Background: Asian dust storms (ADS) have affected several Asian countries and have been major issues due to adverse effects in the perspective of public health. The occurrence of Asian dust storms is different by each country according to its geograph...
Background: Asian dust storms (ADS) have affected several Asian countries and have been major issues due to adverse effects in the perspective of public health. The occurrence of Asian dust storms is different by each country according to its geographical features and distance from the origins. Considerable studies have found significant associations between ADS and morbidity. However, about the association between ADS and mortality, no study found the association in Japan and in Taiwan and Korea, only a few studies showed statistically significant ADS effects.
Aims: This study aims to examine the effect of ADS on daily mortality in three Asian cities (Seoul, South Korea; Taipei, Taiwan; and Kitakyushu, Japan) and to examine the difference in the extent of ADS effects by each city.
Methods: We performed time-series analyses using generalized additive model with Quasi-Poisson regressions. Deaths due to accidents or external causes were excluded. Temperature, relative humidity, barometric pressure, day of the week, and season and time trends were controlled for in a basic model. We considered single day lag effects of ADS up to 7 days and performed subgroup analyses stratified by disease (cardiovascular, respiratory) and age(<65, ≥65). Additional analyses controlling for NO2, SO2, and PM10, as the source of local air pollution were done respectively. We reported results as Excess mortality by percentage due to Asian dust storms.
Results: We found consistent Excess mortality in Seoul and Kitakyushu as follows. In Seoul, ADS showed adverse effects on non-accidental mortality below 65 years old (lag 2: 4.61%; 95% CI: 1.07, 8.28, lag 3: 5.17%; 95% CI: 1.63, 8.82, lag 4: 4.55%; 95% CI: 1.02, 8.2), respiratory mortality below 65 years old (lag 3: 37.53%; 95% CI: 12.85, 67.62), and cardiovascular mortality above 65 years old (lag 5: 5.32%; 95%CI: 0.48, 10.39). In Kitakyushu, ADS had adverse effects on respiratory mortality (lag 2: 21.89%; 95%CI: 4.17, 42.63), respiratory mortality below 65 years old (lag 2: 126.15%; 95% CI: 45.32, 251.93, lag 3: 137.77; 95% CI: 52.48, 270.75), cardiovascular mortality (lag 6: 12.45%; 95% CI: 0.14, 26.28) and cardiovascular mortality below 65 years old (lag 3: 44.36%; 95% CI: 4.38, 99.65). Contradictory to results in Seoul and Kitakyushu, ADS seemed to have a protective effect in Taipei: non-accidental mortality above 65 years old (lag 0: -3.61%; 95% CI: -6.34, -0.80, lag 1: -3.46%, 95% CI: -6.19, -0.65), respiratory mortality above 65 years old (lag 0: -9.95%; 95% CI: -16.99, -2.3, lag 1: -8.15%; 95% CI: -15.31, -0.39), and cardiovascular mortality below 65 years old (lag 5: -10.59%; 95% CI: -18.78, -1.58). Sensitivity analyses showed similar findings as main results.
Conclusion: Our results showed significant adverse effects of Asian dust storms on daily mortality in cities in Korea and Japan, and a positive effect in a city in Taiwan. The results were still significant after controlling for NO2, SO2, and PM10.
Key words: Asian dust storms; non-accidental mortality; lag effects; generalized additive model