The Effect and Attributable Risk of Daily Temperature on Category C Infectious Diarrhea in Guangdong Province, China

Abstract

\n Previous studies have explored the effect between ambient temperature and infectious diarrhea (ID) mostly using the relative risk, which provides limited information in practical applications. There is limited study focused on the disease burden of ID caused by temperature, especially for different subgroups and cities in a multi-city setting. This study aims to estimate the effects and attributable risks of temperature on category C ID and explore potential modifiers among various cities in Guangdong. At first, the distributed lag non-linear models (DLNMs) were used to explore city-specific associations between daily mean temperature and category C ID from 2014 to 2016 in Guangdong and pooled them by applying the multivariate meta-analysis. Then, the multivariate meta-regression was further implemented to analyze the potential heterogeneity among various cities. Finally, we assessed the attributable burden of category C ID due to temperature, low (below the 5th percentile of temperature) and high temperature (above the 95th percentile of temperature) for each city and subgroup population. Comparing with the 50th percentile of daily mean temperature, adverse effects on category C ID were found when temperature was lower than 12.27℃ in Guangdong province. Some city-specific factors (longitude, urbanization rate, population density, disposable income per capita, and the number of medical technicians and beds per thousand persons) could modify the relationship of temperature-category C ID. During the study period, there were 60,505 category C ID cases (17.14% of total cases) attributable to the exposure of temperature, with the attributable fraction (AF) of low temperature (4.23%, 95% empirical confidence interval (eCI): 1.79%-5.71%) was higher than high temperature (1.34%, 95%eCI: 0.86%-1.64%). Males, people under 5 years and workers appeared to be more vulnerable to temperature, with an AF of 29.40%, 19.25%, and 21.49%, respectively. The AF varied substantially at the city level, with the largest AF of low temperature occurred in Shaoguan (9.58%, 95%eCI: 8.36%-10.09%), and that of high temperature occurred in Shenzhen (3.16%, 95%eCI: 2.70%-3.51%). Low temperature was an important risk factor for category C ID in Guangdong province, China. The exposure-response relationship could be modified by city-specific characteristics. Considering the whole population, the attributable risk of low temperature was much higher than high temperature, and males, people under 5 years and workers were vulnerable populations.