Association between Atmospheric Particulate Pollutants and Mortality for Cardio-Cerebrovascular Diseases in Chinese Korean Population: A Case-Crossover Study

Background: Air pollution in large Chinese cities has led to recent studies that highlighted the relationship between particulate matters (PM) and elevated risk of cardio-cerebrovascular mortality. However, it is unclear as to whether: (1) The same adverse relations exist in cities with relatively low levels of air pollution; and (2) the relationship between the two are similar across ethnic groups. Methods: We collected data of PM2.5 (PM with an aerodynamic diameter ≤ 2.5 µm) and PM10 (aerodynamic diameter ≤ 10 µm) in the Yanbian Korean Autonomous Prefecture between 1 January 2015 and 31 December 2016. Using a time-stratified case-crossover design, we investigated whether levels of particulate pollutants influence the risk of cardio-cerebrovascular disease mortality among ethnic Korean vs. ethnic Han residents residing in the Yanbian Korean Autonomous Prefecture. Results: Under the single air pollutant model, the odds ratios (ORs) of cardio-cerebrovascular disease were 1.025 (1.024–1.026) for each 10 μg/m3 increase in PM2.5 at lag0 day, 1.012 (1.011–1.013) for each 10 μg/m3 increase in PM10 at lag1 day. In the multi-pollutant model adjusted by PM10, SO2, and NO2, the ORs of cardio-cerebrovascular disease were 1.150 (1.145–1.155) for ethnic Koreans and 1.154 (1.149–1.158) for ethnic Hans for each 10 μg/m3 increase in PM2.5. In the multi-pollutant model adjusted by PM2.5, SO2, and NO2, the ORs of cardio-cerebrovascular disease were 1.050 (1.047–1.053) for ethnic Koreans and 1.041 (1.039–1.043) for ethnic Hans for each 10 μg/m3 increase in PM10. Conclusion: This study showed that PM2.5 and PM10 were associated with increased risks of acute death events in residential cardio-cerebrovascular disease in Yanbian, China.


Introduction
Air pollution has been demonstrated to be a major risk factor for the development of cardiovascular diseases worldwide. A large number of studies have shown that air pollutants have an effect on cardiovascular morbidity and mortality [1][2][3][4][5]. Many epidemiological studies have showed that air pollution is a complex mixture of gaseous particulate compounds and that air pollution has negative health effects [6][7][8][9][10][11]. Experiments have indicated that particulate matter with a diameter ≤ 2.5 µm (PM 2.5 ), a very small particle, can enter the body through the trachea, and then into the alveolar of lung tissue, and may spread to the capillaries and into the blood [12]. The PM 2.5 can also contain carcinogens such as polycyclic aromatic hydrocarbons (PAHs) [13][14][15][16]. Thus, PM 2.5 might have a serious negative impact on human health.
Studies in China on PM 2.5 exposure have indicated that elevated levels of PM 2.5 exposure can lead to an increased risk of cardiovascular mortality [17,18]. However, all of these studies have focused on the areas where PM 2.5 exposure was higher than the national secondary standard (35 µg/m 3 ).
On the other hand, numerous studies in Europe and America have focused on areas where the range of PM 2.5 exposure was lower than 35 µg/m 3 (specifically 6.7-25 µg/m 3 ), and showed that the increase in PM 2.5 levels can also lead to an increase in the risk of cardiovascular mortality [19][20][21][22].

Mortality Data
For the same time period, we obtained mortality data related to cardiovascular and cerebrovascular diseases along with demographic data (e.g., age, sex, date of death, and address) from the Yanbian Korean Autonomous Prefecture Center for Disease Control and Prevention. The classification of the disease in the death cases was coded according to the International Classification of Diseases (ICD-10). For the present study, ICD-10 codes associated with cardiovascular disease (I09-I11, I20-I27, and I47-I52) and cerebrovascular disease (I60-I69) were included. The present study is restricted to ethnic Koreans and ethnic Hans, because there were limited data on other ethnic groups.

