Distributed lag nonlinear model-based analysis of effects of apparent temperature on the number of hypertension-associated hospitalizations in Urumqi, China

Objective To investigate the relationship between apparent temperature and the number of hypertension-associated hospitalizations as well as the lag effects, and to analyze the effects of apparent temperature for different grades of hypertension.

Methods The information on hypertension-associated hospitalizations in 10 hospitals in Urumqi, China from 2019 to 2021 as well as the meteorological and air pollution data (daily average barometric pressure, ambient temperature, relative humidity, wind speed, sunshine duration, NO₂, SO₂, CO, O₃, PM₁₀, and PM_2.5) from Urumqi Meteorological Service and Environmental Monitoring Station were collected. A distributional lag nonlinear model was used to explore the exposure-lag-response relationship between apparent temperature and the number of hypertension-associated hospitalizations.

Results The exposure-response curves revealed nonlinear associations of daily average apparent temperature with the total number of hypertension-associated hospitalizations, the number of grade-2 hypertension-associated hospitalizations, and the number of grade-3 hypertension-associated hospitalizations. At the extreme low daily average apparent temperature (P₅=-17.2 ℃), the cumulative lag effects of daily average apparent temperature on daily total hypertension-associated hospitalizations and grade-3 hypertension-associated hospitalizations increased with the number of lag days, reaching the highest cumulative relative risks (CRRs) at lag 14 days, which were 2.025 (95% confidence interval CI: 1.191-3.442) and 2.171 (95% CI: 1.268-3.716), respectively. At the extreme high apparent temperature (P₉₅=25.0 ℃), the daily average apparent temperature had an impact on daily total hypertension-associated hospitalizations for hypertension, the number of hospitalizations for grade-2 hypertension, and the number of hospitalizations for grade-3 hypertension. The cumulative lag effect reached its strongest on day 2, day 9, and day 2, with CRRs of 0.72 (95% CI: 0.55-0.94), 0.57 (95% CI: 0.33-0.99), and 0.69 (95% CI: 0.53-0.91), respectively.

Conclusion Apparent temperature has a lagged and nonlinear association with the number of hypertension-associated hospitalizations. Public health departments can formulate hypertension warnings based on predicted apparent temperature to inform populations susceptible to hypertension to take precautions against apparent temperature changes, providing new ideas for reducing the incidence of hypertension.