at: inbox
Serum osmolal gap - UpToDate
All topics are updated as new evidence becomes available and our peer review process is complete.
Literature review current through: Jan 2023. | This topic last updated: Sep 21, 2022.
OSMOLALITY VERSES OSMOLARITY — Osmolality is a measurement of the number of moles of dissolved particles per kg of solvent (for clinical purposes, water is the solvent). Osmolarity is a measure of the number of dissolved particles per liter of solution. For clinical purposes, these two measurements, despite different units, are virtually the same and are often used interchangeably. However, it should be understood that direct measurements (ie, measurements reported from an osmometer) measure osmolality (mosm/kg water) while calculated measurements, as described below, are estimates of osmolarity (mosm/L solution). For the practical clinical purposes described below, these measurements can be considered interchangeable. For consistency, we will refer to these measurements as osmolality throughout this topic.
DETERMINANTS OF SERUM OSMOLALITY — The serum (or plasma) osmolality is determined by the concentrations (in mmol/L) of the different solutes in the plasma. In most individuals, the solutes that normally exist in high enough concentration to significantly affect the osmolality are sodium salts (mainly chloride and bicarbonate), glucose, and urea. If no other solutes are present in serum at high millimolar concentrations (greater than 5 mmol/L), then these three solute concentrations can be used to predict the measured osmolality. A variety of formulas have been evaluated for this purpose, and the following relationship is acceptable for practical clinical purposes [1-5]:
Calculated Sosm = (2 x Serum [Na, in mmol/L]) + [Glucose, in mg/dL]/18 + [Blood urea nitrogen, in mg/dL]/2.8
Or, with international units (all of which are in mmol/L):
Calculated Sosm = (2 x Serum [Na]) + [Glucose] + [Urea]
The serum sodium is multiplied by two to account for the osmolal contributions of the accompanying anions (chloride and bicarbonate), and, in the first formula, the divisors 18 and 2.8 convert units of mg/dL to mosmol/kg (calculator 1 and calculator 2) [6].
MEASUREMENT OF SERUM OSMOLALITY — Osmolality is a colligative property that is related to the ratio of the number of solute particles to the number of solvent molecules in the solution. Colligative properties affect the freezing point, boiling point, and vapor pressure or dew point of any solution. Increasing the osmolality reduces the temperature at which the solution freezes, reduces the solution’s vapor pressure or dew point, and raises the boiling point temperature. The osmolality of serum (or any other aqueous solution) can be determined by measuring any of these colligative properties. The most commonly used clinical osmometers measure the temperature at which the solution freezes. Solute-free water freezes at 0ºC. In general, if 1,000 mosmol of any relatively small solute (or combination of solutes) is added to 1 kg of water, the freezing point of this solution will be depressed by 1.86ºC. This relationship can be used to calculate the osmolality of a solution. As an example, the freezing point of the plasma water is normally approximately -0.521ºC. This represents an osmolality of 0.280 osmol/kg (0.521 ÷ 1.86) or 280 mosmol/kg.
Osmometers that measure vapor pressure depression are much less frequently utilized by clinical laboratories. These devices are less useful in certain circumstances because volatile solutes do not reduce vapor pressure as much as equimolar concentrations of nonvolatile solutes. Consequently, the osmolal contribution of ethanol, methanol, ethylene glycol, and isopropyl alcohol in serum is underestimated by vapor pressure osmometers. Thus, vapor pressure osmometry should not be utilized diagnostically when these poisonings are suspected [7,8]. (See ‘Clinical utility of the serum osmolal gap’ below.)
CLINICAL UTILITY OF THE SERUM OSMOLAL GAP — The measured osmolality and the calculated osmolality should be similar (generally within approximately 6 mosmol/kg) [2-4]. An elevated serum osmolal gap exists if the measured osmolality is more than 10 mosmol/kg greater than the calculated osmolality (using the equations shown above) (calculator 1 and calculator 2). The cause of the elevated osmolal gap will often be apparent from the clinical context (table 1). (See ‘Determinants of serum osmolality’ above.)
There are two mechanisms, other than sporadic laboratory errors, which cause the measured serum osmolality to be significantly higher than the calculated value:
●There may be an additional solute or solutes other than a sodium salt, glucose, or urea that is present at a concentration high enough to raise the osmolality (table 1) [3,4,6,9-12]. Some of these solutes readily penetrate the cell membrane (ethanol, methanol, ethylene glycol, propylene glycol, isopropanol), some can generate water shifts because they exert “tonicity” (the infusion of mannitol, immune globulin, sucrose, maltose), and some produce hyponatremia because they are absorbed as hypotonic solutions (glycine- or sorbitol-based irrigant solutions used during transurethral prostate resection or hysteroscopy).