Data Analysis
The case-crossover design was developed as a variant of the case-control design to study effects of transient exposures on acute events [27,28]. The method can be regarded as a special type of case-control study in which each case serves as his/her own control. Since there is nearly perfect matching on all measured or unmeasured personal characteristics that do not vary over time, there can be no confounding by those characteristics. The method uses cases with cardiovascular or cerebrovascular diseases as their own controls, thereby effectively controlling all factors (including age, gender, race, smoking, genetic, and socio-economic status) except for the exposure of interest. We thus applied the time-stratified case-crossover design to assess relations between levels of air pollutants and the risk of cardiovascular and cerebrovascular deaths. The case period for each death was defined as the day of death. Control periods were selected by matching the same day of the week within the same calendar month and year for each case. For example, if a case was on the second Thursday of April 2016 (14 April 2016), all other Thursdays within April 2016 were assigned as controls (7 April 21, and 28 April 2016). This approach resulted in three or four controls for each case. This approach allows unbiased conditional logistic regression estimates [29,30]. Previous studies have shown that there is a lagging effect of air pollution on death [31]. Therefore, this study used to model lagging to observe the impact of PM 2.5 and PM 10 on the day (lag0) and delay 1-3 days (lag1-3) of death for cardio-cerebrovascular disease in residents, respectively. Lag1 is defined as the pollutant exposure value of the one day before the deaths, lag2 as the pollutant exposure value for two days before deaths, and so on.
The relations among the measures of PM 2.5 , PM 10 , SO 2 , and NO 2 were assessed by Pearson correlation analysis. The effects of atmospheric PM 2.5 and PM 10 on the exposed population were summarized by the difference between the health effects of each layer and its 95% confidence interval based on the formula: where Q 1 and Q 2 were the estimates of the two health effects and SE1 and SE2 were their corresponding standard errors. The conditional logistic regression analysis was used to examine the risk of the single-pollutant and multi-pollutant exposure conditions, and atmospheric PM 2.5 and PM 10 exposures in the case period and control period on cardio-cerebrovascular disease deaths. To take into account the impact of meteorological factors on cardio-cerebrovascular mortality, the average daily temperature, relative humidity and barometric pressure were included in the regression model. Subgroup analysis was also carried out with the categorized factors of sex, age, and season: Sex (male vs. female), age (≤ 65 years vs. > 65 years), and season (cold vs. warm season). Warm season was defined as May to October, cold season was defined as November to April. Odds ratios (ORs) for PM 2.5 and PM 10 were computed for each unit increase of 10 µg/m 3 and 95% confidence intervals (CI) were provided. Data analysis was performed with SPSS 17.0 software (SPSS Inc, Chicago, IL, USA) and all statistical tests were two-sided at the significance level 0.05.

Results
The total number of deaths from cardio-cerebrovascular disease among the residents of Yanbian Korean Autonomous Prefecture during the study period was 16 10 , SO 2 , and NO 2 were lower than the national secondary ambient air quality standard in China. The national secondary ambient air quality standard in China of PM 2.5 , PM 10 , SO 2 , and NO 2 were 35 µg/m 3 , 70 µg/m 3 , 60 µg/m 3 , and 40 µg/m 3 , respectively. The mean value of each pollutant in the cold season was higher than the mean value of each pollutant in the warm season (Table 2). Pearson correlation analysis on PM 2.5 , PM 10 , SO 2 , NO 2 , temperature, relative humidity, and barometric pressure showed that air pollutants were significantly positively correlated with barometric pressure, but were significantly negatively correlated with temperature and relative humidity. As shown in Table 3, increased levels of PM 2.5 and PM 10 yielded significant increases in the cardio-cerebrovascular disease mortality. The ORs were 1.025 (1.024-1.026) for each 10 µg/m 3 increase in PM 2.5 at lag0 day, 1.012 (1.011-1.013) for each 10 µg/m 3 increase in PM 10 at lag one day. Based on single pollutant model, multi-pollutant models were built for the ethnic Koreans and ethnic Hans, respectively (Table 4). In the multi-pollutant model adjusted by PM 10 , SO 2 , and NO 2 , the ORs of cardio-cerebrovascular disease were 1.150 (1.145-1.155) for ethnic Koreans and 1.154 (1.149-1.158) for ethnic Hans for each 10 µg/m 3 increase in PM 2.5 . In the multi-pollutant model adjusted by PM 2.5 , SO 2 , and NO 2 , the ORs of cardio-cerebrovascular disease were 1.050 (1.047-1.053) for ethnic Koreans and 1.041 (1.039-1.043) for ethnic Hans for each 10 µg/m 3 increase in PM 10 . The results indicate that the effect of PM 2.5 on cerebrovascular disease mortality were slightly higher for ethnic Hans than that for ethnic Koreans. While the multi-pollutant models showed that the effect of adjusted PM 10 on the risk of cardiovascular and cerebrovascular deaths in ethnic Koreans were greater than in ethnic Hans (Table 4). Table 5 showed the results from subgroup analyses by gender, age, and season. It showed that there was difference in results with the seasonal stratification, but no difference in results with the gender and age stratifications. Table 6 summarizes available results from studies similar to our study, which shows a consistent significant risk of PM 2.5 and PM 10 on cardiovascular mortality.

