Differentiating transient from persistent diabetic range hyperglycemia in a cohort of people completing tuberculosis treatment in Dhaka, Bangladesh

Glycemic control during TB treatment among Filipinos: The Starting Anti-Tuberculosis Treatment Cohort Study

Poor TB treatment outcomes are observed in patients with type 2 diabetes mellitus (DM) comorbidity and glycemic control throughout treatment may play a role. The objective of this study was to investigate glycemic control longitudinally among Filipino adults undergoing TB treatment using mixed-effects linear and logistic regression. Analyses were conducted in 188 DM-TB patients out of 901 enrolled in the Starting Anti-TB Treatment (St-ATT) cohort, with a median baseline glycosylated hemoglobin (HbA1c) of 8.2% (range 4.5-13.3%). Previous versus new DM diagnosis was associated with higher mean HbA1c (worse glycemic control) during treatment, with a smaller effect amongst those with central obesity (coefficient 0.80, 95% confidence interval [CI] 0.26, 1.57, P = 0.043) than amongst those without central obesity (coefficient 3.48, 95% CI 2.16, 4.80, P<0.001). In those with a new DM diagnosis, central obesity was associated with higher blood glucose (coefficient 1.62, 95% CI 0.72, 2.53, P = 0.009). Of 177 participants with ≥2 HbA1c results, 40% had uncontrolled glycemia (≥2 HbA1c results ≥8%). Of 165 participants with ≥3 HbA1c results, 29.9% had consistently-controlled glycemia, 15.3% had initially-uncontrolled glycemia, and 18.6% had consistently-uncontrolled glycemia. Previous versus new DM diagnosis and glucose-lowering medication use versus no use were associated with having uncontrolled versus controlled glycemia (adjusted odds ratio [aOR] 2.50 95%CI 1.61, 6.05, P = 0.042; aOR 4.78 95% CI 1.61,14.23, P<0.001) and more likely to have consistently-uncontrolled versus consistently-controlled glycemia (adjusted relative risk ratio [aRRR] 5.14 95% CI 1.37, 19.20, P = 0.015; aRRR 10.24 95% CI 0.07, 0.95, P = 0.003). Relapse cases of TB were less likely than new cases to have uncontrolled (aOR 0.20 95%CI 0.06, 0.63, P = 0.031) or consistently-uncontrolled (aRRR 0.25 95%CI 0.07, 0.95, P = 0.042) versus controlled glycemia. Those with long-term DM, suggested by previous diagnosis, glucose-lowering medication use and possibly central obesity, may require additional support to manage blood glucose during TB treatment.

Rethinking the syndemic of tuberculosis and dysglycaemia: a Kenyan perspective on dysglycaemia as a neglected risk factor for tuberculosis

Background

The END TB 2035 goal has a long way to go in low-income and low/middle-income countries (LICs and LMICs) from the perspective of a non-communicable disease (NCD) control interaction with tuberculosis (TB). The World Health Organization has identified diabetes as a determinant for, and an important yet neglected risk factor for tuberculosis. National guidelines have dictated testing time points, but these tend to be at an isolated time point rather than over a period of time. This article aims to give perspective on the syndemic interaction of tuberculosis and dysglycaemia and how the gaps in addressing the two may hamper progress towards END TB 2035.

Main text

Glycated haemoglobin (HbA1C) has a strong predictive association with the progression to subsequent diabetes. Therefore, screening using this measure could be a good way to screen at TB initiation therapy, in lieu of using the random blood sugar or fasting plasma glucose only. HbA1C has an observed gradient with mortality risk making it an informative predictor of outcomes. Determining the progression of dysglycaemia from diagnosis to end of treatment and shortly after may offer information on the best time point to screen and follow-up. Despite TB and Human Immunodeficiency Virus (HIV) disease care being free, hidden costs remain. These costs are additive if there is accompanying dysglycaemia. Regardless of receiving TB treatment, it is estimated that almost half of persons affected by pulmonary TB develop post-TB lung disease (PTLD) as an outcome and the contribution of dysglycaemia is not well described.

Conclusions

Establishing costs of treating TB with diabetes/prediabetes alone and in the additional context of HIV co-infection will inform policy makers on what it takes, financially, to treat these patients and subsidize dysglycaemia care. In Kenya, cardiovascular disease is only rivalled by infectious disease as a cause of mortality, and diabetes is a well-described risk factor for cardiac disease. In poor countries, communicable diseases are responsible for majority of the mortality burden, but societal shifts and rural-urban migration may have contributed to the observed increase of NCDs.