Abstract
PURPOSE: To conduct a systematic review and meta-analysis to examine the strength of associations between social network size and clinical and functional outcomes in schizophrenia. METHOD: Studies were identified from a systematic search of electronic databases (EMBASE, Medline, PsycINFO, and Web of Science) from January 1970 to June 2016. Eligible studies included peer-reviewed English language articles that examined associations between a quantitative measure of network size and symptomatic and/or functional outcome in schizophrenia-spectrum diagnoses. RESULTS: Our search yielded 16 studies with 1,929 participants. Meta-analyses using random effects models to calculate pooled effect sizes (Hedge's g) found that smaller social network size was moderately associated with more severe overall psychiatric symptoms (N = 5, n = 467, g = - 0.53, 95% confidence interval (CI) = - 0.875, - 0.184, p = 0.003) and negative symptoms (N = 8, n = 577, g = - 0.75, 95% CI = - 0.997, - 0.512, p = 0.000). Statistical heterogeneity was observed I(2) = 63.04%; I(2) = 35.75%,) which could not be explained by low-quality network measures or sample heterogeneity in sensitivity analyses. There was no effect for positive symptoms (N = 7, n = 405, g = - 0.19, 95% CI = 0.494, 0.110, p = 0.213) or social functioning (N = 3, n = 209, g = 0.36, 95% CI = - 0.078, 0.801, p = 0.107). Narrative synthesis suggested that larger network size was associated with improved global functioning, but findings for affective symptoms and quality of life were mixed. CONCLUSION: Psychosocial interventions which support individuals to build and maintain social networks may improve outcomes in schizophrenia. The review findings are cross-sectional and thus causal direction cannot be inferred. Further research is required to examine temporal associations between network characteristics and outcomes in schizophrenia and to test theoretical models relating to explanatory or mediating mechanisms.