Diabetes insipidus (DI) is characterized by an inability to form a concentrated urine. This, in turn, causes the two most common symptoms, polyuria and polydipsia. The ability of the kidney to form a concentrated urine is dependent on adequate production and secretion of vasopressin (ADH) by the hypothalamic/pituitary axis and an intact collecting tubule vasopressin (V2) receptor/aquaporin 2 (AQP2) water channel signaling mechanism.

DI occurs in two main forms, central and nephrogenic. In central (neurohypophyseal) DI there is impaired production or secretion of ADH. In nephrogenic DI (NDI), ADH is produced but the kidney is unable to respond. NDI occurs in two forms: congenital or acquired (secondary). The presence of NDI is suggested by the production of large volumes of dilute urine and hypernatremia and unresponsiveness to the administration of exogenous ADH.

Typical findings:

Continue Reading

  • Polyuria and polydipsia

  • Failure to thrive, fever, recurrent episodes of dehydration, vomiting

Congenital NDI – Congenital NDI is a genetic condition that typically presents in the first year of life with failure to thrive, fever, vomiting and recurrent episodes of dehydration. The most common form is X-linked, in which males are predominantly affected. In rare cases, heterozygote (carrier) females can display phenotypes similar to the affected males. Rare Autosomal Recessive and Autosomal Dominant forms have been reported.

Acquired (Secondary) NDI – Acquired NDI is a condition that results from tubular dysfunction due to a variety of factors, including acute or chronic kidney disease, obstructive uropathy, medications, electrolyte imbalances and systemic diseases.

A number of disorders, which present with polydipsia and polyuria, can mimic NDI. These include:

1. Central (neurohypophyseal) DI

2. Psychogenic polydipsia

3. Dipsogenic DI

4. Bartter syndrome

Central DI – Central (or neurohypophyseal) DI results from the inability to make or secrete vasopressin (antidiuretic hormone, ADH). This is distinguished from NDI by responsiveness to administration of exogenous vasopressin. Central DI can be caused by a variety of central nervous system diseases involving the hypothalamus and/or pituitary. These include brain tumors, trauma, infection, vascular disease medication effects. Central DI can also be caused by inherited or congenital disorders involving the hypothalamic/pituitary axis.

Psychogenic polydipsia- Psychogenic polydipsia, also called compulsive water intake, results from chronic ingestion of large volumes of water due to an underlying psychiatric condition.

Dipsogenic DI – Dipsogenic DI is due to disruption in the normal thirst mechanism, resulting in ingestion of large volumes of water.

Bartter syndrome – Bartter syndrome is an inherited disorder (autosomal recessive) that typically presents in infants or young children with polyuria, recurrent episodes of volume depletion and failure to thrive. It is due to a mutation in one of several genes that encode transporters involved with sodium, chloride or potassium handling in the Loop of Henle.

Unlike nephrogenic DI, patients with Bartter’s show evidence of both sodium and water wasting (typically isotonic fluid). Thus, they present with volume depletion but normal serum sodium levels. In this disorder, potassium homeostasis is also abnormal, due to renal losses. The renin-angiotensin system is also upregulated due to chronic volume depletion from sodium loss. As a result, patients invariably demonstrate some degree of hypokalemia, hypochloremia and metabolic alkalosis, findings that are typically absent in nephrogenic DI.

Congenital NDI is an inherited disorder caused by mutations in the genes encoding either the vasopressin (ADH) V2 receptor in the collecting duct of the kidney, or less commonly, the aquaporin 2 water channels. Both are necessary for the kidney to be able to absorb water in the collecting tubule and form a concentrated urine.

Acquired (secondary) NDI is caused by one of numerous factors outlined below. The patient’s history (e.g., chronic lithium therapy) will often suggest that etiology and will readily distinguish congenital and acquired forms of NDI.

