Show Summary Details
Page of

Disorders of water and sodium homeostasis 

Disorders of water and sodium homeostasis
Disorders of water and sodium homeostasis

Michael L. Moritz

and Juan Carlos Ayus



Prevention of hospital-acquired hyponatraemias—recommendation that use of hypotonic intravenous fluids should be restricted to patients with hypernatraemia (Na >145 mmol/L) or those with ongoing urinary or extrarenal free water losses.

Treatment of suspected hyponatraemic encephalopathy—recommendation of a 2 ml/kg intravenous bolus of 3% sodium chloride to maximum of 100 ml to produce a controlled and immediate rise in serum sodium with little or no risk of inadvertent over-correction. Discussion of management guidelines to prevent overcorrection of hyponatraemia.

Chronic hyponatraemia—increased recognition that this is associated with significant morbidity, particularly falls and bone fractures in the elderly. Use of vasopressin receptor antagonists discussed.

Updated on 25 May 2011. The previous version of this content can be found here.
Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2016. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 21 November 2019


Regulation of water balance and sodium disorders

Water intake and the excretion of water are tightly regulated processes that are able to maintain a near-constant serum osmolality. Sodium disorders (dysnatraemias—hyponatraemia or hypernatraemia) are almost always due to an imbalance between water intake and water excretion.

Understanding the aetiology of sodium disorders depends on understanding the concept of electrolyte-free water clearance—this is a conceptual amount of water that represents the volume that would need to be subtracted (if electrolyte-free water clearance is positive) or added (if negative) to the measured urinary volume to make the electrolytes contained within the urine have the same tonicity as the plasma electrolytes. It is the concentration of the electrolytes in the urine, not the osmolality of the urine, which ultimately determines the net excretion of water.


Hyponatraemia, defined as serum sodium concentration of less than 135 mmol/litre, is a common electrolyte disorder. It is almost invariably due to impaired water excretion, often in states where ADH release is: (1) a normal response to a physiological stimulus such as pain, nausea, volume depletion, postoperative state, or congestive heart failure; or (2) a pathophysiological response as occurs with thiazide diuretics, other types of medications, or in the syndrome of inappropriate diuresis; with both often exacerbated in hospital by (3) inappropriate iatrogenic administration of water (or 5% dextrose).

Clinical features—these can range from the patient who is entirely asymptomatic at one end of the spectrum to hyponatraemic encephalopathy—most commonly manifesting with nausea, vomiting, and headache—at the other. Cerebral demyelination is a serious complication associated with hyponatraemia and its treatment, at its worst manifesting as pseudocoma with a ‘locked in’ state. Children and premenopausal women are at particular risk of poor outcomes, as are those who are hypoxic at presentation.

Management approach—the first priority is to exclude a hyperosmolar state and verify whether the patient is hypotonic, by (when possible) measuring the serum osmolality. The diagnostic approach is further based on the history, clinical assessment of the patient’s volume status, and estimation of urinary electrolytes. Key issues are to recognize that: (1) hyponatraemic encephalopathy is a medical emergency that should be diagnosed and treated promptly with hypertonic saline to prevent death or devastating neurological complications; but also (2) that patients who are asymptomatic do not require treatment with hypertonic saline, whatever their level of serum sodium. Precipitating causes (e.g. thiazide diuretics) should be withdrawn when possible.

Practical management—algorithms, even if complex, cannot accurately predict a patient’s response to treatment of hyponatraemia: close monitoring of serum sodium is essential. Patients with suspected hyponatremic encephalopathy, with either mild or advanced symptoms, children or adult, should receive a 2 ml/kg bolus of 3% NaCl with a maximum volume of 100 ml. A single bolus would result in at most a 2 mmol/L acute rise in serum sodium, which would quickly reduce brain edema. The bolus could be repeated 1 - 2 times if symptoms persist. The advantage of this approach over a continuous infusion of 3% NaCl is that there is a controlled and immediate rise in serum sodium and there is little or no risk of inadvertent overcorrection, as can occur if a 3% NaCl infusion runs at an excessive rate or for too long.

Cerebral demyelination—this is a serious complication that has been associated with the correction of hyponatraemia, hence all patients receiving an infusion of 3% saline should have their serum sodium measured at least every 2 h until they are clinically stable and the serum sodium values are stable, with appropriate modification of treatment in response to the measurements. Failure to do so, and reliance on a calculated infusion rate, can lead to significant patient injury.

Prevention—hyponatraemia is usually iatrogenic and can be avoided or detected as follows: (1) hypotonic fluids should never be administered following surgery unless used to correct a free-water deficit—0.9% (normal) saline (NaCl) should be given postoperatively if parenteral fluids are indicated; (2) all hospitalized patients should be considered at risk for the development of hyponatraemia and should not be given hypotonic fluids unless a free-water deficit is present or if ongoing free-water losses are being replaced; (3) patients taking thiazide diuretics, especially older people, should be weighed before and after starting therapy and serum electrolytes monitored to detect water retention and the development of hyponatraemia.


Hypernatraemia, defined as serum sodium concentration greater than 145 mmol/litre, is a common electrolyte disorder that occurs when water intake is inadequate to keep up with water losses. Since the thirst mechanism is such a powerful stimulus, the almost invariable context is illness and care that restrict the patient’s access to water.

Clinical features—these are mainly related to central nervous system dysfunction caused by cerebral dehydration and cell shrinkage.

Management approach—the first step in evaluation is to take a detailed history focusing on fluid intake and losses. To assess urinary water losses, it is necessary to measure the urinary cationic electrolytes (sodium and potassium) and the urinary osmolality, remembering that the urinary osmolality alone cannot always determine the presence or absence of electrolyte-free water losses in the urine, the reason being that water can be excreted with nonelectrolyte osmoles or with electrolyte osmoles.

Practical management—needs to be guided by the following principles: (1) correction of underlying deficits in circulatory blood volume by infusion of 0.9% saline; (2) correction of chronic hypernatraemia at a pace that avoids therapy-induced cerebral oedema, which requires an understanding of both the initial water deficit and of ongoing water losses if the patient is polyuric; (3) administration of water by drinking or feeding tube is preferable to treatment with intravenous fluids if possible; (4) glucose-containing solutions should be avoided if possible; (5) as for the treatment of hyponatraemia, algorithms cannot accurately predict the response to treatment of hypernatraemia, hence regular monitoring of serum sodium with appropriate adjustment of treatment in response to the values obtained is mandatory.

Prevention—(1) patients with impaired access to water (e.g. infants, elderly, and hospitalized patients) should be considered at risk for the development of hypernatraemia, and their serum sodium should be monitored; (2) urinary electrolytes should be measured in conjunction with urinary osmolality in patients with polyuria to assess water losses in the urine and urinary concentrating ability.

Disorders of water metabolism

Hyponatraemia and hypernatraemia occur when there is a breakdown of the normal homeostatic mechanisms that keep water intake and excretion precisely balanced to prevent the development of disturbances in the serum sodium. There are numerous causes of impairment in this homeostatic function, such as renal failure, use of diuretics, and nonosmotic release of ADH due to nausea, pain, or other stimuli. Poor outcomes are still common among patients with hypernatraemia and hyponatraemia, in many cases due to failure to promptly recognize a life-threatening condition and initiate appropriate treatment. In this chapter the pathophysiology of sodium disturbances is addressed, with a focus on understanding clinical presentations of the diseases.

Regulation of water balance

Dysnatraemias (hyponatraemia or hypernatraemia) occur when there is an imbalance between water intake and water excretion. Extracellular fluid tonicity is reflected by the concentration of the serum sodium. Nearly all cell membranes are permeable to water, hence water will equilibrate between the intracellular space and the extracellular space to maintain the same osmolality in both compartments, and intracellular electrolyte concentrations will approximate the extracellular electrolyte concentrations. This means that the serum sodium concentration (Nase) is proportional to the total body exchangeable sodium (Nae) plus the total body exchangeable potassium (Ke): (Equation


Water intake and the excretion of water are tightly regulated processes and therefore a near-constant serum osmolality is maintained (Fig. Because of this tight regulation, disturbances in serum sodium are nearly always caused by perturbations in water balance, not of electrolytes.

Fig. Regulation of water intake and excretion to maintain normonatraemia.

Regulation of water intake and excretion to maintain normonatraemia.

Achinger SG, Ayus JC. Fluid and Electrolytes. In Civetta, Taylor, and Kirby’s Critical Care, 4th edition.

Renal water handling

It is through the actions of ADH (also known as AVP) that the kidney regulates water excretion. The normal kidney has the ability to vary urinary concentration significantly, from as low as 50 mOsm/kg to as high as 1200 mOsm/kg when ADH activity is maximal, although when there is renal insufficiency (especially tubulointerstitial disease) this range is much more restricted. This means that the kidney can either excrete a large water load in very dilute urine or conserve water significantly. A mathematical illustration can make this clear. If a daily solute load is taken to be 800 mOsm (mainly electrolytes and urea, the latter due to protein catabolism), then this amount must be excreted in 24 h in order to maintain solute balance. Under conditions of maximal urinary concentration this could be excreted in approximately 667 ml of urine ([800 mOsm/1200 mOsm] × kg−1), which would be the expected response to a hypernatraemic state. Conversely, under conditions of maximal urinary dilution this osmolar load would be excreted in 16 litres of urine ([800 mOsm/50 mOsm] × kg−1), which would be an expected response to water intoxication or could occur in the setting of diabetes insipidus. Therefore, the body has the ability, under normal conditions, to achieve water balance across a very wide range of water intake. Disorders in water balance usually occur when there is a disruption in these processes that allow water intake and water excretion to be exquisitely matched.

