Study of the relationship between occurrence of Kawasaki disease and air pollution in Chengdu by parametric and semi-parametric models

Abstract Kawasaki disease (KD) is a pediatric vasculitis of unknown etiology which mainly associates the development of coronary artery aneurysms. The etiology of KD seems to be multifactorial, but rare research on the association between KD and potential environmental risk factors. So, we would like to examine the correlation between KD and potential environmental risk factors in west China. We included KD patients in Chengdu from 2015 to 2021, and analyzed the correlation with air pollution indexes and climate condition indexes. The autocorrelation of the data was eliminated by first-order difference, and the risk factors were screened by stepwise regression with AIC criterion and the multiple regression model was established. Random forest and Winsorize were used to test the robustness of the screening results, and it was found that particulate matter with diameter less than or equal to 2.5μm (PM2.5) had a significant positive effect on the incidence of KD. In addition, several variables were positively correlated with KD incidence, but not statistically significant. GAM model was used to explore the nonlinear correlation between PM2.5 and KD incidence. The results showed that PM2.5 concentration was positively correlated with KD incidence, and the effects were varied among different concentration levels of PM2.5. Fisher's exact test was used to explore the influence of PM2.5 on the incidence of coronary tumors. It is found that PM2.5 may be a risk factor for it. This study suggested that exposure to high concentrations of PM2.5 may significantly increase the risk of KD. The evidence for the association between other environmental factors and KD incidence, as well as the association between PM2.5 and coronary tumors, was limited and needed further verification.