Prediabetes and the treatment outcome of tuberculosis: A meta-analysis.

Abstract

Diabetes has been related to higher risk and poor prognosis of patients with tuberculosis, while the influence of prediabetes on the treatment outcome of patients with tuberculosis remains not determined. A meta-analysis was performed to evaluate the influence of prediabetes on treatment outcome of patients with tuberculosis. Relevant cohort studies were acquired through a search of Medline, Embase, and Web of Science databases. To minimise the influence of between-study heterogeneity, a randomised-effects model was used to pool the results. Eight prospective cohort studies including 3001 patients with tuberculosis were available for the meta-analysis. Among them, 752 (25.1%) were with prediabetes at baseline, and the patients were followed for a mean duration of 17.7 months. It was shown that compared to patients with normoglycemia, those with prediabetes were associated with a higher incidence of unfavourable treatment outcome (risk ratio [RR]: 1.41, 95% confidence interval [CI]: 1.02 to 1.96, p = 0.04; I2 = 56%). Subgroup analysis did not support that difference in study country (Asian or non-Asian), diagnosis (pulmonary tuberculosis only or also with extrapulmonary tuberculosis), mean age, follow-up duration, or study quality score had significant influence on the results (p for subgroup difference all >0.05). However, prediabetes at baseline was not associated with an increased risk of all-cause mortality during follow-up (RR: 1.59, 95% CI: 0.75 to 3.38, p = 0.23; I2 = 54%). Patients with tuberculosis and prediabetes may have a higher risk of unfavourable treatment outcome compared to patients with normoglycemia.