Characteristics of parks associated with depression in women only: a cross-sectional study of 329,363 adults.

backgroundTo examine associations between public park characteristics within different walking distances from residential locations and depression, to distinguish between features within parks (e.g. amenities, attractions, facilities, tree cover) and park metrics in the home area (e.g. number of parks, size, and total area), and to employ rigorous geospatial analysis linking the best available objectively measured park and urban green space (UGS) exposures to validated depression outcomes across multiple scales.

methodsThis population-based cross-sectional study utilised baseline data from 329,363 UK Biobank participants resident in urban areas. Prevalent diagnosed depression was defined as an ICD-10 code of F32 (depressive episode) or F33 (recurrent depressive disorder). Park characteristics and urban green space data were derived from Ordnance Survey Great Britain datasets and spatially linked to participants' residential addresses. Three definitions of Home Catchment Area size were tested for every individual respondent: 400 m (m), 800 m, and 1600 m, as proxies for a 10-,20- and 40-min return walk respectively. Logistic regression models assessed associations with robust statistical approaches including assessment of interaction, correction for multiple testing, confounder adjustment, and sensitivity analyses.

resultsSpecific park characteristics within 20-min and 40-min catchments were associated with reduced depression likelihood among women only. Within 40-min catchments, protective associations were observed for recreational amenities (cafés: odds ratio (OR) 0.89, 95% confidence interval (CI) 0.85-0.93; toilets: OR 0.85, 95% CI 0.79-0.91), attractions (OR 0.83, 95% CI 0.80-0.87), sports facilities (OR 0.84, 95% CI 0.79-0.90), and tree canopy coverage (e.g. > 20%, OR 0.88, 95% CI 0.85-0.91). In a 20-min catchment, each 1% increase in urban greenspace classified as parks was associated with 11% reduced depression odds among women (OR 0.89, 95% CI 0.82-0.95). No significant protective associations were observed among men, with some paradoxical adverse associations identified.

conclusionsThis study provides robust evidence for protective associations between park characteristics and depression among women, but not men. Findings support proximity-based planning concepts but challenge the current policy and practice focus on 20-min neighbourhood and identify park features which optimise preventive potential. Results have direct implications for evidence-based urban planning policy internationally, providing a framework for developing mental health-supporting green infrastructure that recognises sex-based differences.