12
CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004 639 YPONATREMIA (a low plasma sodium con- centration) is among the most common laboratory abnormalities that physicians encounter, yet its causes are among the more complex in medicine. In a study of more than 2,800 hospitalized patients, 15% had plasma sodium concentra- tions below the lower end of the reference range of 134 mmol/L. 1 Almost 5% had levels less than 125 mmol/L. In some instances, hyponatremia repre- sents a true medical emergency, requiring immediate intervention to prevent cerebral edema. More commonly, hyponatremia may reflect the severity of a causative underlying illness and be a marker of morbidity and mor- tality. For example, the plasma sodium con- centration is a powerful predictor of severity of heart failure 2 and mortality in hepatic cirrho- sis. 3 This paper reviews the pathophysiology and diagnostic evaluation of hyponatremia. Armed with a sound understanding of body fluid physiology, clinicians should be profi- cient in treating this common disorder and avoid the potentially lethal complications of mistreatment. A LITTLE PHYSIOLOGY The plasma sodium concentration is the main determinant of plasma osmolality and there- fore of the total water balance among the body’s fluid compartments. Perturbations of plasma sodium concentration reflect an under- lying disorder involving water homeostasis. The challenge for the physician is to understand the disorder that is causing the hyponatremia. The mechanism, etiology, man- BENJAMIN J. FREDA, DO Department of Nephrology and Hypertension, The Cleveland Clinic Foundation Evaluation of hyponatremia: A little physiology goes a long way REVIEW ABSTRACT Hyponatremia is common in hospitalized patients. By taking a careful and logical approach, one can promptly recognize the causative factor or factors in nearly all cases. Most cases of hyponatremia are due to impaired renal water excretion, and recognizing the cause and pathophysiologic process makes it possible to provide focused individualized care and avoid mistreatment. KEY POINTS In evaluating hyponatremia, the clinician should determine the chronicity or acuity of the hyponatremia, confirm that the patient has true plasma hypo-osmolality, determine the extracellular fluid volume status, assess laboratory tests and urine sodium values, and assess any exogenous free water intake. In heart failure and hepatic cirrhosis, effective circulating blood volume is low, triggering release of antidiuretic hormone and water retention regardless of plasma osmolality. If hyponatremia is severe (plasma sodium < 120 mmol/L), quickly examine the patient for signs or symptoms of acute neurologic changes (seizures, altered mental status, or focal neurologic signs) that may signal the need for immediate therapy with hypertonic intravenous saline and furosemide. Plasma glucose should be measured in all cases of hyponatremia, and the plasma sodium concentration should be corrected for hyperglycemia. H MICHAEL B. DAVIDSON, DO Department of Internal Medicine, The Cleveland Clinic Foundation PHILLIP M. HALL, MD Staff Consultant, Department of Nephrology and Hypertension, The Cleveland Clinic Foundation

Evaluation of hyponatremia: A little physiology goes a long way

Embed Size (px)

Citation preview

CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004 639

YPONATREMIA (a low plasma sodium con-centration) is among the most common

laboratory abnormalities that physiciansencounter, yet its causes are among the morecomplex in medicine.

In a study of more than 2,800 hospitalizedpatients, 15% had plasma sodium concentra-tions below the lower end of the referencerange of 134 mmol/L.1 Almost 5% had levelsless than 125 mmol/L.

In some instances, hyponatremia repre-sents a true medical emergency, requiringimmediate intervention to prevent cerebraledema. More commonly, hyponatremia mayreflect the severity of a causative underlyingillness and be a marker of morbidity and mor-tality. For example, the plasma sodium con-centration is a powerful predictor of severity ofheart failure2 and mortality in hepatic cirrho-sis.3

This paper reviews the pathophysiologyand diagnostic evaluation of hyponatremia.Armed with a sound understanding of bodyfluid physiology, clinicians should be profi-cient in treating this common disorder andavoid the potentially lethal complications ofmistreatment.

■ A LITTLE PHYSIOLOGY

The plasma sodium concentration is the maindeterminant of plasma osmolality and there-fore of the total water balance among thebody’s fluid compartments. Perturbations ofplasma sodium concentration reflect an under-lying disorder involving water homeostasis.

The challenge for the physician is tounderstand the disorder that is causing thehyponatremia. The mechanism, etiology, man-

BENJAMIN J. FREDA, DODepartment of Nephrology and Hypertension,The Cleveland Clinic Foundation

Evaluation of hyponatremia:A little physiology goes a long way

REVIEW

■ ABSTRACT

Hyponatremia is common in hospitalized patients. Bytaking a careful and logical approach, one can promptlyrecognize the causative factor or factors in nearly allcases. Most cases of hyponatremia are due to impairedrenal water excretion, and recognizing the cause andpathophysiologic process makes it possible to providefocused individualized care and avoid mistreatment.

■ KEY POINTS

In evaluating hyponatremia, the clinician shoulddetermine the chronicity or acuity of the hyponatremia,confirm that the patient has true plasma hypo-osmolality,determine the extracellular fluid volume status, assesslaboratory tests and urine sodium values, and assess anyexogenous free water intake.

In heart failure and hepatic cirrhosis, effective circulatingblood volume is low, triggering release of antidiuretichormone and water retention regardless of plasmaosmolality.

If hyponatremia is severe (plasma sodium < 120 mmol/L),quickly examine the patient for signs or symptoms ofacute neurologic changes (seizures, altered mental status,or focal neurologic signs) that may signal the need forimmediate therapy with hypertonic intravenous salineand furosemide.

Plasma glucose should be measured in all cases ofhyponatremia, and the plasma sodium concentrationshould be corrected for hyperglycemia.

