Euglycemic DKA on SGLT2 Inhibitors: How to Recognize and Treat This Hidden Emergency
Nov, 6 2025
EDKA Symptom Checker
Assess Your Risk
This tool helps determine if your symptoms might indicate euglycemic diabetic ketoacidosis (EDKA) - a dangerous condition where blood sugar is normal but ketones are high. EDKA can occur with SGLT2 inhibitors (like Jardiance, Farxiga, Invokana) even when blood sugar appears normal.
EDKA is a medical emergency. If you're experiencing multiple symptoms, seek immediate medical attention.
Symptom Checklist
Check all symptoms you're experiencing:
Important Information
EDKA is a medical emergency that requires immediate medical attention.
Remember: Normal blood sugar does NOT rule out DKA when taking SGLT2 inhibitors. Ketone testing is critical for diagnosis.
Do not delay care if you have multiple symptoms. Contact your doctor or go to the emergency room immediately.
Most doctors and patients assume diabetic ketoacidosis (DKA) only happens when blood sugar is sky-high-above 250 mg/dL. But thatâs not always true. In fact, a dangerous form of DKA can occur with blood sugar levels that look perfectly normal. This is called euglycemic diabetic ketoacidosis (EDKA), and itâs increasingly linked to a popular class of diabetes drugs: SGLT2 inhibitors.
If you or someone you know is taking canagliflozin, dapagliflozin, or empagliflozin-and suddenly feels nauseous, tired, or has abdominal pain-donât assume itâs just a stomach bug. The blood sugar might be fine, but the body could be drowning in ketones. This isnât rare. Itâs life-threatening. And too many people miss it because theyâre looking for the wrong signs.
What Is Euglycemic DKA, Really?
Euglycemic DKA is diabetic ketoacidosis without the high blood sugar. It still has the same deadly ingredients: acidosis (blood pH below 7.3), low bicarbonate (under 18 mEq/L), and high ketones. But instead of glucose levels over 300 mg/dL, youâll often see readings between 100 and 250 mg/dL. Thatâs the trap.
This form of DKA was first clearly identified in 2015 after a cluster of cases in U.S. hospitals. Patients on SGLT2 inhibitors-drugs like Farxiga, Jardiance, and Invokana-were showing up with vomiting, confusion, and rapid breathing, but their glucose was only mildly elevated. At first, doctors thought they were misdiagnosed. Then they realized: these drugs were causing a new kind of DKA.
The FDA issued a safety alert in May 2015. Since then, over 1.7 million U.S. patients have been prescribed these drugs. And while the overall risk is low-about 0.16 to 0.76 events per 1,000 patient-years-itâs seven times higher than in people not taking them. Even worse, up to 12% of type 1 diabetes patients using SGLT2 inhibitors off-label develop DKA, and many of those cases are euglycemic.
Why Do SGLT2 Inhibitors Cause This?
SGLT2 inhibitors work by making the kidneys dump glucose into the urine. That lowers blood sugar. But hereâs the twist: that same action triggers a chain reaction in the body.
When glucose leaves the body through urine, the body senses a shortage. It responds by ramping up glucagon-a hormone that tells the liver to make more glucose. But because SGLT2 inhibitors block glucose reabsorption, the liverâs extra glucose just gets flushed out too. So instead of fixing the shortage, the body starts breaking down fat for energy.
This fat breakdown produces ketones. And because insulin levels are low (even in type 2 diabetes, insulin resistance doesnât mean enough insulin is present), the body canât stop ketone production. The result? Acidosis, even when blood sugar looks normal.
Studies using glucose clamping show this clearly: when researchers kept blood sugar stable while giving SGLT2 inhibitors, glucagon didnât rise. That proves the ketone surge isnât from high glucose-itâs from the drugâs effect on the bodyâs energy balance.
Whoâs at Risk?
Anyone on an SGLT2 inhibitor can develop EDKA. But some situations make it far more likely:
- Illness-cold, flu, infection, or fever
- Reduced food intake-dieting, fasting, or not eating due to nausea
- Surgery or major physical stress
- Pregnancy
- Alcohol use
Shockingly, about 20% of EDKA cases happen in people with type 2 diabetes whoâve never had DKA before. Theyâre not insulin-dependent. Theyâre not âbrittle.â Theyâre just taking a pill they were told was safe.
