Objectives This study aims to measure incidence and persistence of depression and to investigate the influence of self-reported antenatal social support and traditional/nuclear family structure on incidence and persistence of depression between the third trimester of pregnancy and following childbirth. We hypothesised that lower antenatal social support would be associated with incidence and persistence of case-level depressive symptoms and the family structure would have an effect on the incidence and persistence of depressive symptoms.

Settings The cohort study described here was carried out in and around Ankara the capital of Turkey, because of the considerable heterogeneity of the population in terms of traditional Middle Eastern and ‘modern’ Western lifestyle and social environment. Samples were drawn from 20 urban and rural antenatal clinics (mainly primary care settings) within the geographic catchment.

Participants Of 730 women recruited in their third trimester, 578 (79.2%) were re-examined between 2 and 6 months after childbirth. Exclusion criteria were as follows: aged younger than 18 years, illiteracy, significant health problems and refusal to participate.

Primary and secondary outcome measures Close Persons Questionnaire items enquired about relationships with the husband, mother and mother-in-law and depression was ascertained using the Edinburgh Postnatal Depression Scale at the each assessments.

Results In those followed, onset of postnatal depression occurred in 13.9% and persistence of antenatal depression in 49.7%. After adjustment, worse emotional support from the mother-in-law was significantly associated with postnatal depression incidence (OR=0.93, 95% CI 0.87 to 0.99) and worse emotional support from the husband with postnatal persistence (OR=0.89, 95% CI 0.83 to 0.96) of antenatal depression. Family structure was not a risk or modifying factor.

Conclusions The incidence and persistence of postnatal depression in this Middle Eastern cohort were comparable to international findings. Certain family relationships predicted incidence and persistence of postnatal depression but no role of traditional/nuclear family structure was found.