Diabetic ketoacidosis

Diabetic ketoacidosis is an acute metabolic emergency caused by marked absolute or relative insulin deficiency associated with an excess of counterregulatory hormones, particularly glucagon, catecholamines, cortisol, and growth hormone. The result is a well-defined pathophysiologic triad: hyperglycemia, hyperketonemia, and high-anion-gap metabolic acidosis. This condition is traditionally considered typical of type 1 diabetes mellitus, but it may also occur in type 2 diabetes mellitus in the presence of severe metabolic stress, infections, interruption of insulin therapy, pancreatitis, myocardial infarction, stroke, pregnancy, or treatment with sodium-glucose cotransporter 2 (SGLT2) inhibitors. Clinical injury does not arise from a single abnormal parameter, but from the convergence of dehydration, total electrolyte loss, progressive acidosis, impaired tissue perfusion, and activation of systemic catabolic mechanisms that, unless rapidly interrupted, may progress to shock, organ failure, and death.

From an epidemiologic standpoint, diabetic ketoacidosis is one of the leading causes of emergency hospitalization in people with diabetes, especially young individuals with type 1 diabetes, but its clinical burden is now also substantial among adults with type 2 diabetes, people with newly diagnosed diabetes, and patients exposed to iatrogenic or intercurrent triggers. In recent years, several studies have documented an increase in hospitalizations for diabetic ketoacidosis, a phenomenon related both to the rising prevalence of diabetes and to the longer survival of medically complex patients, the use of drugs that may promote euglycemic forms, and the persistence of educational, social, and organizational barriers to prevention. In well-resourced healthcare systems, mortality is now generally below 1% among adults treated promptly, but it rises with advanced age, diagnostic delay, severe comorbidities, sepsis, organ failure, pregnancy, and mixed presentations with a hyperosmolar state. In pediatric patients, prognosis is generally favorable when treatment is appropriate, but the most feared risk remains brain injury associated with ketoacidosis, a rare event nevertheless associated with significant mortality and possible permanent neurologic sequelae.

The causal sequence leading to diabetic ketoacidosis almost always begins with clinically significant insulin deficiency. In type 1 diabetes mellitus, this may result from complete discontinuation or marked reduction of exogenous insulin, insulin pump failure, errors in treatment management, or the first manifestation of the disease in a person not yet diagnosed. In type 2 diabetes mellitus, residual insulin reserve usually protects against overt ketogenesis, but this barrier may be overcome when insulin requirements rise abruptly or when endogenous secretion is already markedly reduced. In these circumstances, well-recognized precipitating factors come into play: infections, intercurrent febrile illnesses, pneumonia, urinary tract infections, sepsis, acute myocardial infarction, stroke, acute pancreatitis, surgery, trauma, alcohol misuse, use of glucocorticoids, sympathomimetics, or antipsychotics, persistent vomiting, prolonged fasting, pregnancy, and use of SGLT2 inhibitors.

Once insulin deficiency develops, metabolism shifts from an anabolic to a catabolic state. Reduced insulin action in peripheral tissues decreases glucose utilization in muscle and adipose tissue, while increased glucagon, epinephrine, cortisol, and growth hormone amplify hepatic gluconeogenesis and glycogenolysis. The liver therefore becomes an organ that overproduces glucose precisely when insulin-sensitive tissues lose the ability to take it up. The resulting rapid rise in blood glucose exceeds the renal threshold, causing glycosuria and osmotic diuresis. Urinary water loss is accompanied by losses of sodium, potassium, chloride, phosphate, magnesium, and other solutes, producing progressive depletion of extracellular volume and reduced renal blood flow. This further worsens hyperglycemia because the kidneys excrete less glucose and fewer ketone bodies, creating a self-perpetuating vicious cycle.