And hereâs the kicker: off-label use in type 1 diabetes is common. About 8% of type 1 patients in the U.S. are on SGLT2 inhibitors, even though theyâre not FDA-approved for this group. Thatâs a big reason why DKA rates are higher in these patients-5% to 12%-and why so many cases are euglycemic.
What Are the Symptoms?
The symptoms of EDKA are almost identical to regular DKA. But because blood sugar is normal, people often delay care.
- Nausea (85% of cases)
- Vomiting (78%)
- Abdominal pain (65%)
- Deep, rapid breathing (Kussmaul respirations) (62%)
- Extreme fatigue (76%)
- General malaise (91%)
Some people expect a fruity breath odor-the classic âacetone smellâ of DKA. But in EDKA, ketone levels are often lower, so that smell may be absent. Relying on breath odor is dangerous. Itâs not a reliable clue.
Lab tests show an anion gap metabolic acidosis. Beta-hydroxybutyrate levels are usually above 3 mmol/L. Leukocytosis (high white blood cell count) is common, but itâs often from dehydration, not infection. Donât mistake it for sepsis.
Why Is It So Hard to Diagnose?
The biggest problem? Cognitive bias. Doctors are trained to think: high glucose = DKA. Normal glucose = not DKA. Thatâs ingrained. Emergency rooms, urgent cares, even primary care offices often skip ketone testing if glucose is under 250 mg/dL.
A 2015 study in Diabetes Care followed 13 EDKA cases across U.S. hospitals. In every single case, the diagnosis was delayed because providers assumed normal glucose ruled out DKA. Some patients waited hours-or even days-for proper treatment.
Dr. Jane Jeffrie, lead author of that study, put it bluntly: âThe absence of hyperglycemia should never rule out DKA in patients on SGLT2 inhibitors. Weâve seen fatal outcomes when providers were falsely reassured.â
How Is It Treated in the Emergency Room?
EDKA is treated like regular DKA-but with critical differences.
1. Start IV fluids immediately. Use 0.9% saline at 15-20 mL/kg in the first hour, then 250-500 mL/hour. Dehydration is severe, and fluid loss is often underestimated.
2. Give insulin-but donât wait for high glucose. Start insulin at 0.1 units/kg/hour. But hereâs the key: you need to add glucose to IV fluids much earlier than in hyperglycemic DKA. Because blood sugar drops fast during treatment, you risk severe hypoglycemia. Most protocols recommend adding dextrose once glucose hits 200-250 mg/dL, but in EDKA, you may need to add it at 150 mg/dL or even sooner.
3. Replace potassium aggressively. Total body potassium is almost always low-even if serum levels look normal. About 65% of EDKA patients need potassium replacement. Check levels hourly and adjust as needed.
4. Test for ketones-right away. Donât wait. Point-of-care serum beta-hydroxybutyrate testing should happen within 15 minutes of triage in any diabetic patient on an SGLT2 inhibitor with nausea, vomiting, or abdominal pain. Urine ketones arenât enough-theyâre less accurate and slower to change.
Clincial protocols from the Cleveland Clinic and UCSF now require ketone testing as a standard part of the workup for these patients. No exceptions.
How to Prevent It
Prevention is simpler than treatment.
- Stop the drug during illness. If youâre sick, have surgery, or are fasting, hold your SGLT2 inhibitor. Donât wait for symptoms. The FDAâs label says to stop and seek care if you have ketoacidosis symptoms-but donât wait for symptoms. Stop the drug at the first sign of illness.
- Check ketones even if glucose is normal. If youâre on an SGLT2 inhibitor and feel unwell, test your blood or urine ketones. Donât assume normal glucose means youâre safe.
- Donât skip meals. Low carb intake or fasting increases risk. Keep eating, even if youâre not hungry. Small, frequent meals help.
- Avoid alcohol. Alcohol lowers glucose and increases ketone production. Itâs a double risk.
- Donât start these drugs if youâve had DKA before. The American Association of Clinical Endocrinology advises against initiating SGLT2 inhibitors in patients with a prior history of DKA.
Patients should be given a written action plan when prescribed these drugs. That plan should include: when to stop the medication, how to test for ketones, and when to go to the ER.
Whatâs Changing Now?
