Abstract
BACKGROUND: Diabetes mellitus (DM) is a well-known risk factor for tuberculosis (TB). Metformin, which is an essential anti-diabetic drug, has been shown to exhibit anti-TB effects in patients with DM. Its effect on preventing the development of TB among patients who are newly diagnosed with DM remains unclear. METHODS: This was a retrospective cohort study using the claims database of the Korean Health Insurance Review and Assessment Service. The study population included patients who were newly diagnosed with type 2 DM and who were treated with anti-diabetic drugs between 1 January 2003 and 31 March 2011. A patient was defined as a metformin user if he/she had taken metformin for more than 28 days within 6 months since cohort entry, and as a metformin non-user if he/she had never been treated with metformin. The development of TB within 2 years after the index date was compared by Cox proportional hazard regression models between metformin users and 1:1 propensity score (PS)-matched non-users. RESULTS: Among 76,973 patients who were newly diagnosed with type 2 DM, 13,396 were classified as metformin users, 52,736 were classified as metformin non-users, and 10,841 were excluded from the final analysis. PS-matched Cox proportional hazard regression models revealed that metformin use was not associated overall with the prevention of TB development (HR 1.17; 95% CI 0.75-1.83; P = 0.482). There was a trend, however, towards a reduction in the development of TB among patients taking a higher cumulative dose of metformin. Patients who were in the highest quartile (Q4) of cumulative metformin dose had only a 10% risk of developing TB compared to metformin non-users. In contrast, during the early phases of metformin treatment, patients in the second quartile (Q2) of cumulative metformin use had a higher risk of developing TB than patients in the first quartile (Q1). CONCLUSIONS: Only the highest cumulative doses of metformin were protective against the development of TB among patients who were newly diagnosed with type 2 DM; lower cumulative doses of metformin did not appear to reduce the incidence of active TB infection.