2379

Background: Urban design features, such as walkable neighbourhoods and greenspace, might protect human health by enabling healthy behaviours, including increased physical activity. By definition, neighbourhoods that are ‘walkable’ are high density, whereas green spaces (e.g., parks) are inherently low density environments. We aimed to assess the impact of greenspace cover on associations of neighbourhood walkability and physical activity. Methods: We created a novel ‘green walkability’ score that integrated greenspace cover (within a 1000m walkable network buffer) with standard walkability metrics – population density, street junction density and destination density. We applied the green ‘walkability score’ to the residential addresses of UK Biobank participants living in Greater London (n= 58,587). We used logistic regression models to assess associations of ‘green walkability’ versus standard walkability scores with physical activity outcomes, adjusted for age, sex, household income and neighbourhood deprivation. Results: Higher ‘green walkability’ was beneficially associated with active commuting behaviour, active non-commute transport, International Physical Activity Questionnaire (IPAQ) responses, and achieving weekly UK physical activity guideline recommendations. Associations of ‘green walkability’ with physical activity outcomes, however, were attenuated compared to standard walkability score associations. For example, difference in odds of walking for non-commute transport between the least walkable and most walkable quintile in UK Biobank were smaller for ‘green walkability’ (odds ratio (OR) 4.46; 95% CI 4.12, 4.84) than standard walkability (OR 6.37; 95% CI 5.85, 6.94). Conclusion: Our ‘green walkability’ approach highlights the need to account for environmental correlates of physical activity (e.g., density of features) in greenspace and physical activity assessments, and might partially explain inconsistencies in direction and size of effect in the greenspace-physical activity literature. Prioritising neighbourhood density and greenspace exposure, which are inversely related in space, to support active, healthy lifestyles, is an important urban design challenge.

2383

Background: Some cohort studies have shown a protective association of greenness and non-injury mortality, after adjustment for personal contextual and area-level confounding factors, though evidence in the UK is predominantly from ecological studies. We aimed to assess the association of greenspace and non-injury mortality in the UK, using rich confounder data available in the UK Biobank cohort. Methods: We used Ordnance Survey MasterMap Greenspace data, categorized into 18 greenspace functions (e.g., private garden, public park, allotment, cemetery, playing field, etc.) to assess associations of survival and total greenspace cover (18 categories combined) in multiple buffer sizes (100 m, 500 m, 1000 m) surrounding UK Biobank residential addresses (n = 277,236). We used Cox-proportional hazards models, with age as the underlying timescale, adjusted for sex, household income, pack years smoking, alcohol consumption and neighbourhood deprivation. Results: We showed protective associations for UK Biobank participants living in the most green compared to the least green quintile of total greenspace cover in a 100 m circular distance buffer for non-injury mortality (HR = 0.91; 95% CI = 0.84, 0.99). Results were consistent with previous studies, which have shown a small protective association, after adjustment for confounders. Function-specific analyses showed that protective associations of private (residential) garden and mortality remained when other functional greenspace categories were excluded from analysis. Across all buffer sizes, over 50% of total greenspace cover was classified as private (residential) garden, which has important implications for environmental equality. Conclusion: Further analyses of greenspace categories, grouped by function (e.g., private (residential) gardens versus public parks, playing fields and sports facilities) will offer novel insight into specific mechanistic pathways connecting greenspace and health (e.g., the contribution of physical activity), which is critical for designing and/or retrofitting urban environments for health.