Background: Oral DPP-4 inhibitors provide adjunctive therapy to insulin in hospitalized patients with blood glucose (BG) < 400 mg/dL. Results from trials comparing sitagliptin or linagliptin to basal-bolus insulin showed mixed results, with no head-to-head comparison between sitagliptin and linagliptin. The aim of this study is to provide new evidence on their comparative efficacy and safety.

Methods: This was a retrospective, single-center, observational study of hospitalized general medicine and surgery adult patients with type 2 diabetes. Patients who received at least 2 doses of linagliptin or sitagliptin between January and December 2019 were assigned to either arms. Primary endpoint was the difference in mean daily BG. Secondary endpoint was a composite of 3-point major adverse cardiac events (MACE) and heart failure (HF) hospitalization. Safety endpoints included the number of hypoglycemic events (BG < 70 mg/dL) and the appropriateness of dosing based on eGFR.

Results: A total of 100 patients were included. The mean daily BG was similar in the linagliptin arm (167.6 ± 43.5 mg/dL) compared to the sitagliptin arm (162.1 ± 37.5 mg/dL) (p=0.5) . In a subgroup analysis, patients who received linagliptin plus correctional insulin had more BG readings within the range of 140-180 mg/dL (50% vs. 36%, p=0.2) . The composite of 3-point MACE and HF hospitalization was similar between both arms. Hypoglycemia occurred more frequently with linagliptin versus sitagliptin (4 vs. 1 event, p=0.2) . Only 70% of sitagliptin orders were appropriately dosed based on eGFR.

Conclusion: There was no difference between linagliptin and sitagliptin in regards to mean daily BG and cardiovascular outcomes. However, there were numerically more BG readings within range with linagliptin. Our results favor the utilization of linagliptin in the inpatient setting considering it does not require renal dose adjustments.


S. Barakat: None. H. A. Habib: None. C. Chun: None. E. Oh: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at