The concept of electrolyte-free water

The concept of electrolyte-free water is a good approach to understanding patients with disturbances in water balance. The electrolyte-free water clearance is a conceptual amount of a fluid that represents the volume that would need to be subtracted (if electrolyte-free water clearance is positive) or added (if negative) to the measured urinary volume to make the electrolytes contained within the urine have the same tonicity as the serum electrolytes (Fig. (Equation


Fig. Resolution of urinary volume into a proportion of electrolytes isotonic to serum, with the remainder as ‘electrolyte-free water’.

Resolution of urinary volume into a proportion of electrolytes isotonic to serum, with the remainder as ‘electrolyte-free water’.

where [Na+]u is urinary sodium concentration, [K+]u is urinary potassium concentration, [Na+]se is serum sodium concentration, and [K+]se is serum potassium concentration.

The electrolyte-free water represents the amount of water lost in excess of electrolytes and which would therefore—if not replaced—have an effect on serum osmolality. A few key points must be made about the electrolyte-free water. First, it is truly a conceptual volume because, as can be seen in Equation, it can take on a negative value, which occurs when the electrolyte concentration in the urine exceeds that in the serum: when the electrolyte-free water clearance is negative, there is net retention of electrolyte-free water. Second, the concept of electrolyte-free water clearance highlights the fact that it is the concentration of the electrolytes in the urine, not the osmolality of the urine, which ultimately determines the net excretion of water. In other words, the urine osmolality may be high but, if the urine contains mainly urea and very few electrolytes, there will still be a net loss of water. Electrolyte-free water clearance can therefore be calculated as shown in Equation as a convenient clinical tool for assessing water need in a patient.

Clinical utility of electrolyte-free water clearance

A critical point to understand is that the urine electrolytes and not the urine osmolality determine the amount of free water excreted in the urine. Typically, if the relationship between the serum electrolytes and the urine electrolytes is understood, it is not necessary to calculate a value for the electrolyte-free water clearance. In the case where the concentration of electrolytes in the urine exceeds the concentration of electrolytes in the serum, then free water is not being excreted in the urine. Conversely, when the concentration of electrolytes in the urine is less than that in the serum, then free water is being excreted in the urine. Fig. illustrates this relationship: much can be learned regarding water excretion by simply examining the concentration of the electrolytes in the urine.

Fig. Measurement of serum and urinary electrolytes can determine whether the patient is retaining or excreting electrolyte-free water. Nau, urine sodium (spot); Ku, urine potassium (spot); Nase, serum sodium; Kse, serum potassium.

Measurement of serum and urinary electrolytes can determine whether the patient is retaining or excreting electrolyte-free water. Nau, urine sodium (spot); Ku, urine potassium (spot); Nase, serum sodium; Kse, serum potassium.


Hyponatraemia is defined as serum sodium lower than 135 mmol/litre, which is a common condition in hospital settings and is increasingly recognized in outpatients. Hyponatraemia can be asymptomatic, although careful neurological evaluation has detected subtle abnormalities in patients with chronic hyponatraemia and serum sodium as high as 132 mmol/litre. At the other end of the spectrum, presentation with hyponatraemic encephalopathy (central nervous system symptoms secondary to cerebral oedema) is a medical emergency that must be diagnosed promptly and treated quickly, or death or devastating neurological complications can result. It is critical to differentiate between these two extremes because the management is much different, depending on the symptoms. It is recognized that risk factors for hyponatraemic encephalopathy play a critical role in the determining whether or not patients are likely to develop this condition as a consequence of hyponatraemia, these risk factors for poor outcome being young age, premenopausal women, and hypoxia. The pathogenic mechanisms that are responsible for these risk factors are discussed later in this chapter.


Most cases of hospital-acquired hyponatremia can be prevented by avoiding hypotonic intravenous fluids and administering 0.9% sodium chloride when indicated. Significant morbidity and mortality from hyponatremic encephalopathy has occurred in hospitalized patients receiving hypotonic intravenous fluids, in particular post-operative patients. In 2003, we proposed that 0.9% NaCl (Na 154 mmol/l) be administered for the prevention of hospital-acquired hyponatremia in patients at risk for AVP excess, and that the routine practice of administration of hypotonic and near isotonic intravenous fluids (Na ≤130 mmol/l) be abandoned. Hospitalized patients are at high risk for hospital-acquired hyponatremia from numerous physiologic stimuli for AVP production, such as nausea, vomiting, pain, stress, volume depletion and disease states associated with high AVP production such as respiratory illnesses, CNS disease, and the post-operative state. Numerous studies have demonstrated that hypotonic fluids result in a high incidence of hospital acquired hyponatremia, whereas 0.9% NaCl effectively prevents the development of hyponatremia. We recommended that hypotonic fluids be restricted in their use to patients with either hypernatremia (Na >145 mmol/l), or ongoing urinary or extrarenal free water losses.


The main defence against the development of hyponatraemia is the ability of the kidney to dilute the urine and excrete free water. The typical adult (assuming normal renal function) can excrete approximately 15 litres of free water per day in the urine, hence excess ingestion of water as the sole cause of hyponatraemia is rare outside of the setting of mental illness. An underlying condition that impairs free-water excretion (Box is typically necessary in conjunction with free-water intake for the development of hyponatraemia. States of impaired water excretion are often states where ADH release is a normal response to a physiological stimulus such as pain, nausea, volume depletion, postoperative state, or congestive heart failure. ADH release may also be pathophysiological such as occurs with thiazide diuretics or with other types of medications such as antiepileptic drugs, or in the syndrome of inappropriate diuresis.

Brain defences against cerebral oedema

Hyponatraemia leads to an osmotic gradient favouring water movement intracellularly, which—if allowed to act unopposed—could lead to cerebral oedema and severe neurological injury. The first-line defence against this is the blood–brain barrier, which impedes the entry of water. This starts with tight junctions between vascular endothelial cells of the brain capillaries and their interface with the foot processes of astrocytes, the latter being a highly specialized subtype of glial cell that performs many supporting functions in maintenance of the fluid environment and electrolyte milieu of the extracellular space of the brain.

The astrocytes are the main regulator of brain water content: they swell during hypotonic stress, whereas neurons do not, with this capacity largely due to the presence of a water channel specific to astrocytes, aquaporin 4. Mice with targeted deletion of aquaporin 4 are protected from cytotoxic cerebral oedema caused by water intoxication, brain ischaemia, or meningitis, but are particularly vulnerable to vasogenic cerebral oedema caused by e.g. cerebral abscess or tumour, or hydrocephalus. The response of the astrocyte is critical in determining the degree of cerebral oedema in response to hypo-osmolar stress, and modulation of aquaporin 4 production or function may prove useful in the management of a variety of cerebral disorders, including those associated with hyponatraemia, in the future.

However, progressive and increasing swelling of astrocytes in the face of hyponatraemia would not protect the brain against adverse consequences, and there are several other protective mechanisms. There is shunting of cerebrospinal fluid from within the brain: this is a rapid response, but its capacity to buffer significant volume change is limited. Ultimately, cell volume regulatory mechanisms in the cerebral astrocytes must be active to decrease the brain size. This is accomplished by reduction in cellular osmolyte content (mainly electrolytes) using an ATP-dependent mechanism that requires Na+,K+-ATPase to extrude ions (electrolytes) from within, with water obligatorily following to reduce brain volume. In animal models of acute hyponatraemia, brain water content is returned to near the baseline value 6 h after induction of acute hyponatraemia. As will be discussed later, several clinical factors have been shown to impair these glial cell adaptive responses, and these are the chief risk factors for poor patient outcome.

Clinical manifestations

Advanced symptoms of hyponatremic encephalopathy include seizures, respiratory arrest, and non-cardiogenic pulmonary edema.

The symptoms of hyponatraemia are attributable to osmotic swelling of the brain, with pressure on the brain parenchyma arising because of the rigid structures encasing the central nervous system. The manifestations can be varied and not necessarily related to the degree of decrease in serum sodium concentration, which is frequently less than 120 or 115 mmol/litre in congestive heart failure and in cirrhosis with very few—if any—overt symptoms. Conversely, life-threatening cerebral oedema can be the presentation of a patient with a serum sodium as high as 128 mmol/litre. Hyponatraemic encephalopathy is defined as symptomatic cerebral oedema secondary to hyponatraemia: the early signs are usually nonspecific—nausea, vomiting, headache—and can often go unrecognized, with advanced symptoms being signs of brainstem herniation—including seizures, respiratory arrest, non-cardiogenic pulmonary edema, dilated pupils, and decorticate posture—which can lead to death if left untreated.

Hyponatraemic encephalopathy

Risk factors

Not all patients are equal in terms of risk of morbidity and mortality following the development of hyponatraemia. Children are at particular risk for poor outcome following the development of hyponatraemia due to their high ratio of brain size to skull size, the skull not reaching its full size until age 16 years, whereas the brain reaches its adult size at approximately age 6 years. This means that children cannot accommodate as much increase in brain size as adults: there is less capacity for brain expansion before pressure is exerted on the brain parenchyma. For this reason, the long-standing practice of administering hypotonic fluids to children is being challenged: normal (0.9%) saline is the most appropriate fluid to use to prevent the development of iatrogenic hyponatraemic encephalopathy in children.

