INTRODUCTION AND DEFINITION — Polyuria has generally been defined as a urine output exceeding 3 L/day in adults and 2 L/m² in children. It must be differentiated from the more common complaints of frequency or nocturia, which may not be associated with an increase in the total urine output.

The evaluation of patients with polyuria is discussed in this topic. The causes, clinical manifestations, and treatment of polyuria due to central or nephrogenic diabetes insipidus (DI) are presented separately:

●(See "Clinical manifestations and causes of central diabetes insipidus".)

●(See "Clinical manifestations and causes of nephrogenic diabetes insipidus".)

●(See "Treatment of central diabetes insipidus".)

●(See "Treatment of nephrogenic diabetes insipidus".)

CAUSES OF POLYURIA — Polyuria may be caused by a solute (osmotic) diuresis or a water diuresis (table 1 and algorithm 1). (See 'Solute (osmotic) diuresis' below and 'Water diuresis' below.)

The most common cause of an osmotic diuresis is glucose-induced diuresis in patients with uncontrolled diabetes mellitus. Three other major causes of polyuria are all due to water diuresis, in which large volumes of dilute urine (urine osmolality usually below 250 mosmol/kg) are excreted: primary polydipsia, which is primarily seen in adults and adolescents; central diabetes insipidus (DI); and nephrogenic DI [1].

Solute (osmotic) diuresis — In some patients with polyuria, particularly of acute onset, the increase in urine output may be due to a solute or osmotic diuresis in which the primary abnormality is an inability to reabsorb a substantial proportion of the filtered solute (table 1).

Major causes of polyuria due to a solute diuresis include:

●Glucosuria – Glucosuria usually results from hyperglycemia in patients with diabetes mellitus.

●Urea diuresis – A urea diuresis most often occurs in patients with resolving acute kidney injury.

●Sodium diuresis – A sodium diuresis is usually caused by administration of large volumes of intravenous saline or after relief of bilateral urinary tract obstruction.

Glucosuria (usually due to hyperglycemia) — Glucosuria, usually due to uncontrolled diabetes mellitus, is the major cause of an osmotic diuresis in outpatients. In addition, inpatients with severe central DI who are treated with large volumes of intravenous dextrose and water can develop hyperglycemia, glucosuria, and polyuria that is antidiuretic hormone (ADH) resistant. (See "Treatment of central diabetes insipidus".)

Glucosuria can also develop in diabetic patients treated with sodium-glucose co-transporter 2 inhibitors. These drugs inhibit renal glucose absorption and produce glucosuria and an osmotic diuresis [2]. (See "Sodium-glucose co-transporter 2 inhibitors for the treatment of hyperglycemia in type 2 diabetes mellitus".)

Other causes of solute diuresis — Aside from glucosuria, other causes of solute diuresis are:

●Urea diuresis – Polyuria due to a urea diuresis (in which urea acts as an osmotic agent) may develop in several settings. Large amounts of urea may be derived from:

•Resolution from azotemia.

•Administration of urea as therapy in patients with hyponatremia [3]. (See "Treatment of hyponatremia: Syndrome of inappropriate antidiuretic hormone secretion (SIADH) and reset osmostat", section on 'Urea'.)

•Tissue catabolism (eg, due to glucocorticoid therapy). Lean body tissues are approximately 20 percent protein by weight, and protein catabolism results in the production of urea.

•Administration of high amounts of protein orally, via feeding tube, or intravenously, which will be catabolized to urea.

●Sodium diuresis – Volume expansion due, for example, to large volumes of intravenous saline or to the release of bilateral urinary tract obstruction may produce a sodium diuresis and polyuria [1,4-6].

●Mannitol – Mannitol, which is usually given to treat increased intracranial pressure, can induce an osmotic diuresis and polyuria. (See "Complications of mannitol therapy".)

Polyuria in which there is a marked increase in sodium excretion (ie, sodium salts make up most of the increased solute excretion) is almost always appropriate, resulting from volume expansion induced either by saline loading or, following relief of urinary obstruction, by fluid retained during the period of obstruction [5,6]. In addition to the excretion of retained sodium, increased urea excretion can contribute to a postobstructive diuresis when urea has also been retained as evidenced by a substantial increase in blood urea nitrogen. (See "Clinical manifestations and diagnosis of urinary tract obstruction and hydronephrosis", section on 'Prognosis and recovery of renal function'.)

Many clinicians faced with a urine output that may initially exceed 1000 mL per hour in adults feel compelled to replace the urine output with intravenous fluids. This will only prolong the polyuria driven by increased sodium excretion since volume expansion will persist. Optimal therapy of a postobstructive diuresis consists of fluid infusion at a maintenance level, such as 75 mL of one-half isotonic saline per hour. The development of volume depletion, as evidenced by hypotension or a rise in the blood urea nitrogen, is unusual with this regimen [5,6].

An exception to this general rule can occur in infants with congenital obstructive uropathy in whom the associated ADH resistance is sufficiently severe that the maximum urine osmolality is well below that of the plasma. In these infants, relief of the obstruction can lead to a marked water diuresis with subsequent hypernatremia [7,8]. Following relief of obstruction, fluid status should be carefully monitored and fluid therapy adjusted as necessary.

It is possible that some of these patients have congenital nephrogenic DI, with polyuria, not obstruction, being responsible for urinary tract and bladder dilation [9-11]. (See "Treatment of nephrogenic diabetes insipidus".)

Not causes of true polyuria: Salt wasting and typical diuretic use — As a general rule, underlying kidney disease can impair sodium conservation in the presence of volume depletion. However, except in rare cases of Bartter syndrome, salt-wasting nephropathies do not cause true polyuria. This is in part due to the phenomenon of tubuloglomerular feedback in which increased sodium chloride delivery to the macula densa (due to decreased reabsorption in the more proximal segments) results in afferent arteriolar constriction and a fall in glomerular filtration rate, thereby limiting the degree of sodium chloride loss (figure 1) [12-14]. (See "Inherited hypokalemic salt-losing tubulopathies: Pathophysiology and overview of clinical manifestations", section on 'Clinical manifestations'.)

In addition, sodium excretion induced by routine, chronic use of diuretic medications should not produce persistent polyuria (eg, hypertensive patients taking a thiazide or loop diuretic). Such patients who retain sodium when the effect of the diuretic dissipates may complain of nocturia as extracellular volume is redistributed during the night, but true polyuria does not result if the dose of diuretic is constant.

