heterogeneity statistics. Moderator and multivariate analyses were largely underpowered. Publication bias was not significantly related to outcome. Clinical implications for postpartum depression treatments and directions for future research were identified.