Association of Occupation and Exposures With Development of Post-tuberculosis Lung Disease in a Ugandan Cohort

Abstract Background: 10.6 million people were infected with Tuberculosis (TB) in 2022. Cure rates have increased to approximately 80% globally and estimates suggest there are 155 million survivors of TB globally. Unfortunately, 40-60% of those cured of TB will have persistent symptoms or lung dysfunction, termed post-TB lung disease (PTLD). There is a growing interest in understanding risk factors and pathogenesis of PTLD. Methods: Adults with TB confirmed by sputum testing and abnormal x-ray were prospectively enrolled from a consortium of 5 TB clinics in Uganda. At TB diagnosis patients completed questionnaires including modified BOLD exposure and occupation questionnaires with 12 specific occupations and 6 specific exposures known to increase risk for lung disease. Each questionnaire contained an additional option for other exposures or occupations of concern to subjects. After microbiologic cure, generally at 6-months, pulmonary function testing including diffusion capacity testing and total lung capacity was performed. Subjects with FEV1/FVC ratio z-score <-1.64 or diffusion capacity or total lung capacity <80% predicted using GLI race-neutral reference values were classified as having PTLD. Chi-Square testing was utilized to analyze associations between exposures and PTLD. Results: After TB cure and pulmonary function testing 109 subjects were reviewed, 66 controls and 43 cases with PTLD. The groups did not differ by age or gender (p>0.05) for both comparisons. A significantly higher proportion of cases (6/43=11.6% than controls (3/66=4.55%) reported being employed as a cook using biomass (p=0.041). The proportion reporting any occupation of risk compared with those reporting no occupation of risk showed no difference between cases and controls (30/43=69.8%) vs. (44/66=66.7%) p=0.74. In regards to exposures, more cases reported exposure to chemical fumes (3/43=7%) vs. (0/66=0%) p=0.03 and when analyzing all exposures of risk vs. no exposure of risk cases reported a higher proportion of exposures (11/43=25.6%) vs. (7/66=10.6%) p=0.040. Discussion: We found an association between employment as a cook using biomass, exposure to chemical fumes and exposure to substances with known pulmonary risk in subjects with PTLD compared to controls at TB diagnosis. Our study is limited by lack of baseline lung function before enrollment. It is possible that exposures caused pre-existing lung disease without contributing to risk for PTLD development. Additionally, reliance on interviews raises risk of recall bias. Despite these limitations, this data suggests a possible contribution of environmental exposures to PTLD development. Further study in larger cohorts with objective measurements of exposures is warranted.