SA7. Different Longitudinal Relationships Between Childhood Adversity, Adolescent/Adult Social Support, and Dimensions of Depressive and Psychotic Symptoms in 2 General Population Cohorts

Abstract

Background: Childhood adversity and poor social support are implicated in the development of psychosis but have rarely been investigated together. This study investigated longitudinal relationships in adolescents and young adults between childhood adversity, adolescent/adult social support, and later dimensions of depressive symptoms (DS) and psychotic experiences, measured as paranoid ideation (PI) and nonparanoid anomalous experiences and beliefs (AEB). Methods: Data came from 2 longitudinal general population cohorts (Cohort 1: ‘ROOTS’, N = 1238, assessed at age 14 (baseline) and 17; Cohort 2: ‘NSPN’, N = 1919, age 16–25, assessed at baseline and 1-year follow-up). Self-report DS, AEB, and PI were measured at age 17 in Cohort 1 and at follow-up in Cohort 2. Childhood adversity was measured retrospectively at baseline with a structured interview with caregivers in Cohort 1 and a self-report questionnaire in Cohort 2. Self-report social support from friends and family were measured at baseline in both the cohorts. Longitudinal relationships were tested by estimating structural equation models with robust full-information maximum likelihood. The models had direct paths from childhood adversity to DS, PI, and AEB, from childhood adversity to social support and from social support to symptom outcomes. Sex, nonwhite ethnicity, age (Cohort 2), socioeconomic deprivation, cannabis use, family history of mental disorder, and maternal years of education were included as covariates. Results: Childhood adversity had direct effects promoting all symptom outcomes, even when considering its effects on later social support and the effects of support on symptoms. Childhood adversity impaired later support from family in Cohort 1 and support from family and friends in Cohort 2. Social support from family and friends reduced depressive symptoms and paranoid ideation in both cohorts. AEB was less affected by social support; social support from friends reduced AEB in Cohort 2 but with a smaller effect size than for DS and PI. No other social support variables affected AEB. Part of the effects of childhood adversity on symptoms were mediated by poorer family support in Cohort 1 (mediated by family support: DS = 16.5%, PI = 14.2%) and by poorer family and friendship support in Cohort 2 (mediated by family support: DS = 29.5%, PI = 14.7%; mediated by friendship support: DS = 15.4%, PI = 20%, AEB = 10.8%). Conclusion: Childhood adversity may have long-lasting effects predisposing to unusual experiences and beliefs that may not be greatly modifiable by positive social relationships in later life. Paranoid ideation and depressive symptoms are more affected by recent social support. Part of the observed effects of childhood adversity on mental health are explained by poorer quality of relationships later in life. Good social support might not eliminate all psychotic experiences but may lower paranoia and distress, mitigating some effects of early life adversity.