At the same time, loss of insulin-mediated inhibition in adipose tissue activates hormone-sensitive lipase and accelerates lipolysis. Free fatty acids released into the circulation reach the liver, where they undergo mitochondrial beta-oxidation. In the presence of a high glucagon-to-insulin ratio, hepatic metabolism directs large amounts of acetyl coenzyme A toward ketogenesis rather than the Krebs cycle. Acetoacetate and beta-hydroxybutyrate are thus formed, with secondary production of acetone. In diabetic ketoacidosis, beta-hydroxybutyrate often becomes the predominant ketone because the mitochondrial redox environment favors conversion of acetoacetate to its reduced form. This has major clinical importance because common nitroprusside-based urine tests primarily detect acetoacetate and acetone but underestimate beta-hydroxybutyrate; direct blood measurement is therefore more reliable for diagnosis and monitoring.

Accumulation of ketone bodies produces high-anion-gap metabolic acidosis. Acidemia reduces myocardial contractility, promotes vasodilation and hemodynamic compromise, stimulates the respiratory center with development of characteristic Kussmaul breathing, and contributes to nausea, vomiting, abdominal pain, and altered mental status. The body attempts to buffer the acid load by consuming bicarbonate, but when ketone production exceeds compensatory capacity, acidosis progresses rapidly. The condition is aggravated by volume depletion, which increases lactate production through relative hypoperfusion and reduces renal clearance of organic acids.

Potassium pathophysiology deserves particular attention because the initial serum concentration may be normal or elevated despite severe total-body depletion. Insulin deficiency and acidosis promote movement of potassium from the intracellular to the extracellular compartment, while osmotic diuresis increases urinary potassium loss. When insulin therapy is started and acidosis is corrected, potassium moves back into cells and the serum concentration may fall rapidly unless it is monitored and replaced. This is one of the most critical aspects of management because iatrogenic hypokalemia can cause arrhythmias, muscle weakness, and respiratory failure.

Some pathophysiologic variants require specific clinical reasoning. The most important is euglycemic diabetic ketoacidosis, in which acidosis and ketosis are present even though blood glucose is not particularly elevated. This form is promoted by fasting, vomiting, pregnancy, reduced carbohydrate intake, partial insulin administration, and especially SGLT2 inhibitors, which increase urinary glucose loss and may mask the severity of the crisis. The clinical risk in these patients is diagnostic delay: the absence of marked hyperglycemia may lead clinicians to underestimate fully evolving ketoacidosis. Another scenario is a mixed presentation with hyperosmolar hyperglycemic state, in which severe dehydration, hyperosmolality, and significant ketosis coexist. In such cases, prognosis tends to be worse and monitoring must be even more intensive.

In children and adolescents, the pathophysiologic process is similar, but the developing brain appears more vulnerable to rapid osmotic, inflammatory, and perfusion changes. Brain injury associated with ketoacidosis cannot be explained simply by isolated osmotic edema; it is now understood as the result of a combination of initial cerebral hypoperfusion, reperfusion, blood-brain barrier dysfunction, inflammatory mediators, altered water and solute transport, and, in some cases, iatrogenic factors related to overly aggressive correction of specific variables. Pediatric management therefore requires rigorous protocols, gradual correction, and continuous neurologic surveillance.

From the history, diabetic ketoacidosis often presents as the end result of a prodromal phase of metabolic deterioration. In the preceding hours or days, patients typically report polyuria, intense polydipsia, progressive fatigue, rapid weight loss, xerostomia, and worsening glycemic control. As the crisis evolves, nausea, vomiting, loss of appetite, abdominal pain, marked weakness, and reduced exercise tolerance develop. In patients with newly diagnosed diabetes, these symptoms may have been present for weeks without being recognized, whereas in patients with known diabetes the clinical history often reveals missed insulin doses, malfunction of the infusion device, a recent infectious illness, dehydration, or a condition of acute stress.

Respiratory symptoms are often characteristic. Patients may describe dyspnea or a sensation of air hunger, but physical examination primarily reveals Kussmaul breathing, a deep, regular, labored respiratory pattern representing the compensatory attempt to eliminate carbon dioxide and lessen acidemia. The breath may have the classic fruity odor caused by acetone, although this finding is neither constant nor sufficient on its own for diagnosis. Abdominal pain may sometimes be very severe and mimic an acute surgical abdomen; however, in the presence of severe acidosis and marked ketosis, it is often part of the metabolic presentation and tends to improve as the crisis is corrected. It remains essential not to attribute the pain automatically to ketoacidosis alone, because pancreatitis, abdominal sepsis, or another surgical emergency may be the precipitating cause.

