Xiaowei Yi1, Zhaorui Chang2, Xing Zhao1, Yue Ma1, Fengfeng Liu2, Xiong Xiao3. 1. Department of Epidemiology and Biostatistics, West China School of Public Health and West China Fourth Hospital, Sichuan University, Chengdu, China. 2. Division of Infectious Disease & Key Laboratory of Surveillance and Early Warning on Infectious Disease, Chinese Centre for Disease Control and Prevention, Beijing, China. 3. Department of Epidemiology and Biostatistics, West China School of Public Health and West China Fourth Hospital, Sichuan University, Chengdu, China. Electronic address: xiaoxiong.scu@scu.edu.cn.
Abstract
BACKGROUND: Previous studies have thoroughly elucidated the exposure-response relationship between ambient temperature and hand, foot, and mouth disease (HFMD), whereas very little concern has been to the lag-response relationship and related key time points. OBJECTIVES: We aimed to clarify the temporal characteristics of the lag-response relationship between ambient temperature and HFMD and how they may vary spatially. METHODS: We retrieved the daily time series of meteorological variables and HFMD counts for 143 cities in mainland China between 2009 and 2014. We estimated the city-specific lag-response curve between ambient temperature and HFMD and related key time points by applying common distributed lag nonlinear models (DLNM) and Monte Carlo simulation methods. Then, we pooled the city-specific estimates by performing a meta-regression with the city-specific characteristics as meta-predictors to explain the potential spatial heterogeneity. RESULTS: We found a robust lag pattern between temperature and HFMD for different levels of temperatures. The temporal change of risk obtained its maximum value on the current day but dropped sharply thereafter and then rebounded to a secondary peak, which implied the presence of a harvesting effect. By contrast, the estimation of key time points showed substantial heterogeneity, especially at high temperature (the I2 statistics ranged from 47% to 80%). With one unit increase in the geographic index, the secondary peak would arrive 0.37 (0.02, 0.71) days later. With one unit increase in the economic index and climatic index, the duration time of the lag-response curve would be lengthened by 0.36 (0.1, 0.62) and 0.92 (0.54, 1.29) days, respectively. CONCLUSION: Our study examined the lag pattern and spatial heterogeneity of the lag-response relationship between temperature and HFMD. Those findings gave us new insights into the complex association and the related mechanisms between weather and HFMD and important information for weather-based disease early warning systems.
BACKGROUND: Previous studies have thoroughly elucidated the exposure-response relationship between ambient temperature and hand, foot, and mouth disease (HFMD), whereas very little concern has been to the lag-response relationship and related key time points. OBJECTIVES: We aimed to clarify the temporal characteristics of the lag-response relationship between ambient temperature and HFMD and how they may vary spatially. METHODS: We retrieved the daily time series of meteorological variables and HFMD counts for 143 cities in mainland China between 2009 and 2014. We estimated the city-specific lag-response curve between ambient temperature and HFMD and related key time points by applying common distributed lag nonlinear models (DLNM) and Monte Carlo simulation methods. Then, we pooled the city-specific estimates by performing a meta-regression with the city-specific characteristics as meta-predictors to explain the potential spatial heterogeneity. RESULTS: We found a robust lag pattern between temperature and HFMD for different levels of temperatures. The temporal change of risk obtained its maximum value on the current day but dropped sharply thereafter and then rebounded to a secondary peak, which implied the presence of a harvesting effect. By contrast, the estimation of key time points showed substantial heterogeneity, especially at high temperature (the I2 statistics ranged from 47% to 80%). With one unit increase in the geographic index, the secondary peak would arrive 0.37 (0.02, 0.71) days later. With one unit increase in the economic index and climatic index, the duration time of the lag-response curve would be lengthened by 0.36 (0.1, 0.62) and 0.92 (0.54, 1.29) days, respectively. CONCLUSION: Our study examined the lag pattern and spatial heterogeneity of the lag-response relationship between temperature and HFMD. Those findings gave us new insights into the complex association and the related mechanisms between weather and HFMD and important information for weather-based disease early warning systems.
Authors: Yibo Gao; Hongwei Wang; Suyan Yi; Deping Wang; Chen Ma; Bo Tan; Yiming Wei Journal: Int J Environ Res Public Health Date: 2021-05-05 Impact factor: 3.390