OSMOLALITY

Osmolality is the measure of osmoles of solute per kilogram of solvent and includes both permeant (e.g. urea) and impermeant solutes (e.g. sodium) and thus osmolality may change due to both water movement and the movement of permeant solute. Osmolarity is the measure of osmoles per litre of solute and thus is temperature-dependent. Osmolality is temperature-independent. Normal plasma osmolality is in the range of 275–295 mosmol/kg. Plasma osmolality can be estimated from several equations.¹ The formula of Worthley et al.² below is both simple and correlates well with measured values.³

All units of solute are mmol/l.

Where a patient is markedly uraemic, a value of 8 mmol/l is substituted for the actual urea. Actual osmolality may differ markedly from estimated osmolality in the presence of unmeasured, osmotically active solutes such as mannitol, ethanol, bicarbonate, lactate and amino acids. Whenever concern exists that a patient may be hyper- or hypo-osmolal, osmolality should be measured by assessing the freezing-point depression of the plasma or urine. Increasing the osmolality results in depression of the freezing point.

The osmolal gap is the difference between the calculated and measured osmolality in plasma or urine.The normal osmolal gap is < 10 mosmol/l. An increased osmolal gap indicates the presence of an unmeasured osmotically active solute.

TONICITY

Tonicity describes the ability of a solution of impermeant solute (such as sodium) to cause the movement of water between itself and another fluid compartment. It may be considered an index of the water concentration rather than the solute concentration as the solute is impermeant. The tonicity of plasma is largely determined by its sodium content, the main solute in extracellular fluid, which is not freely permeable to cross into the intracellular space. This characteristic facilitates the control of extracellular fluid volume through the regulation of sodium balance.

It is possible for a solution to be both hypotonic and iso-osmolar. Dextrose solution 5% is an example of this; the solution contains no impermeant solute (assuming no absence of insulin, glucose freely enters the cell), but is iso-osmolar with the intracellular milieu.

SOLUTE AND WATER INTAKE AND LOSSES

In order to understand disorders of osmolality, tonicity, fluid balance and urine output, it is necessary to be aware of the essential physiological mechanisms which maintain the fluid and osmolal states of the body in health. Assuming a normal diet, in a 75-kg man, every day there is an obligatory loss of around 800 mmol of solute – approximately 300 mmol of urea and 500 mmol of cations and anions. The maximum concentrating ability of the healthy kidney is around 1200 mosmol/kg; consequently, a minimum of 666 ml of urine a day is required to excrete osmotically active solutes. Additionally, insensible losses of water (respiratory water, faecal water and sweat) approximate to 10 ml/kg per day and this rises markedly with fever and hot dry climates. Thus obligatory water losses of around 1.5 l/day arise due to insensible losses and obligatory solute excretion.

NORMAL URINARY OSMOLALITY

In health, urinary osmolality is usually maintained between 500 and 700 mosmol/kg. As the obligatory solute load to be excreted is relatively constant in value, urine osmolality will fall in response to an increased intake in free water and rise in response to dehydration or water restriction (Figure 51.1). The minimum osmolality of urine achievable inhumans is around 25 mosmol/kg. Diuresis refers to the passage of a high volume of urine (> 1.5 ml/kg per hour) and may be transient or persistent, physiological or pathological. The production of > 3 l/day is arbitrarily defined as polyuria. Water diuresis occurs when the total solute in the urine excreted per day is within the normal range but the osmolality of the urine passed is low. An osmotic or solute diuresis occurs when the total solute passed per day is higher than normal, and the urine passed is usually iso-osmolar to plasma if the extracellular fluid volume is expanded or hyperosmolar if the patient is hypo- or euvolaemic.

Figure 51.1 Urine output rises and urine osmolality falls in health as a function of increasing water intake. Assuming normal solute ingestion, normal solute excretion (around 800 mmol/day) is preserved between water intakes of 1.5 and 32 litres/day.

