Non-Diabetic Ketoacidosis: Causes & Symptoms

Ketoacidosis without diabetes is a real and sometimes life-threatening condition that catches both patients and clinicians off guard. Most people associate ketoacidosis with type 1 diabetes, but heavy alcohol use, prolonged fasting, very low-carbohydrate diets, breastfeeding, and certain medications can all push a non-diabetic person into the same dangerous acid buildup. Because blood sugar often stays normal in these cases, the condition can be harder to recognize than its diabetic counterpart.

How Ketoacidosis Develops Without Diabetes

The body normally runs on glucose. When glucose becomes unavailable or the hormonal signals change to mimic a glucose shortage, the liver starts breaking down fat for fuel. That fat breakdown produces molecules called ketone bodies, which many tissues, including the brain, can use for energy. In small amounts, ketones are harmless. The problem starts when they accumulate faster than the body can use or clear them, turning the blood acidic.

The biochemical trigger is a combination of low insulin and elevated glucagon. Both drive the body to release stored fat and convert it into ketones at a rate that overwhelms the blood’s natural buffering system, producing a high anion gap metabolic acidosis.¹ In diabetic ketoacidosis, the missing ingredient is insulin itself, because the pancreas can’t make enough. In non-diabetic ketoacidosis, the pancreas works fine under normal conditions, but something else has disrupted the balance: starvation, alcohol, extreme carbohydrate restriction, or a drug that alters glucose handling.

Alcoholic Ketoacidosis

Alcoholic ketoacidosis, often called AKA, is probably the most well-known form of non-diabetic ketoacidosis. It typically develops in people with chronic heavy drinking who suddenly stop eating, often because of nausea and vomiting from a binge. The combination of alcohol’s metabolic effects, dehydration, and absent food intake creates a perfect storm: insulin drops, counterregulatory hormones rise, and the liver churns out ketones.

A key distinguishing feature compared to diabetic ketoacidosis is the blood sugar level. In one study comparing the two conditions, patients with diabetic ketoacidosis presented with an average plasma glucose around 32 mmol/L, while those with AKA averaged just 6.6 mmol/L, essentially near normal.² That near-normal sugar level is exactly what makes AKA tricky. A clinician running a quick glucose check might not suspect ketoacidosis at all. Lab work typically reveals elevated ketones (specifically beta-hydroxybutyrate), signs of dehydration, and acidosis.³

AKA can also be difficult to distinguish from toxic alcohol poisoning, such as methanol or ethylene glycol ingestion, since both produce an anion gap acidosis. A retrospective analysis found that higher measured ethanol levels shifted the odds toward AKA rather than toxic alcohol ingestion, but there was limited ability of common clinical factors alone to reliably tell the two apart.⁴ That ambiguity matters, because the treatments are very different.

The mortality risk from AKA is not trivial. A retrospective analysis of clinical patients in Korea found a mortality rate as high as roughly 35%.⁵ That figure likely reflects the broader medical fragility of the population affected, including liver disease, malnutrition, and infections, but it underscores that AKA is a serious emergency, not just a metabolic footnote.

Starvation Ketoacidosis

Any extended period without eating can theoretically push the body into ketoacidosis, though it usually takes more than just skipping a meal. Healthy adults have enough glycogen stores to last roughly a day before the body shifts heavily to fat metabolism. Most people who fast for 24 to 48 hours develop mild ketosis, a modest rise in ketone levels that stays well within safe limits. Full-blown starvation ketoacidosis tends to require a longer fast, an added physiological stressor, or both.

A reported case involved a young, otherwise healthy woman who developed starvation ketoacidosis after prolonged religious fasting.¹ Cases like these are uncommon but tend to catch everyone off guard precisely because the patient has no obvious risk factor like diabetes or alcoholism. An illness that reduces appetite, an eating disorder, or simply a prolonged voluntary fast can all set the stage, especially if dehydration compounds the problem.

