Research On The Response Of Circulatory System Diseases To Weather And Climate Change In Zunyi Area And The Application Of Prediction

In recent years,cardiovascular and cerebrovascular diseases are the primary killers threatening human health,and the impact of weather and climate changes on these diseases cannot be ignored.The "Healthy China 2030" strategy puts forward that the development pattern of China’s health protection should give priority to prevention rather than medical treatment.To solve these two problems,the medical data from 2012 to 2017 and meteorological data from 1961 to 2016 of Zunyi area are used to analyze the characteristics of climate and circulatory system diseases of Zunyi area in detail,explore the relationship between disease incidence and meteorological conditions,set and select optimal disease prediction model,finally design the Zunyi Meteorological Sensitive Disease Prediction System for application and research.The main conclusion are as follows:The trend of climate change of Zunyi area during 1961-2016 was characterized as rising-temperature,decreased-relative humidity and less-rain.The study of apparent temperature and comfort level of Zunyi during 2005-2016 suggests that Zunyi were generalized comfort during the most days,followed by cold discomfort.As for apparent temperature,the ‘comfortable’ level accounts for the highest proportion,‘cool’ level for the second,and there were no ‘hot,cold and freezing’ levels.Consequently,the Zunyi area is very suitable to carry out health-care tourism with summer resort as the main part.From 2012 to 2017,hypertension ranked the first in the spectrum of circulatory diseases in Zunyi,accounting for 67.22% of the total number of patients,which is related to the dietary habits to a large extent.The proportion of male to female of circulatory system diseases was 1.26:1,55.84% for male and 44.16% for female.The age distribution was bimodal,the first peak occurred 49-53 years old,which is the emerging senility period of middle age,the second peak occurred 59-70 years old,which was the aggravating senility period of old age and had a higher number than the first peak.Comparing to the biggest city,it was found that the sex ratio of the disease incidence of in Zunyi is greater and the age of the high-risk population is younger.The number of circulatory diseases has its own characteristics at different time scale.In the day-to-day population changes,2012-2014 has a more obvious periodicity,showing a bimodal distribution.From 2015-2017,however,there were a linear upward trend under the influence of other factors.The peak patient number of circulatory diseases occurred in summer,and hypertension occurred in spring.The number of visits in March,May,November and December were much more,when the seasons alternated.The largest patient number of stroke and heart attacks occurred in December.According to the distributed lag nonlinear model in this study,when the temperature is extreme low(-0.5℃)or decreased after 24 hours,or when relative humidity is 39%-60%,or when air pressure is higher or lower,the meteorology factors affected diseases attacks instantly,and had more intensive lag impact than extreme high temperature(above 28℃)and higher relative humidity.The cumulative effect of low-temperature,high-humidity and high-pressure weather conditions is the most harmful to onset of the diseaseAccording to the daily variation of the number of visits,the data from 2012 to 2014 is regarded as data set 1,and the data from 2015 to 2017 as data set 2.In the prediction model with the multivariable stepwise regression(MSR),the forecasting value of some samples in both datasets are greater than the actual values.In dataset1,the annual prediction equation in dataset 1 has the best back-substitution performance,followed by summer prediction equation,the accuracy of back substitution is 63.1% and 61.4%,respectively.In dataset 2,the fitting degree and prediction performance of summer prediction equation are the best,followed by spring prediction equation,the prediction accuracy is 89.2% and 79.1% respectively.The annual prediction accuracy is 70.4%.In the prediction model with Back-Propagation neural network(BPNN),the forecast accuracy of dataset 1 is 60.07% and 74.19%,and that of dataset 2 is 64.86% and 78.75%.The model of dataset 2 had the better prediction performance and the method to select training set according to time is more helpful to improve the accuracy of the model.As for the annual prediction equation of dataset 2,the BPNN model performed better than the MSR model.Based on the above results,a Zunyi weather-sensitive disease risk prediction system is constructed in this paper.The system integrates meteorological,environmental and public health data effectively,which is convenient for analyzing climate,environmental characteristics of cycle,trend and mutation,as well as population morbidity,mortality and disease composition in Zunyi area.And the different weather-sensitive disease prediction models are optimized and applied to disease risk prediction service which is released to public based on a four-level forecasting and warning system.The system expands the field of meteorological services,improves the depth of medical services,and unites these two services,which is important to guide other cities to develop the medical meteorological forecast services in the future.