PLASMA OSMOLALITY AND PLASMA VOLUME REGULATION

Changes in plasma osmolality and changes in plasma volume can occur in tandem or independently of one another. Although normal plasma osmolality lies in a population range of 275–295 mosmol/kg (265–285 mosmol/kg in pregnancy), individuals tend to vary less than ± 1% around their set value. In pregnancy this set value falls but the limited variability around it does not. The body regulates osmolality and volume by separate mechanisms. Water balance and osmolality are maintained via osmoreceptors which mediate their control via control of thirst and ADH production. Control of the plasma volume is maintained via volume receptors and sodium receptors, which mediate their actions through the sympathetic nervous system, the renin–angiotensin–aldosterone system and atrial natriuretic peptide (ANP). Additionally, the volume receptors have inputs to the hypothalamus via which they too can mediate ADH release and the sensation of thirst.

THIRST

In health fluid intake is determined by the sensation of thirst and the subsequent ingestion of fluid; a plasma osmolality of > 290 mosmol, elevated angiotensin II concentration, sympathetic nervous system activation and circulating volume depletion of 5–10% are all associated with the onset of thirst. Fluid and solute excretion are largely regulated through the kidney, although some solutes such as ethanol and glucose are largely cleared through metabolism rather than excretion.

In the intensive care unit (ICU), the patient loses control over intake of fluids and solute and frequently has impaired excretory mechanisms. Thus both volume and solute homeostasis may become heavily dependent upon the skills of attending clinicians.

OSMORECEPTORS AND OTHER INPUTS TO THE SUPRAOPTIC AND PARAVENTRICULAR NUCLEI

Detection of osmolality occurs largely at osmo- (Na⁺) receptors sited around the anterior aspect of the third ventricle of the brain. These are sensitive to plasma osmolality and cerebrospinal fluid sodium concentration. Hypertonic saline is a more potent stimulus than equi-isotonic equiososmolar solutions of other solutes.⁴ These osmoreceptors link to the cells of the paraventricular nuclei (PVN) and supraoptic nuclei (SON), the sites of ADH synthesis. The axons of the cells in the PVN and SON form part of the pituitary stalk linking the hypothalamus to the pituitary gland in which they terminate. A smaller proportion of the axons terminate in the median eminence where they release ADH and oxytocin, which is transported to the anterior lobe of the pituitary by portal vessels. The ADH and oxytocin so released cause release of adrenocorticotrophic hormone (ACTH) and prolactin respectively; the ADH acts synergistically with corticotrophin-releasing factor (CRF) but is also believed to have ACTH secretagogue properties in its own right.^5,6

Direct inputs from the sympathetic nervous system to the PVN and SON can stimulate ADH release via α-adrenoreceptors. Other central osmoreceptors lie outside the blood–brain barrier in the subfornical organ and come into contact with plasma. It is believed that ANP and angiotensin II⁷ act via these receptors to inhibit or elicit ADH synthesis and ADH release and to modify the sensation of thirst. Additional osmoreceptors in the mouth, stomach and liver are believed to play a role in the anticipation of an osmolal load following ingestion of food and can pre-emptively stimulate ADH synthesis in the hypothalamus.

As the baroreceptor and osmoreceptor inputs to the PVN and SON are distinct, it is possible to lose the normal ADH response to hyperosmolality but maintain a normal ADH response to hypovolaemia.⁸ Additionally, in animal experiments, when hypotension increases the basal plasma ADH concentration, there is a simultaneous resetting of the osmomolality–plasma ADH response curve in an attempt to preserve osmoregulatory function from the new higher baseline.⁹ If this did not occur, the ADH response to hypotension would always result in the development of a hypo-osmolal state in addition to causing vasoconstriction.

The normal response of osmoreceptors to changing plasma osmolality in terms of ADH is illustrated in Figure 51.2. At plasma osmolalities of < 275 mosmol/kg, the osmoreceptors remain hyperpolarised and virtually no ADH release occurs via them. At osmolalities > 295 mosmol/kg, the osmoreceptors are maximally depolarised and plasma concentrations of ADH of > 5 pg/ml are attained. Other inputs and influences upon ADH release are summarised in Figure 51.3 and Table 51.2.

Figure 51.2 Plasma antidiuretic hormone (ADH) and plasma osmolality in health and partial cranial diabetes insipidus.

