Brenner and Rector's The Kidney, 8th ed.

CHAPTER 13. Disorders of Water Balance

Joseph G. Verbalis   Tomas Berl



Body Fluids: Compartmentalization, Composition, and Turnover, 459



Water Metabolism, 460



Arginine Vasopressin Synthesis and Secretion, 461



Thirst, 468



Integration of Arginine Vasopressin Secretion and Thirst, 469



Disorders of Insufficient Arginine Vasopressin or Arginine Vasopressin Effect, 469



Central Diabetes Insipidus, 469



Osmoreceptor Dysfunction, 473



Gestational Diabetes Insipidus, 475



Nephrogenic Diabetes Insipidus, 475



Primary Polydipsia, 477



Clinical Manifestations, 477



Differential Diagnosis, 478



Therapy, 480



Disorders of Excess Arginine Vasopressin or Arginine Vasopressin Effect, 484



Variables That Influence Water Excretion, 484



Etiology of Hyponatremia, 485



Hyponatremia with Extracellular Fluid Volume Depletion, 486



Hyponatremia with Excess Extracellular Fluid Volume, 487



Hyponatremia with Normal Extracellular Fluid Volume, 488



Syndrome of Inappropriate Antidiuretic Hormone Secretion, 491

Disorders of body fluids are among the most commonly encountered problems in clinical medicine, largely because many different disease states can potentially disrupt the finely balanced mechanisms that control the intake and output of water and solute. Because body water is the primary determinant of the osmolality of the extracellular fluid (ECF), disorders of water metabolism can be broadly divided into hyperosmolar disorders, in which there is a deficiency of body water relative to body solute, and hypoosmolar disorders, in which there is an excess of body water relative to body solute. Because sodium is the main constituent of plasma osmolality, these disorders are typically characterized by hypernatremia and hyponatremia, respectively. Before discussing specific aspects of these disorders, this chapter first briefly reviews the regulatory mechanisms underlying water metabolism, which, in concert with sodium metabolism, maintains body fluid homeostasis.


Water constitutes approximately 55% to 65% of body weight (BW), varying with age, sex, and amount of body fat, and therefore constitutes the largest single constituent of the body. Total body water (TBW) is distributed between the intracellular fluid (ICF) and the ECF compartments. Estimates of the relative sizes of these two pools differ significantly depending on the tracer used to measure the ECF volume, but most studies in animals and humans have indicated that 55% to 65% of TBW resides in the ICF and 35% to 45% in the ECF. Approximately 75% of the ECF compartment is interstitial fluid and only 25% is intravascular fluid (i.e., blood volume). [1] [2] Figure 13-1summarizes the estimated body fluid spaces of an average weight adult.



FIGURE 13-1  Schematic representation of body fluid compartments in humans. The shaded areas depict the approximate size of each compartment as a function of body weight. The numbers indicate the relative sizes of the various fluid compartments and the approximate absolute volumes of the compartments (in liters) in a 70-kg adult. ECF, extracellular fluid; ICF, intracellular fluid; ISF, interstitial fluid; IVF, intravascular fluid; TBW, total body water.  (From Verbalis JG: Body water and osmolality. In Wilkinson B, Jamison R [eds]: Textbook of Nephrology. London, Chapman & Hall, 1997, pp 89-94.)




The solute composition of the ICF and ECF differs considerably because most cell membranes possess multiple transport systems that actively accumulate or expel specific solutes. Thus, membrane-bound Na+/K+-ATPase maintains Na+ in a primarily extracellular location and K+ in a primarily intracellular location.[3] Similar transporters effectively result in confining Cl- largely to the ECF and Mg2+, organic acids, and phosphates to the ICF. Glucose, which requires an insulin-activated transport system to enter most cells, is present in significant amounts only in the ECF because it is rapidly converted intracellularly to glycogen or metabolites.[4] HCO3- is present in both compartments, but is approximately three times more concentrated in the ECF. Urea is unique among the major naturally occurring solutes in that it diffuses freely across most cell membranes[5]; therefore, it is present in similar concentrations in virtually all body fluids, except in the renal medulla where it is concentrated by urea trans-porters (see Chapter 9 ).

Despite very different solute compositions, both the ICF and the ECF have an equivalent osmotic pressure,[6] which is a function of the total concentration of all solutes in a fluid compartment because most biologic membranes are semipermeable (i.e., freely permeable to water but not to aqueous solutes). Thus, water will flow across membranes into a compartment with a higher solute concentration until a steady state is reached at which the osmotic pressures have equalized on both sides of the cell membrane.[7] An important consequence of this thermodynamic law is that the volume of distribution of body Na+ and K+ is actually the TBW rather than just the ECF or ICF volume, respectively.[8] For example, any increase in ECF sodium concentration (Na+) will cause water to shift from the ICF to the ECF until the ICF and ECF osmotic pressures are equal, thereby effectively distributing the Na+ across both extracellular and intracellular water.

Osmolality is defined as the concentration of all of the solutes in a given weight of water. The total solute concentration of a fluid can be determined and expressed in several different ways. The most common method is to measure its freezing point or vapor pressure, because these are colligative properties of the number of free solute particles in a volume of fluid, [9] [10] and to express the result relative to a standard solution of known concentration using units of either osmolality (milliosmoles of solute per kilogram of water, mOsm/kg H2O), or osmolarity (milliosmoles of solute per liter of water, mOsm/L H2O). Plasma osmolality can be measured directly as described previously or calculated by summing the concentrations of the major solutes present in the plasma:

Posm (mOsm/kg H2O)=2×plasma Na+ (mEq/L)±glucose (mg/dL)/18± BUN (mg/dL)/2.8

where BUN is the blood urea nitrogen. Both methods produce comparable results under most conditions (the value obtained using this formula is generally within 1% to 2% of that obtained by direct osmometry), as will simply doubling the plasma Na+ because sodium and its accompanying anions are the predominant solutes present in plasma. However, the total osmolality of plasma is not always equivalent to the effective osmolality, often referred to as the tonicity of the plasma, because the latter is a function of the relative solute permeability properties of the membranes separating the two compartments. Solutes that are impermeable to cell membranes (e.g., Na+, mannitol) are restricted to the ECF compartment and are effective solutes because they create osmotic pressure gradients across cell membranes, leading to osmotic movement of water from the ICF to the ECF compartments. Solutes that are permeable to cell membranes (e.g., urea, ethanol, methanol) are ineffective solutes because they do not create osmotic pressure gradients across cell membranes and therefore are not associated with such water shifts.[11] Glucose is a unique solute because at normal physiologic plasma concentrations, it is taken up by cells via active transport mechanisms and therefore acts as an ineffective solute, but under conditions of impaired cellular uptake (e.g., insulin deficiency), it becomes an effective extracellular solute.[12]

The importance of this distinction between total and effective osmolality is that only the effective solutes in plasma are determinants of whether clinically significant hyperosmolality or hypoosmolality is present. An example of this is uremia: A patient with a urea concentration that has increased by 56 mg/dL will have a corresponding 20 mOsm/kg H2O elevation in plasma osmolality, but the effective osmolality will remain normal because the increased urea is proportionally distributed across both the ECF and the ICF. In contrast, a patient whose plasma Na+ has increased by 10 mEq/L will also have a 20 mOsm/kg H2O elevation of plasma osmolality, because the increased cation must be balanced by an equivalent increase in plasma anions, but in this case, the effective osmolality will also be elevated by 20 mOsm/kg H2O because the Na+ and accompanying anions will largely remain restricted to the ECF owing to the relative impermeability of cell membranes to Na+ and other ions. Thus, elevations of solutes such as urea, unlike elevations of sodium, do not cause cellular dehydration and, consequently, do not activate mechanisms that defend body fluid homeostasis by increasing body water stores.

Both body water and solutes are in a state of continuous exchange with the environment. The magnitude of the turnover varies considerably depending on physical, social, and environmental factors, but in healthy adults, it averages 5% to 10% of the total body content each day. For the most part, daily intake of water and electrolytes is not determined by physiologic requirements but is more a function of dietary preferences and cultural influences. Healthy adults have an average daily fluid ingestion of approximately 2 to 3 L, but with considerable individual variation; approximately one third of this is derived from food or the metabolism of fat and the rest from discretionary ingestion of fluids. Similarly, of the 1000 mOsm of solute ingested or generated by the metabolism of nutrients each day, nearly 40% is intrinsic to food, another 35% is added to food as a preservative or flavoring, and the rest is mostly urea. In contrast to the largely unregulated nature of basal intakes, the urinary excretion of both water and solute is highly regulated to preserve body fluid homeostasis. Thus, under normal circumstances, almost all ingested Na+, Cl- and K+, as well as both ingested and metabolically generated urea, are excreted in the urine under the control of specific regulatory mechanisms. Other ingested solutes, for example, divalent minerals, are excreted primarily by the gastrointestinal tract. Urinary excretion of water is also tightly regulated by the secretion and renal effects of arginine vasopressin (AVP), which is discussed in greater detail in Chapters 8 and 9 and the following section.


Water metabolism is responsible for the balance between the intake and the excretion of water. Each side of this balance equation can be considered to consist of a regulated and an unregulated component, the magnitudes of which can vary quite markedly under different physiologic and pathophysiologic conditions. The unregulated component of water intake consists of the intrinsic water content of ingested foods, the consumption of beverages primarily for reasons of palatability or desired secondary effects (e.g., caffeine), or for social or habitual reasons (e.g., alcoholic beverages), whereas the regulated component of water intake consists of fluids consumed in response to a perceived sensation of thirst. Studies of middle-aged subjects have shown mean fluid intakes of 2.1 L/24 hr, and analysis of the fluids consumed indicated that the vast majority of the fluid ingested is determined by influences such as meal-associated fluid intake, taste, or psychosocial factors rather than true thirst.[13]

The unregulated component of water excretion occurs via insensible water losses from a variety of sources (cutaneous losses from sweating, evaporative losses in exhaled air, gastrointestinal losses) as well as the obligate amount of water that the kidneys must excrete to eliminate solutes generated by body metabolism, whereas the regulated component of water excretion comprises the renal excretion of free water in excess of the obligate amount necessary to excrete metabolic solutes. Unlike solutes, a relatively large proportion of body water is excreted by evaporation from skin and lungs. This amount varies markedly depending on several factors, including dress, humidity, temperature, and exercise.[14] Under the sedentary and temperature-controlled indoor conditions typical of modern urban life, daily insensible water loss in healthy adults is minimal at approximately 10 ml/kg BW (0.7 L in a 70-kg man or woman). However, insensible losses can increase to twice this level (i.e., 20 ml/kg BW) simply under conditions of increased activity and temperature; and if environmental temperature or activity is even greater, such as in arid environments, the rate of insensible water loss can even approximate the maximal rate of free water excretion by the kidney.[14] Thus, in quantitative terms, insensible loss and the factors that influence it can be just as important to body fluid homeostasis as regulated urine output. Another major determinant of unregulated water loss is the rate of urine solute excretion, which cannot be reduced below a minimal obligatory level required to excrete the solute load. The volume of urine required depends not only on the solute load but also on the degree of antidiuresis. At a typical basal level of urinary concentration (urine osmolality =600 mOsm/kg H2O) and a typical solute load of 900 to 1200 mOsm/day, a 70-kg adult would require a total urine volume of 1.5 to 2.0 L (21-28 ml/kg BW) to excrete the solute load. However, under conditions of maximal antidiuresis (urine osmolality = 1200 mOsm/kg H2O), the same solute load would require a minimal obligatory urine output of only 0.75 to 1.0 L/day, and conversely, a decrease in urine concentration to minimal levels (urine osmolality=60 mOsm/kg H2O) would obligate a proportionally larger urine volume of 15 to 20 L/day to excrete the same solute load.

The previous discussion serves to emphasize that both water intake and water excretion have very substantial un-regulated components, and these can vary tremendously as a result of factors that are unrelated to maintenance of body fluid homeostasis. In effect, the regulated components of water metabolism are those that act to maintain body fluid homeostasis by compensating for whatever perturbations result from unregulated water losses or gains. Within this framework, it is clear that the two major mechanisms responsible for regulating water metabolism are pituitary secretion and renal effects of AVP and thirst, each of which is discussed in greater detail.

Arginine Vasopressin Synthesis and Secretion

The primary determinant of free water excretion in animals and humans is the regulation of urinary water excretion by circulating levels of AVP in plasma. The renal effects of AVP are covered extensively in Chapters 8 and 9 . This chapter focuses on the regulation of AVP synthesis and secretion from the neurohypophysis.

Structure and Synthesis

Before AVP was biochemically characterized, early studies used the general term “antidiuretic hormone” (ADH) to describe this substance. Now that AVP is known to be the only naturally occurring antidiuretic substance, it is more appropriate to refer to it by its correct hormonal designation. AVP is a 9–amino acid peptide synthesized in the hypothalamus. It is a composed of a 6–amino acid ringlike structure formed by a disulfide bridge, with a 3–amino acid tail at the end of which the COOH-terminal group is amidated. Substitution of lysine for arginine in position 8 yields lysine vasopressin, the ADH found in pigs and other members of the suborder Suina. Substitution of isoleucine for phenylalanine at position 3 and of leucine for arginine at position 8 yields oxytocin, a hormone found in all mammals as well as many submammalian species.[15] Oxytocin has weak antidiuretic activity[16] but is a potent constrictor of smooth muscle in mammary glands and uterus. As implied by their names, AVP and lysine vasopressin also cause constriction of blood vessels, which was the property that led to their original discovery in the late 19th century,[17] but this pressor effect occurs only at concentrations many times those required to produce antidiuresis and is probably of little physiologic or pathologic importance in humans except under conditions of severe hypotension and hypovolemia, in which it acts to supplement the vasoconstrictive actions of angiotensin II (Ang II) and the sympathetic nervous system.[18] The multiple actions of AVP are mediated by different G-protein-coupled receptors, designated V, V, and V2.

AVP and oxytocin are produced by the neurohypophysis, often referred to as the posterior pituitary gland because the neural lobe is located centrally and posterior to the adenohypophysis, or anterior pituitary gland, in the sella turcica. However, it is important to understand that the posterior pituitary gland consists only of the distal axons of the magnocellular neurons that compose the neurohypophysis. The cell bodies of these axons are located in specialized (magnocellular) neural cells located in two discrete areas of the hypothalamus, the paired supraoptic (SON) and paraventricular (PVN) nuclei ( Fig. 13-2 ). In adults, the posterior pituitary is connected to the brain by a short stalk through the diaphragm sellae. The neurohypophysis is supplied with blood by branches of the superior and inferior hypophysial arteries, which arise from the posterior communicating and intracavernous portion of the internal carotid artery. In the posterior pituitary, the arterioles break up into localized capillary networks that drain directly into the jugular vein via the sellar, cavernous, and lateral venous sinuses. Many of the neurosecretory neurons that terminate higher in the infundibulum and median eminence originate in parvicellular neurons in the PVN and are functionally distinct from the magnocellular neurons that terminate in the posterior pituitary because they primarily enhance secretion of adrenocorticotropin hormone (ACTH) from the anterior pituitary. AVP-containing neurons also project from parvicellular neurons of the PVN to other areas of the brain, including the limbic system, the nucleus tractus solitarius, and the lateral gray matter of the spinal cord. The functions of these extrahypophysial projections are still under study.



FIGURE 13-2  Summary of the main anterior hypothalamic pathways that mediate secretion of arginine vasopressin (AVP) and oxytocin (OT). The vascular organ of the lamina terminalis (OVLT) is especially sensitive to hyperosmolality. Hyperosmolality also activates other neurons in the anterior hypothalamus, such as those in the subfornical organ (SFO) and median preoptic nucleus (MnPO), and magnocellular neurons, which are intrinsically osmosensitive. Circulating angiotensin II (Ang II) activates neurons of the SFO, an essential site of Ang II action, as well as cells throughout the lamina terminalis and MnPO. In response to hyperosmolality or Ang II, projections from the SFO and OVLT to the MnPO activate excitatory and inhibitory interneurons that project to the supraoptic nucleus (SON) and paraventricular nucleus (PVN) to modulate direct inputs to these areas from the circumventricular organs. Cholecystokinin (CCK) acts primarily on gastric vagal afferents that terminate in the nucleus of the solitary tract (NST), but at higher doses, it can also act at the area postrema (AP). Although neurons are apparently activated in the ventrolateral medulla (VLM) and NST, most neurohypophyseal secretion appears to be stimulated by monosynaptic projections from A2/C2 cells, and possibly also noncatecholaminergic somatostatin/inhibin B cells, of the NST. Baroreceptor-mediated stimuli, such as hypovolemia and hypotension, are more complex. The major projection to magnocellular AVP neurons appears to arise from A1 cells of the VLM that are activated by excitatory interneurons from the NST. Other areas, such as the parabrachial nucleus (PBN), may contribute multisynaptic projections. Cranial nerves IX and X, which terminate in the NST, also contribute input to magnocellular AVP neurons. It is unclear whether baroreceptor-mediated secretion of oxytocin results from projections from VLM neurons or from NST neurons. AC, anterior commissure; OC, optic chiasm; PIT, anterior pituitary.  (From Stricker EM, Verbalis JG: Water intake and body fluids. In Squire LR, Bloom FE, McConnell SK, et al [eds]: Fundamental Neuroscience. San Diego, Academic Press, 2003, pp 1011-1029.)




The genes encoding the AVP and oxytocin precursors are located in close proximity on chromosome 20 but are expressed in mutually exclusive populations of neurohypophyseal neurons.[19] The AVP gene consists of approximately 2000 base pairs and contains three exons separated by two intervening sequences ( Fig. 13-3 ). Each exon encodes one of the three functional domains of the pre-prohormone, although small parts of the nonconserved sequences of neurophysin are located in the first and third exons that code for AVP and the C-terminal glycoprotein, respectively. The untranslated 5′-flanking region, which regulates expression of the gene, shows extensive sequence homology across several species but is markedly different from the otherwise closely related gene for oxytocin. This promoter region of the AVP gene in the rat contains several putative regulatory elements, including a glucocorticoid response element, a cyclic adenosine monophosphate (cAMP) response element, and four activating protein (AP)-2–binding sites.[20] Recent experimental data indicated that the DNA sequences between the AVP and the oxytocin genes, the intergenic region, contain critical sites for cell-specific expression of these two hormones.[21]



FIGURE 13-3  The arginine vasopressin (AVP) gene and its protein products. The three exons encode a 145–amino acid prohormone with an NH2-terminal signal peptide. The prohormone is packaged into neurosecretory granules of magnocellular neurons. During axonal transport of the granules from the hypothalamus to the posterior pituitary, enzymatic cleavage of the prohormone generates the final products: AVP, neurophysin, and a COOH-terminal glycoprotein. When afferent stimulation depolarizes the AVP-containing neurons, the three products are released into capillaries of the posterior pituitary.  (Adapted from Richter D, Schmale H: The structure of the precursor to arginine vasopressin, a model preprohormone. Prog Brain Res 60:227-233, 1983.)




The gene for AVP is also expressed in a number of other neurons, including but not limited to the parvicellular neurons of the paraventricular and suprachiasmatic nuclei. Oxytocin and AVP genes are also expressed in several peripheral tissues, including the adrenal medulla, ovary, testis, thymus, and certain sensory ganglia.[22] However, the AVP mRNA in these tissues appears to be shorter (620 bases) than its hypothalamic counterpart (720 bases), apparently because of tissue-specific differences in the length of the polyA tails. More importantly, the levels of AVP in peripheral tissues are generally two to three orders of magnitude lower than in the neurohypophysis, suggesting that AVP in these tissues likely has paracrine rather than endocrine functions. This is consistent with the observation that destruction of the neurohypophysis essentially eliminates AVP from the plasma despite the presence of these multiple peripheral sites of AVP synthesis.

Secretion of AVP and its associated neurophysin occurs by a calcium-dependent exocytotic process similar to that described for other neurosecretory systems. Secretion is triggered by propagation of an electrical impulse along the axon that causes depolarization of the cell membrane, an influx of Ca2+, fusion of secretory granules with the cell membrane, and extrusion of their contents. This view is supported by the observation that AVP, neurophysin, and the copeptin glycoprotein are released simultaneously by many stimuli.[23] However, at the physiologic pH of plasma, there is no binding of either AVP or oxytocin to their respective neurophysins, so after secretion each peptide circulates independently in the bloodstream.[24]

Stimuli for secretion of AVP or oxytocin also stimulate transcription and increase the mRNA content of both prohormones in the magnocellular neurons. This has been well documented in rats, in which dehydration, which stimulates secretion of AVP, accelerates transcription and increases the levels of AVP (and oxytocin) mRNA, [25] [26] and hypoosmolality, which inhibits secretion of AVP, produces a decrease in the content of AVP mRNA.[27] These and other data indicate that the major control of AVP synthesis resides at the level of transcription.[28]

Antidiuresis occurs via interaction of the circulating hormone with AVP V2 receptors in the kidney, which results in increased water permeability of the collecting duct through the insertion of the aquaporin-2 (AQP2) water channel into the apical membranes of collecting tubule principal cells (see Chapters 8 and 9 ). The importance of AVP for maintaining water balance is underscored by the fact that the normal pituitary stores of this hormone are very large, allowing more than a week's supply of hormone for maximal antidiuresis under conditions of sustained dehydration.[28] Knowledge of the different conditions that stimulate pituitary AVP release in humans is therefore essential for understanding water metabolism.

Osmotic Regulation

AVP secretion is influenced by many different stimuli, but since the pioneering studies of antidiuretic hormone secretion by Verney, it has been clear that the most important under physiologic conditions is the osmotic pressure of plasma. With further refinement of radioimmunoassays for AVP, the unique sensitivity of this hormone to small changes in osmolality, as well as the corresponding sensitivity of the kidney to small changes in plasma AVP levels, has become apparent. Although the magnocellular neurons themselves have been found to have intrinsic osmoreceptive properties,[29] research over the last several decades has clearly shown that the most sensitive osmoreceptive cells that are able to sense small changes in plasma osmolality and transduce these changes into AVP secretion are located in the anterior hypothalamus, likely in or near the circumventricular organ called the organum vasculosum of the lamina terminalis (OVLT) (see Fig. 13-2 ). Perhaps the strongest evidence for location of the primary osmoreceptors in this area of the brain are the multiple studies that have demonstrated that destruction of this area disrupts osmotically stimulated AVP secretion and thirst without affecting the neurohypophysis or its response to nonosmotic stimuli. [30] [31]

Although some debate still exists with regard to the exact pattern of osmotically stimulated AVP secretion, most studies to date have supported the concept of a discrete osmotic threshold for AVP secretion above which a linear relationship between plasma osmolality and AVP levels occurs ( Fig. 13-4 ).[32] At plasma osmolalities below a threshold level, AVP secretion is suppressed to low or undetectable levels; above this point, AVP secretion increases linearly in direct proportion to plasma osmolality. The slope of the regression line relating AVP secretion to plasma osmolality can vary significantly across individual human subjects, in part because of genetic factors[33] but also in relation to other factors. In general, each 1 mOsm/kg H2O increase in plasma osmolality causes an increase in plasma AVP level ranging from 0.4 to 1.0 pg/mL. The renal response to circulating AVP is similarly linear, with urinary concentration that is directly proportional to AVP levels from 0.5 to 4 to 5 pg/mL, after which urinary osmolality is maximal and cannot increase further despite additional increases in AVP levels ( Fig. 13-5 ). Thus, changes of as little as 1% in plasma osmolality are sufficient to cause significant increases in plasma AVP levels with proportional increases in urine concentration, and maximal antidiuresis is achieved after increases in plasma osmolality of only 5 to 10 mOsm/kg H2O (i.e., 2%–4%) above the threshold for AVP secretion.



FIGURE 13-4  Comparative sensitivity of AVP secretion in response to increases in plasma osmolality versus decreases in blood volume or blood pressure in human subjects. The arrow indicates the low plasma AVP concentrations found at basal plasma osmolality Note that AVP secretion is much more sensitive to small changes in blood osmolality than to changes in volume or pressure.  (Adapted from Robertson GL: Posterior pituitary. In Felig P, Baxter J, Frohman LA [eds]: Endocrinology and Metabolism. New York, McGraw Hill, 1986, pp 338-386.)






FIGURE 13-5  Relationship of plasma osmolality, plasma AVP concentrations, urine osmolality, and urine volume in humans. Note that the osmotic threshold for AVP secretion defines the point at which urine concentration begins to increase, but the osmotic threshold for thirst is significantly higher and approximates the point at which maximal urine concentration has already been achieved. Note also that, because of the inverse relation between urine osmolality and urine volume, changes in plasma AVP concentrations have much larger effects on urine volume at low plasma AVP concentrations than at high plasma AVP concentrations.  (Adapted from Robinson AG: Disorders of antidiuretic hormone secretion. J Clin Endocrinol Metab 14:55-88, 1985.)




However, even this analysis underestimates the sensi-tivity of this system to regulate free water excretion. Urinary osmolality is directly proportional to plasma AVP levels as a consequence of the fall in urine flow induced by the AVP, but urine volume is inversely related to urine osmolality (see Fig. 13-5 ). An increase in plasma AVP concentration from 0.5 to 2 pg/mL has a much greater relative effect to decrease urine flow than does a subsequent increase in AVP concentration from 2 to 5 pg/mL, thereby magnifying the physiologic effects of small initial changes in plasma AVP levels. Furthermore, the rapid response of AVP secretion to changes in plasma osmolality coupled with the short half-life of AVP in human plasma (10-20 min) allows the kidneys to respond to changes in plasma osmolality on a minute-to-minute basis. The net result is a finely tuned osmoregulatory system that adjusts the rate of free water excretion accurately to the ambient plasma osmolality primarily via changes in pituitary AVP secretion.

The set-point of the osmoregulatory system also varies from person to person. In healthy adults, the osmotic threshold for AVP secretion ranges from 275 to 290 mOsm/kg H2O (averaging approximately 280-285 mOsm/kg H2O). Similar to sensitivity, individual differences in the set-point of the osmoregulatory system are relatively constant over time and appear to be genetically determined.[33] However, multiple factors, in addition to genetic influences, can alter either the sensitivity and/or the set-point of the osmoregulatory system for AVP secretion.[33] Foremost among these are acute changes in blood pressure, effective blood volume or both, which are discussed in the following section. Aging has been found to increase the sensitivity of the osmoregulatory system in multiple studies. [34] [35] Metabolic factors such as serum Ca2+ and various drugs can alter the slope of the plasma AVP-osmolality relationship as well.[36] Lesser degrees of shifting of the osmsensitivity and set-point for AVP secretion have been noted with alterations in gonadal hormones. Some studies have found increased osmosensitivity in women, particularly during the luteal phase of the menstrual cycle,[37] and in estrogen-treated men,[38] but these effects were relatively minor; others have found no significant sex differences.[33] The set-point of the osmoregulatory system is reduced more dramatically and reproducibly during pregnancy.[39] Recent evidence has suggested the possible involvement of the placental hormone relaxin[40] rather than gonadal steroids or human chorionic gonadotropin hormone in pregnancy-associated resetting of the osmostat for AVP secretion. That multiple factors can influence the set-point and sensitivity of osmotically regulated AVP secretion is not surprising in view of the fact that AVP secretion reflects a balance of bimodal in-puts, that is, both inhibitory as well as stimulatory,[41] from multiple different afferent inputs to the neurohypophysis ( Fig. 13-6 ).[42]



FIGURE 13-6  Schematic model of the regulatory control of the neurohypophysis. The secretory activity of individual magnocellular neurons is determined by an integration of the activities of both excitatory and inhibitory osmotic and nonosmotic afferent inputs. Superimposed on this are the effects of hormones and drugs, which can act at multiple levels to modulate the output of the system.  (Adapted from Verbalis JG: Osmotic inhibition of neurohypophyseal secretion. Ann N Y Acad Sci 689:227-233, 1983.)




Understanding the osmoregulatory mechanism also requires addressing the observation that AVP secretion is not equally sensitive to all plasma solutes. Sodium and its anions, which normally contribute more than 95% of the osmotic pressure of plasma, are the most potent solutes in terms of their capacity to stimulate AVP secretion and thirst, although certain sugars such as mannitol and sucrose are also equally effective when infused intravenously.[11]In contrast, increases in plasma osmolality caused by noneffective solutes such as urea or glucose cause little or no increase in plasma AVP levels in humans or animals. [11] [43] These differences in response to various plasma solutes are independent of any recognized nonosmotic influence, indicating that they are a property of the osmoregulatory mechanism itself. According to current concepts, the osmoreceptor neuron is stimulated by osmotically induced changes in its water content. In this case, the stimulatory potency of any given solute would be an inverse function of the rate at which it moves from the plasma to the inside of the osmoreceptor neuron. Solutes that penetrate slowly, or not at all, create an osmotic gradient that causes an efflux of water from the osmoreceptor, and the resultant shrinkage of the osmoreceptor neuron activates a stretch-inactivated noncationic channel that initiates depolarization and firing of the neuron.[44] Conversely, solutes that penetrate the cell readily create no gradient and, thus, have no effect on the water content and cell volume of the osmoreceptors. This mechanism agrees well with the observed relationship between the effect of certain solutes like Na+, mannitol, and glucose on AVP secretion and the rate at which they penetrate the blood-brain barrier.

Many neurotransmitters have been implicated in mediating the actions of the osmoreceptors on the neurohypo-physis. The SON is richly innervated by multiple pathways, including acetylcholine, catecholamines, glutamate, gamma-aminobutyric acid (GABA), histamine, opioids, Ang II, and dopamine (see review[45]). Studies have supported a potential role for all of these, and yet others, in the regulation of AVP secretion, as has local secretion of AVP into the hypothalamus from dendrites of the AVP-secreting neurons.[46] Although it remains unclear which of these are involved in the normal physiologic control of AVP secretion, in view of the likelihood that the osmoregulatory system is bimodal and integrated with multiple different afferent pathways (see Fig. 13-6 ), it seems likely that magnocellular AVP neurons are influenced by a complex mixture of neurotransmitter systems rather than only a few.

Nonosmotic Regulation

Hemodynamic Changes.

Not surprisingly, hypovolemia is also a potent stimulus for AVP secretion in humans, [32] [47] because an appropriate response to volume depletion should include renal water conservation. In humans as well as multiple animal species, lowering blood pressure suddenly by any of several methods increases plasma AVP levels by an amount that is proportional to the degree of hypotension achieved. [32] [48] This stimulus-response relationship follows a distinctly exponential pattern, such that small reductions in blood pressure, of the order of 5% to 10%, usually have little effect on plasma AVP, whereas blood pressure decreases of 20% to 30% result in hormone levels many times those required to produce maximal antidiuresis (see Fig. 13-4 ). The AVP response to acute reductions in blood volume appears to be quantitatively and qualitatively similar to the response to blood pressure. In rats, plasma AVP increases as an exponential function of the degree of hypovolemia. Thus, little increase in plasma AVP can be detected until blood volume falls by 5% to 8%; beyond that point, plasma AVP increases at an exponential rate relation to the degree of hypovolemia and usually reaches levels 20 to 30 times normal when blood volume is reduced by 20% to 40%. [49] [50] The volume-AVP relation has not been as thoroughly characterized in other species, but it appears to follow a similar pattern humans.[51] Conversely, acute increases in blood volume or pressure suppress AVP secretion. This response has been characterized less well than that of hypotension or hypovolemia, but it seems to have a similar quantitative relationship (i.e., relatively large changes, of the order of 10%–15%, are required to alter hormone secretion appreciably).[52]

The minimal to absent effect of small changes in blood volume and pressure on AVP secretion contrasts sharply with the extraordinary sensitivity of the osmoregulatory system (see Fig. 13-4 ). Recognition of this difference is essential for understanding the relative contribution of each system to control AVP secretion under physiologic and pathologic conditions. Because day-to-day variations of TBW rarely exceed 2% to 3%, their effect on AVP secretion must be mediated largely, if not exclusively, by the osmoregulatory system. Nonetheless, modest changes in blood volume and pressure do, in fact, influence AVP secretion indirectly, even though they are weak stimuli by themselves. This occurs via shifting the sensitivity of AVP secretion to osmotic stimuli so that a given increase in osmolality will cause a greater secretion of AVP during hypovolemic conditions than during euvolemic states ( Fig. 13-7 ). [53] [54] In the presence of a negative hemodynamic stimulus, plasma AVP continues to respond appropriately to small changes in plasma osmolality and can still be fully suppressed if the osmolality falls below the new (lower) set-point. The retention of the threshold function is a vital aspect of the interaction because it ensures that the capacity to regulate the osmolality of body fluids is not lost even in the presence of significant hypovolemia or hypotension. Consequently, it is reasonable to conclude that the major effect of moderate degrees of hypovolemia on both AVP secretion and thirst is to modulate the gain of the osmoregulatory responses, with direct effects on thirst and AVP secretion occurring only during more severe degrees of hypovolemia (e.g., >10%–20% reductions in blood pressure or volume).



FIGURE 13-7  The relation between the osmolality of plasma and the concentration of AVP in plasma is modulated by blood volume and pressure. The line labeled N shows plasma AVP concentration across a range of plasma osmolality in an adult with normal intravascular volume (euvolemic) and normal blood pressure (normotensive). The lines to the left of N show the relationship between plasma AVP concentration and plasma osmolality in adults whose low intravascular volume (hypovolemia) or blood pressure (hypotension) is 10%, 15%, and 20% below normal. Lines to the right of N indicate volumes and blood pressures 10%, 15%, and 20% above normal. Note that hemodynamic influences do not disrupt the osmoregulation of AVP but rather raise or lower the set-point, and possibly the sensitivity as well, of AVP secretion in proportion to the magnitude of the change in blood volume or pressure.  (Adapted from Robertson GL, Athar S, Shelton RL: Osmotic control of vasopressin function. In Andreoli TE, Grantham JJ, Rector FC Jr [eds]: Disturbances in Body Fluid Osmolality. Bethesda, MD, American Physiological Society, 1977, p 125.)




These hemodynamic influences on AVP secretion are mediated at least in part by neural pathways that originate in stretch-sensitive receptors, generally called baroreceptors, in the cardiac atria, aorta, and carotid sinus (see Fig. 13-2). Afferent nerve fibers from these receptors ascend in the vagus and glossopharyngeal nerves to the nuclei of the tractus solitarius (NTS) in the brainstem.[55] A variety of postsynaptic pathways from the NTS then project, both directly and indirectly via the ventrolateral medulla and the lateral parabrachial nucleus, to the PVN and SON in the hypothalamus.[56] Early studies suggested that the input from these pathways was predominantly inhibitory under basal conditions, because interrupting them acutely resulted in large increases in plasma AVP levels as well as in arterial blood pressure.[57] However, as for most neural systems including the neurohypophysis, innervation is complex and consists of both excitatory and inhibitory inputs. Consequently, different effects have been observed under different experimental conditions.

The baroreceptor mechanism also appears to mediate a large number of pharmacologic and pathologic effects on AVP secretion ( Table 13-1 ). Among them are diuretics, isoproterenol, nicotine, prostaglandins, nitroprusside, trimethaphan, histamine, morphine, and bradykinin, all of which stimulate AVP at least in part by lowering blood volume or pressure,[47] and norepinephrine, which suppresses AVP by raising blood pressure.[58] In addition, upright posture, sodium depletion, congestive heart failure, cirrhosis, and nephrosis likely stimulate AVP secretion by reducing effective circulating blood volume. [59] [60] Symptomatic orthostatic hypotension, vasovagal reactions, and other forms of syncope more markedly stimulate AVP secretion via greater and more acute decreases in blood pressure, with the exception of orthostatic hypotension associated with loss of afferent baroregulatory function.[61] Almost every hormone, drug, or condition that affects blood volume or pressure will also affect AVP secretion, but in most cases, the degree of change of blood pressure or volume is modest and will result in a shift of the set-point and/or sensitivity of the osmoregulatory response rather than marked stimulation of AVP secretion (see Fig. 13-7 ).

TABLE 13-1   -- Drugs and Hormones That Affect Vasopressin Secretion









Morphine (high doses)







Opioid agonists


Morphine (low doses)





Cyclophosphamide IV








Angiotensin II




Corticotropin-releasing factor










Peripheral neural sensors other than baroreceptors can also affect AVP secretion. In humans as well as dogs, drinking lowers plasma AVP before there is any appreciable decrease in plasma osmolality or serum Na+. This is clearly a response to the act of drinking itself because it occurs independently of the composition of the fluid ingested, [62] [63] although it may be influenced by the temperature of the fluid because the degree of suppression appears to be greater in response to colder fluids.[64] The pathways responsible for this effect have not been delineated, but likely include sensory afferent originating in the oropharynx and transmitted centrally via the glossopharyngeal nerve.


Among other nonosmotic stimuli to AVP secretion in humans, nausea is the most prominent. The sensation of nausea, with or without vomiting, is by far the most potent stimulus to AVP secretion known in humans. Whereas 20% increases in osmolality will typically elevate plasma AVP levels to the range of 5 to 20 pg/mL, and 20% decreases in blood pressure to 10 to 100 pg/mL, nausea has been described to cause AVP elevations in excess of 200 to 400 pg/mL.[65] The pathway mediating this effect has been mapped to the chemoreceptor zone in the area postrema of the brainstem in animal studies (see Fig. 13-2 ). It can be activated by a variety of drugs and conditions, including apomorphine, morphine, nicotine, alcohol, and motion sickness. Its effect on AVP is instantaneous and extremely potent ( Fig. 13-8 ), even when the nausea is transient and not accompanied by vomiting or changes in blood pressure. Pretreatment with fluphenazine, haloperidol, or promethazine in doses sufficient to prevent nausea completely abolishes the AVP response. The inhibitory effect of these dopamine antagonists is specific for emetic stimuli, because they do not alter the AVP response to osmotic and hemodynamic stimuli. Water loading blunts, but does not abolish, the effect of nausea on AVP release, suggesting that osmotic and emetic influences interact in a manner similar to that for osmotic and hemodynamic pathways. Species differences also affect emetic stimuli. Whereas dogs and cats appear to be even more sensitive than humans to emetic stimulation of AVP release, rodents have little or no AVP response but release large amounts of oxytocin instead.[66]



FIGURE 13-8  Effect of nausea on AVP secretion. Apomorphine was injected at the point indicated by the vertical arrow. Note that the rise in plasma AVP coincided with the occurrence of nausea and was not associated with detectable changes in plasma osmolality or blood pressure. (Adapted from Robertson GL: The regulation of vasopressin function in health and disease. Recent Prog Horm Res 33:333, 1977.)




