Zhou Bin, Gohlsch Katrin, Ranpal Surendra, Wang Jiancong, Knote Christoph
Chair of Model-Based Environmental Exposure Science, Faculty of Medicine, University of Augsburg, Augsburg, Germany.
J Epidemiol Glob Health. 2025 Aug 30;15(1):113. doi: 10.1007/s44197-025-00459-x.
Air pollution is among the top five environmental risk factors for human health worldwide. However, our understanding of the physiological responses to PM exposure over medium-term lag periods remains limited. This study aims to examine the medium-term lag-response associations-using lagging time windows of up to 21 days-between PM exposure and all-cause mortality in Valencia and London from 2002 to 2006.
We used a time-stratified case-crossover design; building on the methodologies of Tobias et al. and Bhaskaran et al., we applied a fixed-effects conditional quasi-Poisson regression model to quantify the association between PM exposure and all-cause mortality. We also analyzed three different temporal lag methodological models for the exposure-mortality relationships.
We found distinct differences in the relative risk (RR) patterns of PM exposure and all-cause mortality. In Valencia, the RR varied significantly, with confidence intervals that were wider than in London, where the RR remained more stable, fluctuating closely around 1. Significant associations were observed at early lag periods in both cities, consistent with Tobias et al. Notably, Valencia showed a significant peak in RR at lag 14, which was not observed in London. Subgroup analysis in Valencia also indicated delayed effects in younger populations. Scenario 3 (cumulative lag model) is conceptually closer to the cumulative progression of health risks associated with PM exposure and produces higher RR estimates compared to Scenario 1 and 2.
This study highlights the critical importance of addressing medium-term lag-response associations and methodological variations in environmental epidemiology. The findings have important clinical and public health implications and offer insights for risk assessment, healthcare planning, and the development of policies to mitigate the health impacts of PM exposure.
空气污染是全球人类健康面临的五大环境风险因素之一。然而，我们对中期滞后时间段内暴露于细颗粒物（PM）的生理反应的了解仍然有限。本研究旨在考察2002年至2006年期间，瓦伦西亚和伦敦在长达21天的滞后时间窗内，PM暴露与全因死亡率之间的中期滞后反应关联。
我们采用了时间分层病例交叉设计；基于托拜厄斯等人和巴斯卡兰等人的方法，我们应用固定效应条件拟泊松回归模型来量化PM暴露与全因死亡率之间的关联。我们还分析了三种不同的时间滞后方法模型用于暴露 - 死亡率关系。
我们发现PM暴露与全因死亡率的相对风险（RR）模式存在明显差异。在瓦伦西亚，RR变化显著，其置信区间比伦敦更宽，伦敦的RR保持更稳定，在1左右紧密波动。两个城市在早期滞后时间段均观察到显著关联，这与托拜厄斯等人的研究一致。值得注意的是，瓦伦西亚在滞后14天时RR出现显著峰值，而伦敦未观察到。瓦伦西亚的亚组分析还表明年轻人群存在延迟效应。情景3（累积滞后模型）在概念上更接近与PM暴露相关的健康风险的累积进展，并且与情景1和2相比产生更高的RR估计值。
本研究强调了处理环境流行病学中中期滞后反应关联和方法学差异的至关重要性。这些发现具有重要的临床和公共卫生意义，并为风险评估、医疗保健规划以及制定减轻PM暴露对健康影响的政策提供了见解。