Premenopausal women are another significant risk group in terms of neurological outcomes following hyponatraemia, being 25 times more likely to die following hyponatraemic encephalopathy than other groups of patients. This striking difference is not accounted for by differences in clinical presentation, and anatomical factors in terms of the brain size—cranial vault size ratio (as is seen with children)—cannot explain the disparity in outcomes. Differences in adaptive responses to hyponatraemia must exist. As described above, it is known that ATP-dependent mechanisms are important for the response to hypo-osmolar stress in the brain. Oestrogens have a similar steroidal structure to ouabain and other cardiac glycosides (such as digoxin), which are known to inhibit the Na+,K+-ATPase, and female sex hormones have been shown to inhibit the activity of this pump in diverse tissues such as mammalian heart, diaphragm, red blood cells, and liver. Sex steroids and gender also play a significant role in brain adaptation and in animal models of hyponatraemia. There is increased morbidity from hyponatraemia in female rats, and isolated synaptosomes from female hyponatraemic rats have increased uptake of sodium compared with male hyponatraemic rats, suggesting impairment in sodium extrusion. Regulatory volume decrease is also inhibited by the presence of oestrogen/progesterone in rat astrocytes treated in vitro. These studies support the notion that the presence of female sex hormones can impair the critical energy-dependent astrocyte cell volume regulatory processes, with this impairment leading to more severe cerebral oedema. Finally, female rats have more intense vasoconstriction than male rats in response to ADH, which may lead to tissue hypoxia, which is another possible factor in producing poor outcomes.

Role of hypoxia

Animal studies have demonstrated that survival is severely impaired and brain adaptation is significantly impaired following hyponatraemia and simultaneous brain hypoxia. Epidemiological studies have shown that patients with hypoxia at presentation of hyponatraemic encephalopathy have poor outcome compared with those who are not hypoxic, even after adjustment for comorbid conditions. Since impairment of astrocyte adaptive mechanisms can explain poor outcome in premenopausal females, it has been proposed that hypoxia may similarly have an effect on astrocyte volume regulation. Impairment of energy utilization in the brain—a common phenomenon following asphyxiation or cardiac arrest—can lead to diffuse cerebral oedema, termed ‘cytotoxic’ cerebral oedema, related to the impairment of cell-volume regulatory mechanisms. Hence, if hyponatraemia—which will by itself induce cerebral oedema—is compounded with impairment in volume regulatory mechanisms through hypoxia, then this is likely to lead to more severe cerebral oedema than if hypoxia were not present and a poor outcome.

Hypoxia develops in patients with hyponatraemic encephalopathy through two mechanisms: hypercapneic respiratory failure and neurogenic pulmonary oedema. Hypercapneic respiratory failure is secondary to central respiratory depression and is a first sign of impending brainstem herniation, with the hypoxaemia that then develops being due to the central respiratory depression further worsening astrocyte cell-volume regulatory mechanisms and leading to worsening of brain oedema. Neurogenic pulmonary oedema, caused by increased vascular permeability and increased catecholamine release that occurs secondary to elevated intracranial pressure, is a complication of cerebral oedema and can occur in the setting of hyponatraemic encephalopathy as well (Fig. This form of noncardiogenic pulmonary oedema is known as the Ayus-Arieff Syndrome, and is secondary to increased intracranial pressure due to cerebral oedema. Hypoxaemia is therefore both a risk factor and a pathogenic mechanism in severe cerebral oedema, as once hypoxaemia is established the underlying cerebral oedema will worsen because the hypoxia will initiate a vicious cycle that—unless broken—results in worsening of the underlying cerebral oedema (Fig.

Fig. Mechanism of non-cardiogenic pulmonary edema in hyponatremic encephalopathy.

Mechanism of non-cardiogenic pulmonary edema in hyponatremic encephalopathy.

Fig. The possible clinical outcomes of a hypo-osmolar state.

The possible clinical outcomes of a hypo-osmolar state.

Achinger, S.G., Ayus J.C. (2008). Fluid and Electrolytes. In Civetta, Taylor, andKirby’s Critical Care, 4th edition.

Diagnostic approach to hyponatraemic patients

With a patient with hyponatraemia, the first priority is to exclude a hyperosmolar state and verify that a hypotonic state exists by (when possible) measuring the serum osmolality. An osmotically active substance confined to the extracellular fluid—most typically glucose, mannitol, or glycine—leads to translocation of water from the intracellular space and to a decreased serum sodium concentration despite a net increase in serum osmolality. In assessing for a disturbance of sodium when hyperglycaemia is present, the serum sodium must be ‘corrected’ for the presence of hyperglycaemia by adding 1.4 mmol/litre for every 5 mmol/litre increase of the serum glucose above 5 mmol/litre (or 1.6 mmol/litre for every 100 mg/dl above 100 mg/dl).

The possibility of pseudohyponatraemia should also be kept in mind. Hyperproteinaemia and hyperlipidaemia can lead to spuriously low serum sodium measurements if samples are diluted prior to measurement of the serum sodium. By contrast, pseudohyponatraemia due to hyperproteinaemia and hyperlipidaemia alone is not a concern when a potentiometric method is used. Measured serum osmolality will be normal if pseudohyponatraemia is present.

The diagnostic approach is further based on the history, urinary electrolytes, and clinical assessment of the patient’s volume status (Fig., which will be further demonstrated in the case discussions to follow.

Fig. Treatment of hyponatraemia.

Treatment of hyponatraemia.

Adapted with permission with permission from Achinger SG, Moritz ML, Ayus JC. Dysnatremias: Why are patients still dying? South Medical J 2006; 99(4): 353–362.

Treatment of hyponatraemic encephalopathy

A 100 ml 3% sodium chloride bolus is the preferred treatment for symptomatic hyponatremia.

Hyponatraemic encephalopathy is a life-threatening medical emergency that must be treated appropriately and in a timely manner to avoid death or severe neurological impairment. As described above, early symptoms are headache, nausea, and vomiting, with seizures commonly seen if cerebral oedema worsens. The final stages, if not corrected, are coma, respiratory arrest, and death.

The aim of treatment of hyponatraemic encephalopathy is to: (1) remove patients with severe manifestations of cerebral oedema from immediate danger, (2) correct serum sodium to a mildly hyponatraemic level, and (3) maintain this level of serum sodium to allow for the brain to adapt to the change in serum osmolality. Prompt treatment is essential in all patients with hyponatraemic encephalopathy, with the definitive therapy being administration of hypertonic saline (3% sodium chloride, 513 mmol/L). Most of the morbidity associated with this condition results from insufficient therapy rather than overcorrection. Fluid restriction alone is inadequate therapy for symptomatic hyponatremia, and 0.9% sodium chloride, 1.8% sodium chloride, and V2 receptor antagonists are also inappropriate for the treatment of patients who are encephalopathic. 0.9% and 1.8% sodium chloride are not sufficiently hypertonic to consistently induce the rapid rise in plasma osmolality necessary for the reduction in cerebral edema central to the management of this condition. The only consistent way of acutely increasing the plasma sodium and to most effectively treat hyponatraemic encephalopathy is to administer 3% sodium chloride, which has a sodium concentration that exceeds the kidney’s ability to generate free water, but there is significant controversy about the indications and appropriate use of 3% sodium chloride for the treatment of this condition, and 3% sodium chloride is not readily available in many hospitals.

We recommend that any patient with suspected hyponatremic encephalopathy, with either mild or advanced symptoms, children or adult, should receive a 2 ml/kg bolus of 3% sodium chloride to a maximum volume of 100 ml (Box A single bolus results in at most a 2 mEq/L acute rise in serum sodium, which would quickly reduce brain edema, and the bolus can be repeated 1 - 2 times if symptoms persist. The advantage of this approach over a continuous infusion of 3% sodium chloride is that there is a controlled and immediate rise in serum sodium and little or no risk of inadvertent overcorrection, as can occur if a 3% sodium chloride infusion runs for too long.

Reproduced from Moritz ML, Ayus JC. New aspects in the pathogenesis, prevention, and treatment of hyponatremic encephalopathy in children. Pediatr Nephrol 2009.

At times the diagnosis of hyponatremic encephalopathy can be difficult to establish, such as in patients with either: a) hepatic encephalopathy, b) CNS infections, tumors, or trauma, or c) post-operative nausea and vomiting with associated hyponatremia. Bolus therapy with 3% sodium chloride can serve as a diagnostic maneuver, as a patient who does not show some clinical improvement after 2 – 3 boluses is most likely is not suffering from hyponatremic encephalopathy. As long as it does not lead to significant delay in pursuing other diagnostic possibilities, no harm could come from using this approach in a patient with suspected hyponatremic encephalopathy, even if they subsequently prove not to have this condition, and a therapeutic trial of a bolus of 3% sodium chloride should precede radiological investigations because a) neurological deterioration could occur if there is a delay in therapy, and b) a CT scan cannot exclude the possibility of hyponatremic encephalopathy.

Recommended safe limits for the correction of hyponatremia vary among experts depending on the setting of hyponatremia, including 6 – 8 mEq/l in 24 hours or 20 mEq/l in 48 hours, as do recommendations for using hypertonic saline. Our recommendation to use bolus therapy is an approach that would stay well within widely recommended limits of correction and can be used safely in any setting—for children or adults, in chronic or acute symptomatic hyponatremia, and in the outpatient or inpatient setting. The approach is simple: it does not rely on formulas or complicated calculations, and it can be administered quickly in the emergency department or at the bedside, prior to transfer to a monitored setting.

A few precautionary points must be understood to prevent therapy-induced brain injury: (1) the serum sodium should never be corrected to a normonatraemic or hypernatraemic level in a patient treated for hyponatraemic encephalopathy; (2) following correction, patients should be maintained at mildly hyponatraemic levels for a few days following hyponatraemic encephalopathy (this maintenance period will allow the patient to adjust to the new serum tonicity); and (3) if the patient has decreased cardiac output and pulmonary oedema may develop with vigorous saline volume expansion, then furosemide should be given in addition to hypertonic saline—this should prevent volume overload and pulmonary oedema, but such a patient requires very close monitoring.