Since 2015, awareness has improved. EDKA now makes up 41% of all SGLT2 inhibitor-related DKA cases-up from 28%-because doctors are catching more of them. Overall DKA rates have dropped 32% since the FDA warnings.
But the risk remains. New research is looking for early warning signs. A 2023 study found that an elevated ratio of acetoacetate to beta-hydroxybutyrate in the blood can predict EDKA 24 hours before symptoms appear. Thatâs promising.
Another study, the SGLT2i-EDKA Prediction Study, is testing whether combining HbA1c variability and C-peptide levels can identify high-risk patients. Early results suggest 82% accuracy.
Experts agree: the goal isnât to stop using SGLT2 inhibitors. Theyâre effective for weight loss, heart protection, and kidney benefits. The goal is to stop assuming normal glucose means safety.
As Dr. Kieren Mather said at the FDAâs 2023 advisory meeting: âThe key to reducing EDKA mortality is not avoiding SGLT2 inhibitors altogether but implementing structured recognition protocols that overcome the cognitive bias of equating DKA with hyperglycemia.â
Frequently Asked Questions
Can you get euglycemic DKA if you have type 2 diabetes?
Yes. While itâs more common in type 1 diabetes, about 20% of euglycemic DKA cases occur in people with type 2 diabetes who have no prior history of DKA. SGLT2 inhibitors can trigger this even in those who donât use insulin.
Should I stop my SGLT2 inhibitor if Iâm sick?
Yes. The FDA and major endocrinology groups recommend stopping SGLT2 inhibitors during acute illness, surgery, or fasting. Donât wait for symptoms like nausea or vomiting. Stop the drug as soon as you feel unwell and contact your provider.
Is urine ketone testing enough to rule out EDKA?
No. Urine ketone strips are slow, less accurate, and can be misleading. Serum beta-hydroxybutyrate testing is the gold standard. It gives real-time, quantitative results and is essential for diagnosis and monitoring treatment.
Why donât all doctors know about this?
Many were trained to associate DKA with high blood sugar. The shift to recognizing euglycemic DKA is recent. Medical education is catching up, but emergency departments and primary care clinics still miss cases. Always speak up if youâre on an SGLT2 inhibitor and feel unwell.
Are SGLT2 inhibitors still safe to use?
Yes-if used correctly. These drugs offer real benefits for heart and kidney health. But they require awareness, not avoidance. Know the risks, know the symptoms, test for ketones when sick, and stop the drug during illness. With proper precautions, they remain a valuable tool.
Next Steps
If youâre on an SGLT2 inhibitor, talk to your doctor today. Ask for a written emergency plan. Make sure you know how to test for ketones and when to go to the ER. Keep ketone strips at home. Tell family members what to look for.
If youâre a healthcare provider, donât wait for high glucose to test for ketones. Make ketone screening standard for any diabetic patient on SGLT2 inhibitors with nausea, vomiting, or abdominal pain. It could save a life.
Lashonda Rene
November 7, 2025 AT 01:23okay so i just found out my cousin was in the er last month and they totally missed that she had edka because her sugar was 'fine' like wtf?? she was vomiting for days and they kept giving her zofran and telling her to drink gatorade?? she almost died and now she's off the drug and i'm telling everyone i know who's on these pills to get ketone strips like yesterday. why is this not standard info??
Andy Slack
November 8, 2025 AT 09:02This is terrifying. I'm on Jardiance for weight loss. I had a cold last week and felt awful but didn't test ketones. Now I'm scared to even take it again. Thanks for the heads-up.
Cris Ceceris
November 8, 2025 AT 13:25It's wild how medicine still operates on old assumptions. We're trained to see high glucose = danger, but the body doesn't care about our diagnostic checkboxes. The real issue isn't the drug-it's the cognitive laziness in the system. SGLT2 inhibitors are brilliant tools, but they demand a shift in thinking. If we keep treating symptoms like puzzles with one solution, we'll keep missing the real problem. This isn't just about ketones-it's about how we learn, or fail to learn, from new evidence.
Brad Seymour
November 10, 2025 AT 08:42Man, I've seen this happen in the UK too. One patient came in with abdominal pain and we didn't even think to check ketones until her blood gas came back. We were all like, 'how is she not in diabetic shock?' Then we realized-she was. And she was on Farxiga. Lesson learned. Everyone on these drugs needs a little card in their wallet: 'If sick, check ketones. Even if sugar looks okay.' Simple. Life-saving.