Figure 51.3 A schematic representation of anatomical and physiological connections of the supraoptic and paraventricular nuclei (SON and PVN), the principal sites of antidiuretic hormone (ADH) synthesis within the hypothalamus. Blue arrows represent the transport of ADH or its precursor between anatomically distinct sites. Baroreceptor inputs into these nuclei travel via the vagus nerve to reach the central nervous system. Whilst the majority of ADH produced in the SON and PVN is transported to the posterior pituitary (PP) for storage, some is transported to the median eminence (ME) then onwards to the anterior pituitary (AP), where it acts synergistically as a secretagogue for adrenocorticotropic hormone (ACTH). APo, area postrema; SFO, subfornical organ; CRF, corticotrophin-releasing factor.

Table 51.2 Factors influencing antidiuretic hormone (ADH) release

ANTIDIURETIC HORMONE/ARGININE VASOPRESSIN (AVP)

ADH (8-arginine vasopressin) is a nine-amino-acid peptide which differs from oxytocin at only two residues but shares the disulphide bond between the first and sixth ones. This similar structure and conformation results in some cross-reactivity at receptors and in function.¹⁰ It is synthesised in the SON and PVN, bound to neuorophysin, transferred through axons to the posterior pituitary gland and stored in granules prior to release. Synthesis to replace any released stores is a rapid process (1–2 hours from synthesis to storage) and patients with damage to the pituitary can achieve near-normal plasma concentrations of ADH, in terms of osmoregulatory function, via release of newly synthesised ADH via the axons terminating in the median eminence. However, the higher plasma concentrations associated with hypovolaemia cannot be achieved. Normal osmoregulatory plasma ADH concentrations are in the range of 1–8 pg/ml but rise as high as 40 pg/ml in hypovolaemic patients under the influence of the sympathetic nervous system, baroreceptor responses and angiotensin II.

Once released from the pituitary, ADH has a plasma half-life of around 10–35 minutes.¹¹ It is metabolised by hepatic and renal vasopressinases and around 10% of the active hormone is excreted unchanged in the urine.

VOLUME RECEPTORS

Volume homeostasis takes precedence over sodium homeostasis and so rises and falls in sodium will occur in order to preserve the circulating volume. In euvolaemic patients sodium homeostasis is maintained. Sodium concentration is detected by both the osmoreceptors of the subfornical organ outside the blood–brain barrier and also by the juxtaglomerular apparatus which secretes renin in response to reduced GFR and a lower sodium load in the tubule.²³

The predominant determinants of sodium balance, however, are the high-pressure baroreceptors in the pulmonary veins, left atrium, carotid sinus and aortic arch.²⁴ Reduced stretch of these receptors increases sympathetic nervous system activity and activation of the renin–angiotensin–aldosterone system, resulting in reduced sodium excretion (via reduced GFR) and increased reabsorption of sodium in the proximal and distal convoluted tubules. Additionally, release of ADH can be stimulated, resulting in concomitant water retention. Conversely, stretch of the baroreceptors will result in a fall in sodium retention through reduced activity of the sympathetic nervous and renin–angiotensin–aldosterone systems. Stretch additionally results in the release of ANP and a natriuresis through reduced sodium reabsorption in the distal convoluted tubule and collecting duct. ADH release is reduced by the fall in sympathetic nervous tone from the baroreceptors. ADH secretion may also be inhibited by the action of ANP on cerebral osmoreceptors lying outside the blood–brain barrier.²⁵

The role of low-pressure baroreceptors in the systemic venous circulation and right atrium is less clearly defined. When venodilatation occurs, as is seen in sepsis, or when there is a reduction in cardiac output, reduced baroreceptor signalling in the high-pressure system will result in both sodium and water retention, as outlined previously. This will expand the extracellular fluid compartment and potentially cause tissue oedema. As sepsis resolves, venous tone is restored, capillary leak reduces, an increase in the loading of the high-pressure baroreceptors results and a natriuresis takes place. Patients may become transiently polyuric as they clear the excess salt and water accumulated whilst shocked. During this physiological diuresis, plasma osmolality remains tightly within the normal range, provided that renal concentrating mechanisms have not been injured during the septic episode or by drug administration.

