Objectives To identify the impact and the spatial distribution of PM2.5 exposure on daily mortality of residents in Guangzhou.
Methods PM2.5 concentration in the air, meteorological data and daily mortality of residents were collected from 31 atmospheric monitoring points in 12 districts of Guangzhou. Kriging analysis and time series analysis with distributed lag non-linear model based on conventional models (generalized linear model, generalized additive model) were used to explore the status of PM2.5 pollution, the impact and hysteretic effect of PM2.5 exposure on daily overall non-accidental death (T), cardiovascular death (I) and respiratory death (J) of residents.
Results The average concentration of PM2.5 in the air in Guangzhou was (49.5±25.6) μg/m3, which was higher in southwest region and lower in northeast region; the RRs of maximum cumulative effect on non-accidental death (T) was 1.017(1.001, 1.034). The RRs of daily best lag effect in 12 districts on T, I and J were 1.010 1.014, 1.006 1.021 and 1.004 1.034, respectively; the RRs of maximum cumulative effect on T, I and J were 1.010 1.057, 1.017 1.059 and 1.018 1.110, respectively.
Conclusions The impact of PM2.5 on daily death of local residentsin different regions of Guangzhou city was different. The population in both end of south and north part and west part of Guangzhou city were dense. There was a higher risk of PM2.5 in traffic congestion areas, but the spatial distribution characteristics in other areas were not obvious.