A Case-crossover Study On Particulate Air Pollution And Cardiovascular Disease Mortality

Cardiovascular disease (CVD) refers to a group of heart and vascular diseases including myocardial infarction, stroke, hypertension, disease of pulmonary circulation and peripheral vascular disease. CVD constitutes a major threat to public health and is commonly characterized as the disease of affluence. The etiology of CVD is complicated. Besides the biological factors, the psychological and environmental factors are also playing an important role.According to the existing states, air pollutants could be classified as two categories. The one is gaseous pollutants mainly including sulfur dioxide (SO₂), nitrogen dioxide (NO₂), carbon monoxide (CO) as well as ozone (O₃). Particulate matter (PM) is another major and complex ambient pollutant. Based on differnt aerodynamic diameter size, PM could be mainly referred to as total suspended particles (TSP) of aerodynamic diameter less than 100 μm, PM₁₀ of aerodynamic diameter less than 10μm, or PM_2.5 of aerodynamic diameter less than 2.5μm, among which PM₁₀ has been the regulatory focus of the worldwide environmental protection agencies (EPA). With the development of economics and industrialization, particulate air pollution is increasingly worseon. Recently, many toxicological and epidemiological studies have indicated that PM air pollution not only affects the respiratory system, but also has the long-term or short-term effect on the cardiovascular system.Because several factors differ among countries or districts such as the level of environmentalpollution, the proportion of elderly, and socioeconomics, and with the development of exposure measurement and epidemiological study design, it's necessary to accurately investigate the effect of exposure to air pollution on population with proper method.ObjectiveUsing time-stratified case-crossover study design, we intend to evaluate the association between CVD mortality and short-term elevation in PMio daily mean concentration. Although the primary hypotheses related to the PM, it's also to examine the associations of other gaseous pollutants such as SO2 as well as NO2 with CVD mortality. As stroke is the most prevalent subcategory of CVD in China, we then to investigate the association between air pollution and stroke mortality.Material and methodsThe study's population is local people in Hangzhou city. The information of CVD mortality case were extracted from the death register system of center for disease prevention and control in Hangzhou city (HZ CDC). The case was defiened as death from CVD causes (International Classification of Disease, ICD code, 10th revision 100-199 ).CVD mortality from Eight administration districts in Hangzhou city between 2002 and 2004 year were included in this study due to the availability of the information.The information about the mean daily concentration of air pollution (PMio, SO2, NO2) during the study's period were provided by the center for environmental surveillance in Hangzhou city. The meteorological information such as mean daily temperature and daily relative humidity were then obtained from center for meteorological agency in Hangzhou city.The time-stratified case-crossover design was conducted, which was proposed as a method for assessing the effects of transient exposures on the subsequent risk of acute events. The exposure of interest was mean daily PMio, SO2, or NO2 concentration. Exposure level during the case period was defined as mean pollutant's concentration either on the day of death (lag 0) or 1⁵ days preceding death (lag 1⁵). Control periods were chosen by using time-stratified approach such that exposures during the case period were compared with exposures occurring on all otherdays of the same month (before or after the event day) that fall on the same day of the week as the case period. So there were several control times before or after the case time in a matched set.The statistical analysis was composed of two parts, the one was to evaluate the acute effect of air pollution on all CVD mortality and the other was the specific death cause (ie stroke mortality). Two parts were analyzed with the same analytical strategy. In addition to the descriptive statistics for the CVD and exposure information, conditional logistic regression was perfomed, stratifying on each death, to obtain estimates of odds ratios and 95 percent confidence intervals associated with an 10 jJ. g/m3 in the mean daily level of each pollutant. The effect of PM)0 was first evaluated separately at lags of 0-5 days in single pollutant models. On the basis of the results of this analysis, a single lag for all subsequent analyses