Diabetic Ketoacidosis (DKA) & HHS

In Diabetic Ketoacidosis (DKA) (classically Type 1 diabetes with autoimmune beta-cell destruction), an absolute deficiency of insulin combined with surging counter-regulatory hormones (glucagon, epinephrine, cortisol, and growth hormone) unleashes uninhibited hormone-sensitive lipase (HSL) in adipose tissue. Massive lipolysis floods the liver with free fatty acids, which enter mitochondria via carnitine palmitoyltransferase-1 (CPT-1, disinhibited because low insulin lowers malonyl-CoA) and undergo beta-oxidation into acetoacetate and beta-hydroxybutyrate, producing a severe High Anion Gap Metabolic Acidosis within 24 hours. Conversely, in Hyperosmolar Hyperglycemic State (HHS) (classically elderly Type 2 diabetes), residual endogenous portal vein insulin is sufficient to inhibit adipose hormone-sensitive lipase (preventing ketogenesis and keeping arterial pH >7.30 and HCO₃^- >18 mEq/L), but is insufficient to stimulate peripheral skeletal muscle GLUT-4 glucose uptake or suppress hepatic gluconeogenesis. Days of insidious osmotic glucosuria produce a massive 8-to-12-liter free-water deficit, extreme hyperglycemia (>600–1,000 mg/dL), and an effective serum osmolality >320 mOsm/kg (2 × Na⁺ + Glucose/18), driving coma and focal neurological deficits (seizures, hemiparesis).

Two laboratory interpretation traps are heavily tested in DKA vignettes: (1) Direct Serum Beta-Hydroxybutyrate vs. Urine Nitroprusside Ketone Dipstick: during acute DKA, high NADH/NAD⁺ ratios shift the mitochondrial equilibrium strongly toward beta-hydroxybutyrate (3:1 to 10:1 ratio over acetoacetate). Standard urine and serum nitroprusside tablets detect only acetoacetate and acetone, NOT beta-hydroxybutyrate! As DKA resolves with insulin therapy, beta-hydroxybutyrate is oxidized back into acetoacetate, causing the urine nitroprusside ketone test to paradoxically turn more positive even as the patient improves—therefore, always monitor DKA resolution using serum beta-hydroxybutyrate and anion gap closure. (2) Hypertonic Translocational Hyponatremia: extracellular glucose draws water out of cells via osmosis, diluting serum sodium by 1.6 to 2.0 mEq/L for every 100 mg/dL of glucose above 100 mg/dL (add 1.6 × [(Glucose - 100)/100] to measured Na⁺ to determine true tonicity before selecting 0.9% vs. 0.45% NaCl).

Although DKA patients frequently present with normal or mildly elevated extracellular serum potassium due to hyperosmolality (solvent drag) and insulin deficiency (reduced Na⁺/K⁺-ATPase activity), every DKA patient has a massive total-body potassium deficit of 3 to 5 mEq/kg from osmotic diuresis and vomiting. Once intravenous regular insulin is started, insulin rapidly stimulates cell-membrane Na⁺/K⁺-ATPase to drive potassium back into intracellular compartments. Therefore, never administer insulin before checking serum potassium: if K⁺ is <3.3 mEq/L, HOLD insulin and infuse 0.9% Normal Saline with 20 to 40 mEq/hour IV KCl until K⁺ reaches ≥3.3 mEq/L (giving insulin at K⁺ <3.3 precipitates lethal ventricular fibrillation or respiratory paralysis); if K⁺ is 3.3 to 5.2 mEq/L, add 20 to 30 mEq KCl per liter of IV fluid and initiate IV regular insulin; if K⁺ is >5.2 mEq/L, start IV insulin without KCl and recheck serum K⁺ every 2 hours.

During DKA resuscitation with continuous IV regular insulin (0.1 units/kg/hr), serum glucose falls much faster (typically reaching 200 mg/dL within 4 to 6 hours) than hepatic ketogenesis and the high anion gap resolve (which requires 10 to 18 hours). A classic clinical error is discontinuing or transitioning off intravenous insulin as soon as serum glucose hits 200 mg/dL—because intravenous regular insulin has a half-life of only 4 to 5 minutes, stopping the infusion while the anion gap is still elevated (>12 mEq/L) unleashes hormone-sensitive lipase and triggers immediate rebound ketoacidosis. Instead, when serum glucose reaches 200 mg/dL in DKA (or 250–300 mg/dL in HHS), switch IV fluids to D5 0.45% NaCl (5% Dextrose with half-normal saline) and reduce the insulin infusion to 0.02–0.05 units/kg/hr to clamp blood glucose at 150–200 mg/dL while continuing IV insulin until Anion Gap is ≤12 mEq/L, serum HCO₃^- is ≥18 mEq/L, venous pH is >7.30, and the patient can eat—always overlapping subcutaneous basal glargine by 2 to 4 hours before turning off the IV insulin pump.

