RESUMEN
Introduction: Social disconnectedness and isolation are risk factors for poor mental health. Community-based friendship group interventions have been designed to increase an individual's social capital and consequently their mental wellbeing. Structured and unstructured friendship groups reflect two distinct approaches to friendship group interventions. Methods: This meta-analysis investigated whether structured or unstructured community friendship groups are more effective for mental health and social capital outcomes. A systematic search of quantitative studies was conducted across seven databases and study quality was assessed using the Effective Public Health Practice Project (EPHPP) tool. Eight studies (2 unstructured and 6 structured friendship groups) were included in the review, published between 2005 and 2020. Results: Structured friendship groups had a small significant effect on reducing participant depression symptoms. There was not enough available data to compare the effectiveness of structured and unstructured groups for mental health outcomes. There was substantial heterogeneity between studies to calculate pooled effect sizes for any social capital outcomes. Data synthesis indicated mixed reviews for social capital outcomes, likely due to the large heterogeneity and limited studies. Discussion: This meta-analysis provides limited support for positive mental health outcomes following structured community-based friendship group interventions. There is a need for additional research as a large research gap remains, particularly for unstructured friendship groups. Systematic Review Registration: https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=260639, CRD42021260639.