The emetic response probably mediates many pharmacologic and pathologic effects on AVP secretion. In addition to the drugs and conditions already noted, it may be responsible at least in part for the increase in AVP secretion that has been observed with vasovagal reactions, diabetic ketoacidosis, acute hypoxia, and motion sickness. Because nausea and vomiting are frequent side effects of many other drugs and diseases, many additional situations likely occur as well. The reason for this profound stimulation is not known (although it has been speculated that the AVP response assists evacuation of stomach contents via contraction of gastric smooth muscle, AVP is not necessary for vomiting to occur), but it is probably responsible for the intense vasoconstriction that produces the pallor often associated with this state.


Acute hypoglycemia is a less potent but reasonably consistent stimulus for AVP secretion. [67] [68] The receptor and pathway that mediate this effect are unknown; however, they appear separate from those of other recognized stimuli, because hypoglycemia stimulates AVP secretion even in patients who have selectively lost the capacity to respond to hypernatremia, hypotension, or nausea.[68] The factor that actually triggers the release of AVP is likely intracellular deficiency of glucose or ATP, because 2-deoxyglucose is also an effective stimulus.[69] Generally, more than 20% decreases in glucose are required to significantly increase plasma AVP levels; the rate of fall in glucose is probably the critical stimulus, however, because the rise in plasma AVP is not sustained with persistent hypoglycemia.[67] However, glucopenic stimuli are of unlikely importance in the physiology or pathology of AVP secretion, because there are probably few drugs or conditions that lower plasma glucose rapidly enough to stimulate release of the hormone, and furthermore, because this effect is transient.

Renin-Angiotensin System.

The renin-angiotensin sys-tem has also been intimately implicated in the control of AVP secretion.[70] Animal studies have indicated dual sites of action. Blood-borne Ang II stimulates AVP secretion by acting in the brain at the circumventricular subfornical organ (SFO),[71] a small structure located in the dorsal portion of the third cerebral ventricle (see Fig. 13-2 ). Because circumventricular organs lack a blood-brain barrier, the densely expressed Ang II AT1 receptors of the SFO can detect very small increases in blood levels of Ang II.[72] Neural pathways from the SFO to the hypothalamic SON and PVN mediate AVP secretion and also appear to use Ang II as a neurotransmitter.[73] This accounts for the observation that the most sensitive site for angiotensin-mediated AVP secretion and thirst is intracerebroventricular injection into the cerebrospinal fluid. Further evidence in support of Ang II as a neurotransmitter is that intraventricular administration of angiotensin receptor antagonists inhibits the AVP response to osmotic and hemodynamic stimuli.[74] The level of plasma Ang II required to stimulate AVP release is quite high, leading some to argue that this stimulus is active only under pharmacologic conditions. This is consistent with observations that even pressor doses of Ang II increase plasma AVP only about two- to fourfold[70] and may account for the failure of some investigators to demonstrate stimulation of thirst by exogenous angiotensin. However, this procedure may underestimate the physiologic effects of angiotensin, because the increased blood pressure caused by exogenously administered Ang II appears to blunt the induced thirst via activation of inhibitory baroreceptive pathways.[75]


Nonspecific stress caused by factors such as pain, emotion, or physical exercise has long been thought to cause AVP secretion, but it has never been determined whether this effect is mediated by a specific pathway or is secondary to the hypotension or nausea that often accompanies stress-induced vasovagal reactions. In rats[76] and humans,[77] a variety of noxious stimuli capable of activating the pituitary-adrenal axis and sympathetic nervous system do not stimulate AVP secretion unless they also lower blood pressure or alter blood volume. The marked rise in plasma AVP elicited by manipulation of the abdominal viscera in anesthetized dogs has been attributed to nociceptive influences,[78] but mediation by emetic pathways cannot be excluded in this setting. Endotoxin-induced fever stimulates AVP secretion in rats, and recent data support possible mediation of this effect by circulating cytokines such as interleukin-1 (IL-1) and IL-6.[79] Clarification of the possible role of nociceptive and thermal influences on AVP secretion is particularly important in view of the frequency with which painful or febrile illnesses are associated with osmotically inappropriate secretion of the hormone.

Hypoxia and Hypercapnia.

Acute hypoxia and hypercapnia also stimulate AVP secretion. [80] [81] In conscious humans, however, the stimulatory effect of moderate hypoxia (arterial partial pressure of oxygen [Pao2]>35 mm Hg) is inconsistent, and seems to occur mainly in subjects who develop nausea or hypotension. In conscious dogs, more severe hypoxia (Pao2<35 mm Hg) consistently increases AVP secretion with-out reducing arterial pressure.[82] Studies of anesthetized dogs suggest that the AVP response to acute hypoxia depends on the level of hypoxemia achieved. At a Pao2 of 35 mm Hg or lower, plasma AVP increases markedly even though there is no change or even an increase in arterial pressure, but less severe hypoxia (Pao2>40 mm Hg) has no effect on AVP levels.[83] These results indicate that there is likely a hypoxemic threshold for AVP secretion and suggest that severe hypoxemia alone may also stimulate AVP secretion in humans. If so, it may be responsible, at least in part, for the osmotically inappropriate AVP elevations noted in some patients with acute respiratory failure.[84] In conscious or anesthetized dogs, acute hypercapnia, independent of hypoxia or hypotension, also increases AVP secretion. [82] [83] It has not been determined whether this response also exhibits threshold characteristics or otherwise depends on the degree of hypercapnia, nor is it known whether hypercapnia has similar effects on AVP secretion in humans or other animals. The mechanisms by which hypoxia and hypercapnia release AVP remain undefined, but they likely involve peripheral chemoreceptors and/or baroreceptors, because cervical vagotomy abolishes the response to hypoxemia in dogs.[85]


As is discussed more extensively in the clinical disorders, a variety of drugs, including nicotine, also stimulate AVP secretion (see Table 13-1 ). Drugs and hormones can potentially affect AVP secretion at many different sites, as depicted in Figure 13-6 . As already discussed, many excitatory stimulants such as isoproterenol, nicotine, high doses of morphine and cholecystokinin act, at least in part, by lowering blood pressure and/or producing nausea. Others, like substance P, prostaglandin, endorphin, and other opioids, have not been studied sufficiently to define their mechanism of action, but they may also work by one or both of the same mechanisms. Inhibitory stimuli similarly have multiple modes of action. Vasopressor drugs like norepinephrine inhibit AVP secretion indirectly by raising arterial pressure. In low doses, a variety of opioids of all subtypes including morphine, met-enkephalin and kappa-agonists inhibit AVP secretion in rats and humans.[86] Endogenous opioid peptides interact with the magnocellular neurosecretory system at several levels to inhibit basal as well as stimulated secretion of AVP and oxytocin. Opioid inhibition of AVP secretion has been found to occur in isolated posterior pituitary tissue, and the action of morphine as well as several opioid agonists such as butorphanol and oxilorphan likely occurs via activation of kappa-opioid receptors located on nerve terminals of the posterior pituitary.[87] The well-known inhibitory effect of alcohol on AVP secretion may be mediated, at least in part, by endogenous opiates, because it is due to an elevation in the osmotic threshold for AVP release[88] and can be blocked in part by treatment with naloxone.[89] Carbamazepine inhibits AVP secretion by diminishing the sensitivity of the osmoregulatory system; this effect occurs independently of changes in blood volume, blood pressure, or blood glucose.[90] Other drugs that inhibit AVP secretion include clonidine, which appears to act via both central and peripheral adrenoreceptors,[91] muscimol,[92] which acts as a GABA antagonist, and phencyclidine,[93] which probably acts by raising blood pressure. However, despite the importance of these stimuli during pathologic conditions, none of them is a significant determinant of physiologic regulation of AVP secretion in humans.

Distribution and Clearance

Plasma AVP concentration is determined by the difference between the rates of secretion from the posterior pituitary gland and removal of the hormone from the vascular compartment via metabolism and urinary clearance. In healthy adults, intravenously injected AVP distributes rapidly into a space equivalent in size to the ECF compartment. This initial, or mixing, phase has a half-life between 4 and 8 minutes and is virtually complete in 10 to 15 minutes. The rapid mixing phase is followed by a second, slower decline that corresponds to the metabolic clearance of AVP. Most studies of this phase have yielded mean values of 10 to 20 minutes by both steady-state and non-steady-state techniques,[32] consistent with the observed rates of change in urine osmolality after water loading and injection of AVP, which also support a short half-life.[94] In pregnant women, the metabolic clearance rate of increases nearly fourfold,[95] which becomes significant in the pathophysiology of gestational diabetes insipidus (GDI) (see later discussion). Smaller animals such as rats clear AVP much more rapidly than humans because their cardiac output is higher relative to their BW and surface area.[94]

Although many tissues have the capacity to inactivate AVP, metabolism in vivo appears to occur largely in liver and kidney.[94] The enzymatic processes by which the liver and kidney inactivate AVP involve an initial reduction of the disulfide bridge followed by aminopeptidase cleavage of the bond between amino acid residues 1 and 2. The extent of further degradation and the peptide products that escape into plasma and urine are currently unknown. Some AVP is excreted intact in the urine, but there is disagreement about the amounts and the factors that affect it. For example, in healthy, normally hydrated adults, the urinary clearance of AVP ranges from 0.1 to 0.6 mL/kg/min under basal conditions and has never been found to exceed 2 mL/kg/min, even in the presence of solute diuresis.[32] The mechanisms involved in the excretion of AVP have not been defined with certainty, but the hormone is probably filtered at the glomerulus and variably reabsorbed at sites along the nephron. The latter process may be linked to the reabsorption of Na+ or other solutes in the proximal nephron, because the urinary clearance of AVP has been found to vary by as much as 20-fold in direct relation to the solute clearance.[32] Consequently, measurements of urinary AVP excretion in humans do not provide a consistently reliable index of changes in plasma AVP, and should be interpreted cautiously when glomerular filtration or solute clearance is inconstant or abnormal.


Thirst is the body's defense mechanism to increase water consumption in response to perceived deficits of body fluids. It can be most easily defined as a consciously perceived desire for water. True thirst must be distinguished from other determinants of fluid intake such as taste, dietary preferences, and social customs, as discussed previously. Thirst can be stimulated in animals and humans either by intracellular dehydration caused by increases in the effective osmolality of the ECF or by intravascular hypovolemia caused by losses of ECF. [96] [97] As would be expected, many of these same variables provoke AVP secretion. Of these, hypertonicity is clearly the most potent. Similar to AVP secretion, substantial evidence to date has supported mediation of osmotic thirst by osmoreceptors located in the anterior hypothalamus of the brain, [30] [31] whereas hypovolemic thirst appears to be stimulated both via activation of low- and/or high-pressure baroreceptors[98] and circulating Ang II.[99]

Osmotic Thirst.

In healthy adults, an increase in effective plasma osmolality of only 2% to 3% above basal levels produces a strong desire to drink.[100] This response is not dependent on changes in ECF or plasma volume, because it occurs similarly whether plasma osmolality is raised by infusion of hypertonic solutions or by water deprivation. The absolute level of plasma osmolality at which a person develops a conscious urge to seek and drink water is called the osmotic thirst threshold. It varies appreciably among individuals, likely as a result of by genetic factors,[33] but in healthy adults, it averages approximately 295 mOsm/kg H2O. Of physiologic significance is the fact that this level is above the osmotic threshold for AVP release and approximates the plasma osmolality at which maximal concentration of the urine is normally achieved (see Fig. 13-5 ).

The brain pathways that mediate osmotic thirst have not been well defined, but it is clear that initiation of drinking requires osmoreceptors located in the anteroventral hypothalamus in the same area as the osmoreceptors that control osmotic AVP secretion are located. [30] [31] Whether the osmoreceptors for AVP and thirst are the same cells or simply located in the same general area remains unknown. However, the properties of the osmoreceptors are very similar. Ineffective plasma solutes such as urea and glucose, which have little or no effect on AVP secretion, are equally ineffective at stimulating thirst, whereas effective solutes such as NaCl and mannitol are. [11] [101] The sensitivities of the thirst and AVP osmoreceptors cannot be compared precisely, but they are probably similar. Thus, in healthy adults, the intensity of thirst increases rapidly in direct proportion to serum Na+ or plasma osmolality and generally becomes intolerable at levels only 3% to 5% above the threshold level.[102] Water consumption also appears to be proportional to the intensity of thirst, in both humans and animals and, under conditions of maximal osmotic stimulation, can reach rates as high as 20 to 25 L/day. The dilution of body fluids by ingested water complements the retention of water that occurs during AVP-induced antidiuresis, and both responses occur concurrently when drinking water is available.

As with AVP secretion, the osmoregulation of thirst appears to be bimodal, because a modest decline in plasma osmolality induces a sense of satiation and reduces the basal rate of spontaneous fluid intake. [102] [103] This effect is sufficient to prevent hypotonic overhydration even when antidiuresis is fixed at maximal levels for prolonged periods, suggesting that osmotically inappropriate secretion of ADH (SIADH) should not result in the development of hyponatremia unless the satiety mechanism is impaired or fluid intake is inappropriately high for some other reason, such as the unregulated components of fluid intake discussed earlier.[103] Also similar to AVP secretion, thirst can be influenced by oropharyngeal or upper gastrointestinal receptors that respond to the act of drinking itself.[63] In humans, however, the rapid relief provided by this mechanism lasts only a matter of minutes and thirst quickly recurs until enough of the water is absorbed to lower plasma osmolality to normal. Therefore, although local oropharyngeal sensations may have a significant short-term influence on thirst, the hypothalamic osmoreceptors ultimately determine the volume of water intake in response to dehydration.

Hypovolemic Thirst.

In contrast, the threshold for producing hypovolemic, or extracellular, thirst is significantly higher in both animals and humans. Studies in several species have shown that sustained decreases in plasma volume or blood pressure of at least 4% to 8%, and in some species 10% to 15%, are necessary to consistently stimulate drinking. [104] [105] In humans, the degree of hypovolemia or hypotension re-quired to produce thirst has not been precisely defined, but it has been difficult to demonstrate any effects of mild to moderate hypovolemia to stimulate thirst independently of osmotic changes occurring with dehydration. This blunted sensitivity to changes in ECF volume or blood pressure in humans probably represents an adaptation that occurred as a result of the erect posture of primates, which predisposes them to wider fluctuations in blood and atrial filling pressures as a result of orthostatic pooling of blood in the lower body; stimulation of thirst (and AVP secretion) by such transient postural changes in blood pressure might lead to overdrinking and inappropriate antidiuresis in situations in which the ECF volume was actually normal but only transiently maldistributed. Consistent with a blunted response to baroreceptor activation, recent studies have also shown that systemic infusion of Ang II to pharmacologic levels is a much less potent stimulus to thirst in humans[106]than in animals, in which it is one of the most potent dipsogens known. Nonetheless, this response is not completely absent in humans, as demonstrated by rare cases of polydipsia in patients with pathologic causes of hyperreninemia.[107] The pathways by which hypovolemia or hypotension produces thirst have not been well-defined, but probably involve the same brainstem baroreceptive pathways that mediate hemodynamic effects on AVP secretion,[98] as well as a likely contribution from circulating levels of Ang II in some species.[108]

Integration of Arginine Vasopressin Secretion and Thirst

A synthesis of what is presently known about the regulation of AVP secretion and thirst in humans leads to a relatively simple but elegant system to maintain water balance. Under normal physiologic conditions, the sensitivity of the osmoregulatory system for AVP secretion accounts for maintenance of plasma osmolality within narrow limits by adjusting renal water excretion to small changes in osmolality. Stimulated thirst does not represent a major regulatory mechanism under these conditions, and unregulated fluid ingestion supplies adequate water in excess of true “need,” which is then excreted in relation to osmoregulated pituitary AVP secretion. However, when unregulated water intake cannot adequately supply body needs in the presence of plasma AVP levels sufficient to produce maximal antidiuresis, then plasma osmolality rises to levels that stimulate thirst (see Fig. 13-5 ), and water intake increases proportional to the elevation of osmolality above this thirst threshold.

In such a system, thirst essentially represents a back-up mechanism called into play when pituitary and renal mechanisms prove insufficient to maintain plasma osmolality within a few percent of basal levels. This arrangement has the advantage of freeing humans from frequent episodes of thirst that would require a diversion of activities toward behavior oriented to seeking water when water deficiency is sufficiently mild to be compensated for by renal water conservation but would stimulate water ingestion once water deficiency reaches potentially harmful levels. Stimulation of AVP secretion at plasma osmolalities below the threshold for subjective thirst acts to maintain an excess of body water sufficient to eliminate the need to drink whenever slight elevations in plasma osmolality occur. This system of differential effective thresholds for thirst and AVP secretion nicely complements many studies that have demonstrated excess unregulated, or “need-free,” drinking in both humans and animals. Only when this mechanism becomes inadequate to maintain body fluid homeostasis does thirst-induced regulated fluid intake become the predominant defense mechanism for the prevention of severe dehydration.


Disorders of insufficient AVP or AVP effect are associated with inadequate urine concentration and increased urine output (polyuria). If thirst mechanisms are intact, this is accompanied by compensatory increases in fluid intake (polydipsia) as a result of stimulated thirst in order to preserve body fluid homeostasis. The net result is polyuria and polydipsia with preservation of normal plasma osmolality and serum electrolyte concentrations. However, if thirst is impaired or if fluid intake is insufficient for any reason to compensate for the increased urine excretion, then hyperosmolality and hypernatremia can result, with the consequent complications associated with these disorders. The quintessential disorder of insufficient AVP is diabetes insipidus (DI), which is a clinical syndrome characterized by excretion of abnormally large volumes of urine (i.e., diabetes) that is dilute (i.e., hypotonic) and devoid of taste from dissolved solutes (e.g., insipid), in contrast to the hypertonic sweet-tasting urine characteristic of diabetes mellitus (i.e., honey, in Greek).

Several different pathophysiologic mechanisms can cause hypotonic polyuria ( Table 13-2 ). Central (also called hypothalamic, neurogenic, or neurohypophyseal) DI (CDI) is due to inadequate secretion, and usually deficient synthesis of, AVP in the hypothalamic neurohypophyseal system. Lack of AVP-stimulated activation of the V2 subtype of AVP receptors in the kidney collecting tubules (see Chapters 8 and 9 ) causes excretion of large volumes of dilute urine. In most cases, thirst mechanisms are intact, leading to compensatory polydipsia. However, in a variant of CDI called osmoreceptor dysfunction, thirst is also impaired, leading to hypodipsia. DI of pregnancy is a transient disorder due to an accelerated metabolism of AVP as a result of increased activity of the enzyme oxytocinase/vasopressinase in the serum of pregnant females, again leading to polyuria and polydipsia; accelerated metabolism of AVP during pregnancy may also cause a patient with subclinical DI from other causes to shift from a relatively asymptomatic state to a symptomatic state as a result of the more rapid AVP degradation. Nephrogenic DI (NDI) is due to inappropriate renal responses to AVP. This produces excretion of dilute urine despite normal pituitary AVP secretion and secondary polydipsia, similar to CDI. The final cause of hypotonic polyuria, primary polydipsia, differs significantly from the other causes because it is not due to deficient AVP secretion or impaired renal responses to AVP, but rather to excessive ingestion of fluids. This can result from either an abnormality in the thirst mechanism, in which case it is sometimes called dipsogenic DI, or to psychiatric disorders, in which case it is generally referred to as psychogenic polydipsia.

TABLE 13-2   -- Etiologies of Hypotonic Polyuria



Central (neurogenic) diabetes insipidus



Congenital (congenital malformations, autosomal dominant, arginine vasopressin (AVP)–neurophysin gene mutations)



Drug-/toxin-induced (ethanol, diphenylhydantoin, snake venom)



Granulomatous (histiocytosis, sarcoidosis)



Neoplastic (craniopharyngioma, germinoma, lymphoma, leukemia, meningioma, pituitary tumor; metastases)



Infectious (meningitis, tuberculosis, encephalitis)



Inflammatory/autoimmune (lymphocytic infundibuloneurohypophysitis)



Trauma (neurosurgery, deceleration injury)



Vascular (cerebral hemorrhage or infarction, brain death)






Osmoreceptor dysfunction



Granulomatous (histiocytosis, sarcoidosis)



Neoplastic (craniopharyngioma, pinealoma, meningioma, metastases)



Vascular (anterior communicating artery aneurysm/ligation, intrahypothalamic hemorrhage)



Other (hydrocephalus, ventricular/suprasellar cyst, trauma, degenerative diseases)






Increased AVP metabolism






Nephrogenic diabetes insipidus



Congenital (X-linked recessive, AVP V2 receptor gene mutations, autosomal recessive or dominant, aquaporin-2 water channel gene mutations)



Drug-induced (demeclocycline, lithium, cisplatin, methoxyflurane)









Infiltrating lesions (sarcoidosis, amyloidosis)



Vascular (sickle cell anemia)



Mechanical (polycystic kidney disease, bilateral ureteral obstruction)



Solute diuresis (glucose, mannitol, sodium, radiocontrast dyes)






Primary polydipsia



Psychogenic (schizophrenia, obsessive-compulsive behaviors)



Dipsogenic (downward resetting of thirst threshold, idiopathic or similar lesions as with central DI)




Central Diabetes Insipidus


CDI is caused by inadequate secretion of AVP from the posterior pituitary in response to osmotic stimulation. In most cases, this is due to destruction of the neurohypophysis by a variety of acquired or congenital anatomic lesions that destroy or damage the neurohypophysis by pressure or infiltration (see Table 13-2 ). The severity of the resulting hypotonic diuresis depends on the degree of destruction of the neurohypophysis, leading to either complete or partial deficiency of AVP secretion.

Despite the wide variety of lesions that can potentially cause CDI, it is much more common to not have CDI in the presence of such lesions than to actually produce the syndrome. This apparent inconsistency can be understood by considering several common principles of neurohypophyseal physiology and pathophysiology that are relevant to all of these etiologies. The first is that the synthesis of AVP occurs in the hypothalamus (see Fig. 13-2 ); the posterior pituitary simply represents the site of storage and secretion of the neurosecretory granules that contain AVP. Consequently, lesions contained within the sella turcica that destroy only the posterior pituitary generally do not cause CDI because the cell bodies of the magnocellular neurons that synthesize AVP remain intact and the site of release of AVP shifts more superiorly, typically into the blood vessels of the median eminence at the base of the brain. Perhaps the best examples of this phenomenon are large pituitary macroadenomas that completely destroy the anterior and posterior pituitary. DI is a distinctly unusual presentation for such pituitary adenomas, because destruction of the posterior pituitary by such slowly enlarging intrasellar lesions merely destroys the nerve terminals, but not the cell bodies, of the AVP neurons. As this occurs, the site of release of AVP shifts more superiorly to the pituitary stalk and median eminence. Sometimes this can be detected on noncontrast magnetic resonance imaging (MRI) as a shift of the pituitary “bright spot” more superiorly to the level of the infundibulum or median eminence,[109] but often, this process is too diffuse to be detected in this manner. The occurrence of DI from a pituitary adenoma is so uncommon, even with macroadenomas that completely obliterate sellar contents sufficiently to cause panhypopituitarism, that its presence should lead to consideration of alternative diagnoses, such as craniopharyngioma, which often causes damage to the median eminence by virtue of adherence of the capsule to the base of the hypothalamus, more rapidly enlarging sellar/suprasellar masses that do not allow sufficient time for shifting the site of AVP release more superiorly (e.g., metastatic lesions), or granulomatous disease with more diffuse hypothalamic involvement (e.g., sarcoidosis, histiocytosis). With very large pituitary adenomas that produce ACTH deficiency, it is actually more likely that patients will present with hypoosmolality from an SIADH-like picture as a result of the impaired free water excretion that accompanies hypocortisolism, as is discussed later.

A second general principle is that the capacity of the neurohypophysis to synthesize AVP is greatly in excess of the body's daily needs for maintenance of water homeostasis. Carefully controlled studies of surgical section of the pituitary stalk in dogs have clearly demonstrated that destruction of 80% to 90% of the magnocellular neurons in the hypothalamus is required to produce polyuria and polydipsia in this species.[110] Thus, even lesions that do cause destruction of the AVP magnocellular neuron cell bodies must cause a large degree of destruction to produce DI. The most illustrative example of this is surgical section of the pituitary stalk in humans. Necropsy studies of these patients have revealed atrophy of the posterior pituitary and loss of the magnocellular neurons in the hypothalamus.[111] This loss of magnocellular cells presumably results from retrograde degeneration of neurons whose axons were cut during surgery. As is generally true for all neurons, the likelihood of retrograde neuronal degeneration depends on the proximity of the axotomy, in this case, section of the pituitary stalk, to the cell body of the neuron. This was shown clearly in studies of human subjects in whom section of the pituitary stalk at the level of the diaphragm sella (i.e., a low stalk section) produced transient but not permanent DI, whereas section at the level of the infundibulum (i.e., a “high” stalk section) was required to cause permanent DI in most cases.[112]

In recent years, several genetic causes of AVP deficiency have also been characterized. Prior to the application of techniques for amplification of genomic DNA, the only ex-perimental model to study the mechanism of hereditary hypothalamic DI was the Brattleboro rat, a strain that was found serendipitously to have CDI.[113] In this animal, the disease demonstrates a classic pattern of autosomal recessive inheritance in which DI is expressed only in the homozygotes. The hereditary basis of the disease has been found to be a single base deletion producing a translational frame shift beginning in the third portion of the neurophysin coding sequence. Because the gene lacks a stop codon, there is a modified neurophysin, no glycopeptide, and a long polylysine tail.[114] Although the mutant prohormone accumulates in the endoplasmic reticulum, sufficient AVP is produced by the normal allele that the heterozygotes are asymptomatic. In contrast, most all families with genetic CDI in humans that have been described to date demonstrate an autosomal dominant mode of inheritance. [115] [116] [117] In this case, DI is expressed despite the expression of one normal allele, which is sufficient to prevent the disease in the heterozygous Brattleboro rats. Numerous studies have been directed at understanding this apparent anomaly. Two potentially important clues as to the etiology of the DI in familial genetic CDI are that (1) severe to partial deficiencies of AVP and overt signs of DI do not develop in these patients until several months to several years after birth and then gradually progress over the ensuing decades, [115] [118] suggesting adequate initial function of the normal allele with later decompensation and (2) a limited number of autopsy studies suggested that some of these cases are associated with gliosis and a marked loss of magnocellular AVP neurons in the hypothalamus,[119] although other studies have shown normal neurons with decreased expression of AVP, or no hypothalamic abnormality. In most of these cases, the hyperintense signal normally emitted by the neurohypophysis in T1-weighted MRI (see later discussion) is also absent, although some exceptions have been reported.[120] Another interesting, but as yet unexplained, observation is that some adults in these families have been described in whom DI was clinically apparent during childhood but who went into remission as adults, without evidence that their remissions could be attributed to renal or adrenal insufficiency or to increased AVP synthesis.[121]

The autosomal dominant form of familial CDI is caused by diverse mutations in the gene that codes for the AVP-neurophysin precursor ( Fig. 13-9 ). All of the mutations identified to date have been in the coding region of the gene and affect only one allele. They are located in all three exons and are predicted to alter or delete amino acid residues in the signal peptide, AVP, and neurophysin moieties of the precursor. Only the C-terminus glycopeptide, or copeptin moiety, has not been found to be affected. Most are missense mutations, but nonsense mutations (premature stop codons) and deletions also occur. One characteristic shared by all the mutations is that they are predicted to alter or delete one or more amino acids known, or reasonably presumed, to be crucial for processing, folding, and oligomerization of the precursor protein in the endoplasmic reticulum. [115] [117] Because of the related functional effects of the mutations, the common clinical characteristics of the disease, the dominant-negative mode of transmission, and the autopsy and hormonal evidence of postnatal neurohypophysial degeneration, it has been postulated that all of the mutations act by causing production of an abnormal precursor protein that accumulates and eventually kills the neurons because it cannot be correctly processed, folded, and transported out of the endoplasmic reticulum. Expression studies of mutant DNA from several human mutations in cultured neuroblastoma cells support this misfolding/neurotoxicity hypothesis by demonstrating abnormal trafficking and accumulation of mutant prohormone in the endoplasmic reticulum with low or absent expression in the Golgi apparatus, suggesting difficulty with packaging into neurosecretory granules.[122] However, cell death may not be necessary to decrease available AVP. Normally, proteins retained in the endoplasmic reticulum are selectively degraded, but if excess mutant is produced and the selective normal degradative process is overwhelmed, an alternate nonselective degradative system (autophagy) is activated. As more and more mutant precursor builds up in the endoplasmic reticulum, normal wild type is trapped with the mutant protein and degraded by the activated nonspecific degradative system. In this case, the amount of AVP that matures and is packaged would be markedly reduced. [123] [124] This explanation is consistent with those cases in which little pathology is found in the magnocellular neurons and also with cases of DI in which some small amount of AVP can still be detected.



FIGURE 13-9  Location and type of mutations in the gene that codes for the AVP-neurophysin precursor in kindreds with the autosomal, dominant form of familial central diabetes insipidus (CDI). Each arrow indicates the location of the mutation in a different kindred. The various portions of the precursor protein are designated by the abbreviations AVP, vasopressin; CP, copeptin; NP, neurophysin; SP, signal peptide. Deletion and missense mutations are those expected to remove or replace one or more amino acid residues in the precursor. Those designated stop codons are expected to cause premature termination of the precursor. Note that none of the mutations causes a frame shift or affects the part of the gene that encodes the copeptin moiety, that all of the stop codons are in the distal part of the neurophysin moiety, and that only one of the mutations affects the AVP moiety. All these findings are consistent with the concept that the mutant precursor is produced but cannot be folded properly because of interference with either (1) the binding of AVP to neurophysin, (2) the formation of intrachain disulfide bonds, or (3) the extreme flexibility or rigidity normally required at crucial places in the protein.  (Adapted from Rittig S, Robertson GL, Siggaard C, et al: Identification of 13 new mutations in the vasopressin-neurophysin gene in 17 kindreds with familial autosomal dominant neurohypophyseal DI. Am J Hum Genet 58:107, 1996; and Hansen LK, Rittig S, Robertson GL: Genetic basis of familial neurohypophyseal diabetes insipidus. Trends Endocrinol Metab 8:363, 1997.)




Idiopathic forms of AVP deficiency represent a large pathogenic category in both adults and children. A recent study in children revealed that over half (54%) of all cases of CDI were classified as idiopathic.[125] These patients do not have historical or clinical evidence of any injury or disease that can be linked to their DI, and MRI of the pituitary-hypothalamic area generally reveals no abnormality other than absence of the posterior pituitary bright spot and sometimes varying degrees of thickening of the pituitary stalk. Several lines of evidence have suggested that many of these patients may have had an autoimmune destruction of the neurohypophysis to account for their DI. First, the entity of lymphocytic infundibuloneurohypophysitis has been documented to be present in a subset of patients with idiopathic DI.[126] Lymphocytic infiltration of the anterior pituitary, lymphocytic hypophysitis, has been recognized as a cause of anterior pituitary deficiency for many years, but it was not until an autopsy called attention to a similar finding in the posterior pituitary of a patient with DI that this pathology was recognized to occur in the neurohypophysis as well.[127] Since that initial report, a number of similar cases have been described, including cases in the postpartum period, which is characteristic of lymphocytic hypophysitis.[128] With the advent of MRI, lymphocytic infundibuloneurohypophysitis has been diagnosed based on the appearance of a thickened stalk and/or enlargement of the posterior pituitary mimicking a pituitary tumor. In these cases, the characteristic bright spot on MRI T1-weighted images is lost. The enlargement of the stalk can so mimic a neoplastic process that some of these patients were operated based on a suspicion of a pituitary tumor, but with the finding of lymphocytic infiltration of the pituitary stalk. Since then, a number of patients with a suspicion of infundibuloneurohypophysitis and no other obvious cause of DI have been followed and have shown regression of the thickened pituitary stalk over time. [125] [126] Several cases have been reported with the coexistence of CDI and adenohypophysitis, and these presumably represent cases of combined lymphocytic infundibuloneurohypophysitis and hypophysisis. [129] [130] A second line of evidence supporting an autoimmune etiology in many cases of idiopathic DI stems is the finding of AVP antibodies in the serum of an many as one third of patients with idiopathic DI and two thirds of those with Langerhans cell histiocytosis X, but not in patients with DI caused by tumors.[131] More recently, 878 patients with autoimmune endocrine diseases, but without hypothalamic DI, were screened, and 9 patients were found to have AVP antibodies; upon further testing, 4 of these patients were found to have partial DI and 5 were normal. After a 4-year follow-up, 3 of the normal subjects also had developed partial DI and 1 had progressed to complete DI, but interestingly 2 of the patients who had partial DI at entry were treated with desmopressin (desamino-8-d-arginine vasopressin [DDAVP]) and after 1 year became negative for AVP antibodies and had recovered normal posterior pituitary function.[132]


The normal inverse relationship between urine volume and urine osmolality (see Fig. 13-5 ) means that initial decreases in maximal AVP secretion will not cause an increase in urine volume sufficient to be detected clinically by polyuria. In general, basal AVP secretion must fall to less than 10% to 20% of normal before basal urine osmolality decreases to less than 300 mOsm/kg H2O and urine flow increases to symptomatic levels (i.e., >50 ml/kg BW/day). This resulting loss of body water produces a slight rise in plasma osmolality that stimulates thirst and induces a compensatory polydipsia. The resultant increase in water intake restores balance with urine output and stabilizes the osmolality of body fluids at a new, slightly higher but still normal level. As the AVP deficit increases, this new steady-state level of plasma osmolality approximates the osmotic threshold for thirst (see Fig. 13-5 ). It is important to recognize that the deficiency of AVP need not be complete for polyuria and polydipsia to occur; it is only necessary that the maximal plasma AVP concentration achievable at or below the osmotic threshold for thirst is inadequate to concentrate the urine.[133] The degree of neurohypophysial destruction at which such failure occurs varies considerably from person to person, largely because of individual differences in the set-point and sensitivity of the osmoregulatory system.[33] In general, functional tests of AVP levels in patients with DI of variable severity, duration, and cause indicate that AVP secretory capacity must be reduced by at least 75% to 80% for significant polyuria to occur, which also agrees with neuroanatomic studies of cell loss in the SON of dogs with experimental pituitary stalk section[110] and of patients who had undergone pituitary surgery.[111]

Because renal mechanisms for sodium conservation are unimpaired with impaired or absent AVP secretion, there is no accompanying sodium deficiency. Although untreated DI can lead to both hyperosmolality and volume depletion, until the water losses become severe, volume depletion is minimized by osmotic shifts of water from the ICF compartment to the more osmotically concentrated ECF compartment. This phenomenon is not as evident following increases in ECF Na+ concentration, because such osmotic shifts result in a slower increase in the serum Na+ than would otherwise occur. However, when nonsodium solutes such as mannitol are infused, this effect is more obvious owing to the progressive dilutional decrease in serum Na+ caused by translocation of intracellular water to the ECF compartment. Because patients with DI do not have impaired urine Na+ conservation, the ECF volume is generally not markedly decreased and regulatory mechanisms for maintenance of osmotic homeostasis are primarily activated: stimulation of thirst and AVP secretion (to whatever degree the neurohypophysis is still able to secrete AVP). In cases where AVP secretion is totally absent (complete DI), patients are dependent entirely on water intake for maintenance of water balance. However, in cases in which some residual capacity to secrete AVP remains (partial DI), plasma osmolality can eventually reach levels that allow moderate degrees of urinary concentration (see Fig. 13-10 ).



FIGURE 13-10  Relation between plasma AVP levels, urine osmolality, and plasma osmolality in subjects with normal posterior pituitary function (100%) compared with patients with graded reductions in AVP-secreting neurons (to 50%, 25%, and 10% of normal). Note that the patient with a 50% secretory capacity can achieve only half the plasma AVP level and half the urine osmolality of normal subjects at a plasma osmolality of 293 mOsm/kg H2O, but with increasing plasma osmolality, this patient can nonetheless eventually stimulate sufficient AVP secretion to reach a near maximal urine osmolality. In contrast, patients with more severe degrees of AVP-secreting neuron deficits are unable to reach maximal urine osmolalities at any level of plasma osmolality.  (Adapted from Robertson GL: Posterior pituitary. In Felig P, Baxter J, Frohman LA [eds]: Endocrinology and Metabolism. New York, McGraw Hill, 1986, pp 338-386.)




