Objective To understand the impact of extreme temperatures on the risk of stroke mortality among residents in Nanchang city, Jiangxi province, from 2018 to 2022, and to provide data support for the formulation of corresponding preventive measures and health policies.

Methods Data on stroke mortality among Nanchang residents from January 1, 2018, to December 31, 2022, as well as data on air pollutants and meteorology in Nanchang during the same period, were collected from the Chinese Center for Disease Control and Prevention’s Population Mortality Registration Information System. A generalized additive model (GAM) combined with a distributed lag non-linear model (DLNM) was used to analyze the cumulative lag effect of extreme temperatures on the risk of stroke mortality among residents. Subgroup analyses were conducted for different genders, ages, and stroke subtypes to identify sensitive populations.

Results A total of 23 630 stroke deaths were reported in Nanchang city from 2018 to 2022, including 12 269 hemorrhagic strokes and 11 361 ischemic strokes. The distribution relationship between daily average temperature and stroke mortality in Nanchang city showed an approximate "J" shaped curve, with the minimum mortality temperature (MMT) being 8.5 °C. DLNM analysis showed that the effect of extreme high temperature on the risk of stroke mortality reached its peak on the same day (lag0 d), with a cumulative relative risk (cumRR) of 1.328 (95%CI: 1.136–1.553), and the difference was not statistically significant after lag2 d (P>0.05). The effect of extreme low temperature on the risk of stroke mortality had a lag effect, reaching its peak at lag6 d, with a cumRR of 1.031 (95%CI: 1.013–1.049). Subgroup analysis showed that the distribution trends of the effects of extreme high temperature on the risk of stroke mortality in each subgroup were similar, and all reached their peak at lag0 d. The cumRR for males 1.389 (95%CI: 1.149–1.679) was higher than that for females 1.263 (95%CI: 1.027–1.553), the cumRR for residents <60 years old 1.166 (95%CI: 0.844–1.611) was lower than that for residents ≥60 years old 1.373 (95%CI: 1.158–1.630), and the cumRR for hemorrhagic stroke 1.316 (95%CI: 1.090–1.589) was higher than that for ischemic stroke 1.305 (95%CI: 1.059–1.601). Under extreme low temperature, the cumRR at lag5 d for males 1.033 (95%CI: 1.013–1.054) was higher than that at lag7d for females 1.032 (95%CI: 1.004–1.060), the cumRR at lag5 d for residents <60 years old 1.042 (95%CI: 1.005–1.079) was higher than that at lag6 d for residents ≥60 years old 1.028 (95%CI: 1.010–1.046), and the cumRR at lag6 d for hemorrhagic stroke 1.045 (95%CI: 1.015–1.076) was higher than that at lag6 d for ischemic stroke 1.026 (95%CI: 1.003–1.049).

Conclusions Extreme temperatures can increase the risk of stroke mortality in Nanchang city residents. The effect of extreme high temperature is acute and short-lived, while the effect of extreme low temperature has a certain lag. Males and residents with hemorrhagic stroke have a higher risk of stroke mortality under both extreme high and low temperatures. Residents <60 years old and ≥60 years old have a higher risk of stroke mortality under extreme low and high temperatures, respectively.