A variety of factors, particularly medications, have been implicated as potential causes of secondary (acquired) NDI:

Primary renal diseases/electrolyte abnormalities:

  • Diuresis phase of acute tubular necrosis (ATN)

  • Chronic kidney disease

  • Postobstructive diuresis

  • Hypercalcemia

  • Hypokalemia


  • Amphotericin B

  • Cidofovir

  • Colchicine

  • Contrast agents

  • Cyclophosphamide

  • Demeclocycline

  • Dexamethasone

  • Dopamine

  • Epirubicin

  • Ethanol

  • Foscarnet

  • Ifosfamide

  • Indinavir

  • Lithium

  • Lobenzarit

  • Mesalazine

  • Methoxyflurane

  • Netilmicin

  • Ofloxacin

  • Orlistat

  • Pimozide

  • Rifampin

  • Streptozocin

  • Tenofovir

  • Triamterene-hydrochlorothiazide

Other disorders:

  • Amyloidosis

  • Hyperthyroidism

  • Sarcoidosis

  • Sickle cell disease or trait

  • Sjogren’s syndrome

  • Serum chemistries

  • Urinalysis, including urine specific gravity

  • Simultaneously-measured serum and urine osmolarity

  • Genetic testing if congenital NDI is suspected

Renal ultrasonography is useful for assessing the presence of secondary causes of NDI, such as chronic or acute kidney disease. In those instances, the kidneys may be small (e.g., in chronic kidney disease) or enlarged (e.g., in acute kidney injury) and demonstrate increased echogenicity.

For congenital NDI, there are no imaging studies that confirm the diagnosis. Renal ultrasonography is helpful, however, in identifying the presence of hydronephrosis, a well described complication of congenital NDI.

A number of different approaches to confirming the diagnosis of nephrogenic diabetes insipidus have been proposed.

A. Traditional water deprivation test

  • Restrict water until 3%-5% body weight loss or 3 consecutive urine osmolalities within 10%

  • Obtain serum chemistry to confirm resultant hypernatremia (serum sodium must be >144 meq/L)

  • Obtain urine osmolarity (osm)

    Interpretation: Urine Osm > 800 mOsm/kg = normal;

    300-800 mOsm/kg = partial DI (suggestive, may not be diagnostic, simultaneous serum ADH level may be helpful in this instance)

    i. <300 mOsm/kg = complete DI

  • If the urine osm is low and the serum sodium is elevated, administer of DDAVP (vasopressin) at a dose of 10-20 mcg intranasally (for patients >3 months of age)

    Interpretation: Urine Osm >800 mOsm/kg – positive response (= central DI)

    300-800 mOsm/kg – equivocal

    <300 mOsm/kg – no response (= NDI)

B. 7-hour water deprivation test. This is recommended for infants and young children in whom prolonged water deprivation may be dangerous. Although maximal urinary concentrating ability may not be achieved, it can readily distinguish central and complete nephrogenic DI.

  • Measure urine and serum osmolarity (osm) at the start of the test

  • Begin 7-hour water restriction and monitor urine volume and urine specific gravity hourly

  • After completion of the 7 hours, obtain urine and serum osmolarities and administer DDAVP 10-20 mcg intranasally

  • Monitor for an additional 4 hours with hourly urine output and urine specific gravity. During that time the patient can have oral fluids equal to the volume of urine output

  • At the completion of the post DDAVP 4 hours of monitoring (or sooner if the urine specific gravity is 1.014 or greater), obtain urine and serum osmolarities

    Interpretation: Urine osm/serum osm ratio of <1.5 after 7 hours of water deprivation (before DDAVP administration) = abnormal; a value of ≥1.5 is normal

    Urine osm/serum osm ratio of >1.5 after DDAVP is suggestive of a positive response (central DI), <1.5 is a negative response (NDI)

C. DDAVP challenge in patients with polyuria and hypernatremia. In a patient who has hypernatremia and polyuria with a low urine specific gravity in whom DI is suspected, DDAVP has been given to distinguish central versus nephrogenic DI. Serial urine and serum osmolarities are then monitored for a response.

Patients with clinical evidence of shock should receive immediate fluid resuscitation with normal saline in order to stabilize vital signs.

Continued administration of isotonic fluids when vital signs are stabilized or in the absence of overt shock (tachycardia, hypotension, and/or impaired peripheral perfusion) is not recommended, since the defect results in the loss of free (non-electrolyte-containing) water.