was selected. Pollutants were first considered separately by using single-pollutant models and then jointly by using multi-pollutant models. In all models, meteorological variable such as daily mean temperature and relative humidity were included with lags paralleled to the modelled air pollutant's lags. The sensitivity analysis was conducted in three ways. The first was to replace the missing values of exposure measurement by linear interpolation and then model the variables. The second was to specifically analyze the effect of air pollution on population aged over 65 old years. The third was to categorize the environmental variables into four levels which then been modelled.All reported p values were based on two-sided tests. A p value of <0.05 was considered statistically significant. All the data management and statistical analyses were performed by SPSS for Windows 13.0 software package. ArcGIS9.0 software was used for information visualization. Time-stratified referent selection was realized by programming macro with VBA in the EXCEL 2003.Results1. The Characteristics of CVD mortality in Hangzhou city A total of 17124 cases dying of CVD from 2002 to 2004 year among Hangzhou city were included in the study. The proportion of male was somewhat more than that of female . The three leading subcategories of CVD mortality were stroke, coronary heart disease, and hypertension, which accounted for 57.85%,19.15%, and 15.20% respectively. The mortality of hypertension increased with theyear.2. The Characteristics of air pollution in Hangzhou city The locations of air pollution monitoring stations were evenly scattered in the area of interest, of which the mean exposure measurements could relatively substituted population exposure level for individual's. Each environmental variables was presented as abnormal distribution. The mean daily concentration of PM10 was the highest among the three pollutants during the study periods. The days when PMjo levels exceeded current limits(150i)g/ m3) accounted for 23.4% in all monitoring days, however decreased with year. The SO2 level was under the criteria but NO2 level was sometimes over the limits, the variation of which neither could be found statistically significance.3. The single-pollutant models After being adjusted for meteorological factors, the single-pollutant models showed that a 10 u g / m3 increase in PM10, SO2 and NO2 on the two days before the event was associated with 0.61 % (95 %Ch 0.28% 0.94%), 1.73 % (95%C7: 0.68% -2.79%), and 2.02%(95%C7: 0.85% 3.20%) increase in the CVD mortality as well as 0.56%(95%C7: 0.14% 0.99%), 1.62% (95%C7: 0.26 % 3.01%) and 2.07%(95%C7: 0.54 % 3.62 %) increase in stroke mortality repectively.4. The multi-pollutant models In multi-pollutant models, only PM10 remained the short-term effect on CVD mortality, though the estimate seemed to be less after being adjusted for SO2, other co-pollutants had no statistically significant effect on all cardiovascular deaths or specific stroke mortality.5. The sensitivity analyises In sensitivity analysis, the associations were found in all single-pollutant model but not statistically significant in multi-pollutant models after replacing the missing values or only focusing on elderly population. When the environmental variables transforming into categorical variables, the ORs for CVD mortality associated with an interquartile range (IQR) increase in PM10, SO2 or NO2 level were 1.127(95%C7: 1.067-1.189), 1.102 (95%C7:1.037 -1.078) and 1.091(95%C7:1.027 -1.159) and for stroke mortality were 1.102 (95%C7:1.027 -1.182), 1.081 (95%C7: 0.999-1.170) and 1.085 (95%C7:1.003-1.173) respectively.Conclusion1 After being adjusted for meteorological factors, a 10 u g / m3 increase in PMi0, SO2 and NO2 on the two days before the event was associated with 0.61 %(95 %CI: 0.28% 0.94%), 1.73 %(95%CL 0.68% -2.79%), and 2.02%(95%C7: 0.85% 3.20%) increase in the CVD mortality respectively.2 After being adjusted for meteorological factors, a 10 P g / m3 increase in PM10, SO2 and NO2 on the two days before the event was associated with 0.56%(95%CI: 0.14% 0.99%), 1.62%(95%CI: 0.26 % 3.01%)and 2.07%(95%CI: 0.54 % 3.62%)increase in stroke mortality repectively.It can be concluded that short-term elevation in daily mean concentration of particulate as well as other gaseous air pollution are associated with the increase in all cardiovascular disease mortality and stroke mortality in Hangzhou city, which provided some envidences for the studies on the cardiovascular effct of air pollution. This study however is ecological study in nature due to the exposure measurement and further study is still needed to improve and verify the conculsions.