Diabetic ketoacidosis (DKA) remains one of the most serious acute complications of diabetes, traditionally associated with type 1 diabetes but increasingly seen in type 2 diabetes under certain conditions. Over the past decade, large clinical trials and post-marketing surveillance have revealed a concerning link between DKA and the side effects of specific diabetes medications, most notably sodium-glucose cotransporter 2 (SGLT2) inhibitors. Understanding this connection is critical for clinicians, patients, and caregivers to prevent morbidity and mortality. This article provides a comprehensive, evidence-based exploration of the relationship between DKA and medication side effects, covering pathophysiology, specific drug risks, clinical studies, prevention strategies, and management protocols.

Understanding DKA in Depth

Diabetic ketoacidosis arises from an absolute or relative deficiency of insulin combined with elevated counter-regulatory hormones including glucagon, catecholamines, cortisol, and growth hormone. This hormonal imbalance forces the body to switch from glucose to fat as its primary energy source. Free fatty acids are oxidized in the liver to produce ketone bodies—acetoacetate, beta-hydroxybutyrate, and acetone. When ketone production exceeds the body's ability to utilize or excrete them, metabolic acidosis develops. The classical diagnostic triad includes hyperglycemia with blood glucose greater than 250 mg/dL, ketonemia or ketonuria, and metabolic acidosis with pH below 7.3 and serum bicarbonate below 18 mEq/L.

Symptoms progress from polyuria, polydipsia, and weight loss to nausea, vomiting, abdominal pain, rapid deep breathing known as Kussmaul respiration, fruity breath odor, and altered mental status. DKA can be life-threatening if not recognized and treated promptly with intravenous fluids, insulin, and electrolyte replacement. While historically most common in type 1 diabetes, DKA now occurs in type 2 diabetes, especially during severe illness, surgery, or with the use of certain medications. The biochemical drivers of DKA involve not only insulin deficiency but also the unopposed action of glucagon, which stimulates hepatic gluconeogenesis and ketogenesis. Understanding this dual hormonal defect is key to appreciating how certain medications can tip the balance toward DKA even when glucose levels are not markedly elevated.

Euglycemic DKA: A Special Consideration

One of the most clinically relevant developments is the recognition of euglycemic DKA (euDKA), where blood glucose levels are near normal, below 250 mg/dL, or even below 200 mg/dL. This phenomenon is strongly associated with SGLT2 inhibitors. Patients may not present with the expected hyperglycemia, leading to delayed diagnosis. euDKA accounts for a significant proportion of SGLT2-inhibitor-associated DKA cases and requires a high index of suspicion. Clinicians must measure serum ketones or beta-hydroxybutyrate in any patient on these medications who presents with metabolic acidosis, regardless of glucose level. The pathophysiology of euDKA involves the renal threshold for glucose being lowered by SGLT2 inhibitors, so glucose spills into urine even at normal blood levels. Meanwhile, ketogenesis proceeds unchecked because of relative insulin deficiency and glucagon excess. The absence of marked hyperglycemia can mislead clinicians into thinking DKA is not present, which delays appropriate treatment and increases the risk of adverse outcomes.

Diabetes Medications Associated with DKA Risk

Several classes of diabetes medications carry a recognized risk of precipitating DKA. The most prominent are SGLT2 inhibitors, but other drugs and treatment patterns also contribute. Understanding the mechanisms by which each drug class can provoke DKA allows clinicians to stratify risk and implement preventive measures.

SGLT2 Inhibitors: Mechanism and Side Effects

SGLT2 inhibitors, including canagliflozin, dapagliflozin, empagliflozin, and ertugliflozin, lower blood glucose by blocking glucose reabsorption in the proximal renal tubule, promoting glycosuria. The resulting osmotic diuresis can lead to volume depletion and electrolyte disturbances. More importantly, these drugs alter the body's fuel metabolism in ways that increase ketogenesis. By lowering plasma glucose and insulin levels while increasing glucagon, SGLT2 inhibitors shift the liver toward fatty acid oxidation and ketone production. This effect is particularly pronounced during periods of reduced carbohydrate intake, illness, surgery, or alcohol use. The mechanism involves inhibition of SGLT2 in the kidney but also potential direct effects on alpha cells in the pancreas, leading to increased glucagon secretion and reduced insulin secretion from beta cells.