Once perfusion is established, serum sodium should be corrected slowly with hypotonic fluids:

Water deficit should be corrected over a period of 48 hours. In order to determine the rate of fluid delivery, first calculate the water deficit using the formula:

Deficit (ml) = 0.6 x weight (kg) x [current sodium/140) -1]

then determine maintenance water and sodium requirements for 48 hours using the formulas

Water: 100ml/kg/d for the first 10 kg; 50 ml/kg/d for the next 10 kg; 20 ml/kg/day for >20kg

Sodium: 2 meq/kg/day

Add the deficit water + maintenance water (x 2 days) and divide by 48 to get an hourly fluid rate

Take the total volume of deficit + maintenance x 2 days and the total sodium to determine the concentration of the fluid to be delivered:

EXAMPLE: 10 kg child. Serum sodium = 165meq/L.

Water deficit = 0.6 x 10 kg x (165/140 – 1) = 1080 ml

Maintenance water (2 days) = 10 kg x 100 ml/kg/day x 2 days = 2,000 ml

Maintenance sodium (2 days) = 10 kg x 2meq/kg/day x 2 days = 40 meq

Total volume for the 2 days = 1080 + 2000 = 3080 ml /48 hours = 64 ml/hr

Sodium concentration of IVF = 40 meq/3080 ml = D5W with 13 meq/L of sodium chloride (=1/8 NS)

In patients with congenital NDI, chronic treatment is focused on decreasing massive polyuria, maintaining adequate fluid intake and preventing recurrent episodes of dehydration.

Unlimited access to free water is essential for those able to drink, as the thirst mechanism is generally intact in affected patients.

Dietary sodium restriction should be initiated to maximize the effectiveness of the thiazide diuretics.

Dietary protein restriction (<2 g/kg/day) may also decrease the renal water losses but may limit the ability to deliver adequate nutrition.

Medications utilized to decrease polyuria in patients with congenital NDI are listed in Table I. Patients are typically treated with a combination of amiloride and hydrochlorothiazide (HCTZ). However, in some instances indomethacin may be required to achieve sufficient control of polyuria.

Table I.
Medication Dosage
Amiloride 0.3 mg/kg/day(or 20 mg/day/1.73m2)
Hydrochlorothiazide(HCTZ) 2-3 mg/kg/day divided BID-TID

2 mg/kg/day divided TID

Limited data are available regarding the use of these agents in acquired NDI, although indomethacin has been reported to be effective in decreasing urine output in lithium-associated NDI.

Overly aggressive correction of serum sodium may lead to the complication of cerebral edema. Therefore, serum sodium should be corrected at a rate of no more than 0.5 meq/L/hour (12 meq/L/day).

Combination treatment with HCTZ and amiloride is the preferred treatment for congenital NDI; indomethacin with HCTZ is equally effective but associated with significant short and long term side effects.

HCTZ often causes hypokalemia. If used alone or with indomethacin potassium supplementation may be required.

Amiloride (a potassium sparing diuretic) may cause hyperkalemia, but this is unlikely to occur when used in combination with HCTZ.

Indomethacin can cause gastritis and gastrointestinal ulcers, hyperkalemia, acute kidney injury, and chronic renal tubular injury.

Successful treatment with a selective COX-2 inhibitor (Rofecoxib) has been reported in 1 child with congenital NDI. However, that medication is no longer available and COX-2 inhibitors are not approved for use in children.

For patients with congenital NDI, survival is generally excellent if the disease is recognized and treated and if recurrent episodes of potentially life threatening hypernatremia can be prevented. Patients are at risk for poor growth (both weight gain and linear growth) and developmental delays (particularly in patients with recurrent episodes of dehydration and hypernatremia).

The outcomes for acquired NDI depend on the underlying etiology. Lithium associated NDI, for instance, occurs in 30-50% of patients. With short term use, the effects are reversible once the drug is stopped. However, long term use may be associated with irreversible damage, chronic kidney disease and eventual progression to end stage kidney disease.

Congenital NDI is caused by inherited defects in the genes encoding the vasopressin V2 receptor or the AQP2 water channel.

The incidence of congenital NDI is not well established, but is generally believed to be very rare. One study suggests an incidence is 8.8 per 1,000,000 males.

Acquired NDI has multiple etiologies (as listed earlier). Commonly seen causes include acute kidney injury, chronic kidney disease, obstructive uropathy (congenital or acquired), lithium and post-diuresis phase of ATN.

