| BackgroundClimate change has become the biggest threat to human health in the 21st century.Preventing the harmful effects of climate change has become a major issue worldwide.There will be great significance for public health to carrying out comprehensive risk assessment of climate change on human health and developing appropriate intervention strategies and control measures.Many studies conducted both in China and other countries show that infectious diseases are one of the most important impacts of climate change on human health.Incidence of respiratory,enteric and vector-borne infectious diseases which are closely related to meteorological factors is still high in China in recent years.Diarrhea has always been a major public health problem in the world and has been a serious threat to children’s health.According to the global burden of disease study 2017.diarrhea was the eighth leading cause of death for all age groups and the fifth leading cause of death for children under the 5 years old during 2016.Diarrhea has always been a serious infectious disease in China.Other infectious diarrhea refers to the infectious diarrhea except cholera,bacterial and amoebic dysentery,typhoid fever and paratyphoid fever,which is the legal disease among infectious diarrhea and rank first two among class C infectious diseases of China.Current researches about climate change and health are mainly conducted in single city.Further.Systemic and in-depth studies conducted in different meteorological geographic regions are sparse.Few studies explored the risk assessment,forecast,and early warning models of weather sensitive infectious diseases.Our ability of prevention and control of weather sensitive infectious diseases is still weak.Other infectious diarrhea was selected as the target infectious diseases in our study.We aimed to systematically explore the regional differences,interactions.vulnerable groups,effect-modifiers and comprehensive risk assessment,prediction and early warning models of meteorological factors on other infectious diarrhea in China.Our study will help develop appropriate control measures and prevent strategy at country level to reduce the risk of weather sensitive infectious disease and provide scientific evidence and guidance for our government.local health administration,department of emergency.Center for disease control and prevention,and publics.Results will be very important for the health policy of prevention first,people co-construction and people share in China.Objectives1.To analyze the long-term trend of diarrhea in China,and to describe the epidemiological distribution,pathogen characteristics,and epidemiological regions of other infectious diarrhea.2.To explore the regional differences,interactions,vulnerable groups and effect-modifiers of meteorological factors on other infectious diarrhea in China.3.To evaluate the exposure,hazards,vulnerability and comprehensive risk of other infectious diarrhea due to meteorological factors in China.4.To establish a practical prediction and early warning model for other infectious diarrhea based on meteorological factors in high-risk areas.Methods1.Data collection:the incidence of diarrhea during the past 30 years in China were obtained from the database of global burden of disease study 2017.Data of other infectious diarrhea from 2014 to 2016 were obtained from the Chinese center for disease prevention and controls through the National Notifiable Infectious Disease Surveillance System.The meteorological data during the same period were obtained from the China meteorological administration.Data of the population are downloaded from the public health scientific data center,the education,health and economic indicators at prefecture-level are obtained from the China city statistical yearbook 2017 and the national bureau of statistics.284 cities with number of other infectious diarrhea cases greater than P15 were selected from 334 prefecture-level cities as our study areas to avoid uncertainty.The quality of original data has been controlled by relevant departments,and we also checked the data again.After tiding and cleaning these data,the data set for disease-meteorological factors and data set for effect modifiers were established.2.Statistical analysis:(1)the estimated annual percentage change of the incidence of diarrhea in recent 30 years in China was calculated to quantitatively evaluate the long-term trend.Distribution and pathogen characteristics of other infectious diarrhea were described.The epidemiological regions of other infectious diarrhea in China were divided based on a cluster analysis.(2)A two stage model was conducted to get the effect of temperature and humidity on the other infectious diarrhea at city,regional and national level:Distributed lag nonlinear models were conducted at the first stage to obtain the effect of temperature and humidity on other infectious diarrhea at city level,and multivariate meta-analysis was used at second stage to combine the city-level effect get the effect at regional and national level.The best linear unbiased prediction(BLUP)and relative/absolute dimension was used to calculate the precise effect at city level;Residuals test and sensitivity analysis were also performed for the first-stage models.The interaction of temperature and humidity on other infectious diarrhea were analyzed based on binary reaction surface method and meteorological index method.Effect of temperature on other infectious diarrhea was analyzed in different subgroups according to gender,age group and occupation to identify potential vulnerable groups;On the basis of two stage models,the meta-regression method was applied to identify the potential effect modifiers from the demographic,meteorological,economic.health and geographical indicators.