Social capital in pregnancy and postpartum depressive symptoms: a prospective mother-child cohort study (the Rhea study)

Background: Depression, and to a lesser extent postpartum depressive symptoms, have been associated with characteristics of the social environment and social capital. Up to the present, mostly cross-sectional studies have explored such an association without providing a clear temporal relationship between social capital and depression.

Objectives: To estimate prospectively the effect of individual-level self-reported maternal social capital during pregnancy on postpartum depressive symptoms.

Design: Prospective mother-child cohort (Rhea study).

Settings: 4 prenatal clinics in Heraklion, Crete, Greece.

Participants: All women for one year beginning in February 2007. From the 1388 participants, complete data were available for 356 women.

Methods: Women self-completed two questionnaires: The Social Capital Questionnaire at about the 24th week of gestation and the Edinburgh Postnatal Depression Scale (range 0-30) at about the 8-10th week postpartum. Maternal social capital scores were categorized into three groups: the upper 10% was the high social capital group, the middle 80% was the medium and the lowest 10% was the low social capital group that served as the reference category. Multivariable log-binomial and linear regression models were performed for: the whole available sample; for participants with a history of depression and/or prenatal EPDS≥13; for participants without any previous or current depression and prenatal EPDS score<13. Potential confounders included demographic, socio-economic, lifestyle and pregnancy characteristics that have an established or potential association with maternal social capital in pregnancy or postpartum depressive symptoms or both.

Results: Higher maternal social capital was associated with lower EPDS scores (highest vs lowest group: β-coefficient=-3.95, 95% CI -7.75, -0.14). Similar effects were noted for the subscale value of life/social agency (highest vs lowest group: β-coefficient=-5.96, 95% CI -9.52, -2.37). This association remained significant for women with and without past and/or present depression only for the subscale value of life/social agency although with a more imprecise estimate. No effect was found for participation, a structural dimension of social capital.

Conclusions: Women with higher individual-level social capital in mid-pregnancy reported less depressive symptoms 6-8 weeks postpartum. Given the proposed association of perceptions of the social environment with postpartum depressive symptoms, health professionals should consider evidence-based interventions to address depression in a social framework.