Congenital NDI has three forms: (1) X-linked recessive, (90% of cases), caused by mutations in the gene encoding the vasopressin V2 receptor; (2) autosomal recessive (9% of cases), caused by mutations in the gene encoding the AQP2 water channel; and (3) autosomal dominant (1% of cases), caused by mutations in the gene encoding the AQP2 water channel.

Mutations in the genes encoding the vasopressin V2 receptor or the AQP2 water channel result in loss of the ability to respond to ADH by impairing tubular water reabsorption in the collecting duct.

The pathogenesis of acquired NDI is variable based on the underlying etiology. Polyuria in obstructive uropathy (particularly acute) appears to be related primarily to severe downregulation of AQP2 channel production. On the other hand, lithium associated NDI is believed to be multifactorial, caused by: (1) impaired function of the epithelial sodium channel (as the result of the lithium ion competing directly with sodium for binding to the protein) causing loss of the medullary concentrating gradient; (2) inhibition of AQP2 water channel production; and (3) rearrangement of the collecting duct microstructure with changes in the relative number of principal cells and intercalating cells.


Complications of congenital NDI include recurrent episodes of hypernatremia and dehydration, seizures, developmental delays, failure to thrive, poor linear growth, hydronephrosis and bladder dysfunction.

Complications of therapy include electrolyte imbalances (hypokalemia related to thiazide diuretics), acute kidney injury (indomethacin) and chronic tubulointerstitial injury with fibrosis (indomethacin) and gastrointestinal and hematologic complications (indomethacin). In some instances paradoxical water intoxication and hyponatremia can result when treatment is initiated while large volumes of free water are being administered concomitantly.

Complications of acquired (secondary) NDI depend on the underlying etiology. In general, complications are less common because the severity of the NDI may be less and patients may be older and able to compensate with increased water intake. Children with secondary NDI from congenital obstructive uropathy, however, may be at risk for many of the complications of congenital NDI, including hypernatremia, failure to thrive and poor growth. Lithium associated NDI may be irreversible with long-term use, even if the drug is discontinued and chronic kidney disease can result.

Genetic testing for congenital NDI is available (see www.geneclinics.org for a listing of clinical genetics laboratories performing this testing).

Congenital NDI is a genetic disorder that cannot be prevented.

Acquired NDI may be preventable (e.g., by avoiding known agents such as lithium). In many cases (e.g., chronic kidney disease, congenital obstructive uropathies), irreversible tubular damage cannot be prevented.

NDI Review:

Bockenhauer, D, Bichet, DG. “Pathophysiology, diagnosis and management of nephrogenic diabetes insipidus”. Nat Rev Nephrol. vol. 11. 2015 Oct. pp. 576-88. (This review presents an overview of clinical aspects of primary and secondary NDI including diagnosis and management as well as a discussion of the underlying pathophysiology and novel therapies under study/in development.)

Monnens, L, Jonkman, A, Thomas, C. “Response to indomethacin and hydrochlorothiazide in nephrogenic diabetes insipidus”. Cli Sci. vol. 66. 1984. pp. 709-15. (This case series examined the efficacy of indocin, HCTZ or both in decreasing urine output in four patients with X linked NDI. HCTZ was slightly more effective than indocin but the combination of the two was most effective.)

Libber, S, Harrison, H, Spector, D. “Treatment nephrogenic diabetes insipidus with prostaglandin synthesis inhibitors”. J Pediatr. vol. 108. 1986. pp. 305-11. (This study examined the relative efficacy of ibuprofen versus indocin in eight patients with congenital NDI. Ibuprofen had minimal effect on polyuria, whereas indocin decreased urine output. The study highlights the fact that different prostaglandin inhibitors have differential effects in this disease.)

Alon, U, Chan, CM. “Hydrochlorothiazide-amiloride in the treatment of congenital nephrogenic diabetes insipidus”. Am J Nephrol. vol. 5. 1985. pp. 9-13. (This study compares HCTZ alone compared to HCTZ and amiloride in the treatment of two males with congenital NDI. The two drug combination was found to be more effective in preventing hypokalemia. At 10 month follow up this regimen was well tolerated.)