CONGENTIAL CDI

Congenital CDI is rare and usually inherited as an autosomal-dominant characteristic which results from mutations of the gene encoding an ADH precursor – preprovasopressin neurophysin II. The onset of the disease may occur anywhere between 1 year of age and middle adult life and is associated with the final destruction of the ADH-producing cells due to an accumulation of the abnormal ADH precursor. Until the destruction of the SON and PVN cells occurs, ADH secretion (facilitated by expression of the normal gene) and regulation of plasma osmolality are often unaffected.

ACQUIRED CDI

Acquired CDI may be transient or permanent and can arise due to an absolute (complete) or relative (incomplete) lack of ADH. Complete central DI is usually associated with lesions above the level of the median eminence in the SON or PVN or of the neurohypophyseal stalk whereby the production of ADH in the hypothalamus is terminated.²⁶ Permanent central DI tends to be associated with transecting, obliterating or chronic inflammatory lesions, whereas transient DI is more likely to be associated with acute inflammatory or oedematous lesions with some recovery of ADH secretion occurring as the inflammation or oedema resolves. An exception to this is the transient DI seen following excision or destruction of the posterior pituitary; ADH produced in the hypothalamus can still be released into the systemic circulation from capillaries in the median eminence.

In the past the majority of acquired non-traumatic CDI was categorised as idiopathic but it has become apparent that the majority of these cases are associated with abnormality of the inferior hypophyseal arterial system²⁷ or autoimmune reactivity against ADH-producing cells.²⁸ These findings may indicate causality or association.

When the normal release of ADH into the circulation in response to rising plasma osomolality is reduced or absent, inappropriately high urine volumes are passed and the urine osmolality becomes inappropriately low for the state of water depletion being suffered by the patient. Where ADH is entirely absent from the circulation, over 20 litres of very dilute urine (25–200 mosomol/kg) per day may be produced. If patients are unable to drink freely (most ICU patients) or their thirst mechanisms are impaired, profound dehydration will result very rapidly unless appropriate interventions are made by the physician.

Where ADH deficiency is relative rather than absolute, it is possible for the patient to concentrate the urine partially and values of 500–800 mosmol/kg would not be atypical. However, these osmolalities are inappropriately low relative to the plasma osmolality. In partial ADH deficiency volumes of urine as low as 3 l/day may be evidenced. These are still inappropriately high when assessed in terms of the solute excretion of the patient but are more difficult to recognise as being due to DI as there are many other causes of diureses of this magnitude. Additionally, extrinsic stimulants of ADH release (see Table 51.2) may have an antidiuretic effect, further complicating the diagnosis.

The plasma osmolality measured in central DI is usually in the higher regions of the normal range or very slightly supranormal. It is remarkably constant in those with free access to water and intact thirst mechanisms as they will drink huge quantities of water to regulate and maintain their water balance. Hyperosmolality or hypernatraemia suggest impaired sensation of thirst or inability to access water (see water deprivation test later) and can also be seen if patients are administered large quantities of isotonic saline or Hartmann’s solution to replace their hypotonic urine losses. If unrecognised and untreated, hyperosmolality and hypernatraemia may result in death.

CDI is usually associated with reduced production of ADH or damage to the normal release mechanisms of ADH. However, there can be dysfunction of the osmolality-sensing mechanism at receptor or intracellular signalling levels whilst actual ADH production and storage are normal. It is possible to have a normal release of ADH in response to baroreceptor detection of hypotension but subnormal release in response to hyperosmolality. This has been described in association with chronic hypernatraemia.²⁹

The main recognised causes of central DI are listed in Table 51.3. A particularly common cause of DI seen in the ICU is traumatic or postsurgical brain injury. Transsphenoidal surgery for treatment of suprasellar tumours can result in DI in 10–70% of patients; the frequency parallels the magnitude of the tumour being removed. Additionally, transcranial surgery may cause the development of DI in the absence of a fall in plasma ADH. This is postulated to be due to the release of a hypothalamic ADH precursor which acts as a competitive antagonist of both ADH and synthetic analogues. The presence of a competitive antagonist effectively creates an endocrinologice picture similar to NDI with normal or high plasma ADH levels but an inappropriate diuresis of dilute urine.

