RESUMO
Objective: To analyze the effect of air quality index (AQI) on the incidence of tuberculosis (TB) in Beijing, and to provide evidence for setting up a better program regarding prevention and control of tuberculosis. Methods: Generalized additive model (GAM) was used to analyze the association between AQI and the incidence of tuberculosis in Beijing, from January 1, 2014 to November 9, 2016. Confounding factors as meteorological conditions and time trends were under control. Results: In Beijing, a total of 14 533 TB cases with definite dates of onset were collected during the study period, with 36 children excluded from the study. Finally, 14 497 cases were included in the study, including 9 513 men and 4 984 women, with 11 290 adults (15-59 years old) and 3 207 elderly (≥60 years old). Data from the optimal single-day lag effect of GAM showed that,with every 10 increase of AQI, the percent of increase on the onsets of overall, male, female and adult; tuberculosis cases were 0.85% (95%CI: 0.26%-1.44%), 0.83% (95%CI: 0.24%-1.42%), 0.93% (95%CI: 0.24%-1.62%) and 0.88% (95%CI: 0.29%-1.46%), respectively. The optimal lag time of the single-day effects were 15 days (lag15), but 16 days (lag16) for male. The optimal cumulative lag effect showed that with every 10 AQI increase, the percent of increase on the onsets of overall, male, female and adult tuberculosis cases were 1.92% (95%CI: 0.23%-3.16%), 1.94% (95%CI:0.15%-3.72%), 2.04% (95%CI: 0.10%-3.97%) and 2.00% (95%CI: 0.30%-3.69%), respectively, with the optimal lag time of cumulative delayed effects as 17 days (lag0_17), 18 days (lag0_18), 16 days (lag0_16) and 17 days (lag0_17), respectively. However, there were no statistical significances noticed in the elderly cases. Conclusion: There was a positive correlation between AQI and the number of TB cases in Beijing, and the effects of AQI on the number of TB cases in different genders and age groups were different.