Knoers, N, Monnens, LA. “Amiloride-hydrochlorothiazide versus indomethacin-hydrochlorothiazide in the treatment of nephrogenic diabetes insipidus”. J Pediatr. vol. 117. 1990. pp. 499-502. (This prospective cross over study of five children with congenital NDI compared amiloride-HCTZ with indomethacin-HCTZ. Urine volumes were comparable in the two groups but serum sodium levels were significantly lower in the amiloride HCTZ group. In addition, the amiloride HCTZ group did not require potassium supplementation, whereas those in the indomethacin-HCTZ group did. The authors concluded that amiloride-HCTZ was as effective as indomethacin-HCTZ without the need for potassium supplementation or the risks of long-term indomethacin use.)

Uyeki, TM, Barry, FL, Rosenthal, SM, Mathias, RS. “Successful treatment with hydrochlorothiazide and amiloride in an infant with congenital nephrogenic diabetes insipidus”. Pediatr Nephrol. vol. 7. 1993. pp. 554-556. (This study reports the first successful treatment of an infant with congenital NDI with HCTZ and amiloride. It also describes administration of DDAVP to a hypernatremia, polyuric infant to distinguish NDI from central DI.)

Kirchlechner, V, Koller, DY, Seidl, R, Waldhauser, F. “Treatment of nephrogenic diabetes insipidus with hydrochlorothizide and amiloride”. Arch Dis Child. vol. 80. 1999. pp. 548-552. (This case series reports the effect of long term (up to 5 years) of treatment with HCTZ and amiloride in four children with congenital NDI. All patients had normal growth and no electrolyte problems. Four out of five had normal development. Fluid intake was 3.8-7.7 m2/day, urine output was 2.2-7.4 L/m2/day, illustrating that treatment was not able to fully correct the poluria and polydipsia.)

Boussemarrt, T, Nsota, J, Martin-Coignard, D, Champion, G. “Nephrogenic diabetes insipidus: treat with caution”. Pediatr Nephrol. vol. 24. 2009. pp. 1761-63. (This cautionary case report illustrates the potential complication of water intoxication following initiation of treatment w/ HCTZ and indomethacin [with liberal water intake] in a child with congenital NDI.)

Knoers, N, Pagon, RA, Adam, MP, Ardinger, HH. “Nephrogenic diabetes insipidus”. GeneReviews [Internet]. 2000 Feb 12. pp. 1993-2014. (This web-based, peer-reviewed article presents a comprehensive review of the pathogenesis, diagnosis and treatment of congenital NDI, with a focus on the genetic aspects of the disease.)

Allen, HM, Jackson, RL, Winchester, MD, Deck, LV, Allon, M. “Indomethacin in the treatment of lithium-associated nephrogenic diabetes insipidus”. Arch Int Med. vol. 149. 1989. pp. 1123-6. (This case study of one patient reports successful treatment of lithium-associated NDI with indomethacin. At 3 month follow up, there was sustained decrease in urine output.)

Khanna, A. “Acquired nephrogenic diabetes insipidus”. Sem Nephrol. vol. 26. 2006. pp. 244-248. (This review presents an overview of the etiologies of acquired NDI.)

Trepiccione, F, Christensen, BM. “Lithium induced nephrogenic diabetes insipidus: new clinical and experimental findings”. J Nephrol. vol. 23. 2010. pp. S43-48. (This review presents an up to date discussion of the natural history and pathogenesis of lithium associated NDI.)

Bockenhauer, D, Bichet, DG. “Inherited secondary nephrogenic diabetes insipidus: concentrating on humans”. Am J Physiol Renal Physiol. vol. 304. 2013 Apr 15. pp. F1037-42. (This review presents an overview of secondary NDI that occurs as a component of various hereditary tubulopathies, such as Bartter syndrome.)

Ellison, DH. “Disorders of Sodium and Water”. Am J Kidney Dis. vol. 46. 2005. pp. 356-361. (This review outlines the traditional water deprivation test.)

Hendricks, SA, Lippe, B, Kaplan, SA, Lee, WN. “Differential diagnosis of diabetes insipidus: Use of DDAVP to terminate the seven-hour water deprivation test”. J Pediatr. 1981. pp. 244-246. (This report describes a modified 7-hour water deprivation test.)

The point at which lithium treatment will result in irreversible damage is not well defined.

The use of diuretics and/or indomethacin in the treatment of acquired NDI has not been well studied.