The Effects of the School Environment On Student Health: A Systematic Review of Multi-Level Studies

Introduction: Health outcomes vary between schools and it is theorised that this may be partly attributable to variation in the school environment. Existing systematic reviews have not drawn authoritative conclusions because of methodological limitations in the review or studies available; in particular the failure of reviews and/or primary studies to prioritise evidence from multi-level models and the failure to distinguish between potential confounders such as family and neighbourhood factors and potential mediators such as student and school peer factors.

Methods: We undertook a systematic review to address the question: what are the effects of school-level objective measures of the environment on health among school students aged 4-18 years examined via multi-level quantitative designs? Through very comprehensive searching via databases, experts in the field and reference checking we found 82,775 potentially relevant references, of which 20,446 were duplicates. These were screened resulting in the identification of 42 multi-level studies pertinent to our research question, ten of which were judged of sufficient quality. Because these were heterogeneous with regard to measures, these were subject to narrative synthesiis.

Results: We found consistent evidence from cross-sectional and longitudinal studies of middle schools in the USA and secondary schools in the UK that in schools where attainment was higher than would be expected from the social profile of students, and truancy was lower, lower rates of substance use and violence are observed. We also reviewed evidence from one cross-sectional study which reported that students in US middle schools with a larger total campus size and playground areas per student have higher rates of student accelerometer-measured physical activity during school hours. Our review also found evidence from one cross-sectional study of US high school students that the number of unobservable and unsupervised places in and around school was associated with 10^th and 12^th-graders’ use of alcohol in school and 10^th graders use of marijuana in school in the previous 12 months. An attractive school environment was associated with 10^th and 12^th-graders use of alcohol in school in the past year. A final study reported that from a cross-sectional study the following school-level factors were not associated with alcohol use among students age 13-14 in high, junior-high and middle schools in rural USA: whether eighth-graders are located within the same school as high school students or are in separate schools, school size and pupil-to-teacher ratio. 

Conclusion: Future trials of school environment interventions, particularly those modifying school ethos and policies on teaching, behaviour and discipline, should examine the effectiveness of increasing student engagement, attainment and attendance and their health effects. Broader educational policies and school funding need to provide the incentives and resources to ensure that schools are able to promote student health and wellbeing.