Impact of diabetes and smoking on one-year mortality attributable to TB.^*

*Smoking included a ‘missing’ group (not shown) in order to not lose data (5% missing); Age as continuous (linear) and an age indicator (<50 vs. ≥50) were included together for unadjusted age association. Combined Diabetes and smoker are additive main effects (not an interaction); comparison of diabetic smoker vs. non-diabetic non-smoker. **The column TB Deaths Unadjusted shows estimated unadjusted associations of each factor with death. The column TB Deaths Adjusted shows estimated adjusted associations from a single model using all the factors in the table (each factor is adjusted for all other factors in that table). The combined effect at the bottom is a combination of coefficients for diabetes and smoking from the adjusted model. ***Reference value.