Physical activity interventions to improve the health and wellbeing of people experiencing homelessness: a mixed-methods systematic review

Abstract

BackgroundPeople experiencing homelessness have poor health. Physical activity is known to improve health outcomes. No published systematic review of physical activity interventions in this population exists. We aimed to systematically synthesise evidence for physical activity interventions in people experiencing homelessness. MethodsWe searched EMBASE, Web of Science, CINAHL, PubMed, PsycINFO, SPORTDiscus, and Cochrane Library from Jan 1, 1970, to Feb 8, 2021, for studies only published in English. Search terms included “people experiencing homelessness” and “physical interventions”. We included quantitative and qualitative studies of people experiencing homelessness from high-income countries with physical activity interventions in any setting that was compared with or without standard of care, evaluating any health or wellbeing-related outcome—eg, cardiovascular, depression. The Joanna Briggs Institute's (JBI) critical analysis tools were used for quality assessment, with very poor-quality studies rejected. JBI mixed-methods systematic reviews methodology was followed, with qualitisation of quantitative data and convergent synthesis of evidence. The protocol is registered with PROSPERO, number CRD42020216716. FindingsWe screened 2837 records, generating 14 reports (13 studies: eight qualitative and five quantitative [one randomised control trial, three quasi-experimental, one analytical cross-sectional]) of sufficient quality from the UK, USA, Denmark, and Australia, including 403 participants. Interventions included soccer (n=6), group activities (eg, gardening [n=5]), and multi-modal individual interventions (eg, bicycle provision with safety-training [n=2]). Quantitative studies reported 19 outcomes (13 objective, six subjective) quantifying physical health (eg, VO2max; n=13), mental health (eg, Beck Depression Inventory; n=5), and social wellbeing (Outcomes Rating Scale; n=1). Nine quantitative outcomes changed substantially in favour of the intervention. No change was reported in three outcomes, and the relative change of remaining outcomes could not be quantified. Synthesis of qualitative studies found themes reporting predominantly positive perceived impact on physical and mental health and social wellbeing, which corroborated the quantitative findings through convergent synthesis. InterpretationMore qualitative than quantitative studies were identified, with qualitative evidence suggesting physical activity interventions can benefit health and wellbeing. However, quantitative evidence was either positive or inconclusive. Evidence suggests a tentative recommendation for physical activity interventions in people experiencing homelessness. FundingNational Institute for Health Research School for Public Health Research.