南京市大气臭氧浓度与肺癌死亡的时间序列分析

Association between ambient ozone and lung cancer mortality in Nanjing:a time-series analysis

目的 探讨大气臭氧浓度(O_(3-8 h))对南京市居民肺癌死亡风险的影响及滞后效应。方法 收集南京市2013-2020年每日因肺癌死亡人数,收集大气污染物包括O_(3)的每日最大8 h浓度、细颗粒物(PM_(2.5))、可吸入颗粒物(PM_(10))、二氧化氮(NO_(2))、一氧化碳(CO)和二氧化硫(SO_(2))的日均质量浓度,以及日均气温、日均气压、日均风速等气象数据。采用时间序列分析方法,用泊松分布和广义相加模型相连接的方法,控制长期和季节变化趋势、气象因素和“星期几效应”的影响,定量分析大气O_(3-8 h)对人群肺癌死亡的影响及滞后效应,计算大气O_(3-8 h)浓度每升高10μg/m^(3)时肺癌每日死亡数增加的超额危险度(excess risk,ER)及95%CI。并对不同性别、年龄和文化程度居民的肺癌死亡效应影响进行分层分析。结果 2013-2020年南京市因肺癌累计死亡24 107例,平均每日死亡8例,男性占72.20%,女性占27.80%;大气O_(3-8 h)全年日均浓度中位数为92.11μg/m^(3),大气O_(3-8 h)与PM_(2.5)、NO_(2)、CO呈负相关(P<0.05)。在单污染物模型中,大气O_(3-8 h)浓度每升高10μg/m^(3),在滞后1 d(lag1)时对肺癌死亡风险影响最大,ER值为0.50%(95%CI:0.10%~0.90%),其他滞后天数及累积滞后效应均无统计学意义。在双污染物模型中,当纳入其他污染物(PM_(2.5)、NO_(2)、CO、SO_(2))时,大气O_(3-8 h)在lag1时对肺癌死亡的影响均有不同程度变化,仍有统计学意义(P<0.05)。随着大气O_(3-8 h)浓度每升高,≥60岁、文化程度较低居民肺癌死亡影响的超额风险分别增加0.54%(95%CI:0.10%~0.97%)和0.47%(95%CI:0.05%~0.88%),均有统计学意义(P<0.05)。结论 大气O_(3)污染是南京市肺癌死亡影响的独立危险因素,特别是≥60岁、低文化程度的居民肺癌每日死亡人数随大气O_(3-8 h)浓度的增加而呈上升趋势,需重点关注。

Objective To understand the effect of atmospheric ozone(O_(3-8 h)) on lung cancer mortality in Nanjing City and its lag effects. Methods The data from 2013 to 2020 on daily lung cancer deaths in Nanjing City were collected,along with daily maximum 8-hour concentrations of ozone and other pollutants(PM_(2.5),PM_(10),NO_(2),CO,SO_(2))and the meteorological data. Time series analysis methods were used,employing a combination of Poisson distribution and a generalized additive model to control for long-term and seasonal trends,meteorological factors and weekday effects. The excess risk(ER) of lung cancer mortality per10 μg/m~3 increase in atmospheric O3-8 hwas calculated,and stratified analysis was performed by age,gender and education.Results Over the study period,Nanjing City recorded 24 107 lung cancer deaths,averaging 8 daily deaths,with males accounting for 72.20%. The median daily concentration of atmospheric O_(3-8 h)was 92.11 μg/m~3. O_(3-8 h)was negatively correlated with PM_(2.5),NO_(2),and CO(P<0.05). In the single-pollutant model,the highest impact of atmospheric O_(3-8 h)concentration per 10μg/m~3 increase on lung cancer mortality was observed at lag 1 day,with an ER of 0.50%(95%CI:0.10%-0.90%). In the twopollutant model,when other pollutants(PM_(2.5),NO_(2),CO,and SO_(2)) were included,atmospheric O_(3-8 h)showed varying degrees of changes in lung cancer mortality(P<0.05). The excess risks of lung cancer mortality for the residents aged over 60 and those with lower educational levels increased by 0.54%(95%CI:0.10%-0.97%) and 0.47%(95%CI:0.05%-0.88%),respectively,with significant differences(P<0.05). Conclusion Atmospheric O_(3-8 h) pollution is an independent risk factor for lung cancer mortality in Nanjing City,especially among the old residents and those with lower education levels,requiring special attention.