H

MICHAEL B. DAVIDSON, DODepartment of Internal Medicine, The ClevelandClinic Foundation

PHILLIP M. HALL, MDStaff Consultant, Department of Nephrology andHypertension, The Cleveland Clinic Foundation

640 CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004

agement, and complications of hyponatremiaare more easily understood if one considers thephysiology of water regulation by the kidneyand neuroendocrine axis.

Hyponatremia is most commonly dueto decreased renal clearance of waterHyponatremia can occur when there is eithersodium loss or, more commonly, water reten-tion.

Movement of water from cells to theblood can also result in hyponatremia byincreasing the amount of plasma water rela-tive to plasma sodium (see Most body water iswithin cells, on this page).

Total body water can increase as a result ofincreased intake or decreased renal clearance.Hyponatremia due to increased water intakehappens rarely in patients with normal renalfunction because one would have to drinkmore than 20 L/day to overcome the kidneys’capacity to excrete free water.

Thus, in almost all cases of hyponatremia,the problem lies in an impaired ability of thekidneys to excrete free water due to the actionof antidiuretic hormone (ADH, vasopressin).ADH plays a central role in most cases ofrenal-mediated water retention. ADH activitymay be increased appropriately (eg, in hypo-volemia) or inappropriately (eg, in cancer or adrug effect).

How the kidneys regulate water balanceThe fluid that is filtered from the glomerulushas the same concentration of electrolytes asthe plasma. For electrolyte-free water to be

excreted, this ultrafiltrate must be delivered tothe diluting segment of the kidney. Thisrequires an adequate glomerular filtration rateand an intact thick ascending limb of the loopof Henle. In the loop of Henle, salt is reab-sorbed from the ultrafiltrate back into theplasma without water. What remains is a rela-tively dilute (hypotonic) fluid that proceedsinto the distal nephron.

This dilute fluid, made up mostly of water,will be excreted in the urine if it is not reab-sorbed in the collecting ducts. The reabsorp-tion of water depends on the presence andactivity of ADH.

ADH secretion is directly tied to osmolalityIn a healthy patient, the ADH level is sensi-tive to both the plasma osmolality (deter-mined primarily by the plasma sodium con-centration) and the volume status.

As the volume of plasma water increases,the sodium concentration decreases, and thusso does the plasma osmolality. This decrease issensed by osmoreceptors in the hypothalamusthat regulate ADH secretion. These cells aresensitive to changes in plasma osmolality of aslittle as 1% above or below normal values, andthey respond to decreased osmolality by sig-nalling the pituitary to inhibit ADH release.

In the absence of ADH, normally func-tioning kidneys excrete more electrolyte-freewater, restoring plasma osmolality and theplasma sodium concentration to normal lev-els. At a plasma osmolality of less than 275mOsm/kg H2O, ADH secretion ceases.5Conversely, as osmolality increases, ADH

Total bodywater = 60% ofbody weight(men) or 50%(women)

HYPONATREMIA FREDA AND COLLEAGUES

lthough the water content of humansdepends on several factors such as age, gen-

der, and amount of adipose tissue, total body watercan be estimated as 60% of body weight in menand 50% of body weight in women. Older personsand those with more adipose tissue have less leanmuscle mass and therefore less overall body waterthan determined from the above estimate.

The total body water is divided between theintracellular space (40% of body weight) andextracellular space (20% of body weight). Roughly

one third of the extracellular volume is intravas-cular (plasma volume) and two thirds is intersti-tial. Of the total plasma volume (approximately4.6 L in a 70-kg man), 85% is in the venous circu-lation and 15% is in the arterial circulation.

Thus, a 70-kg person has an arterial volume ofroughly 700 cc.4 As we will discuss, this seeminglysmall volume is perhaps the most critical determi-nant of the “fullness” of the body’s circulation thatis critical to the regulation of the body’s sodiumand fluid levels.

A

Most body water is within cells

secretion increases. In a normal kidney, thisresults in the reabsorption of free water andthe reduction of osmolality.

Decreased effective circulating volume:Why patients retain waterin heart failure and hepatic cirrhosisIn some diseases, the normal mechanisms ofwater homeostasis go awry, leading the kidneyto reabsorb water even though the patientappears fluid-overloaded. As a result of waterretention, the plasma osmolality decreases,and hyponatremia ensues.

Patients with heart failure or hepatic cir-rhosis may continue to retain sodium andwater, even though they may have a clinicallyapparent excess of extracellular fluid volume(ie, edema or ascites).

Diminished cardiac output (in heart fail-ure) and peripheral vasodilation (in hepaticcirrhosis) cause baroreceptors in the renalafferent arterioles and the carotid sinus tosense decreased pressure and thereforedecreased “circulatory fullness.” Circulatoryfullness is a combination of cardiac output,intravascular blood volume, and arterial vas-

cular tone, which is sensed by carotid andafferent arteriolar baroreceptors. In anattempt to restore perfusion pressure to thetissues of the body, these baroreceptors signalthe posterior pituitary to release ADH, result-ing in free water reabsorption and hypona-tremia, even if the plasma is already dilute.6The volume of extracellular fluid needed tomaintain perfusion pressure and avoid stimu-lation of ADH release has been called theeffective circulating blood volume.6

Other mechanisms resulting in water andsodium retention in patients with edematousstates are outlined in FIGURE 1.

Thus, the presence of hyponatremia mayindicate that underlying heart failure or cir-rhosis is severe enough to limit tissue perfu-sion as sensed by the kidney and carotidbaroreceptors.

