樊琳, 顾清, 曾强. 广义相加模型在大气污染流行病学研究中的应用进展[J]. 环境与职业医学, 2019, 36(7): 676-681. DOI: 10.13213/j.cnki.jeom.2019.18744
引用本文: 樊琳, 顾清, 曾强. 广义相加模型在大气污染流行病学研究中的应用进展[J]. 环境与职业医学, 2019, 36(7): 676-681. DOI: 10.13213/j.cnki.jeom.2019.18744
FAN Lin, GU Qing, ZENG Qiang. Progress in the application of generalized additive model in epidemiologic studies on air pollution[J]. Journal of Environmental and Occupational Medicine, 2019, 36(7): 676-681. DOI: 10.13213/j.cnki.jeom.2019.18744
Citation: FAN Lin, GU Qing, ZENG Qiang. Progress in the application of generalized additive model in epidemiologic studies on air pollution[J]. Journal of Environmental and Occupational Medicine, 2019, 36(7): 676-681. DOI: 10.13213/j.cnki.jeom.2019.18744

广义相加模型在大气污染流行病学研究中的应用进展

Progress in the application of generalized additive model in epidemiologic studies on air pollution

  • 摘要: 大气污染流行病学研究中探讨大气污染与人群健康效应的暴露-反应关系是一个研究难点,已成为环境卫生工作者面临的一个重要挑战。大气污染物的急性暴露与人群呼吸系统、心血管系统、中枢神经系统等疾病的关系已经被国内外大量研究所证实。时间序列模型是目前研究大气污染急性健康效应最常用的方法,其最主要的优点是对同一研究人群进行反复观察,因此可控制与时间相关变量(例如季节和长期趋势)造成的混杂效应。目前应用最广泛的时间序列模型是广义相加模型(GAM),GAM可以同时评估环境因素对健康效应的线性及非线性关联,可对各种混杂因素进行校正。本文通过对大气污染与人群健康效应关系中GAM的建立及其应用进行综述,为进一步研究提供方法学线索。在今后的研究中需要结合多种研究方法,将空间分析与时间分析结合,开发多水平的空间时间序列模型,充分考虑并控制各种可能对健康结局造成影响的混杂因素,更加准确地评估大气污染所造成的人群健康风险,为相关环境健康政策的制定提供依据。

     

    Abstract: Establishing an exposure-response relationship between air pollution and public health is a difficult problem in the epidemiology of air pollution, and solving this problem has become an important challenge for environmental health researchers. The relationships between acute exposure to atmospheric pollutants and diseases related to respiratory, cardiovascular, and central nervous systems have been evaluated by a large number of studies both at home and abroad. The time-series model is the most commonly used method to study the acute health effects of air pollution, owing to the repeated observation of the same research population and therefore the controllable confounding effects of time-related variables (such as seasonal and long-term trends). The generalized additive model (GAM) has been the most widely used in the time-series model family to simultaneously evaluate the linear and non-linear relationships between air pollution and population health allowing adjustment for various confounding factors. This paper summarized the establishment and application of GAM in air pollution and public health studies, aiming to provide methodological support for further evaluations. In future research, it is necessary to combine various research methods, such as combining spatial analysis with temporal analysis, and develop multi-level spatial and temporal series models after taking into consideration of various confounding factors for health outcomes, so as to accurately assess the population-level health risks caused by air pollution, and ultimately provide scientific evidence for the formulation of environmental health policies.

     

/

返回文章
返回