Water diuresis — Primary polydipsia, central DI, and nephrogenic DI are associated with an increase in water output and the excretion of a relatively dilute urine (ie, water diuresis) (table 1). In primary polydipsia, the polyuria is an appropriate response to enhanced water intake, whereas the water loss is inappropriate with either form of DI.

Primary polydipsia — Primary polydipsia (sometimes called psychogenic polydipsia) is characterized by a primary increase in water intake. This disorder is most often seen in middle-aged women and in patients with psychiatric illnesses, including those taking a phenothiazine, which can lead to the sensation of a dry mouth. Primary polydipsia can also be induced by hypothalamic lesions that directly affect the thirst center, as may occur with an infiltrative disease such as sarcoidosis [1]. (See "Causes of hypotonic hyponatremia in adults".)

Central DI — Central diabetes insipidus (DI; also called neurohypophyseal or neurogenic DI) is due to deficient secretion of ADH. This condition is most often idiopathic (possibly due to autoimmune injury to the ADH-producing cells) or can be induced by trauma, pituitary surgery, or hypoxic or ischemic encephalopathy. Rare familial cases have been described [15]. (See "Clinical manifestations and causes of central diabetes insipidus".)

Nephrogenic DI — Nephrogenic diabetes insipidus (DI) is characterized by normal ADH secretion but varying degrees of renal resistance to its water-retaining effect. This problem, in its mild form, is relatively common since most patients who are older adults or who have underlying kidney disease have a reduction in maximum renal concentrating ability. This defect, however, is usually not severe enough to produce a symptomatic increase in urine output.

Symptomatic polyuria due to ADH resistance is seen primarily in four settings, all of which are discussed in detail elsewhere (see "Clinical manifestations and causes of nephrogenic diabetes insipidus"):

●Nephrogenic DI presenting in childhood is almost always due to inherited defects. The most common are X-linked hereditary nephrogenic DI due to mutations in the AVPR2 gene encoding the ADH receptor V2 and autosomal recessive and dominant nephrogenic DI due to mutations in the aquaporin-2 (water channel) gene.

●Nephrogenic DI presenting in adults is almost always acquired with chronic lithium use and hypercalcemia being the most common causes of a defect severe enough to produce polyuria.

EVALUATION OF SUSPECTED POLYURIA — In addition to nephrogenic and central diabetes insipidus (DI), a variety of conditions may result in the complaint of polyuria, including psychogenic polydipsia, prostatic hypertrophy, or osmotic diuresis (including postobstructive diuresis). (See "Clinical manifestations and diagnosis of urinary tract obstruction and hydronephrosis", section on 'Prognosis and recovery of renal function'.)

The cause of polyuria is often suggested from the history (eg, age of onset and eliciting the possible presence of the different causes of DI) and, rarely, by the plasma sodium concentration. Specific testing is then performed to establish the diagnosis.

Hyperglycemia or another obvious osmotic diuresis — In some patients, the etiology of polyuria is clear from the history and initial laboratory testing, often making further evaluation unnecessary (algorithm 1) (see 'Glucosuria (usually due to hyperglycemia)' above and 'Other causes of solute diuresis' above):

●Diabetic patients with severe hyperglycemia and glucosuria have a solute (osmotic) diuresis caused by filtered glucose that is not reabsorbed.

●Treatment with a sodium-glucose co-transporter 2 inhibitor can produce polyuria; glucosuria and onset after initiation of the drug are clues that the medication is responsible for the polyuria. However, the polyuric effect of this class of medication is relatively modest (eg, approximately 200 mL over six hours after a 25 mg dose of empagliflozin) [16].

●Treatment with exogenous urea or glucocorticoids (which can cause catabolism) or administration of a very high protein diet can produce a urea diuresis and polyuria.

●Administration of large volumes of saline, typically in hospitalized patients, may produce an appropriate sodium diuresis.

●Release of bilateral urinary tract obstruction often results in polyuria due to excretion of retained sodium and urea.

●Administration of mannitol to patients with increased intracranial pressure can produce an osmotic diuresis and polyuria.

If unsure, despite the history and initial laboratory testing, the presence of one of these causes of osmotic diuresis can be confirmed by measuring the urine osmolality (which is typically greater than 400 to 600 mosmol/kg in such patients) and calculating the total daily osmolar output (which should be >1000 mosmol in patients with a solute diuresis) (algorithm 1). The total daily osmolar output is calculated as follows:

If the urine output is not known, it can be estimated in patients with stable glomerular filtration rate from the creatinine concentration of a spot urine sample. If the urine creatinine concentration is expressed in mg/dL and the daily urine creatinine excretion is assumed to be 1 g per day, then:

If the urine output is estimated in this way, then the total daily osmolar output can be estimated by the following equation:

When the cause is not obvious — Solute diuresis is associated with an increase in urine output and the excretion of a concentrated urine. Primary polydipsia generates the appropriate excretion of dilute urine. By contrast, central DI, and nephrogenic DI are associated with an inappropriate increased output of relatively dilute urine.

Measurement of the plasma sodium concentration and the urine osmolality may be helpful in distinguishing among these disorders:

●A low plasma sodium concentration with a low urine osmolality (eg, less than one-half the plasma osmolality) is indicative of water overload due to primary polydipsia.

●A high plasma sodium concentration points toward either an osmotic diuresis or, particularly if the urine osmolality is less than the plasma osmolality, to DI [1].

●A normal plasma sodium concentration is not helpful in diagnosis but, if associated with a urine osmolality more than 600 mosmol/kg in a patient with polyuria, it excludes a diagnosis of DI and indicates that the cause of polyuria is a solute (osmotic) diuresis.

●Total solute excretion (calculated on a 24-hour urine collection from the product of the urine osmolality and the measured [or estimated] urine volume [in liters]) is normal with a water diuresis (600 to 900 mosmol per day on a typical Western diet) but markedly increased with an osmotic diuresis. (See "Patient education: Collection of a 24-hour urine specimen (Beyond the Basics)".)

Confirm the presence of polyuria — In patients with suspected polyuria, 24-hour urine volume should ideally be assessed to confirm the presence of polyuria (algorithm 1). The urine osmolality, sodium, potassium, chloride, creatinine, urea nitrogen, and glucose should be measured in this 24-hour sample.

Polyuria has generally been defined as a urine output exceeding 3 L/day in adults and 2 L/m² in children. It must be differentiated from the more common complaints of frequency or nocturia, which are not associated with an increase in the total urine output.

