Statistical Analyses
We performed a nested case-control study from the IGEDEPP cohort. We identified 250 cases of early PPD, defined as meeting DSM-5 criteria for major depressive episode at the 8 week interview. Similarly, there were 235 cases of late PPD, identified at the one-year visit. Categories were mutually exclusive, participants with early PPD were not considered as having late PPD even if they met criteria for PPD at the one-year visit. Women without any form of perinatal depression (i.e. neither during pregnancy nor in the first year postpartum) were used as unmatched controls (N=2726).
Descriptive statistics for binary and categorical variables included sample sizes and percentages for each category and according to the case/control status. Multivariable logistic regressions were conducted to quantify the association between PPD and associated factors at a given time point. Odds-ratios (ORs), adjusted for socio-demographic variables (i.e. age, marital status, education level, and employment), are reported along with their 95% confidence interval (CI) and p-values. The significance threshold was set at .05. Socio-demographic variables and variables with a p-value below 0.2 in the first analysis were included as adjustment factors for the multivariable model, using stepwise variable selection to optimize the Akaike information criterion (AIC). All analysis were performed with R, v3.6.1.
RESULTS