ADH secretioncan be independent of osmolalityAlthough the prime regulator of ADH secre-tion is osmolality, there are also “osmolality-independent” stimulants of ADH secretion.These include:

CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004 641

Hyponatremiais usually dueto impairedclearance ofwater

BrainIncreased antidiuretic hormonerelease

Advanced cirrhosisArterial vasodilation

Reduced effective circulatingblood volume(decreased tissue perfusion)

Physiologic responses

Continued water intake

Clinical consequencesIncreased plasma free water

and hyponatremiaSodium avidity and low urine sodiumIncreased extracellular fluid volume

(edema)Inability to excrete free water

KidneyDecreased glomerular filtrationIncreased proximal tubular

sodium reabsorptionIncreased proximal tubular

water reabsorptionResultant secondary

hyperaldosteronism

Heart failureDecreased cardiac output

FIGURE 1

How heart failure and cirrhosis cause hyponatremia

642 CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004

• Hypovolemia (as described above).Hypovolemia decreases the osmoticthreshold at which ADH is secreted: dur-ing hypovolemia, ADH secretion will per-sist even if plasma osmolality is less than275 mOsm/kg H2O.

• Postoperative nausea, pain, and stressmay also stimulate ADH secretion.

• Many neoplasms and pulmonary and cen-tral nervous system disease processes canbe powerful stimulants of ADH secretionor even produce ectopic ADH.

• Drugs that stimulate ADH secretioninclude but are not limited to selectiveserotonin reuptake inhibitors, haloperidol,vincristine, cyclophosphamide, and carba-mazepine. Nonsteroidal anti-inflammatorydrugs can also contribute to hyponatremiaby reducing renal prostaglandins, whichnormally act to antagonize the free-waterreabsorption mediated by ADH.5

Water can also move in and out of cellsIn addition to changes in total body water andsodium, we need to consider movement of waterand electrolytes between fluid compartments.Key to understanding this movement is the con-cept of effective and ineffective osmoles.

Water moves between body fluid compart-ments primarily as a result of osmotic forces. Asolute that induces movement of waterbetween fluid compartments is an “effective”osmole. An “ineffective” osmole can readily

cross from one fluid compartment to another,thereby limiting its ability to incite movementof water.

Sodium, which resides primarily in theextracellular fluid due to the action of thesodium-potassium ATPase pump, is an effec-tive osmole—it tends to hold water in theextracellular fluid. Similarly, potassium acts asthe major intracellular effective osmole, act-ing to hold water within the cell.

In contrast, urea is an ineffective osmolebecause it readily crosses cell membranes andhas similar concentrations in the intracellularand extracellular fluid. Thus, although it con-tributes to measured plasma osmolality as ameasured solute, it does not generate anosmotic force to move water between fluidcompartments.

Hyperglycemia can cause hyponatremiaGlucose acts as an effective osmole in hyper-glycemic states by inducing water movementfrom the intracellular to the extracellular fluidspace. Plasma water is increased and hypona-tremia develops as a result of dilution of plasmasodium. Plasma osmolality is actually increasedin severe cases of hyperglycemia, owing tolarge amounts of glucose in the plasma.Therefore, the patient is generally not at riskfor the morbidity (eg, cerebral edema) associ-ated with truly hypo-osmolar hyponatremia.

Although the effect of hyperglycemia onplasma sodium levels is dependent on addi-tional factors (eg, water loss via a glucose-mediated osmotic diuresis), normalization ofglucose levels generally leads to resolution ofhyponatremia.

Hypokalemia can contributeto hyponatremiaLarge losses of total body potassium, as in pro-longed vomiting or diarrhea, resulting in severehypokalemia, will cause potassium to migratefrom the cells to the plasma. This can affect theplasma sodium concentration because sodiumwill then move into cells to maintain elec-troneutrality. In this way, hypokalemia can initself cause a relative hyponatremia as sodiumconcentration in the plasma decreases whilethe concentration in the cells increases.

Thus, one must consider the effect of cor-recting any concomitant hypokalemia on the

Sodium andglucose act aseffectiveosmoles; ureadoes not

Initial evaluation of hyponatremiaExamine the patient for signs of acute neurologic changes

Confirm that the blood draw was accurate (verify result)

Clinically determine his or her extracellular fluid volume status

Rule out hyperglycemia

Determine rate of development of hyponatremia (hours to weeks)

Review all recent and current intravenous fluid orders

Review all intravenous medications for free water content(eg, antibiotics mixed with dextrose and water)

Review all medications (especially use of diuretics)

Confirm true hypo-osmolality (see text and TABLE 3)

Review and order appropriate laboratory tests (see TABLE 2)

T A B L E 1

HYPONATREMIA FREDA AND COLLEAGUES

overall rate of correction of hyponatremia. Inpractical terms, this means that giving normalsaline with added potassium to a patient withhyponatremia and hypokalemia will have agreater effect on increasing the plasma sodiumlevel than giving normal saline by itself.

■ INITIAL EVALUATION:IS IT REAL? IS IT SERIOUS?

TABLE 1 lists the clinical information necessaryto initially evaluate a patient with hypona-tremia.

CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004 643

Laboratory assessment of hyponatremiaTEST DEFINITION AND NORMAL VALUE COMMENTS

Serum osmolality Measured value: amount of Low value (< 275 mOsm/kg H2O)solute in serum water confirms true hypo-osmolality(275–290 mOsm/kg H2O) Osmolal gap < 10 rules out pseudo-Calculated value: hyponatremia or unmeasured solutes

(2 × serum sodium concentration) such as mannitol, glycine, or sorbitol+ (serum glucose concentration / 18) Measured serum osmolality is not+ (blood urea concentration / 2.8) affected by increased protein or lipids

Urine osmolality Amount of solute in urine Appropriate response to hyponatremiaNo “normal” value; depends on fluid is production of maximally diluteand solute intake, kidney function, urine (< 100 mOsm/kg H2O)and water balance Urine osmolality has limited valueRange is 50–1,200 mOsm/kg H2O, but because nearly all patients withis more narrow in older patients and hyponatremia do not make maximallyrenal impairment dilute urine except in psychogenicEstimate = 35 × last two digits of urine polydipsia or reset osmostatspecific gravity

Urine sodium Spot value from fresh urine See TABLE 5 for caveats inUsually < 20 (mEq/L) in low effective interpreting urine sodiumcirculating blood volume statesUsually > 20–40 in euvolemic patientswithout decreased effective circulatingblood volume

Plasma glucose Serum sodium will decrease Mandatory to measure plasma 1.6–2.4 mmol/L for every 100 mg/dL glucose in all cases of hyponatremiaincrease in glucose over 100 mg/dL

Tests of thyroid Thyroid-stimulating hormone (TSH), Useful:and adrenal function Free thyroxine (T4) • When there is a clinical suspicion

adrenocorticotropic hormone (ACTH), of thyroid or adrenal hypofunctionACTH stimulation test • To establish a diagnosis of SIADH*

(document normal thyroid andadrenal function)• When careful evaluationdoes not provide clear etiology

Serum uric acid Usual values: refer to individual Helpful in differentiating euvolemicand blood urea laboratory reference range SIADH from hypovolemicnitrogen (BUN) levels causes of hyponatremia

Low uric acid and BUN levelsare more compatible witheuvolemia and SIADH

*SIADH = syndrome of inappropriate antidiuretic hormone secretion

T A B L E 2

As plasmaosmolalitydecreases, sodoes ADHsecretion, andthe kidneysshould excretemore water

644 CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004

Is it real? With any laboratory abnormal-ity, the clinician must be careful to excludeerrors in collecting the blood sample. This isespecially relevant when an extremely lowsodium level is reported in a patient who isclinically well. In such cases the sodium levelshould be rechecked and verified to be col-lected from a vein that is not being simultane-ously infused with a hypotonic intravenousfluid. In patients with signs or symptoms con-sistent with hyponatremia, the laboratoryvalue should be taken as accurate until provenotherwise.

How quickly did the hyponatremiadevelop? This information is important, as itwill dictate the risk for neurologic morbidity(cerebral edema) and how quickly the plasmasodium concentration must be corrected. Aswill be discussed in a subsequent paper in thisjournal dealing with the treatment of hypona-tremia, overly rapid correction may lead toneuronal damage and central pontine myeli-nolysis.

Are there neurologic signs or symptoms?Patients with severe hyponatremia (< 120mmol/L) should be examined expeditiously forsigns of acute neurologic changes (seizures,altered mental status, focal neurologic signs)that may signal the need for immediate thera-py with hypertonic intravenous saline andfurosemide.

What medications is the patient receiv-ing? All medications should be reviewed, withcareful attention to diuretics (especially thi-azide-type) or medications associated withinappropriate ADH release (see above;

Drugs). All recent and current intravenousfluid orders should also be reviewed, and intra-venous medications (eg, antibiotics, cardio-vascular drips) should be scrutinized for theirfree water content, as many antibiotics arereconstituted in dextrose and free water.

Are additional laboratory tests needed?TABLE 2 lists laboratory tests that are useful dur-ing the evaluation of hyponatremia. As dis-cussed below, the clinician should be able toexplain the hyponatremia on the basis of theclinical history, medications, extracellularfluid volume status, and directed investiga-tions into the presence of comorbid disease ordiseases that predispose to hyponatremia.

Specific laboratory tests are useful in casesin which the etiology is still unclear after care-ful, logical, clinical evaluation. Also, addi-tional laboratory tests may be necessary whenit is difficult to clinically assess extracellularfluid volume.

Rule out hyperglycemiaAn important early step is to rule out hyper-glycemia, and every patient with hyponatrem-ia should have his or her plasma glucose mea-sured. For every 100 mg/dL increase in plasmaglucose, the plasma sodium concentration willdecrease by roughly 1.6 mEq/L, although arecent report7 suggests that the correction fac-tor should be closer to 2.4.

Calculate the osmolal gapto confirm true plasma hypo-osmolalityIn a few uncommon situations, hyponatremiadoes not reflect true underlying plasma hypo-

Thiazidediuretics aremore likely toinducehyponatremiathan are loopdiuretics

How to calculate plasma osmolality and the osmolal gap1 Measure osmolality directly by laboratory assessment

2 Calculate plasma osmolality:

(2 × plasma sodium concentration) + (plasma glucose concentration / 18)+ (blood urea nitrogen concentration / 2.8)

3 Calculate osmolal gap:

Measured value – calculated value*

*An osmolal gap > 10 mOsm/kg H2O suggests a nonsodium effective osmole (mannitol, glycine, or sorbitol) orpseudohyponatremia (severe hyperlipidemia or hyperproteinemia); it may also be seen with ineffective osmoles (eg,ethanol, methanol, ethylene glycol). However, these latter substances do not cause water shifts or hyponatremia.

T A B L E 3

HYPONATREMIA FREDA AND COLLEAGUES

osmolality. These cases of hyponatremia areusually related to the presence of effectiveosmoles other than sodium in the plasma orextreme states of hyperproteinemia or hyper-lipidemia.

