Impacts Of Temperature On The Incidence Of Hand, Foot And Mouth Disease:Based On Distributed Lag Non-linear Model

[BACKGROUND]Hand, foot, and mouth disease (HFMD) is an acute viral infection mainly caused by Coxsackievirus A16 and Enterovirus 71. Infants and children younger than 5 are more likely to acquire the infection. As one of the most serious epidemic areas, China was burdened with the control and prevention of HFMD. Since the first case was reported in Shanghai in 1981, the disease has spread to all the provinces in China so far. Specifically, Shandong Province has been suffering from serious hand, foot and mouth disease epidemic since 2007. The disease has posed a serious threat to the health of children and young adults. This has prompted concerns that, without intervention, the public health impact and spread of the disease will continue to intensify. Therefore, the prevention and control of hand, foot and mouth disease is of great urgency.In the context of climate change and globe warming, there has been increasing interest in assessing the impact of temperatures on the mortality and morbidity. However, the domestic researches concerning the temperature effect of hand, foot and mouth disease are limited, and previous studies examining temperature-disease associations of hand foot and mouth disease mainly focused on a single area. In addition, the conventional time series analysis in environmental epidemiological research usually ignore the delayed effect of exposure, which violates the biomedical plausibility. In recent, the development of distributed lag non-linearity model (DLNM) has proposed solution. Therefore, a multi-city study using this model is required to better understand the impacts of temperature on the childhood incidence of HFMD. The findings can be a practical reference for the early warning and intervention strategies of HFMD.[OBJECTIVE]1. To detect the three-dimension distribution of hand, foot and mouth disease in Shandong Province over the study period.2. To explore the impact of temperature on the incidence of hand, foot and mouth disease in each city.3. To examine whether the temperature-morbidity association varied by place and estimate the pooled effect of temperature on childhood hand, foot and mouth disease incidence.4. To identify the potential effect modifier of temperature-morbidity relationship.[METHODS]1. Data Source:The daily reported morbidity data of HFMD from 1 January 2007 to 1 December 2012 were obtained from the China Information System for Disease Control and Prevention, including the basic information of HFMD cases, and the pathogen type of some laboratory-confirmed cases. Data on daily meteorological variables, including temperature, relative humidity, rainfall, and sunshine duration, were collected from the China Meteorological Data Sharing Service System. The corresponding socio-economic and demographic information of the 17 cities between 2007 and 2012 were obtained from Shandong Statistical Yearbook.2. Descriptive Study:The incidence rate, gender-related and occputation-related constituent ratio, etiological constituent of hand, foot and mouth disease in Shandong Province will be demonstrated. Epidemic curve and disease mapping will be used to detect the high-incidence time and place.3. Temperature-disease Association Study:A time-stratified distributed lag non-linear model with quasi-Poisson distributionu will be used to estimate city-specific effects of temperature on the reported hand foot and mouth disease incidence after controlling potential confunders, seasonal trend and long-term trend.4. Multivariate Meta Analysis:The city-specific estimates will be combined through multivariate meta-analysis to reveal the pooled exposure-response relationship. We then will include some location-specific indicators, including region factor, socio-economic and demographic conditions as meta-predictors to further detect the effect modifiers.[RESULTS]1. Children under 5 years accounted for the majority of hand, foot and mouth disease cases in Shandong Province. A total of 504017 childhood hand, foot and mouth disease cases (0-5 years) were included in the study, which made up 93.84% of total cases. The annual average incidence for children under 5 years was 141.65 per 10 000. The male cases were more than female, and the male-to-female sex ratio was 1.72. According to the statistics, most childhood cases were scattered children and nursery children; EV 71 and Cox A16 were the major pathogens of hand, foot and mouth disease. The dominant pathogen was EV71 (39.86%), followed by CoxA16 (36.66%).2. HFMD demonstrated a clear seasonal trend and the high-incidence periods were roughly consistent, from April to June. The southwest normally reached the incidence peak earlier than the northeast. The disease mapping shows that in 2007 and 2008 the high incidence area mainly clustered in Qingdao, Linyi and Jinan, and the incidence rates in the other cities are relatively lower. In 2009. the disease spreaded throughout the whole province and the incidence of the 17 cities were all above 100 per 10000. In the year of 2010 and 2011, the high incidence area mainly clustered in north and east of Shandong Province. Then the incidence showed a relatively decreasing trend and became more scattering in 2012.3. The temperature-disease associations of the 17 cities demonstrated heterogeneity while the neighboring cities showed some similar trend. Most cities showed an inverted V-shape:within a suitable range, the relative risk increase with the temperature, while after a certain cut-off temperature the relative risk will decrease. In some other cities, such as the north part of the peninsular area (Yantai, Weihai and Dongying) and central region (Laiwu, Taian and Zibo), the risk of incidence approximately showed a linear growth with temperature.4. The pooled result indicates that the effect of temperature on HFMD incidence was an approximately inverted V-shape; the peak RR value was 1.40 (95%CI:1.15-1.69) at 23.2 ℃. According to the I2 statistic, the heterogeneity of temperature-disease relationship between cities is up to 98.31%. Regional indicator, numbers of healthcare institution and annual household income were considered as potential effect-modifiers.[DICUSSION]1. Shandong province has a high incidence rate and children under 5 years old are the high-risk population for HFMD. This infection has posed a threat to the children’s health, and the prevention and control of HFMD should be strengthen. April to July is a high-risk period of HFMD; the health sectors should strengthen the disinfection of places where children congregate, such as nursery, primary school within the high-incidence period. The spatial epidemiological characteristics of HFMD indicated that the hotspots of HFMD wandered around in whole Shandong. Thus, it is important to establish a real-time surveillance system to detect outbreak region and conduct prevention to HFMD timely.2. Temperature has significant association with the HFMD incidence in study area. Patterns of temperature-disease association are likely to be driven by a complex interaction of pathogens, environmental factors and host population. Regional indicator, numbers of healthcare institution and annual household income were considered as potential effect modifiors. However, the city-level modifiers collected only explaina small proportion of the heterogeneity and it is reasonable to expect existence of other unknown or unmeasured modifiers. Therefore, the heterogeneity between different regions should be considered when formulating and optimizing public health policy.3. The DLNM adapted in this study has advantages in modelling detailed lag effects of exposure on response, and providing the estimate of the overall effect. The potential application of this model is including but not limited to assessing the impacts of environmental exposures. It is also applicable for the analysis of the temporal evolution of the risk associated with protracted time-varying exposures in different research fields.