The effectiveness of exercise in the management of post-natal depression: systematic review and meta-analysis

A. Daley, K. Jolly, C. MacArthur
2009 Family Practice  
The effectiveness of exercise in the management of post-natal depression: systematic review and meta-analysis. Family Practice 2009; 26: 154-162. Background. Post-natal depression (PND) is a serious mental health problem that may be reduced by exercise. National Institute for Health and Clinical Excellence in England have recommended that health professions should consider exercise as a treatment for PND. Objective. To evaluate the effectiveness of exercise in the management of PND. Methods.
more » ... tematic review and meta-analysis of randomized controlled trials (RCTs). Data sources involved in the study are Cochrane Library (CENTRAL), MEDLINE, EMBASE, PsycINFO, Science Citation Index and Social Science Citation Index, CINAHL and SPORTDiscus. Review methods. Selection criteria are RCTs and quasi-RCTs that compared any type of exercise intervention with other treatments or no treatment in women with PND. Database searches and abstracts were reviewed independently by two authors. The Delphi criteria were used to assess the quality of included studies. Data were abstracted by two reviewers. Data synthesis is meta-analysis. Main outcome measure is post-natal depression. Results. Five studies fulfilled our inclusion criteria. When compared with no exercise, exercise reduced symptoms of PND {SMD = -0.81 [95% confidence interval (CI): -1.53 to -0.10]}. The overall WMD in Edinburgh Post-natal Depression Scale score was -4.00 points (95% CI: -7.64 to -0.35). However, significant heterogeneity was found. The effect size was reduced considerably (non-significant) when the trial that included exercise as a co-intervention with social support was excluded [SMD = -0.42 (95% CI: -0.90 to 0.05)] and heterogeneity was no longer present. Conclusions. Due to heterogeneity, it is uncertain whether exercise reduces symptoms of PND. Caution is also required when interpreting findings from the main analysis as only five small trials were included and CIs were wide. Further research is evidently required.
doi:10.1093/fampra/cmn101 pmid:19126829 fatcat:rqwduqkbtneffl5qwdo7shl7ni