Abstract
OBJECTIVE: Postpartum psychosis affects 1-2 per 1,000 women following childbirth. Although bipolar disorder and prior postpartum psychosis are established risk factors, the contribution of familial psychiatric history is not well characterized. METHODS: Using Swedish national registers, we identified 1,213,829 women with a full sibling who delivered their first liveborn child between 1980 and 2017. Postpartum psychosis was defined as mania, psychosis, or psychotic depression within 90 days postpartum. Associations between sibling psychiatric diagnoses and postpartum psychosis were estimated using logistic regression with cluster-robust standard errors, including stratification by the woman's own psychiatric history. OUTCOMES: Overall, 1,831 women developed postpartum psychosis (1.5 per 1,000). In sister-sister pairs, the strongest associations were for sister's postpartum psychosis (OR 10.84, 95% CI 5.76-20.40), cyclothymia (8.29, 4.65-14.76), schizoaffective disorder (7.90, 4.99-12.52), bipolar disorder (5.00, 4.23-5.90) and schizophrenia (4.37, 2.75-6.44); brief psychotic disorder, major depressive disorder, and anxiety disorders were also associated with elevated risk. Patterns were broadly similar in brother-sister pairs. Among women with no prior psychiatric diagnosis, sibling schizoaffective disorder (4.82, 2.99-7.74) and bipolar disorder (3.07, 2.59-3.64) were associated with particularly large relative-risk increases. Among women with a history of mania/psychosis, sibling psychiatric information contributed little additional predictive value. INTERPRETATION: Postpartum psychosis shows strong familial aggregation, with risks extending across multiple sibling diagnoses. The pattern of familial associations, with highest associations linked to sibling bipolar, cyclothymia, and schizoaffective disorder, supports conceptualizing postpartum psychosis within the spectrum of severe mood disorders with psychotic features.