Study On Meteorological Epidemiology Of Hemorrhagic Fever With Renal Syndrome

At present, global warming and extreme climate events frequently happening are becoming severe environmental problems faced by human beings. Climate change and ecological environment variation are important causes of incidence rising, endemic area and season extending of infectious diseases, especially vector-borne diseases. To study the relationship between infectious diseases and climate and to explore the rule of meteorological epidemiology are necessary for prevention and control of these diseases. Hemorrhagic fever with renal syndrome (HFRS) is a kind of zoonosis caused by hantavirus (HV) and need to prevent and control emphatically in our country. The occurring and prevalence of HFRS are influenced by climate change. It is important to exploring the quantitative association of endemic law of HFRS with meteorological factors using advanced methods based on theory of meteorological epidemiology.Meteorological epidemiology is a branch of epidemiology to study the relationship between climate and health. Its methods include transverse research based on the data of several different areas and longitudinal research based on time series data of the same area. The longitudinal method is more feasible method at present since the geometric information system (GIS), the main technology used in transverse research, has not been universalized in China. However, the methods of data analysis in present reports on the relationship between incidence rate of HFRS and meteorological factors were mostly limited in simple correlation and regression or traditional multiple statistics such as multiplelinear regression and multiple stepwise discrimination. These methods have two defects: a) The relationship between disease and climate can only be analyzed qualitatively but not quantitatively, b) The confounders can not be controlled through these methods. Therefore, case-crossover design applying in environmental etiology research will be used to clarify the effect of meteorological factors on incidence of HFRS in this study. It is aimed at providing evidence to expound the endemic law of HFRS, to predict its future endemic trends and to make prevention policy and exploring and developing the method of meteorological epidemiology.We chose Junan County of Shandong Province as the research region, where the incidence rate of HFRS was high and a national surveillance spot was established. All the 13676 cases of HFRS from Jan, 1979 to Dec, 2001 were regarded as subjects. Each case's incidence time was recorded according to the report forms about endemic situation of HFRS in Junan County. The period between a case's incidence time and one year before was the case window in which the meteorological factors were observed month by month. The periods of one year both before and after the case window were chosen as its control windows in terms of symmetric bidirectional case-crossover design. So, the case period and its two control periods of each case composed a matching group of 1:2. After stratifying the subjects by endemic seasons, The data analysis of each stratum was from the following two hierarchies for each meteorological variable: ?Analysis month by month: Paired r-test month by month was used to compare the level of the meteorological variable in case window with the average level of the two control windows. Conditional logistic regression model was fitted month by month to reflect the association of the hazard of HFRS with climate change. The purpose of analysis month by month is to reveal that in which period of the recent year the incidence of HFRS is influenced by meteorological factors. ?Analysis of accumulation in time: To change the case window month by month from one month to twelve months before the incidence time and to compare the levels of the meteorological variable in casewindow with the levels of the same periods in the two control windows for each change. Both paired t-test and conditional logistic regression were used every time as in analysis month by month. The purpose of analysis of accumulation in time is to determine if cumulative effect exists by observing the trends of OR with the length of accumulation in time. Then multiple stepwise conditional logistic regression model were fitted to the data using the meteorological variables with statistical significance in single factor analyses for two seasons respectively. At last, the models were verified by the real data.The main results includes: ?The relevant factors to incidence of HFRS in spring and summer includes the mean temperature and the mean air pressure of the recent year, the rainfall of the last autumn, the mean relative humidity of the last autumn to winter, rain days of the last winter, sunshine hours from the last winter to spring and the mean wind speed of the recent year. Warm weather, ample rain, moist air and wind in last autumn to winter are suitable for endemic of HFRS in the following spring and summer, while exceeding long sunshine hours is unfavorable for endemic of HFRS. ?The outbreak of HFRS in spring of 1995 in Junan was probably caused by abnormal warmness of the last winter and the abundant rainfall in last autumn to winter as two times as usual years. However, the endemic of HFRS in spring of 1998 in Junan was mainly associated with higher temperature in last winter and shorter sunshine time in recent half year. (3) The relevant factors to incidence of HFRS in autumn and winter includes the mean temperature of the case month, the mean temperature of last winter to spring, the mean air pressure of the recent eight months, rainfall of the recent half years, rainfall of last half years, rain days of two to five months before, rain days of six to eleven months before and total sunshine hours of two to eleven months before. ?The outbreak of FIFRS from 1984 to 1986 was probably triggered by the coldness of those winters, which forced the A. agrarius moving to residential quarters and increasing contact with inhabitants. Additionally, the endemics in 1984 and 1986 were associated with less rainfall in last spring while more rainfall in autumn paralleled with the usual same period. Differently, the high incidencerate of HFRS in 1985 was probably associated with low air pressure and short sunshine time.To sum up, application of case-crossover design to the meteorological epidemiology of HFRS was feasible. It has prominent advantages to control the confounding of individual characters, secular trends and seasonal variation of exposure and event counts. Moreover, this study solved the problem of ascertaining hazard period resulted from possible carryover effect and cumulative effect through simultaneously carrying out analysis of sectioning time and analysis of accumulative in time. The effect of meteorological factors on incidence rate of HFRS was quantitatively illustrated, which provide evidence to expound the endemic law, to predict the future trends of HFRS and to make prevention policy. The significance of this study in theoretic area lies in exploring and developing the method of meteorological epidemiology.