Association between wet-bulb globe temperature with gastroesophageal reflux disease in different geographic regions in a large Taiwanese population study

Abstract

In this study, we investigated the association between wet-bulb globe temperature (WBGT) and gastroesophageal reflux disease (GERD) in different geographic regions of Taiwan in a large cohort. Participants (n = 120,424) residing in the four main geographical areas of Taiwan (North, Central, South and Eastern Taiwan) were enrolled from the Taiwan Biobank. Questionnaires were used to ascertain the presence of GERD based on self-reported physician diagnosis, and WBGT was assessed separately during working (8:00 AM to 5:00 PM) and noon (11:00 AM to 2:00 PM) periods. In Northern Taiwan, there is no significant association between WBGT and GERD during either the noon or working period. In Central Taiwan, the 1-, 3-, and 5-year WBGT values per 1℃ increase during the noon period (odds ratio [OR], 1.055 [95% CI 1.008-1.105]; 1.062 [95% CI 1.013-1.114]; 1.059 [95% CI 1.009-1.111]) and working period (OR, 1.089 [95% CI 1.034-1.146]; 1.092 [95% CI 1.037-1.150]; 1.084 [95% CI 1.031-1.139]) were significantly associated with GERD. Similarly, in Southern Taiwan, the 1-, 3-, and 5-year WBGT values per 1℃ increase during the noon period (OR, 1.292 [95% CI 1.236-1.351]; 1.323 [95% CI 1.261-1.389]; 1.386 [95% CI 1.316-1.460]) and working period (OR, 1.238 [95% CI 1.190-1.288]; 1.247 [95% CI 1.196-1.301]; 1.259 [95% CI 1.204-1.318]) were significantly associated with GERD. However, in Eastern Taiwan, the 1-, 3-, and 5-year WBGT values per 1℃ increase during the noon period were significantly associated with GERD (OR, 1.067 [95% CI 1.015-1.122]; 1.089 [95% CI 1.041-1.138]; 1.107 [95% CI 1.058-1.158]), whereas the 1-, 3-, and 5-year WBGT values per 1℃ decrease during the working period were significantly associated with GERD. Increases in average WBGT values were significantly associated with GERD during both the noon and working periods in Central and Southern Taiwan, and the impact of WBGT was much stronger in Southern Taiwan. While a similar result was found in Eastern Taiwan during the noon period, a reverse correlation was found during the working period. Our findings suggest that heat stress may be associated with GERD, although the impact may differ according to regional characteristics. The causal relationship could not be confirmed due to the cross-sectional design of the study. Further longitudinal studies are warranted to establish the relationship.

Keywords: Different geographic region; Gastroesophageal reflux disease; Noon and work period; Taiwan biobank; Wet-bulb Globe temperature.

Fig. 1

The distribution of different regions in Taiwan. The climate in the South of Taiwan is tropical. The climate in the North and Center is subtropical. Due to the length of the Eastern region, it encompasses both subtropical and tropical zones.

Fig. 2

Forest plots of the association of WBGT in noon period with GERD using multivariable logistic regression analysis stratification by geographic regions. In Central, Southern and Eastern Taiwan, the 1-, 3-, and 5-year average WBGT values per 1℃ increase were associated with GERD.

Fig. 3

Forest plots of the association of WBGT in work period with GERD using multivariable logistic regression analysis stratification by geographic regions. In Central and Southern Taiwan, the 1-, 3-, and 5-year average WBGT values per 1℃ increase were associated with GERD, whereas in Eastern Taiwan, the 1-, 3-, and 5-year average WBGT values per 1℃ decrease were associated with GERD.

Fig. 4

Nonlinear relationship between the noon period: 1-year (A), 3-year (B), and 5-year (C), and work period: 1-year (D), 3-year (E), and 5-year (F) average WBGT and the risk of GERD across different geographic regions. The relationship between the 1-year, 3-year, and 5-year average WBGT and the risk of GERD exhibited nonlinear effects, with statistical significance observed for both the noon and work period using a restricted cubic spline analysis with five knots was conducted using a logistic regression model.