A sibling study of child neglect and depressive symptoms in adulthood: Evidence from the FAMILY Cohort

Abstract

Introduction: Studies have suggested that the mental health consequences of child neglect may be comparable to physical and emotional abuse. However, these findings are susceptible to confounding by childhood environment and potential genetic disposition. As sibling studies control for confounding from shared measured and unmeasured confounders, we therefore sought to examine the association between child neglect and depressive symptoms using a sibling fixed effects design. Methods: 1,042 participants aged  18 years (482 sibling groups) from the FAMILY Cohort in Hong Kong were administered a child neglect scale including items on health, clothing, food, education and safe living conditions. Depressive symptoms were assessed using the Patient Health Questionnaire-9. Multilevel hurdle models predicting the presence and severity of depressive symptoms from child neglect were fitted. Results: 37.1% of adults reported childhood neglect. Within-sibling effects of child neglect were not associated with the presence (adjusted OR=1.03, 95% CI 0.61 to 1.75) or severity of depressive symptoms (adjusted IRR=1.15, 95% CI 0.89 to 1.48), adjusting for non-shared factors (age, sex, nativity and parental occupation at age 10). Between-sibling effects of child neglect were also not associated with the presence (adjusted OR=0.75, 95% CI 0.49 to 1.48) or severity of depressive symptoms (adjusted IRR=1.17, 95% CI 0.92 to 1.48). Conclusion: The sibling fixed effects model did not find an association between child neglect and depressive symptoms. Future studies should account for potential confounding influences from the childhood environment, examine whether associations might be contextually specific and the role of shared mediators and colliders in a sibling study. Source of funding: The establishment of the original cohort was supported by the Hong Kong Jockey Club Charities Trust from 2007 to 2014. Further support provided by a grant from the Research Grants Council General Research Fund (Ref. No.: 17633216).