2017 HSR&D/QUERI National Conference

1041 — Prior Antibiotic Exposure and Risk of Type 2 Diabetes and Weight Gain Among Veterans

Lead/Presenter: Mark Schwartz
All Authors: Davis J (VA NY Harbor/NYU School of Medicine) Liu M (NYU School of Medicine) Alemi F (VA NY Harbor/George Mason University) Avramovic S (VA NY Harbor/George Mason University) Levy E (NYU School of Medicine) Hayes R (NYU School of Medicine) Sherman S (VA NY Harbor/NYU School of Medicine) Schwartz MD (VA NY Harbor/NYU School of Medicine)

Evidence shows the human intestinal microbiome influences host metabolism through mechanisms related to obesity and insulin resistance. United States Veterans are at excess risk for diabetes. Our objective is to determine the impact of prior antibiotic exposure for risk of type 2 diabetes in this at-risk population.

At the primary care practices of the VA New York Harbor (VA NYHHS) during 2004-2014, we identified 14,361 diabetes-free Veterans who received >=2 or more hemoglobin A1c tests (HbA1c) test results < 6.5% ( < 48 mmol/mol) and characterized these individuals with respect to antibiotic use and selected risk factors. We followed these individuals through 2014 for incident diabetes. Cox proportional hazard regression was used to relate demographic and anthropometric measures, antibiotic medications and comorbid cardiovascular conditions to incident diabetes (Hazard Ratio [HR] and 95% confidence intervals). Models incorporated time varying repeated measures of antibiotic medication to analyze associations by antibiotic class and total cumulative number of courses prescribed during follow-up.

Among 14,361 Veterans, 9,922 (69.1%) were prescribed any antibiotic medication during the course of follow-up. 1,413 (9.8%) individuals developed diabetes. Among this group, 1,033 (73.1%) were prescribed any antibiotic. Increased risk of type 2 diabetes was associated with any exposure HR 1.14 (1.00-1.30), narrow spectrum 1.17 (1.03-1.34) and bacteriostatic antibiotics 1.14 (1.02-1.28) compared to no exposure. Time varying analysis of the total number of cumulative courses prescribed showed increased diabetes risk for azithromycin HR 1.11(1.06-1.16), cephalosporin 1.17 (1.04-1.32), macrolide 1.08 (1.04-1.14) and penicillin 1.05 (1.02-1.07). No association was observed for antivirals or sulfonamides.

This large retrospective cohort study of United States Veterans demonstrated an increased risk of type 2 diabetes among Veterans with prior exposure to any antibiotic, ever use of 6 classes of antibiotics (antifungal, azithromycin, macrolide, penicillin, quinolone and tetracycline) as well as any narrow spectrum and bacteriostatic antibiotic medications. Increased adjusted risks were also shown for cumulative exposures to cephalosporin, macrolide and penicillin antibiotics.

The association between antibiotics and diabetes risk is thought to involve mechanisms of cumulative perturbations of the microbiome and its subsequent effects on inflammation and metabolic functions. Future studies should investigate changes in the human gut microbiome due to antibiotic exposures and how this might translate into metabolic dysfunction. This study provides further rationale to limit antibiotic exposure in veterans to indications where benefits outweigh short-term and potential long-term costs such as diabetes..