Physical examination primarily reveals signs of dehydration and circulatory compromise. The mucous membranes are dry, skin turgor is reduced, the eyes are sunken, heart rate is increased, and blood pressure is low-normal or frankly hypotensive in more severe cases. Urine output, if still present, tends to decrease as hypovolemia worsens. Body temperature may be normal or low even in the presence of infection, so absence of fever does not exclude a septic trigger. Mental status ranges from normal alertness to drowsiness, confusion, obtundation, and coma, with an increasing likelihood of neurologic impairment as the severity of acidosis, hyperosmolality, and associated comorbidities increases.

The clinical presentation varies partly according to context. In newly diagnosed type 1 diabetes, especially in children, ketoacidosis may be the initial event and present with rapid weight loss, secondary enuresis, unquenchable thirst, vomiting, and abnormal breathing. In adults with type 2 diabetes, it may occur in the setting of severe infection, sepsis, myocardial infarction, or treatment with SGLT2 inhibitors and may have a less classic profile, sometimes with less pronounced hyperglycemia but significant ketosis. During pregnancy, the threshold for suspicion must be even lower because ketoacidosis may develop earlier, progress more rapidly, and, in a substantial proportion of cases, present without markedly elevated blood glucose.

In children and adolescents, clinical assessment must seek any early sign of possible cerebral involvement. Worsening headache, irritability, psychomotor slowing, a falling Glasgow Coma Scale score, relative bradycardia, hypertension, intractable vomiting, sudden incontinence, disorientation, and focal signs require immediate reassessment because they may precede clinically overt brain injury. In adults as well, altered mental status is a marker of severity and should prompt care in a more intensive setting.

The diagnostic pathway for diabetic ketoacidosis must be rapid, sequential, and directed at confirming three elements: hyperglycemia or a history of diabetes, significant ketonemia, and metabolic acidosis. Initial investigations include plasma or capillary glucose, venous or arterial blood gas analysis, serum electrolytes, bicarbonate, creatinine, blood urea nitrogen, calculated or measured osmolality, blood beta-hydroxybutyrate, complete blood count, urinalysis, electrocardiogram, and continuous monitoring of vital signs. Blood beta-hydroxybutyrate is now considered the most direct and pathophysiologically appropriate marker of ketosis, whereas urine ketone testing alone may be misleading both at presentation and during treatment.

The differential diagnosis of high-anion-gap metabolic acidosis must always be considered, although in clinical practice the combination of hyperglycemia or a history of diabetes, elevated ketonemia, and low bicarbonate makes diabetic ketoacidosis highly likely. The condition must nevertheless be distinguished from alcoholic ketoacidosis, ketosis due to prolonged fasting, lactic acidosis, advanced kidney failure, salicylate, methanol, or ethylene glycol poisoning, and other forms of high-gap acidosis. Euglycemic diabetic ketoacidosis requires particular attention because blood glucose may not reach classically expected levels. In these cases, suspicion arises from the association of acidosis and ketosis with a compatible clinical context, especially SGLT2 inhibitor therapy, pregnancy, fasting, vomiting, or reduced carbohydrate intake.

After confirming the crisis, its severity, precipitating factor, and possible complications must be established. Sodium should be interpreted after correction for hyperglycemia because the measured value may be spuriously low due to the osmotic shift. Potassium must be assessed with extreme caution, recognizing that a normal serum concentration does not exclude a severe total-body deficit. The anion gap is useful for confirming the presence of unmeasured acids and following the course, but it does not replace measurement of beta-hydroxybutyrate. In the presence of fever, substantial leukocytosis, respiratory symptoms, dysuria, abdominal pain, or hemodynamic instability, triggers are actively sought through blood cultures, urine culture, chest radiography, lipase, troponin, and other investigations guided by the clinical context.