In the absence of insulin, uninhibited adipocyte Hormone-Sensitive Lipase (HSL) hydrolyzes stored triglycerides into glycerol and Non-Esterified Free Fatty Acids (FFAs). Simultaneously, high hepatic glucagon activates PKA, which phosphorylates and inactivates Acetyl-CoA Carboxylase (ACC), causing hepatocyte levels of Malonyl-CoA to plummet. Because Malonyl-CoA is the endogenous allosteric inhibitor of Carnitine Palmitoyltransferase-1 (CPT-1) on the outer mitochondrial membrane, low Malonyl-CoA opens the CPT-1 shuttle wide, flooding hepatocyte mitochondria with long-chain fatty acyl-CoA for β-oxidation. Massive β-oxidation generates more Acetyl-CoA than the TCA cycle can process (especially since oxaloacetate is depleted for gluconeogenesis) and generates a high NADH/NAD⁺ ratio. Excess Acetyl-CoA is condensed via HMG-CoA synthase/lyase into Acetoacetate, which the high NADH/NAD⁺ ratio immediately reduces via β-hydroxybutyrate dehydrogenase into β-Hydroxybutyrate (plus non-enzymatic decarboxylation to acetone, exhaled as fruity breath). Because standard urine/serum nitroprusside dipsticks detect ONLY acetoacetate (and not β-hydroxybutyrate), always order a direct serum β-hydroxybutyrate assay.

Because glucose is an osmotically active solute in the extracellular fluid (ECF) when insulin is absent, severe hyperglycemia creates a hypertonicity gradient that pulls free water out of the intracellular space into the plasma. This osmotic water shift dilutes the extracellular sodium concentration, producing translocational (hyperosmolar) hyponatremia. Before selecting maintenance IV fluids (0.9% NaCl vs. 0.45% NaCl), clinicians must calculate the Corrected Serum Sodium: add 1.6 mEq/L to the measured serum Na⁺ for every 100 mg/dL of glucose above 100 mg/dL (or +2.0 mEq/L when glucose > 400 mg/dL): Corrected Na⁺ = Measured Na⁺ + [0.016 × (Glucose − 100)]. For example, a DKA patient with a measured Na⁺ of 127 mEq/L and glucose of 600 mg/dL has a true corrected Na⁺ of 127 + (1.6 × 5) = 135 mEq/L. Note also that severely hypertriglyceridemic serum in DKA can cause artifactual pseudohyponatremia on indirect ion-selective electrode assays. Calculate Effective Serum Osmolality = (2 × Measured Na⁺) + (Glucose / 18) (BUN is excluded because urea freely crosses cell membranes as an ineffective osmole).

Filtered glucose normally undergoes 100% reabsorption in the proximal convoluted tubule via SGLT2 (90%) and SGLT1 (10%) cotransporters up to a renal transport maximum (T_m ~375 mg/min, corresponding to a plasma threshold of ~180–200 mg/dL). Once plasma glucose exceeds 200 mg/dL, unabsorbed glucose remains in the tubular lumen alongside negatively charged ketoacid anions (β-hydroxybutyrate^- and acetoacetate^-), obligating massive urinary losses of water (3–6 L in DKA; 8–12 L in HHS), sodium, potassium (3–5 mEq/kg deficit), phosphate, and magnesium. Meanwhile, ketoacid accumulation consumes serum bicarbonate (HCO₃^- ≤ 18 mEq/L), lowering arterial pH and stimulating medullary central chemoreceptors to generate Kussmaul respirations—deep, rapid hyperventilation that blows off PaCO₂ according to Winter's formula (Expected PaCO₂ = 1.5 × [HCO₃^-] + 8 ± 2).

Three synergistic mechanisms shift potassium out of the intracellular compartment (where 98% of body K⁺ normally resides at ~140 mEq/L) into the extracellular fluid prior to treatment: (1) Absolute Insulin Deficiency removes tonic stimulation of the cell-membrane Na⁺/K⁺-ATPase pump, allowing intracellular K⁺ to leak passively out of skeletal muscle and liver cells; (2) Extracellular Hypertonicity (Solvent Drag)—as severe hyperglycemia pulls water out of cells via osmosis, intracellular potassium is dragged out with the water across water channels and rises in concentration inside shrinking cells, driving passive K⁺ efflux down its steep concentration gradient; and (3) Acidemia (to a lesser extent in organic ketoacidosis than mineral acidosis) inhibits Na⁺/H⁺ exchange and promotes H⁺/K⁺ transcellular exchange.

