Back

Increased ketoacidosis from co administration of glucose-lowering drugs and Metformin: a post-marketing report study

Abagyan, R.; Richter, L.

2025-10-19 endocrinology
10.1101/2025.10.16.25338182 medRxiv
Show abstract

Sodium-glucose co-transporter 2 inhibitors (SGLT2i) have become widely used not only for glycemic control but also for their protective cardiovascular effects. However, concerns about adverse events including diabetic ketoacidosis and kidney damage, have gained attention, since these agents are used both as a monotherapy and in combination with metformin. In this study, we analyzed 170,000 adverse event reports submitted to the FDA Adverse Event Reporting System for diabetic ketoacidosis (DKA), kidney injury, and lethal outcomes. Reporting odds ratios (RORs) and 95% confidence intervals (CI) revealed substantial variation in adverse event profiles across the six treatments under study. Canagliflozin shows the highest numbers on kidney damage (14.6%) of all SGLT2i monotherapies. Combination with metformin reduces it slightly (14.3%), but results in significantly elevated ketoacidosis (from 15.4% to 22.3%). Empagliflozin adverse effect profile is different, the metformin combination does not change the reported risk of kidney damage, but dramatically increases ketoacidosis from 14.7% to 27.9%, the highest number of all SGLT2is and their combinations. Notably, the DKA risk of each the three combinations exceeded the same parameter for a corresponding SGLT2i monotherapy of metformin alone. Metformin alone, on the other, is not associated with ketoacidosis (<1.1%), but has a high reported risk of death (12.7%) and kidney damage (18.7%), while adding any SGLT2i to the regimen reduces this kidney damage level dramatically. These observations call for individualized risk-benefit assessment of a safe and effective antidiabetic treatment regimen, particularly in patients with ketoacidosis-related comorbidities.

Matching journals

The top 9 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.