According to the 2024 international consensus report on hyperglycemic crises in adults, diagnosis of diabetic ketoacidosis requires the simultaneous presence of hyperglycemia or a history of diabetes, significant ketosis, and metabolic acidosis. In pediatric patients, International Society for Pediatric and Adolescent Diabetes guidelines retain a biochemically comparable definition in principle, with particular emphasis on beta-hydroxybutyrate measurement and identification of cases requiring intensive management.

Official diagnostic criteria for diabetic ketoacidosis

• Plasma glucose ≥200 mg/dL or a known history of diabetes, recognizing that glucose may be lower in euglycemic forms
• Significant ketonemia with beta-hydroxybutyrate ≥3.0 mmol/L, or moderate or high ketonuria if blood measurement is unavailable
• Metabolic acidosis with venous or arterial pH <7.30 or serum bicarbonate <18 mmol/L

Severity classification in adults is based on the depth of acidosis, the degree of hyperketonemia, and mental status. In practical terms, distinguishing mild, moderate, and severe disease is useful for determining the care setting, monitoring intensity, and threshold for admission to an intensive or intermediate care unit. Presentations with altered mental status, hypotension, shock, significant acute kidney injury, hypoxemia, myocardial infarction, sepsis, or overlap with a hyperosmolar state require a higher level of care. For selected cases of mild uncomplicated disease, some guidelines and recent studies allow the use of subcutaneous rapid-acting insulin according to structured protocols, but this strategy requires experienced staff, close monitoring, and absence of instability.

The episode can be considered resolved only when ketosis and acidosis have actually been corrected. Normalization of blood glucose is not sufficient because hyperglycemia often resolves before the ketoacidotic component. Insulin therapy must therefore not be discontinued solely on the basis of the glucose concentration. According to the 2024 consensus report, resolution of diabetic ketoacidosis is defined by achievement of very low ketonemia, with beta-hydroxybutyrate below 0.6 mmol/L, together with venous pH of at least 7.3 or bicarbonate of at least 18 mmol/L. This point is crucial because stopping the insulin infusion too early creates a risk of relapse.

Treatment of diabetic ketoacidosis is based on four inseparable pillars: fluid replacement, insulin administration, electrolyte correction, and identification of the precipitating factor. None of these components is sufficient alone. Initial fluid therapy aims to restore tissue perfusion, improve glomerular filtration, and begin lowering blood glucose through dilution and increased renal glucose clearance. In adults, isotonic crystalloids are generally used, with the type and rate adapted to hemodynamic status, corrected sodium, and cardiac and renal function. In children, rehydration must be more cautious, calculated, and distributed over time according to specific protocols, precisely to limit the risk of brain injury associated with inappropriate correction.

Insulin is the causal treatment because it stops lipolysis and ketogenesis, reduces hepatic glucose production, and promotes peripheral glucose utilization. Continuous intravenous infusion of regular insulin remains the cornerstone for moderate or severe disease, whereas closely spaced subcutaneous rapid-acting analogs may be considered in selected patients with mild uncomplicated disease. Before starting insulin, however, the serum potassium concentration must be known because severe hypokalemia requires insulin to be delayed until appropriate replacement has begun. During treatment, as blood glucose falls, intravenous glucose is added not because the crisis has resolved, but to allow insulin to continue until ketoacidosis is fully corrected.

Potassium management is the most delicate aspect of treatment. Nearly all patients with diabetic ketoacidosis have a total-body potassium deficit even if the serum concentration appears reassuring or is elevated at presentation. Replacement is adjusted according to the initial value, renal function, urine output, and rate of correction. Serum phosphate and magnesium should be monitored, but routine phosphate replacement is not recommended because it has not shown generalized clinical benefit; it becomes reasonable in cases of severe hypophosphatemia, respiratory failure, myocardial dysfunction, or hemolytic anemia. Bicarbonate should likewise not be used systematically: in standard presentations it does not improve outcomes and may increase the risk of hypokalemia and paradoxical intracellular acidosis. Its use remains limited to selected scenarios of extreme acidosis under expert protocols.

