Temporal change in cold and heat-related burden of mortality: an evidence of increasing heat impact in Iran.

Temporal change in heat and cold-related burden of mortality needs to be investigated in Iran. In this study, the burden between 2015 and 2022 was compared with 2008–2014 in eight cities of Kurdistan which is located to the west of Iran. The overall dose–response association between temperature and mortality by different years (i.e., interaction effect between year and temperature) was assessed using bivariate response surface model and the framework of generalized additive model (GAM). The cumulative dose–response as well as lag-response associations in the two periods was compared by a time-varying distributed lag non-linear model. The associations were pooled using a two-stage regression model in which average temperature and temperature range were adjusted to control for the heterogenicity between cities. The attributable fraction (AF) and number (AN) were estimated in each city to ease the interpretation. Compared to 2008–2014, heat and extreme heat caused more mortality than cold and extreme cold in 2015–2022; the relative risk (RR) values of heat in second period were significantly higher than first period, and the cumulative RR of the extreme heat was 1.56 (%95 CI: 1.23–1.98) and 0.88 (%95 CI: 0.66–1.18) in second and first periods, respectively. Also, the cumulative RR of extreme cold was 1.01 (0.87, 1.17) and 0.89 (0.76, 1.05) in second and first periods, respectively. In second period, the extreme heat values approximately caused a minimum of 100 deaths in Marivan and a maximum of 400 in Sanandaj. The AF in Baneh, Kamyaran, and Sanandaj was significantly higher than other cities. All subgroups' people were at risk of heat-related mortality in second period, and the cold had no significant impact in the period. Heat-related mortality was increased in recent years while cold had no significant impact. The results highlight the need for an adaptation or preventive strategy of heat-related mortality in the regions. [ABSTRACT FROM AUTHOR]