| Objective: Over the past 100 years from 1880 to 2012,the global average temperature had increased by 0.85(0.65-1.06)℃.As the trend of climate warming had been confirmed by more and more studies at home and abroad,the research on the relationship between meteorological factors,especially temperature and health,had attracted more and more attention.Because the death cause monitoring data was easier to obtain than other types of health data,the study of meteorological factors,especially the relationship between temperature and death,had become the mainstream field of relevant research.Many studies on the relationship between atmospheric temperature and mortality showed that the relationship between them was nonlinear,showing a typical "U" or "J" type,that is,the increase of mortality was related to both low temperature and high temperature.Most studies showed that the relationship between temperature and death changeed with the change of temperature,with the slow increase of temperature,people’s adaptability to heat also gradually improved,resulting in a downward trend in the number of heat related death risks.However,some studies had come to different conclusions.Some studies believed that the relationship between temperature and death was stable,while others showed that with the increase of temperature,heat related death showed an increasing trend.However,the absolute temperature,geographical distribution and economic level among the above research countries or regions were intertwined.The temperature level,geographical distribution and even economic level were not the main factors affecting the correlation between temperature and death.Similarly,studies on cold related mortality also showed contradictory results.Some studies showed that the risk of cold related death decreased with the increase of temperature,while others showed that there was no change in cold related death.With the rise of global temperature,most studies focused on the change trend of the relationship between temperature and death under the condition of rising temperature.Due to the lack of corresponding research sites in the area of dropping temperature,there was little research on the changes between the relationship on temperature and death under the condition of dropping temperature.Shenyang,located in Northeast China,was a famous heavy industry base in China with a high degree of industrialization.In terms of climate distribution,Shenyang is a severe cold area.The climate is characterized by extremely cold winter and long duration.The average temperature of the coldest month is lower than 10 ℃,and the number of days lower than 5 ℃ throughout the year is not less than 145 days.Some studies showed that the temperature in the northeast of Shenyang showed a temperature decline trend different from global warming in recent 50 years,which provided an excellent place to study the changes in the relationship between temperature and death caused by temperature decline.This study intends to use the method of retrospective analysis and time series analysis to study the exposure response relationship and change trend between temperature change exposure and daily mortality of circulatory diseases,respiratory diseases and total deaths(except accidental deaths)in Shenyang from 1992 to 2018.The main purposes of this study are: 1.The level and trend of MMT;2.The relationship between temperature and main death and its trend;3.Quantitatively evaluate the level and change trend of temperature and heat or cold related death.Methods: Shenyang was selected as the research site,and the research period was 1992-2018.Obtain the meteorological data including daily average temperature,humidity and surface pressure of Shenyang City(Station No.543420)since 1992 through the U.S.Oceanic and Meteorological Administration;The daily death data of circulatory diseases(ICD-10 Code: i0-i99),respiratory diseases(ICD-10 Code:i0-i59.9)and other diseases except accidental injury(ICD-10 Code: i60-i69.9)were obtained through Shenyang Center for Disease Control and prevention.Since the number of daily deaths generally obeyed Poisson distribution,we adopt the generalized additive model(GAM).Because the distribution of daily mortality might have the problem of over dispersion,quasi Poisson was applied in GAM to construct the relationship between temperature and daily mortality.Horizontal visibility,relative humidity,atmospheric pressure,time and other factors are included in GAM to adjust other meteorological factors and the long-term and seasonal trend of daily mortality itself.Previous studies have proved that temperature has a significant lag effect on population mortality.That is,the average temperature of the day might still affect the mortality of the population after a few days or even decades,and the relationship between temperature and mortality was usually nonlinear.We apply the Distributed Lagged Nonlinear Model(DLNM)to construct the "cross-basis" function of temperature and incorporate it into GAM.The advantage of DLNM was to analyze the cumulative effect of temperature exposure in the past few days on the health of the day(backward perspective)after correcting the strong correlation of temperatures in adjacent days,and to explore the lag mode of the health effect of temperature change in a day in the subsequent period(forward perspective).In order to analyze the influence trend of temperature on death,the 27 years of the whole study were divided into three time periods,each of which was 9 years.The following three indicators were mainly analyzed: 1.The level and time trend of MMT,2.RR,and 3.AF(attribution score)and AN(number of attributed deaths).All statistics were bilateral tests,with P < 0.05 as statistically significant righteousness.Results:(1)From 1992 to 2018,the minimum mortality temperature(MMT)of total death under the same day temperature in Shenyang was 19.3 ℃,and that of circulatory system was 20.2 ℃,which was higher than that of respiratory system(16.9 ℃);The MMT of total death in men was 19.7 ℃,higher than that in women(19.0 ℃),and the MMT of people under 65 years old was 20.3 ℃,higher than that of people aged 65 and over(19.0 ℃).