Recent evidence suggests a synergistic acute effect between temperature, ozone and particulate matter (PM) on premature mortality. Several studies reported higher air pollution-related mortality risks during warm days, and higher heat-related mortality risk during polluted days. We investigated if interactions between temperature and air pollution modified the mortality response to an extent that would support the need for joint heat and air pollution warning systems.

We developed a multicentre time-series design for 17 French cities for the period 2000–2015, investigating the influence of season and temperature on the air pollution (PM₁₀ and ozone)-mortality relationship, and the influence of air pollution on the temperature-mortality relationship.

Ozone and PM₁₀ mortality risks exhibit an increasing gradient between spring, summer and heat waves. For instance, a 10 μg/m³ increase in PM₁₀ was associated with a 3% [Confidence interval (CI) 95% 2.1:3.9] increase in mortality during summer, and with a 14.2% [CI 95% 5.6:23.4] increase in mortality during heat waves.

The heat-mortality response was slightly influenced by air pollution, especially during the most extreme heat waves.

Our results suggest that air pollution warnings should take season into account, using lower thresholds during summer and heat waves. Heat warning systems may not be improved by air pollution data, as its added value would be limited, compared to the complexity it would add to the warning systems. Efforts should be made to reduce ozone and PM₁₀ concentrations during heat waves, even when they are already below regulatory thresholds.

最近的证据表明，温度，臭氧和颗粒物（PM）对过早死亡具有协同的急性作用。几项研究报告说，在温暖的日子里，与空气污染相关的死亡风险更高，而在污染的日子里，与热量相关的死亡风险更高。我们调查了温度和空气污染之间的相互作用是否在一定程度上改善了死亡率响应，从而可以满足对联合供热和空气污染预警系统的需求。

我们为2000年至2015年期间的17个法国城市开发了多中心时间序列设计，调查了季节和温度对空气污染（PM ₁₀和臭氧）与死亡率的关系以及空气污染对温度的影响。死亡率关系。

臭氧，PM _10的死亡风险在春季，夏季和热浪之间呈现出越来越大的梯度。例如，10微克/米³增加PM ₁₀用3％[置信区间（CI）95％2.1：3.9]相关联的夏季期间的死亡率增加，并用14.2％[95％CI 5.6：23.4]增加热浪期间的死亡率。

空气污染对热死响应的影响很小，特别是在最极端的热浪中。

我们的结果表明，应在夏季和热浪期间使用较低的阈值来考虑空气污染警告。空气污染数据可能无法改善热警告系统，因为与增加警告系统的复杂性相比，其附加价值有限。即使在已经低于监管阈值的情况下，也应努力降低热浪中的臭氧和PM ₁₀浓度。