Xu, Xiping

In 1995, daily mortality in a district of Chongqing, China, was analyzed from January through
December for associations with daily ambient sulfur dioxide and fine particles (airborne particles
with diameters ≤ 2.5 μm; PM2.5). The mean concentration of PM2.5 was 147 μg/m3 (maximum,
666 μg/m3), and that of SO2 was 213 μg/m3 (maximum, 571 μg/m3). On average, 9.6 persons
died each day. We used a generalized additive model using robust Poisson regression to estimate
the associations of mean daily SO2 and PM2.5 with daily mortality (on the same day and at lags up
to 5 days) adjusted for trend, season, temperature, humidity, and day of the week. The relative
risk of mortality associated with a 100 μg/m3 increase in mean daily SO2 was highest on the second
lag day [1.04; 95% confidence interval (CI), 1.00–1.09] and the third lag day (1.04; 95% CI,
0.99–1.08). The associations between daily mortality and mean daily PM2.5 were negative and statistically
insignificant on all days. The relative risk of respiratory mortality on the second day after
a 100 μg/m3 increase in mean daily SO2 was 1.11 (95% CI, 1.02–1.22), and that for cardiovascular
mortality was 1.10 (95% CI, 1.02–1.20). The relative risk of cardiovascular mortality on the
third day after a 100 μg/m3 increase in mean daily SO2 was 1.20 (95% CI, 1.11–1.30). The relative
risks of mortality due to cancer and other causes were insignificant on both days. The estimated
effects of mean daily SO2 on cardiovascular and respiratory mortality risk remained after controlling for PM2.5.

During the summer of 1999, information about respiratory health outcomes and relevant covariates was collected from 3,709 Chinese adults in Beijing, Anqing City, and rural communities in Anqing Prefecture. Indoor PM10 and SO2 were measured in a random sample of selected households. Using logistic regression and controlling for important covariates (excluding PM10 and SO2) and familial intraclass correlation, highly significant differences were found between study areas in the prevalences of chronic cough, chronic phlegm, wheeze, and shortness of breath, but not physician-diagnosed asthma. Generally, the lowest prevalence of respiratory symptoms was observed in Anqing City, a higher prevalence in rural Anqing, and the highest prevalence in Beijing. Median indoor concentrations of PM10 were similar in Anqing City (239 microg/m3) and rural Anqing (248 microg/m3), but much higher in Beijing (557 microg/m3). Median indoor concentrations of SO2 were similar in all three areas (Beijing: 14 microg/m3, Anqing City: 25 microg/m3, rural Anqing: 20 microg/m3).

The authors analyzed daily mortality data in Shenyang, China, for calendar year 1992 to identify possible associations with ambient sulfur dioxide and total suspended particulates. Both total suspended particulate concentrations (mean = 430 μmlg/m³, maximum = 1,141 μmlg/m³) and sulfur dioxide concentrations (mean 197 = μmlg/m³, maximum = 659 μmlg/m³) far exceeded the World Health Organization's recommended criteria. An average of 45.5 persons died each day. The lagged moving averages of air-pollution levels, calculated as the mean of the nonmissing air-pollution levels of the concurrent and 3 preceding days, were used for all analyses. Locally weighted regression analysis, including temperature, humidity, day of week, and a time variable, showed a positive association between daily mortality and both total suspended particulates and sulfur dioxide. When the authors included total suspended particulates and sulfur dioxide separately in the model, both were highly significant predictors of daily mortality. The risk of all-cause mortality increased by an estimated 1.7% and 2.4% with a 100-μmlg/m³ concomitant increase in total suspended particulate and sulfur dioxide, respectively. When the authors analyzed mortality separately by cause of death, the association with total suspended particulates was significant for cardiovascular disease (2.1%), but not statistically significant for chronic obstructive pulmonary diseases (2.6%). In contrast, the association with sulfur dioxide was significant for chronic obstructive pulmonary diseases (7.4%), but not for cardiovascular disease (1.8%). The mortality from cancer was not associated significantly with total suspended particles or with sulfur dioxide. The correlation between sulfur dioxide and total suspended particulates was high (correlation coefficient = .66). When the authors included sulfur dioxide and total suspended particulates simultaneously in the model, the association between total suspended particulates and mortality from all causes and cardiovascular diseases remained significant. Sulfur dioxide was associated significantly with increased mortality from chronic obstructive pulmonary diseases and other causes. The results of the current study reveal increased mortality associated with both total suspended particulates and sulfur dioxide.

Chongquing is one of the most polluted cities in China. To study the respiratory health effects of air pollution for this city, the authors monitored the ambient levels of particulate matter (PM2.5) and sulfur dioxide (SO2) in urban and suburban areas and performed pulmonary function testing on 1,075 adults 35 to 60 years of age who had never smoked and did not use coal stoves for cooking or heating. The mean concentration of SO2 in the urban areas (213 micrograms/m3) was twice as high as that in suburban areas (103 micrograms/m3). Mean PM2.5 levels were high in both urban (143 micrograms/m3) and suburban (139 micrograms/m3) areas. A generalized additive model was used to estimate the differences between the two areas in FEV1, FVC, and FEV1/FVC%, with adjustment for potential confounding factors, including sex, age, height, education, passive smoking, and occupational exposures to dust, gas, or fumes. Estimated differences in FEV1 between the urban and suburban areas were 199 mL (SE = 50 mL) for men and 87 mL (SE = 30 mL) for women, both statistically significant. When the men and women were pooled, the estimated difference in FEV1 was 126 mL (SE = 27 mL). Similar trends were observed for FVC and FEV1/FVC%. After exclusion of 104 subjects with histories of occupational exposures to dust, gas, or fumes, the estimated difference was some-what smaller than that of the total samples. However, the effects on FEV1 and FEV1/FVC% remained significant for both men and women.