Associations of greenspace, including private residential gardens, with COVID-19 laboratory-confirmed positive test, hospitalisation and mortality in UK Biobank

Abstract

Background and aim: Inequities in access to urban greenspace, such as gardens and parks, were highlighted by the COVID-19 pandemic, which underscored greenspace’s utility for exercising and socialising. Initially, ecological studies suggested a protective association of greenspace with COVID-19 outcomes. The evidence, however, remains limited and lacking adequate adjustment for COVID-19 risk factors. We aimed to investigate associations of greenspace with COVID-19 positive test, hospitalisation and mortality using the UK Biobank cohort study. Methods: COVID-19 positive laboratory-confirmed test results, hospitalisations, and deaths were ascertained via UK Biobank linkage to Public Health England and general practitioner records, Hospital Episode Statistics and Office for National Statistics death records, from March – December 2020. We used Ordnance Survey Greenspace, categorised by function (e.g., private garden, public park, etc.), to assess greenspace in a 100 m distance buffer around residential addresses of UK Biobank participants who resided in urban areas of England (n = 276,805). We used logistic regression models to assess associations of greenspace with COVID-19 outcomes with adjustment for relevant COVID-19 risk factors and confounders, ascertained from baseline UK Biobank questionnaire responses (2006-2010) and general practitioner record linkage, and conducted multiple sensitivity analyses. Results: In age and sex-adjusted models, we found negative associations with all COVID-19 outcomes, however, in fully-adjusted models, associations were not significant. Unexpectedly, in fully-adjusted models, we found that both total greenspace (all types) cover and private residential garden cover within 100 m of addresses were positively associated with positive COVID-19 test (odds ratio (OR) 1.04 (95% confidence intervals (CI) = 1.01, 1.07); OR 1.08 (95% CI = 1.04, 1.13), respectively). Conclusion: We found no evidence that greenspace was associated with COVID-19 hospitalisation or mortality in fully-adjusted models. Results suggest higher COVID-19 positive test result for those residing at addresses surrounded by higher total greenspace and private garden cover.