Early recognition of hyponatraemic encephalopathy and institution of prompt treatment are the factors most associated with good neurological outcomes. Appropriate treatment with hypertonic saline is safe and effective, but improper therapy can have severe consequences. Some authors describe formulas of varying complexity to guide treatment of hyponatraemia: these should not be used at all when determining the amount of hypertonic saline to give because they are fundamentally flawed. Ongoing water losses are unpredictable, and formulas are not able to take these into account because they are based on a closed system assumption (i.e. no ongoing water losses). Significant overcorrection can occur if saline is prescribed according to formula when a patient undergoes a spontaneous water diuresis, which can easily occur when the stimulus for water retention is removed and the body begins to respond appropriately to hypotonicity by suppressing ADH release and excreting dilute urine, , hence the recommendations made in Box

Risk factors for the development of cerebral demyelination

Cerebral demyelination is a potential complication of the overcorrection of chronic hyponatremia.

Cerebral demyelination is a serious complication that has been associated with the overcorrection of severe (serum sodium < 115 mEq/L) chronic hyponatremia (> 48 hours) and is rarely seen in acute hyponatremia. The symptoms often become apparent days to weeks following correction of hyponatraemia, and can vary from being minimal or none to as severe as a pseudocoma with a ‘locked-in stare’. A key point is that the hourly rate of correction of serum sodium by itself is not predictive of cerebral demyelination, whereas the absolute change in serum sodium over 48 h is predictive. This is important because it is not appropriate to treat a patient with respiratory arrest due to hyponatraemic encephalopathy with a ‘slow’ infusion of hypertonic saline to increase the serum sodium by 0.5 to 1 mmol/litre per hour. This type of patient—with impending brainstem herniation—should be treated with a bolus of hypertonic saline to quickly reduce brain volume, after which the hourly rate of correction can be more modest as long as the total change in serum sodium does not exceed 15 to 20 mmol/litre over 48 h and the patient is not corrected to normonatraemic levels. Additionally, there are other clinical factors not related to degree of correction such as liver disease and hypoxia that increase the risk of cerebral demyelination, and particular care must be exercised in treating these patients, for whom the safe degree of correction over 48 h is not known but might be less than 15 to 20 mmol/litre over 48 h.

Preventing overcorrection of hyponatremia

Preventing an extreme rise in serum sodium (> 25 mmol/L in 48 hrs) can be difficult, particularly in the severely hyponatremic patient (SNa ≤ 115 mmol/L). The overall rate of correction of hyponatremia is primarily a determinant of the renal response to fluid therapy, more so than the composition of fluids administered. Under most circumstances, hyponatremia develops due to a state of high AVP production. Once the stimulus for AVP production abates, there will be brisk urinary excretion of free water and hyponatremia will correct rapidly. The main conditions in which correction by fluid therapy will induce a brisk free water diuresis are a) thiazide-induced hyponatremia, b) water intoxication, c) gastroenteritis, d) adrenal insufficiency following replacement therapy, and e) DDAVP-induced hyponatremia following DDAVP withdrawal. Even in patients who are not typically at high risk for overcorrection, such as those with SIADH and post-operative hyponatremia, when the stimuli for AVP production abates, a free water diuriesis will ensue. In general, if the serum sodium is greater then 115 mmol/L, then even if there is a brisk free water diuresis, the absolute rise in serum sodium will not likely to exceed 25 mmol/L, and the risk of brain injury is small. In Box are outlined the measures that should be taken to prevent overcorrection of hyponatremia. Under certain circumstances the administration of DDAVP will be indicated to prevent an overcorrection, and this can also been used to therapeutically re-lower the serum sodium following the overcorrection of chronic hyponatremia.

Asymptomatic hyponatraemia

Asymptomatic hyponatremia is an independent risk factor for falls and fractures in the elderly and for increased hospital mortality.

It has become increasingly clear that any degree of hyponatremia can have dangerous consequences and is associated with increased morbidity and mortality. Mild chronic hyponatremia (sodium < 130 mmol/L) can produce subtle neurological impairment affecting both gait and attention, similar to that of moderate alcohol intake. This may explain why hyponatremia has been increasingly associated with falls and bone fractures in the elderly, in which regard there is additional evidence in both animals and humans that chronic hyponatremia contributes to osteopenia. Hyponatremia appears to induce bone loss by stimulating osteoclastogenesis and bone resorptive activity as a means of preserving sodium homeostasis (Figure 7). It is now well established in adults that hospital-associated hyponatremia is an independent risk factor for all-cause mortality, with studies documenting an association in the ambulatory setting and general medical wards, as well as in patients with community acquired pneumonia, congestive heart failure, and end-stage liver disease. Hyponatremia is also recognized as an independent predictor of increased medical costs in hospitalized patients.

Fig. Mechanism of bone injury from chronic hyponatremia in the elderly.

Mechanism of bone injury from chronic hyponatremia in the elderly.

Reprinted from Ayus JC, Moritz ML. Bone disease as a new complication of hyponatremia: moving beyond brain injury. Clin J Am Soc Nephrol 2010; 5: 167–8.

Treatment of asymptomatic chronic hyponatremia

Regardless of the underlying aetiology and irrespective of the absolute level of serum sodium, asymptomatic hyponatraemia does not require treatment with hypertonic saline. Where possible, the underlying aetiology should be addressed, and fluid restriction may be helpful in some cases. However, asymptomatic and chronic euvolemic hyponatremia due to SIADH, and hypervolemic hyponatremia due to congestive heart failure or cirrhosis, can be difficult to manage and may be unresponsive to fluid restriction.

Demeclocycline, a tetracylcine derivative that produces vasopressin-resistant polyuria, has been given in the past, but is nephrotoxic and should no longer be used. A new class of drugs which hold a promising but as yet undefined role in the management of asymptomatic hyponatremia are the vasopressin receptor antagonists. There are currently two FDA-approved vasopressin antagonists for the management of euvolemic and hypervolumic hyponatremia; conivaptan (Vaprisol), which is a combined V1a and V2 receptor antagonist that is only approved for intravenous use, and Tolvaptan, which is an oral V2 receptor antagonist. Conivaptin is a potent inhibitor of the cytrochrome P450 system, hence its use is restricted as an intravenous agent to minimize adverse drug interactions. A large multicenter, double blinded, placebo controlled trial has demonstrated that tolvaptan is safe and effective in the management of chronic euvolemic and hypervolemic hyponatremia. However, it is not yet proven that undoubted efficacy in correcting the serum sodium concentration will translate into clinically important outcomes, e.g. less falls and fractures, and the high cost of these new drugs will mean that they remain inaccessible to many physicians and their patients. These agents should not be used for the treatment of symptomatic hyponatremia as they do not cause a sufficiently rapid or consistent increase in serum sodium.

Case discussions

Case 1: postoperative hyponatraemia

A 29-year-old woman without significant medical history undergoes an elective laparoscopic cholecystectomy. During the procedure, 5% dextrose in quarter-strength normal saline (0.22% NaCl) is started and maintained at 125 ml/h. There was some bleeding during the procedure, but blood transfusion was not required. The patient is kept in the hospital for observation because of the bleeding. She does not tolerate oral intake, and the intravenous fluids are continued at the current rate. At 4 a.m. the following day the woman complains of headache and she is given paracetamol by the on-call physician. At 10.30 a.m. the attending doctor is notified that the serum sodium is 128 mmol/litre: no new orders are received. The woman is found to be lethargic by the nursing staff and an order is received to withhold pain medications. That afternoon the woman has a seizure and goes into respiratory failure. She is placed on mechanical ventilation and transferred to the intensive care unit. At the time of transfer, the serum sodium is 124 mmol/litre.


The patient has several nonosmotic stimuli for ADH secretion (postoperative, volume depletion, pain, nausea) and the administration of a hypotonic fluid was not appropriate. Postoperative hyponatraemia can be prevented by the administration of 0.9% saline when parenteral fluids are indicated, with avoidance of the use of hypotonic fluids in a postsurgical patient. The induction of hyponatraemic encephalopathy is iatrogenic in this case and therapy needs to be instituted immediately, with a 100 ml 3% sodium chloride bolus as described in Box to try to prevent death or severe neurological impairment.

Case 2: exercise-associated hyponatraemia

A 24-year-old woman collapses 20 min after completing a marathon and is brought to the Emergency Department for evaluation. She has a decreased level of consciousness and is very short of breath. Cardiological examination is normal, but there are crackles in all lung fields. Neurological examination reveals a depressed mental status with no focal signs. The chest radiograph is consistent with pulmonary oedema. Serum electrolytes include sodium of 125 mmol/litre and potassium of 3.3 mmol/litre.


Exercise-associated hyponatraemia has been reported in marathon runners, with those at risk for this problem consuming large amounts of water throughout the course of the race. It is thought that significant amounts of this water remain unabsorbed, sequestered in the gut because blood flow is directed away from the splanchnic circulation while exercising vigorously. At the end of the race the sequestered water is absorbed and hyponatraemia develops rapidly, with water excretion being inhibited by high levels of ADH release secondary to extreme physical exertion.

As with Case 1, treatment needs to be started immediately with a 100 ml 3% sodium chloride bolus. This condition can be prevented by limiting fluid intake during endurance running: salt consumption or the use of hypotonic electrolyte sports drinks do not appear to be effective in prevention.

Case 3: DDAVP withdrawal

A 39-year-old man with a history of central diabetes insipidus following resection of a pituitary tumor is brought into the Emergency Department after a generalized seizure. He has previously been taking deamino-d-arginine vasopressin (DDAVP) 10 µg intranasally twice a day for his condition. In the Emergency Department he is found to be lethargic and unresponsive, with pulse 80 beats/min and blood pressure 135/80 mmHg, and his serum sodium is 119 mmol/litre and serum potassium 4.0 mmol/litre. It is not clear when the patient was last given a dose of DDAVP. Urine sodium is 125 mmol/litre and urine potassium 20 mmol/litre, with urine osmolality 585 mOsm/kg.

