The association between autism and psychosis and the tools used to measure it: An updated systematic review and meta-analysis

OBJECTIVES: Autistic individuals are at increased risk of psychotic experiences and being diagnosed with psychotic disorders. This association may result from methodological issues, including the misinterpretation of psychosis questionnaires by autistic individuals and clinicians' difficulty distinguishing between the conditions. DESIGN: This meta-analysis aimed to review this association and examine whether it is moderated by the assessment measures used. METHODS: Systematic searches were conducted in PsycINFO, MEDLINE, CINAHL, Embase and Web of Science. Included studies required autism and psychosis-spectrum measurements, co-morbidity data, adult participants and quantitative data. Quality and risk of bias were assessed using the AXIS Critical Appraisal of Cross-Sectional Studies tool. Analyses examined correlations, odds ratios and Cohen's d as effects. RESULTS: Sixty-three papers (N = 6,903,960) were included. Associations were found between autistic and overall (r = .435, p < .0001), positive (r = .274, p < .0001), negative (r = .506, p < .0001) and disorganized (r = .366, p < .0001) psychosis-spectrum traits. Individuals with one condition had an increased risk of being diagnosed with the other (OR = 7.03, p < .001) and scored higher on trait measures of the other (d = 1.187, p < .0001). CONCLUSIONS: These meta-analyses evidence a strong association between autism and the psychosis spectrum, at both trait and diagnostic levels. Negative psychosis-spectrum traits were most strongly linked with autistic traits, while measures of positive traits showed weaker correlations, suggesting overlaps in expression and measurement. High heterogeneity and inconsistent reporting, however, hinder the certainty of conclusions, and research is required to better understand this overlap.