(2)MMT showed a downward trend,which was consistent with the change trend of temperature.Taking the lag of 0 days as an example,the comparative analysis of the three time periods shows that the MMT of the three diseases shows a downward trend.Circulatory diseases decreased from 21.6 ℃ in 1992-2000 to19.5 ℃ in 2010-2018,decreased by 2.1 ℃,respiratory diseases decreased from18.5 ℃ to 16.8 ℃,decreased by 1.7℃,and the total death decreased from 20.3 ℃to 19.0 ℃ and decreased by 1.3 ℃.That is,the change trend of the MMT was the same as that of the temperature,which also shows the adaptability of the population to the temperature.With the decrease of the temperature,the MMT also shows a downward trend.The overall downward trend of the MMT was further confirmed by using the time series with different intervals gradually increasing by one year.(3)Compared with the MMT,the impact of the increase or decrease 1 ℃ or extreme temperature on death showed that the risk of heat related death had a downward trend,while the risk of cold death had an upward trend.The correlation between circulatory system,respiratory system and total death and temperature was consistent.The effect of heat on death appeared rapidly and sharply,while cold related death appeared slowly and continuously.The significant effect of cold on death mostly appeared when the lag was 28 days,but it still had a significant effect until the lag was 50 days.There was a nonlinear relationship between temperature and mortality,and this relationship was not fixed.With the passage of time(the decrease of temperature),it gradually transited from "L" type dominated by cold influence to "J" type dominated by heat influence.Taking the daily average temperature as an example,the RR value of very low temperature of circulatory system(below 1%temperature)relative to MMT decreased from 1.369 to 1.174,the RR value of respiratory system decreased from 1.918 to 1.340,and the total mortality decreased from 1.304 to 1.140;The extreme heat(99% temperature and above)related RR value showed an obvious upward trend,the circulatory system increased from 1.170 to1.523,the respiratory system increased from 1.388 to 1.570,and the total mortality increased from 1.182 to 1.386.This change might occur in a relatively short time,such as 9 years.It was suggested that the adaptability of the population should be fully considered in the study of predicting the impact of short-term temperature on death in the future.This study also found that even in the same city,the absolute temperature level could not represent the impact of temperature on mortality.The change trend of temperature was the main reason for the increase of mortality.(4)The attributable risk of major causes of death in Shenyang was higher than that in similar studies.The attributable risk of circulation and total death showed an upward trend,while that of respiratory system showed a downward trend.The cold related attributable risk was significantly higher than that of heat related risk.The attributable fraction(AF)of total death in 27 years was 21.8%(95% CI: 19.8-24.2),26.2%(95% CI: 23.3-28.6)and 25.9(95% CI: 22.0-29.1).The current research showed that the AF caused by temperature in China was higher than that in other countries,and compared with other studies at home and abroad,the AF of this study was at a high level.The attributable number(AN)in the past 27 years was 148565(95% CI: 133091-163185),5502 per year for total death,83958(95% CI:74490-92004),3110 per year for circulatory system.18673(95% CI: 15782-21199),and 692 per year for respiratory system.From the analysis of AF and AN,the overall effect,cold and heat effect of circulatory system and total death showed a significant upward trend;The overall effect and cold effect of respiratory system showed a downward trend,and the heat effect showed an upward trend.The respiratory system may be more sensitive to the changes of temperature.By gender and age,women and people under 65 had better adaptability to temperature changes.In all study periods and different time periods,the AF and AN of cold related deaths were significantly higher than those of heat related deaths.Conclusions:(1)MMT generally showed a downward trend,and the trend was the same as that of the main temperature indicators,indicating that the change of MMT was an adaptive mechanism of people to temperature change.(2)The correlation between temperature and death was nonlinear,and its correlation changed with the decrease of temperature,from heat effect to cold effect;The effect of temperature on death was not caused by the absolute level of temperature,but by the change trend of temperature.(3)With the decrease of temperature,the thermal effects of circulation,respiration and total death all showed upward trends,which were consistent with the changes trend of relative risk.However,in addition to the respiratory system,the cooling effect,the circulation system and total death also showed upward trends,indicating that the relative and absolute risk changes of temperature and death,especially the cooling effect of circulation system and total death,were not completely consistent. |
PDF Full Text Request