GRADE Trial Reveals Which Diabetes Drug Cuts Hospitalizations Most

Hospitalization is one of the most burdensome and costly outcomes in type 2 diabetes management, yet randomized head-to-head data comparing second-line agents on this endpoint are scarce. The GRADE (Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness) study provides a rare opportunity to examine this question rigorously.

GRADE enrolled 5,047 adults with type 2 diabetes inadequately controlled on metformin and randomized them to one of four second-line agents: insulin glargine, glimepiride, liraglutide, or sitagliptin. Participants were followed for a median of approximately five years across 36 U.S. clinical sites. The current analysis examined all-cause hospitalizations as well as hospitalizations categorized by cause, including cardiovascular, hypoglycemia, gastrointestinal, and other categories.

The headline finding is that liraglutide was associated with significantly fewer total hospitalizations compared to sitagliptin and glimepiride, with a numerical reduction versus glargine as well. Glimepiride showed the highest hospitalization burden, partly driven by hypoglycemia-related events—an expected finding given sulfonylurea pharmacology. Sitagliptin also fared worse than liraglutide on total hospitalizations. Cardiovascular hospitalizations were numerically lower in the liraglutide arm, consistent with its known cardiovascular benefits observed in dedicated CVOT trials. Hypoglycemia-driven hospitalizations were notably elevated in the glimepiride group.

These results carry meaningful implications for clinical practice. While glycemic efficacy was relatively similar across treatment arms in the main GRADE publication, the hospitalization data suggest that liraglutide's benefits extend well beyond blood sugar lowering. Choosing a GLP-1 receptor agonist as second-line therapy may reduce downstream healthcare utilization, which matters both for patient quality of life and health system costs.

Important caveats apply. Hospitalization was not the primary prespecified endpoint of GRADE, so these analyses are secondary and hypothesis-generating. Differences in drug tolerability (e.g., GI side effects of liraglutide) could influence both treatment adherence and hospitalization patterns in complex ways. The study population was predominantly established metformin users with relatively early-stage diabetes, limiting generalizability to more advanced disease. Nonetheless, as one of the largest and longest head-to-head RCTs of second-line diabetes therapies, GRADE provides uniquely credible real-world-relevant evidence on this hospitalization question.