Article Text

other Versions

Systematic review and meta-analysis
Psychosocial and psychological interventions reduce the risk of postnatal depression compared with standard care
  1. Michael W O'Hara
  1. Department of Psychology, University of Iowa, Iowa City, Iowa, USA
  1. Correspondence to: Dr Michael W O'Hara, Department of Psychology, University of Iowa, 11 Seashore Hall E, Iowa City, IA 52242, USA; mike-ohara{at}

Statistics from

Commentary on OpenUrlPubMed.

Implications for practice and research

  • Interventions that target women ‘at risk’ for postnatal depression are efficacious.

  • The evidence is weak for interventions that target the general population of women.

  • Future research should focus on developing better tools to identify ‘at-risk’ women, improving the efficacy of these and making them more accessible to women in need.


Postpartum depression is a significant mental health problem that impacts not only women, but also their children and families. It is prevalent, affecting up to 19% of women (7% major depression alone) in the first 3 months after delivery.1 A number of efficacious interventions have been developed to treat postpartum depression.2 Despite the availability of effective interventions, women experience considerable suffering prior to the onset of treatment and many women who are depressed postpartum receive no treatment.3 As a consequence, over the past 20 years there have been numerous studies of interventions delivered in pregnancy and the early postpartum period aimed at preventing the onset of depression and its deleterious consequences.


The authors searched the trials register of the Cochrane Pregnancy and Childbirth Group, checked secondary references and contacted experts working in the field. All published and unpublished trials of acceptable quality comparing psychosocial or psychological interventions with usual care were included. This yielded a total of 28 trials, including almost 17 000 women. A fixed-effect meta-analysis model was used except where there was significant clinical or statistical heterogeneity, in which case a random-effect meta-analysis was used. The risk of bias was carefully assessed in numerous ways. Depression was reported as categorical (depressed vs not depressed) and as a continuous variable (severity). Subgroup analyses were conducted based on the type of intervention, intervention provider, intervention duration and onset of intervention (antenatal vs postnatal). Finally, sensitivity analyses were conducted in cases of high risk of bias and significant missing data.


The principal finding was that a psychosocial or psychological intervention was effective in preventing postpartum depression relative to usual care (RR=0.78, 95% CI 0.66 to 0.93) based on 20 trials and over 14 000 women. Individualised postpartum home visits provided by public health nurses or midwives, peer-based telephone support and interpersonal psychotherapy were identified as promising interventions. Lay and professionally delivered interventions were both effective. Interventions for ‘at-risk’ women produced particularly good results (RR=0.66, 95% CI 0.50 to 0.88, n=1853). However, interventions aimed at the general population of pregnant women did not have a statistically significant impact (RR=0.83, 95% CI 0.68 to 1.02, n=12 874).


Postpartum depression takes an enormous toll on women, their children and families. In the USA alone, between 300 000 and 800 000 women are afflicted each year.1 Numerous interventions have been developed to treat postpartum depression and many of these have been adapted for prevention efforts. There are two large take home messages from this fine review by Dennis and Dowswell. The first is that postpartum depression can be prevented, or its symptoms mitigated by interventions that begin in pregnancy or soon after delivery. However, this optimistic conclusion is tempered by the results of the subanalyses reported by the authors. Second, it would appear that the strong evidence for a positive impact for prevention is only for interventions that target ‘at-risk’ women. In the aggregate, these did not have a statistically significant impact in reducing the risk of postpartum depression. This is an important finding because these studies included about 75% of the approximately 17 000 women included in this meta-analysis.

These findings present something of a public health dilemma. On the one hand, universal prevention ensures that every woman likely to develop postpartum depression will receive a preventive intervention. However, for the vast majority of women the intervention will be unnecessary. Moreover, universal interventions are likely to be ‘low dose’ because of cost considerations. Targeting ‘at-risk’ women (ie, women with depressive symptoms in pregnancy) is much more efficient and allows for more intensive interventions. However, screening tools are imperfect and many women who cannot afford a preventive intervention will go on to experience postpartum depression. It is imperative that we develop better tools to identify ‘at-risk’ women and improve the accessibility and effectiveness of interventions. Women, their children and families will greatly benefit from this work.


View Abstract


  • Competing interests None.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.