The development of DI following surgical or traumatic injury to the neurohypophysis represents a unique situation and can follow any of several different well-defined patterns. In some patients, polyuria develops 1 to 4 days after injury and resolves spontaneously. Less often, the DI is permanent and continues indefinitely (see previous discussion on the relation between the level of pituitary stalk section and the development of permanent DI). Most interestingly, a “triphasic” response can occur as a result of pituitary stalk transection.[112] The initial DI (first phase) is due to axon shock and lack of function of the damaged neurons. This phase lasts from several hours to several days, and then is followed by an antidiuretic phase (second phase) that is due to the un-controlled release of AVP from the disconnected and de-generating posterior pituitary or from the remaining severed neurons.[134] Overly aggressive administration of fluids during this second phase does not suppress the AVP secretion and can lead to hyponatremia. The antidiuresis can last from 2 to 14 days, after which DI recurs following depletion of the AVP from the degenerating posterior pituitary gland (third phase).[135] Recently, transient hyponatremia without preceding or subsequent DI has been reported following transphenoidal surgery for pituitary microadenomas,[136] which generally occurs 5 to 10 days postoperatively. The incidence may be as high as 30% when such patients are carefully followed, although majority of cases are mild and self-limited. [137] [138] This is due to inappropriate AVP secretion via the same mechanism as in the triphasic response, except that in these cases only the second phase occurs (“isolated second phase”) because the initial neural lobe/pituitary stalk damage is not sufficient to impair AVP secretion sufficiently to produce clinical manifestations of DI.[139]

Once a deficiency of AVP secretion has been present for more than a few days or weeks, it rarely improves even if the underlying cause of the neurohypophysial destruction is eliminated. The major exception to this is in patients with postoperative DI, in which spontaneous resolution is the rule. Although recovery from DI that persists more than several weeks postoperatively is less common, nonetheless well-documented cases of long-term recovery have been reported.[135] The reason for amelioration and resolution is apparent from pathologic and histologic examination of neurohypophyseal tissue following pituitary stalk section. [140] [141] Neurohypophyseal neurons that have intact perikarya are able to regenerate axons and form new nerve terminal endings capable of releasing AVP into nearby capillaries. In animals, this may be accompanied by a bulbous growth at the end of the severed stalk, which represents a new, albeit small, neural lobe. In humans, the regeneration process appears to proceed more slowly, and formation of a new neural lobe has not been noted. Nonetheless, histologic examination of a severed human stalk from a patient 18 months after hypophysectomy has demonstrated reorganization of neurohypophyseal fibers with neurosecretory granules in close proximity to nearby blood vessels, closely resembling the histology of a normal posterior pituitary.[141]

Recognition of the fact that almost all patients with CDI retain a limited capacity to secrete some AVP allows an understanding some otherwise perplexing features of the disorder. For example, in many patients, restricting water intake long enough to raise plasma osmolality by only 1% to 2% induces sufficient AVP secretion to concentrate the urine (Figs. 13-10 and 13-11 [10] [11]). As the plasma osmolality increases further, some patients with partial DI can even secrete enough AVP to achieve near maximal urine osmolalities ( Fig. 13-12 ). However, this should not cause confusion about the diagnosis of DI, because in such patients, the urine osmolality will still be inappropriately low at plasma osmolalities within normal ranges, and they will respond to exogenous AVP administration with further increases in urine osmolality. These responses to dehydration illustrate the relative nature of the AVP deficiency in most cases and underscore the importance of the thirst mechanism to restrict the use of residual secretory capacity under basal conditions of ad libitum water intake.



FIGURE 13-11  Relation between plasma AVP and concurrent plasma osmolality in patients with polyuria of diverse causes. All measurements were made at the end of a standard dehydration test. The shaded area represents the range of normal. In patients with severe (◆) or partial (▴) central DI, plasma AVP was almost always subnormal relative to plasma osmolality. In contrast, the values from patients with dipsogenic (○) or nephrogenic (▪) DI were consistently within or above the normal range.  (From Robertson GL: Diagnosis of diabetes insipidus. In Czernichow AP, Robinson A [eds]: Diabetes Insipidus in Man: Frontiers of Hormone Research. Basel, S Karger, 1985, p 176.)






FIGURE 13-12  Relation between urine osmolality and concurrent plasma AVP in patients with polyuria of diverse causes. All measurements were made at the end of a standard dehydration test. The shaded area represents the range of normal. In patients with severe (◆) or partial (▴) central DI, urine osmolality is normal or supranormal relative to plasma AVP when the latter is submaximal. In patients with nephrogenic DI (▪), urine osmolality is always subnormal for plasma AVP. In patients with dipsogenic DI (○), the relation is normal at submaximal levels of plasma AVP but is usually subnormal when plasma AVP is high.  (From Robertson GL: Diagnosis of diabetes insipidus. In Czernichow AP, Robinson A [eds]: Diabetes Insipidus in Man: Frontiers of Hormone Research. Basel, S Karger, 1985, p 176.)




CDI is also associated with changes in the renal response to AVP. The most obvious change is a reduction in maximal concentrating capacity, which has been attributed to washout of the medullary concentration gradient caused by the chronic polyuria. The severity of this defect is proportional to the magnitude of the polyuria and is independent of its cause.[133] Because of this, the level of urinary concentration achieved at maximally effective levels of plasma AVP is reduced in all types of DI. In patients with CDI, this concentrating abnormality is offset to some extent by an apparent increase in renal sensitivity to low levels of plasma AVP (see Fig. 13-12 ). The cause of this supersensitivity is unknown, but it may reflect upward regulation of AVP V2 receptor expression or function secondary to a chronic deficiency of the hormone.[142]

Osmoreceptor Dysfunction


Extensive literature in animals indicates that the primary osmoreceptors that control AVP secretion and thirst are located in the anterior hypothalamus; lesions of this region in animals, the so-called AV3V area, cause hyperosmolality through a combination of impaired thirst and osmotically stimulated AVP secretion. [30] [31] Initial reports in humans described this syndrome as “essential hypernatremia,”[143] and subsequent studies used the term “adipsic hypernatremia” in recognition of the profound thirst deficits found in most of the patients.[144] Based on the known pathophysiology, all of these syndromes can be grouped together as disorders of osmoreceptor dysfunction.[145]Although the pathologies responsible for this condition can be quite varied, all of the cases reported to date have been due to various degrees of osmoreceptor destruction associated with a variety of different brain lesions, as summarized in Table 13-2 . Many of these are the same types of lesions that can cause CDI, but in contrast to CDI, these lesions usually occur more rostrally in the hypothalamus, consistent with the anterior hypothalamic location of the primary osmoreceptor cells (see Fig. 13-2 ). One lesion unique to this disorder is an anterior communicating cerebral artery aneurysm. Because the small arterioles that feed the anterior wall of the third ventricle originate from the anterior communicating cerebral artery, an aneurysm in this region,[146] but more often following surgical repair of such an aneurysm that typically involves ligation of the anterior communicating artery,[147] produces infarction of the part of the hypothalamus containing the osmoreceptor cells.


The cardinal defect of patients with this disorder is lack of the osmoreceptors that regulate thirst. With rare exceptions, the osmoregulation of AVP is also impaired, although the hormonal response to nonosmotic stimuli remains intact ( Fig. 13-13 ). [148] [149] Four major patterns of osmoreceptor dysfunction have been described as characterized by defects in thirst and/or AVP secretory responses: (1) upward resetting of the osmostat for both thirst and AVP secretion (normal AVP and thirst responses but at an abnormally high plasma osmolality), (2) partial osmoreceptor destruction (blunted AVP and thirst responses at all plasma osmolalities), (3) total osmoreceptor destruction (absent AVP secretion and thirst regardless of plasma osmolality), and (4) selective dysfunction of thirst osmoregulation with intact AVP secretion.[145] Regardless of the actual pattern, the hallmark of this disorder is an abnormal thirst response in addition to variable defects in AVP secretion. Because of this, such patients fail to drink sufficiently as their plasma osmolality rises, and as a result, the new set-point for plasma osmolality rises far above the normal thirst threshold. Unlike patients with CDI whose polydipsia maintains their plasma osmolality within normal ranges, patients with osmoreceptor dysfunction typically have osmolalities in the range of 300 to 340 mOsm/kg H2O. This again underscores the critical role played by normal thirst mechanisms in maintaining body fluid homeostasis; intact renal function alone is insufficient to maintain plasma osmolality within normal limits in such cases.



FIGURE 13-13  Plasma AVP responses to arterial hypotension produced by infusion of trimethephan in patients with central DI (“cranial diabetes insipidus”) and osmoreceptor dysfunction (“adipsic diabetes insipidus). Normal responses in healthy volunteers are shown by the shaded area. Note that despite absent or markedly blunted AVP responses to hyperosmolality, patients with osmoreceptor dysfunction respond normally to baroreceptor stimulation induced by hypotension.  (From Baylis PH, Thompson CJ: Diabetes insipidus and hyperosmolar syndromes. In Becker KL [ed]: Principles and Practice of Endocrinology and Metabolism. Philadelphia, JB Lippincott, 1995, p 257.)




The rate of development and the severity of hyperosmolality and hypertonic dehydration in patients with osmoreceptor dysfunction are influenced by a number of factors. First is the ability to maintain some degree of osmotically stimulated thirst and AVP secretion, which will determine the new set-point for plasma osmolality. Second are environmental influences that affect the rate of water output. When physical activity is minimal and ambient temperature is not elevated, the overall rates of renal and insensible water loss are low and the patient's diet may be sufficient to maintain a relatively normal balance for long periods of time. Anything that increases perspiration, respiration, or urine output greatly accelerates the rate of water loss and thereby uncovers the patient's inability to mount an appropriate compensatory increase in water intake.[14] Under these conditions, severe and even fatal hypernatremia can develop relatively quickly. When the dehydration is only moderate (plasma osmolality 300-330 mOsm/kg H2O), the patient is usually asymptomatic and signs of volume depletion are minimal, but if the dehydration becomes severe, the patient can exhibit symptoms and signs of hypovolemia, including weakness, postural dizziness, paralysis, confusion, coma, azotemia, hypokalemia, hyperglycemia, and secondary hyperaldosteronism (see subsequent section on Clinical Manifestations). In severe cases, there may also be rhabdomyolysis with marked serum elevations in muscle enzymes and occasionally acute renal failure.

However, a third factor also influences the degree of hyperosmolality and dehydration present in these patients. For all cases of osmoreceptor dysfunction, it is important to remember that afferent pathways from the brainstem to the hypothalamus remain intact; therefore, these patients will usually have normal AVP and renal concentrating responses to baroreceptor-mediated stimuli such as hypovolemia and hypotension (see Fig. 13-13 )[149] or to other nonosmotic stimuli such as nausea (see Fig. 13-8 ). [144] [148] This has the effect of preventing severe dehydration, because, as hypovolemia develops, this will stimulate AVP secretion via baroreceptive pathways through the brainstem (see Fig. 13-2 ). Although protective, this effect often causes confusion, because at some times, these patients appear to have DI, yet at other times, they can concentrate their urine quite normally. Nonetheless, the presence of refractory hyperosmolality with absent or inappropriate thirst should alert clinicians to the presence of osmoreceptor dysfunction regardless of apparent normal urine concentration at some times.

In a few patients with osmoreceptor dysfunction, forced hydration has been found to lead to hyponatremia in association with inappropriate urine concentration. [143] [144] This paradoxical defect resembles that seen in the SIADH and has been postulated to be due to two different pathogenic mechanisms. One is continuous or fixed secretion of AVP because of loss of the capacity for osmotic inhibition and stimulation of hormone secretion. These observations, as well as electrophysiologic data,[41] strongly suggest that the osmoregulatory system is bimodal (i.e., it is composed of inhibitory as well as stimulatory input to the neurohypophysis [see Fig. 13-6 ]). The other cause of the diluting defect appears to be hypersensitivity to the antidiuretic effects of AVP, because in some patients, urine osmolality may remain high even when the hormone is undetectable.[144]

Hypodipsia is also a common occurrence in elderly persons in the absence of any overt hypothalamic lesion.[150] In such cases, it is not clear whether the defect is in the hypothalamic osmoreceptors, in their projections to the cortex, or in some other regulatory mechanism. However, in most cases, the osmoreceptor is likely not involved, because both basal and stimulated plasma AVP levels have been found to be normal, or even hyperresponsive, in relation to plasma osmolality in aged humans, with the exception of only a few studies that showed decreased plasma levels of AVP relative to plasma osmolality.[151]

Gestational Diabetes Insipidus


A relative deficiency of plasma AVP can also result from an increase in the rate of AVP metabolism. [95] [152] This condition has been observed only in pregnancy, and therefore, it is generally referred to as gestational DI. It is due to the action of a circulating enzyme called cysteine aminopeptidase (“oxytocinase” or “vasopressinase”) that is normally produced by the placenta in order to degrade circulating oxytocin and prevent premature uterine contractions.[153] Because of the close structural similarity between AVP and oxytocin, this enzyme degrades both peptides. In some patients, plasma levels of oxytocinase/vasopressinase are markedly elevated above those found normally in pregnancy. [152] [154] In others, however, oxytocinase/vasopressinase levels are relatively normal, but the effect of the increase in AVP metabolism may be exacerbated by an underlying subclinical deficiency of AVP secretion.[155]Some of these patients have been noted to have accompanying preeclampsia, acute fatty liver, and coagulopathies, but causal relations between the DI and these abnormalities have not been identified. The relationship of this disorder to the transient NDI of pregnancy[156] is not clear.


The pathophysiology of GDI is similar to that of CDI. The only exception is that the polyuria is usually not corrected by administration of AVP, because this is rapidly degraded just as is endogenous AVP, but it can be controlled by treatment with DDAVP, the AVP V2 receptor agonist that is more resistant to degradation by oxytocinase/vasopressinase.[153] It should be remembered that patients with partial CDI in whom only low levels of AVP can be maintained, or patients with compensated NDI in whom the lack of response of the kidney to AVP may be not be absolute, can be relatively asymptomatic with regard polyuria, but with accelerated destruction of AVP during pregnancy, the underlying DI may become manifest. Consequently, patients presenting with GDI should not be assumed so simply have excess oxytocinase/vasopressinase; rather, these patients should be evaluated for other possible underlying pathologic diagnoses (see Table 13-2 ).[155]

Nephrogenic Diabetes Insipidus


Resistance to the antidiuretic action of AVP is usually due to some defect within the kidney, and is commonly referred to as NDI. It was first recognized in 1945 in several patients with the familial, sex-linked form of the disorder. Subsequently, additional kindreds with the X-linked form of familial NDI were identified. Clinical studies of NDI indicate that symptomatic polyuria is present from birth, plasma AVP levels are normal or elevated, resistance to the antidiuretic effect of AVP can be partial or virtually complete, and the disease affects mostly males and is usually, although not always,[157] mild or absent in carrier females. More than 90% of cases of congenital NDI are caused by mutations of the AVP V2 receptor (see review[158] and Chapter 40 ). Most mutations occur in the part of the receptor that is highly conserved among species and/or is conserved among similar receptors, for example, homologies with AVP V or oxytocin receptors. The effect of some of these mutations on receptor synthesis, processing, trafficking, and function has been studied by in vitro expression. [159] [160] These types of studies show that the various mutations cause several different defects in cellular processing and function of the receptor but can be classified into four general categories based on differences in transport to the cell surface and AVP binding and/or stimulation of adenylyl cyclase: (1) the mutant receptor is not inserted in the membrane; (2) the mutant receptor is inserted in the membrane but does not bind or respond to AVP; (3) the mutant receptor is inserted in the membrane and binds AVP but does not activate adenylyl cyclase; or (4) the mutant protein is inserted into the membrane and binds AVP but responds subnormally in terms of adenylyl cyclase activation. Several recent studies have shown a relation between the clinical phenotype and the genotype and/or cellular phenotype. [159] [161] Approximately 10% of the V2 receptor defects causing congenital NDI are believed to be de novo. This high incidence of de novo cases coupled with the large number of mutations that have been identified hinders the clinical use of genetic identification, because it is necessary to sequence the entire open reading frame of the receptor gene rather than short sequences of DNA; nonetheless, use of automated gene sequencing techniques in selected families has been shown to successfully identify mutations in both patients with clinical disease and asymptomatic carriers.[162] Although most female carriers of the X-linked V2 receptors defect have no clinical disease, some females have been reported with symptomatic NDI.[157] Carriers can have a decreased maximum urine osmolality in response to plasma AVP levels, but are generally asymptomatic because of absence of overt polyuria. Occasionally, a girl manifests severe NDI due to a V2 receptor mutation, which is likely due to inactivation of the normal X chromosome.[163]

Congenital NDI can also result from mutations of the autosomal gene that codes for AQP2, the protein that forms the water channels in renal medullary collecting tubules. When the proband is a girl, it is likely the defect is a mutation of the AQP2 gene on chromosome 12, Q12-13[164] More than 20 different mutations of the AQP2 gene have been described (see review[165] and Chapter 40 ). The patients may be heterozygous for two different recessive mutations[166] or homozygous for the same abnormality from both parents.[167] Because most of these mutations are recessive, the patients usually do not present with a family history of DI unless consanguinity is present. Functional expression studies of these mutations show that all of them result in varying degrees of reduced water transport, because the mutant aquaporins either are not expressed in normal amounts, are retained in various cellular organelles, or simply do not function effectively as water channels. Regardless of the type of mutation, the phenotype of NDI from AQP2 mutations is identical to that produced by V2 receptor mutations. Some of the defects in cellular routing and water transport can be reversed by treatment with chemicals that act like “chaperones,”[168] suggesting that misfolding of the mutant AQP2 may be responsible for misrouting. Similar salutary effects of chaperones have been found to reverse defects in cell surface expression and function of selected mutations of the AVP V2 receptor.[169]

NDI can also be caused by a variety of drugs, diseases, and metabolic disturbances, among them lithium, hypokalemia, and hypercalcemia (see Table 13-2 ). Some of these disorders (e.g., polycystic kidney disease) act to distort the normal architecture of the kidney and interfere with the normal urine concentration process. However, experimental studies in animal models have suggested that many have in common a down-regulation of AQP2 expression in the renal collecting tubules ( Fig. 13-14 , see also Chapters 8 and 9 ). [170] [171] The polyuria associated with potassium deficiency develops in parallel with decreased expression of kidney AQP2, and repletion of potassium reestablishes the normal urinary concentrating mechanism and normalizes renal expression of AQP2.[172] Similarly, hypercalcemia has also been found to be associated with down-regulation of AQP2.[173] A low-protein diet diminishes the ability to concentrate the urine primarily by a decreased delivery of urea to the inner medulla, thus decreasing medullary concentration gradient, but rats on a low-protein diet also appear to down-regulate AQP2, which could be an additional component of the decreased ability to concentrate the urine.[174] Bilateral urinary tract obstruction causes inability to produce a maximum concentration of the urine, and rat models have demonstrated a down-regulation of AQP2, which persists for several days after release of the obstruction.[175] However, it is not yet clear which of these effects on AQP2 expression are primary or secondary and what cellular mechanism(s) are responsible for the down-regulation of AQP2 expression.



FIGURE 13-14  Kidney expression of the water channel aquaporin-2 in various animal models of polyuria and water retention. Note that kidney aquaporin-2 expression is uniformly down-regulated relative to levels in controls in all animal models of polyuria, but up-regulated in animal models of inappropriate antidiuresis. DI +/+, genetic diabetes insipidus; Hyper-Ca, hypercalcemia; Hypo-K, hypokalemia; Urinary obstr, ureteral obstruction.  (From Nielsen S, Kwon TH, Christensen BM, et al: Physiology and pathophysiology of renal aquaporins. J Am Soc Nephrol 10:647-663, 1999.)




Administration of lithium to treat psychiatric disorders is the most common cause of drug-induced NDI and illustrates the multiple mechanisms likely involved in produc-ing this disorder. As many as 10% to 20% of patients on chronic lithium therapy develop some degree of NDI.[176] Lithium is known to interfere with the production of cAMP[177] and produces a dramatic (95%) reduction in kidney AQP2 levels in animals.[178] The defect of aquaporins is slow to correct both in experimental animals and in humans, and in some cases, it can be permanent[179] in association with glomerular or tubulointerstitial nephropathy.[180] Several other drugs that are known to induce renal concentrating defects have also been associated with abnormalities of AQP2 synthesis.[181]


Similar to CDI, renal insensitivity to the antidiuretic effect of AVP also results in the excretion of an increased volume of dilute urine, a decrease in body water, and a rise in plasma osmolality, which by stimulating thirst induces a compensatory increase in water intake. As a consequence, the osmolality of body fluid stabilizes at a slightly higher level that approximates the osmotic threshold for thirst. As in patients with CDI, the magnitude of polyuria and polydipsia varies greatly depending on a number of factors, including the degree of renal insensitivity to AVP, individual differences in the set-points and sensitivity of thirst and AVP secretion, as well as total solute load. It is important to note that the renal insensitivity to AVP need not be complete for polyuria to occur; it is only necessary that the defect is great enough to prevent concentration of the urine at plasma AVP levels achievable under ordinary conditions of ad libitum water intake (i.e., at plasma osmolalities near the osmotic threshold for thirst). Calculations similar to those used for states of AVP deficiency indicate that this requirement is not met until the renal sensitivity to AVP is reduced by more than 10-fold. Because renal insensitivity to the hormone is often incomplete, especially in cases of acquired rather than congenital NDI, many patients with NDI are able to concentrate their urine to varying degrees when they are deprived of water or given large doses of DDAVP.

New knowledge about the renal concentration mechanism from studies of AQP2 expression in experimental animals (see Chapters 8 and 9 ) has suggested that a form of NDI is likely associated with all types of DI, as well as with primary polydipsia. Brattleboro rats have been found to have low levels of kidney AQP2 expression compared to Long-Evans control rats; AQP2 levels are corrected by treatment with AVP or DDAVP, but this process takes 3 to 5 days, during which time urine concentration remains subnormal despite pharmacologic concentrations of AVP.[182] Similarly, physiologic suppression of AVP by chronic overadministration of water produces a down-regulation of AQP2 in the renal collecting duct.[182] Clinically, it is well known that patients with both CDI and primary polydipsia often fail to achieve maximally concentrated urine when they are given DDAVP during a water deprivation test to differentiate among the various causes of DI. This effect has long been attributed to a “washout” of the medullary concentration gradient as a result of the high urine flow rates in polyuric patients, but based on the results of animal studies, it seems certain that at least part of the decreased response to AVP is due to a down-regulation of kidney AQP2 expression. This also explains why it takes time, typically several days, to restore normal urinary concentration after patients with primary polydipsia and CDI are treated with water restriction or antidiuretic therapy.[183]

Primary Polydipsia


Excessive fluid intake also causes hypotonic polyuria and, by definition, polydipsia. Consequently, this disorder must be differentiated from the various causes of DI. Furthermore, it is apparent that despite normal pituitary and kidney function, patients with this disorder share many characteristics of both CDI (i.e., AVP secretion is suppressed as a result of the decreased plasma osmolality) and NDI (kidney AQP2 expression is decreased as a result of the suppressed plasma AVP levels). Many different names have been used to describe patients with excessive fluid intake, but primary polydipsia remains the best descriptor, because it does not presume any particular etiology for the increased fluid intake. Primary polydipsia is often due to a severe mental illness such as schizophrenia, mania, or an obsessive-compulsive disorder,[184] in which case, it is called psychogenic polydipsia. These patients usually deny true thirst and attribute their polydipsia to bizarre motives such as a need to cleanse their body of poisons. Series of polydipsic patients in psychiatric hospital have shown an incidence as high as 42% of patients with some form of polydipsia, and in most reported cases, there is no obvious explanation for the polydipsia.[185] However, primary polydipsia can also be caused by an abnormality in the osmoregulatory control of thirst, in which case it has been termed dipsogenic DI.[186] These patients have no overt psychiatric illness and invariably attribute their polydipsia to a nearly constant thirst. Dipsogenic DI is usually idiopathic, but it can also be secondary to organic structural lesions in the hypothalamus identical to any of the disorders described as causes of CDI, such as neurosarcoidosis of the hypothalamus, tuberculous meningitis, multiple sclerosis, or trauma. Consequently, all polydipsic patients should be evaluated with an MRI scan of the brain before concluding that excessive water intake is due to an idiopathic or psychiatric cause. Primary polydipsia can also be produced by drugs that cause a dry mouth or by any peripheral disorder causing pathologic elevations of renin and/or angiotensin.[107] Finally, primary polydipsia is sometimes caused by physicians, nurses, lay practitioners, or health writers who recommend a high fluid intake for valid (e.g., recurrent nephrolithiasis) or unsubstantiated reasons of health.[187] These patients lack overt signs of mental illness but also deny thirst and usually attribute their polydipsia to habits acquired from years of adherence to their drinking regimen.


The pathophysiology of primary polydipsia is essentially the reverse of that in CDI: The excessive intake of water expands and slightly dilutes body fluids, suppresses AVP secretion, and dilutes the urine. The resultant increase in the rate of water excretion balances the increase in intake, and the osmolality of body water stabilizes at a new, slightly lower level that approximates the osmotic threshold for AVP secretion. The magnitude of the polyuria and polydipsia varies considerably, depending on the nature or intensity of the stimulus to drink. In patients with abnormal thirst, the polydipsia and polyuria are relatively constant from day to day. However, in patients with psychogenic polydipsia, water intake and urine output tend to fluctuate widely, and at times can be quite large.

Occasionally, fluid intake rises to such extraordinary levels that the excretory capacity of the kidneys is exceeded and dilutional hyponatremia develops.[188] There is little question that excessive water intake alone can sometimes be suffici-ent to override renal excretory capacity and produce severe hyponatremia. Although the water excretion rate of normal adult kidneys can generally exceed 20 L/day, maximum hourly rates rarely exceed 1000 mL/hr. Because many psychiatric patients drink predominantly during the day or during intense drinking binges,[189] they can transiently achieve symptomatic levels of hyponatremia with total daily volumes of water intake under 20 L if it is ingested sufficiently rapidly. This likely accounts for many of the cases in which such patients present with maximally dilute urine, accounting for as many as 50% of patients in some studies, and correct quickly via a free water diuresis.[190] The prevalence of this disorder based on hospital admissions for acute symptomatic hyponatremia may have been underestimated, because studies of polydipsic psychiatric patients have shown a marked diurnal variation in serum Na+ (from 141 mEq/L at 7 am to 130 mEq/L at 4 pm), suggesting that many such patients drink excessively during the daytime but then correct themselves via a water diuresis at night.[191] This and other considerations have led to defining this disorder as the psychosis, intermittent hyponatremia, polydipsia (PIP) syndrome.[189]

However, many other cases of hyponatremia with psychogenic polydipsia have been found to meet the criteria for a diagnosis of SIADH, suggesting the presence of nonosmotically stimulated AVP secretion. As might be expected, in the face of much higher than normal water intakes, virtually any impairment of urinary dilution and water excretion can ex-acerbate the development of a positive water balance and thereby produce hypoosmolality. Acute psychosis itself can also cause AVP secretion,[192] which often appears to take the form of a reset osmostat.[184] It is therefore apparent that no single mechanism can completely explain the occurrence of hyponatremia in polydipsic psychiatric patients, but the combination of higher than normal water intakes plus modest elevations of plasma AVP levels from a variety of potential sources appears to account for a significant portion of such cases.

Clinical Manifestations

The characteristic clinical symptoms of DI are the polyuria and polydipsia that result from the underlying impairment of urinary concentrating mechanisms, which have already been covered in the previous section discussing pathophysiology of specific types of DI. Interestingly, patients with DI typically describe a craving for cold water, which appears to quench their thirst better.[64] Patients with CDI also typically describe a precipitous onset of their polyuria and polydipsia, which simply reflects the fact that urinary concentration can be maintained fairly well until the number of AVP-producing neurons in the hypothalamus decreases to 10% to 15% of normal, after which plasma AVP levels decrease to the range at which urine output increases dramatically.

However, patients with DI, and particularly those with osmoreceptor dysfunction syndromes, can also present with varying degrees of hyperosmolality and dehydration, depending on their overall hydration status. It is therefore important to be aware of the clinical manifestations of hyperosmolality as well. These can be divided into the signs and symptoms produced by dehydration, which are largely cardiovascular, and those caused by the hyperosmolality itself, which are predominantly neurologic and reflect brain dehydra-tion as a result of osmotic water shifts out of the central nervous system (CNS). Cardiovascular manifestations of hypertonic dehydration include hypotension, azotemia, acute tubular necrosis secondary to renal hypoperfusion or rhabdomyolysis, and shock. [193] [194] Neurologic manifestations range from nonspecific symptoms such as irritability and cognitive dysfunction to more severe manifestations of hypertonic encephalopathy such as disorientation, decreased level of consciousness, obtundation, chorea, seizures, coma, focal neurologic deficits, subarachnoid hemorrhage, and cerebral infarction. [193] [195] The severity of symptoms can be roughly correlated with the degree of hyperosmolality, but individual variability is marked, and for any single patient, the level of serum Na+ at which symptoms will appear cannot be accurately predicted. Similar to hypoosmolar syndromes, the length of time over which hyperosmolality develops can markedly affect the clinical symptomatology. Rapid development of severe hyperosmolality is frequently associated with marked neurologic symptoms, whereas gradual development over several days or weeks generally causes milder symptoms. [193] [194] [195] [196] In this case, the brain counteracts osmotic shrinkage by increasing intracellular content of solutes. These include electrolytes such as potassium and a variety of organic osmolytes, which previously had been called “idiogenic osmoles”; for the most part, these are the same organic osmolytes that are lost from the brain during adaptation to hypoosmolality.[197] The net effect of this process is to protect the brain against excessive shrinkage during sustained hyperosmolality. However, once the brain has adapted by increasing its solute content, rapid correction of the hyperosmolality can produce brain edema, because it takes a finite time (24-48 hr in animal studies) to dissipate the accumulated solutes, and until this process has been completed, the brain will accumulate excess water as plasma osmolality is normalized.[198] This effect is most often seen in dehydrated pediatric patients who can develop seizures with rapid rehydration,[199] but it has been described only rarely in adults, including the most severely hyperosmolar patients with nonketotic hyperglycemic hyperosmolar coma.

Differential Diagnosis

Before beginning involved diagnostic testing to differentiate among the various forms of DI and primary polydipsia, the presence of true hypotonic polyuria should be established by measurement of a 24-hour urine for volume and osmolality. Generally accepted standards are that 24-hour urine volume should exceed 50 mL/kg BW with an osmolality less than 300 mOsm/kg H2O.[200] Simultaneously, there should be a determination of whether the polyuria is due to an osmotic agent such as glucose, or intrinsic renal disease. Routine laboratory studies and the clinical setting will generally distinguish these disorders; diabetes mellitus and other forms of solute diuresis usually can be excluded by the history, a routine urinalysis for glucose, or measurement of the solute excretion rate (urine osmolality × urine volume in liters <15 mOsm/kg BW/day). There is universal agreement that the diagnosis of DI requires stimulating AVP secretion osmotically and then measuring the adequacy of the secretion by either direct measurement of plasma AVP levels or indirect assessment by urine osmolality.

In a patient who is already hyperosmolar with submaximally concentrated urine (i.e., urine osmolality <800 mOsm/kg H2O), the diagnosis is straightforward and simple: Primary polydipsia is ruled out by the presence of hyperosmolality,[200] confirming a diagnosis of DI. CDI can then be distinguished from NDI by evaluating the response to administered AVP (5 units SC) or, preferably, the AVP V2 receptor agonist DDAVP (1-2 mg subcutaneously or intravenously). A significant increase in urine osmolality within 1 to 2 hours after injection indicates insufficient endogenous AVP secretion and, therefore, CDI, whereas an absent response indicates renal resistance to AVP effects and, therefore, NDI. Although conceptually simple, interpretational difficulties can arise because the water diuresis produced by AVP deficiency in DDI produces a washout of the renal medullary concentrating gradient as well as down-regulation of kidney AQP2 water channels, as discussed previously, so that initial increases in urine osmolality in response to administered AVP or DDAVP are not as great as would be expected. Generally, increases of urine osmolality greater than 50% reliably indicate CDI and responses of less than 10% indicate NDI, but responses between 10% and 50% are indeterminate.[133] For this reason, plasma AVP levels should be measured to aid in this distinction: Hyperosmolar patients with NDI will have clearly elevated AVP levels, whereas those with CDI will have absent (complete) or blunted (partial) AVP responses relative to their plasma osmolality ( Fig. 13-15 ). Because it will not be known beforehand which patients will have diagnostic versus indeterminate responses to AVP or DDAVP, a plasma AVP level should be drawn prior to AVP or DDAVP administration in patients presenting with hyperosmolality and inadequately concentrated urine without a solute diuresis.



FIGURE 13-15  Effects of fluid deprivation and subsequent AVP (Pitressin) administration on urine osmolality in 156 patients with polyuria of diverse causes. The shaded areas indicate the range of values in healthy adults. Note that, although AVP responses tended to be greater in patients with central (neurogenic) DI, the overlap between the three groups was significant.



Because patients with DI have intact thirst mechanisms, most often they do not present with hyperosmolality, but rather with a normal plasma osmolality and serum Na+ and symptoms of polyuria and polydipsia. In these cases, it is most appropriate to perform a fluid deprivation test. The relative merits of the indirect fluid deprivation test (the Miller-Moses test[201]) versus direct measurement of plasma AVP levels after a period of fluid deprivation[133] has been debated in the literature for the last 2 decades, with substantial pros and cons in support of each of these tests. The standard indirect test has a long track record of successfully making an appropriate diagnosis in the large majority of cases, generally yields interpretable results by the end of the test, and does not require sensitive assays for the notoriously difficult measurement of plasma AVP levels. [202] [203] However, maximum urine concentrating capacity is well known to be variably reduced in all forms of DI as well as primary polydipsia,[133] and as a result, the absolute levels of urine osmolality achieved during fluid deprivation and after AVP administration are reduced to overlapping degrees in patients with partial CDI, partial NDI, and primary polydipsia (see Fig. 13-15 ). Measurements of basal plasma osmolality or serum Na+ are of little use, because they also overlap considerably among these disorders.[200] And although association with certain diseases, surgical procedures, or family history often helps to differentiate among these disorders, sometimes the clinical setting may not be helpful because certain diseases such as sarcoidosis, tuberculous meningitis, and other hypothalamic pathologies can cause more than one type of DI (see Table 13-2 ). Consequently, a simpler approach that has been proposed is to measure plasma or urine AVP before and during a suitable osmotic stimulus such as fluid restriction or hypertonic NaCl infusion and plot the results as a function of the concurrent plasma osmolality or plasma sodium concentration (see Figs. 13-11 and 13-12 [11] [12]).[204] [205] Using a highly sensitive and validated research assay for plasma AVP determinations, this approach has been shown to provide a definite diagnosis of most cases if the final level of plasma osmolality or sodium achieved is above the normal range (>295 mOsm/kg H2O or 145 mmol/L, respectively). The diagnostic effectiveness of this approach derives from the fact that the magnitude of the AVP response to osmotic stimulation is not appreciably diminished by chronic overhydration[184] or dehydration. Hence, the relationship of plasma AVP to plasma osmolality is usually within or above normal limits in NDI and primary polydipsia. In most cases, these two disorders can then be distinguished by measuring urine osmolality before and after the dehydration test and relating these values to the concurrent plasma AVP concentrations (see Fig. 13-12 ). However, because maximal concentrating capacity can be severely blunted in patients with primary polydipsia, it is often better to analyze the relationship under basal, nondehydrated conditions when plasma AVP is not elevated. Because of the solute diuresis that often ensues following infusion of hypertonic NaCl, measurements of urine osmolality or AVP excretion are unreliable indicators of changes in hormone secretion and are of little or no diagnostic value when this procedure is used to increase osmolality to greater than 295 mOsm/kg H2O. Given the proven usefulness of both the indirect and the direct approaches, a combined fluid deprivation test that synthesizes the crucial aspects of both tests can easily be performed (Table 13-3 ) and, in many cases, will allow interpretation of both the plasma AVP levels and the response to an AVP challenge.

TABLE 13-3   -- Fluid Deprivation Test for the Diagnosis of Diabetes Insipidus




Initiation of the deprivation period depends on the severity of the DI; in routine cases, the patient should be made NPO after dinner, whereas in cases with more severe polyuria and polydipsia, this may be too long a period without fluids and the water deprivation should be begun early on the morning (e.g., 6 AM) of the test.



Obtain plasma and urine osmolality, serum electrolytes and a plasma AVP level at the start of the test.



Measure urine volume and osmolality hourly or with each voided urine.



Stop the test when body weight decreases by ≥3%, the patient develops orthostatic blood pressure changes, the urine osmolality reaches a plateau (i.e., <10% change over two or three consecutive measurements), or the serum Na+ >145 mmol/L.



Obtain plasma and urine osmolality, serum electrolytes, and a plasma AVP level at the end of the test, when the plasma osmolality is elevated, preferably >300 mOsm/kg H2O.



If the serum Na+ <146 mmol/L or the plasma osmolality <300 mOsm/kg H2O when the test is stopped, then consider a short infusion of hypertonic saline (3% NaCl at a rate of 0.1 ml/kg/min for 1–2 hr) to reach these endpoints.



If hypertonic saline infusion is not required to achieve hyperosmolality, administer AVP (5 U) or DDAVP (1 μg) SC and continue following urine osmolality and volume for an additional 2 hr.




An unequivocal urine concentration after AVP/DDAVP (>50% increase) indicates CDI and an unequivocal absence of urine concentration (<10%) strongly suggests nephrogenic DI (NDI) or primary polydipsia (PP).



Differentiating between NDI and PP, as well as for cases in which the increase in urine osmolality after AVP/DDAVP administration is more equivocal (e.g., 10%–50%), is best done using the relation between plasma AVP levels and plasma osmolality obtained at the end of the dehydration period and/or hypertonic saline infusion and the relation between plasma AVP levels and urine osmolality under basal conditions (see Figs. 13-11 and 13-12 [11] [12]).




With use of the fluid deprivation test with plasma AVP determinations, greater than 95% of all cases of polyuria and polydipsia can be diagnosed accurately. A useful approach in the remaining indeterminate cases is to conduct a closely monitored trial with standard therapeutic doses of DDAVP. If this treatment abolishes thirst and polydipsia as well as polyuria for 48 to 72 hours without producing water intoxication, the patient most likely has uncomplicated CDI. Conversely, if the treatment abolishes the polyuria but has no or a lesser effect on thirst or polydipsia and results in the development of hyponatremia, it is more likely that the patient has some form of primary polydipsia. If DDAVP has no effect over this time interval, even when given by injection, it is virtually certain that the patient has some form of NDI.