The man is given 2 litres of 0.9% saline in the Emergency Department. Six hours after presentation, the serum sodium is 127 mmol/litre and the man is admitted for management of hyponatraemia. The admitting physician continues to withhold the DDAVP and stops the intravenous fluids. The urine output increases significantly over the ensuing night, and the following morning the serum sodium is 158 mmol/litre, urine sodium 17 mmol/litre, urine potassium 10 mmol/litre, and urine osmolality 70 mOsm/kg.


DDAVP by itself is not a cause of hyponatraemia. DDAVP will cause retention of free water and therefore dosing must be titrated in conjunction with the patient’s fluid intake. The patient must be closely monitored and the serum electrolytes closely followed. If DDAVP is withheld following DDAVP-associated hyponatraemia, then a free-water diuresis will ensue and dangerous overcorrection of the serum sodium hypernatraemia may occur, as observed here. This is especially a concern in a patient such as this who has central diabetes insipidus and can rapidly excrete a large volume of dilute urine.

An appropriate approach to this patient with diabetes insipidus and hyponatraemic encephalopathy due to DDAVP-associated hyponatraemia would have been to continue DDAVP and restrict all enteral fluid intake. To correct the patient to the desired serum sodium level, 3% saline could have been used, and then discontinued. During this time, absolutely no hypotonic fluids would be given, and the patient would be monitored closely to restrict all enteral intake. A slow infusion of 0.9% saline could have been continued after the 3% saline was stopped if necessary to support volume status. This approach, coupled with frequent monitoring of the serum sodium, would have prevented overcorrection secondary to water diuresis as happened in this case.

Case 4: hyponatraemia due to syndrome of inappropriate diuresis

A 28-year-old man with HIV and a CD4 count of 75 is seen in follow-up 3 days after being discharged from the hospital where he had been diagnosed with pneumonia, suspected to be due to Pneumocystis jirovecii. His serum sodium had been decreased throughout the hospitalization, which was managed with fluid restriction. Current medications include a taper of prednisone and co-trimoxazole. He is significantly improved since hospital discharge, with physical examination revealing that he is afebrile, with pulse 84 beats/min, blood pressure 104/55 mmHg, and no abnormalities in the cardiac or respiratory systems—the lungs are clear and there is no peripheral oedema. Laboratory values from the morning of the clinic visit reveal the following:

  • serum—sodium 113 mmol/litre, potassium 3.9 mmol/litre, creatinine 64 µmol/litre (0.7 mg/dl), glucose 6.2 mmol/litre (112 mg/dl), osmolality 248 mOsm/kg

  • urine—sodium 105 mmol/litre, potassium 18 mmol/litre, osmolality 590 mOsm/kg


This presentation is consistent with syndrome of inappropriate diuresis, which is defined as hypotonic hyponatraemia, with a urine osmolality above 100 mOsm/kg, in the absence of volume depletion, adrenal insufficiency, congestive heart failure, hypothyroidism, cirrhosis, and/or renal impairment. The laboratory values support syndrome of inappropriate diuresis as the serum osmolality is decreased, which rules out hyperosmolar hyponatraemia (also known as dilutional hyponatraemia) and pseudohyponatraemia, and the urine osmolality is high.

If this man were responding normally to the hypotonicity of the serum, the urine should be dilute. The fact that the urine is concentrated is abnormal, but it is important be sure that the patient does not have another cause of a water-retentive state that is a physiological response. These are most commonly congestive heart failure, cirrhosis, and volume depletion. In these conditions, the low effective circulating blood volume initiates both sodium and water-retentive mechanisms. Hence a similar set of laboratory values may be seen in these conditions, with the exception that the urine sodium should not be 105 mmol/litre. When the kidney is conserving sodium, the urinary sodium is typically below 20 mmol/litre, but many patients with cirrhosis and congestive heart failure receive diuretics and interpretation of the serum sodium must be done cautiously in this setting. The presence of severe congestive heart failure and cirrhosis is typically obvious based on history and physical examination. Volume depletion (not due to diuretic use) can be distinguished from syndrome of inappropriate diuresis based on the urine sodium, which should be less than 20 mmol/litre. Other conditions leading to ADH release should also be considered and ruled out before syndrome of inappropriate diuresis is diagnosed: these include postoperative stress, medications, trauma, pain, and nausea. Pulmonary disease, especially pneumonia, is a common cause of syndrome of inappropriate diuresis, as seen in this case: other causes are given in Table

Table Causes of the syndrome of inappropriate antidiuresis

Neoplastic disease

Carcinoma (bronchus, pancreas, bladder, prostate, duodenum)



Lymphoma, leukaemia

Ewing’s sarcoma


Bronchial adenoma

Neurological disorders

Head injury, neurosurgery

Brain abscess

Brain tumour

Meningitis, encephalitis

Guillain Barré syndrome

Cerebral haemorrhage

Cavernous sinus thrombosis


Cerebellar and cerebral atrophy

Shy-Drager syndrome

Peripheral neuropathy


Subdural haematoma

Alcohol withdrawal

Chest disorders




Cystic fibrosis






Vincristine, cis-platinum



Dopamine antagonists

Tricyclic antidepressants



‘Ecstasy’ (3,4-MDMA)






Abdominal surgery’

MAOIs = monoamine oxidase inhibitors; SSRIs = selective serotonin reuptake inhibitors; 3,4-MDMA, 3,4-methylenedioxymetamphetamnine.

The diagnosis of the syndrome of inappropriate diuresis is often made incorrectly: the syndrome is a diagnosis of exclusion and it is essential that the diagnostic approach described in Fig. is rigorously applied to prevent wrong diagnosis and treatment.

Despite the profound hyponatraemia, this man is clinically well and thus he does not require treatment with hypertonic saline. Fluid restriction should be advised, with regular monitoring of the serum sodium, which should increase as syndrome of inappropriate diuresis resolves with recovery from pneumonia. If hyponatraemia is persistent and refractory to fluid restriction, then a V2 receptor blocker should be considered.

Case 5: hyponatraemic encephalopathy in an elderly woman presenting with a fall

A 77-year-old woman is brought to the Emergency Department after falling at home. Her past medical history is significant only with regard to osteoporosis, but the patient’s daughter states that 2 weeks ago she was started by her primary care physician on a blood pressure medication and that she has been slightly confused over the last few days. Physical examination reveals that she is afebrile, with pulse 70 beats/min and blood pressure 120/60 mmHg. She is confused and not answering questions appropriately, but cardiac examination is normal, the lungs are clear, and she does not have any pedal oedema.

Laboratory investigation reveals the following:

  • serum—sodium 110 mmol/litre, potassium 2.7 mmol/litre, creatinine 118 µmol/litre (1.3 mg/dl), urea 7.9 mmol/litre (22 mg/dl), glucose 6.2 mmol/litre (108 mg/dl), chloride 78 mmol/litre, bicarbonate 20 mmol/litre


Hydrochlorothiazide can lead to significant hyponatraemia and is one of the more common causes of hyponatraemia in an outpatient setting. Thiazide diuretics (but not loop diuretics) act at the level of the distal convoluted tubule and impair urinary concentrating capacity. ADH secretion is stimulated by a state of relative volume depletion, and the result is increased urinary concentration and water retention. Loop diuretics, by contrast, act in the ascending limb of loop of Henle on the Na+/K+/2Cl cotransporter and lead to impairment of both urinary concentrating and diluting capacity and are less likely to lead to hyponatraemia.

Depending on the degree of this woman’s confusion, her immediate management could either be with the administration of 0.9% sodium chloride or with the administration of a 100 ml 3% sodium chloride bolus as described in Box Her thiazide should be stopped, (almost) needless to say.

Depending on definition, about 10% of patients develop hyponatraemia when given a thiazide, and older people are particularly susceptible. A proposed measure to detect those who might be retaining water and thereby becoming hyponatraemic is to have the patient weigh themselves before and 48 h after starting the medication. If they fail to lose weight, or they actually gain weight, then the medication should be stopped and serum electrolytes checked; and all patients given thiazide diuretics should have their electrolytes measured after the onset of therapy or dose adjustments.


Hypernatraemia, defined as serum sodium of greater than 145 mmol/litre, is a commonly encountered problem. It occurs when water intake is inadequate to keep up with water losses, and, since the thirst mechanism is such a powerful stimulus, restricted access to water is nearly always necessary for its development. This occurs in a variety of settings, usually in the very young or very old, or in patients whose illness inhibits their access to water. Several other clinical factors typically seen in the hospital setting can contribute to hypernatraemia, including water losses due to solute diuresis (typically urea or glucose), loop diuretics, gastrointestinal fluid losses, and excessive hypertonic sodium bicarbonate administration. Most patients who develop hypernatraemia have some combination of factors that lead to both impaired access to water and ongoing free-water losses. Hypernatraemia is common in the hospital setting and is frequently iatrogenic during critical illness, typically involving the failure to recognize significant water losses in the urine and to provide the appropriate amount of replacement in either parenteral or enteral solutions.


