Analysis of Maternal Postnatal Depression, Socioeconomic Factors, and Offspring Internalizing Symptoms in a Longitudinal Cohort in South Africa

IMPORTANCE: Few studies from low-and-middle income countries have investigated long-term associations between maternal postnatal depression and offspring internalizing (ie, depressive and anxiety) symptoms, and none have investigated interactions in this association. OBJECTIVE: To investigate the association between maternal postnatal depression and offspring internalizing symptoms from adolescence to adulthood and the interaction with exposure to socioeconomic adversity and with the child’s sex. DESIGN, SETTING, AND PARTICIPANTS: This secondary analysis used data from Birth to Twenty Plus (BT20+), a prospective birth cohort study of children born in Soweto, South Africa, and followed up until age 28 years. Data were collected from 1990 to 2018, and data were analyzed for this study from February 16 through December 15, 2020. EXPOSURES: Maternal postnatal depression self-reported by mothers 6 months after childbirth. MAIN OUTCOMES AND MEASURES: The main outcome was offspring internalizing symptoms, assessed at offspring ages 14 years, 22 years, and 28 years and modeled longitudinally. Participants with the highest probability of experiencing high internalizing symptoms (ie, those in the top 20% of the distribution) from age 14 to 28 years were categorized as belonging to the high internalizing symptoms trajectory (vs the low trajectory). Socioeconomic adversity was measured with an index (continuous variable) including low maternal education, household crowding, low assets, and low maternal age. This variable was further stratified into more than 1 SD above the mean index, more than 1 SD below the mean index, and from 1 SD below to 1 SD above the mean index to conduct subgroup analyses. Associations were investigated using multivariable regression models. RESULTS: Among 1087 participants born in Soweto, South Africa (543 [50.0%] male participants; 544 [50.0%] female participants), 118 individuals (10.8%) showed a high trajectory of internalizing symptoms from age 14 to 28 years vs 969 individuals (89.1%) with a low trajectory. Children exposed to maternal postnatal depression had statistically significantly increased odds of following the high trajectory (adjusted odds ratio [aOR] per 1-SD increase in maternal postnatal depression, 1.20; 95% CI, 1.02-1.41). This increase in odds differed by exposure to socioeconomic adversity and by child sex: for male participants, the increase in odds was greater in a context of higher vs lower socioeconomic adversity (eg, for >1 SD above the mean: aOR, 3.28; 95% CI, 1.06-10.14 vs for >1 SD below the mean: aOR, 0.98; 95% CI, 0.64-1.50; P for interaction = .12), while for female participants, the increase in odds was greater in a context of lower vs higher socioeconomic adversity (eg, for >1 SD below the mean: aOR, 1.82; 95 % CI, 1.12-2.98 vs for >1 SD above the mean: aOR, 0.59; 95 % CI, 0.30-1.17; P for interaction = .002) (P for 3-way interaction = .003). CONCLUSIONS AND RELEVANCE: This study found that postnatal depression was associated with higher odds of persistently increased internalizing symptoms among offspring from adolescence to adulthood in a middle-income country, with variation by socioeconomic adversity and sex. These findings suggest that better understanding of these associations is needed to implement targeted interventions and maximize the impact of public health initiatives aimed at breaking the intergenerational transmission of mental health problems.