Total and cause-specific mortality attributable to cold spells in China: A multicity and multicounty study

Climate change has resulted in an increase in the frequency and intensity of extreme cold weather events, but few multicity or multicounty researches have explored the association between cold spells and mortality risk and burden. We collected daily data on climate, sociodemographic factors and mortality in 18 cities/counties across 11 geographical regions for the period of November to March 2014-2018. A distributed lag nonlinear model was used to examine the association between cold spells and mortality after adjustment for confounding factors. Twelve definitions of cold spells were used. Multi -meta regression analysis was applied to pool the impacts over different regions. Cold spells were significantly associated with all-cause mortality at lag 0-21 (CRR:1.38, 95%CI:1.21,1.57). In addition to respiratory diseases and circulatory system diseases, digestive, endocrine and nervous system diseases and injury were also affected by cold spells. The magnitude of the impacts of cold spells on mortality varied among the diseases investigated, with the highest risk estimate found for influenza and pneumonia (CRR: 2.00, 95%CI: 1.45, 2.76) and the lowest estimate found for injury (CRR: 1.26, 95%CI: 1.09,1.46). The fraction of all-cause mortality attributable to cold spells was 2.31% (95%CI: 0.90%, 3.46%). Among the regional differences, the attributable burden of all-cause mortality was higher in rural areas and subtropical monsoon climate zone, with attributable fractions of 2.85% (95%CI: 1.23%, 4.11%) and 3.36% (95%CI: 0.55%, 5.35%), respectively. Cold spells increased mortality from a range of diseases. Women, older adults and residents of rural areas and subtropical monsoon climate zone were more vulnerable to cold spells impacts. The findings may help to formulate preventive strategies and early warning response plans to reduce mortality burden of extreme cold events.