Side Effects of SGLT2 Inhibitors That Contribute to DKA

The side effects of SGLT2 inhibitors that contribute to DKA include:

  • Dehydration – Glycosuria induces osmotic diuresis, which can lead to intravascular volume contraction and impaired renal clearance of ketones. Volume depletion also stimulates counter-regulatory hormones that further promote ketogenesis.
  • Electrolyte imbalances – Sodium, potassium, and magnesium losses can disrupt acid-base homeostasis and impair the kidney's ability to excrete acid loads.
  • Increased ketone production – Directly via hormonal changes that favor hepatic ketogenesis, including increased glucagon-to-insulin ratio.
  • Genitourinary infections – Though not directly causing DKA, infections can stress the body and precipitate DKA through the release of stress hormones and cytokines.
  • Reduced glucagon suppression – Some studies suggest SGLT2 inhibitors blunt the normal suppression of glucagon by glucose, perpetuating ketosis even when glucose levels are normal.
  • Increased renal ketone reabsorption – There is evidence that SGLT2 inhibitors may increase the renal threshold for ketone excretion, leading to higher circulating ketone levels.

The U.S. Food and Drug Administration (FDA) has issued multiple safety communications regarding DKA with SGLT2 inhibitors (FDA Drug Safety Communication). Post-marketing data revealed cases of serious DKA, including euDKA, in patients with both type 1 and type 2 diabetes. The FDA now requires warnings on labels and advises that these drugs should not be used for type 1 diabetes due to the risk of DKA. Recent meta-analyses confirm an approximately twofold increased risk of DKA with SGLT2 inhibitors compared to placebo or other glucose-lowering agents (JAMA study).

Other Diabetes Medications and DKA

While SGLT2 inhibitors are the most discussed, other drug classes and treatment patterns can contribute to DKA. A comprehensive understanding of medication-associated DKA requires awareness of these less common but still important triggers:

  • Insulin omission or insufficient dosing – The most common cause of recurrent DKA in type 1 diabetes. Any medication that reduces insulin secretion or sensitivity may indirectly increase risk if insulin therapy is not adjusted appropriately. This includes medications that promote weight loss and reduce insulin requirements, which can lead to inadvertent insulin underdosing.
  • Off-label use of GLP-1 receptor agonists – Some reports suggest a potential risk, though it is much lower than with SGLT2 inhibitors. The mechanism may involve reduced insulin secretion and appetite suppression leading to ketosis, especially when combined with reduced carbohydrate intake.
  • Thiazolidinediones (TZDs) – Rarely associated with DKA, usually in the context of insulin deficiency. The mechanism is not well understood but may involve fluid retention and dilutional acidosis in some cases.
  • Metformin – Lactic acidosis is the primary concern; DKA is not typically a metformin complication, but it can coexist with metformin-associated acidosis in renal impairment, creating a mixed acid-base disorder.
  • Dipeptidyl peptidase-4 (DPP-4) inhibitors – No established link to DKA, though rare case reports exist, likely representing coincidental occurrences.
  • Insulin pump therapy – While not a medication itself, insulin pump malfunction or infusion site issues can rapidly lead to DKA, especially in type 1 diabetes, because of the rapid loss of basal insulin delivery.

Off-label use of certain therapies, such as insulin sensitizers in type 1 diabetes, has been reported to precipitate DKA when insulin doses are reduced too aggressively. The combination of multiple agents that promote ketosis can create a dangerous synergistic effect.

Clinical Evidence and Studies Linking DKA to Medication Side Effects

Multiple large observational studies and randomized controlled trials have quantified the risk of DKA with SGLT2 inhibitors. A 2020 meta-analysis of 38 trials including over 100,000 patients found a hazard ratio of 2.2 for DKA with SGLT2 inhibitors compared to non-SGLT2 inhibitor treatments. The risk was highest in patients with type 1 diabetes, with a hazard ratio of 4.3, but still significantly elevated in type 2 diabetes at 1.8. The absolute risk remains low, approximately 0.1 to 0.5 percent per year, but given the large number of patients using these drugs, the public health impact is substantial. When extrapolated to the millions of patients taking SGLT2 inhibitors worldwide, even a small absolute risk translates into thousands of preventable DKA episodes annually.

The FDA's Adverse Event Reporting System (FAERS) database has identified hundreds of cases of DKA, including deaths, in patients taking SGLT2 inhibitors. Many cases occurred in situations that would have been considered low-risk for hyperglycemic DKA, such as mild illness, reduced carbohydrate intake, or after surgery. This pattern underscores the importance of recognizing predisposing conditions and educating patients on sick-day rules. A notable finding from FAERS data is that a significant proportion of cases occurred within the first few weeks of therapy, suggesting that the risk is highest during the initiation phase when metabolic adaptations are occurring.

