Effects and Interaction of Meteorological Factors on Pulmonary Tuberculosis in Urumqi, China, 2013-2019

Abstract

Background: Most existing studies have only investigated the delayed effect of meteorological factors on pulmonary tuberculosis (PTB). However, the effect of extreme climate and the interaction between meteorological factors on PTB has been rarely investigated.

Methods: Newly diagonsed PTB cases and meteorological factors in Urumqi in each week between 2013 and 2019 were collected. The lag-exposure-response relationship between meteorological factors and PTB was analyzed using the distributed lag non-linear model (DLNM). The generalized additive model (GAM) was used to visualize the interaction between meteorological factors. Stratified analysis was used to explore the impact of meteorological factors on PTB in different stratification and RERI, AP and SI were used to quantitatively evaluate the interaction between meteorological factors.

Results: A total of 16,793 newly diagnosed PTB cases were documented in Urumqi, China from 2013 to 2019. The median (interquartile range) temperature, relative humidity, wind speed, and PTB cases were measured as 11.3°C (-5.0-20.5), 57.7% (50.7-64.2), 4.1m/s (3.4-4.7), and 47 (37-56), respectively. The effects of temperature, relative humidity and wind speed on PTB were non-linear, which were found with the "N"-shaped, "L"-shaped, "N"-shaped distribution, respectively. With the median meteorological factor as a reference, extreme low temperature was found to have a protective effect on PTB. However, extreme high temperature, extreme high relative humidity, and extreme high wind speed were found to increase the risk of PTB and peaked at 31.8°C, 83.2%, and 7.6 m/s respectively. According to the existing monitoring data, no obvious interaction between meteorological factors was found, but low temperature and low humidity (RR = 1.149, 95%CI: 1.003-1.315), low temperature and low wind speed (RR = 1.273, 95%CI: 1.146-1.415) were more likely to cause the high incidence of PTB.

Conclusion: Temperature, relative humidity and wind speed were found to play vital roles in PTB incidence with delayed and non-linear effects. Extreme high temperature, extreme high relative humidity, and extreme high wind speed could increase the risk of PTB. Moreover, low temperature and low humidity, low temperature and low wind speed may increase the incidence of PTB.

Keywords: distributed lag non-linear model (DLNM); interaction; meteorological; pulmonary tuberculosis (PTB); seasonally.

Figure 1

3D plots of temperature, relative humidity, wind speed on PTB at 0–12 lag weeks. The median value was reference.

Figure 2

Contour plots of relative risks of meteorological factors on PTB at 0–12 lag weeks. The median value was reference.

Figure 3

Cumulative effects between meteorological factors and the risk of pulmonary tuberculosis at 0–12 lag weeks. The median value was reference.

Figure 4

Lag-response curves for P1, P5, P95, P99 of meteorological variables on PTB. P1, the 1th percentile; P5, the 5th percentile; P95, the 95th percentile; P99, the 99th percentile. The median value was reference. Red asterisk and blue asterisk indicate the mean relative risk, and the black vertical line corresponds to 95%CI.

Figure 5

The interaction effect of meteorological variables on PTB after a lag of 12 weeks.