Discussion
Our study showed that levels of PM 2.5 were significantly associated with cardiovascular mortality in the Yanbian Korean Autonomous Prefecture, even though the daily mean value of PM 2.5 concentration was lower than China's national secondary standard of 35 µg/m 3 for PM 2.5 during the study period. To our knowledge, this is the first study on the relationship between PM 2.5 and the death of cardiovascular disease in China with the data that the PM 2.5 exposure concentration is lower than that of the national secondary standard. In addition, this study extends the significant public health findings by investigating whether the relationship between PM 2.5 and cardiovascular mortality differs by ethnicity. In our study, a 10 µg/m 3 increment in the lag one day concentration of PM 2.5 corresponded to 2.6% (95%CI: 2.5%, 2.8%) increase of cardiovascular mortality for total population in Yanbian Korean Autonomous Prefecture. Our estimates of cardiovascular mortality in the Yanbian Korean Autonomous Prefecture were higher than those in studies of Guo [39], Ma [18], and Venners [40] in other parts of China, including Beijing, Shenyang, and Chongqing, respectively. Our results were similar to those from studies conducted in developed countries with lower levels of PM 2.5 that showed a significant impact on cardiovascular mortality [22,32,33]. A possible explanation of this phenomenon could lie in the exposure-response curve of air pollution that often tends to become flat at higher concentration levels [41].
In the single pollutant model, the results show that PM 2.5 is more harmful than PM 10 for the entire population residing in the Yanbian Korean Autonomous Prefecture, which was similar to the results of Guo's [42] and Dai's [43]. This might be due to the fact that PM 2.5 has a smaller particle size and has a larger specific surface area, so it can absorb more harmful substances than PM 10 , and PM 2.5 can stay longer in the air. In addition, PM 2.5 can be inhaled directly into the alveoli, participate in blood circulation, and may ultimately have a direct impact on the cardiovascular system.
It is interesting to see that ORs of PM 2.5 and PM 10 in the multi-pollutant models were higher than that in the single pollutant model in ethnic Korean Chinese and ethnic Han Chinese. Additionally, the effects of adjusted PM 10 on the risk of cardiovascular and cerebrovascular deaths in ethnic Koreans were greater than those of ethnic Hans (Table 4). This may be due to the combined effects of particulate contaminants and gaseous pollutants on cardio-cerebrovascular disease. Our analysis also showed that there might be an interaction between PM 2.5 and PM 10 , and both have synergistic effects on the health of the human body.
In addition, the effect of PM 2.5 in ethnic Hans had slightly higher risk for cerebrovascular disease mortality compared to the ethnic Koreans (Table 4). Similarly, Ostro and colleagues [21], using data from nine California Counties, reported that elevated PM 2.5 in Caucasians was associated with a higher risk of cardio-respiratory disease mortality compared with the cardio-respiratory disease mortality in African Americans. Potentially, this may be due to the differences in dietary habits and genetic factors between ethnic groups, but further studies are needed.
Our subgroup analysis showed that the effects of PM 2.5 and PM 10 on the cardio-cerebrovascular mortality significant differed by seasons. The PM 2.5 and PM 10 in the warm season have higher impact on cardio-cerebrovascular diseases mortality than those in the cold season. It is consistent with the results from the work of Ma and colleagues in Shenyang [18]. It might be due to the fact that temperature is higher in the warm season and PM 2.5 is more likely to diffuse in the air, resulting in more inhalation of the body and eventually into the human blood circulation. There is no consensus on the effects of seasonal factors on the relationship between atmospheric PM 10 exposure and population death, as demonstrated by Peng [44], Nawrot [45], and Lu [46]. These studies have shown that the warm season PM 10 pollution has a higher impact on population mortality than the cold season. It is believed that air pollution affects the biological pathway of cardiovascular mechanism [47,48], but it is not clear in which way.
Our study observed no significant differences in health effects of PM 2.5 and PM 10 exposure on the cardio-cerebrovascular mortality when stratified by age and gender. It is consistent with the results of the studies performed in Guangzhou and Shanghai by Kan and colleagues [34,49]. However, Franklin and colleagues [32], using data from 27 US communities, reported that an increase in PM 2.5 concentration had a stronger impact for cardiovascular mortality in women and seniors (75 years or older). Hong and colleagues [50] also found that elderly women were most susceptible to the adverse effects of PM 10 on the risk of acute mortality from stroke. The difference might be due to the difference in study population and in the criteria for inclusion, the number of cases, age stratification, composition of air pollutants, and degree of contamination. Further studies are needed.
Finally, comparison to the studies similar to ours showed higher PM 2.5 cardiovascular mortality in the Yanbian Korean Autonomous Prefecture compared to those in Shenyang, Shijiazhuang, Wuhan, Guangzhou, and Shanghai in China. In previous studies, it appeared that PM 2.5 exposure levels in developed countries were lower than in many Chinese cities, and some cities were even lower than China's national secondary standards. However, the PM 2.5 cardiovascular mortalities in those countries were higher than those in Chinese studies. It is interesting to see that the increase in cardiovascular disease death caused by PM 2.5 exposure in the Yanbian Korean Autonomous Prefecture was also higher than that in other cities in China. One possible explanation might lie in different research backgrounds, such as local PM levels, population sensitivity to PM 2.5 , age structure, and particle composition and toxicity [17]. Another possibility might be based on a reasoning that, at high exposure concentrations, the risk of death per unit increase of pollutant concentrations often tends to be reduced, possibly because vulnerable subjects may have died before the concentration had reached the maximum level [51]. In addition, the PM 10 cardiovascular mortality in our study was higher than those in studies in Shanghai and Sao Paulo, but was lower than that in a European study [26,35,36]. These inconsistent findings might be due to the regional differences in the study, such as the population living at different latitudes and the sensitivity of the population to PM, which is of interest for further investigation.
The research has some limitations. The data we collected was limited to only one city with a limited sample size and study duration period. While our study has showed that the increase in the concentration of atmospheric particulate pollutants would increase the risk of cardio-cerebrovascular mortality in the Yanbian Korean Autonomous Prefecture, the causal relationship between the two is unclear. Further investigation is needed to explore the biological mechanism.