Ketogenic Diets and Diet-Induced Ketoacidosis

The popularity of very low-carbohydrate and ketogenic diets has introduced a newer category of non-diabetic ketoacidosis. Under normal circumstances, a ketogenic diet produces a mild, controlled state of ketosis. Blood ketone levels rise modestly, and the body’s buffering system handles it without difficulty. The danger arrives when a second stressor piles on top of the dietary carbohydrate restriction.

One case described a 53-year-old non-diabetic woman who presented with nausea, abdominal pain, and dry mouth after losing 17 pounds in about three weeks on a low-carbohydrate ketogenic diet. Lab work confirmed anion gap metabolic acidosis with ketosis.⁶ A review of the literature on diet-associated ketoacidosis noted that while “dietary ketosis” is generally harmless, added stress can lead to massive overproduction of ketone bodies that overwhelms the body’s acid buffering.⁷

Those additional stressors can include an infection, surgery, intense exercise, or simply the cumulative effect of very aggressive caloric restriction on top of near-zero carbohydrate intake. The takeaway for anyone following a ketogenic diet is that mild ketosis and ketoacidosis are not the same thing, but the gap between them narrows when the body faces extra demands.

Lactation Ketoacidosis

Breastfeeding is metabolically expensive. Producing milk burns a significant number of calories and demands a steady supply of glucose. When a lactating woman also restricts carbohydrates heavily, the combination can be enough to tip the balance toward ketoacidosis. In lactating women, a mismatch between increased metabolic demand and reduced carbohydrate intake suppresses insulin, raises counterregulatory hormones, and promotes runaway ketone production.⁸

Multiple case reports describe life-threatening ketoacidosis in non-diabetic lactating mothers who were following strict ketogenic diets.⁹ In at least one case, the woman was also fighting a pneumonia, adding yet another metabolic drain on top of the calorie demands of breastfeeding and the carbohydrate restriction.¹⁰ The pattern that keeps emerging across these cases is that a single factor rarely causes ketoacidosis on its own, but two or three factors stacked together can push the body past its ability to compensate.

This is worth highlighting because ketogenic diets are heavily marketed for postpartum weight loss. A new mother who is breastfeeding, sleeping poorly, and eating very few carbohydrates could unknowingly be approaching a threshold. An added illness or even a few days of particularly poor intake could tip the situation from safe ketosis into acidosis.

Medication-Induced Ketoacidosis in Non-Diabetic Patients

A growing body of case reports involves medications originally developed for diabetes that are now prescribed more widely. SGLT-2 inhibitors, which work by making the kidneys excrete more glucose into the urine, are increasingly used for heart failure and kidney disease in people who do not have diabetes. By continuously removing glucose from the bloodstream, these drugs create a state that mimics carbohydrate deprivation even when the patient is eating normally. The body responds by shifting toward fat metabolism and ketone production. Residual insulin keeps blood sugar from climbing, but that same insulin level may not be enough to stop ketone production when an added stress like illness or surgery strikes.¹¹

The result is a condition sometimes called euglycemic ketoacidosis: full-blown acidosis with completely normal blood sugar. A narrative review of SGLT-2 inhibitor use in non-diabetic patients described the mechanism as a state of relative starvation combined with mild baseline ketonemia from the drug itself.¹² Patients with chronic kidney disease appear to be at extra risk because their kidneys are less able to clear ketones from the blood.¹¹

Newer GLP-1 receptor agonists like tirzepatide, widely used for weight loss, have also been linked to ketoacidosis in non-diabetic patients. In a series of cases, patients on tirzepatide developed hypoglycemic ketoacidosis, likely because the drug’s powerful appetite suppression led to severely reduced food intake, effectively triggering starvation physiology. All patients recovered after stopping the medication and resuming adequate nutrition, with symptoms resolving within about a month.¹³ As these drugs become more popular for weight management in the general population, clinicians and patients will need to watch for signs that extreme appetite suppression is tipping into dangerous territory.