However, potential problems with a timed urine collection include an incomplete collection and the necessity for multiple containers if the patient has severe polyuria. The latter can be minimized by obtaining an 8-hour rather than a 24-hour collection. Measurement of urinary creatinine excretion can help determine if the collection is complete. (See "Patient education: Collection of a 24-hour urine specimen (Beyond the Basics)".)

Patients with a normal serum sodium

Determining if further testing is necessary — The first step in the evaluation of polyuric patients with a normal serum sodium is to distinguish between a solute (osmotic) diuresis and a water diuresis (algorithm 1).

Patients with a water diuresis should undergo water restriction or infusion of hypertonic saline to achieve a serum sodium concentration high enough (>145 mEq/L) to stimulate vasopressin sufficiently to produce a maximally concentrated urine. (See 'Goal of water restriction (or hypertonic saline)' below.)

Once a serum sodium >145 mEq/L has been achieved, there are two options to help distinguish primary polydipsia from DI and to classify DI:

●Measurement of urine osmolality before and after the administration of desmopressin (algorithm 2). This is the most common approach and what we do in our practice.

●Determination of plasma copeptin or vasopressin. However, reliable assays for copeptin and vasopressin are infrequently available. (See 'Water restriction (or hypertonic saline) test' below.)

Exceptions to this general rule are patients with a dilute urine (ie, urine osmolality well below that of the plasma) who are strongly suspected of having nephrogenic DI (eg, long-term lithium use) and newborns and young infants who are thought to have hereditary nephrogenic DI. In these patients who are resistant to antidiuretic hormone (ADH), the response to desmopressin can be evaluated without prior water restriction. (See 'Protocol in infants and children' below.)

Measurements of plasma vasopressin or copeptin are an alternative to the assessment of the response to desmopressin or aqueous vasopressin. However, reliable vasopressin levels are only available in research laboratories, and reliable plasma copeptin assays are not yet available in many regions, including in the United States.

If reliable plasma copeptin assays are available, then additional methods for evaluating normonatremic patients with polyuria and a water diuresis are possible (algorithm 3):

●If the baseline plasma copeptin is >21.4 pmol/L, then the patient has nephrogenic DI [17]. Water restriction (or hypertonic saline) is unnecessary in such cases.

●Assuming that nephrogenic DI has been excluded, the plasma copeptin can be measured again after water restriction (or hypertonic saline), once the serum sodium is >145 mEq/L; central DI is diagnosed if the plasma copeptin is ≤4.9 pmol/L, and primary polydipsia is diagnosed if a higher value is obtained [18]. Administration of, and judging the response to, desmopressin is unnecessary.

●Water restriction (or hypertonic saline) plus desmopressin administration can be avoided altogether, after exclusion of nephrogenic DI, by measuring the plasma copeptin after infusion of arginine (0.5 g/kg over 30 minutes). Arginine provides a nonosmotic stimulus to the posterior pituitary. If, at 60 minutes after infusion, the plasma copeptin is ≤3.8 pmol/L, then the diagnosis is central DI [19]. However, if the plasma copeptin is >3.8 pmol/L, then the diagnosis is likely to be primary polydipsia (ie, with 92 percent specificity). The arginine-stimulated plasma copeptin test has not been systematically evaluated in children [20].

If suspicion for central DI remains high despite an arginine-stimulated plasma copeptin >3.8 pmol/L, then water restriction (or hypertonic saline) plus desmopressin administration can be performed. (See 'Water restriction (or hypertonic saline) test' below.)

Urine osmolality >600 mosmol/kg — Patients with a urine osmolality >600 mosmol/kg and a normal serum sodium have a solute (osmotic) diuresis. The cause of solute diuresis can often be elucidated from the history. (See 'Solute (osmotic) diuresis' above.)

Water restriction and administration of desmopressin are not necessary in such patients.

Urine osmolality 300 to 600 mosmol/kg — Patients with an intermediate urine osmolality of 300 to 600 mosmol/kg may have a solute diuresis, primary polydipsia, or DI.

To determine if a solute diuresis is present, the total daily osmolar output can be calculated as the urine osmolality x the 24-hour urine volume. If the total daily osmolar output is >1000 mosmol, then a solute diuresis is present. (See 'Solute (osmotic) diuresis' above.)

However, if the total daily osmolar output is <900 mosmol (a normal value), then the patient has either primary polydipsia or DI. In such patients, the history and the response to the water restriction test can help distinguish primary polydipsia from DI and help classify the type of DI. As noted below, hypertonic saline can be used in addition to or in place of water restriction if water restriction is unsuccessful (often in cases of primary polydipsia or partial DI) or cannot be performed. (See 'Water restriction (or hypertonic saline) test' below.)

Urine osmolality <300 mosmol/kg — Polyuric patients with a low urine osmolality of <300 mosmol/kg have either primary polydipsia or DI. Most of these patients should undergo water restriction and, if necessary, administration of desmopressin in order to distinguish primary polydipsia from DI and to classify the type of DI (algorithm 2). Hypertonic saline can be used in addition to or in place of water restriction if water restriction is unsuccessful (often in cases of partial primary polydipsia or DI) or cannot be performed. (See 'Water restriction (or hypertonic saline) test' below.)

However, water restriction (or hypertonic saline) may be unnecessary if the history and initial evaluation point to a particular cause of nephrogenic DI:

●Bilateral urinary tract obstruction – Resolution of the polyuria with correction of the obstruction confirms the diagnosis of nephrogenic DI; if obstruction is chronic, the diagnosis can be made by evaluating the response to desmopressin (without water restriction). (See "Clinical manifestations and causes of nephrogenic diabetes insipidus", section on 'Kidney disease'.)

●Persistent hypercalcemia – A diagnosis of nephrogenic DI can be made by documenting the resolution of polyuria with correction of the hypercalcemia or by evaluating the response to desmopressin (without water restriction). (See "Clinical manifestations and causes of nephrogenic diabetes insipidus", section on 'Hypercalcemia'.)

●Onset of polyuria at an early age in patients with a family history of autosomal dominant or X-linked hereditary nephrogenic DI – The diagnosis of nephrogenic DI can be made by evaluating the response to desmopressin (without water restriction) or by genetic testing. (See "Clinical manifestations and causes of nephrogenic diabetes insipidus", section on 'Hereditary nephrogenic DI'.)

