Do the early social environment and persistent peripartum depressive symptoms shape toddlers' expressive language?

Extensive research suggests that peripartum depression is a risk factor for children's early language development. Yet, previous research on this association shows mixed results, often lacking information on the persistence of depression and the social context. This population-based cohort study addresses this gap by investigating the longitudinal influence of peripartum depressive symptoms on toddlers' expressive language. Specifically, we systematically examined the influences of timing, severity, and persistence of depressive symptoms during pregnancy and the first 6 months postpartum on child expressive language development, while accounting for important social and environmental factors.

This study is part of a prospective, population-based investigation conducted within the follow-up Uppsala Birth Cohort study in Uppsala, Sweden. The final analysis included 2176 mother-infant dyads (1122 boys, mean age = 18.3 months, SD = 0.7). Perinatal depressive symptoms were assessed at gestational weeks 17 and 32 and at postpartum six weeks and six months, using the Edinburgh Postnatal Depression Scale (cut-off >12). At 6 months postpartum, mothers were also invited to fill out the Postnatal Bonding Difficulty Questionnaire. At 18 months postpartum, mothers completed the Language Development Survey, which assessed expressive vocabulary and word combinations. Multivariable linear regression models were applied to examine the associations between peripartum depressive symptoms and child language development. Adjusted models incorporated background and social context variables to account for potential confounding factors.

Depressive symptoms during prenatal and postnatal periods were not significantly associated with language outcomes. Our final model identified negative associations with second-born status, family history of late talkers, countryside residence, and maternal age at childbirth. Positive correlations were found for sex (girl) and pregnancy length. The final model explained 8.4% of the variance (F(22, 1566) = 6.525, p < 0.001). Furthermore, we found that persistent depressive symptoms were not significantly related to language outcomes (Kruskal-Wallis test: H = 2.227, df = 2, p = 0.21).

Our findings found no negative link between peripartum depressive symptoms and expressive language in toddlers, even after considering timing, severity, and persistence. While no immediate direct negative influence of peripartum depressive symptoms was observed, the long-term cumulative effects later in life remain unclear.