(3)The comprehensive risk assessment framework based on exposure,hazards,and vulnerability was used to assess the comprehensive risk of other infectious diarrhea due to meteorological factor in China then identify the high-risk areas.(4)High risk areas from study of comprehensive risk assessment were selected to develop predict and early warning models.Generalized additive model was used to predict other infectious diarrhea cases.Warning threshold value for predicted case numbers was calculated based on warning threshold value for the actual case number and ROC curve method.Early warning models of other infectious diarrhea give out the warning signal or not depends on the predicted case numbers and warning threshold value for predicted case numbers.Results1.High incidence of diarrhea in the world are mainly occurred in developing countries and regions including Africa,South Asia and South America.The incidence of diarrhea in China is at a low level.Results showed that incidence trend of diarrhea in China was decreasing,but the long-term trend of diarrhea in children under 5 years old was increasing obviously.Distribution characteristics of other infectious diarrhea indicated that incidence in Jiangsu province,Zhejiang province,Northern China,and Southern China was obviously higher than in other regions.An obvious difference of seasonal distribution for other infectious diarrhea in different regions was observed,with a summer peak in most Northern provinces,a winter peak mostly in southern provinces,and a mixed summer or winter peak in most central regions.Among cases with other infectious diarrhea,males were higher than females(56.6%vs.43.4%).laboratory diagnosed cases were less(22.3%),and most of them were clinically diagnosed cases.More than half of them are children under 5 years old(53.7%),more than half of them are scattered children(53.2%),followed by farmers(21.4%).Results of pathogen characteristics of other infectious diarrhea showed a significant difference of pathogen distribution among different age groups.Rotavirus accounted for the absolute proportion in children under 5 years old,while vibrio parahaemolyticus and salmonella increased significantly in people over 20 years old.Rotavirus accounted for the absolute proportion of scattered children in different occupational groups.The proportion of norovirus in preschool children began to increase and the proportion of norovirus in students was the highest.The proportion of vibrio parahaemolyticus and salmonella in farmers,retired people,domestic workers and unemployed people was higher.The seasonal distribution of different pathogens is very different,rotavirus has a high incidence in winter,while salmonella,vibrio parahaemolyticus.escherichia coli have a high incidence in summer.Adenovirus has a high incidence in summer and autumn,while norovirus mainly occur in autumn and winter;Result of cluster analysis shows that there are three epidemiological regions in the whole country,including summer high incidence area,mixed area and winter high incidence area from north to south regions of China.Though Heilongjiang province is located in the north,it is identified as winter high incidence area.2.Effect of temperature on other infectious diarrhea at country level indicated an approximately "U" type,with both high temperature and low temperature can increase the risk of other infectious diarrhea,effect difference at city level is big.At national level the lag effect of low temperature increases first over time then reduced,the lag effect reached the maximum at lag 15 day and disappeared at lag 35 day with a relative long duration of lag effect.Effect of high temperature reaches maximum at lag 0 day then decrease until not significant with a relative short duration.Effect of temperature on other infectious diarrhea varies significantly in different regions.In most northern regions(including Northern China.Northeast,Inner Mongolia.Huanghuai and Northwest),high temperature increases the risk of other infectious diarrhea,while in southern regions(including Southern China,Jianghuai and Jiangnan),low temperature significantly increases the risk.Sensitivity analysis shows that the coefficient estimation is still robust after varying different modeling strategies.Our results are reliable.Results of interaction analysis showed that there were significant regional differences in the effects of meteorological indicators on other infectious diarrhea in China which is similar with effect of temperature.Interaction varies significantly at different cities.Results of vulnerable groups analysis showed that the vulnerable groups affected by low temperature