Evaluating urban greenspace characteristics and their associations with heart- and dementia-related disease prevalence

Urban greenspaces are increasingly linked to a range of health outcomes. However, evidence remains limited on variation in these associations by greenspace dimension and position in the outcome distribution, particularly when spatial dependence is considered. This study used spatial filter unconditional quantile regression to assess the associations of greenspace availability, network distance to the nearest greenspace, and total canopy coverage with age-standardised heart- and dementia-related disease prevalence at the Statistical Area Level 2 scale in Melbourne. The results showed that greenspace availability was significantly and negatively associated with heart-related disease prevalence across quantiles, and this association was generally maintained across alternative buffer sizes. Spatial filtering also reduced residual spatial autocorrelation in the heart-related disease models. In contrast, greenspace availability and canopy coverage showed no clear associations with dementia-related disease prevalence, while greater network distance to the nearest greenspace was associated with lower prevalence at some quantiles. In addition, the association between greenspace availability and dementia prevalence varied with community age structure. Overall, the associations with disease prevalence differed by greenspace characteristic and between the two health outcomes. These findings suggest the importance of considering spatial dependence, different dimensions of greenspace, and local demographic context in urban environmental health research.