Abstract
Adolescence is a vulnerable period for the onset of mental disorders and risk behaviours. Based on the Health-Promoting Schools Framework, whole-school interventions offer a promising strategy in this developmentally-sensitive cohort, through championing a systems-based approach to promotion and prevention that involves the key stakeholders in an adolescent's life. The evidence-base surrounding the effectiveness of whole-school interventions, however, remains inconclusive, partly due to the insufficient number of studies in previous meta-analyses. An updated systematic review and meta-analysis was thus conducted on the effectiveness of whole-school interventions promoting mental health and preventing risk behaviours in adolescence. From 12,897 search results, 28 studies reported in 58 publications were included. Study characteristics and implementation assessments were synthesized across studies, and quality appraisals and meta-analyses performed. Analyses identified a significant reduction in the odds of cyber-bullying by 25%, regular smoking by 31% and cyber-aggression by 37% in intervention participants compared to the control. Whole-school interventions thus offer substantial population health benefits through the reduction of these highly-prevalent issues affecting adolescents. The non-significant findings pertaining to the remaining eleven outcomes, including alcohol use, recreational drug use, anxiety, depression and positive mental health, are likely attributable to suboptimal translation of the Health-Promoting Schools Framework into practice and inadequate sensitivity to adolescents' local developmental needs. Given the ongoing challenges faced in the implementation and evaluation of these complex interventions, this study recommends that future evaluations assess the implementation of health-promoting activities in both intervention and control conditions and actively use this implementation data in the interpretation of evaluation findings.Preregistration: A pre-registered PROSPERO protocol (ID: CRD42023491619) informed this study.