Few evidences are available about the impact of temperature variation on childhood asthma in different seasons. This study aimed to assess the influence of temperature changes between neighboring days (TCN) on the exacerbation of asthma among children. Daily outpatient visits for childhood asthma (DOVCA) were collected from 17 main hospitals in Shanghai, China, from 2016 to 2018. A quasi-Poisson regression combined with distributed lagged nonlinear models was employed to estimate the association between TCN and asthma visits in cool or warm seasons, after controlling for short- and long-term trends, day of week, holidays, daily mean temperature, daily mean relative humidity, and air pollutants. The TCN varied from - 9.6 to 6.7 °C. The relationship between TCN and DOVCA greatly varied by season. In warm seasons, positive TCN (temperature rise) was associated with higher risks of asthma outpatient visits and negative TCN (temperature drop) was associated with lower risks; the associations were present on lag 1 day and lasted for 2 weeks; the cumulative relative risk of childhood asthma over 0 to 14 days was 1.98 (95% confidence interval: 1.42, 2.76) and 0.31 (95% confidence intervals: 0.21, 0.44) comparing a TCN of 2.5 °C (5th percentile) and - 3.2 °C (95th percentile) with 0 °C, respectively. In cool seasons, neither negative nor positive TCN showed significant risks. In conclusion, temperature rise might increase the risk of childhood asthma exacerbation and temperature drop might decrease the risks in warm seasons. There were no statistically significant influences in cool seasons.
Keywords: Asthma; Children; Distributed lag nonlinear model; Outpatient visits; Temperature variability.