As might be expected, most patients with DI will also exhibit a subnormal increase in AVP secretion in response to nonosmotic stimuli, such as hypotension, nausea, and hypoglycemia.[204] For diagnostic purposes, however, these nonosmotic tests of neurohypophyseal function do not provide any advantage over dehydration or hypertonic NaCl infusion, because orthostatic, emetic, and glucopenic stimuli are difficult to control or quantitate and generally cause a markedly variable AVP response. A more fundamental disadvantage with all nonosmotic stimuli is the possibility of a false-positive or false-negative results, because there are patients who exhibit little or no rise in AVP after hypotension or emesis yet lack polyuria and have a normal response to osmotic stimuli. Conversely, patients with osmoreceptor dysfunction exhibit little or no AVP response to hypertonic NaCl but have a normal increase in response to induced hypotension (see Fig. 13-13 ).[149]

MRI has also proved to be useful in diagnosing DI. In normal subjects, the posterior pituitary produces a characteristic bright signal in the posterior part of the sella turcica similar on T1-weighted images, usually best seen in sagittal views.[206] This was originally believed to represent fatty tissue, but more recent evidence indicates that the bright spot is actually due to the stored hormone in neurosecretory granules.[207] An experimental study done in rabbits subjected to dehydration for varying periods of time showed a linear correlation between pituitary AVP content and the signal intensity of the posterior pituitary by MRI.[208] As might be expected from the fact that destruction of more than 85% to 90% of the neurohypophysis is necessary to produce clinical symptomatology of DI, this signal has been found to be most always absent in patients with CDI in multiple studies.[209] However, as with any diagnostic test, clinical usefulness is dependent on the sensitivity and specificity of the test. Although earlier studies using small numbers of subjects demonstrated the presence of the bright spot in all normal subjects, subsequent larger studies reported an age-related absence of a pituitary bright spot in up to 20% of normal subjects.[210] Conversely, some studies have reported the presence of a bright spot in patients with clinical evidence of DI.[211] This may be because some patients with partial CDI have not yet progressed to the point of depletion of all neurohypophyseal reserves of AVP or because a persistent bright spot in patients with DI might be due to pituitary content of oxytocin rather than AVP. In support of this, it is known that oxytocinergic neurons are more resistant to destruction by trauma than are vasopressinergic neurons in both rats[212] and humans.[23] The presence of a positive posterior pituitary bright spot has been variably reported in other polyuric disorders. In primary polydipsia, the bright spot is usually seen,[209] consistent with studies in animals in which even prolonged lack of secretion of AVP caused by hyponatremia did not cause a decreased content of AVP in the posterior pituitary.[27] In NDI, the bright spot has been reported to be absent in some patients but present in others.[120] Consequently, specificity is lacking to use the MRI routinely as a diagnostic screening test for DI. Nonetheless, the sensitivity is sufficient to allow an approximately 95% probability that a patient with a bright spot on MRI does not have CDI. Thus, MRI is more useful for ruling out than for ruling in a diagnosis of CDI.

Additional useful information can be gained through the MRI via assessment of the pituitary stalk. Enlargement of the stalk beyond 2 to 3 mm is generally considered to be pathologic[213] and can be due to multiple disease processes.[214] Consequently, when the MRI reveals thickening of the stalk, especially with absence of the posterior pituitary bright spot, systemic diseases should be searched for diligently, including cerebrospinal fluid and plasma beta-human chorionic gonadotropin and alpha-fetoprotein measurement for evaluation of suprasellar germinoma, chest imaging and cerebrospinal fluid and plasma angiotensin-converting enzyme (ACE) levels for evaluation of sarcoidosis, and bone and skin surveys for evaluation of histiocytosis. When a diagnosis is still in doubt, the MRI should be repeated every 3 to 6 months. Continued enlargement, especially in children over the first 3 years of follow-up, suggest a germinoma and mandates a biopsy, whereas a decrease in the size of the stalk over time is more indicative of an inflammatory process such as lymphocytic infundibuloneurohypophysitis.[215]


The general goals of treatment of all forms of DI are (1) a correction of any preexisting water deficits and (2) a reduction in the ongoing excessive urinary water losses. The specific therapy required ( Table 13-4 ) will vary according to both the type of DI present and the clinical situation. Awake, ambulatory patients with normal thirst have relatively little body water deficit, but benefit greatly by alleviation of the polyuria and polydipsia that disrupt their normal activities. In contrast, comatose patients with acute DI after head trauma are unable to drink in response to thirst, and in these patients, progressive hyperosmolality can be life-threatening.

TABLE 13-4   -- Therapies for the Treatment of Diabetes Insipidus


Antidiuretic agents
Arginine vasopressin (Pitressin)
1-Deamino-8-D-arginine vasopressin (Desmopressin; DDAVP)

Antidiuresis-enhancing agents
Prostaglandin synthetase inhibitors (indomethacin, ibuprofen, tolmetin)

Natriuretic agents
Thiazide diuretics




The TBW deficit in a hyperosmolar patient can be estimated using the formula:

TBW deficit = 0.6 × premorbid weight × (1 - 140/Na+)

where Na+ is the serum sodium concentration in mmol/L and weight is in kg. This formula depends on three assumptions: (1) TBW is approximately 60% of the premorbid BW, (2) no body solute was lost as the hyperosmolality developed, and (3) the premorbid serum Na+ was 140 mEq/L.

To reduce the risk of CNS damage from protracted exposure to severe hyperosmolality, in most cases, the plasma osmolality should be rapidly lowered in the first 24 hours to the range of 320 to 330 mOsm/kg H2O, or by approximately 50%. Plasma osmolality may be estimated most easily as twice the serum Na+ if there is no hyperglycemia, and measured osmolality may be substituted if azotemia is not present. As discussed earlier, the brain increases intracellular osmolality by increasing content of a variety organic osmolytes as a protection against excessive shrinkage during hyperosmolality.[197] Because these osmolytes cannot be immediately dissipated, further correction to a normal plasma osmolality should be spread over the next 24 to 72 hours to avoid producing cerebral edema during treatment.[198] This is especially important in children,[216] in whom several studies have indicated that limiting correction of hypernatremia to a maximal rate of no greater than 0.5 mmol/L/hr prevents the occurrence of symptomatic cerebral edema with seizures. [199] [217] In addition, the possibility of associated thyroid or adrenal insufficiency should also be kept in mind, because patients with CDI caused by hypothalamic masses can have associated deficiencies of anterior pituitary function.

The previous formula does not take into account ongoing water losses and is, at best, a rough estimate. Frequent serum and urine electrolyte determinations should be made, and the administration rate of oral water, or intravenous 5% dextrose in water, should be adjusted accordingly. Note, for example, that the estimated deficit of a 70-kg patient whose serum Na+ is 160 mEq/L is 5.25 L of water. In such an individual, administration of water at a rate greater than 200 mL/hr would be required simply to correct the established deficit over 24 hours. Additional fluid would be needed to keep up with ongoing losses until a definitive response to treatment has occurred. The therapeutic agents available for the treatment of DI are shown in Table 13-4 . Water should be considered a therapeutic agent because, when ingested or infused in sufficient quantity, there is no abnormality of body fluid volume or composition.

As noted previously, in most patients with DI, thirst remains intact, and the patients will drink sufficient fluid to maintain a relatively normal fluid balance. Patients with known DI should therefore be treated in order to decrease the patient's polyuria and polydipsia to acceptable levels that allow the patient to maintain a normal lifestyle. Because the major goal of therapy is improvement in symptomatology, the therapeutic regimen prescribed should be individually tailored to each patient in order to accommodate her or his needs. The safety of the prescribed agent and use of a regimen that avoids potential detrimental effects of overtreatment are primary considerations because of the relatively benign course of DI in most cases, and the potential adverse consequences of hyponatremia. Available treatments are summarized later, and their use is discussed separately for different types of DI.

Arginine Vasopressin (Pitressin).

Pitressin is a synthetic form of naturally occurring human AVP. The aqueous solution contains 20 units/mL. Because of the drug's relatively short half-life (2-4-hour duration of antidiuretic effect) and propensity to cause acute increases in blood pressure when given as a bolus intravenously, this route of administration should generally be avoided. This agent is mainly used for acute situations such as postoperative DI. However, repeated dosing is required, unless a continuous infusion is used, and the frequency of dosing or infusion rate must be titrated to achieve the desired reduction in urine output (see subsequent discussion of postoperative DI).


DDAVP is an agonist of the AVP V2 receptor that was developed for therapeutic use because it has a significantly longer half-life than AVP (8-20-hr duration of antidiuretic effect) and is devoid of the latter's pressor activity by virtue of absence of activation of AVP V receptors on vascular smooth muscle.[218] As a result of these advantages, it is the drug of choice for both acute and chronic administration in patients with CDI.[219] Several different preparations are available. The intranasal form is provided as an aqueous solution containing 100 mg/mL in a bottle with either a calibrated rhinal tube, which requires specific training to use appropriately, or a nasal spray delivering a metered dose of 10 mg in 0.1 mL. An oral preparation is also available in doses of 0.1 or 0.2 mg. Neither the intranasal or the oral preparations should be utilized in an acute emergency setting, where it is essential that the patient achieve a therapeutic dose of the drug; in this case, the parenteral form should always be used. This is supplied as a solution containing 4 mg/mL and may be given by intravenous, intramuscular, or subcutaneous route. The parenteral form is approximately 5 to 10 times more potent than the intranasal preparation, and the recommended dosage is 1 to 2 mg every 8 to 12 hours. For both the intranasal and the parenteral preparations, increasing the dose generally has the effect of prolonging the duration of antidiuresis for several hours rather than increasing its magnitude; consequently, altering the dose can be useful to reduce the required frequency of administration. However, given the cost of the drug, it is often more cost-effective, and sometimes more efficacious, to use a smaller dose more frequently than a larger dose less frequently.

Chlorpropamide (Diabinese).

Primarily used as an oral hypoglycemic agent, this sulfonylurea also potentiates the hydro-osmotic effect of AVP in the kidney. Chlorpropamide has been reported to reduce polyuria by 25% to 75% in patients with CDI. This effect appears to be independent of the severity of the disease and is associated with a proportional rise in urine osmolality, correction of dehydration, and elimination of the polydipsia similar to that caused by small doses of AVP or DDAVP.[200] The major site of action of chlorpropamide appears to be at the renal tubule to potentiate the hydro-osmotic action of circulating AVP, but there is also evidence of a pituitary effect to increase release of AVP as well; the latter effect may account for the observation that chlorpropamide can produce significant antidiuresis even in patients with severe CDI and presumed near-total AVP deficiency.[200] The usual dose is 250 to 500 mg/day with a response noted in 1 to 2 days and a maximum antidiuresis in 4 days. It should be remembered that this is an off-label use of chlorpropamide; it should not be used in pregnancy or in children, it should never be used in an acute emergency setting in which achievement of rapid antidiuresis is necessary, and it should be avoided in patients with concurrent hypopituitarism because of the increased risk of hypoglycemia. Other sulfonylureas share chlorpropamide's effect but are generally less potent. In particular, the newer generation oral hypoglycemic agents such as glipizide and glyburide are virtually devoid of any AVP-potentiating effects.

Prostaglandin Synthase Inhibitors.

Prostaglandins have complex effects both in the CNS and in the kidney, most of which are incompletely understood at this time owing to the variety of different prostaglandins and their multiplicity of cellular effects. In the brain, intracerebroventricular infusion of E prostaglandins stimulate AVP secretion[220] and administration of prostaglandin synthase inhibitors attenuate osmotically stimulated AVP secretion.[221] However, in the kidney, prostaglandin E2(PGE2) has been reported to inhibit AVP-stimulated generation of cAMP in the cortical collecting tubule by interacting with Gi.[222] Thus, the effect of prostaglandin synthetase inhibitors to sensitize AVP effects in the kidney likely result from enhanced cAMP generation upon AVP binding to the V2 receptor. The predominant renal effects of these agents is demonstrated by the fact that clinically they successfully reduce urine volume and free water clearance even in patients with NDI of different etiologies. [223] [224]

Natriuretic Agents.

Thiazide diuretics have a paradoxical antidiuretic effect in patients with CDI.[225] However, given the better antidiuretic agents available for treatment of CDI, its main therapeutic use is in NDI. Hydrochlorothiazide at doses of 50 to 100 mg/day usually reduces urine output by approximately 50%, and efficacy can be further enhanced by restricting sodium intake. Unlike DDAVP or the other antidiuresis-enhancing drugs, these agents are equally effective in most forms of NDI (see later).

Central Diabetes Insipidus

Patients with CDI should generally be treated with intranasal or oral DDAVP. Unless the hypothalamic thirst center is also affected by the primary lesion causing superimposed osmoreceptor dysfunction, these patients will develop thirst when the plasma osmolality increases by only 2% to 3%.[200] Severe hyperosmolality is therefore not a risk in the patient who is alert, ambulatory, and able to drink in response to perceived thirst. Polyuria and polydipsia are thus inconvenient and disruptive, but not life-threatening. However, hypo-osmolality is largely asymptomatic and may be progressive if water intake continues during a period of continuous antidiuresis. Therefore, treatment must be designed to minimize polyuria and polydipsia but without an undue risk of hyponatremia from overtreatment.

Treatment should be individualized to determine optimal dosage and dosing interval. Although tablets offer greater convenience and are generally preferred by patients, it is useful to start with the nasal spray initially because of greater consistency of absorption and physiologic effect, and then switch to the oral tablets only after the patient is comfortable with use of the intranasal preparation to produce antidiuresis; having tried both preparations, the patient can then chose which he or she prefers for long-term usage. Because of variability in response among patients, it is desirable to determine the duration of action of individual doses in each patient.[226] A satisfactory schedule can generally be determined using modest doses, and the maximum dose needed is rarely above 0.2 mg orally or 10 mg (one nasal spray) given two, or occasionally three, times daily.[227] Even in these cases, multiple small doses may be preferred because of the relatively high cost of DDAVP. In selected cases, chlorpropamide may lower the required DDAVP dose and produce an additional, though limited, economy. These doses generally produce plasma DDAVP levels many times those required to produce maximum antidiuresis but obviate the need for more frequent treatment. Rarely, once-daily dosing suffices. In a few patients, the effect of intranasal or oral DDAVP is erratic, probably as a result of variable interference with absorption from the gastrointestinal tract or nasal mucosa. This variability can be reduced and the duration of action prolonged by administering the drug on an empty stomach[228] or after thorough cleansing of the nostrils. Resistance caused by antibody production has not been reported to date.

Hyponatremia is a rare complication of DDAVP therapy and occurs only if the patient is continually antidiuretic while maintaining a fluid intake sufficient to become volume expanded and natriuretic. Absence of thirst in this circumstance is protective; but also, most patients with DI on standard therapy are not continuously maximally antidiuretic. There are reports of hyponatremia in patients with normal AVP function, and presumably normal thirst, when they are given DDAVP to treat hemophilia and von Willebrand's disease,[229] and in children treated with DDAVP for primary enuresis.[230] In these cases, the hyponatremia can develop rapidly and is often first noted by the onset of convulsions and coma.[231] Severe hyponatremia in patients with DI being treated with DDAVP can be avoided by monitoring serum electrolytes frequently during initiation of therapy. Patients who show a tendency to develop low serum sodium concentrations that do not respond to recommended decreases in fluid intake should then be instructed to delay a scheduled dose of DDAVP once or twice a week so that polyuria recurs, thereby allowing any excess retained fluid to be excreted.[203]

Acute postsurgical DI occurs relatively frequently following surgery that involves the suprasellar hypothalamic area, but several confounding factors must be considered. These patients often receive stress doses of glucocorticoids, and the resulting hyperglycemia with glucosuria may confuse a diagnosis of DI. Thus, the blood glucose must first be brought under control to eliminate an osmotic diuresis as the cause of the polyuria. In addition, excess fluids administered intravenously may be retained perioperatively, but then excreted normally postoperatively. If this large output is matched with continued intravenous input, an incorrect diagnosis of DI may be made based on the resulting polyuria. Therefore, if the serum Na+ is not elevated concomitantly with the polyuria, the rate of parenterally administered fluid should be slowed with careful monitoring of serum Na+ and urine output to establish the diagnosis. Once a diagnosis of DI is confirmed, the only acceptable pharmacologic therapy is an antidiuretic agent. However, because many neurosurgeons fear water overload and brain edema after this type of surgery, the patient is sometimes treated with only intravenous fluid replacement for a considerable time before the institution of ADH therapy (see the potential benefits of this approach later in this chapter). If the patient is awake and able to respond to thirst, one can treat with an ADH and allow the patient's thirst to be the guide for water replacement. However, if the patient is unable to respond to thirst, either because of a decreased level of consciousness or from hypothalamic damage to the thirst center, fluid balance must be maintained by intravenously administered fluid. The urine osmolality and serum Na+ must be checked every several hours during the initial therapy, and then at least daily until stabilization or resolution of the DI. Caution must also be exercised regarding the volume of water replacement, because excess water administered during continued administration of AVP or DDAVP can create a syndrome of inappropriate antidiure-sis and potentially severe hyponatremia. Recent studies in experimental animals have indicated that DDAVP-induced hyponatremia markedly impairs survival of vasopressin neurons after pituitary stalk compression,[212] suggesting that overhydration with subsequent decreased stimulation of the neurohypophysis may also increase the likelihood of permanent DI postoperatively.

Postoperatively, DDAVP may be given parenterally in a dose of 1 to 2 mg subcutaneously, intramuscularly, or intravenously. The intravenous route is preferable, because it obviates any concern about absorption, is not associated with significant pressor activity, and has the same total duration of action as the other parenteral routes. A prompt reduction in urine output should occur, and the duration of antidiuretic effect is generally 6 to 12 hours. Usually, the patient is hypernatremic with relatively dilute urine when therapy is started. One should follow the urine osmolality and urine volume to be certain the dose was effective, and check the serum Na+ at frequent intervals to ensure some improvement of hypernatremia. It is generally advisable to allow some return of the polyuria before administration of subsequent doses of DDAVP, because postoperative DI is often transient and return of endogenous AVP secretion will become apparent by a lack of return of the polyuria. Also, in some cases, transient postoperative DI is part of a “triphasic” pattern that has been well described following pituitary stalk transection (see previous discussion). Because of this possibility, allowing a return of polyuria before redosing with DDAVP will allow earlier detection of a potential second phase of inappropriate antidiuresis and decrease the likelihood of producing symptomatic hyponatremia by continuing antidiuretic therapy and intravenous fluid administration when it is not required. Some clinicians have recommended using a continuous intravenous infusion of a dilute solution of AVP to control DI postoperatively. Algorithms for continuous AVP infusion in postoperative and posttraumatic DI in pediatric patients have begun at infusion rates of 0.25 to 1.0 mU/kg/hr and titrated the rate using urine specific gravity (goal of 1.010-1.020) and urine volume (goal 2-3 mL/kg/hr) as a guide to adequacy of the antidiuresis. [232] [233] Although pressor effects have not been reported at these infusion rates and the antidiuretic effects are quickly reversible in 2 to 3 hours, it should be remembered that use of continuous infusions versus intermittent dosing will not allow an assessment of when the patent has recovered from transient DI or entered the second phase of a triphasic response. If DI persists, the patient should eventually be switched to maintenance therapy with intranasal or oral preparations of DDAVP for treatment chronic DI.

Acute traumatic DI can occur after injuries to the head, usually a motor vehicle accident. DI is more common with deceleration injuries that result in a shearing action on the pituitary stalk and/or cause hemorrhagic ischemia of the hypothalamus and/or posterior pituitary.[135] Similar to the onset of postsurgical DI, posttraumatic DI is usually recognized by hypotonic polyuria in the face of an increased plasma osmolality. The clinical management is similar to that of postsurgical DI as outlined previously, except that the possibility of anterior pituitary insufficiency must also be considered in such cases, and the patient should be given stress doses of glucocorticoids (e.g., hydrocortisone, 100 mg intravenously every 8 hr) until anterior pituitary function can be definitively evaluated.

Osmoreceptor Dysfunction

Acutely, patients with hypernatremia due to osmoreceptor dysfunction should be treated the same as any hyperosmolar patient by replacing the underlying free water deficit as described at the beginning of this section. The long-term management of osmoreceptor dysfunction syndromes requires a thorough search for a potentially treatable causes (see Table 13-2 ) in conjunction with the use of measures to prevent recurrence of dehydration. Because the hypodipsia cannot be cured, and rarely if ever improves spontaneously, the mainstay of management is education of the patient and her or his family about the importance of continuously regulating her or his fluid intake in accordance with the hydration status. This is never accomplished easily in such patients, but can be done most efficaciously by establishing a daily schedule of water intake based on changes in BW and regardless of the patient's thirst. In effect, a “prescription” for daily fluid intake must be written for these patients, because they will not drink spontaneously. In addition, if the patient has polyuria, DDAVP should also be given as for any patient with DI. The success of this regimen should be monitored periodically (weekly at first, later every month depending on the stability of the patient) by measuring serum Na+. In addition, the target weight (at which hydration status and serum Na+concentration are normal) may need to be recalculated periodically to allow for growth in children or changes in body fat in adults.

Gestational Diabetes Insipidus

The polyuria of GDI is usually not corrected by administration of AVP itself because this is rapidly degraded by high circulating levels of oxytocinase/vasopressinase just as is endogenous AVP. The treatment of choice is DDAVP, because this synthetic AVP V2 receptor agonist is not destroyed by the cysteine aminopeptidase (oxytocinase/vasopressinase) in the plasma of pregnant women[234] and to date appears to be safe for both the mother and the child.[235] [236] DDAVP has only 2% to 25% the oxytocic activity of AVP[219] and can be used with minimal stimulation of the oxytocin receptors in the uterus. Doses should be titrated to individual patients, because higher doses and more frequent dosing intervals are sometimes required because of the increased degradation of the peptide. However, physicians should remember that the naturally occurring volume expansion and reset osmostat that occurs in pregnancy maintains the serum Na+ at a lower level during pregnancy.[39] During delivery, these patients can maintain adequate oral intake and continued administration of DDAVP, but physicians should be cautious about overadministration of fluid parenterally during delivery because these patients will not be able to excrete the fluid and will be susceptible to the development of water intoxication and hyponatremia. After delivery, oxytocinase/vasopressinase decreases in plasma within several days and, depending on the etiology of the DI, the patients may have disappearance of the disorder or become asymptomatic with regard to fluid intake and urine volume.[237]

Nephrogenic Diabetes Insipidus

By definition, patients with NDI are resistant to the effects of AVP. Some patients with NDI can be treated by eliminating the drug (e.g., lithium) or disease (e.g., hypercalcemia) responsible for the disorder. For many others, however, including those with the genetic forms, the only practical form of treatment at present is to restrict sodium intake and administer a thiazide diuretic either alone[225] or in combination with prostaglandin synthetase inhibitors[238] or amiloride. [239] [240] The natriuretic effect of the thiazide class of diuretics is conferred by their ability to block sodium absorption in the renal cortical diluting site. When combined with dietary sodium restriction, the drugs cause modest hypovolemia. This stimulates isotonic proximal tubular solute reabsorption and diminishes solute delivery to the more distal diluting site, where ex-perimental studies have indicated that thiazides also act to enhance water reabsorption in the inner medullary collecting duct independently of AVP.[241] Together, these effects markedly diminish renal diluting ability and free water clearance independently of any action of AVP. Thus, agents of this class are the mainstay of therapy for NDI. Monitoring for hypokalemia is recommended, and potassium supplementation is occasionally required. Any drug of the thiazide class may be used with equal potential for benefit, and the clinicians should use the one with which they are most familiar from use in other conditions. Care must be exercised when treating patients taking lithium with diuretics, because the induced contraction of plasma volume may increase lithium concentrations and worsen potential toxic effects of the therapy. In the acute setting, diuretics are of no use in NDI and only free water administration can reverse hyperosmolality.

Indomethacin, tolmetin, and ibuprofen have been used in this setting, [238] [242] [243] although the last may be less effective than the others. The combination of thiazides and a nonsteroidal anti-inflammatory agent will not increase urinary osmolality above that of plasma, but the lessening of polyuria is nonetheless beneficial to patients. In many cases, the combination of thiazides with the potassium-sparing diuretic amiloride is preferred in order to lessen the potential side effects associated with long-term use of nonsteroidal anti-inflammatory agents. [239] [240] Amiloride also has the advantage of decreasing lithium entrance into cells in the distal tubule, and because of this, may have a preferable action for the treatment of lithium-induced NDI. [244] [245]

Although DDAVP is generally not effective in NDI, a few patients may have receptor mutations that allow partial responses to AVP or DDAVP,[246] with increases in urine osmolality following much higher doses of these agents than typically used to treat central DI (e.g., 6-10 mg), and it is generally worth a trial of DDAVP at these doses to ascertain whether this is a potential useful therapy in selected patients in whom the responsivity of other affected family members is not already known. Potential therapies involving administration of chaperones to bypass defects in cellular routing of misfolded aquaporin[168] and AVP V2 receptor[169] proteins is an exciting, but uncertain, future possibility.[158]

Primary Polydipsia

At present, no completely satisfactory treatment for primary polydipsia exists. Fluid restriction would seem to be the obvious treatment of choice. However, patients with a reset thirst threshold will be resistant to fluid restriction because of the resulting thirst from stimulation of brain thirst centers at higher plasma osmolalities.[247] In some cases, the use of alternative methods to ameliorate the sensation of thirst (e.g., wetting the mouth with ice chips or using sour candies to increase salivary flow) can help to reduce fluid intake. Fluid intake in patients with psychogenic causes of polydipsia is driven by psychiatric factors that have responded variably to behavioral modification and pharmacologic therapy. Several recent reports have suggested potential efficacy of the antipsychotic drug clozapine as an agent to reduce polydipsia and prevent recurrent hyponatremia in at least a subset of these patients.[248]Administration of any ADH or thiazides to decrease polyuria is hazardous because they invariably produce water intoxication. [200] [249] Therefore, if the diagnosis of DI is uncertain, any trial of antidiuretic therapy should be conducted with close monitoring, preferably in the hospital with frequent evaluation of fluid balance and serum electrolytes. If a patient with primary polydipsia is troubled by nocturia, this may be reduced or eliminated by administering a small dose of DDAVP at bedtime; because thirst and fluid intake are reduced during sleep, this treatment is less likely to cause water intoxication provided the dose is titrated to allow resumption of a water diuresis as soon as the patient awakens the next morning. However, this approach cannot be recommended for patients with psychogenic polydipsia because of the unpredictability of their fluid intake.


The disorders of the renal concentrating mechanism that have been described may be associated with water depletion and hypernatremia. In contrast, disorders in the renal diluting mechanism frequently present as hyponatremia and hypo-osmolality. Hyponatremia is among the most common electrolyte disorders encountered in clinical medicine, with an incidence of 0.97% and a prevalence of 2.48% in hospitalized adult patients when plasma Na+ concentration below 130 mEq/L is the diagnostic criterion,[250] and as high as 15% to 30% if a sodium of less than 135 mEq/L is used.[251] The prevalence may be somewhat lower in the hospitalized pediatric population (between 0.34% and 1.38%),[252] but the incidence is higher than originally recognized in the geriatric population. [251] [253]

As serum osmolality is most often measured to assist in the evaluation of hyponatremic disorders, it is useful to bear in mind the basic relationship of plasma osmolality to Na+ concentration. As reviewed in the introduction to this chapter, Na+ and its associated anions account for nearly all of the osmotic activity of plasma. Therefore, changes in plasma Na+ are usually associated with comparable changes in plasma osmolality. The osmolality calculated from the concentrations of Na+, urea, and glucose is usually in close agreement with that obtained from a measurement of osmolality.[254] When the measured osmolality exceeds the calculated osmolality by more than 10 mOsm/kg H2O, an osmolar gap is said to be present.[254] This occurs in two circumstances: (1) with a decrease in the water content of serum and (2) on addition of a solute other than urea or glucose to the serum. A decrease in the water content of serum is usually due to its displacement by excessive amounts of protein or lipids, as may occur in severe hyperglobulinemia or hyperlipidemia. Normally, 92% to 94% plasma volume is water, the remaining 6% to 8% being lipids and protein. Because of its ionic nature, Na+ dissolves only in the water phase of plasma. Thus, when a greater than normal proportion of plasma is accounted for by solids, the concentration of Na+ in plasma water remains normal, but the concentration in the total volume, as measured by flame photometry, is artifactually low. Such a discrepancy can be avoided if the Na+ concentration is measured with an ion-selective electrode that is now widely available.[255] The sample needs to remain undiluted (direct potentiometry) for accurate measurement of the serum Na+ concentration. Whereas the flame photometer measures the concentration of Na+ in the total volume, the ion-selective electrode measures it in plasma water. Normally, the difference is only 3 mEq/L, but in the setting under discussion, the difference may be much greater. Likewise, because the large lipid and protein molecules contribute only minimally to the total osmolality, the measurement of osmolality by freezing point depression remains normal in these patients. The hyponatremia associated with normal osmolality has been termed factitious or pseudohyponatremia. The most common causes of pseudohyponatremia are primary or secondary hyperlipidemic disorders. The serum need not appear lipemic as increments in cholesterol alone can cause the same discrepancy.[255] Plasma protein elevations above 10 g/dL, as seen in multiple myeloma or macroglobulinemia, can also cause pseudohyponatremia. More recently the administration of intravenous immune globulin has been reported to be associated with hyponatremia without hypo-osmolality in several patients.[256]

The second setting in which an osmolar gap occurs is the presence in plasma of an exogenous low-molecular-weight substance such as ethanol, methanol, ethylene glycol, or mannitol.[257] Undialyzed patients with chronic renal failure, as well as critically ill patients,[258] also have an increment in the osmolar gap of unknown cause. Whereas all of these exogenous substances, as well as glucose and urea, elevate measured osmolality, the effect they have on the plasma concentration of Na+ and intracellular hydration depends on the solute in question. As previously discussed, substances such as glucose, in the presence of relative insulin deficiency, do not penetrate cells readily and remain in the ECF. As a consequence, they draw water from the cellular compartment, causing cell shrinkage, and this translocation of water commensurately decreases the extracellular concentration of Na+. In this setting, therefore, the plasma Na+ concentration may be low while plasma osmolality is high. It is said that for every 100 mg/dL rise in plasma glucose, the osmotic shift of water causes plasma Na+ to drop by 1.6 mEq/L. However, a more recent assessment suggests that this may represent an underestimate of the decrease caused by hyperglycemia, and suggests a 2.4-mEq/L correction factor.[259] Similar “translocational” hyponatremia occurs with mannitol or maltose or with the absorption of glycine during transurethral prostate resection, as well as in gynecologic and orthopedic procedures. A potential toxicity for glycine in this setting also requires consideration.[260] The introduction of new bipolar retroscopes that allow for the use of NaCl as irrigants should translate into the disappearance of this clinical entity. When the plasma solute is readily permeable (e.g., urea, ethylene glycol, methanol, ethanol), it enters the cell and so does not establish an osmotic gradient for water movement. There is no cellular dehydration despite the hypertonic state, and the plasma Na+ concentration remains unchanged. The relationship between plasma osmolality and the plasma Na+ concentration in the presence of various substances is summarized in Table 13-5 .

TABLE 13-5   -- Relationship Between Serum Tonicity and Sodium Concentration in the Presence of Other Substances

Condition or Substance

Serum Tonicity

Serum Sodium


Mannitol, maltose, glycine

Azotemia (high blood urea)

Ingestion of ethanol, methanol, ethylene glycol

Elevated serum lipid/protein




Variables That Influence Water Excretion

In considering clinical disorders that result from excessive or inappropriate secretion of AVP, it is also helpful to remember the other variables that influence water excretion ( Fig. 13-16 ). These factors fall into three categories.



FIGURE 13-16  Urinary dilution mechanisms. Normal determinants of urinary dilution and disorders causing hyponatremia.  (From Cogan M: Normal water homeostasis. In Cogan M [ed]: Fluid and Electrolytes. Norwalk, CT, Appleton & Lange, 1991, pp 98-106.)




Fluid Delivery from the Proximal Tubule.

In spite of the fact that proximal fluid reabsorption is iso-osmotic and therefore does not contribute directly to urine dilution, the volume of tubule fluid that is delivered to the distal nephron determines in large measure the volume of dilute urine that can be excreted. Thus, if glomerular filtration is decreased or proximal tubule reabsorption is greatly enhanced, the resulting diminution in the amount of fluid delivered to the distal tubule itself limits the rate of renal water excretion even if other components of the diluting mechanism are intact.

Dilution of Tubular Fluid.

The excretion of urine that is hypotonic to plasma requires that some segment of the nephron reabsorb solute in excess of water. The water impermeability of the entire ascending limb of Henle, as well as the capacity of its thick segment to reabsorb NaCl, actively endows this segment of the nephron with the characteristics required by the diluting process. Thus, the transport of NaCl by the Na+/K+/2Cl- cotransporter converts the hypertonic tubule fluid that is delivered from the descending limb of the loop of Henle to a distinctly hypotonic fluid (100 mOsm/kg H2O). Interference with the reabsorption of Na+ and Cl- in the ascending limb therefore impairs urine dilution.

Water Impermeability of the Collecting Duct.

The excretion of urine, which is more dilute than the fluid that is delivered to the distal convoluted tubule, requires continued solute reabsorption and minimal water reabsorption in the terminal segments of the nephron. Because the water permeability of the collecting duct epithelium is primarily dependent on the presence or absence of AVP, the hormone plays a pivotal role in determining the fate of the fluid delivered to the collecting duct and, thus, the concentration or dilution of the final urine. In the absence of AVP, the collecting duct remains essentially impermeable to water, even though some water is still reabsorbed. The continued reabsorption of solute then results in the excretion of a maximally dilute urine (∼50 mOsm/kg H2O). As the medullary interstitium is always hypertonic, the absence of circulating AVP, which renders the collecting duct impermeable to water, is critical to the normal diluting process.

This diluting mechanism allows the intake and subsequent excretion of large volumes of water without major alterations in the tonicity of body water.[261] Rarely, this limit can be exceeded, causing water intoxication. Much more commonly, however, hyponatremia occurs at lower rates of water intake, owing either to an intrarenal defect in urine dilution or to the persistent secretion of AVP in the circulation. In the latter case, because hypo-osmolality normally suppresses AVP secretion,[262] the hypo-osmolar state frequently reflects the persistent secretion of AVP in response to hemodynamic or other nonosmolar stimuli.[262]


The serum Na+ concentration is determined by the body's content of sodium, potassium, and TBW. Thus:


This formula has been simplified from the observations made by Edelman in the 1950s. This simplification introduces some errors in the prediction of changes in serum sodium based on the previous formula and has been subject of some reinterpretation by Nguyen and Kurtz.[263] Whereas their revision of the formula is more accurate, as pointed out by Sterns, there are so many inaccuracies in the measurements of sodium, potassium, and water losses as well as intake that there is no substitute for frequent measurements of serum sodium concentration in rapidly changing clinical settings.[264]

As the previous relationship depicts, hyponatremia can therefore occur by an increase in TBW, a decrease in body solutes (either Na+ or K+), or any combination of these. In most cases, more than one of these mechanisms is operant. Therefore, an alternative approach is presented here. In approaching the hyponatremic patient, the physician's first task is to ensure that hyponatremia in fact reflects a hypo-osmotic state and is not a consequence of the causes of pseudohyponatremia or translocational hyponatremia, discussed earlier. Thereafter, an assessment of ECF volume provides a useful working classification of hyponatremia as it can be associated with decreased, normal, or high total body sodium [265] [266]: (1) hyponatremia with ECF volume depletion, (2) hyponatremia with excess ECF volume, and (3) hyponatremia with normal ECF volume.

Hyponatremia with Extracellular Fluid Volume Depletion

Patients with hyponatremia who have ECF volume depletion have sustained a deficit in total body Na+ that exceeds the deficit in water. The decrease in ECF volume is manifested by physical findings such as flat neck veins, decreased skin turgor, dry mucous membranes, orthostatic hypotension, and tachycardia.

If sufficiently severe, volume depletion is a potent stimulus to AVP release. When the osmoreceptor and volume receptor receive opposing stimuli, the former remains fully active but the set-point of the system is lowered. Thus, in the presence of hypovolemia, AVP is secreted and water is retained despite hypo-osmolality. Whereas the hyponatremia in this setting clearly involves a depletion of body solutes, a concomitant failure to excrete water is critical to the process.

As shown in Figure 13-17 , an examination of the urinary Na+ concentration is helpful in assessing whether the fluid losses are renal or extrarenal in origin. A urinary Na+ concentration of less than 20 mEq/L reflects a normal renal response to volume depletion and points to an extrarenal source of fluid loss. This is most commonly seen in patients with gastrointestinal disease with vomiting or diarrhea. Other causes include loss of fluid into the third space, such as the abdominal cavity in pancreatitis or the bowel lumen with ileus. Burns and muscle trauma can also be associated with large fluid and electrolyte losses. Because many of these pathologic states are associated with thirst, an increase in either orally or parenterally taken free water leads to hyponatremia. Hypovolemic hyponatremia in patients whose urinary Na+ concentration is greater than 20 mEq/L points to the kidney as the source of the fluid losses.



FIGURE 13-17  Diagnostic approach to the hyponatremic patient.  (Modified from Halterman R, Berl T: Therapy of dysnatremic disorders. In Brady H, Wilcox C [eds]: Therapy in Nephrology and Hypertension. Philadelphia, WB Saunders, 1999, p 256.)


Diuretic-induced hyponatremia, a commonly observed clinical entity, accounts for a significant proportion of symptomatic hyponatremia in hospitalized patients. It occurs almost exclusively with thiazide rather than loop diuretics, most likely because the former have no effect on urine concentrating ability but the latter do. The hyponatremia is usually evident within 14 days but can occur up to 2 years later in most patients.[267] Underweight women appear to be particularly prone to this complication,[268] and advanced age has been found to be a risk factor in some, [267] [269] but not all,[268] studies. A careful study on diluting ability in the elderly revealed that thiazide diuretics exaggerate the already slower recovery from hyponatremia induced by water ingestion in this population.[270] Diuretics can cause hyponatremia by a variety of mechanisms[271]: (1) volume depletion, which results in impaired water excretion by both enhanced AVP release and decreased fluid delivery to the diluting segment; (2) a direct effect of diuretics on the diluting segment; and (3) K+ depletion causing a decrease in the water permeability of the collecting duct as well as an increase in water intake. K+ depletion leads to hyponatremia independent of the Na+ depletion that frequently accompanies diuretic use.[272] The concomitant administration of K+-sparing diuretics does not prevent the development of hyponatremia. Although the diagnosis of diuretic-induced hyponatremia is frequently obvious, surreptitious diuretic abuse is being increasingly recognized and should be considered in patients in whom other electrolyte abnormalities and high urinary Cl- excretion suggest this possibility.