Studies have also examined the mechanism behind euDKA. Research suggests that SGLT2 inhibitors lower the renal threshold for glucose and stimulate glucagon secretion while decreasing insulin secretion, creating a ketotic state even at normal blood glucose levels. The risk is amplified when patients follow very low-carbohydrate or ketogenic diets, as these diets independently increase ketone production (Diabetes Care article). Additionally, alcohol consumption can worsen the risk because alcohol metabolism produces acetate and shifts the liver toward ketogenesis.

Patient Populations at Highest Risk

Not all patients on SGLT2 inhibitors develop DKA. Identifying those at heightened risk allows targeted prevention and monitoring. The following populations require special attention:

  • Patients with type 1 diabetes, especially if SGLT2 inhibitors are used off-label without appropriate insulin dose adjustments
  • Those with reduced oral intake or who are on low-carbohydrate or ketogenic diets
  • Patients undergoing surgery or procedures requiring prolonged fasting, including colonoscopy preparations
  • Individuals with acute illness such as infection, gastroenteritis, or myocardial infarction
  • Patients with history of alcohol use disorder or substance abuse, particularly with binge drinking
  • Those with impaired renal function, specifically eGFR below 60 mL/min/1.73 m²
  • Patients taking concomitant medications that affect fluid balance or insulin secretion, including diuretics and corticosteroids
  • Older adults who may have reduced thirst sensation and are more susceptible to dehydration
  • Patients with a history of recurrent DKA or poor diabetes self-management skills

Preventive Measures and Recommendations

Healthcare providers play a central role in preventing medication-induced DKA. The cornerstone is patient and caregiver education combined with proactive monitoring and clear protocols for high-risk situations. Prevention must be multifaceted, addressing medication management, patient behavior, and system-level factors.

Patient Education

  • Teach recognition of DKA symptoms: nausea, vomiting, abdominal pain, lethargy, rapid breathing, fruity breath, unusual thirst or fatigue. Emphasize that euDKA may present without extreme thirst or frequent urination.
  • Instruct patients to check blood glucose and urine or blood ketones during illness or when symptoms appear. Provide written sick-day action plans.
  • Emphasize the risk of euDKA: normal blood glucose does not rule out DKA. If symptoms occur, ketones must be measured using blood beta-hydroxybutyrate meters rather than urine strips, which may be less reliable.
  • Advise against very low-carbohydrate or ketogenic diets while on SGLT2 inhibitors. If patients choose to follow such diets, discuss the increased risk and consider alternative medications.
  • Provide clear sick-day guidelines: stay hydrated with water or sugar-free fluids, continue medications unless instructed otherwise, and contact a healthcare provider if vomiting or symptoms persist for more than six hours.
  • Educate about alcohol consumption: advise moderation and caution about drinking on an empty stomach, which can precipitate ketosis.

Clinical Monitoring

  • Measure baseline renal function, electrolytes, and ketone levels before starting SGLT2 inhibitors. Repeat these measurements after dose adjustments or if clinical status changes.
  • For patients with type 2 diabetes, reassess the need for SGLT2 inhibitors periodically, especially if they develop risk factors such as declining renal function or dietary changes.
  • Advise withholding SGLT2 inhibitors at least 24 to 48 hours before elective surgery, procedures requiring prolonged fasting, or during acute illness. The exact timing depends on the drug's half-life, with longer-acting agents requiring earlier discontinuation.
  • Consider checking beta-hydroxybutyrate levels in any patient on SGLT2 inhibitors admitted with metabolic acidosis, regardless of glucose level. This should be part of the initial laboratory workup.
  • Use caution when combining SGLT2 inhibitors with other drugs that may increase ketone production, such as insulin, alcohol, or corticosteroids.
  • Implement a system for flagging high-risk patients in electronic health records, with alerts for clinicians when prescribing SGLT2 inhibitors to patients with type 1 diabetes or other risk factors.

Medication Adjustments

  • In type 1 diabetes, SGLT2 inhibitors are not approved and should generally be avoided due to high DKA risk. If used off-label under specialist care, strict monitoring and insulin dose adjustments are mandatory, with explicit documentation of informed consent.
  • During illness, patients may need to temporarily increase insulin doses and ensure adequate carbohydrate intake. SGLT2 inhibitors should be held until the patient is eating and drinking normally.
  • If DKA develops, the offending medication should be discontinued until the episode resolves and the cause is identified. Restarting the medication should be considered only after a careful risk-benefit analysis and with stricter monitoring protocols in place.
  • Consider alternative glucose-lowering agents with lower DKA risk for patients with multiple risk factors, such as DPP-4 inhibitors or GLP-1 receptor agonists.

