Spatial-temporal epidemiological characteristics of influenza and its correlation with meteorological factors in...

To describe the spatial-temporal epidemiological characteristics of influenza outpatient visits in Gansu province, and to explore their correlation with temperature and humidity using the 24 solar terms (24-STs) as the time dimension.

Using influenza visit data from Gansu (2019-2024), spatiotemporal analysis employed gravity center shift, standard deviational ellipses, and seasonal decomposition. A distributed lag non-linear model (DLNM) assessed the effects of daily mean temperature (DMT, °C) and daily mean relative humidity (DMRH, %) on influenza visits, using a bivariate model to estimate joint effects.

A total of 57,443 influenza visits were recorded. Visits peaked between Beginning of Winter and Winter Solstice, with Great Snow accounting for 10.62%. The northwest-southeast ellipse orientation was consistent with Gansu's geography. DMT and DMRH exhibited significant nonlinear lag effects, with solar term-specific characteristics. The temperature risk window spanned Cold Dew to Frost's Descent (6-16 °C; RR = 1.459, 95% CI: 1.246-1.709). Humidity risk increased above 50% in spring, peaking at 50% in winter (RR = 1.869, 95% CI: 1.582-2.208). Synergistic high-risk combinations and multiple visit peaks were identified across specific solar terms.

The 24-STs may serve as a useful exploratory time frame for characterizing influenza activity. Daily mean temperature, daily mean relative humidity, and their joint effects are key meteorological factors associated with influenza visits, with notable differences across solar terms.