Therefore, it is necessary to calculate theosmolal gap by subtracting the calculated plas-ma osmolality from the measured laboratoryvalue (TABLE 3). If the osmolal gap is more than10 mOsm/kg H2O, the clinician should sus-pect that other effective osmoles are presentin the plasma (see below).

Sodium-free irrigation fluids such asglycine or mannitol, used during transurethralresection of the prostate or hysteroscopicendometrial surgery, can be systemicallyabsorbed in large amounts, leading to severehyponatremia.8 In these cases the measuredplasma osmolality is normal but the osmolalgap is large, reflecting the mannitol or glycinein the blood.

Although the exact mechanism is notcompletely understood, mannitol and glycineare thought to have an osmotic effect, draw-ing water from the cells into the extracellularfluid (plasma). There is also a dilutional com-ponent later in the postoperative course asglycine shifts into the cells, leaving sodium-free water behind in the extracellular fluid.

Very high protein or lipid levels (usuallytriglycerides with lactescent serum) can causethe plasma sodium concentration to appearfalsely low if it is measured by older flow pho-tometry methods. These methods measure theconcentration of sodium in the whole plasma(including both solid and liquid phases), inwhich the amount of sodium in the sample ofwhole plasma is diluted by the elevated levelsof protein and lipid. In this situation, calledpseudohyponatremia, serum osmolality is normalas measured using direct ion-specific electrodes.The calculated plasma osmolality will be low,and there will be an elevated osmolal gap.9

Newer methods measure the concentra-tion of sodium in serum water independent ofthe solid phase of plasma, using ion-specificelectrodes,10 and the problem of pseudohy-ponatremia is much rarer now.

Urine osmolality offers little diagnostic aidEven though the urine osmolality is routinelyevaluated, it offers very little in narrowing the

differential diagnosis of hyponatremia.Almost all cases of hyponatremia are asso-

ciated with an improperly concentrated urine(> 100 mOsm/kg H2O). Less common causesof hyponatremia that are associated withappropriately dilute urine (urine osmolality <100 mOsm/kg) include psychogenic polydip-sia, in which patients drink excessive amountsof water, and a reset osmostat.

Although it is necessary to demonstratethat the urine is inappropriately concentratedto diagnose the syndrome of inappropriateantidiuretic hormone secretion (SIADH; seebelow), there is no “diagnostic” correlationbetween the osmolality of the urine and serumin this disorder.

■ NEXT, EVALUATE THE VOLUME STATUS

Once hyponatremia is verified and true hypo-osmolality is confirmed, the critical next stepis to determine the patient’s extracellular fluidvolume status. True hypo-osmolar hypona-tremia can be associated with euvolemia,hypervolemia, or hypovolemia.

While the exact cause of the hyponatrem-ia may not be immediately apparent (see TABLE 4

CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004 645

It is mandatoryto measureplasma glucosein all cases ofhyponatremia

Causes of hyponatremia based onextracellular fluid volume status

HypovolemicGastrointestinal solute loss (diarrhea, emesis)Third-spacing (ileus, pancreatitis)Diuretic useAddison diseaseSalt-wasting nephritis

EuvolemicSyndrome of inappropriate antidiuretic hormone secretion (SIADH)Diuretic useGlucocorticoid deficiencyHypothyroidismBeer-drinker’s potomania, psychogenic polydipsiaReset osmostat

Hypervolemic with decreased effectivecirculating blood volume

Decompensated heart failureAdvanced liver cirrhosisRenal failure with or without nephrosis

T A B L E 4

646 CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004

for major causes), assessing the extracellularfluid volume is critical in determining the prop-er initial treatment (ie, fluid restriction vs resus-citation with isotonic saline).

Careful evaluation of the history (diar-rhea, vomiting, thirst, and polyuria), nursingrecords (daily weights, intake and output),and physical examination (orthostasis, neckveins, peripheral edema, and ascites) shouldbe adequate to determine the patient’s extra-cellular fluid volume status.

Peripheral edema or ascites indicates ele-vated total body sodium and therefore elevat-ed extracellular fluid volume. On the otherhand, patients with circulatory compromise(orthostatic changes or supine hypotensionand tachycardia) and no edema have lowextracellular fluid volume. Other patients mayhave hypovolemia not detectable by initialexamination or may be truly euvolemic.

Although it is usually easy to determine ifa patient is hypervolemic, it may be harder todifferentiate euvolemia from subtle hypo-volemic states.11 Other data such as urinesodium, serum uric acid, and blood urea nitro-gen (BUN) levels and response to isotonicintravenous saline may be helpful in thesecases (see If the extracellular fluid volume isnormal, below).

If the extracellular fluid volume is elevatedMost hospitalized patients with hyponatremia

and increased extracellular fluid volume haveeither decompensated heart failure oradvanced hepatic cirrhosis.

Although edema or ascites always indicateelevated total body sodium, and therefore ele-vated extracellular fluid volume, total bodywater is elevated to an even greater degree inthese patients due to reduced effective circu-lating volume.

In the absence of diuretic use or othervariables affecting renal handling of sodium(TABLE 5), patients with hyponatremia, expand-ed extracellular fluid volume, and reducedeffective circulating blood volume are pro-foundly “sodium-avid” (their kidneys willretain sodium). This will be manifested by alow urine sodium level (< 10–20 mEq/L).

A urine sodium level greater than 20mEq/L in a patient with expanded extracellu-lar fluid volume and no recent diuretic use isconsistent with adequate effective circulatingvolume.

TABLE 5 reviews important caveats that mayaffect interpretation of urinary sodium valuesin patients with hyponatremia.