When water intake falls below the level of ongoing water losses, the relative amount of exchangeable electrolytes in the body compared with water increases, and this leads to hypernatraemia. The thirst mechanism and the kidney’s ability to concentrate the urine are the defences against this. However, in patients with normal mental status it is rare for hypernatraemia to develop, irrespective of the degree of ongoing water losses, if access to water is not limited because the thirst mechanism will lead to increased water intake to match ongoing losses. The common causes of hypernatraemia are shown in Table

Table Common causes of hypernatraemia

Lack of water intake

  • Decreased thirst, e.g. dementia, neurological impairment

  • Bowel rest/nasogastic suction

  • Dependent on others, e.g. mechanical ventilation, infants

Increased water losses

  • Solute diuresis, e.g. hyperglycaemia, urea loading from tube feeds, or hyperalimentation

  • Loop diuretics

  • Gastrointestinal water losses

  • Diabetes insipidus

Hypernatraemia leads to osmolar forces that cause movement of water out of cells, which in particular subjects the brain to stress that can lead to significant damage. The brain attempts to counteract the osmolar stress during hypernatraemia through a series of adaptations, the principal among these being accumulation of osmotically active ions and de novo generation of osmotically active idiogenic osmoles. The earliest response involves accumulation of the osmotically active cations sodium and potassium. Idiogenic osmoles are a heterogeneous group of substances—including glycerophosphocholine, choline, myoinositol, and sorbitol—that are generated intracellularly to exert an osmotic effect and counteract osmotic forces favouring water removal from the cells. These responses are seen very quickly, and after 1 week of hypernatraemia no further changes in brain osmolality are observed. They serve to maintain brain volume during elevations in serum osmolality and prevent significant decrease in brain size due to osmotic water losses. During correction of chronic hypernatraemia it must noted that idiogenic osmoles are not rapidly dissipated, and correction of chronic hypernatraemia over 24 h can lead to cerebral oedema. For this reason, chronic hypernatraemia should be treated cautiously to prevent the development of cerebral oedema.

Clinical manifestations

Clinical manifestations are mainly related to central nervous system dysfunction as cerebral dehydration and cell shrinkage occurs. Hypernatraemia, perhaps owing to the underlying conditions that lead to its development, is associated with an overall mortality between 40 and 70%. Groups at particular risk for complications and poor outcomes from hypernatraemia are older people and patients with endstage liver disease. In the latter case, the use of lactulose in the treatment of hepatic encephalopathy frequently leads to an osmotic diarrhoea and significant water losses in the stool: if this is not appreciated and free water is not given (many encephalopathic patients are obtunded and unable to drink), then hypernatraemia can develop quickly and lead to severe morbidity. It is therefore mandatory to monitor the serum electrolytes closely in this setting, particularly given that patients with liver disease are at increased risk for cerebral demyelination during changes in the serum sodium.

Diagnostic approach to hypernatraemic patients

The first step in evaluating a patient with hypernatraemia is to take a detailed history focusing on fluid intake and losses. Various potential sources of water loss need to be assessed. This is generally straightforward in the outpatient setting, where these are mainly in the urine, but in the patient in hospital many sources of water losses may need to be considered (Fig. from the gastrointestinal tract (diarrhoea, nasogastric suction, bowel fistulae), from the urine, and from insensible losses (fever, sepsis, massive diaphoresis, burns). Whenever practical, these losses should be calculated or estimated. To assess urinary water losses it is necessary to measure the urinary cationic electrolytes (sodium and potassium) and the urinary osmolality, these pieces of information giving complementary but different information. However, a word of caution is necessary in the interpretation of urinary osmolality as errors are frequent in this area. The urinary osmolality alone cannot always determine the presence or absence of electrolyte-free water losses in the urine, the reason for this being that water can be excreted with nonelectrolyte osmoles or with electrolyte osmoles. Both of these contribute to the osmolality of the urine, but their excretion will have different effects on water balance. Recall the relationship that the serum sodium is proportional to total body electrolytes relative to total body water (Equation Therefore, when water is excreted with very few electrolytes, the loss of water is in excess of the loss of electrolytes and hypernatraemia can develop if this water is not adequately replaced. This situation of a high urine osmolality, but very few electrolytes in the urine, most typically occurs when there is a significant amount of urea or glucose (e.g. with poorly controlled diabetes) in the urine. By contrast, when water is excreted with a significant amount of electrolyte osmoles, this will tend not to affect the serum sodium, as long as the concentration of electrolytes in the urine and serum are similar, because loss of water is proportional to the loss of electrolytes and therefore the value of the serum sodium does not change.

Fig. Sources of water intake and loss.

Sources of water intake and loss.

When there is a high urea or glucose load, tremendous quantities of water can be lost in the urine despite maximal urinary concentration. This is what occurs during a solute diuresis and such a patient is typically polyuric. However, if there is a failure to concentrate the urine during a time of hypernatraemia when the patient does not have a solute diuresis, then this should raise suspicion of a urinary concentrating defect. The most common causes of such urinary concentrating defects are renal failure, loop diuretics, tubulointerstitial renal disease, and diabetes insipidus.

Treatment of hypernatraemia

Patients with hypernatraemia typically have significant intravascular volume depletion, hence the initial goal of treatment is to restore this, which is best accomplished with 0.9% saline or colloid. Focus then switches to correction of the serum sodium with free-water replacement (Box The rate of fluid administration required by the patient will depend significantly on the degree of ongoing water losses, so that the appropriate amount of replacement water can be given for these in addition to that required for correction of the hypernatraemic state. If there are extrarenal fluid losses, these will need to be estimated because accurate monitoring is typically not possible, and it is necessary to assess any ongoing water losses in the urine to determine whether the kidneys are appropriately conserving water, or whether they are inappropriately continuing to excrete it. As described previously, electrolyte-free water losses in the urine can be calculated with the formula:

where [Na+]u is urinary sodium concentration, [K+]u is urinary potassium concentration, [Na+]se is serum sodium concentration, and [K+]se is serum potassium concentration.

Patients with hypernatraemia may be insulin resistant such that hyperglycaemia can result if dextrose-containing solutions are given. For this reason, glucose-containing solutions are potentially harmful and should be avoided if possible, but if they must be used (e.g. 5% dextrose in water), then plasma glucose should be monitored closely. When possible, the enteral route should be used before use of parenteral fluid administration.

As with the treatment of patients with symptomatic hyponatraemia, patients with neurological impairment due to hypernatraemia require serial measurement of electrolytes, every 2 h, until they are neurologically stable. In patients without evidence of encephalopathy, the serum sodium should not be corrected more quickly than 0.5 to 1 mmol/litre per hour or 15 mmol/litre over 24 h, and, in severe cases (serum sodium above 170 mmol/litre), sodium should not be corrected to below 150 mmol/litre in the first 48 to 72 h.

Case discussions

The evaluation of a polyuric patient and differentiation of primary polydipsia, central diabetes insipidus, nephrogenic diabetes insipidus (Table, and hypernatraemia due to a solute diuresis can be complex and daunting to the general physician, but should be approached as described in the following case studies.

Table Causes of polyuria-polydipsia syndromes

Cranial diabetes insipidus


Autosomal dominant inheritance

DIDMOAD* (autosomal recessive)



Inflammatory (lymphophocytic infiltration, sarcoidosis, histiocytosis X autoimmunity, Guillain Barré syndrome)

Trauma (neurosurgery, head injury)

Neoplasmsa (craniopharyngioma, germinoma, pinealoma, hypothalamic metastasis, large pituitary tumour)

Infection (meningitis, encephalitis)

Vascular (sickle cell anaemia, aneurysms of anterior communicating artery, Sheehan’s syndrome)

Pregnancy (associated with vasopressinase)

Nephrogenic diabetes insipidus


X-linked inheritance

Autosomal recessive inheritance



Metabolic (hypercalcaemia, hypokalaemia)

Vascular (sickle cell disease)

Osmotic diuresis (glycosuria, post-obstructive uropathy)

Chronic renal disease (renal failure, amyloid, myeloma, sarcoidosis, pyelonephritis)

Drugs (lithium, demeclocycline, amphotercin, glibenclamide, methofluorane)

Primary polydipsia

Unknown aetiology

Psychogenic (compulsive water drinking)

Psychotic (schizophrenia, mania)


Granuloma (sarcoidosis)


TB meningitis

Multiple sclerosis

Drugs (phenothiazines, tricyclic antidepressants)

*DIDMOAD=diabetes insipidus, diabetes mellitus, optic atrophy, deafness (Wolfram syndrome).

Case 6: central diabetes insipidus

A forty-five year old man weighing about 70 kg is involved in a motor vehicle accident and suffers a closed head injury, following which he is admitted to the intensive care unit. He is administered large amounts of 0.9% sodium chloride for fluid resucitation and then prescribed continued 0.9% sodium chloride as maintenance fluid. He develops raised intracranial pressure with evidence of cerebral oedema, which is treated by placement of an extraventricular device and infusion of 3% sodium chloride. He then develops polyuria with urine output exceeding 500 ml/hr. His serum sodium increases to 184 mEq/L, with a spot urinary osmolality being 120 mOsm/kg/H2O and the combined urinary sodium plus urinary potassium concentrations being 50 mEq/L. A continuous dDAVP infusion is started and the urine osmolality increases to 800 mOsm/kg/H2O.


This patient’s hypernatremia is multifactorial. There is a brisk free water diuresis, and the patient is receiving both 0.9% saline and hypertonic saline, which in combination with urinary retention of sodium has resulted in severe hypernatremia. The initiation of dDAVP has stopped the free water diuresis and so the hypernatremia should not worsen, indeed—because the patient has a fixed inability to excrete free water while on dDAVP—even 0.9% saline could in theory lead to a fall in serum sodium concentration, and this should be administered at a restricted rate of 50 ml/hr to maintain water and sodium homeostasis.

