The association between green space and cause-specific mortality in urban New Zealand: an ecological analysis of green space utility

This was an observational study on a population of 1,546,405 living in 1009 small urban areas in New Zealand. A neighbourhood-level classification was developed to distinguish between usable (i.e., visitable) and non-usable green space (i.e., visible but not visitable) in the urban areas. Negative binomial regression models were fitted to examine the association between quartiles of area-level green space availability and risk of mortality from cardiovascular disease (n = 9,484; 1996 - 2005) and from lung cancer (n = 2,603; 1996 - 2005), after control for age, sex, socio-economic deprivation, smoking, air pollution and population density.Deprived neighbourhoods were relatively disadvantaged in total green space availability (11% less total green space for a one standard deviation increase in NZDep2001 deprivation score, p < 0.001), but had marginally more usable green space (2% more for a one standard deviation increase in deprivation score, p = 0.002). No significant associations between usable or total green space and mortality were observed after adjustment for confounders.Contrary to expectations we found no evidence that green space influenced cardiovascular disease mortality in New Zealand, suggesting that green space and health relationships may vary according to national, societal or environmental context. Hence we were unable to infer the mechanism in the relationship. Our inability to adjust for individual-level factors with a significant influence on cardiovascular disease and lung cancer mortality risk (e.g., diet and alcohol consumption) will have limited the ability of the analyses to detect green space effects, if present. Additionally, green space variation may have lesser relevance for health in New Zealand because green space is generally more abundant and there is less social and spatial variation in its availability than found in other contexts.Whilst individual characteristics are undoubtedly an important determinant of population health in an a