Table 51.3 Causes of cranial diabetes insipidus

* Doczi T, Tarjanyi J, Kiss J. Syndrome of inappropriate antidiuretic syndrome after head injury. Neurosurgery 1982; 10: 685–8.

† Repaske DR, Medlej R, Gulteken EK et al. Heterogeneity in clinical manifestation of autosomal dominant neurohypophyseal diabetes insipidus caused by a mutation encoding Ala¹-Val in the signal peptide of the arginine vasopressin/neurophysin II/copeptin precursor. J Clin Endocrinol Metab 1997; 82: 51–6.

Following surgery or traumatic brain injury, several different patterns of polyuria can be seen: immediate permanent or transient polyuria, initial normal or low urine production followed by transient or permanent polyuria, or initial low followed by normal urine output. Additionally, a classical triphasic pattern of urine output may be observed with:

During the second phase of this pattern, administration of fluids may result in volume overload and hyponatraemia as the ADH release is not under feedback control from osmoreceptors but occurring in an uncontrolled manner as a result of pituitary degeneration. Effectively, there is a transient syndrome of inappropriate ADH (SIADH) secretion. The triphasic pattern is usually associated with sudden severe damage to the hypothalamus or pituitary from trauma, surgery or intracranial bleed, and careful, regular clinical and biochemical assessment is essential to ensure normal water balance and osmolality during this transition from DI to SIADH and back to DI again.

The exact nature of the urinary pattern seen in DI is relatively unimportant and gradual resolution may occur over several months in those with transient DI. What is essential is that meticulous assessments of patients and their plasma and urinary biochemistry as well as fluid inputs and outputs are made to prevent the development of unnecessary fluid and solute imbalances which could lead to worsening morbidity or mortality.

It is important to maintain a high level of suspicion for the development of DI in anyone who is suffering from pituitary disease or who has suffered a pituitary injury as the symptoms may have gradual onset. Anterior pituitary failure can lessen the impact of central DI because of the deficiency of ACTH and cortisol, which reduces GFR and free water loss. Additionally, the loss of feedback inhibition may stimulate increased release of ADH from the median eminence. Thus in Sheehan’s syndrome and pituitary apoplexy, the presenting symptoms tend not to be those of DI with polyuria. However, once corticosteroid therapy is commenced, polyuria indicating DI may become apparent or exacerbated. Conversely, patients with persistent DI of idiopathic origin should have long-term endocrine follow-up as a number will go on to develop tumours of the pituitary several years following the diagnosis of DI.^30,31

TREATMENT OF CDI

Four separate problems have to be addressed when treating CDI:

In all ICU patients with DI, hourly urine measurements, hourly fluid losses and fluid inputs and at least twice-daily urine and plasma osmolalities are recommended. In shocked patients and those with hypernatraemia, hourly monitoring of plasma sodium is recommended to prevent worsening of hyperosmolality or over-rapid correction of hypernatraemia.

CONGENITAL NDI

Some 80–90% of patients with congenital NDI have an X-linked recessive abnormality of the AVPR₂ gene coding for the V₂ receptor. Different mutations of the gene are described but the majority result in trapping of the V₂ receptor intracellularly, unable to integrate into the membrane of the collecting duct cell. Drugs have been developed which can facilitate receptor integration into the membrane, restoring some of the urine-concentrating abilities of ADH.³³ The sex linkage results in the vast majority of affected patients being male. However, female children can also present less severe polyuria and polydipsia due to expression of the abnormal gene.

Non-sex-linked genetic abnormalities can also cause NDI. Approximately 10% of cases of congenital NDI have mutations of the AQP₂ (acquaporin 2) gene which codes for the AQP₂ channel. Over 40 mutations, both autosomal-dominant and recessive, have been described to date.

The remainder of cases of congenital NDI have a variety of pathologies which result in failure to generate a hypertonic renal medulla, with inability to reabsorb water even if the V₂ receptor and acquaporin₂ channels are normal.