Recognizing the Symptoms

The symptoms of non-diabetic ketoacidosis overlap considerably across its various causes, though the presentation can be subtle early on. Common early complaints include nausea, vomiting, abdominal pain, and loss of appetite. These are vague enough to be mistaken for a stomach bug, food poisoning, or a hangover, which is one reason the condition often goes unrecognized until it worsens.

As acidosis deepens, more specific signs appear. Kussmaul breathing, a pattern of deep, rapid, labored breathing, is the body’s attempt to blow off carbon dioxide and correct the acidic blood. Dehydration is almost always present, due to some combination of vomiting, poor intake, and the osmotic effects of ketones in the urine. A pediatric case series found that every child with non-diabetic ketoacidosis presented with dehydration, poor appetite, and Kussmaul breathing.¹⁴ Adults show the same pattern but may compensate longer before seeking care.

A fruity or acetone-like odor on the breath is a classic clue, though not always easy to detect in a busy emergency department. Confusion, lethargy, and eventually altered consciousness can develop if the acidosis is not corrected. The severity of neurological symptoms tends to track with the degree of acidosis, though the exact mechanisms behind the brain effects remain incompletely understood.¹⁵

Why Normal Blood Sugar Makes Diagnosis Harder

The single most important thing to understand about non-diabetic ketoacidosis is that blood sugar is often normal or even low. In diabetic ketoacidosis, sky-high glucose is a blazing red flag that immediately triggers the right tests. In alcoholic, starvation, and SGLT-2 inhibitor-associated ketoacidosis, glucose levels can sit comfortably in the normal range, or dip below it. The study comparing diabetic and alcoholic ketoacidosis found that while DKA patients had glucose levels averaging about 32 mmol/L, AKA patients averaged around 6.6 mmol/L.²

This matters enormously in clinical practice. A patient walks into an emergency room vomiting and dehydrated. A finger-stick glucose reads 90. Without a high index of suspicion, the next thought is probably a gastrointestinal infection or food poisoning, not ketoacidosis. The right diagnosis depends on measuring ketones directly, either through blood beta-hydroxybutyrate or urine ketones, and checking blood gas or electrolytes for an anion gap acidosis. In patients who are acutely ill and have a plausible trigger like heavy drinking, prolonged fasting, a very low-carb diet, breastfeeding, or SGLT-2 inhibitor use, clinicians need to think of ketoacidosis even when glucose looks reassuring.

Treatment Approach

The good news is that non-diabetic ketoacidosis tends to respond quickly to treatment, often faster than its diabetic counterpart. The cornerstone of treatment is giving the body what it’s missing: glucose and fluids. Intravenous dextrose solutions provide the carbohydrate the liver needs to stop pouring out ketones, while fluids address dehydration and help the kidneys clear the accumulated acid. In the diet-induced case of the 53-year-old woman, she was treated with a dextrose-containing saline solution and insulin coverage, and her acidosis corrected over a three-day hospital stay.⁶

In pediatric cases, treatment with insulin plus glucose supplementation led to faster recovery from acidosis compared to bicarbonate infusion or dialysis-type approaches.¹⁴ This makes physiological sense: insulin is the hormone that shuts down ketone production at its source. In non-diabetic patients, the pancreas can make insulin, but it needs the right stimulus (glucose) and the removal of whatever was suppressing it.

For medication-induced cases, stopping or adjusting the offending drug is a key part of the plan. The tirzepatide patients, for instance, recovered within about four weeks after discontinuing the medication and resuming normal eating.¹³ For alcoholic ketoacidosis, treatment also involves thiamine supplementation and addressing the underlying alcohol use disorder and its complications.