Severe renal impairment. The hallmarkof renal impairment is reduced functionalnephron mass and impaired glomerular filtra-tion. Smaller volumes of fluid are filtered, andif water intake exceeds the amount that can befiltered, water will be retained and hypona-tremia will develop. An acute worsening of

Caveats in interpreting spot urinary sodium levels

SETTING EFFECT ON COMMENTSURINE SODIUM

Diuretics Increase Urine sodium may be falsely elevated by diureticaction, but may still be low if there is severe concomitant hypovolemia

Osmotic diuresis Increase Solute diuresis (eg, glucosuria) obligates sodium andwater loss even in the setting of hypovolemia

Chronic kidney disease, Increase Sodium handling is altered due to renal disease,tubulointerstitial disease and urine sodium may be elevated even in the

presence of hypovolemia

Bicarbonaturia Increase Bicarbonate loss obligates sodium loss to balancenegative charge in urine (eg, proximal renal tubularacidosis and emesis-induced metabolic alkalosis)

T A B L E 5

HYPONATREMIA FREDA AND COLLEAGUES

In truehyponatremia,determine iffluid volumeis high, normal,or low

renal function may also impair free waterexcretion independent of changes in waterintake.

Patients with nephrotic syndrome maydevelop hyponatremia regardless of the pres-ence or degree of renal impairment. Loss ofplasma oncotic pressure due to hypoalbumin-emia favors a shift of fluid from the plasmainto the interstitial compartment. Thisreduces the effective circulating blood volumeand triggers ADH release. However, thismechanism may not apply in all patients withnephrotic syndrome who develop hypona-tremia.12

If the extracellular fluid volume is lowNonedematous patients with hyponatremiaand signs of extracellular fluid volume con-traction are losing sodium and water andreplacing it with hypotonic fluids. Solute lossand resulting extracellular fluid volume con-traction stimulate ADH release.

The sodium loss is nearly always hypoton-ic or isotonic to plasma; therefore, the devel-opment of hyponatremia is usually caused byreplacement of the fluid losses with intra-venous or oral free water.

Measure urine sodiumto determine the source of sodium lossThe evaluation of patients with hyponatrem-ia and decreased extracellular fluid volumebegins by determining the source of sodiumloss.

Sodium loss usually occurs via the gas-trointestinal tract or kidney. Common causesinclude diarrhea and emesis. Third-spacing offluid and sodium, seen in pancreatitis, burns,rhabdomyolysis, or ileus, may also lead to sodi-um loss and contracted extracellular fluid vol-ume.

The cause of sodium loss may not alwaysbe clinically obvious, however. In these cases,it is critical to measure urine sodium in a spotsample to evaluate the renal response (or con-tribution) to extracellular fluid volume con-traction.

If urine sodium is low. Urine sodiumshould be low (< 20 mEq/L) if extrarenal fac-tors (ie, gastrointestinal losses or third-spac-ing) account for solute loss. However, as out-lined in TABLE 5, renal sodium handling may be

dysfunctional in patients with chronic kidneydisease or tubulointerstitial disease. In thesecircumstances the kidney may not be able toavidly retain sodium even in hypovolemicstates.

If urine sodium is high or normal inpatients with clinically suspected extracellularfluid volume contraction, then the patient iseither taking diuretics or the kidney is wastingsodium.

Diuretic use is the most common cause ofhypovolemic hyponatremia, which in mostcases occurs within the first 2 weeks of thera-py.13 A combination of hypokalemia, con-tracted extracellular fluid volume, andimpaired urinary dilution contribute to itsoccurrence. Clinicians must be careful tointerpret the urine sodium excretion in thecontext of the timing of the last diuretic dose.

Patients with hyponatremia who are tak-ing diuretics can also present with hyper-volemia or euvolemia, depending on sodiumintake and other underlying diseases (eg,heart failure).

Thiazide diuretics are more likely toinduce hyponatremia than are loop diureticsbecause they impair both diluting and con-centrating mechanisms in the kidney. As aresult of tubular sodium loss from the distalnephron, thiazide diuretics preserve themedullary osmotic gradient; therefore, freewater can still be reabsorbed in the presenceof ADH.

Acting at a more proximal location, loopdiuretics decrease the medullary osmotic gra-dient and therefore impair free water reab-sorption (urinary concentration) and urinarydilution.

Other causes. Salt-wasting nephritis isassociated with several uncommon renal dis-orders, including renal cystic diseases, anal-gesic nephropathy, obstructive uropathy, andchronic pyelonephritis.14

These states are associated with theinability to conserve sodium even in the pres-ence of hypovolemia, most often when renalimpairment is severe (glomerular filtrationrate < 10–15 mL/minute). In patients withadrenal insufficiency due to Addison disease(mineralocorticoid and glucocorticoid defi-ciency), there is sodium wasting due to aldos-terone deficiency.

CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004 647

Edema orascitesindicateselevated totalbody sodium

648 CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004

When sodium intake cannot keep up withsodium loss (eg, during diarrhea), hypo-volemia develops in the face of elevated con-tinuing renal sodium loss (ie, Addisonian cri-sis). Renal salt loss and hypovolemia-inducedstimulation of ADH cause the hyponatremia.4

If the extracellular fluid volume is normalThe causes of hyponatremia associated withclinical euvolemia include the syndrome ofinappropriate antidiuretic hormone secretion(SIADH), hypocortisolism, diuretic use, beer-drinker’s potomania, psychogenic polydipsia,and hypothyroidism. Diuretic use and SIADHcontribute the majority of cases. In makingthe diagnosis of SIADH, euvolemia must bedemonstrated (see below).