A lack of the Kidd antigen (a blood group antigen) results in an inability to concentrate urine to more than 800 mosmol/kg even with water deprivation and exogenous ADH administration. This is because the antigen is also expressed in the collecting duct epithelium where it functions as a urea transporter (urea transport B protein) and facilitates movement of urea from urine into the medullary interstitium maintaining some of the gradients required to facilitate water reabsorption. Similarly, patients with mutations in chloride channel genes, potassium channel genes or the sodium–potassium–chloride co-transporter gene resulting in the Bartter syndrome are unable to generate a hypertonic medullary interstium. However, in these patients the defect is more marked and urine can rarely be concentrated above 350 mosmol/kg.

With congenital NDI, early diagnosis and management are essential as avoidance of hypernatraemia and dehydration facilitates the achievement of normal developmental milestones and avoids the cerebral damage once commonly accepted as an inevitable association of NDI.

ACQUIRED NDI

Lithium-associated nephrotoxicity remains the commonest form of acquired NDI with more than 20% of patients on chronic lithium therapy developing polyuria. Lithium is taken up into the principal cells of the collecting duct via sodium channels and inhibits intracellular adenylate cylcase antagonising the effects of ADH. Additionally, it reduces the medullary interstitial hypertonicity, possibly through reducing expression of urea transport protein B. If patients with early lithium-related NDI are prescribed amiloride, some reversal of both the polyuria and lithium-mediated toxicity is possible. Amiloride’s natriuretic action is achieved through closure of the luminal sodium channels in the collecting duct; the sodium channels through which lithium enters the cells.³⁴ Indomethacin increases intracellular cAMP and counteracts the diminution of this and AQP₂ caused by lithium, resulting in a marked and immediate drop in urine output. However, care is required as non-steroidal anti-inflammatory drugs (NSAIDs) may worsen renal failure, reducing GFR and lithium excretion, thereby worsening toxicity.

Hypercalcaemia, hypokalaemia, release of ureteric or urethral obstruction and hypoproteinaemia are also recognised as causing NDI and are associated with reduced expression of AQP₂ channels, urea transport proteins and a loss of interstitial hypertonicity. These defects normally cause milder polyuria than that associated with lithium toxicity.

TREATMENT OF NDI

The treatment of NDI aims to minimise the occurrence of hypernatraemia and hypovolaemia and wherever possible to remove the underlying cause.

MEASURE AND CALCULATE PLASMA OSMOLALITY AND URINE OSMOLALITY

A high urine osmolality in the presence of a high plasma osmolality is appropriate and, if the plasma osmolality is higher than 295 mosmol/l, urine osmolality should reach 1000–1200 mosmol/kg. Urine osmolalities of less than this imply that there is a urine-concentrating defect and lead to consideration of the causes of DI and also the use of medications that might reduce renal interstitial hypertonicity, such as loop diuretics. Urine osmolalities of < 150 mosmol/kg in this circumstance are sufficient to make the diagnosis of DI provided there is not obvious gross fluid and solute overload of the patient.

Where the patient is polyuric with a high plasma osmolality and maximum urine osmolalities are being achieved, an osmotic diuresis is implied. The diuresis may be inappropriate in as much as it is leading to dehydration but appropriate in that the kidneys are retaining as much water as they can for the large solute load being excreted. If there is a normal osmolal gap (the difference between measured and calculated plasma osmolality, normal < 10 mosmol/kg), then hyperglycaemia, hypernatraemia or hyperkalaemia is implied. If the osmolal gap is greater than 10 mosmol/kg, investigation should be undertaken to look for an unmeasured solute such as ethanol, mannitol, ethylene glycol, sorbitol or methanol.

If plasma osmolality is less than 280 mosmol/kg, urine osmolality would also be expected to be lower than this as the body attempts to clear free water. This picture implies water overload which may be iatrogenic or patient-mediated. Low plasma osmolality with high urine osmolality implies SIADH and would not normally be associated with polyuria.

WATER DEPRIVATION AND ADH TESTS

In health, being deprived of water rapidly results in an increase in plasma osmolality which causes ADH release and an increase in urine osmolality to between 1000 and 1200 mosmol/kg in order to preserve water and reduce plasma osmolality back towards its normal value. When the cause of polyuria is unclear and the patient is not already clinically dehydrated, a water deprivation test may be useful to determine the cause of the diuresis. In patients with severe polyuria, the test is potentially dangerous as dehydration and hyperosmolality can develop very rapidly, resulting in permanent cerebral damage and cardiovascular collapse. It is therefore wise to undertake the test under very close supervision during daylight hours.

