Diabetes Insipidus

01Pathophysiology

Diabetes insipidus is a disorder of water homeostasis, not glucose. The central regulator is arginine vasopressin (AVP), also called antidiuretic hormone, synthesized in the supraoptic and paraventricular nuclei of the hypothalamus and stored in the posterior pituitary. Rising plasma osmolality stimulates AVP release. AVP binds V2 receptors on the basolateral membrane of renal collecting duct principal cells, activating adenylyl cyclase and triggering insertion of aquaporin-2 (AQP2) channels into the apical membrane. Water then flows down its osmotic gradient into the hypertonic renal medulla, producing concentrated urine.

In central DI, the hypothalamus or posterior pituitary fails to secrete adequate AVP. Importantly, the posterior pituitary has substantial functional reserve, and roughly 80 to 90 percent of AVP-producing magnocellular neurons must be lost before overt symptoms appear. This explains why partial lesions produce a milder phenotype while complete stalk transection produces fulminant polyuria. Common causes include neurosurgery (especially transsphenoidal resection of pituitary adenoma), head trauma, tumors (craniopharyngioma, germinoma, metastases), infiltrative disease (sarcoidosis, Langerhans cell histiocytosis), autoimmune hypophysitis, and idiopathic disease.

In complete central DI, AVP secretion is essentially abolished and the collecting duct receives no signal at all. AQP2 channels remain entirely sequestered, urine osmolality stays close to that of plasma water (typically 50 to 200 mOsm/kg), and daily urine output may reach 8 to 20 L. These patients develop hypernatremia rapidly when water access is interrupted.

In partial central DI, residual AVP-secreting neurons still respond to osmotic stimuli, but the response is blunted. The patient can achieve modest urinary concentration under stress, but cannot reach the normal maximum of 800 to 1200 mOsm/kg. Daily urine volume is typically 3 to 6 L, and symptoms may be limited to nocturia or mild polyuria. The diagnostic difficulty is that the water deprivation response overlaps with primary polydipsia, in which chronic AVP suppression also produces medullary washout.

In nephrogenic DI, AVP is present but the kidney does not respond. Lithium is the most frequent acquired cause: it accumulates in principal cells via ENaC channels and downregulates AQP2 expression. Hypercalcemia and hypokalemia impair the medullary concentrating gradient. Hereditary forms involve X-linked V2 receptor mutations (most common) or autosomal AQP2 mutations.

In gestational DI, placental vasopressinase (cysteine aminopeptidase) degrades circulating endogenous AVP, typically presenting in the third trimester. Primary polydipsia is a mimicker rather than true DI: excessive water intake suppresses AVP physiologically, washes out the medullary gradient, and produces polyuria with dilute urine.

The cardinal physiologic consequence of any form of DI is free water loss. When patients have access to water, thirst drives compensatory drinking and serum sodium remains near the upper limit of normal. The danger emerges when access to water is restricted (intubated patients, infants, the elderly, the cognitively impaired), where rapid hypernatremic dehydration develops and may cause seizures, coma, or intracranial hemorrhage.

02Classification and Clinical Manifestation

Clinically, all forms present with polyuria (>3 L/day or >50 mL/kg/day), polydipsia, and nocturia. A preference for ice-cold water is a classic vignette clue. Pediatric patients may present with enuresis, irritability, failure to thrive, and unexplained fevers.

Patients post-pituitary surgery may follow a triphasic pattern: initial DI (days 1 to 5) from axonal shock, transient SIADH (days 6 to 11) from release of stored AVP from degenerating neurons, then permanent DI if magnocellular damage is extensive. This pattern is testable and management-altering.

03Diagnostic Workup

The workup begins by confirming that polyuria is true and hypotonic. Order paired serum and urine osmolality. The diagnostic signature of DI is inappropriately dilute urine in the face of elevated serum osmolality, which a healthy kidney would never permit. A urine osmolality below 300 mOsm/kg with serum osmolality above 295 mOsm/kg essentially establishes the diagnosis of a concentrating defect. Always check fingerstick glucose first to exclude osmotic diuresis from diabetes mellitus.

