Short-term association between sulfur dioxide and daily mortality: the Public Health and Air Pollution in Asia (PAPA) study

Abstract

Sulfur dioxide (SO(2)) has been associated with increased mortality and morbidity, but only few studies were conducted in Asian countries. Previous studies suggest that SO(2) may have adverse health effects independent of other pollutants. In the Public Health and Air Pollution in Asia (PAPA) project, the short-term associations between ambient sulfur dioxide (SO(2)) and daily mortality were examined in Bangkok, Thailand, and three Chinese cities: Hong Kong, Shanghai, and Wuhan. Poisson regression models incorporating natural spline smoothing functions were used to adjust for seasonality and other time-varying covariates. Effect estimates were obtained for each city and then for the cities combined. The impact of alternative model specifications, such as lag structure of pollutants and degree of freedom (df) for time trend, on the estimated effects of SO(2) were also examined. In both individual-city and combined analysis, significant effects of SO(2) on total non-accidental and cardiopulmonary mortality were observed. An increase of 10 microg/m(3) of 2-day moving average concentrations of SO(2) corresponded to 1.00% [95% confidence interval (CI), 0.75-1.24], 1.09% (95% CI, 0.71-1.47), and 1.47% (95% CI, 0.85-2.08) increase of total, cardiovascular and respiratory mortality, respectively, in the combined analysis. Sensitivity analyzes suggested that these findings were generally insensitive to alternative model specifications. After adjustment for PM(10) or O(3), the effect of SO(2) remained significant in three Chinese cities. However, adjustment for NO(2) diminished the associations and rendered them statistically insignificant in all four cities. In conclusion, ambient SO(2) concentration was associated with daily mortality in these four Asian cities. These associations may be attributable to SO(2) serving as a surrogate of other substances. Our findings suggest that the role of outdoor exposure to SO(2) should be investigated further in this region.

Figure 1

Smoothing plots of SO₂ against total mortality risk (df = 3) in each city. X-axis is the 2-day average (lag01) SO₂ concentrations (μg/m³). The solid lines indicate the estimated mean percentage of change in daily mortality, and the dotted lines represent twice the standard error

Figure 2

Percent increase of total mortality associated with 10 μg/m³ increase of SO₂, using different lag structures (lag0, lag01, and lag 04) – individual and combined effects. X-axis is lag structures. Y-axis is percent increase of total mortality.

Figure 3

Percent increase of total mortality associated with 10 μg/m³ increase of 2-day moving average SO₂, using different df/yr for time trend

Figure 4

Percent increase of mortality outcomes associated with 10 μg/m³ increase of 2-day moving average SO₂ with single and two-pollutant models ^{a a} S: sing-pollutant model; C1: two-pollutant model (SO₂ + NO₂); C2: two-pollutant model (SO₂ + PM₁₀); C3: two-pollutant model (SO₂ + O₃)