The limitations of the test should also be appreciated:

Figure 51.4 Urine versus plasma osmolality during water deprivation testing. DDAVP, 1-deamino-8-O-arginine vasopressin; CDI, cranial diabetes insipidus; NDI, nephrogenic diabetes insipidus. Adapted from Sands JM, Bichet DG. Ann Intern Med 2006; 144: 186–94.

Figure 51.5 (a) As plasma osmolality increases as a result of water deprivation or hypertonic saline infusion, the measured concentration of antidiuretic hormone (ADH) should rise. Failure to do so suggests cranial diabetes insipidus. (b) At a given value of plasma ADH, whether endogenous or exogenous in origin, urine osmolality is expected to lie within a corresponding range of osmolality. However, the range of expected osmolalities is high as osmoreceptors adjust regulation around the basal level of ADH, which is also partly determined by volume receptors. In nephrogenic diabetes insipidus (NDI), the urine osmolality remains low even when high plasma ADH concentrations are measured.

INTERPRETATION OF THE TESTS (SEE FigureS 51.4 and 51.5)

In complete DI, plasma osmolality will rise but urine osmolality will not rise above 300 mosmol/kg. Upon administration of ADH, patients with complete CDI will raise their urine osmolality to 500 mosmol/kg or higher whereas there will be no rise in urine osmolality in complete NDI.

In complete CDI, the original plasma ADH measurement will be zero, whereas in complete NDI, the initial ADH measurement will be normal or high depending on the corresponding plasma osmolality at the time of measurement.

In partial DI, plasma osmolality will rise and urine osmolality will also increase but usually plateaus between 400 and 800 mosmol/kg. In partial CDI the ADH will initially be normal or low and will rise with increasing plasma osmolality but is unlikely to rise above 4–5 pg/ml. In partial NDI, the ADH will initially be normal or high and will increase with plasma osmolality to > 8 pg/ml but without achieving a correspondingly appropriate rise in urine concentration.

Following administration of ADH or DDAVP, urine osmolality is expected to double at least in complete CDI and rise by 10–50% in partial CDI or NDI. In complete NDI, there will be no rise in urine osmolality. Partial CDI may be inferentially differentiated from partial NDI on the basis of urine osmolality alone; in the former, urine osmolality rises above plasma osmolality whereas in partial NDI it tends to remain hypo-osmolal or iso-osmolal to plasma following ADH/DDAVP administration. However, this generalisation is indicative only and for greater certainty, it is preferred to have measured sequential ADH concentrations to assess hypothalamic–pituitary function independently of the renal concentrating ability.

Water deprivation tests should not be conducted in infants or patients with pre-existing hypovolaemia or hyperosmolality. In the latter two categories, treatment of the fluid deficit and/or solute excess should precede further investigations. In infants and in adults with equivocal water deprivation test results, an infusion of hypertonic saline should be considered.

HYPERTONIC SALINE INFUSION

In patients with equivocal results from a water deprivation test and in those at high risk of dehydration and hypovolaemia from water deprivation, a hypertonic saline infusion test may be undertaken to establish the cause of a water diuresis. Interpretation of the test is as for the water deprivation test but there is a clearer demarcation between partial CDI and primary polydipsia and the risk of missing a diagnosis of CDI secondary to osmoreceptor dysfunction is reduced.⁵⁹

Hypertonic saline is available in multiple different concentrations. Care is required in calculating the appropriate infusion rate to use for the test as this varies with the concentration of the hypertonic preparation stocked. Hypertonic sodium chloride solution (0.0425 mmol/kg per min) is infused for up to 3 hours or until a plasma osmolality of 300 mosmol/kg is achieved.

Blood samples are taken 30 minutes before and at 30-minute intervals throughout the duration of the test and plasma sodium, osmolality and ADH are measured. Urine samples are collected before and where possible at 60-minute intervals throughout the test period and measurements of osmolality and sodium are performed. Thirst and blood pressure are recorded at 30-minute intervals.