In patients with hyponatremia whoappear euvolemic, the challenge is to rule outhypovolemia that is not revealed by clinicalexamination. The clinical suspicion of euvo-lemia should be confirmed by appropriatelaboratory assessment with random urinesodium, BUN, and serum uric acid measure-ments.

Clinical euvolemia with elevated urinesodium. If a spot urine sodium measurement isnot low and there are no conditions that mayaffect the interpretation of urine sodium val-ues (TABLE 5), the patient is likely truly eu-volemic.

Clinical euvolemia with low urine sodi-um. In patients who appear euvolemic, urinesodium may be low if:• There was recent extracellular fluid vol-ume loss that was replaced by hypotonic fluids,or• Dietary solute intake is very low but waterintake is adequate (eg, in beer-drinker’s poto-mania, as detailed below), or • It is early in the course of hypovolemiawhen clinical signs (eg, tachycardia, orthostat-ic changes) are lagging behind the onset ofsodium retention by the kidney.

In these situations, repeat measurement isadvised.

If there is uncertaintyabout the patient’s volume statusIf there is uncertainty about the patient’sextracellular fluid volume status after initialexamination and urine sodium measurement:

Monitor response to isotonic saline. Incases in which it is still unclear if the patientis euvolemic or hypovolemic (equivocalexamination and urine sodium), the cliniciancan monitor the renal response to isotonicintravenous fluids (1–2 L of 0.9% saline over24–48 hours).

If the patient is truly hypovolemic, theplasma sodium level should increase as fluidresuscitation corrects the hypovolemia anddecreases nonosmotic ADH release, resultingin excretion of free water.

On the other hand, euvolemic patientswould be expected to excrete relatively moreof the sodium load from the intravenous salinethan patients with hypovolemia. An increasein the fractional excretion of sodium of morethan 0.5% after isotonic saline infusion hasbeen proposed to distinguish truly euvolemicpatients (ie, those with SIADH) from patientswith hypovolemia.15

Serum uric acid and BUN can also beused to distinguish between euvolemic andhypovolemic hyponatremia.

Patients with normal or slightly expandedextracellular fluid volume are more likely tohave low levels of serum uric acid and nonel-evated BUN. The converse is true in cases ofhypovolemia in which uric acid is normal andBUN is elevated due to decreased renal perfu-sion. In euvolemic patients with SIADH,serum uric acid is often low due to increasedurate excretion.16

■ SYNDROME OF INAPPROPRIATEANTIDIURETIC HORMONE SECRETION

In patients with hyponatremia, euvolemia,and true hypo-osmolality, the continuedrelease of ADH is inappropriate. However,release can continue in SIADH. Most cases ofSIADH are associated with tumors, pul-monary and central nervous system disease, orvarious drugs, although the cause is not alwaysevident when hyponatremia is detected.17

While most cases of SIADH appear to beassociated with abnormal and inappropriatelevels of ADH (relative to plasma osmolality),10% to 20% of patients do not have measur-able levels of the hormone in their serum,despite continued renal free water reabsorp-tion.18

HYPONATREMIA FREDA AND COLLEAGUES

Confirmeuvolemia bymeasuringurine sodium,BUN, and serumuric acid

Diagnosis of SIADHSIADH is a diagnosis of exclusion. As withother cases of hyponatremia, true plasmahypo-osmolality must be confirmed. Theremust also be failure to excrete a maximallydilute urine and documentation of euvolemia.As discussed previously, the latter can some-times prove difficult. In these cases we recom-mend measuring serum uric acid and BUNand testing the response to isotonic saline torule out hypovolemia. The diagnostic criteriafor SIADH are listed in TABLE 6.18

During the course of chronic hypona-tremia due to SIADH, a steady state isachieved in which urinary sodium excretiontypically matches dietary sodium intake,resulting in a urine sodium of around 40mEq/L.5 In this case, the persistence ofhyponatremia depends on continued waterintake, as patients will often become nor-monatremic on proper water restriction.

However, if a patient with SIADHbecomes hypovolemic, urine sodium may fallto levels below 20 mEq/L, confusing the clin-ical picture. Furthermore, in patients witheuvolemia and SIADH, serum uric acid levelsand BUN are commonly reduced.

Other causes of euvolemic hyponatremiaPatients with psychogenic polydipsia ingestmassive amounts of water (typically 20–30L/day) and overcome the renal capacity forfree water excretion. They have true plasmahypo-osmolality, but most importantly, they

also make an appropriately dilute urine (< 100mOsm/kg H2O). They can appear euvolemicor volume-expanded, depending on soluteintake and other comorbidities.

Beer-drinker’s potomania refers to thedevelopment of hyponatremia from waterintake (in this case in the form of beer) that isdisproportionate to a severely reduced dietarysodium intake.19 At reduced levels of soluteintake, there is a decrease in the ability of thenephron to excrete free water. As most adultsingest approximately 1,000 mOsm of soluteper day and can dilute their urine to 50mOsm/kg H2O, the kidney can excretearound 20 L (1,000/50) of free water per day.

If dietary solute intake is severely reducedto 300 mOsm/day and the resulting maximumcapacity for renal dilution is 50 mOsm/kg H2O,any water intake from beer over 6 L will resultin free water retention and hyponatremia.