The optimum rate of correction of hypernatremia is difficult to determine. A serum sodium of > 180 mmol/L could lead to brain injury, but a fall in the serum sodium could aggravate cerebral oedema. In this situation, the rate of sodium correction should probably not exceed 10 mmol/L/24 hr, and 5 mmol/24 hr should be initially attempted. If intracranial pressures increase with correction of hypernatremia, a 100 ml bolus of 3% sodium chloride should be administered to acutely raise the serum sodium and decrease the cerebral edema. The free water required to correct the serum sodium by 5 mEq is 1.25 L (See Box [5/140] X 35) or about 50 ml/hr over 24 hours. An appropriate management strategy would therefore be (in addition to giving the 0.9% saline described above) to administer 50 ml/hr of 5% dextrose in water, checking the serum sodium every two hours. If the serum sodium were to fall faster than anticipated or the intracranial pressure to rise, the rate of the free water infusion would be adjusted accordingly. The rate of correction may end up being greater than predicted if a natriuresis were to ensue.

Case 7: diarrheal dehydration

A 77 year old woman who is a nursing home resident develops vomiting and diarrhea for three days, presenting to the Emergency Department with intravascular volume depletion and dehydration. On presentation her weight is 46 kg, down from 50 kg just two weeks earlier. Her blood pressure is 90/40 mmHg with a pulse of 126/min. Biochemical testing reveals serum sodium 156 mEq/L, potassium 5.6 mEq/L, blood urea nitrogen 18 mmol/l (50 mg/dl) and creatinine 130 μ‎ mol/l (1.4 mg/dl). Urinary tests reveal sodium < 5 mEq/L, potassium 20 mEq/L and osmolality 800 mOsm/kg/H2O.


This woman has an estimated volume deficit of approximately 4 liters, based on her recent weight loss, which would have a composition of approximately 0.9% NaCl (154 mEq/L). To correct her serum sodium by 10 mmol/L in 24 hours, approximately 1.8L of free water would have to be administered (See Box [10/140] × 25). She should initially be given 2L of 0.9% saline to acutely correct her intravascular volume depletion and restore circulatory perfusion. This would leave the remaining deficit of approximately 2L of isotonic fluid to be corrected over 24 hours. Her typical maintenance fluid requirement would otherwise be about 2L for the next 24 hours, which with the addition to 2L of deficit therapy would result in a total volume of 4L or 166 ml/hr. A total of 4L of 0.45% NaCl with 2.5% dextrose in water would provide the equivalent of 2L of free water and 2L of 0.9% saline and would be adequate therapy to correct both the remaining volume deficit and free-water deficit and to provide for urinary losses.

Polyuria and polydipsia

The evaluation of a polyuric patient and differentiation of primary polydipsia, central diabetes insipidus, nephrogenic diabetes insipidus (Table, and hypernatraemia due to a solute diuresis can be complex and daunting to the general physician, but should be approached as described in the following case studies.



Joint involvement

1 large joint (‘large’ refers to shoulder, elbows, hips, knees and ankles)


2-10 large joints


1-3 small joints (with or without involvement of large joints) (small refers to MCP, PIP, 2-5 MTP, wrists)

4-10 small joints (with or without involvement of large joints


>10 joints (at least 1 small joint)



Serology (at least 1 test result is needed for classification)

Negative RF and negative ACPA


Low-positive RF or low-positive ACPA


High-positive RF or high-positive ACPA



Acute-phase reactants (at least 1 test result is needed for classification)

Normal CRP and normal ESR


Abnormal CRP or abnormal ESR



Duration of symptoms

<6 weeks


≥6 weeks


From Aletaha D, et al (2010). 2010 Rheumatoid Arthritis Classification Criteria. Arthritis Rheum, 62, 2569–81. Reprinted with permission of John Wiley and Sons Inc.

Case 8: primary polydipsia

A 27-year-old man with schizophrenia is being evaluated prior to admission to a psychiatric hospital. His only complaint is of frequent urination, approximately 15 times per day according to his family, and that he is always thirsty. He has had no recent seizures and his level of consciousness is normal. Routine physical examination is unremarkable. Laboratory values are as follows:

  1. 1 serum—sodium 131 mmol/litre, potassium 4.0 mmol/litre, chloride 96 mmol/litre, bicarbonate 24 mmol/litre, urea 5.7 mmol/litre (16 mg/dl), creatinine 118 µmol/litre (1.3 mg/dl), glucose 5.4 mmol/litre (98 mg/dl)

  2. 2 urine—sodium 10 mmol/litre, potassium 8 mmol/litre, osmolality 65 mOsm/kg


This patient is very likely to be polyuric, given the history. Blood tests reveal mild hyponatraemia and near-normal renal function (CKD stage 3). Urinary parameters are consistent with a water diuresis. Calculation of the electrolyte-free water clearance shows that he is losing significant amounts of water in the urine (Equation ([Na+]u + [K+]u)/([Na+]se + [K+]se) = (10 + 8)/(131 + 4) = 0.13. This means that 87% of the patient’s urine output is electrolyte-free water. The low urinary osmolality signifies that this is a water diuresis, rather than being driven by the presence of nonelectrolyte solute, e.g. glucose.

The question now becomes whether the water diuresis is an appropriate response to excessive water intake or whether it is is pathological, leading to excessive water losses that must then be replaced. In this case, the most likely answer is excessive water intake because of the hyponatraemia and decreased serum osmolality. If a urine concentrating defect were the primary cause of the polyuria, then the patient should not be hyponatraemic unless they had both a urinary concentrating defect and excessive water intake.

Case 9: primary polydipsia vs diabetes insipidus

A 39-year-old mother is concerned because her 12-year-old daughter has noted frequent urination and says that she is always thirsty. The patient is a well-adjusted adolescent with no past medical history and normal development up to this point. Her physical examination is normal. She also has had no recent seizures and her level of consciousness is normal. Serum electrolytes and the results of a 24-h urinary collection are as follows:

  • serum—sodium 140 mmol/litre, potassium 4.5 mmol/litre, chloride 103 mmol/litre, bicarbonate 25 mmol/litre, urea 5.4 mmol/litre (15 mg/dl), creatinine 109 µmol/litre (1.2 mg/dl), glucose 5.7 mmol/litre (103 mg/dl)

  • urine (24 h)—total volume 9 litres, sodium 15 mmol/litre, potassium 8 mmol/litre, osmolality 70 mOsm/kg


The 24-h urinary collection volume clearly demonstrates that this girl is polyuric, and the urinary studies—similar to Case 8—are consistent with a water diuresis. However, the patient is normonatraemic and thus the serum electrolytes are not helpful in reaching a diagnosis: based on the information that we currently have, it is impossible to tell whether her polyuria is due to excessive water intake or to a urinary concentrating defect, which is an important determination to make in this seemingly healthy adolescent. In order to distinguish between these two possibilities, a water deprivation test can be performed. This is usually done in a hospital setting because a patient with diabetes insipidus can rapidly develop hypernatraemia if water intake is restricted. The details of different protocols for water deprivation tests are beyond the scope of this chapter, but a typical test and its interpretation are shown in Box, the basic principle being that if a patient with diabetes insipidus is deprived of water and allowed to become mildly hypernatraemic, then such a patient will not have concentrated urine at that time. By contrast, a patient with primary polydipsia will begin to concentrate the urine if allowed to become mildly hypernatraemic.

Case 10: nephrogenic diabetes insipidus

A 41-year-old man presents for a routine physical examination. His past medical history is significant only for bipolar disorder, for which he has taken lithium carbonate for the last 15 years. This information leads to further questioning, and he admits to frequent urination and excessive thirst, but denies any symptoms of hesitancy or dysuria. His physical examination is normal. Serum chemistry profile and urine studies are as follows:

  • serum—sodium 147 mmol/litre, potassium 3.8 mmol/litre, chloride 110 mmol/litre, bicarbonate 26 mmol/litre, urea 5.4 mmol/litre (15 mg/dl), creatinine 73 µmol/litre (0.8 mg/dl), glucose 6.9 mmol/litre (124 mg/dl)

  • urine—sodium 25 mmol/litre, potassium 22 mmol/litre, osmolality 160 mOsm/kg


The laboratory data in this case are most consistent with diabetes insipidus. The history suggests polyuria, and, as nephrogenic diabetes insipidus is a complication of lithium therapy, it is appropriate to rule out this diagnosis in a patient such as this. In contrast with Case 9, we have a definite differentiation between primary polydipsia and diabetes insipidus because the serum sodium is mildly elevated and the urinary osmolality is simultaneously low, which confirms the diagnosis of diabetes insipidus. In a sense, by demonstrating the failure to concentrate the urine despite having hypernatraemia, we have the results of a water deprivation test. However, it is important to note that—based solely on the information given above—it is not known whether the patient has central or nephrogenic diabetes insipidus, although the latter would clearly be anticipated in a patient taking lithium. To make this distinction would require formalized testing to assess the response to exogenously administered DDAVP: if the patient fails to concentrate the urine following administration of DDAVP, then they have nephrogenic diabetes insipidus.

