RESUMEN
AIMS: We aim to determine the association between prediabetes and diabetes with latent TB using National Health and Nutrition Examination Survey data. METHODS: We performed a cross-sectional analysis of 2011-2012 National Health and Nutrition Examination Survey data. Participants ≥20â¯years were eligible. Diabetes was defined by glycated hemoglobin (HbA1c) as no diabetes (≤5.6% [38â¯mmol/mol]), prediabetes (5.7-6.4% [39-46â¯mmol/mol]), and diabetes (≥6.5% [48â¯mmol/mol]) combined with self-reported diabetes. Latent TB infection was defined by the QuantiFERON®-TB Gold In Tube (QFT-GIT) test. Adjusted odds ratios (aOR) of latent TB infection by diabetes status were calculated using logistic regression and accounted for the stratified probability sample. RESULTS: Diabetes and QFT-GIT measurements were available for 4958 (89.2%) included participants. Prevalence of diabetes was 11.4% (95%CI 9.8-13.0%) and 22.1% (95%CI 20.5-23.8%) had prediabetes. Prevalence of latent TB infection was 5.9% (95%CI 4.9-7.0%). After adjusting for age, sex, smoking status, history of active TB, and foreign born status, the odds of latent TB infection were greater among adults with diabetes (aOR 1.90, 95%CI 1.15-3.14) compared to those without diabetes. The odds of latent TB in adults with prediabetes (aOR 1.15, 95%CI 0.90-1.47) was similar to those without diabetes. CONCLUSIONS: Diabetes is associated with latent TB infection among adults in the United States, even after adjusting for confounding factors. Given diabetes increases the risk of active TB, patients with co-prevalent diabetes and latent TB may be targeted for latent TB treatment.