Temperature change has become an important factor affecting human health. Exploring the health effect of diurnal temperature difference (DTR), one of the important indicators to measure temperature change, is conducive to develop strategies for population adaptation to climate change.

This study aims to evaluate the effect of DTR on years of life lost (YLLs) in Hunan Province.

The exposure-response relationship between DTR and years of life lost rate (YLLsR, per 100 000 population) among 69 counties from 2013 to 2017 in Hunan Province was evaluated by distributed lag nonlinear model (DLNM). The pooled effect was estimated by multivariate meta-analysis. Selected category-specific disease burdens due to DTR were estimated including warm or cold season, gender, age groups and DTR intervals.

Over 50 million population was covered in this study, and the mean daily YLLsR of non-accidental deaths was (22.6±13.4)/10⁵, and the median DTR in warm season (from May to October) (8.1℃) was higher than that in cold season (from November to next April) (6.4℃). In warm season, the pooled relationship between DTR and attributable YLLsR was nearly U-shaped, and the attributable fractions (AF) of YLLs increased by high level of DTR in total non-accidental deaths (DTR ≥ 14.3℃) and respiratory system disease deaths (DTR ≥ 15.8℃) were 0.2% (95% CI:0.0%-0.05%) and 0.1% (95% CI:0.0%-0.3%), respectively. Women and seniors (≥ 65 years old) were the vulnerable subgroups in warm season. In cold season, the pooled relationship between DTR and attributable YLLsR was nearly J-shaped, and the minimum attributable YLLsR diurnal temperature differences (MYD) were found between 2.1-2.2℃. The attributable YLLs of cold season were higher than that of warm season, and both DTR<MYD and DTR>MYD increased residents' YLLs from total non-accidental deaths and respiratory system disease deaths. The AF of total non-accidental deaths in each subgroup was significantly larger than zero in cold season. The AF of men (13.8%, 95% CI:4.5%~23.0%) was larger than that of women (12.3%, 95% CI: 1.7%-22.9%). The attributable YLLsR among the ≥ 65 years old subgroup13.1 (95% CI:2.7-23.6)/10⁵ was significantly higher than that of the 0-64 years old subgroup1.8 (95% CI:0.5-3.1)/10⁵, but the AF of the latter14.6% (95% CI:4.1%-25.0%) was larger.

The change of DTR exposure would increase the YLLs of residents in Hunan Province. Women and the elderly are vulnerable groups in warm season, while general residents in cold season are generally affected, and the adverse effect size in cold season is larger than that in warm season.