Abstract
IMPORTANCE: Childhood bereavement increases the risk of common psychiatric disorders later in life. However, the role of stress resilience in this association remains underexplored. OBJECTIVE: To assess whether stress resilience mediates the association between childhood bereavement and psychiatric disorder risk in adulthood. DESIGN, SETTING, AND PARTICIPANTS: Three matched cohort studies were performed using data recorded in the Swedish Military Conscription Register. Individuals with childhood loss of either a parent or a sibling (19 162 participants), a parent (16 247 participants), or a sibling (3023 participants) due to death from 1987 to 2020, together with 10 unexposed individuals per exposed individual, were matched on sex, birth year, and county of birth. All participants had available stress resilience measure at conscription. Data were analyzed from February to December 2024. EXPOSURES: Childhood bereavement was ascertained from the Swedish Multi-Generation and Causes of Death Registers. MAIN OUTCOMES AND MEASURES: Incident diagnosis of depression, anxiety, substance use disorder, and stress-related disorder was ascertained from the Swedish Patient Register. Cox models were used to estimate the association between childhood bereavement and risk of postconscription psychiatric disorders after multivariable adjustment. Causal mediation analysis was conducted to examine if stress resilience measured at conscription mediated this association. RESULTS: Among 1 733 085 conscripted individuals (median [IQR] age at conscription 18.2 [18.0-18.5] years; 1 707 960 [98.5%] male), the median (IQR) age of individuals exposed to loss of either a parent or a sibling, parent, or sibling was 13.4 (9.6-15.8), 13.7 (10.4-15.9), and 10.7 (5.5-14.9) years, respectively. The crude incidence rate of any psychiatric disorder was 7.9, 8.1, and 6.6 per 1000 person-years among the 3 groups (5.3, 5.8, and 5.5 per 1000 person-years among the respective unexposed groups). A positive association was noted for loss of either a parent or a sibling (HRs ranged from 1.13; 95% CI, 1.06-1.20 for anxiety to 1.31; 95% CI, 1.23-1.40 for substance abuse disorder) and loss of a parent (HRs ranged from 1.10; 95% CI, 1.01-1.20 for stress-related disorders to 1.19; 95% CI, 1.12-1.27 for depression). For loss of a sibling, the statistically significant associations were for any common psychiatric disorder (HR, 1.12; 95% CI, 1.00-1.25) and stress-related disorders (1.27; 95% CI, 1.04-1.55). Stress resilience partially mediated the associations (proportions for loss of either a parent or a sibling ranged from 10.6%-19.4%, for a parent ranged from 15.6%-21.7%, and for a sibling ranged from 6.2% for stress-related disorders to 18.4% for any common psychiatric disorder). CONCLUSIONS AND RELEVANCE: In this cohort study of a nationwide Swedish sample, altered stress resilience was found to be one mechanism through which childhood bereavement is associated with risk of psychiatric disorders later in life.