Salt-losing nephropathy occurs in some patients with advanced renal insufficiency. In the majority of these patients, the Na+-wasting tendency is not one that manifests itself at normal rates of sodium intake; however, some patients with interstitial nephropathy, medullary cystic disease, polycystic kidney disease, or partial urinary obstruction with sufficient Na+ wasting exhibit hypovolemic hyponatremia.[273] Patients with proximal renal tubular acidosis exhibit renal sodium and potassium wasting despite modest renal insufficiency because bicarbonaturia obligates these cation losses.

It has long been recognized that adrenal insufficiency is associated with impaired renal water excretion and hyponatremia. This diagnosis should be considered in the volume-contracted hyponatremic patient whose urinary Na+concentration is not low, particularly when the serum K+, BUN, and creatinine levels are elevated. Separate mechanisms for mineralocorticoid and glucocorticoid deficiency have been defined.[274]

Observations in glucocorticoid-replete adrenalectomized experimental animals provide evidence to support a role of mineralocorticoid deficiency in the abnormal water excretion, as both AVP release and intrarenal factors appear to be causal mechanisms. Thus, conscious adrenalectomized dogs given physiologic doses of glucocorticoids develop hyponatremia. Either saline or physiologic doses of mineralocorticoids corrected the defect in association with both ECF volume repletion and improvement in renal hemodynamics. Immunoassayable AVP levels were elevated in a similarly treated group of mineralocorticoid-deficient dogs despite hypo-osmolality.[275] The decreased ECF volume thus provides the nonosmotic stimulus of AVP release. More direct evidence for the role of AVP was provided in studies employing an AVP antagonist. When glucocorticoid-replete, adrenally insufficient rats were given an AVP antagonist, the minimal urine osmolality was significantly lowered.[276] Urine dilution was not corrected, in contrast to mineralocorticoid-replete rats, supporting a role for an AVP-independent mechanism. This is in concert with studies of adrenalectomized homozygous Brattleboro rats, which also have a defect in water excretion that can be partially corrected by mineralocorticoids or by normalization of volume. In summary, therefore, the mechanism of the defect in water excretion associated with mineralocorticoid deficiency is mediated by AVP and by AVP-independent intrarenal factors, both of which are activated by decrements of ECF volume, rather than by deficiency of the hormone per se.

The presence in the urine of an osmotically active nonreabsorbable or poorly reabsorbable solute causes renal excretion of Na+ and culminates in volume depletion. Glycosuria secondary to uncontrolled diabetes mellitus, mannitol infusion, or urea diuresis after relief of obstruction is a common setting for this disorder. In patients with diabetes, the Na+ wasting caused by the glycosuria can be aggravated by ketonuria because hydroxybutyrate and acetoacetate also cause urinary electrolyte losses. In fact, ketonuria can contribute to the renal Na+ wasting and hyponatremia seen in starvation and alcoholic ketoacidosis. Na+ and water excretion are also increased when a nonreabsorbable anion appears in the urine. This is observed principally with the metabolic alkalosis and bicarbonaturia that accompany severe vomiting or nasogastric suction. In these patients, the excretion of HCO3- requires, for the maintenance of electroneutrality, the excretion of cations, including Na+ and K+. Whereas the renal losses in these clinical settings may be hypotonic, the volume contraction-stimulated thirst and water intake can result in the development of hyponatremia.

Cerebral salt wasting is a rare syndrome described primarily in patients with subarachnoid hemorrhage; it leads to renal salt wasting and volume contraction.[277] Although hyponatremia is increasingly reported in these patients, true cerebral wasting is probably less common than reported.[278] In fact, one critical review found no conclusive evidence for volume contraction or renal salt wasting in any of the patients.[279] The mechanism of this natriuresis is unknown but the increased release of natriuretic peptides has been suggested.[280]

Hyponatremia with Excess Extracellular Fluid Volume

In the advanced stages, the edematous states listed in Figure 13-17 are associated with a decrease in plasma Na+ concentration. Patients have an increase in total body Na+ content, but the rise in TBW exceeds that of Na+. With the exception of renal failure, these states are characterized by avid Na+ retention (urinary Na+ concentration <10 mEq/L). This avid retention may be obscured by the concomitant use of diuretics, which are frequently used in treating these patients. In fact, these agents can further contribute to the abnormal water excretion seen in these states.

Congestive Heart Failure

The common association between congestive heart failure and Na+ and water retention is well established. A mechanism mediated by decreased delivery of tubule fluid to the distal nephron or increased release of AVP has been proposed. In an experimental model of low cardiac output, both AVP and diminished delivery to the diluting segment were found to be important in mediating the abnormality in water excretion. It thus appears that the decrement in “effective” blood volume and the decrease in arterial filling are sensed by aortic and carotid sinus baroreceptors and most likely stimulate AVP release.[281]

This stimulation must supersede the inhibition of AVP release that accompanies at least acute distention of the left atrium. In fact, there is evidence that chronic distention of the atria blunts the sensitivity for the receptor, so high-pressure baroreceptors can act in an uninhibited manner to stimulate AVP release. The importance of AVP in the abnormal dilution in experimental models of heart failure is underscored by the total correction of the water excretory defect by an AVP antagonist in rats with inferior vena cava constriction.[282]

High plasma AVP levels have been demonstrated in patients with congestive heart failure, in both the presence and the absence of diuretics.[283] Likewise, the hypothala-mic mRNA message for the AVP pre-prohormone is elevated in rats with chronic cardiac failure.[284] Although these human studies do not exclude a role for intrarenal factors in the pathogenesis of the abnormal water retention, they complement the experimental observations that demonstrate an important role for AVP in the process. It is most likely that nonosmotic pathways, whose activation is suggested by the increase in sympathetic activity seen in congestive heart failure,[285] are the mediators of AVP release. These hormonal factors, by decreasing the glomerular filtration rate (GFR) and enhancing tubular Na+ reabsorption, decrease distal fluid delivery, further contributing to the hyponatremia. The degree of neurohumoral activation correlates with the clinical severity of left ventricular dysfunction.[286] The degree of hyponatremia is a powerful prognostic factor in these patients.[287] The role of the vasopressin-regulated water channel (AQP2) has been examined in heart failure as well. Two groups have described an up-regulation of this water channel in rats with heart failure. [288] [289] In the latter study,[289] the V2 receptor antagonist OPC31260 reversed the up-regulation, suggesting that a receptor-mediated function, most likely enhanced cAMP generation, is responsible for the process. In fact, a selective V2 antagonist decreases AQP2 excretion[290] and increases urine flow in patients with heart failure.

Hepatic Failure

Patients with advanced cirrhosis and ascites frequently present with hyponatremia as a consequence of their inability to excrete a water load.[291] The classic view suggests that a decrement in effective arterial volume leads to avid Na+ and water retention in an attempt to restore volume toward normal.[292] In this regard, a number of the pathologic derangements in cirrhosis—including splanchnic venous pooling, diminished plasma oncotic pressure secondary to hypoalbuminemia, and the decrease in peripheral resistance—could all contribute to a decrease in effective blood volume.[293] This classic theory was challenged by observations that suggest primary renal Na+ retention—the overflow hypothesis.[294] A proposal that unifies these views has been put forth: Na+ retention occurs early but is a consequence of the severe vasodilation-mediated arterial underfilling.[295]

As with cardiac failure, the relative role of intrarenal and extrarenal factors in impaired water excretion has been a matter of controversy. The observation that expansion of intravascular volume with saline, mannitol, ascites fluid, water immersion, or peritoneovenous shunting improves water excretion in cirrhosis could be interpreted as implicating an intrarenal mechanism in the impaired water excretion, as these maneuvers increase GFR and improve distal delivery. Such maneuvers could also suppress baroreceptor-mediated AVP release and cause an osmotic diuresis, which would also improve water excretion.[292] Experimental models of deranged liver function, including acute portal hypertension by vein constriction, bile duct ligation, and chronic cirrhosis produced by administration of carbon tetrachloride, have demonstrated a predominant role for AVP secretion in the pathogenesis of the disorder. In this latter model, an increment in hypothalamic AVP mRNA has been demonstrated.[296] A study employing an AVP antagonist also points to a central role for AVP in the process.[297] As was the case in heart failure, increased expression of AQP2 has also been reported in the cirrhotic rat,[298] but dysregulation of AQP1 and AQP3 is also present in carbon tetrachloride (CCl4-)–induced cirrhosis.[299] In contrast, in the common bile duct model of cirrhosis, no increase in AQP2 is observed.[300]

Although patients with cirrhosis who have no edema or ascites excrete a water load normally, those with ascites usually do not. Several studies have demonstrated elevated AVP levels in such patients.[291] Patients who had a defect in water excretion had higher levels of AVP, plasma renin activity, plasma aldosterone, and norepinephrine,[301] as well as lower rates of PGE2 production. Likewise, their serum albumin was lower, as was their urinary excretion of Na+, all suggesting a decrease in effective blood volume. As is the case in heart failure, sympathetic tone is high in cirrhosis.[302] In fact, the plasma concentration of norepinephrine, a good index of baroreceptor activity in humans, appears to correlate well with the levels of AVP and the excretion of water. These studies, therefore, offer strong support for the view that effective arterial blood volume is contracted, rather than expanded, in decompensated cirrhosis.[295] This view is further strengthened by observations of subjects during head-out water immersion. This maneuver, which translocates fluid to the central blood volume, caused a decrease in AVP levels and improved water excretion,[303] but in this study, peripheral resistance decreased further. By combining head-out water immersion with norepinephrine administration in an effort to increase systemic pressure and peripheral resistance, water excretion was completely normalized.[304] Such observations underline the critical role of peripheral vasodilation in the process. The observation that inhibition of nitric oxide corrects the arterial hyporesponsiveness to vasodilators[305] and the abnormal water excretion in cirrhotic rats provides evidence for a role of nitric oxide in the vasodilation. [306] [307]

Nephrotic Syndrome

The incidence of hyponatremia in the nephrotic syndrome is lower than in either congestive heart failure or cirrhosis, most likely as a consequence of the higher blood pressure, higher GFR, and more modest impairment in Na+ and water excretion than in the other groups of patients.[308] As lipids are frequently elevated, a direct measurement of plasma osmolality should be done. Diminished excretion of free water was first noted in children with the nephrotic syndrome, and since then, other investigators[309] have noted elevated levels of AVP in these patients. In view of the alterations in Starling forces that accompany hypoalbuminemia and allow transudation of salt and water across capillary membranes to the interstitial space, patients with the nephrotic syndrome have been believed to have intravascular volume contraction. Increased levels of humoral markers of decreased effective blood volume also support this underfilling theory.[310] The possibility that this nonosmotic pathway stimulates AVP release was suggested by studies in which head-out water immersion and blood volume expansion[309] increased water excretion in nephrotic subjects. However, these pathogenic events may not be applicable to all patients with the disorder. Some patients with the nephrotic syndrome in fact have increased plasma volumes with suppressed plasma renin activity and aldosterone levels.[311] The cause of these discrepancies is not immediately evident, but this overfill view has been subject to some criticism.[312] It is most likely that the underfilling mechanism is operant in patients with normal GFR and with the histologic lesion of minimal-change disease and that hypervolemia may be more prevalent in patients with underlying glomerular pathology and decreased renal function. In such patients, an intrarenal mechanism probably causes Na+ retention, as has been described in an experimental model of nephrotic syndrome.[313] Also, in contrast to the increase in AQP2 found in the previously described Na2+- and water-retaining states, in two models of nephrotic syndrome induced with either puromycin aminonucleoside[314] or doxorubicin (Adriamycin),[315] the expression of the water channel was decreased. The animals were not hyponatremic and most likely had expanded ECF volumes to explain the discrepancy.

Renal Failure

Hyponatremia with edema can occur with either acute or chronic renal failure. It is clear that in the setting of either experimental or human renal disease, the ability to excrete free water is maintained better than the ability to reabsorb water. Nonetheless, the patient's GFR rate still determines the maximal rate of free water formation; thus, whenever minimal urine osmolality is reduced to 150 to 250 mOsm/kg H2O and fractional water excretion approaches 20% to 30% of the filtered load, the uremic patient with a GFR of 2 mL/min can excrete only 300 mL/day. Intake of more fluid culminates in hyponatremia. Thus, a decrement in GFR rate with an increase in thirst underlies the hyponatremia of patients with renal insufficiency.[316]

Hyponatremia with Normal Extracellular Fluid Volume

Figure 13-17 lists the clinical entities that have to be considered in patients with hyponatremia whose volume is neither contracted nor expanded and who are, at least by clinical assessment, euvolemic. These entities are considered individually.

Glucocorticoid Deficiency

Considerable evidence exists for an important role for glucocorticoids in the abnormal water excretion of adrenal insufficiency.[317] The water excretory defect of anterior pituitary insufficiency, and particularly corticotropin deficiency, is associated with elevated AVP levels [318] [319] and corrected by physiologic doses of glucocorticoids. Likewise, adrenalectomized dogs receiving replacement of mineralocorticoids have abnormal water excretion. The relative importance of intrarenal factors and AVP in defective water excretion has been a matter of considerable controversy. Studies employing a sensitive radioimmunoassay for plasma AVP and the Brattleboro rat with hypothalamic DI have provided evidence that both factors are involved. Support for a role for AVP has been obtained in studies of conscious adrenalectomized, mineralocorticoid-replaced dogs[320] and rats[321] and with the use of an inhibitor of the hydro-osmotic effect of AVP.[276] As plasma AVP was elevated despite a fall in plasma osmolality, the hormone's release may have been nonosmotically mediated. Although in both of these studies ECF volume was normal, a decrease in systemic pressure and cardiac function [320] [321] could well have provided the hemodynamic stimulus for AVP release. In addition, there may be a direct effect of glucocorticoids that inhibits vasopressin secretion. In this regard, vasopressin gene expression is increased in glucocorticoid-deficient rats.[322] The presence of a glucocorticoid-responsive element on the AVP gene promoter may be responsible for the inhibition of vasopressin gene transcription by glucocorticoids.[323] Also, glucocorticoid receptors are present in magnocellular neurons and they are increased during hypo-osmolality.[324]

A role for AVP-independent intrarenal factors was defined in the antidiuretic-deficient, adrenalectomized Brattleboro rat[321] and with the AVP inhibitor.[276] It appears that prolonged glucocorticoid deficiency (14-17 days) is accompanied by decreases in renal hemodynamics that impair water excretion. A direct effect of glucocorticoid deficiency that enhances water permeability of the collecting duct has been proposed, but such a view is not supported by studies of anuran membranes that suggest that glucocorticoids enhance rather than inhibit water movement. Also, in vitro perfusion studies of the collecting duct of adrenalectomized rabbits show an impaired rather than enhanced AVP response,[325] a defect that may be related to enhanced cAMP metabolism.[326] In fact, AQP2 and AQP3 abundance appears not to be sensitive to glucocorticoid.[327] In summary, the defect in glucocorticoid deficiency is primarily AVP-dependent, and an AVP-independent pathway becomes evident with more prolonged hormone deficiency. It appears likely that alterations in systemic hemodynamics account for the nonosmotic release of AVP, but a direct effect of glucocorticoid hormone in AVP release has not been entirely excluded. The AVP-independent renal mechanism is probably caused by alterations in renal hemodynamics and not by a direct increase in collecting duct permeability. It must be noted that secondary hypoadrenalism, as occurs in hypopituitarism, can also be associated with hyponatremia. [328] [329]


Patients and experimental animals with hypothyroidism often have impaired water excretion and sometimes develop hyponatremia. [317] [330] The dilution defect is reversed by treatment with thyroid hormones. Both decreased delivery of filtrate to the diluting segment and persistent secretion of AVP, alone or combination, have been proposed as mechanisms responsible for the defect.

Hypothyroidism has been shown to be associated with decreases in GFR and renal plasma flow.[330] In the AVP-free Brattleboro rat, the decrement in maximal free water excretion can be entirely accounted for by the decrease in GFR. The osmotic threshold for AVP release appears not to be altered in hypothyroidism.[331] The normal suppression of AVP release with water loading and the normal response to hypertonic saline,[332] coupled with the failure to observe up-regulation of hypothalamic AVP gene expression in hypothyroid rats,[333] supports an AVP-independent mechanism. There is, however, also evidence for a role of AVP in impairing water excretion in hypothyroidism. Thus, in both experimental animals[334] and humans with advanced hypothyroidism,[330] elevated AVP levels were measured in the basal state and after a water load. Although increased sensitivity to AVP in hypothyroidism has been proposed, experimental evidence suggests the contrary, as urine osmolality is relatively low for the circulating levels of the hormone,[334] and AVP-stimulated cyclase is impaired in the renal medulla of hypothyroid rats,[335]possibly leading to decreased AQP2 expression.[336] However, the predominant defect is one of water of excretion with increased AQP2 expression and reversal with a V2 antagonist.[337] It appears, therefore, that diminished distal fluid delivery and persistent AVP release mediate the impaired water excretion in this disorder, but the relative contributions of these two factors remain undefined and may depend on the severity of the endocrine disorder.

Psychosis—Primary Polydipsia

It has long been recognized that patients with psychiatric disease demonstrate generous water intake. Although such polydipsia is normally not associated with hyponatremia, it has been observed that these patients are at increased risk of developing hyponatremia when they are acutely psychotic.[338] Most patients have schizophrenia, but some have psychotic depression. The frequency of hyponatremia in this population of patients is unknown, but in a survey conducted in one large psychiatric hospital, 20 polydipsic patients with a plasma Na+ concentration below 124 mEq/L were reported,[339] and another survey found hyponatremia in 8 of 239 patients.[340] Elucidation of the mechanism of the impaired water excretion has been confounded by antipsychotic drug treatment (see later). The relative contributions of the pharmacologic agent and the psychosis are therefore difficult to define, as thiazides and carbamazepine are frequently implicated.[341] Nonetheless, there are several reports of psychotic patients who suffered water intoxication when free of medication.[342]

The mechanism responsible for the hyponatremia in psychosis appears to be multifactorial.[184] In a comprehensive study of water metabolism in eight psychotic hyponatremic patients and seven psychotic normonatremic control subjects, no unifying defect emerged. The investigators found a small defect in osmoregulation that caused AVP to be secreted at plasma osmolalities somewhat lower than those of the control group, but they did not observe a true resetting of the osmostat. Also, the hyponatremic patients had a mild urine dilution defect even in the absence of AVP. When AVP was present, the renal response was somewhat enhanced, suggesting increased renal sensitivity to the hormone. Psychotic exacerbations appear to be associated with increased vasopressin levels in schizophrenic patients with hyponatremia.[343] Finally, thirst perception is also increased, as excessive water intake that exceeds excretory capacity is responsible for most episodes of hyponatremia in these patients. However, concurrent nausea caused increased vasopressin levels in some of the subjects.[344] Although each of these derangements by itself would remain clinically unimportant, it is possible that, during exacerbation of the psychosis, the defects are more pronounced and that, in combination, they can culminate in hyponatremia.[345]

Hyponatremia also supervenes in beer drinkers (so-called beer potomania). Although this has been ascribed to an increase in fluid intake in the setting of very low solute intake,[346] a recent report suggests that such patients may also have sustained significant solute losses.[347]

Postoperative Hyponatremia

The incidence of hospital-acquired hyponatremia is high, both in adults[205] and in children,[348] and it is particularly prevalent in the postoperative stage [349] [350] (incidence ∼4%). The majority of affected patients appear clinically euvolemic and have measurable levels of AVP in their circulation. [349] [351] Although this occurs primarily as a consequence of administration of hypotonic fluids,[352] a decrease in serum Na+ can occur in this high AVP state, even when isotonic fluids are given.[353] Hyponatremia has also been reported following cardiac catherization in patients receiving hypotonic fluids.[354] Although the presence of hyponatremia is a marker for poor outcome, this is a consequence not of the hyponatremia per se but of the severe underlying diseases associated with it. As discussed in more detail later, there is, however, a subgroup of postoperative patients—almost always premenstrual women—who develop catastrophic neurologic events, frequently accompanied by seizures and hypoxia. [355] [356]

Strenuous Exercise

There is increasing recognition that strenuous exercise, such as military training[357] and marathons and triathlons,[358] can cause hyponatremia that is frequently symptomatic. A review of 57 such patients found a mean serum Na+of 121 mEq/L.[359] A recent prospective study of 488 runners in the Boston Marathon revealed that 13% of the runners had a sodium level less than 130 mEq/L. The multivariate analysis revealed that weight gain related to excessive fluid intake was the strongest single predictor of the hyponatremia. Longer racing times and very low body mass indexes (BMIs) were also predictors.[360] Composition of the consumed fluids and use of nonsteroidal anti-inflammatory agents was not predictive. Symptomatic hyponatremia is even more frequent in ultraendurance events.[361]

Pharmacologic Agents

Table 13-6 lists drugs associated with water retention. Some of the more clinically important ones are discussed here. An increasing number of patients who are receive vasopressin for indications such as von Willebrand disease[362]and nocturnal enuresis [363] [364] are developing severe hyponatremia.

TABLE 13-6   -- Drugs Associated with Hyponatremia

Antidiuretic hormone analogs
Desmopressin acetate

Drugs that enhance arginine vasopressin (AVP)
Carbamazepine, oxycarbazepine
Narcotics (μ-opioid receptors)
Antipsychotics, antidepressants

Drugs that potentiate renal action of AVP
Nonsteroidal anti-inflammatory drugs

Drugs that cause hyponatremia by unknown mechanisms
Selective serotonin reuptake inhibitors
Ecstasy (amphetamine-related)

Data from Berl T, Schrier RW: Disorders of water metabolism. In Schrier RW (ed): Renal and Electrolyte Disorders, 6th ed. Philadelphia, Lippincott Williams & Wilkins, 2003.





The incidence of at least mild hyponatremia in patients taking chlorpropamide may be as high as 7%, but severe hyponatremia (<130 mEq/L) occurs in 2% of patients so treated.[365] As noted earlier, the drug exerts its action primarily by potentiating the renal action of AVP.[366] Studies of toad urinary bladder have demonstrated that, although chlorpropamide alone has no effect, it enhances both AVP- and theophylline-stimulated water flow but decreases cAMP-mediated flow. The enhanced response may be due to up-regulation of the hormone's receptor.[367] Alternatively, studies of chlorpropamide-treated animals suggest that the drug enhances solute reabsorption in the medullary ascending limb (thereby increasing interstitial tonicity and the osmotic drive for water reabsorption) rather than a cAMP-mediated alteration in collecting duct water permeability.[368]

Carbamazepine and Oxycarbazepine.

The anticonvulsant drug carbamazepine is known to possess antidiuretic properties. The incidence of hyponatremia in carbamazepine-treated patients was believed to be as high as 21%, but a survey of patients with mental retardation reported an incidence of 5%.[369] Cases continue to be reported.[370] The antiepileptic oxcarbazepine, of the same class as carbamazepine, has also been reported to cause hyponatremia.[371] Evidence exists for both a mechanism mediated by AVP release[90] and for renal enhancement of the hormone's action[372] to explain carbamazepine's antidiuretic effect. The drug also decreases the sensitivity of the vasopressin response to osmotic stimulation.[373]

Psychotropic Drugs.

An increasing number of psychotropic drugs have been associated with hyponatremia, and in fact, they are frequently implicated to explain the water intoxication in psychotic patients. Among the agents implicated are the phenothiazines,[374] the butyrophenone haloperidol,[375] and the tricyclic antidepressants.[376] Recently, an increasing number of cases of amphetamine (Ecstasy)-related hyponatremia have been described. [377] [378] Likewise, the widely used antidepressants fluoxetine,[379] sertraline,[380] and paroxetine[381] have been associated with hyponatremia. In this latter study involving 75 patients, 12% developed hyponatremia (Na<135 mmol/L). The elderly appear to be particularly susceptible, [382] [383] with an incidence as high as 22% to 28%. [384] [385] The tendency for these drugs to cause hyponatremia is further compounded by their anticholinergic effect. By drying the mucous membranes, they stimulate water intake. The role of the drugs in impaired water excretion has not, in most cases, been dissociated from the role of the underlying disorder for which the drug is given. Furthermore, evaluation of the effect of the drugs on AVP release has frequently revealed a failure to increase the levels of the hormone, particularly if mean arterial pressure remained unaltered. Therefore, although a clinical association between antipsychotic drugs and hyponatremia is frequently encountered, the pharmacologic agents themselves may not be the principal factors responsible for the water retention.[184]

Antineoplastic Drugs.

Several drugs used in cancer therapy cause antidiuresis. The effect of vincristine may be mediated by the drug's neurotoxic effect on the hypothalamic microtubule system, which then alters normal osmoreceptor control of AVP release.[386] A recent retrospective survey suggests that this may be more common in Asians given the drug.[387] The mechanism of the diluting defect that results from cyclophosphamide administration is not fully understood. It may act, at least in part, to enhance action, because the drug does not increase hormone levels.[388] It is known that the antidiuresis has its onset 4 to 12 hours after injection of the drug, lasts as long as 12 hours, and seems to be temporally related to excretion of a metabolite. The importance of anticipating potentially severe hyponatremia in cyclophosphamide-treated patients who are vigorously hydrated to avert urologic complications cannot be overstated. The synthetic analog of cyclophosphamide, ifosfamide, has also been associated with hyponatremia and AVP release.[389]


Since the 1940s, it has been known that the administration of opioid agonists, such as morphine, reduces urine flow by causing the release of an antidiuretic substance. The possibility that endogenous opioids could serve as potential neurotransmitters has been suggested by the finding of enkephalins in nerve fibers projecting from the hypothalamus to the pars nervosa. However, the reported effects vary and range from stimulation to no change and even to inhibition of AVP release. The reasons for these diverse observations may be that the opiates and their receptors are widely distributed in the brain, implying that the site of action of the opiate can differ markedly depending on the route of administration. Likewise, there are multiple opiate peptides and receptor types. It has now been defined that agonists of m-receptors have antidiuretic properties whereas D receptors have the opposite effect.


Several case reports suggest an association between the use of ACE inhibitors and hyponatremia. [390] [391] [392] Of interest is that all three reported patients were women in their 60s. The use of ACE inhibition was also a concomitant risk factor for the development of hyponatremia in a survey of veterans who received chlorpropamide.[365] However, given the widespread use of these agents, the incidence of hyponatremia must be vanishingly low. Likewise, an association with angiotensin receptor blockers has to date not been reported.

Recently four patients have been reported to develop hyponatremia during amiodorone loading.[393]

Syndrome of Inappropriate Antidiuretic Hormone Secretion

Clinical Characteristics

SIADH is the most common cause of hyponatremia in hospitalized patients.[250] As first described by Schwartz and associates[394] in two patients with bronchogenic carcinoma and later further characterized by Bartter and Schwartz,[395] patients with this syndrome have serum hypo-osmolality when excreting urine that is less than maximally dilute (>50 mOsm/kg H2O). Thus, a diagnostic criterion for this syndrome is the presence of inappropriate urinary concentration. The development of hyponatremia with a dilute urine (<100 mOsm/kg H2O) should raise suspicion of a primary polydipsic disorder. Although large volumes of fluid need to be ingested to overwhelm normal water excretory ability, if there are concomitant decreases in solute intake, this volume need not be excessively high.[396] In SIADH, the urinary Na+ is dependent on intake, because Na+ balance is well maintained. As such, urinary Na+ concentration is usually high, but it may be low in patients with the syndrome who are receiving a low-sodium diet. The presence of Na+ in the urine is helpful in excluding extrarenal causes of hypovolemic hyponatremia, but low urinary Na+ concentration does not exclude SIADH. Before the diagnosis of SIADH is made, other causes for a decreased diluting capacity, such as renal, pituitary, adrenal, thyroid, cardiac, or hepatic disease, must be excluded. In addition, nonosmotic stimuli for AVP release, particularly hemodynamic derangements (e.g., due to hypotension, nausea, or drugs), need to be ruled out. Another clue to the presence of the syndrome is the finding of hypouricemia. In one study, 16 of 17 patients had levels below 4 mg/dL, whereas in 13 patients with hyponatremia of other causes the level was greater than 5 mg/dL. This hypouricemia appears to occur as a consequence of increased urate clearance.[397] The measurement of an elevated level of AVP confirms the clinical diagnosis. It must be noted, however, that the majority of patients with SIADH have AVP levels in the “normal” range (≥10 pg/mL); the presence of any AVP is, however, abnormal in the hypo-osmolar state. As the presence of hyponatre-mia is itself evidence for abnormal dilution, a formal urine-diluting test need not be performed. The water test is helpful in determining whether an abnormality remains in a patient whose serum Na+ has been corrected by water restriction. Because Brattleboro rats receiving vasopressin,[398] as well as an animal model of SIADH, display up-regulation of AQP2 expression, the excretion of AQP2 has been investigated as a marker for the persistent secretion of ADH. The excretion of the water channel remains elevated in patients with SIADH; however, this is not specific to this entity, as a similar pattern was observed in patients with hyponatremia due to hypopituitarism.[399]


In 1953, Leaf and associates[400] described the effects of chronic AVP administration on Na+ and water balance. They noted that high-volume water intake was required for the development of hyponatremia. Concomitant with the water retention, an increment in urinary Na+ excretion was noted. The relative contributions of the water retention and Na+ loss to the development of hyponatremia were subsequently investigated. Acute water loading causes transient natriuresis but, when water intake is increased more slowly, no significant negative Na+ loss can be documented. Such studies clearly demonstrate that the hyponatremia is in large measure a consequence of water retention; however, it must be noted that the net increase in water balance fails to account entirely for the decrement in serum Na+.[400] In a carefully studied model of SIADH secretion in rats, the retained water was found to be distributed in the intracellular space and to be in equilibrium with the tonicity of ECF.[401] The natriuresis and kaliuresis that occur early in the development of this model contribute to a decrement of body solutes and in part account for the observed hyponatremia.[402] Studies involving analysis of whole-body water and electrolyte content demonstrate that the relative contributions of water retention and solute losses vary with the duration of induced hyponatremia; the former is central to the process, but with more prolonged hyponatremia, Na+ depletion becomes predominant.[403] In this regard, it has even been suggested that the natriuresis and volume contraction are an important component of the syndrome that maintains the secretion of AVP[404] with atrial natriuretic peptide as a mediator of the Na+ loss.[405] Therefore, although natriuresis frequently accompanies the syndrome, the secretion of ADH is essential. Finally, patients with the syndrome must also have a defect in thirst regulation whereby the osmotic inhibition of water intake is not operant. The mechanism of this failure to suppress thirst is not fully understood.

After the initial retention of water, loss of Na+, and development of hyponatremia, continued administration of AVP is accompanied by the reestablishment of Na+ balance and a decline in the hydro-osmotic effect of the hormone. The integrity of renal regulation of Na+ balance is manifested by the ability to conserve Na+ during Na+ restriction and by the normal excretion of an Na+ load. Thus, the mechanisms that regulate Na+ excretion are intact. Loss of the hydro-osmotic effect of AVP, albeit to varying degrees, is evident in many studies, [400] [402] because urine flow increased and urine osmolality decreased despite continued administration of the hormone ( Fig. 13-18 ). This effect has been termed vasopressin escape.[406] Several studies have demonstrated that hypotonic expansion rather than chronic administration of AVP per se is needed for escape to occur, as the escape phenomenon is seen only when positive water balance is achieved.[406]



FIGURE 13-18  Effects of desmopressin (DDAVP) and water administration to two normal rats (circles and triangles). Note that urine osmolality decreases and serum Na+ stabilizes. Sham-treated control subjects are depicted by squares.  (From Verbalis JG, Drutarosky M: Adaptation to chronic hypo-osmolality in rats. Kidney Int 34:351, 1988.)




The cellular mechanisms responsible for vasopressin escape have been the subject of some investigation. Studies of broken epithelial cell preparations of the toad urinary bladder revealed down-regulation of AVP receptors[407] as well as vasopressin binding in the inner medulla.[408] Post-cAMP mechanisms are probably also operant. In this regard, a decrease in expression of AQP2 has been reported in the process of escape from DDAVP-induced antidiuresis, without a concomitant change in basolateral AQP3 and AQP4. [409] [410] The decrement in AQP2 was associated with decreased V2 responsiveness.[409] The distal tubule also has an increase in sodium transporters, including the α- and γ-subunits of the epithelial sodium channel and the thiazide sensitive Na+/Cl- cotransporter.[411] In addition to a renal mechanism, it appears that chronic hyponatremia causes a decrement in hypothalamic mRNA production, a process that could ameliorate the syndrome in the clinical setting.[27]

Clinical Settings

It is now apparent that the previously described pathophysiologic sequence occurs in a variety of clinical settings characterized by persistent AVP secretion. Since the original report of Schwartz and co-workers,[394] the syndrome has been described in an increasing number of clinical settings ( Table 13-7 ). These fall into three general categories[412]: (1) malignancies, (2) pulmonary disease, and (3) CNS disorders. In addition, an increasing number of patients with acquired immunodeficiency syndrome have been reported to have hyponatremia. The frequency may be as high as 35% of hospitalized patients with the disease, and in as many as two thirds, SIADH may be the underlying cause.[413] As was noted previously, hyponatremia caused by excessive water repletion can occur after moderate and severe exercise. [358] [359] [414] [415] Finally, it is increasingly recognized that an idiopathic form is common in the elderly. [416] [417] [418] [419] As many as 25% of elderly patients admitted to a rehabilitation center had serum Na+ values less than 135 mEq/L.[417] In a significant proportion of these patients, no underlying cause is unveiled. This may be related to an increase in AQP2 production and excretion in this age group.[420]

TABLE 13-7   -- Disorders Associated with the Syndrome of Inappropriate Antidiuretic Hormone Secretion


Pulmonary Disorders

Central Nervous System Disorders


Bronchogenic carcinoma

Viral pneumonia

Encephalitis (viral or bacterial)


Carcinoma of the duodenum


Meningitis (viral, bacterial, tuberculous, fungal)

Prolonged exercise

Carcinoma of the pancreas


Carcinoma of the ureter

Idiopathic (in elderly)


Pulmonary abscess

Head trauma


Carcinoma of the stomach


Brain abscess




Guillain-Barré syndrome


Ewing sarcoma

Positive pressure breathing

Acute intermittent porphyria


Carcinoma of the bladder


Subarachnoid hemorrhage or subdural hematoma


Prostatic carcinoma


Cerebellar and cerebral atrophy


Oropharyngeal tumor






Cystic fibrosis



Cavernous sinus thrombosis



Neonatal hypoxia



Shy-Drager syndrome



Rocky Mountain spotted fever



Delirium tremens



Cerebrovascular accident (cerebral thrombosis or hemorrhage)



Acute psychosis



Peripheral neuropathy



Multiple sclerosis


Data from Berl T, Schrier RW: Disorders of water metabolism. In Schrier RW (ed): Renal and Electrolyte Disorders, 6th ed. Philadelphia, Lippincott Williams & Wilkins, 2003.




A material with antidiuretic properties has been extracted from some of the tumors or metastases of patients with malignancy-associated SIADH. Not all patients with the syndrome have AVP in their tumors. A number of the tumors have also been found to produce the carrier hormone of AVP, neurophysin, suggesting that repression of normal genetic information has occurred. Of the tumors that cause SIADH secretion, bronchogenic carcinoma, and particularly small cell lung cancer, is the most common, with a reported incidence of 11%.[421] It appears that patients with bronchogenic carcinoma have higher plasma AVP levels in relation to plasma osmolality, even if they do not manifest the full-blown SIADH, although in patients with the syndrome, the levels of the hormone are higher. The possibility that the hormone could serve as a marker of bronchogenic carcinoma has been suggested, and in fact, SIADH has been reported occasionally to precede the diagnosis of the tumor by several months.[422] In view of the potential to treat patients with this tumor, it is important that patients with unexplained SIADH be fully investigated and evaluated for the presence of this malignancy. Head and neck malignancies are the second most common tumors associated with the syndrome, as it occurs in approximately 3% of such patients.

The mechanism whereby AVP is produced in other pulmonary disorders is not known, but the associated abnormalities in blood gases could act as mediators of the effect. Antidiuretic activity has also been assayed in tuberculous lung tissue. The syndrome can also occur in the setting of miliary rather than only lung-limited tuberculosis.[423] In CNS disorders, AVP is most likely released from the neurohypophysis. Studies of monkeys have shown that elevations of intracranial pressure cause AVP secretion, and this may be the mechanism that mediates the syndrome in at least some CNS disorders. The magnocellular vasopressin secreting cells in the hypothalamus are subject to numerous excita-tory inputs, and therefore, it is conceivable that a large variety of neurologic disorders can cause the secretion of the hormone.

Finally, hyponatremia was described recently in two infants with undetectable AVP levels who were found to have a gain of function mutation at the X-linked vasopressin receptor wherein in codon 137 a missense mutation resulted in the change from arginine to cysteine or leucine. The authors termed this nephrogenic syndrome of inappropriate antidiuresis (NSIAD).[424]

Robertson and colleagues have studied osmoregulation of AVP secretion in a large group of patients with SIADH.[425] In the great majority, the plasma AVP concentration was inadequately suppressed relative to the hypotonicity present. In most patients, the plasma AVP concentration ranged between 1 and 10 pg/mL, the same range as in normally hydrated healthy adults. Inappropriate secretion, therefore, can often be demonstrated only by measuring AVP under hypotonic conditions. Even with this approach, however, abnormalities in plasma AVP were not apparent in almost 10% of the patients with clinical evidence of SIADH. To better define the nature of the osmoregulatory defect in these patients, plasma AVP was measured during infusion of hypertonic saline. When this method of analysis was applied to 25 patients with SIADH, four different types of osmoregulatory defects were identified.

As shown in Figure 13-19 , in the type A osmoregulatory defect, infusion of hypertonic saline was associated with large and erratic fluctuations in plasma AVP, which bore no relationship to the rise in plasma osmolality. This pattern was found in 6 of 25 patients studied, who had acute respiratory failure, bronchogenic carcinoma, pulmonary tuberculosis, schizophrenia, or rheumatoid arthritis. This pattern indicates that the secretion of AVP either had been totally divorced from osmoreceptor control or was responding to some periodic nonosmotic stimulus.