Alternatively, plasma copeptin can be measured to diagnose nephrogenic DI, although assays for copeptin are frequently unavailable (algorithm 3). Copeptin, the C-terminal glycoprotein moiety of pro-arginine vasopressin (AVP), is a stable surrogate marker of vasopressin secretion. Without prior water deprivation, a single baseline copeptin level >21.4 pmol/L was found to differentiate nephrogenic DI from other etiologies with 100 percent sensitivity and specificity, rendering water deprivation testing and desmopressin administration unnecessary [17].

Water restriction (or hypertonic saline) test — If the diagnosis is unclear, patients should be evaluated by raising the serum sodium concentration and plasma osmolality either by water restriction (supplemented by hypertonic saline if necessary) or solely by the administration of hypertonic saline. In general, we use administration of hypertonic saline only if the water restriction test is unsuccessful (often in cases of primary polydipsia or partial DI) or cannot be performed.

Once a serum sodium >145 mEq/L and serum osmolality >295 mosmol/kg is attained, exogenous ADH is often given to distinguish primary polydipsia from DI and to determine the type of DI. Desmopressin is usually preferred to aqueous vasopressin because it produces fewer side effects. However, desmopressin has a longer half-life and can result in hyponatremia in patients with primary polydipsia. If there is a strong suspicion of primary polydipsia, aqueous vasopressin can be substituted for desmopressin.

Combining hypertonic saline infusion with serial measurements of copeptin (the C-terminal glycoprotein moiety of the AVP prohormone and therefore a surrogate marker of ADH) can accurately distinguish DI from primary polydipsia [18]; provided that assays for copeptin are available, this may be performed as an alternative to judging the response to desmopressin. (See 'If water restriction is nondiagnostic' below.)

Goal of water restriction (or hypertonic saline) — It is important to achieve a serum sodium concentration >145 mEq/L and serum osmolality >295 mosmol/kg with water restriction and/or hypertonic saline to differentiate central DI from primary polydipsia; simply giving exogenous ADH before the serum sodium is >145 mEq/L will not distinguish between these two conditions, since both have submaximal plasma ADH levels and both will respond to desmopressin therapy.

Thus, we continue water restriction and/or hypertonic saline until the serum sodium concentration is >145 mEq/L and serum osmolality is >295 mosmol/kg or until the urine osmolality reaches a clearly normal value (ie, approximately 700 mosmol/kg or greater) [21,22].

Protocol in adults and adolescents — Our approach to the water restriction test in adults and adolescents depends upon the severity of the polyuria and the urine osmolality:

●Most patients with a urine osmolality >100 mosmol/kg can safely undergo overnight fluid restriction and be evaluated the following morning. This duration of water deprivation often provides sufficient time for the plasma osmolality and serum sodium concentration to reach a level high enough (ie, 295 to 300 mosmol/kg and 145 mEq/L) to stimulate enough ADH release to maximally concentrate the urine of normal subjects.

●By contrast, individuals with severe polyuria and a urine osmolality <100 mosmol/kg (who likely have complete DI) should not undergo overnight water restriction. Rather, they should refrain from fluids for two to three hours prior to evaluation. In such patients, two to three hours without fluids typically provides enough time for the plasma osmolality and serum sodium concentration to increase sufficiently (ie, 295 to 300 mosmol/kg and 145 mEq/L) to stimulate enough ADH release to maximally concentrate the urine in normal subjects.

Overnight fluid restriction should be avoided in these patients with severe polyuria since potentially severe volume depletion and hypernatremia can be induced. As an example, we saw a patient with undiagnosed DI who was scheduled for surgery in the morning and was told to stop drinking and eating after going to sleep the night before. Surgery was cancelled when routine preoperative laboratory testing revealed a plasma sodium of 162 mEq/L.

The duration of water restriction that will be required to achieve an adequate stimulus for endogenous ADH secretion can be estimated from the patient's body weight and known hourly urine volume [23]:

When the patient is evaluated after water restriction (whether overnight or just two to three hours), the urine osmolality, plasma osmolality, and plasma sodium should be measured:

●If the urine osmolality has reached a clearly normal value (above approximately 700 mosmol/kg), this indicates that both ADH release and effect are intact and the diagnosis is primary polydipsia. Patients with partial DI may have a substantial rise in urine osmolality but not to this extent.

●If the urine osmolality is <700 mosmol/kg, the plasma osmolality exceeds 295 to 300 mosmol/kg, and the plasma sodium is 145 mEq/L or higher, we administer desmopressin (or, less commonly, aqueous vasopressin) and judge the response. The typical dose of desmopressin is 10 mcg by nasal insufflation or 2 to 4 mcg subcutaneously or intravenously. The typical dose of aqueous vasopressin is 5 units subcutaneously. (See 'Judging the response to desmopressin' below.)

As noted above, a plasma osmolality of 295 to 300 mosmol/kg and plasma sodium of 145 mEq/L should maximally stimulate ADH release. Thus, such patients with a urine osmolality <700 mosmol/kg may have either abnormal ADH release (central DI) or ADH effect (nephrogenic DI).

●If the urine osmolality is <700 mosmol/kg, the plasma osmolality is <295 mosmol/kg, and the plasma sodium is <145 mEq/L, we continue water deprivation and also infuse 3 percent (hypertonic) saline, 0.1 mL/kg/min intravenously. While giving hypertonic saline, we monitor the urine osmolality and plasma sodium every hour. We stop the infusion when the serum sodium reaches 145 mEq/L or higher or when the urine osmolality is above 700 mosmol/kg (this may take several hours):

•If the urine osmolality reaches a value greater than 700 mosmol/kg, then the diagnosis is primary polydipsia. Such patients are often nonadherent to fluid restriction, which can explain why they may present with a urine osmolality <600 mosmol/kg and plasma osmolality <295 mosmol/kg despite instructions to avoid fluids overnight.

•If the serum sodium reaches 145 mEq/L or higher, we administer desmopressin (or, less commonly, aqueous vasopressin) and judge the response. The typical dose of desmopressin is 10 mcg by nasal insufflation or 2 to 4 mcg subcutaneously or intravenously. The typical dose of aqueous vasopressin is 5 mcg subcutaneously. (See 'Judging the response to desmopressin' below.)

The normal physiologic response to the water restriction test (or the administration of hypertonic saline) is based upon the following observations [1,24,25]:

●Raising the plasma osmolality (and serum sodium concentration) leads to a progressive elevation in ADH release and an increase in urine osmolality in normal individuals (figure 2).

