103 4 Diabetes-tuberculosis comorbidity in a low-income setting high school education (AOR 0.32; 95% CI: 0.16, 0.64; p=0.001) was a significant negative predictor of unfavourable TB outcome. The odds of an unfavourable TB outcome increased with alcohol use (AOR 1.68; 95% CI: 0.97, 2.89), but this was only marginally significant (p=0.062). Discussion The prevalence of elevated blood glucose for patients commencing TB treatment at baseline was 8% (95% CI: 5.5, 11.3). It was highest in the Hhohho region, higher in males than females, increased with age and highest in those with a reactive HIV status at 9.5% compared to the non-reactive group. The proportion of patients with an elevated blood glucose measurement reduced at the second-month visit and even further at the fifthmonth visit. At multivariate analysis, a family history of DM and a reactive HIV status were significant predictors of an elevated baseline blood glucose. Three-quarters of the participants (75.4%) had a favourable treatment outcome. Elevated baseline blood glucose was not associated with unfavourable treatment outcomes; instead, hypertension and unemployment predicted unfavourable treatment outcomes, while high school education was protective. Comparison with other studies Our reported prevalence of 8.0% elevated blood glucose at diagnosis is similar to a pooled prevalence of 9% (95% CI:6.0%, 12.0%) of DM for patients in Sub-Saharan Africa from a 2019 systematic review and meta-analysis [26]. This review similarly reported a higher prevalence in HIV-infected patients at 8.9%, with a DM prevalence of 15%, 11% and 10% in Nigeria, Tanzania and Ethiopia, respectively, indicating variations across countries [26]. Other studies from Tanzania, Uganda and Ethiopia have reported a prevalence of 9.2%, 8.5% and 5.1%, respectively. Our study prevalence is less than 17.7% reported from a global meta-analysis on the common comorbid conditions with TB [27] and in Asian countries of Nepal [28], India [29,30], Iran [31], Vietnam [32] and Pakistan [33]. This higher prevalence in Asian countries is expected as they are known to have a higher prevalence of DM and TB than the rest of the world. In contrast, a much lower prevalence of 1.9% and 4.5% have been reported in Benin [34] and Brazil [35], respectively. Our reported DM prevalence of 12% from a sub-analysis of 33 patients should be interpreted cautiously, as this could have been due to selection bias. Consistent with our findings, a study from Tanzania [14] found that positive HIV status and a family history of DM were significant predictors of DM in patients receiving treatment for TB. Studies from India, Vietnam and Iran also corroborate this finding [32,36,37]. Another
RkJQdWJsaXNoZXIy MTk4NDMw