FIGURE 13-19  Plasma vasopressin as a function of plasma osmolality during the infusion of hypertonic saline in four groups of patients with clinical syndrome of inappropriate antidiuretic hormone (SIADH). The shaded area indicates the range of normal values. See text for description of each group.  (From Zerbe R, Stropes L, Robertson G: Vasopressin function in the syndrome of inappropriate antidiuresis. Annu Rev Med 31:315, 1980.)




A completely different type of osmoregulatory defect is exemplified by the type B response, as depicted in Figure 13-19 . The infusion of hypertonic saline resulted in prompt and progressive rises in plasma osmolality. Regression analysis showed that the precision and sensitivity of this response were essentially the same as those in healthy subjects, except that the intercept or threshold value at 253 mOsm/kg was well below the normal range. This pattern, which reflects the resetting of the osmoreceptor, was found in 9 of the 25 patients who had a diagnosis of bronchogenic carcinoma, cerebrovascular disease, tuberculous meningitis, acute respiratory disease, or carcinoma of the pharynx. Another patient has been reported with hyponatremia and acute idiopathic polyneuritis who reacted in an identical manner to the hypertonic saline infusion and was determined to have resetting of the osmoreceptor. Because their threshold function is retained when they receive a water load, this patient and others with reset osmostats have been able to dilute their urine maximally and sustain a urine flow sufficient to prevent a further increase in body water. Thus, an abnormality in AVP regulation can exist in spite of the ability to maximally dilute the urine and excrete a water load.

In the type C response (see Fig. 13-19 ), plasma AVP was elevated initially but did not change during the infusion of hypertonic saline until plasma osmolality reached the normal range. At that point, plasma AVP began to rise appropriately, indicating a normally functioning osmoreceptor mechanism. This response was found in 8 of the 25 patients with the diagnosis of CNS disease, bronchogenic carcinoma, carcinoma of the pharynx, pulmonary tuberculosis, or schizophrenia. Its pathogenesis is unknown, but the authors speculate that it may be due to a constant, nonsuppressible leak of AVP despite otherwise normal osmoregulatory function. Unlike type B, the resetting type of defect, the type C response results in impaired urine dilution and water excretion at all levels of plasma osmolality.

In the type D response (see Fig. 13-19 ), the osmoregulation of AVP appears to be completely normal despite a marked inability to excrete a water load. The plasma AVP is appropriately suppressed under hypotonic conditions and does not rise until plasma osmolality reaches the normal threshold level. When this procedure is reversed by water loading, plasma osmolality and plasma AVP again fall normally, but urine dilution does not occur, and the water load is not excreted. This defect was present in 2 of 25 patients with the diagnosis of bronchogenic carcinoma, indicating that, in these patients, the antidiuretic defect is caused by some abnormality other than SIADH. It could be due either to increased renal tubule sensitivity to AVP or to the existence of an antidiuretic substance other than AVP. Alternatively, it is possible that the presently available assays are not sufficiently sensitive to detect significant levels of AVP. Perhaps some of these subjects have the nephrogenic syndrome of antidiuresis described previously.[424]

It is of interest that patients with bronchogenic carcinoma, which has generally been believed to be associated with ectopic production of AVP, manifested every category of osmoregulatory defect, including the reset osmostat. It has been suggested that many of these tumors probably cause SIADH secretion not by producing the hormone ectopically but rather by interfering with the normal osmoregulation of AVP secretion from the neurohypophysis through direct invasion of the vagus nerve, metastatic implants in the hypothalamus, or some other more generalized neuropathic changes.

Symptoms, Morbidity, and Mortality

The majority of patients with hyponatremia appear to be asymptomatic. However, a recent case-control study of 122 elderly patients with a mean serum sodium of 126 mmol/L suggests that such patients are not truly asymptomatic. Thus, when compared with 244 age-matched controls, they had a 67-fold greater risk for sustaining falls, and neuropsychiatric testing revealed subtle abnormalities.[426] Other clinical manifestations of hyponatremia usually occur only at a serum Na+ concentration below 125 mmol/L. Although gastrointestinal complaints occur early, the majority of the manifestations are neuropsychiatric, including lethargy, psychosis, and seizures, designated as hyponatremic encephalopathy. [427] [428] In its severe form, hyponatremic encephalopathy can cause brainstem compression leading to pulmonary edema and hypoxemia, [94] [429] which may be, at least in part, mediated by AQP4.[430] In fact, in a retrospective study of 168 hyponatremic patients, most of them acute, a strong association existed between the development of hypoxemia and the risk of mortality (13-fold).[431] Finally, a number of hyponatremic patients have been reported to also develop rhabdomyolysis.[414]

The development of symptoms also depends on the age, gender, and magnitude and acuteness of the process. Elderly persons and young children with hyponatremia are most likely to develop symptoms. It has also become apparent that neurologic complications occur more frequently in menstruating women. In a case-control study, Ayus and colleagues[356] noted that, despite an approximately equal incidence of postoperative hyponatremia in males and females, 97% of those with permanent brain damage were women and 75% of them were menstruant. However, this view is not universally held, as others have not found increased postoperative hyponatremia in this population,[432]and the aforementioned retrospective study did not reveal a gender or age association with mortality.[431]

The degree of clinical impairment is related not to the absolute measured level of lowered serum Na+ concentration but to both the rate and the extent of the drop in ECF osmolality. In a survey of hospitalized hyponatremic patients (serum Na+ level <128 mEq/L), 46% had CNS symptoms and 54% were asymptomatic.[433] It is of note, however, that the authors believed that the hyponatremia was the cause of the symptoms in only 31% of the symptomatic patients. In this subgroup of symptomatic patients, the mortality was no different from that of asymptomatic patients (9%–10%). In contrast, the mortality of patients whose CNS symptoms were not caused by hyponatremia was high (64%), suggesting that the mortality of these patients is more often due to the associated disease than to the electrolyte disorder itself. This is in agreement with the report of Anderson,[250] who noted a 60-fold increase in mortality in hyponatremic patients over that of normonatremic control subjects. In the hyponatremic patients, death frequently occurred after the plasma Na+ concentration was returned toward normal and was due to progression of severe underlying disease; this suggests that the hyponatremia is an indicator of severe disease and poor prognosis. In fact, a number of recent studies further point out that even mild hyponatremia is an independent predictor of higher mortality in a number of disorders. These include patients with acute ST elevation myocardial infarctions,[434] heart failure,[435] and liver disease. [436] [437] In fact, these authors have concluded that the addition of the hyponatremia to the model end stage liver disease (MELD) score (employed as a system to assign priority for liver transplantation) predicts outcome more precisely than the now-used MELD score.

The mortality of acute symptomatic hyponatremia has been noted to be as high as 55% and as low as 5%. [438] [439] The former reflects the observation of few symptomatic hyponatremic patients in a consultative setting, the latter the estimate from a broad-based literature survey. Equally controversial is the mortality rate associated with hyponatremia in children. One series found no in-hospital deaths attributable to hyponatremia, but others described an 8.4% mortality in such postoperative children and estimated that more than 600 children die as a result of hyponatremia in the United States yearly.[252] Hospital-acquired hyponatremia may have contributed to the morbidity and mortality associated with La Crosse encephalitis in children.[440] The mortality associated with chronic hyponatremia has been reported to be between 14% and 27%. [441] [442]

The observed CNS symptoms are most likely related to the cellular swelling and cerebral edema that result from acute lowering of ECF osmolality, which leads to movement of water into cells. In fact, such cerebral edema occasionally causes herniation, as has been noted in postmortem examination of both humans and experimental animals. The increase in brain water is, however, much less marked than would be predicted from the decrease in tonicity were the brain to operate as a passive osmometer. The volume regulatory responses that protect against cerebral edema, and which probably occur throughout the body, have been extensively studied and reviewed.[443]Studies of rats demonstrate a prompt loss of both electrolyte and organic osmolytes after the onset of hyponatremia. [443] [444] Some of the osmolyte losses occur within 24 hours,[445] but the loss of water becomes more marked in subsequent days ( Fig. 13-20 ). The rate at which the brain restores the lost electrolytes and osmolytes when hyponatremia is corrected is of great pathophysiologic importance. Na+ and Cl- recover quickly and even overshoot. [444] [445] [446] However, the reaccumulation of osmolytes is considerably delayed (see Fig. 13-20 ). This process is likely to account for the more remarked cerebral dehydration that accompanies the correction in previously adapted animals.[447] It has been observed that urea may prevent the myelinosis associated with this pathology. This may well be due to the more rapid reaccumulation of organic osmolytes, and particularly myoinositol, in the azotemic state.[448]



FIGURE 13-20  Comparison of changes in brain electrolyte (A) and organic osmolyte (B) contents during adaptation to hyponatremia and after rapid correction of hyponatremia in rats. Both electrolytes and organic osmolytes are lost quickly after the induction of hyponatremia beginning on day 0. Brain content of both solutes remains depressed during maintenance of hyponatremia from days 2 through 14. After rapid correction of the hyponatremia on day 14, electrolytes reaccumulate rapidly and overshoot normal brain contents on the first 2 days after correction, before returning to normal levels by the 5th day after correction. In contrast, brain organic osmolytes recover much more slowly and do not return to normal brain contents until the 5th day after correction. The dashed lines indicate±SEM from the mean values of normonatremic rats on day 0. *P<.01 compared with brain contents of normonatremic rats. DBW, dry brain weight.  (Data from Verbalis JG, Gullans SR: Hyponatremia causes large sustained reductions in brain content of multiple organic osmolytes in rats. Brain Res 567:274, 1991; and Verbalis JG, Gullans SR: Rapid correction of hyponatremia produces differential effects on brain osmolyte and electrolyte reaccumulation in rats. Brain Res 606:19, 1993.)





Treatment of hyponatremia is a subject of considerable interest and has been discussed in a companion to this chapter[449] and in other reviews.[450] The therapeutic strategy is dictated by the underlying cause of the disorder, as well as (1) the presence or absence of symptoms, (2) the duration of the disorder, and (3) the risk for neurologic complications. Although the ultimate goal, whenever possible, is to identify and treat the underlying pathologic condition, this is not always entirely possible (especially in SIADH), so a general approach to the management of categories of hyponatremia has been developed.[449]

Symptomatic Hyponatremia.

The therapeutic approach to symptomatic hyponatremia has been a subject of great controversy [438] [439] [447] [448] [449] [450] [451] [452] [453] [454] emanating from the observation that neurologic disorders supervene both in untreated acutely hyponatremic patients and occasionally in the course of treatment of the hyponatremia. The neurologic symptoms that occur in acutely hyponatremic patients [355] [356] and in elderly persons taking thiazide diuretics[455]usually include apathy, confusion, nausea, vomiting, and frequently seizures. The mortality rate in this group is high and the majority of survivors have significant neurologic residua. The possibility that this permanent neurologic damage is a consequence of postanoxic encephalopathy has been suggested. In fact, there is evidence from experimental animals to support the view that, when hypoxia is combined with hyponatremia, the adaptive mechanisms are abrogated, leading to an increase in brain edema and mortality,[456] perhaps explaining the previously mentioned association of hypoxemia with high mortaltity.[431] The cause of the female preponderance among those who suffer permanent neurologic damage is unknown, but it has been suggested that the adaptive mechanisms whereby the brain decreases its volume in acute hyponatremia are less efficient and could be inhibitied by sex hormones. In this regard, it is of interest that estrogens appear to alter the function of the Na+,K+-ATPase in rat brain synaptosomes,[457] a process that could delay cell volume regulation in response to hypotonicity. Nonetheless, a combination of factors, including AVP via a V1 receptor[458] atrial natriuretic peptide,[459] hypoxia, and sex hormones, all contribute to alterations in cellular adaptation and vascular reactivity culminating in severe, often fatal, cerebral edema.

In view of the devastating neurologic consequences that can be associated with acute symptomatic hyponatremia, it has been suggested that such patients' metabolic disorder should be corrected rapidly [431] [460] ( Fig. 13-21 ). In fact, the observation of Ayus and associates[461] demonstrated the safety of this approach. Nonetheless, the increasing number of patients who have been reported to develop a neurologic syndrome suggestive of central pontine myelinolysis in the course of treatment of hyponatremia cannot be ignored.[451] This syndrome has been recently reviewed.[462] The clinical picture is not that of the classic syndrome described in malnourished alcoholics, and the demyelinating lesions are frequently extrapontine in both humans,[451] including children,[463] and experimental animals.[448] It appears that the demyelinating syndrome is most likely to occur in patients whose hyponatremia is more chronic and is corrected once the adaptive process has set in.[447] A disruption of the blood-brain barrier and alterations in cortical and subcortical blood flow may be involved in the pathogenesis of this disorder.



FIGURE 13-21  Treatment of severe euvolemic hyponatremia.  (From Thurman J, Halterman R, Berl T: Therapy of dysnatremic disorders. In Brady H, Wilcox C [eds]: Therapy in Nephrology and Hypertension, 2nd ed. Philadelphia, Saunders, 2003.)


Although small lesions produce minimal symptoms, patients with more extensive disease have flaccid quadriplegia, dysphagia, and dysarthria. Patients with extrapontine demyelination can present with an atypical picture that includes disorders of movements, mutism, and catatonia. The disorder was considered uniformly fatal, but it is now recognized that most patients survive, including some with complete recovery.[464] The diagnosis is now best made with MRI, with diffusion-weighted MRI showing lesions within 24 hours of symptoms.[465] The factors that predispose to these neurologic complications have not been fully delineated. In the view of some, the rate of correction is critical,[438] a view supported by data from experimental animals.[461] Other data suggest that the process is independent of the rate of correction and is due to the absolute change in serum Na+ over a given time period.[451] This concept is also supported by experimental data.[466] These two variables are not entirely independent, and therefore, attention to both correction rate and magnitude is indicated. In this regard, in experimental animals,[467] correction at an excessive rate (>2 mEq/L/hr) and magnitude (>20 mEq/L/24 hr) was associated with the development of cerebral lesions ( Fig. 13-22 ). Others would propose more conservative correction rates of approximately 0.4 mEq/L/hr and 12 mEq/L in any 24-hour period.[468]



FIGURE 13-22  Incidence of demyelinative lesions in individual rats as a function of both the maximal rate of correction of hyponatremia and the magnitude of the increase in plasma Na+ concentration over the first 24 hours of the correction. Each rat in one of the three correction groups (water restriction, water diuresis, and hypertonic saline) is plotted as a function of the maximal rate of increase in plasma Na+ achieved in any 4-hour period during the correction (abscissa) and the total magnitude of the increase in plasma Na+ achieved during the first 24 hours of the correction (ordinate). Rats are represented by their neuropathology score (0, no demyelinative lesions; 1, focal demyelinative lesions; 2, diffuse demyelinative lesions).  (Data from Verbalis JG, Martinez AJ: Neurological and neuropathological sequelae of correction of chronic hyponatremia. Kidney Int 39:1274, 1991.)




In summary, rapid correction is indicated for patients with acute (<48 hr) and symptomatic hyponatremia (see Fig. 13-21 ). This should probably aim to raise serum Na+ concentrations by approximately 1 to 2 mEq/L/hr until seizures subside. This correction can be achieved by administration of hypertonic saline with the concomitant administration of furosemide, which impairs free water reabsorption and lowers urine osmolality, induces excretion of Na+ in a much larger volume of urine, and leads to a much greater negative water balance. This allows more rapid correction of the plasma Na+ concentration. Although full correction in this setting is safe, it is not necessary. Several formulas have been suggested to predict the increased serum Na+ that accompanies administration of intravenous fluids.[266] These formulas operate well in static conditions but fail to account for ongoing water and solute losses. A comprehensive formula that incorporates these variables has been proposed.[469] Although probably accurate, its complexity makes it difficult to employ in practice. A tonicity, solutes, and water balance approach that monitors the Na+, K+, and water infused and excreted best predicts changes in serum sodium.[470]

Example: A 70-kg man has a serum Na+ concentration of 110 mEq/L. Assuming that 60% of BW is water,



In this patient,


Over the next 2 hours, the patient receives 200 mL of 3% NaCl and excretes 1000 mL of urine. The urinary Na+ is 70 mEq/L and the urinary K+ is 39 mEq/L. In this case, there is no net cation gain or loss, as 100 mEq of Na+ was given (the Na+ content of 200 mL of 3% NaCl) and 100 mEq of cation was excreted. However, TBW has decreased by 800 mL (1000 mL excreted -200 mL given with the hypertonic saline). Thus, the 4620 mEq of body cations is now in 41.2 L, which would increase serum Na+ to 112 mEq/L (4620/41.2). The rate at which serum Na+ will rise therefore depends not only on the volume of hypertonic saline administered but also on the volume and cation content of urine excreted. Because Na+ and K+ balance is maintained, cation excretion is dependent on the amount infused, but the urine volume is in large measure determined by the kidney's ability to generate free water. Concomitant administration of loop diuretics would increase free water excretion and therefore the rate of correction, if needed.

The excess water that must be excreted to achieve an increment in Na+ can likewise be calculated. Assuming that a patient weighs 70 kg when the serum Na+ is 115 mEq/L, the excess water that needs to be excreted to correct the hyponatremia to 130 mEq/L can be estimated as follows:


or in this case,



During therapy, serum and urine electrolytes should be monitored frequently to avoid overcorrection. Once symptoms have decreased and a small degree of correction has been achieved, further therapy can proceed more slowly with either decreased rates of saline or fluid restriction alone.

Because of the increased risk of osmotic demyelination syndrome, the treatment of symptomatic patients with chronic hyponatremia calls for careful monitoring and restraint. In general, treatment with hypertonic saline to raise serum Na+ concentrations by approximately 10 mEq/L (i.e., 10%), at a rate not to exceed 1 mEq/L/hr is probably safe and usually sufficient for amelioration of symptoms. This is rarely necessary, and should be employed only when the patient has seizures. Even more so than in the acutely symptomatic group, care should be taken to avoid an absolute increase in Na+ concentration greater than 12 mEq/L/day. After this degree of correction, therapy should continue with water restriction. Such an approach should be accompanied by a vanishingly low incidence of myelinolysis. The issue of spontaneous correction at an undesirably rapid rate by the onset of a water diuresis is deserving of consideration because brain damage can ensue in this setting. There are some clinical settings in which correction of hyponatremia can occur at a very rapid rate with institution of therapy. If the previous parameters have been exceeded and correction has been proceeded more rapidly than discussed (usually because of excretion of hypotonic urine), the events leading to demyelination may be reversed by readministration of hypotonic fluids and DDAVP. This is suggested from animal studies[471] and case reports in humans,[472] even when overtly symptomatic.[473]

Asymptomatic Hyponatremia.

The cornerstone of the treatment of “asymptomatic” hyponatremia is water restriction (see Fig. 13-21 ). The extent of this restriction depends on the particular patient's degree of diluting impairment. Thus, for some patients, restriction to 1000 mL/day is adequate and achieves negative water balance and an increase in serum Na+ concentration. If, in another patient, such a volume exceeds the total renal and extrarenal water losses, no improvement will be noted and even greater restriction of water intake is necessary. A guide to the degree of water restriction required to improve the hyponatremia can be assessed from the concentration of Na+ and K+ in the urine.[474]

In practical terms, severe water restriction is difficult to enforce for prolonged periods, especially in the out-patient setting. Pharmacologic agents that antagonize AVP action and maneuvers that increase solute excretion have allowed patients with SIADH to drink more water. The two most commonly employed agents are lithium and dimethylchlortetracycline.

Despite the well-established effect of lithium to antagonize AVP action, its use in SIADH has been superseded by the less toxic and perhaps more effective use of demeclocycline.[475] This agent, in doses between 600 and 1200 mg/day, is effective in inhibiting AVP action and restores serum Na+ to normal levels within 5 to 14 days, permitting unrestricted water intake. The mechanism whereby demeclocycline exerts this effect is not known. Toxicity in patients with cirrhosis has been well recognized. A nonpharmacologic alternative to the treatment of these patients involves an increase in solute intake and excretion. Because the level of urine concentration is more or less fixed in many patients with SIADH, urine flow is extremely dependent on solute excretion and increasing the solute load therefore increases fluid loss. Thus, administration of urea (30-60 g dissolved in 100 mL of water) once a day was successful in patients with the syndrome.[476] With the exception of one patient who had headaches but who responded to an alteration in dosing schedule, no reactions were noted, although gastrointestinal disturbances would not be unexpected.

The use of furosemide (40 mg/day) and high salt intake (200 mEq/day), an extension of the treatment of acute symptomatic hyponatremia to the chronic management of euvolemic hyponatremia, has also been reported to be successful.[477] In general, all these treatments are in one manner or another unsatisfactory, and therefore, the emergence of vasopressin antagonists to treat these disorders represents an important improvement in the armamentarium.

Vasopressin Antagonists.

The development of antagonists to the vasopressor and antidiuretic properties of AVP has potential therapeutic implications for the management of patients with excess AVP. Initial attempts to develop such agents were met with problems related to agonist effects and species specificity. More recently nonpeptide AVP selective V2 and combined V/V2 have been developed. These agents antagonize both endogenous and exogenous AVP, causing water diuresis in the absence of alterations in filtration rate or solute excretion.[478] The antagonists block the action of AVP to stimulate adenylate cyclase in cortical, medullary, and papillary collecting ducts; to interfere with binding of radioactive AVP in papillary membranes; and to block AVP-mediated water flow in isolated rabbit collecting ducts.[479] The antagonist appears to bind to the transmembrane region of the receptor blocking the binding of AVP to the receptor.[480] The development of a number of oral antagonists of the V2 receptors [478] [481] is extremely promising, particularly as they have been found effective in a rat model of SIADH,[482] as well as in patients with SIADH.[483] [484] [485] In one study, the V1 and V2 antagonist conivaptan has been administered for 3 months.[486] The agent was well tolerated and maintained a normal serum Na+. An intravenous preparation of this drug has now been approved by the U.S. food and Drug Administration (FDA) for use in hospitalized patients with euvolemic and hypervolemic hyponatremia. Vasopressin antagonists appear to also cause an aquaresis in other disorders characterized by high AVP levels such as cirrhosis [487] [488] as well as in heart failure. [489] [490] A potential role for these compounds in the treatment of heart failure has been considered.[491] A large scale trial (EVEREST) demonstrated short-term improvement in dyspnea,[492] but no long-term survival benefit.[493] The potential therapeutic role of the antagosits in the treatment of water-retaining disorders will await further experience and the delineations of potential toxicities. Nonetheless, they clearly usher in a new era in the treatment of these disorders.


1. Edelman IS, Leibman J: Anatomy of body water and electrolytes.  Am J Med  1959; 27:256.

2. Fanestil DD: Compartmentation of body water.   In: Narins RG, ed. Clinical Disorders of Fluid and Electrolyte Metabolism,  New York: McGraw-Hill; 1994:3-20.

3. Thomas RC: Electrogenic sodium pump in nerve and muscle cells.  Physiol Rev  1972; 52:563.

4. Czech MP: The molecular basis of insulin action.  Annu Rev Biochem  1977; 46:359.

5. Wolf AV, McDowell ME: Apparent and osmotic volumes of distribution of sodium, chloride, sulfate and urea.  Am J Physiol  1954; 176:207.

6. Maffly RH, Leaf A: The potential of water in mammalian tissues.  J Gen Physiol  1959; 42:1257.

7. Leaf A, Chatillon JY, Tuttle EPJ: The mechanism of the osmotic adjustment of body cells as determined in vivo by the volume of distribution of a large water load.  J Clin Invest  1954; 33:1261.

8. Rose BD: New approach to disturbances in the plasma sodium concentration.  Am J Med  1986; 81:1033-1040.

9. Foster KG: Relation between the colligative properties and chemical composition of sweat.  J Physiol (Lond)  1961; 155:490.

10. Hendry EB: Osmolarity of human serum and of chemical solutions of biologic importance.  Clin Chem  1961; 7:156.

11. Zerbe RL, Robertson GL: Osmoregulation of thirst and vasopressin secretion in human subjects: Effect of various solutes.  Am J Physiol  1983; 244:E607-E614.

12. Vokes TP, Aycinena PR, Robertson GL: Effect of insulin on osmoregulation of vasopressin.  Am J Physiol  1987; 252:E538-E548.

13. de Castro J: A microregulatory analysis of spontaneous fluid intake in humans: Evidence that the amount of liquid ingested and its timing is mainly governed by feeding.  Physiol Behav  1988; 3:705-714.

14. Adolph EF: Physiology of Man in the Desert,  New York, Hafner Publishing, 1969.

15. Du Vigneaud V: Hormones of the posterior pituitary gland: Oxytocin and vasopressin.   In: DuVigneaud V, Bing RJ, Oncley JL, ed. The Harvey Lectures, 1954-55,  New York: Academic Press; 1956:1-28.

16. Edwards BR, LaRochelle Jr FT: Antidiuretic effect of endogenous oxytocin in dehydrated Brattleboro homozygous rats.  Am J Physiol  1984; 247:F453-F465.

17. Oliver G, Schaefer EA: On the physiological actions of extracts of the pituitary body and certain other glandular organs.  J Physiol (Lond)  1895; 18:277-279.

18. Cowley Jr AW: Vasopressin and blood pressure regulation.  Clin Physiol Biochem  1988; 6:150-162.

19. Mohr E, Bahnsen U, Kiessling C, et al: Expression of the vasopressin and oxytocin genes in rats occurs in mutually exclusive sets of hypothalamic neurons.  FEBS Lett  1988; 242:144-148.

20. Mohr E, Richter D: Sequence analysis of the promoter region of the rat vasopressin gene.  FEBS Lett  1990; 260:305-308.

21. Gainer H, Yamashita M, Fields RL, et al: The magnocellular neuronal phenotype: Cell-specific gene expression in the hypothalamo-neurohypophysial system.  Prog Brain Res  2002; 139:1-14.

22. Richter D: Molecular events in expression of vasopressin and oxytocin and their cognate receptors.  Am J Physiol  1988; 255:F207-F219.

23. Robinson AG, Haluszczak C, Wilkins JA, et al: Physiologic control of two neurophysins in humans.  J Clin Endocrinol Metab  1977; 44:330-339.

24. Nowycky MC, Seward EP, Chernevskaya NI: Excitation-secretion coupling in mammalian neurohypophysial nerve terminals.  Cell Mol Neurobiol  1998; 18:65-80.

25. Herman JP, Schafer MK, Watson SJ, et al: In situ hybridization analysis of arginine vasopressin gene transcription using intron-specific probes.  Mol Endocrinol  1991; 5:1447-1456.

26. Majzoub JA, Rich A, van Boom J, et al: Vasopressin and oxytocin mRNA regulation in the rat assessed by hybridization with synthetic oligonucleotides.  J Biol Chem  1983; 258:14061-14064.

27. Robinson AG, Roberts MM, Evron WA, et al: Hyponatremia in rats induces downregulation of vasopressin synthesis.  J Clin Invest  1990; 86:1023-1029.

28. Fitzsimmons MD, Roberts MM, Robinson AG: Control of posterior pituitary vasopressin content: Implications for the regulation of the vasopressin gene.  Endocrinology  1994; 134:1874-1878.

29. Leng G, Mason WT, Dyer RG: The supraoptic nucleus as an osmoreceptor.  Neuroendocrinology  1982; 34:75-82.

30. Buggy J, Jonhson AK: Preoptic-hypothalamic periventricular lesions: Thirst deficits and hypernatremia.  Am J Physiol  1977; 233:R44-R52.

31. Thrasher TN, Keil LC, Ramsay DJ: Lesions of the organum vasculosum of the lamina terminalis (OVLT) attenuate osmotically-induced drinking and vasopressin secretion in the dog.  Endocrinology  1982; 110:1837-1839.

32. Robertson GL: The regulation of vasopressin function in health and disease.  Rec Prog Horm Res  1976; 33:333-385.

33. Zerbe RL, Miller JZ, Robertson GL: The reproducibility and heritability of individual differences in osmoregulatory function in normal human subjects.  J Lab Clin Med  1991; 117:51-59.

34. Helderman JH, Vestal RE, Rowe JW, et al: The response of arginine vasopressin to intravenous ethanol and hypertonic saline in man: The impact of aging.  J Gerontol  1978; 33:39-47.

35. Ledingham JGG, Crowe MJ, Forsling ML: Effects of aging on vasopressin secretion, water excretion, and thirst in man.  Kidney Int Suppl  1987; 32:S90.

36. Baylis PH: Osmoregulation and control of vasopressin secretion in healthy humans.  Am J Physiol  1987; 253:R671-R678.

37. Vokes TJ, Weiss NM, Schreiber J, et al: Osmoregulation of thirst and vasopressin during normal menstrual cycle.  Am J Physiol  1988; 254:R641-R647.

38. Vallotton MB, Merkelbach U, Gaillard RC: Studies of the factors modulating antidiuretic hormone excretion in man in response to the osmolar stimulus: Effects of oestrogen and angiotensin II.  Acta Endocrinol (Copenh)  1983; 104:295-302.

39. Davison JM, Gilmore EA, Durr J, et al: Altered osmotic thresholds for vasopressin secretion and thirst in human pregnancy.  Am J Physiol  1984; 246:F105-F109.

40. Weisinger RS, Burns P, Eddie LW, et al: Relaxin alters the plasma osmolality-arginine vasopressin relationship in the rat.  J Endocrinol  1993; 137:505-510.

41. Leng G, Brown CH, Bull PM, et al: Responses of magnocellular neurons to osmotic stimulation involves coactivation of excitatory and inhibitory input: An experimental and theoretical analysis.  J Neurosci  2001; 21:6967-6977.

42. Verbalis JG: Osmotic inhibition of neurohypophysial secretion.  Ann N Y Acad Sci  1993; 689:146-160.

43. Thrasher TN: Osmoreceptor mediation of thirst and vasopressin secretion in the dog.  Fed Proc  1982; 41:2528-2532.

44. Bourque CW, Voisin DL, Chakfe Y: Stretch-inactivated cation channels: Cellular targets for modulation of osmosensitivity in supraoptic neurons.  Prog Brain Res  2002; 139:85-94.

45. Sladek CD, Kapoor JR: Neurotransmitter/neuropeptide interactions in the regulation of neurohypophyseal hormone release.  Exp Neurol  2001; 171:200-209.

46. Ludwig M, Sabatier N, Dayanithi G, et al: The active role of dendrites in the regulation of magnocellular neurosecretory cell behavior.  Prog Brain Res  2002; 139:247-256.

47. Schrier RW, Berl T, Anderson RJ: Osmotic and nonosmotic control of vasopressin release.  Am J Physiol  1979; 236:F321-F332.

48. Raff H, Merrill D, Skelton M, et al: Control of ACTH and vasopressin in neurohypophysectomized conscious dogs.  Am J Physiol  1985; 249:R281-R284.

49. Dunn FL, Brennan TJ, Nelson AE, et al: The role of blood osmolality and volume in regulating vasopressin secretion in the rat.  J Clin Invest  1973; 52:3212-3219.

50. Stricker EM, Verbalis JG: Interaction of osmotic and volume stimuli in regulation of neurohypophyseal secretion in rats.  Am J Physiol  1986; 250:R267-R275.

51. Goldsmith SR, Francis GS, Cowley AW, et al: Response of vasopressin and norepinephrine to lower body negative pressure in humans.  Am J Physiol  1982; 243:H970-H973.

52. Goldsmith SR, Cowley Jr AW, Francis GS, et al: Effect of increased intracardiac and arterial pressure on plasma vasopressin in humans.  Am J Physiol  1984; 246:H647-H651.

53. Robertson GL, Athar S: The interaction of blood osmolality and blood volume in regulating plasma vasopressin in man.  J Clin Endocrinol Metab  1976; 42:613-620.

54. Quillen Jr EW, Cowley Jr AW: Influence of volume changes on osmolality-vasopressin relationships in conscious dogs.  Am J Physiol  1983; 244:H73-H79.

55. Andresen MC, Doyle MW, Jin YH, et al: Cellular mechanisms of baroreceptor integration at the nucleus tractus solitarius.  Ann N Y Acad Sci  2001; 940:132-141.

56. Renaud LP: CNS pathways mediating cardiovascular regulation of vasopressin.  Clin Exp Pharmacol Physiol  1996; 23:157-160.

57. Blessing WW, Sved AF, Reis DJ: Destruction of noradrenergic neurons in rabbit brainstem elevates plasma vasopressin, causing hypertension.  Science  1982; 217:661-663.

58. Berl T, Cadnapaphornchai P, Harbottle JA, et al: Mechanism of suppression of vasopressin during alpha-adrenergic stimulation with norepinephrine.  J Clin Invest  1974; 53:219-227.

59. Schrier RW: Pathogenesis of sodium and water retention in high-output and low-output cardiac failure, nephrotic syndrome, cirrhosis, and pregnancy (1).  N Engl J Med  1988; 319:1065-1072.

60. Schrier RW: Pathogenesis of sodium and water retention in high-output and low-output cardiac failure, nephrotic syndrome, cirrhosis, and pregnancy (2).  N Engl J Med  1988; 319:1127-1134.

61. Zerbe RL, Henry DP, Robertson GL: Vasopressin response to orthostatic hypotension. Etiologic and clinical implications.  Am J Med  1983; 74:265-271.

62. Seckl JR, Williams TD, Lightman SL: Oral hypertonic saline causes transient fall of vasopressin in humans.  Am J Physiol  1986; 251:R214-R217.

63. Thompson CJ, Burd JM, Baylis PH: Acute suppression of plasma vasopressin and thirst after drinking in hypernatremic humans.  Am J Physiol  1987; 252:R1138-R1142.

64. Salata RA, Verbalis JG, Robinson AG: Cold water stimulation of oropharyngeal receptors in man inhibits release of vasopressin.  J Clin Endocrinol Metab  1987; 65:561-567.

65. Rowe JW, Shelton RL, Helderman JH, et al: Influence of the emetic reflex on vasopressin release in man.  Kidney Int  1979; 16:729-735.

66. Verbalis JG, McHale CM, Gardiner TW, et al: Oxytocin and vasopressin secretion in response to stimuli producing learned taste aversions in rats.  Behav Neurosci  1986; 100:466-475.

67. Baylis PH, Robertson GL: Rat vasopressin response to insulin-induced hypoglycemia.  Endocrinology  1980; 107:1975-1979.

68. Baylis PH, Zerbe RL, Robertson GL: Arginine vasopressin response to insulin-induced hypoglycemia in man.  J Clin Endocrinol Metab  1981; 53:935-940.

69. Thompson DA, Campbell RG, Lilavivat U, et al: Increased thirst and plasma arginine vasopressin levels during 2-deoxy-d-glucose-induced glucoprivation in humans.  J Clin Invest  1981; 67:1083-1093.

70. Keil LC, Summy-Long J, Severs WB: Release of vasopressin by angiotensin II.  Endocrinology  1975; 96:1063-1065.

71. Ferguson AV, Renaud LP: Systemic angiotensin acts at subfornical organ to facilitate activity of neurohypophysial neurons.  Am J Physiol  1986; 251:R712-R717.

72. McKinley MJ, McAllen RM, Pennington GL, et al: Physiological actions of angiotensin II mediated by AT1 and AT2 receptors in the brain.  Clin Exp Pharmacol Physiol Suppl  1996; 3:S99-S104.

73. McKinley MJ, Allen AM, Mathai ML, et al: Brain angiotensin and body fluid homeostasis.  Jpn J Physiol  2001; 51:281-289.

74. Yamaguchi K, Sakaguchi T, Kamoi K: Central role of angiotensin in the hyperosmolality- and hypovolaemia-induced vasopressin release in conscious rats.  Acta Endocrinol (Copenh)  1982; 101:524-530.

75. Stocker SD, Stricker EM, Sved AF: Acute hypertension inhibits thirst stimulated by ANG II, hyperosmolality, or hypovolemia in rats.  Am J Physiol Regul Integr Comp Physiol  2001; 280:R214-R224.

76. Keil LC, Severs WB: Reduction in plasma vasopressin levels of dehydrated rats following acute stress.  Endocrinology  1977; 100:30-38.

77. Edelson JT, Robertson GL: The effect of the cold pressor test on vasopressin secretion in man.  Psychoneuroendocrinology  1986; 11:307-316.

78. Ukai M, Moran Jr WH, Zimmermann B: The role of visceral afferent pathways on vasopressin secretion and urinary exeretory patterns during surgical stress.  Ann Surg  1968; 168:16-28.

79. Chikanza IC, Petrou P, Chrousos G: Perturbations of arginine vasopressin secretion during inflammatory stress. Pathophysiologic implications.  Ann N Y Acad Sci  2000; 917:825-834.

80. Baylis PH, Stockley RA, Heath DA: Effect of acute hypoxaemia on plasma arginine vasopressin in conscious man.  Clin Sci Mol Med  1977; 53:401-404.

81. Claybaugh JR, Hansen JE, Wozniak DB: Response of antidiuretic hormone to acute exposure to mild and severe hypoxia in man.  J Endocrinol  1978; 77:157-160.

82. Rose Jr CE, Anderson RJ, Carey RM: Antidiuresis and vasopressin release with hypoxemia and hypercapnia in conscious dogs.  Am J Physiol  1984; 247:R127-R134.

83. Raff H, Shinsako J, Keil LC, et al: Vasopressin, ACTH, and corticosteroids during hypercapnia and graded hypoxia in dogs.  Am J Physiol  1983; 244:E453-E458.

84. Farber MO, Weinberger MH, Robertson GL, et al: Hormonal abnormalities affecting sodium and water balance in acute respiratory failure due to chronic obstructive lung disease.  Chest  1984; 85:49-54.

85. Anderson RJ, Pluss RG, Berns AS, et al: Mechanism of effect of hypoxia on renal water excretion.  J Clin Invest  1978; 62:769-777.

86. Miller M: Role of endogenous opioids in neurohypophysial function of man.  J Clin Endocrinol Metab  1980; 50:1016-1020.

87. Oiso Y, Iwasaki Y, Kondo K, et al: Effect of the opioid kappa-receptor agonist U50488H on the secretion of arginine vasopressin. Study on the mechanism of U50488H-induced diuresis.  Neuroendocrinology  1988; 48:658-662.

