Childhood parental psychiatric hospitalization and offspring suicidal behavior, violent crime, and violent Victimization: A Swedish population-based cousin comparison study

Population-level studies consistently report associations between parental psychiatric disorders and violence-related outcomes in offspring, but it remains unclear whether these reflect causal effects of parental illness or shared familial vulnerabilities. Using Swedish national registers, we followed 2,978,645 individuals born 1975-2005 from age 15 until the first violence-related outcome, emigration, death, age 25, or end of 2020. We estimated associations between exposure to parental hospitalization for severe mental illness (SMI), affective disorders (AD), and substance use disorders (SUD) during offspring ages 0-14 and three offspring outcomes: suicidal behavior, violent crime conviction, and violent victimization. Parental hospitalization for accidental injury served as a comparator to distinguish psychiatric-specific transmission from generic effects of parental absence. Cousin comparisons addressed unmeasured familial confounding, and subsequent models examined mediation by offspring sociodemographic status. At the population level, all parental psychiatric hospitalizations were associated with elevated risks across outcomes, with hazard ratios ranging from 1.56 (parental SMI hospitalization and offspring violent victimization) to 2.76 (parental SUD hospitalization and offspring violent crime conviction). Associations attenuated substantially in cousin-comparison models, and crucially, this attenuation was greater for psychiatric than for accidental hospitalization. Offspring sociodemographic status partly mediated the remaining within-family associations. Patterns were consistent across developmental periods of exposure. These findings indicate that the elevated risks associated with parental psychiatric hospitalization are driven largely by shared familial vulnerabilities rather than hospitalization-related disruption alone. Population-level estimates may overstate the causal environmental effects of parental psychiatric illness. Interventions may therefore need to target both broader familial risk processes and modifiable socioeconomic pathways in offspring.