How Ketones Evolved as a Backup Fuel

It might seem strange that a normal metabolic process can become so dangerous. The explanation lies in how our bodies evolved. Humans have unusually large, energy-hungry brains, and throughout evolutionary history, food supply was unpredictable. Ketone bodies evolved as a critical backup fuel system, allowing the brain to keep functioning during famine when glucose was scarce. Under ordinary conditions in modern, regularly fed humans, blood ketone levels sit around 0.1 mmol/L, a fraction of what our distant ancestors likely experienced during routine food shortages.¹⁶

The system works beautifully within its intended range: mild ketosis during a short fast or overnight sleep is completely normal and harmless. The danger zone arrives when the usual braking mechanisms, mainly insulin, are overridden by illness, drugs, or extreme metabolic demand. What evolved as a survival advantage during famine becomes a liability when the buffering capacity of the blood is overwhelmed. Understanding this evolutionary context also explains why ketogenic diets can feel effective for many people: they are tapping into a metabolic pathway that humans used for millennia. The hazard only emerges at the extreme end, when something prevents the body from throttling ketone production back to safe levels.

Who Should Be Especially Aware

Several groups carry a higher risk of non-diabetic ketoacidosis and should know what to watch for:

People with alcohol use disorder: especially during or after a binge when food intake drops and vomiting leads to dehydration.
Breastfeeding mothers on low-carb diets: the combination of high caloric output and low carbohydrate intake can narrow the margin between safe ketosis and acidosis, particularly if an illness strikes.
Anyone on SGLT-2 inhibitors: whether prescribed for diabetes, heart failure, or kidney disease. Surgical procedures, acute illness, or reduced food intake while on these drugs substantially raises the risk.
Patients using GLP-1 agonists like tirzepatide: especially those who experience severe appetite suppression and eat very little for days at a time.
People practicing prolonged fasting: whether for religious, health, or weight-loss reasons, particularly if they are also exercising heavily or become ill during the fast.
Children with acute illness: younger children have smaller glycogen reserves and can shift to fat metabolism more quickly during illness-related fasting.

For anyone in these groups, the warning signs to take seriously are persistent nausea and vomiting, inability to keep food or fluids down, rapid or deep breathing, confusion, and a general sense that something is wrong beyond an ordinary stomach illness. These symptoms warrant a medical evaluation that includes checking for ketones and acidosis, not just a glucose level.

1
PubMed Central. Starvation Ketoacidosis in a Young Healthy Female After Prolonged Religious Fasting
2
Journal of Critical Care. Differences in metabolic and hormonal milieu in diabetic- and alcohol-induced ketoacidosis
3
The Journal of Emergency Medicine. Alcoholic Ketoacidosis: Etiologies, Evaluation, and Management
4
PubMed. Distinguishing between toxic alcohol ingestion vs alcoholic ketoacidosis: how can we tell the difference?
5
The American Journal of Emergency Medicine. Non-diabetic ketoacidosis: A case of alcoholic ketoacidosis accompanied by hyperglycemia
6
PubMed Central. Diet-induced Ketoacidosis in a Non-diabetic: A Case Report
7
PubMed Central. Another “D” in MUDPILES? A Review of Diet-Associated Nondiabetic Ketoacidosis
8
JCEM Case Reports. A Case of Severe Lactation Ketoacidosis in a Nondiabetic Mother on a Ketogenic Diet
9
PubMed Central. Ketogenic Diet-Induced Severe Ketoacidosis in a Lactating Woman: A Case Report and Review of the Literature
10
PubMed Central. Case Report: Lactation Ketoacidosis Can Complicate the Ketogenic Diet
11
Frontiers in Endocrinology. Case Report: Euglycemic ketoacidosis in a non-diabetic patient: a rare adverse effect of SGLT2 inhibitor therapy
12
PubMed Central. Euglycemic Ketoacidosis Associated with SGLT-2 Inhibitors in Non-diabetic Patients-A Narrative Review
13
American Journal of Case Reports. Exploring Hypoglycemic Ketoacidosis in Nondiabetic Patients on Tirzepatide: Is Starvation the Culprit?
14
PubMed Central. Pediatric non-diabetic ketoacidosis: a case-series report
15
Metabolism. Neurologic manifestations of diabetic comas: Correlation with biochemical alterations in the brain
16
PubMed Central. Brain Ketone Bodies in Health, Evolution and Disease