Disparities of weather type and geographical location in the impacts of temperature variability on cancer mortality: A multicity case-crossover study in Jiangsu Province, China.

Considering the serious health burden caused by adverse weather events, increasing researches focused on the relationship between temperature variability (TV) and cause-specific mortality, but its association with cancer was not well explored. We aimed to investigate the impacts of TV on cancer mortality and examine the modifying effects of weather type and geographical location as well as other characteristics. Daily city-specific data of cancer deaths, mean temperature (Tmean), maximum and minimum temperatures (Tmax and Tmin), relative humidity (RH), rainfall, and air pollutants were collected during 2016–2017 in 13 cities in Jiangsu Province, China. TV 0-t was defined as the standard deviation of the daily Tmax and Tmin on the exposure 0-t days. A two-stage analysis was applied. First, a time-stratified case-crossover design was used to examine the odds ratio (OR) and attributable fraction of cancer mortality per 1 °C increase in TV by adjusting for potential confounders. Random effect meta-analysis was used to summarize the pooled ORs. Second, stratified analysis was performed for weather type, geographical location, demographics, and other city-level characteristics. The weather was defined as four types according to days during warm or cold season combined with high or low RH. A total of 303670 cases were included in our study. Meta-analysis showed that the ORs of cancer mortality per 1 °C increase in TV 0-t significantly increased and peaked in TV 0-2 (OR=1.0098, 95% CI: 1.0039–1.0157). The attributable fraction of TV 0-2 on cancer mortality was 4.74%, accounting for 14395 deaths in the study period. Significant ORs of TV-related cancer mortality were found during the warm season combined with high RH and in the northern region of Jiangsu. Susceptible groups of TV-related cancer mortality were identified as female patients, patients aged 45–65 years, and those living in cities with lower per capita green area. TV can significantly increase the risk of cancer mortality, especially during warm and humid days and in the northern region of Jiangsu. Findings are of great significance to formulate urban planning, resource allocation, and health intervention to prolong the life of cancer patients. • Temperature variability (TV) significantly increased the risk of cancer mortality. • TV-related cancer deaths were higher on warm and humid days and northern region. • Females patients and patients aged 45–65 years were more susceptible to TV. • A high level of per capita green area decreased the impacts of TV on cancer death. [ABSTRACT FROM AUTHOR]