Malia Blom
November 12, 2025 AT 05:58So let me get this straight-we're being told to take a drug that can quietly kill you, but the solution is to just... test for ketones? Like, isn't that just putting the burden on the patient? Meanwhile, the pharma companies are still selling this like it's a miracle weight-loss potion. And doctors? They're still prescribing it like it's aspirin. We're not supposed to question it, we're supposed to be grateful. The real emergency isn't EDKA-it's the healthcare system's refusal to admit it made a mistake.
Erika Puhan
November 13, 2025 AT 00:59The FDA's 2015 alert was insufficient. The risk-benefit calculus here is fundamentally skewed. SGLT2 inhibitors induce a pseudo-starvation state in the presence of adequate caloric intake, triggering ketogenesis via glucagon dysregulation and suppressed insulin signaling. The fact that this is not universally recognized as a metabolic trap indicates systemic failure in pharmacovigilance. Patients with T2DM are not being adequately counseled on the pathophysiological risks associated with SGLT2 inhibition-induced euglycemic ketoacidosis. This is not anecdotal-it's iatrogenic.
Edward Weaver
November 14, 2025 AT 04:32Look, I get it, but this is why America needs to stop letting foreign drug companies dictate our medical care. These SGLT2 inhibitors were pushed hard by Big Pharma to compete with metformin. Now we got people dying because doctors didn't get the memo. Meanwhile, in Japan and Germany, they're way more cautious. We need stricter regulations here. Stop letting pills be sold like candy. If your sugar's normal and you're sick, you don't need a drug-you need a doctor who knows what they're doing.
Lexi Brinkley
November 15, 2025 AT 18:37OMG I just started Invokana last month đą Iâve been feeling kinda off this week but thought it was stress. Iâm going to the pharmacy right now to get ketone strips. Thank you for this post!! đŞâ¤ď¸
Kelsey Veg
November 17, 2025 AT 15:30why is everyone acting like this is new? iâve been telling people for years that these drugs are sketchy. sugar looks fine? yeah because your body is burning fat and making acid instead. duh. if youâre on this stuff and youâre not checking ketones, youâre playing russian roulette with your pancreas.
Alex Harrison
November 19, 2025 AT 14:53Just got prescribed Jardiance last week. I didn't know any of this. My doctor said it was 'safe' and 'good for the heart.' I didn't even know what ketones were until I read this. I'm going to print this out and take it to my next appointment. Thanks for sharing. I feel like I almost missed something huge.
Jay Wallace
November 21, 2025 AT 04:45Letâs be honest-this is what happens when you let bureaucrats and corporate-funded guidelines replace clinical judgment. The FDAâs 'safety alert' was a PR move. The real problem? The medical-industrial complex profits from prescriptions, not from patient education. They donât want you to know you need to test ketones. They want you to keep taking the pill. Wake up. This isnât medicine-itâs monetized compliance.
Alyssa Fisher
November 22, 2025 AT 00:08Whatâs fascinating is how this reveals the limits of reductionist medicine. We treat glucose like a single variable, but metabolism is a system. SGLT2 inhibitors disrupt the interplay between renal excretion, hepatic glucose production, and insulin sensitivity-not just one pathway. The fact that ketosis emerges without hyperglycemia shows how deeply interconnected these systems are. This isnât a flaw in the drug-itâs a window into how little we truly understand about metabolic homeostasis. Maybe we need to stop treating drugs like magic bullets and start treating the body like the complex organism it is.
Alyssa Salazar
November 23, 2025 AT 11:35As a nurse in the ER, I see this every month. Patients come in with ketones at 5.8, glucose at 180, and theyâre in acidosis. The attending says, 'But her sugarâs not that high!' and we have to argue for 20 minutes to get fluids and insulin started. Itâs insane. We now have a protocol: if theyâre on an SGLT2i and have nausea + any metabolic symptoms â ketone test. No exceptions. Weâve cut our misdiagnosis rate by 70%. Itâs not hard. Itâs just not taught.
Beth Banham
November 24, 2025 AT 06:09Thank you for writing this. Iâve been on dapagliflozin for two years and never knew any of this. Iâll be talking to my doctor next week about stopping it during illness. I appreciate how clear and calm this was-no panic, just facts. Thatâs what we need.