Ketoacidosis reports in non-diabetic GLP-1 users are climbing sharply
A five-year FAERS analysis found both semaglutide and tirzepatide carry disproportionate ketoacidosis signals in people without diabetes, with roughly three-quarters of reported cases requiring hospitalization.
Euglycemic ketoacidosis — a dangerous buildup of blood acids that occurs without the very high blood sugar that typically accompanies diabetic ketoacidosis — has emerged as a recognized, if uncommon, complication of GLP-1 receptor agonist use, including in people who don't have diabetes. A study published this month in Cureus is the first to directly compare ketoacidosis reporting signals between semaglutide and tirzepatide specifically in non-diabetic patients [s1].
The design
Researchers conducted a disproportionality analysis of the FDA Adverse Event Reporting System (FAERS) from January 2021 through December 2025 [s1]. After deduplication, they filtered for reports naming semaglutide or tirzepatide as the primary suspect medication, then identified non-diabetic patients by excluding reports listing diabetes-related indications, and identified ketoacidosis cases using standardized MedDRA terminology [s1]. Reporting odds ratios (RORs) and proportional reporting ratios (PRRs) with 95% confidence intervals were calculated for each drug [s1].
What it found
After deduplication, 7,349,591 unique FAERS reports were analyzed, of which 48,082 were semaglutide reports and 98,295 were tirzepatide reports in non-diabetic patients [s1]. Ketoacidosis was identified in 261 semaglutide cases and 209 tirzepatide cases [s1]. Semaglutide showed a stronger disproportionality signal (ROR 3.15, 95% CI 2.78–3.56; PRR 3.14, 95% CI 2.78–3.54) than tirzepatide (ROR 1.22, 95% CI 1.06–1.39; PRR 1.21, 95% CI 1.06–1.39), though both were statistically significant [s1]. Hospitalization was required in 74.3% of semaglutide-associated ketoacidosis cases and 71.3% of tirzepatide-associated cases [s1]. Tirzepatide-associated ketoacidosis reports showed a marked upward trend over the study period, rising from just 1 to 2 cases per quarter in 2022 to 28 to 34 cases per quarter by 2025 [s1].
Reading the trend line, not just the totals
The rising quarterly report count for tirzepatide is the detail worth focusing on alongside the headline disproportionality numbers, because it likely reflects at least two things happening simultaneously: tirzepatide's rapidly expanding use for obesity since its approval, and — separately — growing clinical and public awareness of ketoacidosis as a possible GLP-1-related complication, which tends to increase reporting rates independent of any change in true underlying risk. The study doesn't disentangle how much of the rising trend reflects genuinely more cases occurring, versus more cases being recognized and reported as awareness has grown; both are plausible contributors, and pharmacovigilance databases can't distinguish between them.
Why semaglutide's higher signal doesn't necessarily mean higher true risk
Semaglutide's disproportionality signal (ROR 3.15) is notably higher than tirzepatide's (ROR 1.22), but the case counts were actually similar in absolute terms — 261 versus 209 — while tirzepatide's non-diabetic report volume (98,295) was roughly double semaglutide's (48,082) in this dataset. That means, even setting aside the disproportionality statistics, the raw rate of reported ketoacidosis per non-diabetic report was in fact higher for semaglutide (261/48,082) than tirzepatide (209/98,295) in this specific sample — consistent with, not contradicting, the disproportionality finding, though neither figure establishes a true incidence rate given the well-documented underreporting inherent to FAERS.
What this doesn't establish
This is a spontaneous-reporting pharmacovigilance analysis, not a controlled study — it cannot establish true incidence of ketoacidosis in non-diabetic patients taking either drug, only that reports of it occur more often than statistical background expectation would predict. It cannot rule out reporting bias, and the "non-diabetic" classification relies on excluding reports with a diabetes-related indication listed, which may imperfectly capture true diabetes status if that information was incompletely reported.
What the authors recommend
The study's authors describe the findings as highlighting "an emerging safety concern requiring clinical vigilance" given the high hospitalization rates and the rapidly increasing tirzepatide report volume specifically [s1] — a call for awareness among prescribers and patients using these drugs off-label for obesity without diabetes, not a claim that either drug is unsafe for its approved uses.
What to watch
Whether prospective studies establish a true incidence rate for euglycemic ketoacidosis in non-diabetic GLP-1 users, and whether regulatory labeling is updated to reflect this signal more explicitly. This article describes drug safety data; it is not medical advice.
Sources
Sources
More on
Weight loss modifies some diseases and is untested in most of the ones it is blamed for
A Lancet review triangulates the evidence linking adiposity to multisystem disease and finds robust trial support in a short list — while an expert panel argues the drugs must be lifelong.
Semaglutide users had 51% lower suicidal ideation risk than older GLP-1 users
Tirzepatide and semaglutide showed comparable psychiatric risk profiles against each other over two years. Against earlier GLP-1 drugs, semaglutide came out ahead on all three measures.
Bariatric surgery still beats GLP-1 drugs on weight loss, 44,000 patients show
Gastric bypass patients lost roughly 28% of body weight at three years, against 7-11% for tirzepatide or semaglutide — even among patients who stayed on the drugs continuously for at least a year.
In 160,000 nondiabetic adults, GLP-1 users had 41% lower obesity-linked cancer risk
The association held across most demographic subgroups, but not among Black patients, a gap the study doesn't explain. Two-year follow-up limits what it can say about longer-term risk.