Table 4.
Multivariable cox regression analysis of predictors associated with pulmonary tuberculosis treatment success among artisanal miners and relatives.
| Treatment success | ||||||
|---|---|---|---|---|---|---|
| Variables | Model 1 | Model 2 | ||||
| All participants (n = 122) | Non-miners (n = 41) | Miners (n = 81) | ||||
| aHR (95% CI) | p-value | aHR (95% CI) | p-value | aHR (95% CI) | p-value | |
| Mining-related activities | ||||||
| No | Ref | |||||
| Yes | 1.17 (0.84–1.63) | 0.35 | – | – | – | – |
| Doubling of TL | 1.27 (1.05–1.44) | 0.019 | 1.32 (0.86–1.59) | 0.14 | 1.36 (1.14–1.52) | 0.003 |
| Doubling of mtDNA content | 1.11 (0.65–1.88) | 0.65 | 1.17 (0.64–2.18) | 0.61 | 1.24 (0.48–3.21) | 0.65 |
Hazard ratios and p-values were computed by Cox regression and expressed for a doubling in TL and mtDNA content at baseline. Model 1 includes all participants. Model 2 stratifies participants by mining related-activities. All models are adjusted for age, sex, tobacco smoking, TB category, years working in artisanal mining (when applicable), HIV-infection status and body mass index. Model 1 also adjusted for Xpert MTB/RIF result.
TL leukocyte telomere length, mtDNA mitochondrial DNA, aHR adjusted hazards ratio, Ref. reference.