The moment intravenous fluids and IV regular insulin are initiated, serum potassium ummets precipitously within 30 to 60 minutes via three mechanisms: (1) IV Insulin directly activates Na⁺/K⁺-ATPase across skeletal muscle, adipose, and hepatic tissue, pumping K⁺ rapidly back inside cells; (2) IV Isotonic Crystalloid expansion dilutes extracellular K⁺ and restores renal perfusion and distal tubular flow, accelerating urinary kaliuresis (aldosterone-mediated K⁺ excretion); and (3) Correction of acidemia and hyperosmolality reverses transcellular efflux. Consequently, if a patient starts with a 'low-normal' K⁺ of 3.1 mEq/L and receives an IV insulin bolus, serum K⁺ can drop to < 2.0 mEq/L within minutes, triggering Torsades de Pointes, ventricular fibrillation, and diaphragmatic respiratory arrest. Hence the ironclad board rule: Fluids first, check K⁺ second, and NEVER start insulin if K⁺ < 3.3 mEq/L.

Exactly like potassium, serum phosphate (PO₄^3-) is severely depleted in total body stores due to osmotic phosphaturia, yet appears normal or elevated on initial DKA presentation due to transcellular shift out of insulin-deficient cells. Once IV insulin is started, phosphate rushes back into cells along with glucose and potassium for ATP and 2,3-BPG synthesis, unmasking hypophosphatemia within 12–24 hours. Randomized trials show that routine IV phosphate administration in mild-to-moderate hypophosphatemia (> 1.0 mg/dL) does not improve clinical outcomes and can precipitate severe hypocalcemia and metastatic tissue calcification. Therefore, IV potassium phosphate is indicated ONLY if serum phosphate falls below 1.0 mg/dL OR if the patient develops life-threatening hypophosphatemic complications: respiratory muscle weakness/ventilator failure, cardiac systolic dysfunction, hemolytic anemia, or rhabdomyolysis (due to intracellular ATP and 2,3-BPG depletion).

During IV regular insulin infusion (0.1 units/kg/hr, or 0.14 units/kg/hr without bolus), plasma glucose declines at a target rate of 50 to 75 mg/dL per hour via renal excretion, volume dilution, and GLUT-4 uptake, typically reaching 200 mg/dL within 4 to 6 hours. However, suppressing adipocyte Hormone-Sensitive Lipase and allowing peripheral tissues to oxidize circulating β-hydroxybutyrate and acetoacetate (which regenerates HCO₃^- and closes the anion gap) requires 10 to 18 hours of continuous insulin signaling. Therefore, when blood glucose reaches 200 mg/dL in DKA (or 250–300 mg/dL in HHS), the anion gap is almost always still elevated (> 12 mEq/L). Clinicians must add 5% Dextrose (D₅W with 0.45% NaCl) to the IV fluids and step the insulin drip down to 0.02–0.05 units/kg/hr—using exogenous dextrose as a 'buffer' so continuous IV insulin can finish shutting down ketogenesis without causing hypoglycemia or rapid osmolar drops.

DKA is officially resolved when the Anion Gap is normal (≤ 12 mEq/L), blood glucose is < 200 mg/dL, serum bicarbonate is ≥ 15–18 mEq/L, venous/arterial pH is > 7.30, and the patient is mentally alert and able to tolerate oral nutrition. Because intravenous regular insulin has an elimination half-life of 4 to 5 minutes, its biological activity vanishes within 15–20 minutes of stopping the infusion pump. Conversely, subcutaneous long-acting basal insulin (insulin glargine or detemir) requires 2 to 4 hours to absorb from subcutaneous depots and reach therapeutic systemic concentrations (and rapid-acting lispro/aspart takes 30–60 minutes). If the nurse turns off the IV insulin drip at the same time subcutaneous glargine is administered, the patient experiences a 3-hour window of zero circulating insulin, triggering immediate rebound lipolysis and recurrent DKA. Always overlap the IV insulin infusion for 2 to 4 hours after injecting subcutaneous basal insulin.