●Once the plasma osmolality reaches 295 to 300 mosmol/kg (normal 275 to 290 mosmol/kg) or the plasma sodium is 145 mEq/L or higher, the effect of endogenous ADH on the kidney is maximal. At this point, administering desmopressin or vasopressin will not further elevate the urine osmolality unless endogenous ADH release is impaired (ie, unless the patient has central DI).

Protocol in infants and children — Water restriction is not performed in newborns or very young infants suspected to have hereditary nephrogenic DI (eg, documented plasma sodium 145 mEq/L or higher with a concomitant urine osmolality ≤200 mosmol/kg). If the diagnosis is unclear in such patients, the preferred diagnostic test is the administration of desmopressin (1 mcg subcutaneously or intravenously infused over 20 minutes, maximum dose 0.4 mcg/kg of body weight) with measurement of the urine osmolality at baseline and at 30-minute intervals over the next two hours. If the urine osmolality does not increase by more than 100 mosmol/kg over baseline, the diagnosis of nephrogenic DI is made and DNA should be obtained for mutation analysis [26]. However, the AVPR2 mutations (D85N, V88M, R104C, R106C, Y128S, L161P,G201D, T273M, F287L M311V, N317K N317S, N321Y, P322S, and the splice mutant c.276A>G) have been associated with a mild phenotype responsive to desmopressin [27].

Water deprivation tests for older infants and children should be performed in the hospital under close medical supervision. The patient should not be allowed to lose more than 5 percent of their body weight. Monitoring of vital signs (temperature, pulse, and blood pressure), body weight, laboratory tests, urine and plasma osmolalities, and the plasma sodium concentration are essential.

A recommended protocol includes the following steps:

●The test is performed after breakfast. It is started after the child voids or, in infants, after the first spontaneous void after the morning feed. Body weight and plasma sodium and osmolality are measured after the patient voids. No further fluid is given until the test is terminated.

●Record each urine void and measure the urine volume and urine osmolality. Some expert centers also measure urine specific gravity.

●Weight and vital signs are obtained every two hours for the first four hours and then hourly. The plasma sodium and osmolality are measured at four hours and then every two hours until the conclusion of the test.

●The test is terminated when one of the following end points are attained:

•Urine osmolality is ≥600 mosmol/kg. A urine specific gravity ≥1.020 can also indicate that urine osmolality greater than this level.

•Plasma osmolality exceeds 295 or 300 mosmol/kg or plasma sodium is 145 mEq/L or higher.

•The patient has lost 5 percent of body weight or exhibits signs of volume depletion.

•If the period of water restriction reaches six hours in infants less than six months of age, eight hours in children from six months to two years of age, or 12 hours in children older than two years of age.

●At the end of the test, weight, vital signs, plasma sodium, plasma osmolality, and urine osmolality should be measured. A specimen should also be obtained for measurement of plasma ADH, which is always elevated during short dehydration tests in patients with hereditary nephrogenic DI.

Children who continue to have impaired urinary concentration (ie, urine osmolality <600 mosmol/kg) despite reaching a plasma osmolality of 295 mosmol/kg or a plasma sodium of 145 mEq/L, children who lose 5 percent or more of their body weight, and if the test is terminated due to time (6 to 12 hours, depending upon the age of the child) desmopressin is administered (5 to 10 mcg by nasal insufflation or 2 to 4 mcg intravenously or subcutaneously). The urine volume and osmolality are then measured every 30 minutes for two hours to detect any antidiuretic response. (See 'Judging the response to desmopressin' below.)

We no longer use aqueous vasopressin, which, due to its vasoconstrictive effect mediated by the V1a receptor, produces sudden and noticeable pallor that raises concerns with the caregiver.

Judging the response to desmopressin — Accurate interpretation of the water restriction test usually requires that desmopressin not be given before the serum sodium is >145 mEq/L and the plasma osmolality has reached 295 mosmol/kg. Below this level, maximum endogenous ADH effect may not be present and an antidiuretic response to exogenous ADH is of no diagnostic benefit, since it will raise the urine osmolality even in normal subjects.

After desmopressin administration, the urine osmolality and volume should be measured every 30 minutes over the next two hours. The two-hour monitoring period is particularly important if there is dilatation of the urinary bladder by previous high urine volumes. In this setting, any concentrated new urine might be diluted with post-micturitional residual urine (which could be as much as 200 to 400 mL).

Each of the causes of polyuria produces a distinctive pattern to water restriction and desmopressin administration [1,24,25,28]:

●Central DI is usually partial, and therefore both ADH release and the urine osmolality may increase as the plasma osmolality rises but submaximally. Desmopressin will more than double the urine osmolality (and generate an equivalent fall in urine output) in complete central DI and increase the osmolality to a lesser degree in partial central DI [24,25].

However, patients with partial central DI may be hyperresponsive to the submaximal rise in ADH induced by water restriction, perhaps due to receptor upregulation. As a result, they may be polyuric at the normal plasma osmolality of 285 to 290 mosmol/kg when ADH levels are very low but have a maximally concentrated urine at a plasma osmolality above 295 mosmol/kg when ADH levels are somewhat higher. In this setting, desmopressin will be without effect, resulting in a pattern suggestive of primary polydipsia [25]. The history may provide important clues in this setting, with abrupt onset favoring central DI and gradual onset, particularly with a history of psychiatric illness, favoring primary polydipsia. (See 'Clues from the history' below.)

●Nephrogenic DI is also associated with a submaximal rise in urine osmolality in response to water restriction. The elevation in plasma osmolality stimulates ADH release, which, since most patients with acquired nephrogenic DI are partially (not completely) resistant to ADH, may induce a modest increase in urine osmolality. The administration of exogenous desmopressin (which increases plasma levels 5- to 10-fold) produces [24,25]:

•No or minimal elevation (ie, <15 percent) in urine osmolality in complete nephrogenic DI.

•A small (up to 45 percent) elevation in urine osmolality in partial nephrogenic DI.

Although the directional change in partial nephrogenic DI is similar to that seen with partial central DI, the absolute numbers are quite different. Patients with the central DI usually achieve a urine osmolality of 300 mosmol/kg or higher after desmopressin, while patients with symptomatic acquired nephrogenic DI typically have a persistently dilute urine with an osmolality that rises, but remains well below 300 mosmol/kg, after desmopressin [24,25]. The history may also be helpful in distinguishing between these disorders. (See 'Clues from the history' below.)

