Urban greenness and survival in lung cancer patients: A registry-based cohort study in Beijing

Previous studies have reported inverse associations between urban greenness and mortality. However, findings regarding contemporaneous and cumulative average greenness exposure are inconsistent. A potential explanation is the weak effects of greenness among the general population. This study assessed the associations between contemporaneous and cumulative average greenness exposure and survival in lung cancer patients.

An open cohort consisting of lung cancer patients enrolled from 1 January 2001 was studied. The cohort was followed up until 31 December 2018. We evaluated contemporaneous and cumulative greenness exposure using the normalized difference vegetation index (NDVI) within a 250 m buffer of each patient's geocoded residential address. The association between NDVI and mortality was evaluated using both time-fixed and time-varying Cox models. Their difference was compared either.

In time-fixed Cox models, a 0.1-unit increment in contemporaneous and cumulative average NDVI was associated with 17% (95% confidence interval [CI], 15-18%) and 7% (95% CI, 5-8%) reductions in mortality rate, respectively. Time-varying Cox models consistently revealed significant associations, but their magnitudes were lower, at 2% (95% CI, 1-3%) for both contemporaneous and cumulative average NDVI. Compared to the time-fixed models, the time-varying ones exhibited better fit for the proportional hazard assumption in Cox regressions with more conservative results.

Urban greenness is positively associated with an increased chance of survival in lung cancer patients. Compared to time-fixed Cox models, the results of time-varying Cox models were more conservative. This study confirms the beneficial health effects of urban greenness, which may facilitate the improvement of public health by urban green space planning.