基于广义相加模型的气温与带状疱疹发病风险的关联分析

尤美莹; 徐成东; 亓晓; 王天奇; 王苗苗; 刘欣雨; 李旭东; 符秋显; 胡跃华; 殷大鹏

doi:10.3784/jbjc.202312270696

基于广义相加模型的气温与带状疱疹发病风险的关联分析

Association between temperature and the risk of herpes zoster based on generalized additive model

摘要

摘要:
目的定量评估气温变化对带状疱疹患者日就诊数的短期影响及滞后效应。
方法收集2016—2021年北京市、2017—2022年湖北省宜昌市带状疱疹患者每日就诊的数据及同期气象监测资料。运用广义相加模型，控制长期趋势、短期波动、星期几效应、相对湿度等气象混杂因素，计算因气温改变引起的带状疱疹患者超额风险值，分析气温与带状疱疹发病风险的关联。
结果本研究共纳入515393例带状疱疹患者。日均气温与带状疱疹日均就诊数呈正相关（P<0.001）。日均气温每升高1 ℃，北京市和宜昌市带状疱疹日均就诊数分别增加0.30%95%置信区间(CI): 0.20%～0.40%、1.03%（95%CI：0.85%～1.21%）。滞后效应结果显示，北京市和宜昌市分别在滞后第11天、第4天超额风险达到峰值0.41%（95%CI：0.27%～0.56%）、1.09%（95%CI：0.91%～1.28%）。分季节亚组结果显示，北京市和宜昌市分别在暖季条件下、冷季条件下效应值更高。
结论气温升高造成带状疱疹发病风险增高，且存在滞后效应。北京市和宜昌市气温效应结果存在区域差异。

Abstract:
Objective To quantitatively assess the short-term effects and lag effects of temperature change on the daily number of visits for herpes zoster (HZ).
Methods The daily number of visits for HZ in Beijing from 2016 to 2021 and in Yichang from 2017 to 2022 were collected, as well as daily meteorological surveillance data during the same period. The generalized additive model was used to calculate the excess risk of patients with HZ due to temperature change adjusted for secular trend, seasonal trend, day of week, relative risk and other meteorological confounders, and analyze the association between temperature and the risk of HZ.
Results A total of 515393 HZ cases were included during the study period. The GAM-based time series showed that outdoor temperature had statistically significant effects on the onset of HZ (P<0.001). The average daily number of visits for HZincreased by 0.30%95% confidence interval(CI): 0.20%−0.40% and 1.03% (95%CI: 0.85%−1.21%) in Beijing and Yichang, respectively, with an increase of 1 ℃ in average daily temperature. The results of lag effect showed that the excess risk in Beijing and Yichang reached a peak of 0.41% (95%CI: 0.27%−0.56%) and 1.09% (95%CI: 0.91%−1.28%) at lag 11 days and lag 4 days, respectively. The results of seasonal subgroups showed that the effect value was higher in Beijing and Yichang in warm and cold seasons respectively.
Conclusion Rising temperature resulted in the risk of HZ increase with lag effects. There are regional differences in the temperature effect results in Beijing and Yichang areas.

HTML全文

参考文献(21)

施引文献

资源附件(0)