●Primary polydipsia will be associated with a rise in urine osmolality as the plasma osmolality is increased, usually to above 500 mosmol/kg, and no response to desmopressin, since endogenous release is intact. Maximum concentrating ability is frequently impaired in this disorder, resulting in a maximum urine osmolality that may reach 500 to 600 mosmol/kg, as compared with 800 mosmol/kg or more in normal subjects. This acquired defect appears to be due to two effects of chronic polydipsia and polyuria: partial wash out the medullary interstitial gradient and downregulation of ADH release [29].

A properly performed test in which ADH is not given until the plasma osmolality exceeds 295 mosmol/kg (and the serum sodium exceeds 145 mEq/L) will usually establish the correct diagnosis.

However, distinguishing between partial central DI and primary polydipsia may be difficult even when the water restriction test is performed properly. If the diagnosis remains uncertain, we evaluate the response to a trial of desmopressin therapy. (See 'If water restriction is nondiagnostic' below.)

If water restriction is nondiagnostic — The water restriction test may fail to distinguish patients with primary polydipsia from those with partial central DI.

If the water restriction test is nondiagnostic, we perform a therapeutic trial of desmopressin, as follows. For three days, desmopressin (10 mcg intranasally) is given in the evening and patients are advised to restrict their fluid intake to less than 1.5 to 2 L/day. Assessments of symptoms (thirst and polyuria), urine osmolality, and serum sodium (obtained as "stat") are made twice daily. Careful monitoring during the trial of desmopressin is essential to promptly detect the development of hyponatremia. Thus, this trial should be performed in an outpatient setting only if the patient can reliably report symptoms and adhere to frequent monitoring. If there are any concerns that the patient cannot be carefully monitored, the trial of desmopressin should be performed as an inpatient.

The trial of desmopressin is interpreted as follows:

●In patients with partial central DI, the polyuria and polydipsia will reverse, the urine will become concentrated and, if the patient was adherent to fluid restriction, the serum sodium should remain normal (although mild hyponatremia may develop, depending upon the level of fluid restriction). (See "Treatment of central diabetes insipidus".)

●In patients with primary polydipsia, desmopressin will not relieve the thirst, although the urine will be concentrated and, because of persistent water intake, the patient can become hyponatremic.

An alternative to performing a therapeutic trial of desmopressin is to measure plasma copeptin levels after infusing hypertonic saline to stimulate ADH release. However, copeptin assays are not yet commercially available in many regions, including in the United States.

Copeptin is the C-terminal glycoprotein moiety of the AVP prohormone and therefore a surrogate marker of ADH [17,30]. To distinguish central DI from primary polydipsia, hypertonic saline is infused to raise the serum sodium to >145 mEq/L, at which point plasma copeptin is measured [18]. If reliable assays are available, the diagnosis is made as follows:

●Central DI is diagnosed if the plasma copeptin is ≤4.9 pmol/L.

●Primary polydipsia is diagnosed if the plasma copeptin is >4.9 pmol/L.

The study that reported this threshold copeptin level of 4.9 pmol/L infused hypertonic saline to raise the serum sodium concentration to >150 rather than >145 mEq/L [18]. However, the UpToDate authors and editors of this topic are concerned about the safety of raising the serum sodium concentration to this level and instead suggest that the target serum sodium during hypertonic saline infusion should be >145 mEq/L. If the copeptin level is close to 4.9 pmol/L, there will still be some uncertainty about the diagnosis. If this occurs, the test can either be repeated (with a target serum sodium >150 mEq/L) or a carefully supervised therapeutic trial of desmopressin can be used to distinguish between primary polydipsia and partial central DI.

If plasma copeptin levels are unavailable, vasopressin (ADH) levels should be measured on plasma samples collected both at baseline and after water deprivation or hypertonic saline infusion (but prior to the administration of desmopressin) [25,31]:

●If there is an increase in plasma ADH in response to the rising plasma osmolality, central DI is excluded.

●If there is an appropriate elevation in urine osmolality as ADH secretion is increased, nephrogenic DI is excluded.

However, measurement of ADH concentrations may be misleading for several reasons [32]:

●The commercially available assays for both plasma and urine ADH are not very sensitive, especially in the range below 3 pg/mL.

●ADH in the circulation is largely bound to platelets, and, without special handling of the specimen, falsely high and falsely low levels can be seen.

●ADH is unstable in isolated plasma, even when stored at -20°C so that send-out assays may be unreliable. Even highly sensitive assays may be misleading in patients with primary polydipsia since chronic overhydration can cause partial suppression of ADH release, mimicking the pattern in central DI [29].

Urinary assays have been suggested since the ADH concentration is higher in urine and the plasma separation issues may not apply. However, there are few data to support this approach.

Patients with hypernatremia

Differentiating solute diuresis from diabetes insipidus — Hypernatremic patients with polyuria have either DI, a solute diuresis, or, in some patients, both disorders. Primary polydipsia is not a consideration, because such patients will be normonatremic or hyponatremic.

Differentiating solute diuresis from DI in hypernatremic patients with polyuria depends upon the urine osmolality:

●Hypernatremic patients with a urine osmolality >600 mosmol/kg have a solute diuresis. However, such patients may also have a component of DI.

If infusion of dilute fluids to treat the hypernatremia does not produce a substantial decrease in the urine osmolality, then an osmotic diuresis is likely to be solely responsible for both the polyuria and hypernatremia. (See "Etiology and evaluation of hypernatremia in adults", section on 'Osmotic diuresis'.)

However, if infusion of dilute fluids to correct the hypernatremia results in a decline in urine osmolality to <600 mosmol/kg before the serum sodium concentration falls below 145 mEq/L, then a component of DI in addition to a solute diuresis should be suspected. If the urine osmolality declines to <300 mosmol/kg, before the serum sodium concentration falls below 145 mEq/L, then a diagnosis of DI can be made. In such patients, the history and the response to exogenous desmopressin can help classify the type of DI. (See 'Classification of diabetes insipidus in hypernatremic patients' below and 'Clues from the history' below.)

●Hypernatremic patients with an intermediate urine osmolality of 300 to 600 mosmol/kg may have a component of both DI and a solute diuresis. If the total daily osmolar output is >1000 mosmol (calculated as the urine osmolality x the 24-hour urine volume), then a solute diuresis is present. Patients with total daily osmolar outputs <900 mosmol have DI. In such patients, the history and the response to exogenous desmopressin can help classify the type of DI. (See 'Classification of diabetes insipidus in hypernatremic patients' below and 'Clues from the history' below.)

