Tuberculosis and Diabetes Linked

Diabetes is an important risk factor for tuberculosis that has hitherto been neglected by clinicians -- they do not routinely screen for it. In fact, many clinicians who treat tuberculosis are completely unaware of this essential link.

This article suggests that patients with both diabetes and tuberculosis may present differently from patients who have only tuberculosis, and that they may have a different prognosis. Diabetes may also complicate the management of tuberculosis.

The link between tuberculosis and diabetes has been known for at least a thousand years. Avicenna noted in the 11th century AD that diabetes was frequently complicated by phthisis (often a terminal event).  In the pre-insulin era, patients with diabetes who did not die as a result of a diabetic coma frequently died of tuberculosis. Yet, although it is standard practice for all tuberculosis patients to be screened for HIV and despite the preponderance of evidence linking tuberculosis to diabetes,^[1] there are not currently any recommendations to screen tuberculosis patients for diabetes mellitus (the WHO 2009 report on tuberculosis does not even mention diabetes).

Uncontrolled diabetes is strongly associated with cardiovascular disease, retinal disease and an increased risk of other infections. All these complications may be prevented by achieving good glycemic control, and this is contingent on early recognition and treatment. This review summarizes the epidemiology and management of patients with diabetes and tuberculosis.

Treatment failure is more common in diabetics and mortality is higher. Diabetes medication may interact with tuberculosis treatment (rifampicin in particular) with corresponding complications in glycemic control. Each year, cases in India account for one-fifth of newly diagnosed tuberculosis cases worldwide, of whom almost half have diabetes. From 1989 to 1995, there was a 40% increase in the prevalence of diabetes in Chennai, India. The prevalence of diabetes worldwide is close to 10% and the relative risk of tuberculosis varies from 3 to > 8 depending on the study.

The increasing prevalence of obesity parallels the current diabetes epidemic. A meta-analysis by Jeon and Murray, looked at 1,786,212 patients in total and found that diabetes patients had a relative risk for TB of 3.11. A cost-effectiveness study for diabetes screening in hypertension found that screening in this patient group was cost effective, and they assumed a prevalence of undiagnosed diabetes in the hypertensive group that was < 2 times that in the non-hypertensive group. To a good approximation, the prevalence of diabetes in most populations is 10%: the Indian data shows a prevalence of diabetes in the TB population that approaches 50%. Given recent disappointing data on the performance of screening tests such as PSA, which were implemented without cost-effectiveness data, the situation for diabetes is quite unlike that for cancer in that the screening test is also the diagnostic test. In screening for prostate cancer, for example, prostate-specific antigen (PSA) and digital rectal exam (DRE) are not diagnostic tests and must be followed up with prostate biopsy, with its attendant costs, morbidity and mortality.

Screening for diabetes is neither time-consuming not expensive, whereas the benefits of early detection of diabetes are well established. For this reason, the U.S. Preventive Services Task Force in 2003 saw fit to recommended diabetes screening for all patients with hypertension or dyslipidaemia without any prior cost-effectiveness analysis but based their recommendation on similar data to what we have now for tuberculosis. Although nay-sayers will argue that we need cost-effectiveness studies before recommending routine diabetes screening for TB patients, such studies seem very unlikely to show diabetes screening not to be cost-effective given the current data.

1. Jeon and Murray, PloS Med 2008, 5:e152 [PMID:18630984].
2. Young et al. Indian J Med Res 2009, 130:1-4 [PMID:19700792].
3. World Health Organization, Global Tuberculosis Control - Epidemiology, Strategy, Financing. WHO Report 2009, Geneva.
4. Hoerger et al. Ann Int Med 2004, 140:689-99 [PMID:15126252].