Postnatal depression has historically been ignored, but there is increasing awareness of the prevalence and impact of depression at this time (Howard et al., 2014; Stein et al., 2014). Depression during pregnancy (known as antenatal or prenatal depression) is also experienced by many women but tends to receive less attention than postnatal depression.

Studies show that antenatal and postnatal depression are common, but it is difficult to make a precise estimate of their prevalence because results from different studies can vary widely. A previous systematic review (Gavin et al., 2005) estimated that the prevalence of depression ranged between 6.5% to 13% at different time points during pregnancy and the postnatal period. However, this review was published over a decade ago and only included 28 studies, so it is important for prevalence estimates to be updated to include more recent research. In addition, Gavin et al.’s review excluded studies conducted in “less developed” countries. This is a major limitation as the vast majority of births occur in low or middle income countries.

This recent systematic review by Woody et al. (2017) aimed to give an up-to-date summary of the prevalence and incidence of perinatal depression. In their review, Woody et al. also examined factors which might explain the difference in depression prevalence estimates between studies, such as whether the study was conducted during pregnancy or the postnatal period, and where the study was conducted.

Methods

Woody et al. (2017) identified relevant published peer-reviewed research papers by searching two databases (PsycINFO and PubMed) using terms related to depression, prevalence and perinatal.

Studies were eligible for the review if they were published between 1980 and 2015 in any language, and gave an estimate of the prevalence or incidence of depression during pregnancy or the postnatal period. Depression could be assessed using a diagnostic interview or a screening scale. To be eligible, studies had to use samples that were representative of the community or region under study, as those using non-representative samples might over or underestimate the prevalence of depression in the general population.

One author performed the database searches, screened the articles and extracted the data, with 10% of papers cross-checked by a second author.

Results

The systematic review searches identified over 10,000 studies, but only 96 of these were eligible to be included in the analyses. Of the 96 studies, 60 were conducted in high income countries, 33 studies in middle income countries and 3 in low income countries.

All of the included studies gave an estimate of the prevalence of perinatal depression but only six examined the incidence of depression. Most of the results in the paper are therefore about the prevalence of perinatal depression. The analyses are quite complicated and can take a while to get your head around! In the paper, the meta-regression is described first then the meta-analysis, but I’ll describe them the other way round as I think it is easier to follow.

Woody et al. calculated prevalence estimates for depression by combining the results of all the eligible studies using meta-analysis. The meta-analyses were adjusted for study characteristics, which is quite unusual.

Pooled adjusted prevalence estimates	High income countries	Low and middle incomes countries
Perinatal depression	11.4%	13.1%
Antenatal depression	9.2%	19.2%
Postnatal depression	9.5%	18.7%

The pooled adjusted prevalence of perinatal depression was estimated to be around 11.9%. The prevalence was higher in studies in low and middle income countries (13.1%) than in high income countries (11.4%).

Separate adjusted pooled prevalence estimates for antenatal and postnatal depression were also calculated: in high income countries these were 9.2% and 9.5% respectively, whilst in low and middle income countries they were 19.2% and 18.7% respectively. The review authors do not clarify why the estimates for low and middle income countries are so much higher when separated into the antenatal and postnatal periods compared to the overall perinatal estimate.

Woody et al. also wanted to examine why studies sometimes find such different estimates for the prevalence of perinatal depression. They used meta-regression to see if prevalence estimates were related to key characteristics of the studies. The characteristics that they examined were period (antenatal vs. postnatal), country income status (high income vs. low or middle income), depression assessment (clinical diagnosis vs. symptom scale), recruitment (community-based vs. healthcare-based) and whether the sample was urban, rural or mixed. They found that all these characteristics together explained approximately one third (31%) of the differences in depression prevalence estimates between studies. Examining the characteristics separately, they found that country income status predicted prevalence, with studies in low or middle income countries showing a significantly higher prevalence of perinatal depression than those in high income countries (similar to the meta-analysis finding described above). In addition, after adjusting for other study characteristics, studies which used symptom scales found a significantly higher prevalence of depression than those which used diagnostic instruments. Interestingly, the period (antenatal or postnatal) did not predict the prevalence of depression. They also found no difference based on how the sample was recruited or whether the sample was urban or rural.

Strengths and limitations

This systematic review provides an up to date estimate of the prevalence of depression during pregnancy and the postnatal period. An important strength of this review is the inclusion of studies from low and middle income countries. However, it is worth highlighting that whilst 33 studies were identified from middle income countries, only 3 included studies were conducted in low income countries. As the authors note, more research is needed in these settings.

The authors identified very few studies on the incidence of perinatal depression, but I don’t think that this is necessarily a major limitation of the review. Recent research has highlighted that perinatal depression often occurs within the context of prior mental health problems: a prospective longitudinal study (Patton et al., 2015) found that 85% of women in their sample who experienced perinatal depression had a history of poor mental health during adolescence or young adulthood.

Almost all of the studies on prevalence in this review estimated point prevalence (i.e. the proportion of women experiencing depression at a particular time point). This is likely to underestimate the proportion of women who will experience depression at any time during pregnancy or the postnatal period (known as the period prevalence). The period prevalence of depression has been estimated at up to 18% in pregnancy and 19% in the first three months after birth (Gavin et al., 2005). In addition, although this study focused on depression it is important to remember that other mental disorders also occur during pregnancy and the postnatal period so the overall prevalence of women experiencing mental health problems during the perinatal period will be higher.

Overall, the review appears to be generally well conducted, although some things could be clarified (e.g. what period they defined as “postnatal”) and the searches might have been improved by including regional databases such as African Index Medicus. The funnel plot (given in the appendix) seems to suggest that there may be some publication bias, but this is not discussed in the paper. The analyses are interesting but there are some issues interpreting the results of the meta-analysis (as discussed above), and it would be helpful for the authors to address these.

Conclusions and implications

This systematic review estimated that the prevalence of perinatal depression is approximately 12%. Estimates of prevalence are important for increasing awareness of depression during pregnancy and the postnatal period, and informing the allocation of health resources. In particular, this study highlights that depression is equally prevalent during pregnancy as in the postnatal period, and that there is a higher prevalence of perinatal depression in low and middle income countries than in high income countries.

Conflicts of interest

None.

Links

Primary paper

Woody C, Ferrari A, Siskind D, Whiteford H, Harris M. (2017) A systematic review and meta-regression of the prevalence and incidence of perinatal depression. Journal of Affective Disorders, Volume 219, 86-92. http://dx.doi.org/10.1016/j.jad.2017.05.003

Other references

Howard LM, Molyneaux E, Dennis C-L, Rochat T, Stein A, Milgrom J. (2014) Non-psychotic mental disorders in the perinatal period. The Lancet 2014; 384: 1775-88.

Stein A, Pearson RM, Goodman SH, Rapa E, Rahman A, McCallum M, et al. (2014) Effects of perinatal mental disorders on the fetus and child. The Lancet. 2014; 384(9956): 1800-19.

Gavin NI, Gaynes BN, Lohr KN, Meltzer-Brody S, Gartlehner G, Swinson T. (2005) Perinatal depression: a systematic review of prevalence and incidence. Obstet Gynecol. 2005; 106(5, Part 1): 1071-83.

Patton GC, Romaniuk H, Spry E, Coffey C, Olsson C, Doyle LW, et al. (2015) Prediction of perinatal depression from adolescence and before conception (VIHCS): 20-year prospective cohort study. The Lancet. 2015; 386(9996): 875-83.

Photo credits

• Photo by Alex Jones on Unsplash
• Photo by Annie Spratt on Unsplash
• U.S. Army photo by Spc. Paxton Busch