Dept.Environmnetal Sciences, Informatics, and Statistics, Università Ca' Foscari di Venezia, Mestre, 30172, Italy; Fondazione Centro Euro-Mediterraneo Sui Cambiamenti Climatici, Regional Model and Geo-Hydrological Impacts (REMHI) Division, Caserta, 81100, Italy.
Barcelona Institute for Global Health (ISGlobal), Universitat Pompeu Fabra, CIBER Epidemiología y Salud Pública, Barcelona, 08003, Spain.
Environ Res. 2022 Nov;214(Pt 3):114082. doi: 10.1016/j.envres.2022.114082. Epub 2022 Aug 11.
To date, little is known about the temporal variation of the temperature-mortality association among different demographic and socio-economic groups. The aim of this work is to investigate trends in cold- and heat- attributable mortality risk and burden by sex, age, education, marital status, and number of household occupants in the city of Turin, Italy.
We collected daily time-series of temperature and mortality counts by demographic and socio-economic groups for the period 1982-2018 in Turin. We applied standard quasi-Poisson regression models to data subsets of 25-year moving subperiods, and we estimated the temperature-mortality associations with distributed lag non-linear models (DLNM). We provided cross-linkages between the evolution of minimum mortality temperatures, relative risks of mortality and temperature-attributable deaths under cold and hot conditions.
Our findings highlighted an overall increase in risk trends under cold and heat conditions. All-cause mortality at the 1st percentile increased from 1.15 (95% CI: 1.04; 1.28) in 1982-2006 to 1.24 (95% CI: 1.11; 1.38) in 1994-2018, while at the 99th percentile the risk shifted from 1.51 (95% CI: 1.41; 1.61) to 1.59 (95% CI: 1.49; 1.71). In relation to social differences, women were characterized by greater values in respect to men, and similar estimates were observed among the elderly in respect to the youngest subgroup. Risk trends by educational subgroups were mixed, according to the reference temperature condition. Finally, individuals living in conditions of isolation were characterized by higher risks, with an increasing vulnerability throughout time.
The overall increase in cold- and heat- related mortality risk suggests a maladaptation to ambient temperatures in Turin. Despite alert systems in place increase public awareness and improve the efficiency of existing health services at the local level, they do not necessarily prevent risks in a homogeneous way. Targeted public health responses to cold and heat in Turin are urgently needed to adapt to extreme temperatures due to climate change.
迄今为止，人们对不同人口统计学和社会经济群体之间的温度-死亡率关联的时间变化知之甚少。本研究的目的是调查意大利都灵市不同性别、年龄、教育程度、婚姻状况和家庭人口数的人群中与冷、热相关的死亡风险和负担的趋势。
我们收集了 1982 年至 2018 年期间按人口统计学和社会经济分组的每日时间序列温度和死亡率数据。我们将标准的拟泊松回归模型应用于 25 年移动子时段的数据子集，并使用分布式滞后非线性模型（DLNM）估计温度-死亡率关联。我们提供了最小死亡率温度演变、冷、热条件下的死亡率相对风险和归因于温度的死亡人数之间的交叉链接。
我们的研究结果突出了冷、热条件下风险趋势的总体增加。在第 1 百分位数的全因死亡率从 1982 年至 2006 年的 1.15（95%置信区间：1.04；1.28）增加到 1994 年至 2018 年的 1.24（95%置信区间：1.11；1.38），而在第 99 百分位数的风险从 1.51（95%置信区间：1.41；1.61）转移到 1.59（95%置信区间：1.49；1.71）。关于社会差异，女性的数值大于男性，而与最年轻的亚组相比，老年人的估计值也相似。根据参考温度条件，教育亚组的风险趋势存在差异。最后，独居的人面临更高的风险，随着时间的推移，脆弱性不断增加。
冷、热相关死亡风险的总体增加表明都灵市对环境温度的适应不良。尽管现有的警报系统提高了公众意识并提高了当地卫生服务的效率，但它们不一定能以相同的方式预防风险。都灵迫切需要针对冷、热天气的公共卫生应对措施，以适应气候变化引起的极端温度。
