Associations of obstetric, sleep and social support variables with perinatal depression in the UK Biobank

BackgroundPerinatal depression (PND) is highly prevalent and can have serious consequences for mother and baby. However, predictors such as obstetric complications, sleep disturbances and social support are underexplored in large population-based cohorts. MethodsWe analysed data from 44,479 UK Biobank participants with at least one live birth to examine associations between obstetric factors (e.g., hypertensive disorders, multiple pregnancy, preterm birth), later self-reported sleep problems (e.g., insomnia, short sleep), and social support (e.g., loneliness) with PND risk. Comparisons were made with a) healthy controls and b) women with non-perinatal depression, using both lifetime history and prospective ICD-coded diagnoses around first birth. Predictors were examined using logistic regression and penalised regression models, adjusting for age, neuroticism, body mass index, deprivation, and education.ResultsLifetime PND was associated with younger age at first birth, higher parity, stillbirth/miscarriage, preterm birth, sleep disorders, and later self-reported insomnia, difficulty getting up, short sleep, loneliness, and evening chronotype vs. controls. Prospective analyses showed increased PND risk vs. controls following multiple or preterm births and in those later reporting difficulty getting up. Preterm birth was associated with PND but not non-perinatal depression. Penalised regression models showed good discrimination of lifetime and prospective PND vs. controls (AUC = 0.78 and 0.77). Strongest predictors included neuroticism, younger maternal age, higher BMI, greater parity, difficulty getting up, stillbirth/miscarriage, and sleep disorder history.ConclusionsMothers with pregnancy-related trauma/complications, tendency towards sleep difficulties and low social support may benefit from targeted preventive resources. PND may also be associated with later sleep problems and loneliness.