●Hypernatremic patients with a low urine osmolality of less than 300 mosmol/kg have DI. The history and the response to exogenous desmopressin can help classify the type of DI. (See 'Classification of diabetes insipidus in hypernatremic patients' below and 'Clues from the history' below.)

Classification of diabetes insipidus in hypernatremic patients — In patients with DI, no cognitive impairment, and access to fluids, true hypernatremia should not occur, because the initial loss of water stimulates thirst, resulting in an increase in intake to match the urinary losses (figure 2). Thus, hypernatremia is seen primarily in patients who cannot experience thirst normally (eg, patients with impaired mental status, critical illness, or adipsia) or who cannot respond to thirst normally (eg, patients with physical constraints, young children who require others to provide fluid intake). (See "Etiology and evaluation of hypernatremia in adults", section on 'The importance of thirst'.)

The diagnostic approach to hypernatremic patients with DI is as follows:

●Adults with a urine osmolality <300 mosmol/kg who are cognitively intact, have unlimited access to fluids, but who lack thirst have adipsic DI.

Adipsic DI is usually due to a central lesion that impairs both thirst and release of ADH, thereby causing a water diuresis and hypodipsia or adipsia; in this setting, the plasma sodium concentration can exceed 160 mEq/L [33]. Adipsic DI is associated with significant morbidity including obesity, sleep apnea, venous thrombosis during episodes of hypernatremia, thermoregulatory dysfunction, seizures, and significant mortality [34]. Adipsic DI is discussed in detail elsewhere. (See "Etiology and evaluation of hypernatremia in adults", section on 'Adipsic diabetes insipidus'.)

●Polyuric infants with hypernatremia and urine osmolality <300 mosmol/kg typically have hereditary nephrogenic DI. Hypernatremia during the first year of life is a common feature in such patients. The intense and constant thirst of these young children is often not understood by adults, particularly if this is the first affected family member due to a de novo mutation or if an X-linked mutation has passed unrecognized through asymptomatic females [35].

●In other patients, we generally give desmopressin (10 mcg by nasal insufflation or 2 to 4 mcg subcutaneously or intravenously) and judge the response by monitoring the urine osmolality every 30 minutes over the next two hours:

•Complete nephrogenic DI – No elevation in urine osmolality and the urine osmolality is <300 mosmol/kg

•Partial nephrogenic DI – A small (up to 45 percent) elevation in urine osmolality to a level that remains <300 mosmol/kg

•Complete central DI – A rise in urine osmolality of more than 100 percent

•Partial central DI – An increase in urine osmolality of 15 to 50 percent to a level >300 mosmol/kg

Patients with hyponatremia — A low plasma sodium concentration in a polyuric patient is almost always indicative of water overload due to primary polydipsia. Such patients will have a low urine osmolality (eg, less than one-half the plasma osmolality).

If the patient has hyponatremia and a high, rather than low, urine osmolality, then an osmotically active substance, such as glucose or mannitol, is present causing both hyperosmolar hyponatremia and osmotic diuresis. Hyperglycemia is simple to diagnose and should be obvious, and infusion of mannitol should be apparent from the history. (See 'Hyperglycemia or another obvious osmotic diuresis' above.)

Clues from the history

Rapidity of onset of polyuria — The patient (or their parent) should be questioned about the rate of onset of the polyuria. In the majority of cases of hereditary nephrogenic DI, severe polyuria (with risk of dehydration and hypernatremia) manifests during the first week of life. In familial central DI (usually an autosomal dominant disease), polyuria may present after the first year of life, sometimes in young adulthood, due to preservation of function of the normal allele [26]. (See "Clinical manifestations and causes of nephrogenic diabetes insipidus" and "Clinical manifestations and causes of central diabetes insipidus".)

In adults, the onset is usually abrupt in central DI ("I suddenly began urinating too much a few days ago") and almost always gradual in acquired nephrogenic DI or primary polydipsia.

The new onset of nocturia in the absence of other causes of nocturia (eg, prostatic enlargement in men over 50 years of age or urinary tract infection in children) is often a first clue to DI. The urine is normally most concentrated in the morning due to lack of fluid ingestion overnight; as a result, the first manifestation of a loss of concentrating ability is often nocturia.

Family history of polyuria or diabetes insipidus — There are familial forms of both central and nephrogenic DI. The defects in these disorders are due to mutations that impair ADH synthesis or the renal response to ADH; the latter defect is most often due to mutations in the AVPR2 gene encoding the V2 receptor but can also result from mutations in the aquaporin-2 (water channel) gene [36]. A family history of polyuria is helpful both for diagnosis and to identify asymptomatic members of affected families who harbor the suspect allele. Thus, all families with hereditary DI should have their molecular defect identified. (See "Clinical manifestations and causes of nephrogenic diabetes insipidus" and "Clinical manifestations and causes of central diabetes insipidus".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Fluid and electrolyte disorders in adults".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Diabetes insipidus (The Basics)")

SUMMARY AND RECOMMENDATIONS

Definition and differential diagnosis

●Polyuria has generally been defined as a urine output exceeding 3 L/day in adults and 2 L/m² in children. It must be differentiated from the more common complaints of frequency or nocturia, which may not be associated with an increase in the total urine output. (See 'Introduction and definition' above and 'Confirm the presence of polyuria' above.)

●Polyuria may be caused by a solute (osmotic) diuresis or a water diuresis (table 1). Major causes of polyuria due to a solute diuresis include glucosuria (eg, in patients with uncontrolled diabetes mellitus), urea diuresis (eg, in patients with resolving acute kidney injury, and sodium diuresis (eg, in patients given larger volumes of intravenous saline). (See 'Solute (osmotic) diuresis' above.)

●Primary polydipsia, central diabetes insipidus (DI), and nephrogenic DI are associated with an increase in water output and the excretion of a relatively dilute urine (ie, water diuresis) (table 1). In primary polydipsia, the polyuria is an appropriate response to enhanced water intake, whereas the water loss is inappropriate with either form of DI. (See 'Water diuresis' above.)

Evaluation

●In some patients, the history and initial laboratory testing may clearly point to a specific osmotic diuresis, often making further evaluation unnecessary (eg, hyperglycemia and glucosuria). If unsure, the presence of an osmotic diuresis can be confirmed by a urine osmolality greater than 400 to 600 mosmol/kg and a total daily osmolar output >1000 mosmol. (See 'Hyperglycemia or another obvious osmotic diuresis' above.)

