Short-term exposure to CO and NO2 and Kawasaki disease onset in a cold region of China: a distributed lag non-linear time-series study.

BACKGROUND

Short-term exposure to ambient air pollution has been associated with Kawasaki disease (KD), however, evidence from cold climates remains limited. We aim to evaluate the association between short-term ambient pollutant exposure and KD onset in a cold region of China.

METHODS

We conducted an ecological time-series study based on aggregated daily counts of hospital-diagnosed KD onset cases. Individual clinical records from 2,562 hospitalized patients diagnosed with KD at two large tertiary medical centers between January 1, 2015, and December 31, 2022, were used to identify eligible cases and determine the dates of first fever onset. Our analysis included a year-round evaluation of multiple ambient pollutants pertinent to both heating and non-heating periods typical of cold regions. We assessed six pollutants (PM2.5, PM10, SO₂, NO₂, CO, and O₃) using a generalized linear model with distributed lag nonlinear models (GLM-DLNM), and performed stratified analyses of CO and NO₂ by gender and heating season.

RESULTS

High CO concentration (RR = 1.172; 95% CI: 1.024-1.340) and high NO₂ concentration (RR = 1.117; 95% CI: 1.009-1.237) were associated with elevated KD risk at lag 0 under single-day lag analysis. Significant cumulative effects were observed for CO (lag 0-6: RR 1.838; 95% CI: 1.022-3.306) and NO2 (lag 0-2: RR 1.298; 95% CI: 1.003-1.680). Cumulative exposure-response curves indicated that KD risk increased at concentrations above approximately 1.85 mg/m3 for CO and 53.6 μg/m3 for NO₂. In stratified analyses, these associations appeared more evident among females and during the heating season. Furthermore, O3, SO2, PM2.5 and PM10 showed no significant links.

CONCLUSION

Short-term exposure to CO and NO₂ was associated with increased daily hospital-diagnosed KD onset cases. Reducing CO and NO2 may have potential children health relevance, although further studies are needed to confirm these associations and clarify underlying mechanisms.