Authors’ reply: We thank Dr Kawada for his interest in our study.^{Reference Perrino, Lombard, Rundek, Wang, Dong and Gutierrez1} We are particularly interested in the findings cited by Kawada on the relationship of greenness to all-cause mortality.^{Reference Ji, Zhu, Bai, Wu, Yan and Tang2} In our own work we have found that among US Medicare beneficiaries over 65 years of age living in Miami-Dade County, Florida, USA, greenness was associated with a broad range of chronic diseases that have the potential to contribute to all-cause mortality. Specifically, an increase in mean block-level Normalized Difference Vegetation Index (NDVI) greenness scores from –1 s.d. below to +1 s.d. above the mean was associated with reductions of 10% in hyperlipidaemia, 13% in hypertension and 14% in diabetes diagnoses.^{Reference Brown, Lombard, Wang, Byrne, Toro and Plater-Zyberk3} These findings have a remarkable population-level impact on health, equivalent to 49 fewer chronic conditions per 1000 individuals – or, the equivalent of a population who is 3 years younger at +1 s.d. NDVI when compared with –1 s.d. NDVI, even when controlling in all of these analyses by individual age, gender, race/ethnicity and neighbourhood income.^{Reference Brown, Lombard, Wang, Byrne, Toro and Plater-Zyberk3} Notably, these results are 50% stronger for the lowest income blocks, suggesting that for the people with the least resources, block-level greenness is so much more protective, possibly because of their inability to afford or have access to other types of healthy environments.^{Reference Brown, Lombard, Wang, Byrne, Toro and Plater-Zyberk3} Moreover, in follow-up analyses, our findings revealed that the same increase in mean-level greenness was also associated with 18% reductions in Alzheimer's disease,^{Reference Brown, Perrino, Lombard, Wang, Toro and Rundek4} 19% reductions in heart disease^{Reference Wang, Lombard, Rundek, Dong, Marinovic Gutierrez and Byrne5} and 20% reductions in stroke (details available from the authors on request).

It is also of interest that the greenness-to-health relationships appears to be robust across different ways of defining the variable, ranging from the block as in our case, to 250 m and 1250 m as in the work of Ji et al.^{Reference Ji, Zhu, Bai, Wu, Yan and Tang2} Urbanisation may play an important role in understanding greenness’ relationship to depression.

Longitudinal studies such as Ji et al^{Reference Ji, Zhu, Bai, Wu, Yan and Tang2} are important for assertions of causality. Indeed, our team currently is working on prospective longitudinal studies to evaluate the longitudinal impact of greenness including tree planting on health outcomes. The addition of randomised controlled studies and examination of place-based interventions is important for identifying and designing the most appropriate greening strategies for promoting the health of populations.

In summary, we agree that the findings from these and other research groups highlight potential approaches to improving the quality of life and health of older adults, as well as identifying mechanisms through which greenness may lead to better outcomes and inform interventions.