Peer-reviewed article
Neighbourhood environment and cardiorespiratory fitness in youth: A cross-sectional random forest analysis.
Dorsa Salimi, Tibor Schuster, Roseane de Fátima Guimarães, Andraea Van Hulst, Mélanie Henderson et al. ·
Summary
Written by AI, not yet AICS's view · claude-haiku-4-5-20251001This study examined whether neighbourhood environmental features predict cardiorespiratory fitness in children aged 8-10 in Montreal, Canada. Researchers measured children's peak oxygen consumption during a cycling test and assessed neighbourhood characteristics through audits and geographic mapping. The analysis found that individual factors like BMI, sex, and physical activity were stronger predictors of fitness than neighbourhood features, which contributed only modestly to the model.
Key points
- The random forest model explained 61.2% of variance in cardiorespiratory fitness when body weight was factored in
- Body mass index z-score, sex, and moderate-to-vigorous physical activity were the top three predictors of fitness
- Neighbourhood features added only 5.7 percentage points additional predictive value beyond individual characteristics
- Study involved 504 children aged 8-10 years in the Greater Montreal Area
- Cardiorespiratory fitness was measured objectively using peak oxygen consumption during incremental cycling
Limitations
This is a cross-sectional analysis, so it cannot establish causal relationships between neighbourhood features and fitness or determine temporal sequence. The authors note their findings are exploratory and call for longitudinal and intervention studies to confirm the relationship.
Abstract
Cardiorespiratory fitness (CRF) in youth is associated with cardiometabolic outcomes and risk of early-onset cardiovascular disease. While the role of the neighbourhood environment in supporting physical activity is recognized, its contribution to CRF requires further investigation. We examined the extent to which a range of neighbourhood features predict youth CRF. Data were from the baseline visit of the Quebec Adipose and Lifestyle Investigation in Youth (QUALITY) cohort study (n = 504, ages 8-10 years, Greater Montreal Area, Canada). CRF was measured as peak oxygen consumption during an incremental cycling test on an electromagnetic bike, expressed relative to body weight as the primary outcome (VO2peak/kg; mL·kg-1·min-1) and relative to fat-free mass as a sensitivity analysis (VO2peak/FFM; mL·kg FFM-1·min-1). Neighbourhood features were assessed through on-site audits and a Geographic Information System. Random forest modeling was applied to explore the predictive contribution of neighbourhood features alongside individual-level characteristics including sex, BMI z-score, Tanner stage, and moderate-to-vigorous physical activity (MVPA). This model explained 61.2% of the variance in VO2peak/kg and 40.7% in VO2peak/FFM based on out-of-bag predictions. The model identified BMI z-score, sex, and MVPA as the top three predictors, followed by several neighbourhood features. In a descriptive block-wise model comparison, out-of-bag variance explained for VO2peak/kg was 5.7 percentage points higher when neighbourhood variables were added to individual characteristics and MVPA. Overall, neighbourhood features suggested a modest contribution to predicting youth CRF. These findings are exploratory and warrant further investigation in longitudinal and intervention studies.