If the patient is compensating by drinking and serum osmolality is not yet elevated, proceed to a water deprivation test, the traditional best initial functional test. Fluids are withheld under monitoring, with hourly measurement of body weight, urine volume, urine osmolality, and serum sodium. A normal patient or one with primary polydipsia will concentrate urine to >600 to 800 mOsm/kg. A patient with DI will continue to produce dilute urine and develop rising serum osmolality. The test is stopped when serum osm exceeds 295 to 300 mOsm/kg, weight loss exceeds 3 to 5 percent, or urine osmolality plateaus.

At that endpoint, administer desmopressin 2 to 4 mcg subcutaneously or 10 mcg intranasally and recheck urine osmolality at 1 and 2 hours. The interpretation is structured around the magnitude of the rise:

Partial central DI and primary polydipsia are the most diagnostically challenging pair, since both can produce intermediate urine osmolality on deprivation and modest rises with desmopressin. The discriminator is baseline serum sodium: high-normal in partial central DI, low-normal in primary polydipsia.

The most accurate test in modern practice is plasma copeptin, the C-terminal fragment of pre-pro-vasopressin that is co-secreted with AVP in equimolar amounts but is far more stable. Hypertonic saline-stimulated copeptin below 4.9 pmol/L identifies central DI, with greater accuracy than the water deprivation test, and can distinguish partial central DI from primary polydipsia where the older test struggles. Where available, copeptin-based testing is becoming the gold standard.

Once central DI is confirmed, an MRI of the pituitary and hypothalamus is mandatory to exclude tumor, infiltrative disease, or stalk lesion. The classic radiologic finding is absence of the posterior pituitary bright spot on T1-weighted images, reflecting depletion of AVP-containing neurosecretory granules. Stalk thickening raises suspicion for germinoma, sarcoidosis, or histiocytosis and warrants further evaluation.

04Management and Treatment

The first management priority in any DI patient is to assess volume status and serum sodium. A hemodynamically stable, alert patient with access to water can usually be managed outpatient. A patient with hypernatremia, altered mentation, or restricted oral intake requires hospitalization and careful free water replacement.

For central DI, the cornerstone is desmopressin, a synthetic AVP analogue with selective V2 activity (no V1 vasoconstrictive effect) and a longer half-life. Dosing differs by severity. Complete central DI typically requires 0.1 to 0.4 mg orally two to three times daily, 10 to 40 mcg intranasally per day divided into one or two doses, or 1 to 2 mcg subcutaneously every 12 hours for inpatients. Partial central DI is managed with lower doses, often 0.05 to 0.1 mg orally twice daily or 5 to 10 mcg intranasally daily, sometimes only at bedtime to control nocturia. Some patients with mild partial DI tolerate the condition well with adequate fluid access alone and do not require chronic pharmacotherapy. The principal complication is iatrogenic hyponatremia, so patients are taught to allow a small daily window of breakthrough polyuria before the next dose to permit free water excretion.

For nephrogenic DI, exogenous AVP will not work because the defect is at the receptor or channel level. Management has three pillars. First, remove the offending agent: stop lithium if possible (in consultation with psychiatry, because abrupt discontinuation can destabilize bipolar disorder), correct hypercalcemia, replete potassium. Second, institute a low-sodium, low-protein diet to reduce solute load and obligate water loss. Third, use medications that paradoxically reduce urine output. Hydrochlorothiazide 25 mg daily induces mild volume contraction, increases proximal tubular sodium and water reabsorption, and decreases delivery to the diluting segments. Amiloride 5 to 10 mg daily is particularly useful for lithium-induced nephrogenic DI because it blocks the ENaC channel through which lithium enters principal cells. Indomethacin or other NSAIDs reduce renal prostaglandin synthesis (which normally antagonizes AVP) and can be added in resistant cases, with attention to renal and gastric toxicity.

