Gender and violent crime modify associations between greenspace and cardiovascular disease mortality in Philadelphia, PA.

Few studies have explored variability of associations between greenspace and cardiovascular (CVD) mortality according to demographic or neighborhood contextual factors. We estimated overall and sex-stratified associations between greenspace and CVD mortality rates in Philadelphia, PA, and quantified effect modification of the sex-stratified associations by neighborhood violent crime rates. Sex- and age-stratified census tract CVD mortality rates (years 2008-2015) were linked with proportion tree canopy cover, grass/shrub cover, and total vegetation cover, and proportion of adult residents reporting access to a park. We used multivariable negative binomial models to estimate overall and sex-stratified associations between neighborhood greenspace and CVD mortality rates, and quantified effect modification by tract-level rates of violent crime. Higher proportion tree canopy cover and higher proportion adults reporting access to parks were associated with modestly lower rates of CVD mortality, with more pronounced estimates in males than females. In tracts with higher rates of crime, higher tree canopy coverage and perceived park access were associated with lower CVD mortality rates. We did not observe strong evidence of sex-based differences in interaction between neighborhood crime rates and greenspace. Results from this study reveal variability of associations between greenspace and CVD mortality rates according to sex categories, and according to neighborhood social environments. There is a need for further research exploring the extent to which differences in perceived safety explain gender-based differences in associations between greenspace and cardiovascular outcomes.