88. Eisenhofer G, Johnson RH: Effect of ethanol ingestion on plasma vasopressin and water balance in humans.  Am J Physiol  1982; 242:R522-R527.

89. Oiso Y, Robertson GL: Effect of ethanol on vasopressin secretion and the role of endogenous opioids.   In: Schrier R, ed. Water Balance and Antidiuretic Hormone,  New York: Raven Press; 1985:265-269.

90. Stephens WP, Coe JY, Baylis PH: Plasma arginine vasopressin concentrations and antidiuretic action of carbamazepine.  BMJ  1978; 1:1445-1447.

91. Reid IA, Ahn JN, Trinh T, et al: Mechanism of suppression of vasopressin and adrenocorticotropic hormone secretion by clonidine in anesthetized dogs.  J Pharmacol Exp Ther  1984; 229:1-8.

92. Iovino M, De Caro G, Massi M, et al: Muscimol inhibits ADH release induced by hypertonic sodium chloride in rats.  Pharmacol Biochem Behav  1983; 19:335-338.

93. Zerbe RL, Bayorh MA, Quirion R, et al: The role of vasopressin suppression in phencyclidine-induced diuresis.  Pharmacology  1983; 26:73-78.

94. Lausen HD: Metabolism of the neurohypophyseal hormones.   In: Greep RO, Astwood EB, Knobil E, et al ed. Handbook of Physiology,  Washington, DC: American Physiological Society; 1974:287-393.

95. Davison JM, Sheills EA, Barron WM, et al: Changes in the metabolic clearance of vasopressin and in plasma vasopressinase throughout human pregnancy.  J Clin Invest  1989; 83:1313-1318.

96. Andersson B: Thirst—and brain control of water balance.  Am Sci  1971; 59:408-415.

97. Fitzsimons JT: Thirst.  Physiol Rev  1972; 52:468-561.

98. Quillen EW, Reid IA, Keil LC: Carotid and arterial baroreceptor influences on plasma vasopressin and drinking.   In: Cowley AWJ, Liard JF, Ausiello DA, ed. Vasopressin: Cellular and Integrative Functions,  New York: Raven Press; 1988:405-411.

99. Stricker EM, Verbalis JG: Water intake and body fluids.   In: Zigmond MJ, Bloom FE, Landis SC, et al ed. Fundamental Neuroscience,  San Diego: Academic Press; 1999:1111-1126.

100. Phillips PA, Rolls BJ, Ledingham JG, et al: Osmotic thirst and vasopressin release in humans: A double-blind crossover study.  Am J Physiol  1985; 248:R645-R650.

101. Szczepanska-Sadowska E, Kozlowski S: Equipotency of hypertonic solutions of mannitol and sodium chloride in eliciting thirst in the dog.  Pflugers Arch  1975; 358:259-264.

102. Robertson GL: Disorders of thirst in man.   In: Ramsay DJ, Booth DA, ed. Thirst: Physiological and Psychological Aspects,  London: Springer-Verlag; 1991:453-477.

103. Verbalis JG: Inhibitory controls of drinking.   In: Ramsay DJ, Booth DA, ed. Thirst: Physiological and Psychological Aspects,  London: Springer-Verlag; 1991:313-334.

104. Fitzsimons JT: Drinking by rats depleted of body fluid without increases in osmotic pressure.  J Physiol (Lond)  1961; 159:297-309.

105. Thrasher TN, Keil LC, Ramsay DJ: Hemodynamic, hormonal, and drinking responses to reduced venous return in the dog.  Am J Physiol  1982; 243:R354-R362.

106. Phillips PA, Rolls BJ, Ledingham JG, et al: Angiotensin II-induced thirst and vasopressin release in man.  Clin Sci (Lond)  1985; 68:669-674.

107. Rogers PW, Kurtzman NA: Renal failure, uncontrollable thirst, and hyperreninemia. Cessation of thirst with bilateral nephrectomy.  JAMA  1973; 225:1236-1238.

108. Stricker EM, Sved AF: Thirst.  Nutrition  2000; 16:821-826.

109. Root AW, Martinez CR, Muroff LR: Subhypothalamic high-intensity signals identified by magnetic resonance imaging in children with idiopathic anterior hypopituitarism. Evidence suggestive of an “ectopic” posterior pituitary gland.  Am J Dis Child  1989; 143:366-367.

110. Heinbecker P, White HL: Hypothalamico-hypophyseal system and its relation to water balance in the dog.  Am J Physiol  1941; 133:582-593.

111. Maccubbin DA, Van Buren JM: A quantitative evaluation of hypothalamic degeneration and its relation to diabetes insipidus following interruption of the human hypophyseal stalk.  Brain  1963; 86:443.

112. Lippsett MB, MacLean IP, West CD, et al: An analysis of the polyuria induced by hypophysectomy in man.  J Clin Endocrinol Metab  1956; 16:183-195.

113. Valtin H, North WG, Edwards BR, Gellai M: Animal models of diabetes insipidus.   In: Czernichow P, Robinson AG, ed. Diabetes insipidus in man,  Basel: Karger; 1985:105-126.

114. Schmale H, Richter D: Single base deletion in the vasopressin gene is the cause of diabetes insipidus in Brattleboro rats.  Nature  1984; 308:705-709.

115. Hansen LK, Rittig S, Robertson GL: Genetic basis of familial neurohypophyseal diabetes insipidus.  Trends Endocrinol Metab  1997; 8:363.

116. Repaske DR, Phillips III JA, Kirby LT, et al: Molecular analysis of autosomal dominant neurohypophyseal diabetes insipidus.  J Clin Endocrinol Metab  1990; 70:752-757.

117. Rittig S, Robertson GL, Siggaard C, et al: Identification of 13 new mutations in the vasopressin-neurophysin ii gene in 17 kindreds with familial autosomal dominant neurohypophyseal diabetes insipidus.  Am J Hum Genet  1996; 58:107-117.

118. Repaske DR, Medlej R, Gultekin EK, et al: Heterogeneity in clinical manifestation of autosomal dominant neurohypophyseal diabetes insipidus caused by a mutation encoding ala-1-val in the signal peptide of the arginine vasopressin/neurophysin ii/copeptin precursor.  J Clin Endocrinol Metab  1997; 82:51-56.

119. Bergeron C, Kovacs K, Ezrin C, et al: Hereditary diabetes insipidus: An immunohistochemical study of the hypothalamus and pituitary gland.  Acta Neuropathol (Berl)  1991; 81:345-348.

120. Maghnie M, Villa A, Arico M, et al: Correlation between magnetic resonance imaging of posterior pituitary and neurohypophyseal function in children with diabetes insipidus.  J Clin Endocrinol Metab  1992; 74:795-800.

121. Kaplowitz PB, D'Ercole AJ, Robertson GL: Radioimmunoassay of vasopressin in familial central diabetes insipidus.  J Pediatr  1982; 100:76-81.

122. Siggaard C, Rittig S, Corydon TJ, et al: Clinical and molecular evidence of abnormal processing and trafficking of the vasopressin preprohormone in a large kindred with familial neurohypophyseal diabetes insipidus due to a signal peptide mutation.  J Clin Endocrinol Metab  1999; 84:2933-2941.

123. Si-Hoe SL, de Bree FM, Nijenhuis M, et al: Endoplasmic reticulum derangement in hypothalamic neurons of rats expressing a familial neurohypophyseal diabetes insipidus mutant vasopressin transgene.  FASEB J  2000; 14:1680-1684.

124. Davies J, Murphy D: Autophagy in hypothalamic neurones of rats expressing a familial neurohypophysial diabetes insipidus transgene.  J Neuroendocrinol  2002; 14:629-637.

125. Maghnie M, Cosi G, Genovese E, et al: Central diabetes insipidus in children and young adults.  N Engl J Med  2000; 343:998-1007.

126. Imura H, Nakao K, Shimatsu A, et al: Lymphocytic infundibuloneurohypophysitis as a cause of central diabetes insipidus.  N Engl J Med  1993; 329:683-689.

127. Kojima H, Nojima T, Nagashima K, et al: Diabetes insipidus caused by lymphocytic infundibuloneurohypophysitis.  Arch Pathol Lab Med  1989; 113:1399-1401.

128. Van Havenbergh T, Robberecht W, Wilms G, et al: Lymphocytic infundibulohypophysitis presenting in the postpartum period: Case report.  Surg Neurol  1996; 46:280-284.

129. Nishioka H, Ito H, Sano T, et al: Two cases of lymphocytic hypophysitis presenting with diabetes insipidus: A variant of lymphocytic infundibulo-neurohypophysitis.  Surg Neurol  1996; 46:285-290.

130. Thodou E, Asa SL, Kontogeorgos G, et al: Clinical case seminar: lymphocytic hypophysitis: Clinicopathological findings.  J Clin Endocrinol Metab  1995; 80:2302-2311.

131. Scherbaum WA, Bottazzo GF, Czernichow P: Role of autoimmunity in central diabetes insipidus.   In: Czernichow P, Robinson AG, ed. Diabetes insipidus in man,  Basel: Karger; 1985:232-239.

132. De Bellis A, Colao A, Di Salle F, et al: A longitudinal study of vasopressin cell antibodies, posterior pituitary function, and magnetic resonance imaging evaluations in subclinical autoimmune central diabetes insipidus.  J Clin Endocrinol Metab  1999; 84:3047-3051.

133. Zerbe RL, Robertson GL: A comparison of plasma vasopressin measurements with a standard indirect test in the differential diagnosis of polyuria.  N Engl J Med  1981; 305:1539-1546.

134. Hollinshead WH: The interphase of diabetes insipidus.  Mayo Clin Proc  1964; 39:92-100.

135. Verbalis JG, Robinson AG, Moses AM: Postoperative and post-traumatic diabetes insipidus.   In: Czernichow P, Robinson AG, ed. Diabetes Insipidus in Man,  Basel: Karger; 1984:247-265.

136. Cusick JF, Hagen TC, Findling JW: Inappropriate secretion of antidiuretic hormone after transsphenoidal surgery for pituitary tumors.  N Engl J Med  1984; 311:36-38.

137. Olson BR, Rubino D, Gumowski J, et al: Isolated hyponatremia after transsphenoidal pituitary surgery.  J Clin Endocrinol Metab  1995; 80:85-91.

138. Olson BR, Gumowski J, Rubino D, et al: Pathophysiology of hyponatremia after transsphenoidal pituitary surgery.  J Neurosurg  1997; 87:499-507.

139. Ultmann MC, Hoffman GE, Nelson PB, et al: Transient hyponatremia after damage to the neurohypophyseal tracts.  Neuroendocrinology  1992; 56:803-811.

140. Daniel PM, Prichard MM: Regeneration of hypothalamic nerve fibres after hypophysectomy in the goat.  Acta Endocrinol (Copenh)  1970; 64:696-704.

141. Daniel PM, Prichard MM: The human hypothalamus and pituitary stalk after hypophysectomy or pituitary stalk section.  Brain  1972; 95:813-824.

142. Block LH, Furrer J, Locher RA, et al: Changes in tissue sensitivity to vasopressin in hereditary hypothalamic diabetes insipidus.  Klin Wochenschr  1981; 59:831-836.

143. DeRubertis FR, Michelis MF, Davis BB: “Essential” hypernatremia. Report of three cases and review of the literature.  Arch Intern Med  1974; 134:889-895.

144. Halter JB, Goldberg AP, Robertson GL, et al: Selective osmoreceptor dysfunction in the syndrome of chronic hypernatremia.  J Clin Endocrinol Metab  1977; 44:609-616.

145. Baylis PH, Thompson CJ: Osmoregulation of vasopressin secretion and thirst in health and disease.  Clin Endocrinol (Oxf)  1988; 29:549-576.

146. Takaku A, Shindo K, Tanaka S, et al: Fluid and electrolyte disturbances in patients with intracranial aneurysms.  Surg Neurol  1979; 11:349-356.

147. McIver B, Connacher A, Whittle I, et al: Adipsic hypothalamic diabetes insipidus after clipping of anterior communicating artery aneurysm.  BMJ  1991; 303:1465-1467.

148. DeRubertis FR, Michelis MF, Beck N, et al: “Essential” hypernatremia due to ineffective osmotic and intact volume regulation of vasopressin secretion.  J Clin Invest  1971; 50:97-111.

149. Smith D, McKenna K, Moore K, et al: Baroregulation of vasopressin release in adipsic diabetes insipidus.  J Clin Endocrinol Metab  2002; 87:4564-4568.

150. Phillips PA, Bretherton M, Johnston CI, et al: Reduced osmotic thirst in healthy elderly men.  Am J Physiol  1991; 261:R166-R171.

151. Hodak SP, Verbalis JG: Abnormalities of water homeostasis in aging.  Endocrinol Metab Clin North Am  2005; 34:1031-1046.

152. Durr JA, Hoggard JG, Hunt JM, et al: Diabetes insipidus in pregnancy associated with abnormally high circulating vasopressinase activity.  N Engl J Med  1987; 316:1070-1074.

153. Durr JA: Diabetes insipidus in pregnancy.  Am J Kidney Dis  1987; 9:276-283.

154. Gordge MP, Williams DJ, Huggett NJ, et al: Loss of biological activity of arginine vasopressin during its degradation by vasopressinase from pregnancy serum.  Clin Endocrinol (Oxf)  1995; 42:51-58.

155. Baylis PH, Thompson C, Burd J, et al: Recurrent pregnancy-induced polyuria and thirst due to hypothalamic diabetes insipidus: An investigation into possible mechanisms responsible for polyuria.  Clin Endocrinol (Oxf)  1986; 24:459-466.

156. Barron WM, Cohen LH, Ulland LA, et al: Transient vasopressin-resistant diabetes insipidus of pregnancy.  N Engl J Med  1984; 310:442-444.

157. van Lieburg AF, Verdijk MA, Schoute F, et al: Clinical phenotype of nephrogenic diabetes insipidus in females heterozygous for a vasopressin type 2 receptor mutation.  Hum Genet  1995; 96:70-78.

158. Morello JP, Bichet DG: Nephrogenic diabetes insipidus.  Annu Rev Physiol  2001; 63:607-630.

159. Sadeghi H, Robertson GL, Bichet DG, et al: Biochemical basis of partial nephrogenic diabetes insipidus phenotypes.  Mol Endocrinol  1997; 11:1806-1813.

160. Wildin RS, Cogdell DE, Valadez V: AVPR2 variants and V2 vasopressin receptor function in nephrogenic diabetes insipidus.  Kidney Int  1998; 54:1909-1922.

161. Pasel K, Schulz A, Timmermann K, et al: Functional characterization of the molecular defects causing nephrogenic diabetes insipidus in eight families.  J Clin Endocrinol Metab  2000; 85:1703-1710.

162. Wildin RS, Cogdell DE: Clinical utility of direct mutation testing for congenital nephrogenic diabetes insipidus in families.  Pediatrics  1999; 103:632-639.

163. Chan Seem CP, Dossetor JF, Penney MD: Nephrogenic diabetes insipidus due to a new mutation of the arginine vasopressin V2 receptor gene in a girl presenting with non-accidental injury.  Ann Clin Biochem  1999; 36(pt 6):779-782.

164. Deen PM, Knoers NV: Vasopressin type-2 receptor and aquaporin-2 water channel mutants in nephrogenic diabetes insipidus.  Am J Med Sci  1998; 316:300-309.

165. Knoers NV, Deen PM: Molecular and cellular defects in nephrogenic diabetes insipidus.  Pediatr Nephrol  2001; 16:1146-1152.

166. Canfield MC, Tamarappoo BK, Moses AM, et al: Identification and characterization of aquaporin-2 water channel mutations causing nephrogenic diabetes insipidus with partial vasopressin response.  Hum Mol Genet  1997; 6:1865-1871.

167. van Os CH, Deen PM: Aquaporin-2 water channel mutations causing nephrogenic diabetes insipidus.  Proc Assoc Am Physicians  1998; 110:395-400.

168. Tamarappoo BK, Verkman AS: Defective aquaporin-2 trafficking in nephrogenic diabetes insipidus and correction by chemical chaperones.  J Clin Invest  1998; 101:2257-2267.

169. Morello JP, Salahpour A, Laperriere A, et al: Pharmacological chaperones rescue cell-surface expression and function of misfolded V2 vasopressin receptor mutants.  J Clin Invest  2000; 105:887-895.

170. Knepper MA, Verbalis JG, Nielsen S: Role of aquaporins in water balance disorders.  Curr Opin Nephrol Hypertens  1997; 6:367-371.

171. Nielsen S, Kwon TH, Christensen BM, et al: Physiology and pathophysiology of renal aquaporins.  J Am Soc Nephrol  1999; 10:647-663.

172. Marples D, Frokiaer J, Dorup J, et al: Hypokalemia-induced downregulation of aquaporin-2 water channel expression in rat kidney medulla and cortex.  J Clin Invest  1996; 97:1960-1968.

173. Earm JH, Christensen BM, Frokiaer J, et al: Decreased aquaporin-2 expression and apical plasma membrane delivery in kidney collecting ducts of polyuric hypercalcemic rats.  J Am Soc Nephrol  1998; 9:2181-2193.

174. Sands JM, Naruse M, Jacobs JD, et al: Changes in aquaporin-2 protein contribute to the urine concentrating defect in rats fed a low-protein diet.  J Clin Invest  1996; 97:2807-2814.

175. Frokiaer J, Christensen BM, Marples D, et al: Downregulation of aquaporin-2 parallels changes in renal water excretion in unilateral ureteral obstruction.  Am J Physiol  1997; 273:F213-F223.

176. Bendz H, Aurell M: Drug-induced diabetes insipidus: Incidence, prevention and management.  Drug Saf  1999; 21:449-456.

177. Christensen S, Kusano E, Yusufi AN, et al: Pathogenesis of nephrogenic diabetes insipidus due to chronic administration of lithium in rats.  J Clin Invest  1985; 75:1869-1879.

178. Marples D, Christensen S, Christensen EI, et al: Lithium-induced downregulation of aquaporin-2 water channel expression in rat kidney medulla.  J Clin Invest  1995; 95:1838-1845.

179. Bendz H, Sjodin I, Aurell M: Renal function on and off lithium in patients treated with lithium for 15 years or more. A controlled, prospective lithium-withdrawal study.  Nephrol Dial Transplant  1996; 11:457-460.

180. Markowitz GS, Radhakrishnan J, Kambham N, et al: Lithium nephrotoxicity: A progressive combined glomerular and tubulointerstitial nephropathy.  J Am Soc Nephrol  2000; 11:1439-1448.

181. Fernandez-Llama P, Andrews P, Ecelbarger CA, et al: Concentrating defect in experimental nephrotic syndrone: Altered expression of aquaporins and thick ascending limb Na+ transporters.  Kidney Int  1998; 54:170-179.

182. Terris J, Ecelbarger CA, Nielsen S, et al: Long-term regulation of four renal aquaporins in rats.  Am J Physiol  1996; 271:F414-F422.

183. Harrington AR, Valtin H: Impaired urinary concentration after vasopressin and its gradual correction in hypothalamic diabetes insipidus.  J Clin Invest  1968; 47:502.

184. Goldman MB, Luchins DJ, Robertson GL: Mechanisms of altered water metabolism in psychotic patients with polydipsia and hyponatremia.  N Engl J Med  1988; 318:397-403.

185. de Leon J, Verghese C, Tracy JI, et al: Polydipsia and water intoxication in psychiatric patients: A review of the epidemiological literature.  Biol Psychol  1994; 35:408-419.

186. Robertson GL: Differential diagnosis of polyuria.  Annu Rev Med  1988; 39:425-442.

187. Valtin H: “Drink at least eight glasses of water a day.” Really? Is there scientific evidence for “8 × 8”?.  Am J Physiol Regul Integr Comp Physiol  2002; 283:R993-R1004.

188. Goldman MB, Robertson GL, Luchins DJ, et al: The influence of polydipsia on water excretion in hyponatremic, polydipsic, schizophrenic patients.  J Clin Endocrinol Metab  1996; 81:1465-1470.

189. Vieweg WV, Carey RM, Godleski LS, et al: The syndrome of psychosis, intermittent hyponatremia, and polydipsia: Evidence for diurnal volume expansion.  Psychol Med  1990; 8:135-144.

190. Cheng JC, Zikos D, Skopicki HA, et al: Long-term neurologic outcome in psychogenic water drinkers with severe symptomatic hyponatremia: the effect of rapid correction.  Am J Med  1990; 88:561-566.

191. Vieweg WV, Robertson GL, Godleski LS, et al: Diurnal variation in water homeostasis among schizophrenic patients subject to water intoxication.  Schizophr Res  1988; 1:351-357.

192. Goldman MB, Robertson GL, Luchins DJ, et al: Psychotic exacerbations and enhanced vasopressin secretion in schizophrenic patients with hyponatremia and polydipsia.  Arch Gen Psychiatry  1997; 54:443-449.

193. Adrogue HJ, Madias NE: Hypernatremia.  N Engl J Med  2000; 342:1493-1499.

194. Palevsky PM, Bhagrath R, Greenberg A: Hypernatremia in hospitalized patients.  Ann Intern Med  1996; 124:197-203.

195. Riggs JE: Neurologic manifestations of fluid and electrolyte disturbances.  Neurol Clin  1989; 7:509-523.

196. Palevsky PM: Hypernatremia.  Semin Nephrol  1998; 18:20-30.

197. Gullans SR, Verbalis JG: Control of brain volume during hyperosmolar and hypoosmolar conditions.  Annu Rev Med  1993; 44:289-301.

198. Ayus JC, Armstrong DL, Arieff AI: Effects of hypernatraemia in the central nervous system and its therapy in rats and rabbits.  J Physiol  1996; 492(pt 1):243-255.

199. Kahn A, Brachet E, Blum D: Controlled fall in natremia and risk of seizures in hypertonic dehydration.  Intensive Care Med  1979; 5:27-31.

200. Robertson GL: Diabetes insipidus.  Endocrinol Metab Clin North Am  1995; 24:549-572.

201. Miller M, Dalakos T, Moses AM, et al: Recognition of partial defects in antidiuretic hormone secretion.  Ann Intern Med  1970; 73:721-729.

202. Moses AM: Clinical and laboratory observations in the adult with diabetes insipidus and related syndromes.   In: Czernichow P, Robinson AG, ed. Diabetes Insipidus in Man,  Karger: Basel; 1985:156-175.

203. Robinson AG: Disorders of antidiuretic hormone secretion.  Clin Endocrinol Metab  1985; 14:55-88.

204. Baylis PH, Gaskill MB, Robertson GL: Vasopressin secretion in primary polydipsia and cranial diabetes insipidus.  Q J Med  1981; 50:345-358.

205. Milles JJ, Spruce B, Baylis PH: A comparison of diagnostic methods to differentiate diabetes insipidus from primary polyuria: A review of 21 patients.  Acta Endocrinol (Copenh)  1983; 104:410-416.

206. Fujisawa I, Asato R, Nishimura K, et al: Anterior and posterior lobes of the pituitary gland: Assessment by 1.5 T MR imaging.  J Comput Assist Tomogr  1987; 11:214-220.

207. Arslan A, Karaarslan E, Dincer A: High intensity signal of the posterior pituitary. A study with horizontal direction of frequency-encoding and fat suppression MR techniques.  Acta Radiol  1999; 40:142-145.

208. Kurokawa H, Fujisawa I, Nakano Y, et al: Posterior lobe of the pituitary gland: Correlation between signal intensity on T1-weighted MR images and vasopressin concentration.  Radiology  1998; 207:79-83.

209. Moses AM, Clayton B, Hochhauser L: Use of T1-weighted MR imaging to differentiate between primary polydipsia and central diabetes insipidus. Comment. [see comments].  AJNR Am J Neuroradiol  1992; 13:1273-1277.

210. Brooks BS, el Gammal T, Allison JD, et al: Frequency and variation of the posterior pituitary bright signal on MR images.  AJNR Am J Neuroradiol  1989; 10:943-948.

211. Maghnie M, Genovese E, Bernasconi S, et al: Persistent high MR signal of the posterior pituitary gland in central diabetes insipidus.  AJNR Am J Neuroradiol  1997; 18:1749-1752.

212. Dohanics J, Hoffman GE, Verbalis JG: Chronic hyponatremia reduces survival of magnocellular vasopressin and oxytocin neurons following axonal injury.  J Neurosci  1996; 16:2372-2380.

213. Bonneville JF, Cattin F, Dietemann JL: The pituitary stalk.   In: Bonneville JF, ed. Computed Tomography of the Pituitary Gland,  New York: Springer-Verlag; 1986:106-114.

214. Leger J, Velasquez A, Garel C, et al: Thickened pituitary stalk on magnetic resonance imaging in children with central diabetes insipidus.  J Clin Endocrinol Metab  1999; 84:1954-1960.

215. Czernichow P, Garel C, Leger J: Thickened pituitary stalk on magnetic resonance imaging in children with central diabetes insipidus.  Horm Res  2000; 53(suppl 3):61-64.

216. Bruck E, Abal G, Aceto Jr T: Pathogenesis and pathophysiology of hypertonic dehydration with diarrhea. A clinical study of 59 infants with observations of respiratory and renal water metabolism.  Am J Dis Child  1968; 115:122-144.

217. Blum D, Brasseur D, Kahn A, et al: Safe oral rehydration of hypertonic dehydration.  J Pediatr Gastroenterol Nutr  1986; 5:232-235.

218. Fjellestad-Paulsen A, Hoglund P, Lundin S, et al: Pharmacokinetics of 1-deamino-8-d-arginine vasopressin after various routes of administration in healthy volunteers.  Clin Endocrinol (Oxf)  1993; 38:177-182.

219. Robinson AG: DDAVP in the treatment of central diabetes insipidus.  N Engl J Med  1976; 294:507-511.

220. Sklar AH, Schrier RW: Central nervous system mediators of vasopressin release.  Physiol Rev  1983; 63:1243-1280.

221. Hoffman PK, Share L, Crofton JT, et al: The effect of intracerebroventricular indomethacin on osmotically stimulated vasopressin release.  Neuroendocrinology  1982; 34:132-139.

222. Nadler SP, Hebert SC, Brenner BM: PGE2, forskolin, and cholera toxin interactions in rabbit cortical collecting tubule.  Am J Physiol  1986; 250:F127-F135.

223. Allen HM, Jackson RL, Winchester MD, et al: Indomethacin in the treatment of lithium-induced nephrogenic diabetes insipidus.  Arch Intern Med  1989; 149:1123-1126.

224. Delaney V, de Pertuz Y, Nixon D, et al: Indomethacin in streptozocin-induced nephrogenic diabetes insipidus.  Am J Kidney Dis  1987; 9:79-83.

225. Magaldi AJ: New insights into the paradoxical effect of thiazides in diabetes insipidus therapy.  Nephrol Dial Transplant  2000; 15:1903-1905.

226. Richardson DW, Robinson AG: Desmopressin.  Ann Intern Med  1985; 103:228-239.

227. Lam KS, Wat MS, Choi KL, et al: Pharmacokinetics, pharmacodynamics, long-term efficacy and safety of oral 1-deamino-8-d-arginine vasopressin in adult patients with central diabetes insipidus.  Br J Clin Pharmacol  1996; 42:379-385.

228. Rittig S, Jensen AR, Jensen KT, et al: Effect of food intake on the pharmacokinetics and antidiuretic activity of oral desmopressin (DDAVP) in hydrated normal subjects.  Clin Endocrinol (Oxf)  1998; 48:235-241.

229. Dunn AL, Powers JR, Ribeiro MJ, et al: Adverse events during use of intranasal desmopressin acetate for haemophilia A and von Willebrand disease: A case report and review of 40 patients.  Haemophilia  2000; 6:11-14.

230. Robson WL, Norgaard JP, Leung AK: Hyponatremia in patients with nocturnal enuresis treated with DDAVP.  Eur J Pediatr  1996; 155:959-962.

231. Schwab M, Wenzel D, Ruder H: Hyponatraemia and cerebral convulsion due to short term DDAVP therapy for control of enuresis nocturna.  Eur J Pediatr  1996; 155:46-48.

232. Lugo N, Silver P, Nimkoff L, et al: Diagnosis and management algorithm of acute onset of central diabetes insipidus in critically ill children.  J Pediatr Endocrinol Metab  1997; 10:633-639.

233. Ralston C, Butt W: Continuous vasopressin replacement in diabetes insipidus.  Arch Dis Child  1990; 65:896-897.

234. Davison JM, Sheills EA, Philips PR, et al: Metabolic clearance of vasopressin and an analogue resistant to vasopressinase in human pregnancy.  Am J Physiol  1993; 264:F348-F353.

235. Kallen BA, Carlsson SS, Bengtsson BK: Diabetes insipidus and use of desmopressin (minirin) during pregnancy.  Eur J Endocrinol  1995; 132:144-146.

236. Ray JG: DDAVP use during pregnancy: An analysis of its safety for mother and child.  Obstet Gynecol Surv  1998; 53:450-455.

237. Iwasaki Y, Oiso Y, Kondo K, et al: Aggravation of subclinical diabetes insipidus during pregnancy.  N Engl J Med  1991; 324:522-526.

238. Libber S, Harrison H, Spector D: Treatment of nephrogenic diabetes insipidus with prostaglandin synthesis inhibitors.  J Pediatr  1986; 108:305-311.

239. Kirchlechner V, Koller DY, Seidl R, et al: Treatment of nephrogenic diabetes insipidus with hydrochlorothiazide and amiloride.  Arch Dis Child  1999; 80:548-552.

240. Uyeki TM, Barry FL, Rosenthal SM, et al: Successful treatment with hydrochlorothiazide and amiloride in an infant with congenital nephrogenic diabetes insipidus.  Pediatr Nephrol  1993; 7:554-556.

241. Cesar KR, Magaldi AJ: Thiazide induces water absorption in the inner medullary collecting duct of normal and Brattleboro rats.  Am J Physiol  1999; 277:F756-F760.

242. Usberti M, Dechaux M, Guillot M, et al: Renal prostaglandin E2 in nephrogenic diabetes insipidus: Effects of inhibition of prostaglandin synthesis by indomethacin.  J Pediatr  1980; 97:476-478.

243. Chevalier RL, Rogol AD: Tolmetin sodium in the management of nephrogenic diabetes insipidus.  J Pediatr  1982; 101:787-789.

244. Batlle DC, von Riotte AB, Gaviria M, et al: Amelioration of polyuria by amiloride in patients receiving long-term lithium therapy.  N Engl J Med  1985; 312:408-414.

245. Singer I, Oster JR, Fishman LM: The management of diabetes insipidus in adults.  Arch Intern Med  1997; 157:1293-1301.

246. Postina R, Ufer E, Pfeiffer R, et al: Misfolded vasopressin V2 receptors caused by extracellular point mutations entail congenital nephrogenic diabetes insipidus.  Mol Cell Endocrinol  2000; 164:31-39.

247. Robertson GL: Abnormalities of thirst regulation.  Kidney Int  1984; 25:460-469.

248. Canuso CM, Goldman MB: Clozapine restores water balance in schizophrenic patients with polydipsia-hyponatremia syndrome.  J Neuropsychiatry Clin Neurosci  1999; 11:86-90.

249. Robertson GL: Dipsogenic diabetes insipidus: A newly recognized syndrome caused by a selective defect in the osmoregulation of thirst.  Trans Assoc Am Phys  1987; 100:241-249.

250. Anderson RJ: Hospital-associated hyponatremia.  Kidney Int  1986; 29:1237.

251. Hawkins RC: Age and gender as risk factors for hyponatremia and hypernatremia.  Clin Chim Acta  2003; 337:169-172.

252. Wattad A, Chiang ML, Hill LL: Hyponatremia in hospitalized children.  Clin Pediatr (Phila)  1992; 31:153.

253. Saito T: Hyponatremia in elderly patients.  Intern Med  2001; 40:851.

254. Kumar S, Berl T: Sodium.  Lancet  1998; 352:220-228.

255. Turchin A, Seifter JL, Seely EW: Clinical problem-solving. Mind the gap.  N Engl J Med  2003; 349:1465-1469.

256. Steinberger BA, Ford SM, Coleman TA: Intravenous immunoglobulin therapy results in post-infusional hyperproteinemia, increased serum viscosity, and pseudohyponatremia.  Am J Hematol  2003; 73:97-100.

257. Perez-Perez AJ, Pazos B, Sobrado J, et al: Acute renal failure following massive mannitol infusion.  Am J Nephrol  2002; 22:573-575.

258. Guglielminotti J, Pernet P, Maury E, et al: Osmolar gap hyponatremia in critically ill patients: Evidence for the sick cell syndrome?.  Crit Care Med  2002; 30:1051-1055.

259. Hillier TA, Abbott RD, Barrett EJ: Hyponatremia: Evaluating the correction factor for hyperglycemia.  Am J Med  1999; 106:399-403.

260. Ayus JC, Arieff AI: Glycine-induced hypo-osmolar hyponatremia.  Arch Intern Med  1997; 557:223.

261. Berl T, Schrier RW: Disorders of water metabolism.   In: Schrier RW, ed. Renal and Electrolyte Disorders,  6th ed. Philadelphia: Lippincott Williams & Wilkins; 2003.

262. Robertson GL: Physiopathology of ADH secretion.   In: Tolis G, Labrie F, Martin JB, et al ed. Clinical Neuroendocrinology: A Pathophysiological Approach,  New York: Raven Press; 1979:247.

263. Nguyen MK, Kurtz I: New insights into the pathophysiology of the dysnatremias: A quantitative analysis.  Am J Physiol Renal Physiol  2004; 287:F172-F180.

264. Sterns R: Sodium and water balance disorders.  Neph SAP  2006; 5:35-50.

265.   Parikh C, Kumar S, Berl T: Disorders of water metabolism. In Johnson RR, Feehaly J (eds): Comprehensive Clinical Nephrology, 3rd ed. St Louis, CV Mosby, 2007 (in press).

266. Androgue HJ, Madias NE: Hyponatremia.  N Engl J Med  2000; 342:1581-1589.

267. Chow KM, Kwan BC, Szeto CC: Clinical studies of thiazide-induced hyponatremia.  J Natl Med Assoc  2004; 96:1305-1308.

268. Sonnenblick M, Friedlander Y, Rosin AJ: Diuretic induced severe hyponatremia. Review and analysis of 129 reported patients.  Chest  1993; 103:601.

269. Sharabi Y, Illan R, Kamari Y, et al: Diuretic induced hyponatraemia in elderly hypertensive women.  J Hum Hypertens  2002; 16:631-635.

270. Clark B, Shannon R, Rosa R, Epstein F: Increased susceptibility to thiazide induced hyponatremia in the elderly.  J Am Soc Nephrol  1994; 5:1106.

271. Berl T: Water metabolism in potassium depletion.  Miner Electrolyte Metab  1980; 4:209.

272. Fichman MP, Vorherr H, Kleeman CR, Tefler N: Diuretic induced hyponatremia.  Ann Intern Med  1971; 75:853.

273. Danovitch GM, Bourgoignie J, Bricker NS: Reversibility of the salt losing tendency of chronic renal failure.  N Engl J Med  1977; 296:14.

274. Schrier RW, Linas SL: Mechanisms of the defect in water excretion in adrenal insufficiency.  Miner Electrolyte Metab  1980; 4:1.

275. Boykin J, McCool A, Robertson G, et al: Mechanisms of impaired water excretion in mineralocorticoid deficient dogs.  Miner Electrolyte Metab  1979; 2:310.

276. Schrier RW: Body water homeostasis: Clinical disorders of urinary dilution and concentration.  J Am Soc Nephrol  2006; 17:1820-1832.

277. Palmer BF: Hyponatremia in patients with central nervous system disease: SIADH versus CSW.  Trends Endocrinol Metab  2003; 14:182-187.

278. Bohn S, Carlotti P, Cusimono M, et al: Cerebral salt wasting: Truth, fallacies, theories and challenges.  Crit Care Med  2002; 30:2575.

279. Oh MS, Carroll HJ: Cerebral salt-wasting syndromes.  Crit Care Clin  2001; 17:125-138.

280. McGirt MJ, Blessing R, Nimjee SM, et al: Correlation of serum brain natriuretic peptide with hyponatremia and delayed ischemic neurological deficits after subarachnoid hemorrhage.  Neurosurgery  2004; 54:1369-1373.discussion 1373-1364

281. Schrier RW, Gurevich AK, Cadnapaphornchai MA: Pathogenesis and management of sodium and water retention in cardiac failure and cirrhosis.  Semin Nephrol  2002; 2:157-172.

282. Ishikawa S, Saito S, Okada K, et al: Effect of vasopressin antagonist on water excretion in vena cava constriction.  Kidney Int  1986; 30:49.

283. Szatalowicz VL, Arnold PE, Chaimovitz C, et al: Radioimmunoassay of plasma arginine vasopressin in hyponatremic patients with congestive heart failure.  N Engl J Med  1981; 305:263.

284. Kim JK, Michel JB, Soubrier F, Schrier RW: Arginine vasopressin gene expression in chronic cardiac failure in rats.  Kidney Int  1990; 38:818.

285. Ferguson DW, Berg WJ, Sanders JS: Clinical and hemodynamic correlates of sympathetic nerve activity in normal humans and patients with heart failure. Evidence from direct microneurographic recordings.  J Am Coll Cardiol  1990; 16:1125.

286. Benedict C, Johnston D, Weiner D, et al: Relation of neurohumoral activation to clinical variables and degrees of ventricular dysfunction.  J Am Coll Cardiol  1994; 23:1410.

287. Lee W, Packer M: Prognostic importance of serum sodium concentration and its modification by converting enzyme inhibitors.  Circulation  1986; 73:257.

288. Nielsen S, Torris D, Andersen C: Congestive heart failure in rats is associated with increased expression and targeting of aquaporin 2 water channel in collecting duct.  Proc Natl Acad Sci U S A  1997; 94:5450.

289. Xu DL, Martin P-Y, Ohara M, et al: Upregulation of aquaporin 2 water channel expression in chronic heart failure rat.  J Clin Invest  1997; 99:1500.

290. Martin PY, Abraham WT, Lieming X, et al: Selective V2-receptor vasopressin antagonism decreases urinary aquaporin-2 excretion in patients with chronic heart failure.  J Am Soc Nephrol  1999; 10:2165-2170.