For gestational DI, desmopressin is safe and effective because it is not degraded by placental vasopressinase. Symptoms resolve within 4 to 6 weeks after delivery.

In acute hypernatremia from DI, calculate the free water deficit using:

$$\text{Water deficit (L)}=0.6\times \text{weight (kg)}\times (\frac{[{\text{Na}}^{+}{]}_{\text{measured}}}{140}-1)\backslash text\{Water\; deficit\; (L)\}\; =\; 0.6\; \backslash times\; \backslash text\{weight\; (kg)\}\; \backslash times\; \backslash left(\backslash frac\{[\backslash text\{Na\}^+]\_\{\backslash text\{measured\}\}\}\{140\}\; -\; 1\backslash right)$$Replace this deficit with 5 percent dextrose or half-normal saline orally or intravenously, while ensuring serum sodium falls no faster than 10 to 12 mEq/L per 24 hours to avoid cerebral edema. Begin desmopressin concurrently in central DI to halt ongoing losses.

The next best step when post-neurosurgical polyuria appears is to check paired serum and urine osmolality before giving any desmopressin. Empiric therapy without confirming the diagnosis can mask the diagnostic picture and precipitate hyponatremia, particularly because postoperative patients are at risk of the triphasic response in which an early DI phase is followed by transient SIADH on days 6 to 11. A short-acting parenteral desmopressin given as needed, rather than a fixed standing regimen, is preferred in the immediate postoperative period.

05Differential Diagnosis and Distractors

06Traps and High-Yield Pearls

The classic trap is confusing diabetes insipidus with diabetes mellitus because the keywords "polyuria" and "polydipsia" appear in both. The discriminator is the glucose: in DM the urine is full of solute and has high osmolality, while in DI the urine is essentially water with osmolality below 300 mOsm/kg. Always check a fingerstick glucose before launching into a DI workup.

The second trap is the partial central DI versus primary polydipsia dilemma. Both produce intermediate urine osmolality on water deprivation and both can show some rise in urine osmolality with desmopressin. Students who memorize only "DI fails to concentrate, polydipsia concentrates normally" miss the partial cases entirely. The discriminator is the baseline serum sodium (high-normal in partial central DI, low-normal in primary polydipsia), the clinical context (psychiatric history versus pituitary lesion or trauma), and stimulated copeptin if available.

The third trap is giving desmopressin without first distinguishing central from nephrogenic DI, or worse, giving it to a patient with primary polydipsia who will then develop dangerous hyponatremia. The water deprivation test followed by desmopressin challenge, or copeptin testing, must come before chronic therapy.

A fourth trap is the post-neurosurgical patient with the triphasic response. A trainee who treats early DI aggressively with standing desmopressin may then push the patient into severe hyponatremia when the SIADH phase begins on day 6 to 11. Treat as needed with short-acting desmopressin and reassess sodium and urine output frequently rather than locking in a fixed regimen.

Another testable subtlety is the lithium patient. The vignette will describe a patient on long-standing lithium for bipolar disorder presenting with polyuria. The right answer is nephrogenic DI, and the management pearl is amiloride, which blocks lithium uptake into principal cells and is more useful here than thiazide alone. Stopping lithium often does not fully reverse the defect if exposure has been prolonged.

In pediatric vignettes, watch for the male infant with failure to thrive, recurrent fevers, vomiting, and hypernatremic dehydration. The answer is X-linked nephrogenic DI from an AVPR2 mutation. These infants do not respond to desmopressin, and management is dietary plus thiazide and amiloride.

Recognize the classic vignette phrase "patient craves ice-cold water", which is near-pathognomonic shorthand for DI and is a tell that the question stem is steering you toward a water balance disorder rather than an endocrine or metabolic alternative.

The core competency being tested is the ability to recognize inappropriate urinary dilution in the setting of plasma hypertonicity, distinguish complete from partial central DI and central from nephrogenic DI through the desmopressin response, and manage acute hypernatremia at a safe correction rate without precipitating cerebral edema.