Hypothyroidism can also impair freewater excretion. Although the mechanismsare not fully understood, giving thyroid hor-mone can correct the low plasma sodium con-centration.20

Pure glucocorticoid deficiency due toadrenocorticotropic hormone deficiency oranterior hypopituitarism can cause hypona-tremia in the presence of preserved mineralo-corticoid function. These patients are usuallyeuvolemic. Cortisol replacement suppressesADH release from the hypothalamus and pro-motes free water release at the level of the kid-ney.21

CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004 649

Diagnostic criteria for the syndrome of inappropriateantidiuretic hormone secretion (SIADH)Essential criteria

True plasma hypo-osmolality (< 275 mOsm/kg H2O)Inappropriate urinary response to hypo-osmolality (urine osmolality > 100 mOsm/kg H2O)Euvolemia; no edema, ascites, or signs of hypovolemiaElevated urine sodium(> 30 mEq/L) during normal sodium and water intakeNo other causes of euvolemic hyponatremia (hypothyroidism, diuretic use, hypocortisolism)

Supplemental criteriaUnable to excrete > 80% of a water load (20 cc/kg) in 4 hours and/or failure to achieve urine

osmolality < 100 mOsm/kg H2ONo significant increase in serum sodium after volume expansion, but improvement with fluid restriction

MODIFIED FROM VERBALIS JG. THE SYNDROME OF INAPPROPRIATE ANTIDIURETIC HORMONE SECRETION AND OTHER HYPO-OSMOLAR DISORDERS.IN: SCHRIER RW, EDITOR. DISEASES OF THE KIDNEY AND URINARY TRACT, 7TH ED. PHILADELPHIA;

LIPPINCOTT WILLIAMS & WILKINS, 2001: 2511–2548.

T A B L E 6

SIADH is adiagnosis ofexclusion

■ REFERENCES1. Gill G, Leese G. Hyponatremia: biochemical and clinical per-

spectives. Postgrad Med J 1998; 74:516–523.2. Lee WH, Packer M. Prognostic importance of serum sodium

concentration and its modification by converting-enzymeinhibition in patients with severe chronic heart failure.Circulation 1986; 73: 257–267.

3. Gines P, Arroyo V, Rodes J. Complications of cirrhosis: ascites,hyponatremia, hepatorenal syndrome, and spontaneousbacterial peritonitis. In: Bacon BR, Di Bisceglie AM, editors.Liver Disease: Diagnosis and Management, 1st ed.Philadelphia; Churchill Livingstone, 2000:238–250.

4. Berl T, Schrier RW. Disorders of water metabolism. In: SchrierRW, editor. Renal and Electrolyte Disorders, 6th ed.Philadelphia; Lippincott Williams & Wilkins, 2003:1–63.

5. Rose BD, Post TW. Clinical Physiology of Acid-Base andElectrolyte Disorders, 5th ed. New York; McGraw-Hill,2001:696–745.

6. Schrier RW. Body fluid volume regulation in health and dis-ease: a unifying hypothesis. Ann Intern Med 1990;113:155–159.

7. Hillier TA, Abbott RD, Barrett EJ. Hyponatremia: evaluatingthe correction factor for hyperglycemia. Am J Med 1999;106:399–403.

8. Agarwal R, Emmett M. The post-transurethral resection ofprostate syndrome: therapeutic proposals. Am J Kidney Dis1994; 24:108–111.

9. Oster JR, Singer I. Hyponatremia, hyposmolality, and hypo-tonicity: tables and fables. Arch Intern Med 1999;159:333–336.

10. James TJ. Ion-selective electrodes. Postgrad Med J 1999;75:254.

11. Chung HM, Kluge R, Schrier RW, Anderson RJ. Clinicalassessment of extracellular fluid volume in hyponatremia.Am J Med 1987; 83:905–908.

12. Dorhout Mees EJ, Geers AB, Koomans HA. Blood volume

and sodium retention in the nephrotic syndrome: a contro-versial pathophysiological concept. Nephron 1984;36:201–211.

13. Sonnenblick M, Friedlander Y, Rosin AJ. Diuretic-inducedsevere hyponatremia. Review and analysis of 129 reportedpatients. Chest 1993; 103:601–606.

14. Uribarri J, Oh MS, Carroll HJ. Salt-losing nephropathy. Clinicalpresentation and mechanisms. Am J Nephrol 1983;3:193–198.

15. Milionis HJ, Liamis GL, Elisaf MS. The hyponatremic patient:a systematic approach to laboratory diagnosis. CMAJ 2002;166:1056–1062.

16. Maesaka JK. An expanded view of SIADH, hyponatremiaand hypouricemia. Clin Nephrol 1996; 46:79–83.

17. Verbalis JG. The syndrome of inappropriate antidiuretic hor-mone secretion and other hypo-osmolar disorders. In:Schrier RW, editor. Diseases of the Kidney and Urinary Tract,7th ed. Philadelphia; Lippincott Williams & Wilkins, 2001:2511–2548.

18. Zerbe R, Stropes L, Robertson G. Vasopressin function in thesyndrome of inappropriate antidiuresis. Annu Rev Med1980; 31:315–327.

19. Thaler SM, Teitelbaum I, Berl T. “Beer potomania” in non-beer drinkers: effect of low dietary solute intake. Am JKidney Dis 1998; 31:1028–1031.

20. Hanna FW, Scanlon MF. Hyponatremia, hypothyroidism, androle of arginine-vasopressin. Lancet 1997; 350:755–756.

21. Linas SL, Berl T, Robertson GL, Aisenbrey GA, Schrier RW,Anderson RJ. Role of vasopressin in the impaired waterexcretion of glucocorticoid deficiency. Kidney Int 1980;18:58–67.

ADDRESS: Phillip M. Hall, MD, Department ofNephrology and Hypertension, The Cleveland ClinicFoundation, A51, 9500 Euclid Avenue, Cleveland, OH44195; e-mail [email protected].

650 CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 71 • NUMBER 8 AUGUST 2004

HYPONATREMIA FREDA AND COLLEAGUES