Abstract
INTRODUCTION: Influenza-like illness (ILI) represents a significant global public health challenge influenced by environmental factors. While previous studies have demonstrated associations, most have been limited to single-city analyses with inconsistent findings. This multi-city study systematically examines the effects of meteorological and air pollution factors on ILI across diverse urban environments. STUDY DESIGN: We analyzed daily ILI surveillance data (2015-2023) from 18 sentinel hospitals across nine Fujian Province cities, combined with concurrent air quality and meteorological data. Using LASSO regression for variable selection, we employed distributed lag non-linear models (DLNMs) to characterize exposure-response relationships in each city, followed by random-effects multivariate meta-analysis to pool estimates and assess heterogeneity. RESULTS: A total of 2,995,909 ILI cases were collected. ILI cases in nine cities of Fujian Province exhibited significant seasonal fluctuations, peaking in winter or early summer. LASSO regression identified temperature, NO₂, and SO₂ as key environmental factors. Our results indicated that the combined cumulative effect of average temperature on ILI across nine cities decreased with rising temperatures, with a risk peak at -0.1°C. The NO₂-ILI association was non-linear, resembling an inverted "U" shape, with a risk peak at 40.5 μg/m(3). SO₂ exposure had a large degree of heterogeneity in its effect on ILI. CONCLUSION: This study provides robust evidence that ambient temperature and NO₂ levels significantly influence ILI risk in Fujian Province, with distinct exposure-response patterns. Public health strategies should prioritize cold-weather preparedness and air quality management, particularly during high-risk seasons. Our two-stage analytical approach addresses previous limitations in multi-city environmental health studies.