Reviews/SGLT2 Inhibitors/SGLT2 inhibitors are associated with euglycaemic diabetic ketoacidosis
· SG-SF-03 · v2
Euglycaemic diabetic ketoacidosis is a form of ketoacidosis that occurs without the high blood sugar usually seen in such emergencies, which can delay its recognition. Across several independent reports from different countries and data systems, this condition is repeatedly documented in people taking SGLT2 inhibitors [10][13][17][14][15][1][3]. Laboratory and clinical studies also show how these drugs raise ketone levels even when blood glucose is normal [5][8], giving the link a coherent biological basis. The claim only asserts that such an association exists, which is a low bar, and the corpus meets it many times over. The main limitations are that most of the direct evidence cannot pin down how often it happens or prove cause, and one large surgical study found no such link [18].
Multiple independent designs document SGLT2 inhibitor-associated euglycaemic DKA: a case series in which 12 of 21 SGLT2i-associated DKA cases were euglycaemic [10], a 13-case clinical series with spontaneous-report data [13], seven SGLT2i-associated euglycaemic DKA cases in a German multicentre study [17], a cardiac-surgery nested case-control in which 13 of 15 euDKA patients were SGLT2i users [14], and a propensity-matched Veterans Affairs case-control finding SGLT2i use associated with eKA (OR 1.11; 95% CI 1.05-1.17) [15]. A cross-database disproportionality analysis found SGLT-2 inhibitors accounted for 85.8% (FAERS) and 91.7% (JADER) of all euDKA reports [3]. Controlled studies provide the mechanism: ketone and breath-acetone concentrations were higher with dapagliflozin in the absence of significant hyperglycaemia [5], and dapagliflozin increased lipid oxidation and plasma ketone concentration in type 2 diabetes [8]. The contested point: The division is not over whether the association exists but over its magnitude and causal status: [3] states its pharmacovigilance signals 'cannot establish incidence, relative risk, or causality', and [18] found no significant association between preoperative SGLT2i use and postoperative DKA (ATE 0.2%; 95% CI -1.7% to 2.2%). The quantified case-control estimate [15] is modest (OR 1.11), so the strength and setting-dependence of the link remain unresolved by this corpus.
All 23 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from SGLT2 Inhibitors, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The directly euDKA-focused evidence is heavily case-series and pharmacovigilance ([3],[10],[13],[17],[23]); [3] states verbatim these are 'hypothesis-generating signals that cannot establish incidence, relative risk, or causality,' so they confirm the association's existence but not its magnitude or causal status.
[18] found no significant association between preoperative SGLT2i use and postoperative DKA (ATE 0.2%, 95% CI -1.7% to 2.2%), showing the link is setting-dependent and not uniformly detectable; several strong sources ([19],[21],[23]) report general DKA, not euglycaemic DKA specifically.
Where it remains uncertain. Much of the direct evidence is hypothesis-generating pharmacovigilance and case-series that cannot quantify incidence or causality [3], and one large surgical cohort found no significant association [18].
Aleatory
euDKA was only 4.1% (FAERS) and 7.8% (JADER) of SGLT2i reports [3], and reports span a wide range of age, weight and treatment duration [23], so individual occurrence is highly variable.
Epistemic
The directly euDKA-focused data are disproportionality and case-series that '[cannot] establish incidence, relative risk, or causality' [3]; the magnitude of the association is not quantified for euDKA specifically.
The load-bearing papers from a corpus of 23, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Cohort of DKA admissions documenting euDKA presentations in an SGLT2i-stratified population.
“A total of 115 patients from all cohorts presented with eu DKA, characterized by lower glucose and bicarbonate levels.”
Panel auditInternal validity: medium · scope of inference: Outcomes of hospitalized DKA in T1DM and T2DM adults, stratified by prior SGLT2i useKey limitations: conditioned on DKA, no exposure denominator · retrospective · euDKA not the primary outcome
Cross-database disproportionality showing euDKA reports concentrated in the SGLT2i class.
“SGLT-2 inhibitors accounted for 85.8% (FAERS) and 91.7% (JADER) of all euDKA reports, though euDKA was only 4.1% and 7.8% of SGLT-2 inhibitor reports.”
Panel auditInternal validity: medium · scope of inference: Spontaneous-report euDKA signal for SGLT-2 inhibitors across cardiorenal and diabetes indicationsKey limitations: reporting bias, no denominator · hypothesis-generating only · cannot establish causality
Controlled demonstration of the euglycaemic-ketosis mechanism underlying euDKA.
“blood and breath ketone concentrations were significantly greater compared to usual care and in the absence of significant hyperglycemia.”
Panel auditInternal validity: medium · scope of inference: Ketone response to dapagliflozin during supervised insulin withdrawal in T1D adultsKey limitations: surrogate outcome not euDKA event · small n · artificial insulin-withdrawal setting
No DKA occurred and the ketosis-event difference (3 vs 10) was non-significant and driven largely by one participant, so direction on the euDKA claim is not determinable.
Panel auditInternal validity: low · scope of inference: Ketosis risk of adjunct dapagliflozin in adults with type 1 diabetesKey limitations: no DKA events observed · underpowered · ketosis events in a single participant
4 of 23 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
The association is confirmed across multiple independent designs and data systems, so it is Established even though incidence and causality are not quantified by this corpus.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Population incidence of euDKA on SGLT2i. No denominator-based incidence rate for euglycaemic DKA exists in the corpus; signals are hypothesis-generating [3]. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A large prospective or population cohort with laboratory-confirmed glucose and ketone values capturing euDKA events per person-year of SGLT2i exposure.
Open question: What is the absolute incidence of euglycaemic DKA among SGLT2 inhibitor users across cardiorenal indications? Why it is still open: The disproportionality and case-series data cannot establish incidence or relative risk [3].
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 23 Oct 2026 · protocol v0.19