Is seasonal variation another risk factor for postpartum depression?

INTRODUCTION Postpartum depression (PPD) occurs in the first 6 months after delivery in 10% to 20% of mothers. Despite the availability of screening tools, there is a general consensus that PPD is under-diagnosed. A number of risk factors contributing to PPD have been investigated, but role of seasonal variability in PPD is unclear. Our purpose was to assess whether seasonal variation is another risk factor for PPD. METHODS This cross-sectional pilot study was conducted at 2 family medicine clinics and an obstetrics/gynecology clinic over 24 months. During their postpartum visit, mothers who gave consent were asked to fill out a survey requesting demographic data, followed by the Edinburgh Postpartum Depression Scale (EPDS). The EPDS is a well-validated tool shown to be highly effective in detecting postnatal depression. A score >12 on the EPDS indicated a likely risk of PPD. RESULTS Of the 556 patients approached, 530 completed the EPDS. Mean (+/-SE) patient age was 24.9 +/- 0.2 years; 71% were African Americans; 74% were single mothers; and 39% had at least some college education. On the EPDS, 17.8% scored > or =13. Of the depressed patients, 18.1% had babies born during the winter, 19.2% had babies born during the spring, 13.4% during the summer, and 21.5% during autumn (chi(2); P = .342). Mothers with very good or excellent support at home had lower EPDS scores (12.6%) than mothers with just adequate support (44.0%) or very little or no support (30.8%; P < .0005). A greater proportion of women with a history of depression (42.9% vs 12.9%) or who were currently taking antidepressives (58.3% vs 15.9%) were in the depressed group (P < .0005). Logistic regression analysis with the above variables, excluding education and income (excess missing data), on the 452 women with complete datasets found 4 significant predictors of an EPDS score >12. Predictors were history of depression (odds ratio [OR], 4.003; 95% CI, 2.016-7.949); parity (OR, 1.431; 95% CI, 0.204-1.701); social support (OR, 3.904; 95% CI, 2.08-7.325); and currently taking medication for depression (OR, 3.613; 95% CI, 1.207-10.817). CONCLUSION The slight seasonal variation in PPD in our pilot study was not statistically significant. Our study was underpowered to detect the projected differences in seasons. Additional patients are needed to diversify the participants and provide an adequate sample to test the projected seasonal differences. The high ORs found for greater parity, weak social support, history of depression, and currently taking antidepressants suggest that new mothers with these characteristics should be questioned about symptoms of PPD.