291. Gines P, Berl T, Bernardi M, et al: Hyponatremia in cirrhosis: From pathogenesis to treatment.  Hepatology  1998; 28:851-864.

292. Schrier RW: Mechanisms of disturbed renal water excretion in cirrhosis.  Gastroenterology  1983; 84:870.

293. Arroyo V, Jimenez W: Complications of cirrhosis: Renal and circulatory dysfunction, light shadows in an important clinical problem.  J Hepatol  2000; 32(suppl 1):157.

294. Unikowsky B, Wexler JJ, Levy M: Dogs with experimental cirrhosis of the liver but without intrahepatic hypertension do not retain sodium or form ascites.  J Clin Invest  1983; 72:1594.

295. Rahman SN, Abraham W, Schrier RW: Peripheral arterial vasodilation in cirrhosis.  Gastroenterol Int  1992; 5:192.

296. Kim J, Summer S, Howard R, Schrier RW: Vasopressin gene expression in rats with experimental cirrhosis.  Hepatology  1993; 17:143.

297. Claria J, Jimenez W, Arroyo V, et al: Blockade of the hydroosmotic effect of vasopressin normalizes water excretion in cirrhotic rats.  Gastroenterology  1989; 97:1294.

298. Fujita N, Ishikawa S, Sasaki S: Role of water channel AQP-CD in water retention in SIADH and cirrhotic rats.  Am J Physiol  1994; 269:F926.

299. Fernandez-Llama P, Jimenez W, Bosch-Marce M, et al: Dysregulation of renal aquaporins and Na-Cl cotransporter in CCI4-induced cirrhosis.  Kidney Int  2000; 58:216-228.

300. Fernandez-Llama P, Turner R, Dibona G, Knepper MA: Renal expression of aquaporins in liver cirrhosis induced by chronic common bile duct ligation in rats.  J Am Soc Nephrol  1999; 10:1950-1957.

301. Bichet D, Van Putten VJ, Schrier RW: Potential role of increased sympathetic activity in impaired sodium and water excretion in cirrhosis.  N Engl J Med  1982; 307:1552.

302. Floras J, Legaut L, Morali GA: Increased sympathetic outflow in cirrhosis and ascites. Direct evidence from intraneural recordings.  Ann Intern Med  1991; 114:373.

303. Bichet DG, Groves BM, Schrier RW: Mechanism of improvement of water and sodium excretion by enhancement of central hemodynamics in decompensated cirrhosis.  Kidney Int  1983; 24:788.

304. Shapiro M, Nichols K, Groves B, Schrier RW: Interrelationship between cardiac output and vascular resistance as determinants of effective arterial blood volume in cirrhotic patients.  Kidney Int  1985; 28:201.

305. Weigert A, Martin P, Niederberger M: Edothelium-dependent vascular hyporesponsiveness without detection of nitric oxide synthase induction in aorta of cirrhotic rats.  Hepatology  1997; 22:1856.

306. Martin P-Y, Ohara M, Gines P: Nitric oxide synthase (NOS) inhibition for one week improves sodium and water excretion in cirrhotic rats with ascites.  J Clin Invest  1998; 201:235.

307. Martin P-Y, Gines P, Schrier RW: Nitric oxide as a mediator of hemodynamic abnormalities and sodium and water retention in cirrhosis.  N Engl J Med  1998; 339:533.

308. Abraham W, Cadnapopornchai M, Schrier RW: Cardiac failure, liver disease and nephrotic syndrome.   In: Schrier RW, Gottschalk CW, ed. Disease of the Kidney,  7th ed. Philadelphia: Lippincott Williams & Wilkins; 2001:2465.

309. Usberti M, Federico S, Mecariello S, et al: Role of plasma vasopressin in the impairment of water excretion in nephrotic syndrome.  Kidney Int  1984; 25:422.

310. Kimagi H, Onayma K, Isehi K: Role of renin-angiotensin-aldosterone in minimal change nephrotic syndrome.  Clin Nephrol  1985; 25:229.

311. Meltzer JI, Keim HJ, Laragh JH, et al: Nephrotic syndrome: Vasoconstriction and hypervolemic types indicated by renin-sodium profiling.  Ann Intern Med  1979; 91:688.

312. Schrier RW, Fasset RG: A critique of the overfill hypothesis of sodium and water retention in the nephrotic syndrome.  Kidney Int  1998; 53:1111.

313. Ichikawa I, Rennke HG, Hoyer JR, et al: Role of intrarenal mechanisms in the impaired salt excretion in experimental nephrotic syndrome.  J Clin Invest  1983; 71:91.

314. Apostol E, Ecelbarger CA, Terris J, et al: Reduced renal medullary water channel expression in puromycin aminonucleoside-induced nephrotic syndrome.  J Am Soc Nephrol  1997; 8:15-24.

315. Fernandez-Llama P, Andrews P, Ecelbarger CA, et al: Concentrating defect in experimental nephrotic syndrone: Altered expression of aquaporins and thick ascending limb Na+ transporters.  Kidney Int  1998; 54:170-179.

316. Gross P, Raascher W: Vasopressin and hyponatremia in renal insufficiency.  Contrib Nephrol  1986; 50:54.

317. Weiss NM, Robertson GL: Water metabolism in endocrine disorders.  Semin Nephrol  1987; 4:303.

318. Ishikawa S, Fujisawa G, Tsuboi Y, et al: Role of antidiuretic hormone in hyponatremia in patients with isolated adrenocorticotropic hormone deficiency.  Endocrinol Jpn  1991; 38:325.

319. Oelkers W: Hyponatremia and inappropriate secretion of vasopressin in patients with hypopituitarism.  N Engl J Med  1989; 321:492.

320. Boykin J, de Torrente A, Erickson A, et al: Role of plasma vasopressin in impaired water excretion of glucocorticoid deficiency.  J Clin Invest  1978; 62:738.

321. Linas SL, Berl T, Robertson GL, et al: Role of vasopressin in the impaired water excretion of glucocorticoid deficiency.  Kidney Int  1980; 18:58.

322. Pyo HI, Summer SN, Kim JK: Vasopressin gene expression in glucocorticoid hormone deficient rats.  Ann N Y Acad Sci  1993; 689:659.

323. Kim JK, Summer SN, Wood WM, et al: Role of glucocorticoid hormones in arginine vasopressin gene regulation.  Biochem Biophys Res Commun  2001; 289:1252-1256.

324. Berghorn KA, Knapp LT, Hoffman GE, Sherman TG: Induction of glucocorticoid receptor expression in hypothalamus neurons during chronic hypoosmolality.  Endocrinology  1995; 136:804.

325. Schwartz MJ, Kokko JP: Urinary concentrating defect of adrenal insufficiency. Permissive role of adrenal steroids on the hydroosmotic response across the rabbit collecting tubule.  J Clin Invest  1980; 66:234.

326. Jackson BA, Braun-Werness J, Kusano E, Dousa T: Concentrating defect in adrenalectomized rat.  J Clin Invest  1983; 72:997.

327. Kwon TH, Nielson J, Masilamani S, et al: Regulation of collection duct AQP3 expression response to mineralocorticoid.  Am J Physiol Renal Physiol  2002; 283:F1403-F1421.

328. Olchovsky D, Ezra D, Vered I, et al: Symptomatic hyponatremia as a presenting sign of hypothalamic-pituitary disease: A syndrome of inappropriate secretion of antidiuretic hormone (SIADH)—like glucocorticosteroid responsive condition.  J Endocrinol Invest  2005; 28:151-156.

329. Diederich S, Franzen NF, Bahr V, et al: Severe hyponatremia due to hypopituitarism with adrenal insufficiency: Report on 28 cases.  Eur J Endocrinol  2003; 148:609-617.

330. Hanna F, Scanlon M: Hyponatremia, hypothyroidism and role of arginine vasopressin.  Lancet  1997; 350:755.

331. Hochberg Z, Benderly A: Normal osmotic threshold for vasopressin release in the hyponatremia of hypothyroidism.  Horm Res  1983; 18:128.

332. Iwasaki Y, Oiso Y, Yamauchi K: Osmoregulation of plasma vasopressin in myxedema.  J Clin Endocrinol Metab  1990; 70:534.

333. Howard R, Summer S, Rossi N: Short term hypothyroidism and vasopressin gene expression in the rat.  Am J Kidney Dis  1992; 19:573.

334. Seif SM, Robinson AG, Zenser TV, et al: Neurohypophyseal peptides in hypothyroid rats: Plasma levels and kidney response.  Metabolism  1979; 28:137.

335. Harckom TM, Kim JK, Palumbo PJ, et al: Medullary effect of thyroid function on enzymes of the vasopressin-sensitive adenosine 3′,5′-monophosphate system in renal medulla.  Endocrinology  1978; 102:1475.

336. Cadnapaphornchai MA, Kim YW, Gurevich AK, et al: Urinary concentrating defect in hypothyroid rats: Role of sodium, potassium, 2-chloride co-transporter, and aquaporins.  J Am Soc Nephrol  2003; 14:566-574.

337. Chen YC, Cadnapaphornchai MA, Yang J, et al: Nonosmotic release of vasopressin and renal aquaporins in impaired urinary dilution in hypothyroidism.  Am J Physiol Renal Physiol  2005; 289:F672-F678.

338. Riggs AT, Dysken MW, Kim SW, Opsahl JA: A review of disorders of water homeostasis in psychiatric patients.  Psychosomatics  1991; 32:133.

339. Hariprasad MK, Eisinger RP, Nadler IM, et al: Hyponatremia in psychogenic polydipsia.  Arch Intern Med  1980; 140:1639-1642.

340. Jose CJ, Perez Crult J: Incidence and morbidity of self-induced water intoxication in state mental hospital patients.  Am J Psychiatry  1979; 136:221.

341. Shah PJ, Greenberg WM: Water intoxication precipitated by thiazide diuretics in polydipsic psychiatric patients.  Am J Psychiatry  1991; 148:1424-1425.

342. Brows RP, Koesis JM, Cohen SK: Delusional depression and inappropriate antidiuretic hormone secretion.  Biol Psychiatry  1983; 18:1059.

343. Goldman MB, Robertson GL, Luchins DJ, et al: Psychotic exacerbations and enhanced vasopressin secretion in schizophrenic patients with hyponatremia and polydipsia.  Arch Gen Psychiatry  1997; 54:443-449.

344. Kawai N, Atsuomi B, Toshihito S, Hiroyasu S: Roles of arginine vasopressin and atrial natriuretic peptide in polydipsia-hyponatremia of schizophrenic patients.  Psychiatry Res  2001; 101:37-45.

345. Berl T: Psychosis and water balance.  N Engl J Med  1988; 318:441.

346. Fenves AZ, Thomas S, Knochel JP: Beer potomania: Two cases and review of the literature.  Clin Nephrol  1996; 45:61-64.

347. Musch W, Xhaet O, Decaux G: Solute loss plays a major role in polydipsia-related hyponatraemia of both water drinkers and beer drinkers.  QJM  2003; 96:421-426.

348. Hoorn EJ, Geary D, Robb M, et al: Acute hyponatremia related to intravenous fluid administration in hospitalized children: An observational study.  Pediatrics  2004; 113:1279-1284.

349. Chung H-M, Kluge R, Schrier RW, Anderson RJ: Post-operative hyponatremia.  Arch Intern Med  1986; 146:333.

350. Tambe AA, Hill R, Livesley PJ: Post-operative hyponatraemia in orthopaedic injury.  Injury  2003; 34:253-255.

351. Anderson RJ, Chung H-M, Kluge R, Schrier RW: Hyponatremia: A prospective analysis of its epidemiology and the pathogenetic role of vasopressin.  Ann Intern Med  1985; 102:164.

352. Shafiee MA, Charest AF, Cheema-Dhadli S, et al: Defining conditions that lead to the retention of water: The importance of the arterial sodium concentration.  Kidney Int  2005; 67:613-621.

353. Steele A, Growishankar A, Abramson S, et al: Postoperative hyponatremia despite near isotonic saline infursion. A phenomenon of desalination.  Ann Intern Med  1997; 126:20.

354. Aronson D, Dragu RE, Nakhoul F, et al: Hyponatremia as a complication of cardiac catheterization: a prospective study.  Am J Kidney Dis  2002; 40:940-946.

355. Arieff AI: Permanent neurological disability from hyponatremia in healthy women undergoing elective surgery.  N Engl J Med  1986; 314:1529.

356. Ayus JC, Wheeler J, Arieff AI: Postoperative hyponatremic encephalopathy in menstruant women.  Ann Intern Med  1992; 117:891.

357. O'Brien KK, Montain SJ, Corr WP, et al: Hypernatremia associated with hyponatremia in US Army trainees.  Mil Med  2001; 166:405-410.

358. Davis DP, Videen JS, Marino A, et al: Exercise associated hyponatremia in marathon runners: A two year experience.  J Emerg Med  2001; 21:47-57.

359. Montain SJ, Sawka MN, Wenger CB: Hyponatremia associated with exercise: Risk factors and pathogenesis.  Exerc Sports Sci Rev  2001; 29:113-117.

360. Almond CS, Shin AY, Fortescue EB, et al: Hyponatremia among runners in the Boston Marathon.  N Engl J Med  2005; 352:1550-1556.

361. Noakes TD, Sharwood K, Collins M, et al: The dipsomania of great distance: Water intoxication in an Ironman triathlete.  Br J Sports Med  2004; 38:E16.

362. Bertholini DM, Butler CS: Severe hyponatraemia secondary to desmopressin therapy in von Willebrand's disease.  Anaesth Intensive Care  2000; 28:199-201.

363. Schwab M, Wenzel D, Ruder H: Hyponatraemia and cerebral convulsion due to short term DDAVP therapy for control of enuresis nocturna.  Eur J Pediatr  1996; 155:46-48.

364. Shindel A, Tobin G, Klutke C: Hyponatremia associated with desmopressin for the treatment of nocturnal polyuria.  Urology  2002; 60:344.

365. Hirokawa CA, Gray DR: Chlorpropamide-induced hyponatremia in the veteran population.  Ann Pharmacother  1992; 26:1243.

366. Mendoza SA, Brown Jr CF: Effect of chlorpropamide on osmotic water flow across toad bladder and the response to vasopressin, theophylline and cyclic AMP.  J Clin Endocrinol Metab  1974; 38:883-889.

367. Durr JA, Hensen J, Ehnis T, et al: Chlorpropamide upregulates antidiuretic hormone receptors and unmasks constitutive receptor signaling.  Am J Physiol Renal Physiol  2000; 278:F799-F808.

368. Kusano B, Brain-Werness JL, Vich DJ, et al: Chlorpropamide action on renal concentrating mechanism in rats with hypothalamic diabetes insipidus.  J Clin Invest  1983; 72:1298.

369. Kastner T, Friedman DL, Pond WS: Carbamazepine-induced hyponatremia in patients with mental retardation.  Am J Ment Retard  1992; 96:536.

370. Cooney JA: Carbamazepine and SIADH.  Am J Psychiatry  1990; 147:1101.

371. Steinhoff BJ, Stoll KD, Stodieck SR, Paulus W: Hyponatremic coma under oxcarbazepine therapy.  Epilepsy Res  1995; 11:67.

372. Meinders HE, Cejka V, Robertson GL: Antidiuretic action of carbamazepine.  Clin Sci Mol Med  1974; 47:289.

373. Gold PW, Robertson GL, Ballenger J, et al: Carbamazepine diminishes the sensitivity of the plasma arginine vasopressin response to osmotic stimulation.  J Clin Endocrinol Metab  1983; 57:952.

374. Kosten TR, Camp W: Inappropriate secretion of antidiuretic hormone in a patient receiving piperazine phenothiazines.  Psychosomatics  1980; 21:351.

375. Peck V, Shenkman L: Haloperidol-induced syndrome of inappropriate secretion of antidiuretic hormone.  Clin Pharmacol Ther  1979; 26:442.

376. Beckstrom D, Reding R, Cerletti J: Syndrome of inappropriate antidiuretic hormone secretion associated with amitriptyline administration.  JAMA  1979; 241:133.

377. Cherney DZ, Davids MR, Halperin ML: Acute hyponatremia and “ecstasy”: Insights from a quantative and integrated analysis.  Q J Med  2002; 95:475-483.

378. Budisavljevic MN, Stewart L, Sahn SA, et al: Hyponatremia associated with 3,4-methylenedioxymethylamphetamine (“Ecstasy”) abuse.  Am J Med Sci  2003; 326:89-93.

379. Vishwanath BM, Vavalgund A, Cusando W, et al: Fluoxetine as a cause of SIADH.  Am J Psychiatry  1991; 148:542.

380. Kessler J, Samuels S: Sertraline and hyponatremia. Letter.  N Engl J Med  1996; 335:524.

381. Fabian TJ, Amico JA, Kroboth PD, et al: Paroxetine-induced hyponatremia in the elderly due to the syndrome of inappropriate secretion of antidiuretic hormone (SIADH).  J Geriatr Psychiatry Neurol  2003; 16:160-164.

382. Spigset O, Hedermalm K: Hyponatremia in relation to treatment with antidepressants.  Pharmacotherapy  1997; 17:348.

383. Kirby D, Harrigan S, Ames D: Hyponatremia in elderly psychiatric patients treated with selective serotonin reuptake inhibitors and venlafaxine: A retrospective controlled study in an inpatient unit.  Int J Geriatr Psychiatry  2002; 17:231-237.

384. Strachan J, Shepherd J: Hyponatraemia associated with the use of selective serotonin reuptake inhibitors.  Aust N Z J Psychiatry  1998; 32:295-298.

385. Bouman WP, Pinner G, Johnson H: Incidence of selective serotonin reuptake inhibitor (SSRI) induced hyponatraemia due to the syndrome of inappropriate antidiuretic hormone (SIADH) secretion in the elderly.  Int J Geriatr Psychiatry  1998; 13:12-15.

386. Robertson GL, Bhoopalam N, Zelkowitz LJ: Vincristine neurotoxicity and abnormal secretion of antidiuretic hormone.  Arch Intern Med  1973; 132:717.

387. Hammond IW, Ferguson JA, Kwong K, et al: Hyponatremia and syndrome of inappropriate anti-diuretic hormone reported with the use of vincristine: An over-representation of Asians?.  Pharmacoepidemiol Drug Saf  2002; 11:229-234.

388. Bode U, Seif SM, Levine AS: Studies on the antidiuretic effect of cyclophosphamide vasopressin release and sodium excretion.  Med Pediatr Oncol  1980; 8:295.

389. Culine S, Ghosn M, Droz J: Inappropriate antidiuretic hormone secretion induced by ifosfamide.  Eur J Cancer  1990; 26:922.

390. Subramanian D, Ayus JC: Case report: Severe symptomatic hyponatremia associated with lisinopril therapy.  Am J Med Sci  1992; 303:177.

391. Castrillon JL, Mediavilla A, Mendez MA, et al: Syndrome of inappropriate antidiuretic hormone secretion (SIADH) and enalapril.  J Intern Med  1993; 233:89.

392. Gonzalex-Martinez H, Gaspard JJ, Espino DV: Hyponatremia due to enalapril in an elderly patient. A case report.  Arch Fam Med  1993; 2:791.

393. Aslam MK, Gnaim C, Kutnick J, et al: Syndrome of inappropriate antidiuretic hormone secretion induced by amiodarone therapy.  Pacing Clin Electrophysiol  2004; 27:831-832.

394. Schwartz WB, Bennett W, Curelop S, Bartter FC: A syndrome of renal sodium loss and hyponatremia probably resulting from inappropriate secretion of antidiuretic hormone.  Am J Med  1957; 23:529.

395. Bartter FE, Schwartz WB: The syndrome of inappropriate secretion of antidiuretic hormone.  Am J Med  1967; 42:790.

396. Thaler S, Teitelbaum I, Berl T: “Beer potamania” in beer drinkers. Effect of low dietary solute intake.  Am J Kidney Dis  1998; 31:1028.

397. Passamonte PM: Hypouricemia, inappropriate secretion of antidiuretic hormone, and small cell carcinoma of the lung.  Arch Intern Med  1984; 144:1569.

398. DiGiovanni SR, Nielsen S, Christensen E, et al: Regulation of collection duct water channel expression by vasopressin in Brattleboro rat.  Proc Natl Acad Sci U S A  1994; 91:8984.

399. Saito T, Ishikawa S, Ando F, et al: Exaggerated urinary excretion of aquaporin 2 in the pathological state of impaired water excretion dependent upon arginine vasopressin.  J Clin Endocrinol Metab  1998; 83:4043.

400. Leaf A, Bartter FC, Santos RF, Woony O: Evidence in man that urine electrolyte loss induced by Pitressin is a function of water retention.  J Clin Invest  1953; 32:868.

401. Verbalis J: An experimental model of syndrome of inappropriate antidiuretic hormone secretion in the rat.  Am J Physiol  1984; 247:E540.

402. Verbalis JG, Drutarosky M: Adaptation to chronic hypoosmolality in rats.  Kidney Int  1988; 34:351.

403. Verbalis JG: Pathogenesis of hyponatremia in an experimental model of inappropriate antidiuresis.  Am J Physiol  1994; 267:R1617.

404. Nelson PB, Seif SM, Maroon JC, Robinson AG: Hyponatremia in intracranial disease: Perhaps not the syndrome of inappropriate secretion of antidiuretic hormone.  J Neurosurg  1991; 55:938.

405. Diringer MN, Lim JS, Kirsch JR, Hanley DF: Suprasellar and intraventricular blood predicts elevated plasma atrial natriuretic factor in subarachnoid hemorrhage.  Stroke  1991; 22:577.

406. Anderson RJ: Arginine vasopressin escape. vivo and in vitro studies.   In: Cowley AW, Liard JK, Ausiello DA, ed. Vasopressin: Cellular and Integrative Function,  New York: Raven Press; 1988:215.

407. Eggena P, Ma CL: Downregulation of vasopressin receptors in toad bladder.  Am J Physiol  1986; 250:C453.

408. Tian Y, Sandberg K, Murase T, et al: VasopressinRx receptor binding is downregulated during renal escape from vasopressin antidiuresis.  Endocrinology  2000; 141:307.

409. Ecelbarger C, Chou C, Lee A, et al: Escape from vasopressin-induced antidiuresis: Role of vasopressin resistance of the collecting duct.  Am J Physiol  1998; 274:F1161.

410. Ecelbarger C, Nielsen S, Olson BR, et al: Role of renal aquaporins in escape from vasopressin antidiuresis in rat.  J Clin Invest  1997; 99:1852.

411. Ecelbarger C, Verbalis J, Knepper M: Increased abundance of distal sodium transporters in rat kidney during vasopressin escape.  J Am Soc Nephrol  2001; 12:207.

412. Berl T, Schrier RW: Disorders of water metabolism.   In: Schrier RW, ed. Renal and Electrolyte Disorders,  6th ed. Philadelphia: Lippincott, Williams & Wilkins; 2003:1.

413. Tang WW, Kaptein EM, Feinstein EI, Massry SG: Hyponatremia in hospitalized patients with the acquired immunodeficiency syndrome and the AIDS related complex.  Am J Med  1993; 94:169.

414. Putterman C, Levy L, Rubinger D: Transient exercise induced water intoxication and rhabdomyolysis.  Am J Kidney Dis  1993; 21:206.

415. Irving RA, Noakes TD, Buck R, et al: Evaluation of renal function and fluid homeostasis during recovery from exercise induced hyponatremia.  J Appl Physiol  1991; 70:342.

416. Miller M, Hecker MS, Friedlander DA, et al: Apparent idiopathic hyponatremia in an ambulatory geriatric population.  J Am Geriatr Soc  1996; 44:404-408.

417. Anpalahan M: Chronic idiopathic hyponatremia in older people due to the syndrome of inappropriate antidiuretic hormone secretion (SIADH) possibly related to aging.  J Am Geriatr Soc  2001; 49:788-792.

418. Hirshberg B, Ben-Yehuda A: The syndrome of inappropriate antidiuretic hormone secretion in the elderly.  Am J Med  1997; 103:270-273.

419. Arinzon Z, Feldman J, Jarchowsky J, et al: A comparative study of the syndrome of inappropriate antidiuretic hormone secretion in community-dwelling patients and nursing home residents.  Aging Clin Exp Res  2003; 15:6-11.

420. Ishikawa SE, Saito T, Fukagawa A, et al: Close association of urinary excretion of aquaporin-2 with appropriate and inappropriate arginine vasopressin-dependent antidiuresis in hyponatremia in elderly subjects.  J Clin Endocrinol Metab  2001; 86:1665-1671.

421. List AF, Hainsworth JD, Davis BW, et al: The syndrome of inappropriate secretion of antidiuretic hormone (SIADH) in small cell lung cancer.  J Clin Oncol  1986; 4:1191.

422. Coyle S, Penney MD, Masters PW, Walker BE: Early diagnosis of ectopic arginine vasopressinRx secretion.  Clin Chem  1993; 39:152.

423. Hussain SF, Irfan M, Abbasi M, et al: Clinical characteristics of 110 miliary tuberculosis patients from a low HIV prevalence country.  Int J Tuberc Lung Dis  2004; 8:493-499.

424. Feldman BJ, Rosenthal SM, Vargas GA, et al: Nephrogenic syndrome of inappropriate antidiuresis.  N Engl J Med  2005; 352:1884-1890.

425. Zerbe R, Stropes L, Robertson G: Vasopressin function in the syndrome of inappropriate antidiuresis.  Annu Rev Med  1980; 31:315.

426. Renneboog B, Musch W, Vandemergel X, et al: Mild chronic hyponatremia is associated with falls, unsteadiness, and attention deficits.  Am J Med  2006; 119:e71-e78.

427. Fraser CL, Arieff AI: Epidemiology, pathophysiology, and management of hyponatremic encephalopathy.  Am J Med  1997; 102:67.

428. Verbalis JG: SIADH and other hypoosmolar disorders.   In: Schrier RW, ed. Diseases of the Kidney and Urinary Tract,  Philadelphia: Lippincott Williams & Wilkins; 2007:2219-2248.

429. Ayus JC, Varon J, Arieff AI: Hyponatremia, cerebral edema, and noncardiogenic pulmonary edema in marathon runners.  Ann Intern Med  2000; 132:711-714.

430. Manley GT, Fujimura M, Ma T, et al: Aquaporin-4 deletion in mice reduces brain edema after acute water intoxication and ischemic stroke.  Nat Med  2000; 6:159-163.

431. Nzerue CM, Baffoe-Bonnie H, You W, et al: Predictors of outcome in hospitalized patients with severe hyponatremia.  J Natl Med Assoc  2003; 95:335-343.

432. Wijdicks EF, Larson TS: Absence of postoperative hyponatremia syndrome in young, healthy females.  Ann Neurol  1994; 35:626.

433. Baran D, Hutchinson TA: The outcome of hyponatremia in a general hospital population.  Clin Nephrol  1984; 22:72.

434. Goldberg A, Hammerman H, Petcherski S, et al: Prognostic importance of hyponatremia in acute ST-elevation myocardial infarction.  Am J Med  2004; 117:242-248.

435. Lee DS, Austin PC, Rouleau JL, et al: Predicting mortality among patients hospitalized for heart failure: Derivation and validation of a clinical model.  JAMA  2003; 290:2581-2587.

436. Ruf AE, Dremers WK, Chavez LL, et al: Addition of serum sodium into the MELD score predicts waiting list mortality better than MELD alone.  Liver Transpl  2005; 11:336-343.

437. Heuman DM, Abou-Assi SG, Habib A, et al: Persistent ascites and low serum sodium identify patients with cirrhosis and low MELD scores who are at high risk for early death.  Hepatology  2004; 40:802-810.

438. Sterns RH: Severe symptomatic hyponatremia: Treatment and outcome.  Ann Intern Med  1987; 107:656.

439. Berl T: Treating hyponatremia. What is all the controversy about?.  Ann Intern Med  1990; 113:417.

440. McJunkin JE, de los Reyes EC, Irazuzta JE, et al: La Crosse encephalitis in children.  N Engl J Med  2001; 34:801-807.

441. Tierney WM, Martin DK, Greenlee MC, et al: The prognosis of hyponatremia in hospital admission.  J Gen Intern Med  1986; 1:380.

442. Sterns RH: The treatment of hyponatremia.  Am J Med  1990; 88:557.

443. Pasantes-Morales H, Franco R, Ordaz B, et al: Mechanisms counteracting swelling in brain cells during hyponatremia.  Arch Med Res  2002; 33:237-244.

444. Lien YH, Shapiro UI, Chan L: Study of brain electrolytes and organic osmolytes during correction of chronic hyponatremia. Implication for the pathogenesis of central pontine myelinolysis.  J Clin Invest  1991; 88:303.

445. Verbalis JG, Gullans SR: Hyponatremia causes large sustained reductions in brain content of multiple organic osmolytes in rats.  Brain Res  1991; 567:274.

446. Verbalis JG, Gullans SR: Rapid correction of hyponatremia produces differential effects on brain osmolyte and electrolyte reaccumulation in rats.  Brain Res  1993; 606:19.

447. Berl T: Treating hyponatremia. Are we damned if we do and damned if we don't?.  Kidney Int  1990; 37:1008.

448. Soupart A, Silver S, Schroeder B, et al: Rapid (24 hour) reaccumulation of brain organic osmolytes (particularly myo-inositol) in azotemic rats after correction of chronic hyponatremia.  J Am Soc Nephrol  2002; 13:1433-1441.

449. Thurman J, Halterman R, Berl T: Treatment of dysnatremic disorders in therapy in nephrology and hypertension.   In: Brady H, Wilcox C, ed. Therapy in Nephrology and Hypertension,  Philadelphia: WB Saunders; 2003.

450. Gross P: Treatment of severe hyponatremia.  Kidney Int  2001; 60:2417-2427.

451. Sterns RH, Riggs JE, Schochet SS: Osmotic demyelination syndrome following correction of hyponatremia.  N Engl J Med  1986; 314:1535.

452. Oh MS, Kim HJ, Carroll HJ: Recommendation for treatment of symptomatic hyponatremia.  Nephrology  1995; 70:143.

453. Ayus JC, Krothapalli RK, Arieff AI: Treatment of symptomatic hyponatremia and its relation to brain damage.  N Engl J Med  1987; 317:1190.

454. Ayus JC, Arieff AI: Chronic hyponatremic encephalopathy in postmenopausal women: Association of therapies with morbidity and mortality.  JAMA  1999; 281:2299-2304.

455. Ashouri OS: Severe diuretic induced hyponatremia in the elderly: A series of eight patients.  Arch Intern Med  1986; 146:1355.

456. Vexler Z, Ayus C, Roberts T, et al: Hypoxic and ischemic hypoxia exacerbates injury associated with metabolic encephalopathy in laboratory animals.  J Clin Invest  1994; 93:256.

457. Fraser CL, Sarnacki P: Na+-K+-ATPase pump function in male rat brain synaptosomes is different from that of females.  Am J Physiol  1989; 257:E284.

458. Dickinson LD, Betz AL: Attenuated development of ischemic brain edema in vasopressin deficient rats.  J Cereb Blood Flow Metab  1992; 12:681.

459. Rosenberg GA, Scremin O, Estrada E, Kyner WT: Arginine vasopressin V1 antagonist and atrial natriuretic peptide reduce hemorrhagic brain edema in rats.  Stroke  1992; 23:1767.

460. Ayus JC, Arieff AI: Pathogenesis and prevention of hyponatremic encephalopathy.  Endocrinol Metab Clin North Am  1993; 22:425.

461. Ayus JC, Krothapalli RK, Armstrong DL: Rapid correction of severe hyponatremia in the rat: Histopathological changes in the brain.  Am J Physiol  1985; 248:F711.

462. Martin RJ: Central pontine and extrapontine myelinolysis: The osmotic demyelination syndromes.  J Neurol Neurosurg Psychiatry  2004; 75(suppl 3):iii22-iii28.

463. Tan H, Onbas O: Central pontine myelinolysis central pontine myelinolysis manifesting with massive myoclonus.  Pediatr Neurol  2004; 31:64-66.

464. Menger H, Jorg J: Outcome of central pontine and extrapontine myelinolysis.  J Neurol  1999; 246:700-705.

465. Ruzek KA, Campeau NG, Miller GM: Early diagnosis of central pontine myelinolysis with diffusion-weighted imaging.  AJNR Am J Neuroradiol  2004; 25:210-213.

466. Soupart A, Penninck R, Stenuit R, et al: Treatment of chronic hyponatremia in rats by intravenous saline: Comparison of rate versus magnitude of correction.  Kidney Int  1992; 41:1662.

467. Verbalis JG, Martinez AJ: Neurological and neuropathological sequelae of correction of chronic hyponatremia.  Kidney Int  1991; 39:1274.

468. Sterns RH: Severe hyponatremia. The case for conservative management.  Crit Care Med  1992; 20:534.

469. Barsoum NR, Levine BS: Current prescriptions for the corrections of hyponatremia and hypernatremia: Are they too simple?.  Nephrol Dial Transplant  2002; 17:1176-1180.

470. Carlotti AP, Bohn D, Mallie JP, Halperin ML: Tonicity balance, and not electrolyte-free water calculations, more accurately guides therapy for acute changes in natremia.  Intensive Care Med  2001; 27:921-924.

471. Soupart A, Pennirek U, Creniar L, et al: Prevention of brain demyelination in rats after excessive correction of chronic hyponatremia by serum sodium lowering.  Kidney Int  1994; 45:193.

472. Goldszmidt MA, Iliescu EA: DDAVP to prevent rapid correction in hyponatremia.  Clin Nephrol  2000; 53:226-229.

473. Oya S, Tsutsumi K, Ueki K, Kirino T: Reinduction of hyponatremia to treat central pontine myelinolysis.  Neurology  2001; 57:1931-1932.

474. Furst H, Hallows KR, Post J, et al: The urine/plasma electrolyte ratio: A predictive guide to water restriction.  Am J Med Sci  2000; 319:240-244.

475. Forrest JN, Cox M, Hong C, et al: Superiority of demeclocycline over lithium in the treatment of chronic syndrome of inappropriate secretion of antidiuretic hormone.  N Engl J Med  1978; 298:173.

476. Decaux G, Genette F: Urea for long term treatment of syndrome of inappropriate secretion of antidiuretic hormone.  BMJ  1981; 283:1081.

477. Decaux G, Waterlot Y, Genette F, Mockel J: Treatment of the syndrome of inappropriate secretion of antidiuretic hormone with furosemide.  N Engl J Med  1981; 304:329.

478. Greenberg A, Verbalis JG: Vasopressin receptor antagonists.  Kidney Int  2006; 69:2124-2130.

479. Kim JK, Dillingham MD, Summer SN, et al: Effects of vasopressin antagonist on the cellular action of vasopressin binding, adenylate cyclase activation and water flux.  J Clin Invest  1985; 76:1530.

480. Macion-Dazard R, Callahan N, Xu Z, et al: Mapping the binding site of six nonpeptide antagonists to the human V2-renal vasopressin receptor.  J Pharmacol Exp Ther  2006; 316:564-571.

481. Ohnishi A, Orita Y, Okahara R: Potent aquaretic agent: A novel nonpeptide selective vasopressin antagonist OPC 31260 in man.  J Clin Invest  1993; 92:2653.

482. Fujisawa G, Ishikawa S, Okada K, Saito T: Therapeutic efficacy of non-peptide ADH antagonist OPC-31260 in SIADH rats.  Kidney Int  1993; 44:19.

483. Saito T, Ishikawa S, Abe K: Acute aquaresis by the nonpeptide arginine vasopressin antagonist OPC 31260 improves hyponatremia in patients with the syndrome of inappropriate secretion of antidiuretic hormone.  J Clin Endocrinol Metab  1997; 82:1054.

484. Schrier RW, Gross P, Gheorghiade M, et al: Tolvaptan, a selective oral vasopressin V2-receptor antagonist, for hyponatremia.  N Engl J Med  2006; 355:2099-2112.

485. Decaux G: Difference in solute excretion during correction of hyponatremic patients with cirrhosis or syndrome of inappropriate secretion of antidiuretic hormone by oral vasopressin V2 receptor antagonist VPA-985.  J Lab Clin Med  2001; 138:18-21.

486. Decaux G: Long-term treatment of patients with inappropriate secretion of antidiuretic hormone by the vasopressin receptor antagonist conivaptan, urea, or furosemide.  Am J Med  2001; 110:582-584.

487. Guyader D, Patat A, Ellis-Grosse EJ, Orczyk GP: Pharmacodynamic effects of a nonpeptide antidiuretic hormone V2 antagonist in cirrhotic patients with ascites.  Hepatology  2002; 36(5):1197-1205.

488. Gerbes AL, Gulberg V, Gines P, et al: Therapy of hyponatremia in cirrhosis with a vasopressin receptor antagonist: A randomized double blind multicenter trial.  Gastroenterology  2003; 124:933-939.

489. Abraham WT, Shamshirsaz AA, McFann K, et al: Aquaretic effect of lixivaptan, an oral, non-peptide, selective V2 receptor vasopressin antagonist, in New York Heart Association functional class II and III chronic heart failure patients.  J Am Coll Cardiol  2006; 47:1615-1621.

490. Gheorghiade M, Niazi I, Ouyang J, et al: Vasopressin V2-receptor blockade with tolvaptan in patients with chronic heart failure: Results from a double-blind, randomized trial.  Circulation  2003; 107:2690-2696.

491. Gheorghiade M, Gattis WA, O'Connor CM, et al: Effects of tolvaptan, a vasopressin antagonist, in patients hospitalized with worsening heart failure: A randomized controlled trial.  JAMA  2004; 291:1963-1971.

492. Gheorghiade M, Konstam MA, Burnett Jr JC, et al: Short-term clinical effects of tolvaptan, an oral vasopressin antagonist, in patients hospitalized for heart failure: the EVEREST Clinical Status Trials.  JAMA  2007; 297:1332-1343.

493. Konstam MA, Gheorghiade M, Burnett Jr JC, et al: Effects of oral tolvaptan in patients hospitalized for worsening heart failure: the EVEREST Outcome Trial.  JAMA  2007; 297:1319-1331.