Case 11: central diabetes insipidus

A 28-year-old man is brought to the Emergency Department by ambulance following a car accident during which he sustained severe head trauma. His past medical history is significant only for asthma as a child and a previous appendicectomy. He is taken immediately to surgery for evacuation of an acute epidural haematoma. During the course of surgery, his urinary output increases from 35 ml/h to over 300 ml/h. Blood tests taken immediately on admission to hospital reveal serum sodium 141 mmol/litre; the findings on serum chemistry profile and urine studies taken just after surgery when he arrives on the intensive care unit are as follows:

  • serum—sodium 148 mmol/litre, potassium 4.5 mmol/litre, chloride 112 mmol/litre, bicarbonate 26 mmol/litre, urea 6.8 mmol/litre (19 mg/dl), creatinine 127 µmol/litre (1.4 mg/dl), glucose 6.4 mmol/litre (115 mg/dl)

  • urine—sodium 17 mmol/litre, potassium 13 mmol/litre, osmolality 120 mOsm/kg


The history is highly suggestive of central diabetes insipidus due to head trauma as the cause of polyuria. The urinary studies, as in the three previous cases, show a water diuresis, and, as in Case 10, we have a definite diagnosis of diabetes insipidus because the patient is simultaneously hypernatraemic and undergoing a water diuresis. Again, based solely on the information above, it is not possible to say whether the patient has central or nephrogenic diabetes insipidus, but since the history is so suggestive of a central cause it is prudent to simply administer DDAVP and assess the response. If the patient fails to concentrate the urine following administration of DDAVP, then he has nephrogenic diabetes insipidus, whereas if the urine becomes concentrated—as would be anticipated in this case—then the diagnosis is central diabetes insipidus. When DDAVP is administered, water intake should be adjusted appropriately to avoid precipitation of significant hyponatraemia, and serum electrolytes should be monitored closely during dose titration.

Central diabetes insipidus should always be suspected when the urine is not concentrated in the setting of hypernatraemia. Severe hypernatraemia can develop rapidly in an individual who has restricted access to fluids, such as a patient in an intensive care unit, and hence early recognition is vital.

Case 12: solute diuresis from excess urea load

A 58-year-old man with long history of alcohol abuse and chronic liver disease is admitted with necrotizing pancreatitis. Among other manoeuvres, a urinary catheter is inserted, which demonstrates that his urinary output is 30 ml/h. Admission laboratory test results are as follows:

  • serum—sodium 138 mmol/litre, potassium 3.9 mmol/litre, chloride 103 mmol/litre, bicarbonate 21 mmol/litre, urea 11.8 mmol/litre (33 mg/dl), creatinine 136 µmol/litre (1.5 mg/dl)

The patient is ordered to have no enteral intake overnight, and he receives 5 litres of 0.9% (normal) saline volume expansion. His abdominal pain worsens 24 h after admission and he is continued without enteral intake. Repeat laboratory tests show that his serum sodium has risen to 146 mmol/litre. Over the following 24 h his urinary output increases and infusion of 0.9% saline is continued at 100 ml/h. Total parenteral nutrition is initiated with a daily regimen that comprises a total volume of 1.5 litres, including 120 mmol of sodium and high amino acid content. Repeat laboratory tests are as follows:

  • serum—sodium 151 mmol/litre, potassium 3.2 mmol/litre, chloride 117 mmol/litre, bicarbonate 26 mmol/litre, urea 22.5 mmol/litre (63 mg/dl), creatinine 100 µmol/litre (1.1 mg/dl), glucose 7.0 mmol/litre (126 mg/dl)

  • urine—volume 150 ml/hour; sodium 50 mmol/litre, potassium 13 mmol/litre, osmolality 650 mOsm/kg


What has occurred in this case is very typical of a solute diuresis leading to hypernatraemia in the critical care setting. The patient is significantly polyuric and has become progressively more and more hypernatraemic. It is important to recognize that, in contrast to previous cases discussed, the urinary osmolality is high, meaning that ADH activity is present and it must be concluded that the patient is losing ‘free water’ (which has to be the situation because the serum sodium is increasing in the absence of administration of any hypertonic sodium solution).

The loss of free water occurring at the same time that the urine is highly concentrated may appear paradoxical, but the answer is evident when the electrolyte-free water is calculated. The ratio of the (sodium + potassium) in the urine to the (sodium + potassium) in the serum is 63/159 = 0.39, hence at his current urinary output he is losing water at a rate of (0.61 × 150) or 91.5 ml/h. Water replacement must be given at least equal to this rate to replace the ongoing urinary water losses. The low urinary sodium and potassium at a time when the urine osmolality is high signifies that there must be a nonelectrolyte osmole in the urine that is ‘obligating’ water loss. The patient is undergoing an osmotic diuresis secondary to a high urea load, this probably being secondary both to the hypercatabolic state of critical illness/stress (protein breakdown is increased, leading to significant urea generation) and to the high amino acid content of the total parenteral nutrition. This scenario is commonly seen in critical illness and is easily preventable if the responsible clinician appreciates the possibility of free-water loss in a patient who becomes polyuric.

Further reading

Abbott NJ, Ronnback L, Hansson E (2006). Astrocyte–endothelial interactions at the blood–brain barrier. Nat Rev Neurosci, 7, 41–53.Find this resource:

Achinger SG, Moritz ML, Ayus JC (2006). Dysnatremias: why are patients still dying? South Med J, 99, 1–12.Find this resource:

Agre P, et al. (1993). Aquaporin CHIP: the archetypal molecular water channel. Am J Physiol, 265, F463–76.Find this resource:

Amiry-Moghaddam M, Ottersen OP (2003). The molecular basis of water transport in the brain. Nat Rev Neurosci, 4, 991–1001.Find this resource:

Arieff AI, Ayus JC (1993). Endometrial ablation complicated by fatal hyponatremic encephalopathy. JAMA, 270, 1230–2.Find this resource:

Arieff AI, Ayus JC (1993). Pathogenesis of hyponatremic encephalopathy. Current concepts. Chest, 103, 607–10.Find this resource:

Arieff AI, Ayus JC, Fraser CL (1992). Hyponatraemia and death or permanent brain damage in healthy children. BMJ, 304, 1218–22.Find this resource:

Ayus JC, Achinger SG, Arieff AI (2008). Brain cell volume regulation in hyponatremia: role of sex age, vasopressin, and hypoxia. Am J Renal Physiol, 295, F619–F624.Find this resource:

Ayus JC, Arieff AI (1995). Pulmonary complications of hyponatremic encephalopathy. Noncardiogenic pulmonary edema and hypercapnic respiratory failure. Chest, 107, 517–21.Find this resource:

Ayus JC, Arieff AI (1999). Chronic hyponatremic encephalopathy in postmenopausal women: association of therapies with morbidity and mortality. JAMA, 281, 2299–304.Find this resource:

Ayus JC, Arieff A, Moritz ML (2005). Hyponatremia in marathon runners. N Engl J Med, 353, 427–8.Find this resource:

Ayus JC, Armstrong D, Arieff A (2006). Hyponatremia with hypoxia: effects on brain adaptation, perfusion and histology in rodents. Kidney Int,69, 1319–25.Find this resource:

Ayus JC, Krothapalli RK, Arieff AI (1987). Treatment of symptomatic hyponatremia and its relation to brain damage. A prospective study. N Engl J Med, 317, 1190–5.Find this resource:

Ayus JC, Moritz ML (2010). Bone Disease as a New Complication of Hyponatremia: Moving Beyond Brain Injury. Clin J Am Soc Nephrol, 5, 167–8.Find this resource:

Ayus JC, Wheeler JM, Arieff AI (1992). Postoperative hyponatremic encephalopathy in menstruant women. Ann Intern Med, 117, 891–7.Find this resource:

Dolman D, et al. (2005). Induction of aquaporin 1 but not aquaporin 4 messenger RNA in rat primary brain microvessel endothelial cells in culture. J Neurochem, 93, 825–33.Find this resource:

Heilig CW, et al. (1989). Characterization of the major brain osmolytes that accumulate in salt-loaded rats. Am J Physiol, 257, F1108–16.Find this resource:

Hew-Butler T, et al. (2005). Consensus statement of the 1st International Exercise-Associated Hyponatremia Consensus Development Conference, Cape Town, South Africa 2005. Clin J Sport Med, 15, 208–13.Find this resource:

Lien YH, Shapiro JI, Chan L (1990). Effects of hypernatremia on organic brain osmoles. J Clin Invest, 85, 1427–35.Find this resource:

Manley GT, et al. (2000). Aquaporin-4 deletion in mice reduces brain edema after acute water intoxication and ischemic stroke. Nat Med, 6, 159–63.Find this resource:

Moritz ML, Ayus JC (2003). Prevention of hospital-acquired hyponatremia: a case for using isotonic saline. Pediatrics, 111, 227–30.Find this resource:

Moritz ML, Ayus JC (2003). The pathophysiology and treatment of hyponatraemic encephalopathy: an update. Nephrol Dial Transplant, 18, 2486–91.Find this resource:

Moritz ML, Ayus JC (2004). Dysnatremias in the critical care setting. Contrib Nephrol, 144, 132–57.Find this resource:

Moritz ML, Ayus JC (2007). Hospital-acquired hyponatremia--why are hypotonic parenteral fluids still being used? Nat Clin Pract Nephrol 3, 374–82.Find this resource:

Moritz ML, Ayus JC (2008). Exercise-associated hyponatremia: why are athletes still dying? Clin J Sport Med 18, 379–81.Find this resource:

Moritz ML, Ayus JC (2009). New aspects in the pathogenesis, prevention, and treatment of hyponatremic encephalopathy in children. Pediatr Nephrol 25, 1225–38.Find this resource:

Moritz ML, Ayus JC. (2010). 100 cc 3% sodium chloride bolus: a novel treatment for hyponatremic encephalopathy. Metab Brain Dis, 25, 91–6.Find this resource:

Nzerue CM, et al. (2003). Predictors of outcome in hospitalized patients with severe hyponatremia. J Natl Med Assoc, 95, 335–43.Find this resource:

Papadopoulos MC, Verkman AS. Aquaporin-4 and brain edema. Pediatr Nephrol, 22, 778–84.Find this resource:

Schrier RW, et al. (2006). Tolvaptan, a selective oral vasopressin V2-receptor antagonist, for hyponatremia. N Engl J Med, 355, 2099–112.Find this resource:

Simard M, Nedergaard M (2004). The neurobiology of glia in the context of water and ion homeostasis. Neuroscience, 129, 877–96.Find this resource: