New-onset diabetes with ketoacidosis: interpreting the acid–base and metabolic picture
Arjun, a 16-year-old previously healthy student, is brought to the emergency department after 24 hours of vomiting and increasing drowsiness. His mother reports three weeks of excessive thirst, frequent urination at night and a 6 kg weight loss; his BMI is now 17.8 kg/m². He is dehydrated, breathes deeply at 28/min and his breath smells of acetone. Capillary glucose is 28.4 mmol/L, venous pH 7.09, PaCO2 18 mmHg, bicarbonate 6 mmol/L, sodium 132 mmol/L, potassium 5.6 mmol/L, chloride 98 mmol/L, urea 9.8 mmol/L and creatinine 110 µmol/L. Urine ketones are 4+, blood beta-hydroxybutyrate 6.8 mmol/L and HbA1c 11.2%. C-peptide is low and anti-GAD65 antibodies are positive. He has never received insulin. The resident calculates the anion gap and asks whether the potassium should be replaced before insulin is started.
1. Calculate the anion gap and classify the acid–base disturbance. What is the diagnosis, and how would you assess its severity biochemically?
Read the worked answer
Anion gap = Na⁺ − (Cl⁻ + HCO₃⁻) = 132 − (98 + 6) = 28 mmol/L, clearly raised (normal roughly 8–12 mmol/L without potassium, or 12–16 if potassium is included; the reference range must be stated). The primary disorder is a high-anion-gap metabolic acidosis. Respiratory compensation is appropriate: Winter formula gives expected PaCO2 = 1.5 × 6 + 8 ± 2 = 17 ± 2 mmHg, and the measured 18 mmHg fits. The delta ratio is (28 − 12)/(24 − 6) = 16/18 ≈ 0.9, consistent with a pure high-anion-gap acidosis without a second metabolic process. The diagnosis is diabetic ketoacidosis: hyperglycaemia, ketonaemia and acidosis with clinical dehydration. Severity is judged by pH, bicarbonate, beta-hydroxybutyrate, conscious level and haemodynamic state, not by glucose alone. The corrected sodium (add 2.4 mmol/L per 10 mmol/L glucose above 5.5) is about 137 mmol/L, suggesting predominant water loss rather than severe hypernatraemia. Effective osmolality is raised, explaining the osmotic diuresis. These calculations matter because a mixed picture, such as a normal anion gap from excessive saline or a concurrent lactic acidosis, changes fluid and insulin strategy.
2. Explain biochemically why insulin deficiency causes hyperglycaemia and ketonaemia, and why his serum potassium is high even though total body potassium is depleted.
Read the worked answer
Insulin normally promotes GLUT4 translocation in muscle and fat, stimulates glycogen synthesis and glycolysis, and suppresses hepatic glucose output. In insulin deficiency, glucose uptake falls while glucagon excess drives glycogenolysis and gluconeogenesis, so hyperglycaemia develops. Glucose above the renal threshold causes osmotic diuresis, producing dehydration and urinary losses of sodium, potassium and phosphate. The low insulin-to-glucagon ratio activates hormone-sensitive lipase in adipose tissue, releasing free fatty acids to the liver. Beta-oxidation generates acetyl-CoA, while low malonyl-CoA (reduced acetyl-CoA carboxylase activity) disinhibits carnitine palmitoyltransferase 1 and accelerates fatty acid entry into mitochondria. Acetyl-CoA exceeds tricarboxylic acid cycle capacity, so ketogenesis produces acetoacetate and beta-hydroxybutyrate, causing the raised anion gap. Potassium is high in serum because insulin deficiency, hyperosmolality and acidosis shift potassium out of cells and reduce renal excretion. Total body potassium is still low because of vomiting and osmotic diuresis. Insulin will drive potassium back into cells and can cause dangerous hypokalaemia, so potassium must be checked frequently and replaced once the serum level is not dangerously high and urine output is adequate.
3. Why are HbA1c and C-peptide useful here, and what are their limitations during this acute presentation?
Read the worked answer
HbA1c of 11.2% indicates sustained hyperglycaemia over the preceding 8–12 weeks and makes stress hyperglycaemia unlikely; it supports a chronic process such as new type 1 diabetes. Its limitations are important: it reflects average glucose, not the speed of onset, and can be falsely low with haemolysis, recent transfusion, haemoglobin variants or pregnancy, and falsely high in iron deficiency. It cannot diagnose ketoacidosis and should not be used alone to classify diabetes in an acute admission. C-peptide is low, consistent with beta-cell failure, and anti-GAD65 positivity strongly supports autoimmune type 1 diabetes. However, C-peptide can be suppressed in any acute illness and in ketoacidosis even when the underlying diabetes is type 2, so a single low value is not definitive. Antibody status is more specific. The combination of young age, lean body habitus, rapid symptom onset, ketosis, low C-peptide and positive autoantibodies makes type 1 diabetes overwhelmingly likely. The practical conclusion is to start insulin, not to repeat C-peptide during the acute phase.