Treatment of the precipitating factor is an integral component of care. Infection requires targeted antimicrobial therapy, acute coronary syndrome requires the appropriate cardiologic pathway, pancreatitis must be recognized and managed accordingly, insulin pump malfunction must be corrected immediately, and SGLT2 inhibitors must be discontinued when ketoacidosis is suspected. During pregnancy, management requires close collaboration among emergency medicine, diabetology, obstetrics, and anesthesia because maternal stabilization is the first measure for improving fetal prognosis as well. In the perioperative period, patients taking SGLT2 inhibitors should discontinue the drug several days before surgery according to current recommendations, specifically to reduce the risk of euglycemic presentations.

The transition from intravenous to subcutaneous insulin must be planned carefully. The infusion must not be stopped until biochemical resolution criteria have been met; the subcutaneous regimen should also be selected according to the anticipated nutritional intake. The first dose of subcutaneous basal insulin should be administered with adequate temporal overlap before discontinuing the intravenous infusion to prevent gaps in insulin coverage and early recurrence. This phase is particularly critical in patients with newly diagnosed diabetes, insulin pump users, and those who developed the crisis because of poor treatment adherence.

Prognosis depends on prompt diagnosis, the appropriateness of initial treatment, age, comorbidities, and the precipitating cause. In well-organized care settings, prognosis for isolated diabetic ketoacidosis is generally good, but it worsens markedly when sepsis, multiorgan failure, a hyperosmolar state, pregnancy, delayed access to care, or brain injury coexist. Long-term prognosis does not end with recovery from the acute episode: an event of ketoacidosis often signals persistent problems in diabetes management, such as social vulnerability, eating disorders, depression, inadequate access to insulin, or insufficient diabetes education. Every hospitalization should therefore conclude with a genuine recurrence-prevention plan.

Complications of diabetic ketoacidosis arise from both the disease itself and its treatment. The most immediate is severe dehydration with hypovolemia, which may progress to hypotension, shock, and prerenal or mixed acute kidney injury. Reduced tissue perfusion contributes to elevated lactate, worsens altered mental status, and impairs renal clearance of glucose and ketone bodies. If the precipitating cause is sepsis or myocardial infarction, hemodynamic instability may be even more pronounced, and distinguishing the effects of the metabolic crisis from those of the associated disease becomes clinically essential.

Potassium abnormalities are among the most dangerous complications. Initial hyperkalemia may cause conduction disturbances and arrhythmias, but the greater risk during treatment is often a rapid fall in potassium due to intracellular shift, especially if insulin is started without adequate monitoring or bicarbonate is used inappropriately. Hypophosphatemia and hypomagnesemia may also become clinically significant, contributing to muscle weakness, respiratory failure, and cardiac and neuromuscular abnormalities.

A classic but often underestimated complication is cerebral edema, or more accurately brain injury associated with diabetic ketoacidosis, which is particularly feared in pediatric patients. Its pathogenesis appears multifactorial and includes initial cerebral hypoperfusion, inflammatory processes, disruption of the blood-brain barrier, and possible effects of inappropriate osmotic correction. Warning signs include worsening headache, bradycardia, deterioration in consciousness, recurrent vomiting, sudden incontinence, focal signs, and seizures. This complication requires immediate recognition and urgent treatment with a reduction in the fluid administration rate and osmotic therapy according to a specialist protocol.

Other complications include pulmonary edema, aspiration in patients with vomiting or impaired consciousness, venous thrombosis in the setting of marked dehydration and immobility, rhabdomyolysis, associated or precipitating pancreatitis, and cardiac arrhythmias. In some patients, especially those with mixed presentations or advanced age, an overlapping hyperosmolar state may develop and amplify neurologic risk and mortality. Early recurrence is also a common management-related complication if the transition from intravenous to subcutaneous insulin is performed incorrectly or the trigger has not truly been removed.

Finally, there is a less conspicuous but highly important systemic prognostic complication: recurrence. Repeated episodes of diabetic ketoacidosis identify high-risk patients, often characterized by financial difficulties, poor access to care, psychiatric disorders, substance use disorders, low health literacy, or adolescent and family problems. In these individuals, secondary prevention cannot be limited to medication prescribing but must include structured diabetes education, psychosocial assessment, simplified access to supplies, sick-day instructions, ketone monitoring, and reassessment of the insulin regimen.