Article Text
Abstract
Background To investigate associations between the food and physical activity (PA) environments and obesity.
Methods Cross-sectional data (n = 22,889) from the Yorkshire Health Study were used. Body mass index (BMI) was calculated using self-reported height and weight; obesity was defined as BMI ≥ 30. Waist circumference (WC) was also self-reported; ‘at risk’ was defined as ≥94 cm and ≥80 cm for males and females respectively. Food outlets (FO) and PA facility locations were mapped using the Ordnance Survey Points of Interest database. Park locations were obtained separately from Open Street Map. Home neighbourhoods were defined using 2 km buffers radiating from each participant’s home postcode. FO or PA opportunities within home neighbourhoods were then summed to indicate availability. FO were categorised as ‘takeaway’, ‘supermarket’ and ‘other food retail’. PA facilities and parks were considered as two separate categories. Multi-level logistic models were used to estimate associations between the food and PA environment and obesity in separate models for each environmental variable. To account for the skewed environment data we modelled availability in quartiles (Q1 least exposed, Q4 most exposed). Age, gender, ethnicity, deprivation and rural/urban classification were included as covariates in all models.
Results For the food environment, 89.1% had immediate access to at least one takeaway and 69.9% to a supermarket. For the PA environment, 97.6% and 77.7% of individuals had one PA facility and park available within their home neighbourhood. Availability varied little by deprivation. Separate multi-level models showed no evidence of an association neither between (i) the number of FO, nor (ii) most PA factors and obesity in those most exposed (Q4) compared to those least exposed (Q1). However, there was evidence of a negative association between PA facilities and obesity in the most exposed quartile (Q4 OR = 0.88 [95% CI 0.78–0.98]) compared to those least exposed (Q1). Findings were substantively the same for BMI and WC.
Conclusion There is little evidence to suggest that availability within the food or PA environment is associated with obesity. The evidence presented here provides little support for policy interventions aiming to modify the local food or PA environment.