Built environment correlates of physical activity in low- and middle-income countries: A systematic review.

Insufficient physical activity (PA) is the fourth major risk factor for many non-communicable diseases and premature mortality worldwide. Features of the built environment (BE) play a considerable role in determining population PA behaviors. The majority of evidence for PA-BE relationships comes from high-income countries and may not be generalizable to low- and middle-income countries (LMICs). We aim to systematically review the literature and assess the associations between perceived and/or objective BE characteristics and PA domains in LMICs. This review adopted a systematic search strategy for English language articles published between January 2000 and June 2019 from four electronic databases-Medline, Embase, Web of Science and PubMed-adhering to the PRISMA guidelines. Studies addressing the associations between self-reported and/or objective BE and PA were only included if they were conducted in LMICs, according to the World Bank classification list. Articles investigating PA-BE relationships across any age groups were included, and all study designs were eligible, except for qualitative studies and reviews. Thirty-three studies were included for evidence synthesis. Cross-sectional studies were the most prevailing study design (97%), revealing a notable gap in longitudinal PA-BE research in LMICs. A majority of the BE factors were not associated with different PA domains while others (e.g., density, proximity to services, aesthetics) exhibited an inconsistent association. Land-use mix diversity was positively associated with transport PA and the presence of recreation facilities resulted in an increase in PA during leisure-time. Increased safety from crime at night consistently increased total PA and walking levels. Research exploring the associations between BE attributes and PA behaviors in LMICs appears to be limited and is primarily cross-sectional. Longitudinal research studies with objective measures are needed for inferring well-grounded PA-BE causal relationships and informing the design of evidence-based environmental interventions for increasing PA levels in LMICs.