OBJECTIVES : No drugs have yet been FDA- approved to treat non- alcoholic fatty liver disease (NAFLD). Several clinical trials have demonstrated beneficial effects of antidiabetic drugs (i.e., pioglitazone, liraglutide, and exenatide) in treating NAFLD. However, whether diagnosis of NAFLD influences clinicians’ choice of antidiabetic agents is unknown. This study assessed the initiation timing of glucagon-like peptide-1 receptor agonists (GLP-1RAs), dipeptidyl peptidase‐4 inhibitors (DPP-4is), and thiazolidinediones (TZDs) relative to a NAFLD diagnosis.

METHODS : We conducted a retrospective cohort study from 2012- 2019 using MarketScan® Explorys® Claims-EMR Data Set. Adult patients with diagnosis of NAFLD who newly initiated GLP- 1RAs, DPP- 4is, or TZDs were selected from the database. The index date was defined as the initiation date of an antidiabetic agent. NAFLD was identified using the earliest date of diagnosis 365 days before and after the index date. We described the raw frequency distribution of antidiabetic drugs initiation relative to a NAFLD diagnosis.

RESULTS : Overall, 4,640 patients with diagnosis of NAFLD contributed 5,180 initiation episodes (n_{GLP- 1RAs} = 2,243, n_{DPP- 4is} = 2,587, n_TZDs = 350). Approximately 60% of the initiations were subsequent to the diagnosis of NAFLD. Among GLP1- RAs, DPP-4is, and TZDs, the distribution of dispensed products was similar before and after the NAFLD diagnosis, with a slight decline in sitagliptin (40.3% prior to NAFLD vs 36.9% after NAFLD) and increase in dulaglutide (10.2% prior to NAFLD vs 12.5% after NAFLD) and linagliptin (7.0% prior to NAFLD vs 8.3% after NAFLD).

CONCLUSIONS : The preliminary findings from this study suggest that patients are more likely to initiate these GLP-1RAs, DPP-4is, and TZDs after a NAFLD diagnosis. Further analysis will aim to establish the direct effect of a NAFLD diagnosis on the initiation of these drugs.