BMC Public Health (May 2024)

The association between humidex and tuberculosis: a two-stage modelling nationwide study in China

  • Wen Li,
  • Jia Wang,
  • Wenzhong Huang,
  • Yu Yan,
  • Yanming Liu,
  • Qi Zhao,
  • Mingting Chen,
  • Liping Yang,
  • Yuming Guo,
  • Wei Ma

DOI
https://doi.org/10.1186/s12889-024-18772-8
Journal volume & issue
Vol. 24, no. 1
pp. 1 – 8

Abstract

Read online

Abstract Background Under a changing climate, the joint effects of temperature and relative humidity on tuberculosis (TB) are poorly understood. To address this research gap, we conducted a time-series study to explore the joint effects of temperature and relative humidity on TB incidence in China, considering potential modifiers. Methods Weekly data on TB cases and meteorological factors in 22 cities across mainland China between 2011 and 2020 were collected. The proxy indicator for the combined exposure levels of temperature and relative humidity, Humidex, was calculated. First, a quasi-Poisson regression with the distributed lag non-linear model (DLNM) was constructed to examine the city-specific associations between humidex and TB incidence. Second, a multivariate meta-regression model was used to pool the city-specific effect estimates, and to explore the potential effect modifiers. Results A total of 849,676 TB cases occurred in the 22 cities between 2011 and 2020. Overall, a conspicuous J-shaped relationship between humidex and TB incidence was discerned. Specifically, a decrease in humidex was positively correlated with an increased risk of TB incidence, with a maximum relative risk (RR) of 1.40 (95% CI: 1.11–1.76). The elevated RR of TB incidence associated with low humidex (5th humidex) appeared on week 3 and could persist until week 13, with a peak at approximately week 5 (RR: 1.03, 95% CI: 1.01–1.05). The effects of low humidex on TB incidence vary by Natural Growth Rate (NGR) levels. Conclusion A J-shaped exposure-response association existed between humidex and TB incidence in China. Humidex may act as a better predictor to forecast TB incidence compared to temperature and relative humidity alone, especially in regions with higher NGRs.

Keywords