2015-2021年海口市和乌鲁木齐市大气污染物与肺癌发病例数的时间序列分析

Correlation of lung cancer incidence with air pollution in Haikou and Urumqi from 2015 to 2021:a time series analysis

目的 探讨海口市和乌鲁木齐市大气污染物与肺癌发病的相关性及污染物对肺癌发病的滞后效应。方法 收集海口市和乌鲁木齐市2015年1月1日-2021年12月31日之间的PM_(2.5)、PM_(10)、NO_(2)、SO_(2)、CO、臭氧日最大8 h平均浓度(O_(3-8h))、气象因素以及居民肺癌发病资料,用时间序列的广义相加泊松回归模型,控制长期趋势、气象因素和星期几效应等因素的影响后,定量分析单污染物模型和双污染物模型中大气污染物与肺癌发病的关联性。结果 在单污染物模型中,海口市PM_(10)在滞后当天每升高10μg/m~3引起肺癌日发病数的超额危险度(ER)最大,为6.33%(95%CI:2.29%~10.53%);SO_(2)在lag6时效应值最大,为41.82%(95%CI:3.19%~94.91%);NO_(2)在lag6时效应值最大,为18.43%(95%CI:7.97%~29.89%);CO在滞后当天效应值最大,即每升高1 mg/m3引起肺癌日发病数的ER为80.08%(95%CI:20.53%~169.06%)。乌鲁木齐市PM_(10)、NO_(2)和CO暴露对肺癌每日发病例数的效应值在lag02、lag04和lag02最大,ER值分别为1.30%(95%CI:0.56%~2.05%)、15.40%(95%CI:8.33%~22.94%)和23.84%(95%CI:7.29%~42.94%)。在双污染物模型中,海口市在引入PM_(10)后CO的效应值消失,而分别引入其他污染物后PM_(10)、SO_(2)、NO_(2)和CO对肺癌发病的影响效应仍有统计学意义(P<0.05);乌鲁木齐市PM_(10)、NO_(2)和CO分别在调整其他污染物后对肺癌发病的影响效应仍有统计学意义(P<0.05)。结论 大气污染物暴露可导致居民肺癌发病风险增加,且在不同城市、不同滞后天数下存在一定差异。

Objective To understand the correlation between air pollutants and lung cancer in Haikou and Urumqi,and the lag effect of air pollutants on lung cancer.Methods The data of fine particles(PM_(2.5)),inhalable particles(PM_(10)),NO_(2),SO_(2),CO and ozone daily maximum 8 h average concentration(O_(3-8 h)),meteorological factors and lung cancer incidence in Haikou and Urumqi from January 1,2015 to December 31,2021 were collected.The generalized additive Poisson regression model of time series was used to control the influence of long-term trend,meteorological factors and day of week effect,and the correlation between air pollutants and lung cancer in single-pollutant model and double-pollutant model was quantitatively analyzed.Results In the single-pollutant model,the excess risk(ER)of daily incidence of lung cancer caused by every 10μg/m^(3) increase of PM_(10)in Haikou was the largest on the lag day,which was 6.33%(95%CI:2.29%-10.53%).The maximum effect value of SO_(2) was 41.82%(95%CI:3.19%-94.91%)at lag6,the maximum effect value of NO_(2) was 18.43%(95%CI:7.97%-29.89%)at lag6,and the maximum effect value of CO was on the lag day.The ER was 80.08%(95%CI:20.53%-169.06%)for every 1 mg/m3increase in lung cancer incidence.The effect values of PM_(10),NO_(2) and CO exposure on the daily incidence of lung cancer in Urumqi were the largest in lag02,lag04 and lag02,with ER values of 1.30%(95%CI:0.56%-2.05%),15.40%(95%CI:8.33%-22.94%)and 23.84%(95%CI:7.29%-42.94%),respectively.In the two-pollutant model,the effect value of CO disappeared after the introduction of PM_(10)in Haikou City,while the effect of PM_(10),SO_(2),NO_(2) and CO on lung cancer incidence remained statistically significant after the introduction of the other pollutants,respectively,and the effect of PM_(10),NO_(2) and CO on lung cancer incidence remained statistically significant after adjusting for the other pollutants,respectively,in Urumqi City.Conclusion Exposure to air pollutants can lead to an increase in the risk of lung cancer in local residents.This risk is different in different cities and different lag days.