What do four decades of research tell us about the association between childhood adversity and psychosis: an updated and extended multi-level meta-analysis

Background A 2012 meta-analysis found that patients with psychosis are almost three times more likely to have been exposed to childhood adversity than controls. Additional studies have been rapidly accumulating for more than a decade. We address the urgent need for an updated evidence synthesis, but also expand the available evidence by investigating a broad array of adversity subtypes and exploring sex and the age of psychosis onset as relevant factors. Methods Following PRISMA guidelines, we searched PubMed, EMBASE, PsycINFO, Web of Science, WANFANG, and China National Knowledge Internet, for case-control studies (adversity in patients vs. controls, or psychotic symptoms in adversity-exposed vs. non-exposed individuals), cross-sectional and cohort studies on the association between adversity and psychotic symptoms/illness. Multi-level meta-analyses, meta-regressions, and study quality assessments were conducted. PROSPERO registration number: CRD42022310002. Findings The main analysis included 183 study samples (N=349 265), with 119 case-control studies (15 186 cases; 14 879 controls), 51 cross-sectional studies (N=299 659), and 13 cohort studies (N=19 541). Significant associations between adversity and psychosis were observed across all study designs, yielding an overall OR of 2·80 (2·18–3·60). Secondary analyses revealed that exposure to each adversity subtype increased the odds of psychosis, with the highest OR of 3·54 (3·04–4·13) for emotional abuse, and the lowest OR of 1·58 (1·48–1·68) for parental antipathy. No statistically significant sex differences were observed, although the OR for sexual abuse was higher for women. Onset of psychosis was earlier in adversity-exposed vs. unexposed individuals (mean difference=-0·79 years, 95%CI -1·47 to -0·12). Interpretation This is the largest meta-analysis to date on the association between childhood adversity and psychosis, confirming but significantly expanding on past findings. The results have broad clinical implications, as they highlight the need for primary prevention of exposure to early adversities and the implementation of trauma-informed and trauma-focused therapies in the treatment of psychosis.Funding None.