Conclusions
In summary, this study shows that airborne particulate pollutants can increase the risk of acute death events in residential cardiovascular disease and cerebrovascular disease in the Yanbian Korean Autonomous Prefecture. The synergies between the various pollutants will increase their impact on cardio-cerebrovascular mortality, especially the combination of PM 2.5 and PM 10 will lead to a significant increase in cardio-cerebrovascular mortality. The cerebrovascular disease mortality risk associated with PM 2.5 was slightly higher in ethnic Hans than in ethnic Koreans. However, due to the small magnitude of the difference, further studies are needed to confirm and elaborate the findings.
Author Contributions: C.Z. conceived and coordinated the study, performed data analysis and drafted the manuscript; L.C. contributed to study design, reviewed and edited the manuscript; Z.Q. provided air pollution information and health data, and drafted the manuscript; Q.W. contributed to statistical analysis; Z.J., J.H.L., C.L. and Y.Z. contributed to review and edit the manuscript. All authors read and approved the final manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.

ICD10
International Classification of Disease, tenth revision PM particulate matter PM 10 particulate matter less than 10 µm in aerodynamic diameter PM 2.5 particulate matter less than 2.5 µm in aerodynamic diameter SO 2 sulfur dioxide NO 2 nitrogen dioxide OR odds ratio CI confidence interval SD standard deviation