Associations of ambient temperature with the CO poisoning risk in China

Although studies have explored the relationship between temperature and CO poisoning, the results are not consistent, and there is still a lack of early warning criteria of temperature related to CO poisoning. In order to comprehensively study the exposure-response relationship between daily average temperature and CO poisoning, and to further explore the early warning criteria of temperature related to CO poisoning, we used daily cases of CO poisoning in 31 National Injury Surveillance System (NISS) surveillance sites in seven administrative geographical regions of China and daily meteorological data obtained from the China Meteorological Science Data Sharing Service Platform from 2009 to 2019 to do the analysis. Daily meteorological data of 698 weather stations across China were interpolated at a 0.01° × 0.01°spatial resolution, which were then applied to extract the daily meteorological data of all included NISS sites. The Distributed Lag Non-linear Model (DLNM) model was applied to estimate the exposure-response associations (relative risk, RR) of daily mean temperature with CO poisoning, which was then further used to identify early warning criteria of temperature related to CO poisoning. A total of 10,618 CO poisoning cases were included in this study, with an average of 0.4 cases per day. There was generally a reverse J-shaped association between temperature and CO poisoning risk, indicating that both low and high temperature may increase the risk of CO poisoning, but low temperature usually has a longer lagged effects than high temperature. Spatialy, the exposure-response associations between temperatue and CO poisoning largely varied among regions, with greater effects of low temperatures in Southern China than in Northern China. The cumulative effects (RR, lag0-6 days) of 10 % percentile temperature ranged from 1.13 (95%CI: 1.01,1.26) in East China to 1.73 (95%CI:1.63,1.83) in South China. We also observed significant spatial variations in the early warning criteria of temperature related to CO poisoning across China. However, the patterns of high temperature effects on CO poisoning and the warning criteria of high temperature were mixed across China. In conclusions, both low temperature and high temperature may increase the risk of CO poisoning in China, and the effect of low temperature is more obvious, especially in South China, Northeast China, and North China. In addition, there is an urgent need to establish air temperature early warning and grading criteria for CO poisoning in different areas of China.