The contributions of neighbourhood design in promoting metabolic health

Cardiovascular diseases remain the leading cause of global mortality, with modifiable clinical risk factors (e.g., hypertension, hyperlipidemia, obesity, hyperglycaemia) prevalent in populations. Metabolic syndrome—a cluster of these risk factors—substantially increases the risk of cardiovascular disease and mortality and is common worldwide. Physical activity is a key behaviour for preventing and managing metabolic syndrome. The built environment can nudge populations toward health-supportive behaviours like physical activity without mandates, suggesting that neighbourhood design may influence metabolic health at scale. The study’s aim was to estimate associations between activity-friendly neighbourhood built characteristics and: (1) being classified as having metabolic syndrome and (2) the number of metabolic syndrome clinical risk factors among Canadian adults.

Conceptual frameworks link transportation systems, land use, pedestrian environments, and green space to chronic disease risk via behaviours such as physical activity, diet, pollution exposure, social interaction, and sedentary time. Systematic reviews and longitudinal studies indicate that improved access to destinations, active transportation infrastructure, and public transport support higher physical activity. Evidence connects built environment characteristics with cardiovascular risk factors, though relatively few studies have examined metabolic syndrome directly and none in Canadian populations in earlier reviews. Past studies in Australia, the UK, and China examined walkability indices or greenness (NDVI/VCF) in relation to metabolic syndrome, but findings are inconclusive, warranting further investigation in diverse contexts, including Canada.

Design: Cross-sectional analysis using data from Alberta’s Tomorrow Project (ATP), a prospective cohort in Alberta, Canada. Participants: Adults aged 35–69 years without prior cancer (except non-melanoma skin cancer) recruited 2000–2008, with follow-up surveys and in 2009–2015 clinic visits for physical measures and biosamples. Present analysis included n=6718 urban residents with complete data who completed the 2008 survey and had physical measures and biological samples. Outcomes: Metabolic syndrome defined per Adult Treatment Panel III as meeting ≥3 of: abdominal obesity (men ≥102 cm; women ≥88 cm), elevated triglycerides (≥1.7 mmol/L), reduced HDL-cholesterol (men <1.03 mmol/L; women <1.30 mmol/L), elevated blood pressure (systolic ≥130 or diastolic ≥85 mm Hg), elevated fasting glucose (≥6.1 mmol/L). As fasting glucose was unavailable, HbA1c ≥5.7% substituted for abnormal blood glucose. Trained staff measured waist circumference and blood pressure; non-fasting blood samples analyzed for lipids and HbA1c. Exposures: Residential greenness via NDVI (Landsat 5) and built attributes (dwelling density, intersection density, points of interest) from CANUE’s Can-ALE 2016 dataset, calculated within 1 km Euclidean buffers around dissemination area centroids and linked to participants by postal code. Points of interest from OpenStreetMap included destinations (e.g., schools, shops, businesses). NDVI and built variables were standardized (z-scores). An Active Living Environment (ALE) index was computed as the sum of z-scores for dwelling density, intersection density, and points of interest (NDVI excluded due to negative correlation with ALE and original index design). Covariates: Age, sex, marital status, education, ethnicity, employment status, annual household income, and current smoking status. Statistical analysis: Descriptive statistics were computed overall and by sex. Group comparisons used t-tests and chi-square tests. Covariate-adjusted regression models examined associations of each exposure (separate models) with (1) number of metabolic syndrome risk factors (reporting unstandardized coefficients β and 95% CI) and (2) presence of metabolic syndrome (logistic regression, reporting OR and 95% CI). Interactions with sex were tested; significant interactions prompted sex-stratified analyses. Complete-case analysis was used. Two-tailed tests with p<0.05; analyses in Stata 15.0.

Sample characteristics (n=6718; 66% women): Mean age 54.3 years (SD 9.4). Prevalence of metabolic syndrome: 34.3%. Mean number of risk factors: 1.95 (SD 1.4). Component prevalence: abdominal obesity 50.8%; elevated triglycerides 43.4%; reduced HDL 25.6%; elevated HbA1c 43.1%; elevated blood pressure 32.0%. Built environment exposures (means): NDVI 0.3 (SD 0.1); dwelling density 881.0/km² (SD 568.4); intersection density 30.7/km² (SD 14.7); points of interest 47.9/km² (SD 41.4); ALE index 0.1 (SD 2.5). NDVI slightly higher among women. Associations with the number of metabolic syndrome risk factors (β, 95% CI):

• Dwelling density: -0.05 (-0.09, -0.01)
• Intersection density: -0.03 (-0.08, 0.01) (not significant)
• Points of interest: -0.11 (-0.16, -0.07)
• ALE index: -0.03 (-0.05, -0.01)
• NDVI: -0.00 (-0.05, 0.04) (null) Associations with having metabolic syndrome (OR, 95% CI):
• Points of interest: 0.89 (0.84, 0.94)
• ALE index: 0.97 (0.95, 0.99)
• Dwelling density: 0.95 (0.90, 1.01) (not statistically significant)
• Intersection density: 0.95 (0.90, 1.00) (borderline)
• NDVI: 1.00 (0.95, 1.06) (null) Sex interaction: Significant only for points of interest regarding number of risk factors; associations were protective in both women (β -0.08, 95% CI -0.14, -0.03) and men (β -0.18, 95% CI -0.26, -0.10).

Findings indicate that activity-supportive neighbourhood design—captured by higher points of interest, greater dwelling density, and a higher ALE index—is associated with fewer metabolic syndrome risk factors and lower odds of metabolic syndrome. This aligns with prior evidence linking walkable urban form and access to destinations with greater physical activity and lower cardiometabolic risk. The ALE index results corroborate Canadian studies associating more supportive active living environments with higher physical activity and better weight-related outcomes. Points of interest and dwelling density may promote transport-related walking and active commuting, influencing metabolic health. No significant association was observed between NDVI-measured greenness and metabolic syndrome, contrasting with some studies; potential reasons include NDVI’s inability to capture greenspace quality, variation in exposure measurement and populations, and pathways dominated by transport-related rather than recreational physical activity. Overall, the results support urban design strategies that increase destination density and walkability to improve metabolic health at the population level.

Activity-friendly neighbourhoods appear beneficial for metabolic health. Living in areas with more destinations (points of interest), higher dwelling density, and higher overall active living environment scores was associated with fewer metabolic syndrome risk factors and lower odds of metabolic syndrome. Residential greenness (NDVI) showed no significant association with metabolic syndrome in this sample, suggesting the need to explore alternative greenness metrics. The evidence can inform policies and planning aimed at promoting population health, while recognizing that local context (e.g., climate, governance, culture) may influence the effectiveness of built environment interventions.

Cross-sectional design limits causal inference. NDVI may not capture quality, type, or usability of greenspaces, potentially underestimating associations with greenness. Built environment variables focused on those related to transportation physical activity; other relevant features (e.g., food environment) were not included. Potential residential self-selection could bias associations if healthier individuals choose more walkable neighbourhoods. Binary classification of metabolic syndrome may mask gradations of cardiometabolic risk among non-cases. Lack of data on medication use (e.g., for hypertension, hyperglycaemia, hyperlipidemia) may have led to misclassification and attenuation of associations. Generalizability is to urban residents in Alberta; external factors may differ elsewhere.