Division of Environmental Health Sciences, School of Public Health, University of Minnesota Twin Cities, Minneapolis, MN, USA.
Department of Health, Environmental Health Division, Minnesota, St. Paul, MN, USA.
Risk Anal. 2018 Oct;38(10):2208-2221. doi: 10.1111/risa.12998. Epub 2018 Apr 10.
Emergency risk communication (ERC) programs that activate when the ambient temperature is expected to cross certain extreme thresholds are widely used to manage relevant public health risks. In practice, however, the effectiveness of these thresholds has rarely been examined. The goal of this study is to test if the activation criteria based on extreme temperature thresholds, both cold and heat, capture elevated health risks for all-cause and cause-specific mortality and morbidity in the Minneapolis-St. Paul Metropolitan Area. A distributed lag nonlinear model (DLNM) combined with a quasi-Poisson generalized linear model is used to derive the exposure-response functions between daily maximum heat index and mortality (1998-2014) and morbidity (emergency department visits; 2007-2014). Specific causes considered include cardiovascular, respiratory, renal diseases, and diabetes. Six extreme temperature thresholds, corresponding to 1st-3rd and 97th-99th percentiles of local exposure history, are examined. All six extreme temperature thresholds capture significantly increased relative risks for all-cause mortality and morbidity. However, the cause-specific analyses reveal heterogeneity. Extreme cold thresholds capture increased mortality and morbidity risks for cardiovascular and respiratory diseases and extreme heat thresholds for renal disease. Percentile-based extreme temperature thresholds are appropriate for initiating ERC targeting the general population. Tailoring ERC by specific causes may protect some but not all individuals with health conditions exacerbated by hazardous ambient temperature exposure.
当环境温度预计超过某些极端阈值时，会启动紧急风险沟通 (ERC) 计划，以管理相关的公共卫生风险。然而，在实践中，这些阈值的有效性很少得到检验。本研究旨在检验基于极端温度阈值（包括寒冷和炎热）的激活标准是否能捕捉到明尼阿波利斯-圣保罗大都市区全因和特定原因死亡率和发病率的升高风险。使用分布滞后非线性模型 (DLNM) 结合准泊松广义线性模型，得出每日最高热指数与死亡率（1998-2014 年）和发病率（急诊科就诊；2007-2014 年）之间的暴露-反应函数。考虑的具体原因包括心血管、呼吸、肾脏疾病和糖尿病。研究了六个极端温度阈值，分别对应于当地暴露历史的第 1-3 百分位数和第 97-99 百分位数。所有六个极端温度阈值均显著提高了全因死亡率和发病率的相对风险。然而，特定原因的分析显示出异质性。极端寒冷阈值可捕捉到心血管和呼吸系统疾病的死亡率和发病率风险增加，而极端炎热阈值可捕捉到肾脏疾病的死亡率和发病率风险增加。基于百分位的极端温度阈值适合启动针对普通人群的 ERC。针对特定原因的 ERC 可能会保护某些但不是所有因危险环境温度暴露而加重健康状况的人。