were people under 5 years old,scattered children and men.while the vulnerable groups due to high temperature were women,farmers and people over 20 years old.Analysis of effect modifiers of temperature on other infectious diarrhea showed that the statistically significant effect modifiers identified by meta-regression model included population growth rate,population density,number of hospitals,per capita GDP,average temperature,average relative humidity,average rainfall,latitude and region.3.Results showed that the areas with high exposure were Northern China,Huanghuai,Jianghuai,Jiangnan,and Southern China,which were concentrated in the east of China.Results of the hazards assessment by season shows the regions with high hazards in winter included Jiangnan.Southern China,and Northwest China,while the regions with low hazards included Northern China,Huanghuai,and Jianghan.In summer,the areas with high hazards are Northern China,Inner Mongolia,Huanghuai,Northwest,and Northeast,while the areas with low risk are Jiangnan,Jianghuai,Jianghan,and Southwest.Results of vulnerability assessment by season indicated the regions with high vulnerability in winter included Southern China,Southwest China,and Northwest China,while the regions with low vulnerability included Northeast China,Inner Mongolia,Northern China,and eastern coastal regions.The areas with high vulnerability in summer include Northwest,Southwest and some central regions,while the areas with low vulnerability are mainly located in the east coast.Results of comprehensive risk assessment show that the comprehensive risk in summer is high in Northern China,Huanghuai,and other northern regions,while the comprehensive risk is low in the Northwest.Southwest.Southern China,Jiangnan,and other southern regions.In winter,the comprehensive risk is high in Jianghuai.Jiangnan.Southern China,and other southeastern regions,while the comprehensive risk is low in the west and north.4.Model fitting of the generalized additive model was good.The adjustment R2 of Beijing reached 0.94,and that of Shenzhen reached 0.97.The RMSE and MAPE of Beijing were 70.19 and 0.07%,respectively,better than that of Shenzhen(RMSE and MAPE were 76.94 and 0.12%,respectively).Results of early warning models showed that the predicted outbreaks of other infectious diarrhea were very close to the gold standard.Result of ROC curve display with the sensitivity of Beijing of 0.87,the specificity of 0.99,the positive predictive value of 0.98,the negative predictive value of 0.92.and the AUC of 0.97.In Shenzhen,the sensitivity reached 1 and the specificity reached 0.99.The positive predictive value was 0.92 and the negative predictive value was 1 with the AUC reached 0.99.All above indicators showed a good warning effect and the model could be applied in other high-risk areas.Conclusions1.Incidence trend of diarrhea in China was decreasing for all age groups but increasing obviously for children under 5 years old.Incidence in Jiangsu province,Zhejiang province.Northern China,and Southern China was obviously higher than that in other regions.Rotavirus infection is the main cause in children under the age of 5 years,with high incidence in winter and scattered children.Three epidemiological regions were clustered including summer high incidence area,mixed area and winter high incidence area from north to south China.2.Effect of temperature on other infectious diarrhea varies greatly in different regions.High temperature increased the risk mainly in northern regions including Northern China,Northeast China,Inner Mongolia,Huanghuai,and Northwest China with vulnerable groups including people over 20 years old,peasants,and women.Low temperature increased the risk mainly in southern regions including Southern China,Jiangnan,and Jianghuai with vulnerable groups including people under 5 years old,scattered children,and men.3.The exposure,hazards,vulnerability and comprehensive risk of other infectious diarrhea due to meteorological factors vary greatly in different regions with different seasonal distribution.Areas with high comprehensive risk are concentrated in Northern China and Huanghuai for summer while in Jiangnan and Southern China for winter.4.The prediction and early warning models of other infectious diarrhea based on meteorological factors are simple,practical,and effective,which could be applied in other high-risk areas.Innovation1.Our multi-city study systematically and comprehensively evaluated the effects of meteorological factor on other infectious diarrhea in a large geographical area of China for the first time.Regional differences,interaction,vulnerable groups and city-level effect modifiers were also explored.2.The risk assessment framework based on exposure,hazards and vulnerability was firstly used to assess the comprehensive risk of other infectious diarrhea due to meteorological factor in a large geographical area.3.Our practical predict and early warning model based on meteorological factors in high-risk areas can help the prevention and control of diarrheal diseases for the local government and publics and also provide a methodological reference for the early warning of other similar infectious diseases. |
PDF Full Text Request