●If the etiology is unclear, the evaluation depends in part upon whether or not the patient has a normal serum sodium, hypernatremia, or hyponatremia (algorithm 1). (See 'When the cause is not obvious' above.)

•Normonatremic patients may have a solute diuresis, primary polydipsia, or DI:

-Normonatremic patients with a urine osmolality >600 mosmol/kg have a solute (osmotic) diuresis. (See 'Solute (osmotic) diuresis' above.)

-In normonatremic patients with an intermediate urine osmolality of 300 to 600 mosmol/kg, the total daily osmolar output should be calculated; if >1000 mosmol, then a solute diuresis is present. However, if the total daily osmolar output is <900 mosmol (a normal value), then the patient has a water diuresis due either to primary polydipsia or DI. (See 'Urine osmolality 300 to 600 mosmol/kg' above.)

-Normonatremic patients with a low urine osmolality of <300 mosmol/kg have a water diuresis due either to primary polydipsia or DI. (See 'Urine osmolality <300 mosmol/kg' above.)

•Hypernatremic patients with a urine osmolality >600 mosmol/kg have a solute diuresis but may also have DI. If infusion of dilute fluids to correct the hypernatremia results in a decline in urine osmolality to <600 mosmol/kg before the serum sodium concentration falls below 145 mEq/L, then a component of DI in addition to a solute diuresis should be suspected; if the urine osmolality declines to <300 mosmol/kg before the serum sodium concentration falls below 145 mEq/L, then a diagnosis of DI can be made. Hypernatremic patients with a urine osmolality of 300 to 600 mosmol/kg should have the total daily osmolar output calculated to differentiate solute diuresis from DI (DI is the diagnosis if the total daily osmolar output is <900 mosmol). Hypernatremic patients with a urine osmolality <300 mosmol/kg have DI. (See 'Patients with hypernatremia' above.)

•Hyponatremia in a polyuric patient is almost always indicative of water overload due to primary polydipsia. However, if the urine osmolality is high, then an osmotically active substance, such as glucose, is present causing both hyperosmolar hyponatremia and osmotic diuresis. (See 'Patients with hyponatremia' above.)

●In normonatremic patients with a water diuresis of unclear etiology, differentiating primary polydipsia from DI requires further evaluation with water restriction (and/or hypertonic saline infusion) to achieve a goal plasma sodium concentration >145 mEq/L (algorithm 2); this will stimulate vasopressin sufficiently to produce a maximally concentrated urine. The protocol for reaching this goal depends upon the age and reliability of the patient. (See 'Goal of water restriction (or hypertonic saline)' above and 'Protocol in adults and adolescents' above and 'Protocol in infants and children' above.)

After a plasma sodium concentration >145 mEq/L has been achieved, the subsequent evaluation depends upon the urine osmolality (see 'Protocol in adults and adolescents' above and 'Protocol in infants and children' above and 'Judging the response to desmopressin' above):

•If the urine osmolality has reached a clearly normal value (above approximately 700 mosmol/kg), the diagnosis is primary polydipsia.

•If the urine osmolality is <700 mosmol/kg, we administer desmopressin (or, less commonly, aqueous vasopressin) and judge the response. Determination of plasma copeptin or vasopressin levels is an alternative to desmopressin administration. However, assays for copeptin and vasopressin are infrequently available.

●In hypernatremic patients with DI, water restriction is unnecessary. Hypernatremic adults with a urine osmolality <300 mosmol/kg who are cognitively intact, have unlimited access to fluids, but lack thirst have adipsic DI. Polyuric infants with hypernatremia and urine osmolality <300 mosmol/kg typically have hereditary nephrogenic DI. In other patients, we give exogenous antidiuretic hormone (ADH; typically desmopressin) and judge the response in order to distinguish between central and nephrogenic DI. (See 'Classification of diabetes insipidus in hypernatremic patients' above.)

●Judging the response to desmopressin helps to differentiate primary polydipsia from DI (in normonatremic patients) and to distinguish central from nephrogenic DI (in normonatremic or hypernatremic patients) (see 'Judging the response to desmopressin' above and 'Classification of diabetes insipidus in hypernatremic patients' above):

•Desmopressin will more than double the urine osmolality (and generate an equivalent fall in urine output) in complete central DI and increase the osmolality to a lesser degree in partial central DI.

•Desmopressin produces no or minimal elevation (ie, <15 percent) in urine osmolality in complete nephrogenic DI. A small (up to 45 percent) elevation in urine osmolality occurs in partial nephrogenic DI. Although the directional change in partial nephrogenic DI is similar to that seen with partial central DI, the absolute numbers are quite different. Patients with the central DI usually achieve a urine osmolality of 300 mosmol/kg or higher after desmopressin, while patients with symptomatic acquired nephrogenic DI typically have a persistently dilute urine with an osmolality that rises, but remains well below 300 mosmol/kg, after desmopressin.

•Primary polydipsia will be associated with no response to desmopressin, since endogenous release is intact. Maximum concentrating ability is frequently impaired in this disorder, resulting in a maximum urine osmolality that may reach 500 to 600 mosmol/kg, as compared with 800 mosmol/kg or more in normal subjects. This acquired defect appears to be due to two effects of chronic polydipsia and polyuria: partial wash out of the medullary interstitial gradient and downregulation of ADH release [29].

However, distinguishing between partial central DI and primary polydipsia may be difficult even when the test is performed properly. If the diagnosis remains uncertain (see 'If water restriction is nondiagnostic' above):

-We perform a three-day therapeutic trial of desmopressin (given in the evening) with assessments of symptoms (thirst and polyuria), urine osmolality, and serum sodium made twice daily. In patients with partial central DI, the polyuria and polydipsia will reverse, the urine will become concentrated, and the serum sodium should remain normal (although mild hyponatremia may develop if the patient does not restrict fluids). In patients with primary polydipsia, desmopressin will not relieve the thirst, although the urine will be concentrated and, because of persistent water intake, the patient can become hyponatremic.

-An alternative to performing a therapeutic trial of desmopressin is to measure plasma copeptin levels (if available) after a plasma sodium >145 mEq/L is attained (with water restriction or hypertonic saline) (algorithm 3). Copeptin is a surrogate marker of ADH. Partial central DI is diagnosed if the plasma copeptin is ≤4.9 pmol/L. Primary polydipsia is diagnosed if the plasma copeptin is >4.9 pmol/L.