During the recovery phase of DKA (12–24 hours into resuscitation), patients frequently develop a transient Normal Anion Gap (Hyperchloremic) Metabolic Acidosis (NAGMA)—where the anion gap has closed (≤ 10 mEq/L) but serum bicarbonate remains low (14–16 mEq/L) and serum chloride is high (> 110 mEq/L). This occurs for two reasons: (1) large volumes of 0.9% Normal Saline contain supraphysiologic chloride (154 mEq/L vs. plasma 100 mEq/L), and (2) ketoacid anions (β-hydroxybutyrate^-), which represent 'potential bicarbonate' once oxidized by the liver, were lost in the urine during osmotic diuresis and replaced by renal reabsorption of chloride. This hyperchloremic acidosis is benign and self-resolves as the kidneys regenerate HCO₃^-. As noted, IV Sodium Bicarbonate is reserved strictly for arterial pH < 6.90 (where severe acidemia depresses myocardial contractility and blunts catecholamine vasopressor responsiveness).

Cerebral edema is the leading cause of mortality in pediatric DKA (occurring in ~1% of children, typically 4 to 12 hours after initiating treatment). During days of untreated hyperosmolar hyperglycemia, neurons and astrocytes protect their cell volume from osmotic shrinkage by synthesizing and accumulating intracellular osmotically active molecules termed idiogenic osmoles (taurine, myo-inositol, betaine, glutamate). Because these idiogenic osmoles dissipate very slowly over 24–48 hours, if a clinician drops the extracellular plasma osmolality too rapidly—by administering excessive hypotonic fluids, giving an IV insulin bolus in children, failing to add D₅W when glucose reaches 200–250 mg/dL, or giving sodium bicarbonate—free water rushes down the osmotic gradient into the hyperosmolar brain cells, causing fatal uncal/tonsillar herniation. Warning signs include new headache, recurrent vomiting, lethargy, confusion, incontinence, and Cushing triad (hypertension, bradycardia, irregular respirations). Do NOT wait for a head CT: immediately elevate the head of the bed to 30°, slow IV maintenance fluids, and administer IV Mannitol (0.5–1 g/kg) or IV 3% Hypertonic Saline (5–10 mL/kg).

Euglycemic DKA is defined as high anion gap metabolic acidosis (pH ≤ 7.30, HCO₃^- ≤ 18 mEq/L, AG > 12) with elevated serum β-hydroxybutyrate (≥ 3.0 mmol/L) in a patient whose blood glucose is < 250 mg/dL (often 130–200 mg/dL). The #1 pharmacologic cause on board exams is Sodium-Glucose Cotransporter-2 (SGLT2) inhibitors (empagliflozin, dapagliflozin, canagliflozin), especially when combined with acute illness, surgery, fasting, low-carbohydrate/ketogenic diets, or excessive insulin dose reduction. SGLT2 inhibitors block proximal tubular glucose reabsorption, forcing continuous urinary glucose excretion (glucosuria) that keeps blood glucose deceptively normal while simultaneously lowering endogenous insulin secretion and directly stimulating pancreatic alpha-cell glucagon release, driving unrestrained hepatic ketogenesis. Euglycemic DKA also occurs in pregnancy (due to fetal glucose consumption and human placental lactogen insulin resistance). Management is identical to classic DKA, except that 5% or 10% Dextrose (D₅W / D₁₀W) must be started IMMEDIATELY alongside the IV regular insulin infusion at hour 1 (since baseline glucose is already < 250 mg/dL), and the SGLT2 inhibitor must be permanently discontinued (hold SGLT2i 3–4 days prior to any elective surgery).

Patients in DKA are uniquely susceptible to Acute Rhinocerebral Mucormycosis, a fulminant opportunistic fungal infection caused by ubiquitous molds of the order Mucorales (Rhizopus, Mucor, Lichtheimia/Absidia). Why does Rhizopus specifically target DKA? First, Rhizopus possesses the enzyme ketoreductase, allowing it to metabolize ketone bodies and thrive in high-glucose, acidic environments. Second, acidemia (low pH) protonates serum transferrin, causing transferrin to release bound ferric iron (free Fe³⁺), which Rhizopus avidly uptakes via high-affinity iron permeases to fuel explosive hyphal growth. Spores inhaled into the nasal turbinates and paranasal sinuses invade blood vessel walls (angioinvasion), causing thrombosis and ischemic tissue infarction—presenting with fever, unilateral facial/periorbital pain, proptosis, cranial nerve palsies (CN III, IV, V₁, V₂, VI via cavernous sinus thrombosis), and a pathognomonic black necrotic eschar on the nasal turbinates or hard palate. Biopsy shows broad (ribbon-like), non-septate (pauciseptate) hyphae branching at right angles (90°) (contrasted with Aspergillus narrow septate 45° branching). Treatment requires emergent radical surgical debridement of all necrotic tissue + IV Liposomal Amphotericin B (or Posaconazole/Isavuconazole step-down; note that Voriconazole is completely ineffective against Mucorales) + rapid correction of DKA acidosis.