Managing DKA When It Occurs

Management of medication-associated DKA follows standard DKA protocols with special considerations for the underlying drug effects. Prompt recognition and appropriate treatment are essential to prevent progression to severe acidosis, coma, or death.

  1. Fluid resuscitation – Correct volume depletion while monitoring for fluid overload, especially in patients with renal impairment or heart failure. Use isotonic saline initially, then switch to half-normal saline once hemodynamic stability is achieved. The goal is to restore perfusion and enhance renal ketone clearance.
  2. Insulin therapy – Use intravenous insulin to suppress ketogenesis and promote glucose utilization. Patients on SGLT2 inhibitors may require higher insulin doses because of ongoing ketone production and relative insulin resistance induced by acidosis. Start with a bolus of 0.1 units per kilogram followed by continuous infusion at 0.1 units per kilogram per hour.
  3. Electrolyte replacement – Pay close attention to potassium, phosphate, and magnesium. SGLT2 inhibitors can cause electrolyte wasting, and acidosis shifts potassium extracellularly, so levels can drop rapidly once insulin therapy begins. Monitor potassium hourly and replace as needed to maintain levels above 4.0 mEq/L.
  4. Ketone monitoring – Beta-hydroxybutyrate levels should be checked every two to four hours until resolution. Do not rely solely on urine ketones, as they may be discordant and can remain positive even after blood ketones have cleared. The goal is a beta-hydroxybutyrate level below 0.6 mmol/L.
  5. Address precipitating factors – Discontinue the suspected offending medication, treat underlying infection or illness, and ensure adequate nutritional intake. Obtain blood cultures, chest X-ray, and urinalysis if infection is suspected.
  6. Transition to subcutaneous insulin – Once DKA resolves with pH above 7.3 and bicarbonate above 18 mEq/L, transition to subcutaneous insulin and consider whether to restart the SGLT2 inhibitor. In many cases, the risk-benefit ratio favors switching to a different class of glucose-lowering medication.

The American Diabetes Association (ADA) Standards of Medical Care in Diabetes provides detailed guidance on DKA management (ADA Standards of Care). Additional resources are available from the Endocrine Society (Endocrine Society Guidelines).

The Role of Off-Label and Emerging Therapies

Off-label use of SGLT2 inhibitors in type 1 diabetes remains a contentious area. Despite the high risk, some patients and clinicians pursue this combination for glycemic control and weight loss. The FDA has not approved any SGLT2 inhibitor for type 1 diabetes, and prescribing requires careful informed consent, frequent ketone monitoring, and specialized management. The risk of DKA in this population is substantially higher than in type 2 diabetes, and even with meticulous monitoring, breakthrough DKA can occur during intercurrent illness or dietary indiscretion.

The same caution applies to newer therapies like dual SGLT1/SGLT2 inhibitors such as sotagliflozin, which may further increase ketone production through additional gastrointestinal effects. Clinical trials of these agents in type 1 diabetes have shown higher rates of DKA compared to placebo, leading to regulatory concerns.

Additionally, the growing popularity of ketogenic diets among people with diabetes creates a synergistic risk when combined with SGLT2 inhibitors. Clinicians should ask about dietary practices and counsel accordingly. Patients who are determined to follow a very low-carbohydrate diet should be offered alternative glucose-lowering medications that do not independently promote ketosis.

Conclusion

The link between DKA and side effects of specific diabetes medications, particularly SGLT2 inhibitors, is a well-established but preventable complication. The key points for clinicians are: recognize the risk of euglycemic DKA, screen for ketones in any patient on these medications presenting with unexplained metabolic acidosis, educate patients about warning signs and sick-day rules, and follow established safety guidelines for perioperative or illness management. With appropriate vigilance, the benefits of SGLT2 inhibitors, including cardiovascular and renal protection, can be preserved while minimizing the risk of DKA. Ongoing research continues to refine our understanding of the mechanisms and may lead to safer use in the future through the development of agents with more selective effects on ketone metabolism or through personalized risk stratification approaches. Until then, a combination of clinical awareness, patient education, and proactive monitoring remains the best defense against this serious but avoidable complication.