●Marked hyperlipidemia or hyperproteinemia does not affect the concentration of sodium in serum water, and it is this concentration that determines the measured serum osmolality. However, if the sodium concentration is measured utilizing certain analytic techniques (such as flame photometry or indirect ion-selective electrode methodology), the result may be spuriously reduced (called pseudohyponatremia) when marked hyperlipidemia or hyperproteinemia exist. (See “Causes of hypotonic hyponatremia in adults”.)
As noted above, the accurate detection of an elevated serum osmolal gap when it is generated by any volatile solute, such as ethanol, methanol, ethylene glycol, or isopropyl alcohol poisoning, requires utilization of a freezing point depression osmometer, not a vapor pressure osmometer. (See ‘Measurement of serum osmolality’ above.)
Issues related to the urine osmolal gap are discussed separately. (See “Urine anion and osmolal gaps in metabolic acidosis”.)
MAJOR CAUSES OF AN ELEVATED SERUM OSMOLAL GAP
Ethanol ingestion — The most common cause of an elevated serum osmolal gap is ethanol ingestion. The serum ethanol concentration can often exceed 100 mosmol/kg [3,5,12], but even moderate levels will generate a significant serum osmolal gap.
Ethanol has a molecular weight of 46 mg/mmol. Thus, its osmolal contribution should be equal to the serum ethanol level in mg/100 mL divided by 4.6. Although some studies have proposed that ethanol may actually have an effect greater than its molecular weight would generate (due to reduction in the serum water content) [4,5,13], contemporary data showed that this was not correct [14].
Thus, if one wishes to include the contribution of ethanol to the calculation of the serum osmolality, the following equations should be used:
Calculated Sosm = (2 x Serum [Na]) + [Glucose]/18 + [BUN]/2.8 + [Ethanol]/4.6
Or, with international units (all of which are in mmol/L):
Calculated Sosm = (2 x Serum [Na]) + [Glucose] + [Urea] + (1.25 x [Ethanol])
This calculation is particularly important when toxic alcohols or glycols are coingested with ethanol, and this occurs frequently [15].
High anion gap metabolic acidosis — Measurement of the serum osmolal gap can be diagnostically helpful in patients with an otherwise unexplained high anion gap metabolic acidosis, particularly when toxic alcohol poisoning is suspected (table 2). (See “Approach to the adult with metabolic acidosis”.)
Toxic alcohols and glycols
Methanol and ethylene glycol — Calculation of the serum osmolal gap is a helpful rapid screening test when methanol or ethylene glycol intoxication is suspected [6,11,16]. In the appropriate clinical setting, an elevated serum osmolal gap (greater than 10 mosmol/kg) is consistent with but not diagnostic of methanol or ethylene glycol intoxication [4].
Both methanol and ethylene glycol are metabolized to strong organic acids (eg, formic acid, oxalic acid, and others). As described below, conversion of these toxic alcohols and glycols to strong acids simultaneously reduces the osmolal gap and generates an anion gap metabolic acidosis.
Although an anion gap metabolic acidosis is not always present (especially early in the disorder, before methanol and ethylene glycol are converted to strong acids such as formic acid and oxalic acid), the combination of both an elevated osmolal gap and elevated anion gap increases the diagnostic likelihood of a toxic alcohol ingestion. Strong suspicion of this diagnosis should trigger therapies that inhibit alcohol dehydrogenase. These interventions block the formation of toxic metabolites and should generally be initiated urgently and before results of toxin assays are available. (See “Methanol and ethylene glycol poisoning: Management”, section on ‘Alcohol dehydrogenase inhibition’.)
Methanol is a small molecule (mol wt of 32), and ingestion (or absorption across the skin) can generate high serum concentrations. A serum methanol concentration of 80 mg/dL is equivalent to approximately 21 mosmol/kg, and this produces a similar increase in the serum osmolal gap. The increase in osmolal gap is less pronounced with similar concentrations of ethylene glycol, which is a larger molecule (mol wt of 62). Thus, at the same serum concentration, ethylene glycol will raise the serum osmolal gap by only approximately one-half as much as methanol (table 3) [17]. (See “Methanol and ethylene glycol poisoning: Pharmacology, clinical manifestations, and diagnosis”.)
A high serum osmolal gap in a patient with an otherwise unexplained high anion gap metabolic acidosis may be an important clue to the presence of methanol or ethylene glycol poisoning [18]. The possibility that the patient has also ingested ethanol must be considered, and, if present, its contribution to the osmolal gap should be identified and quantitated. Co-ingestion of a toxic alcohol or glycol and ethanol is common [5].
Thus, when the osmolal gap calculation is used as a diagnostic clue for the presence of toxic molecules such as methanol or ethylene glycol, the above formulas for estimation of the serum osmolality should be amended to include any contribution from ethanol [4,5] (see ‘Major causes of an elevated serum osmolal gap’ above):
Calculated Sosm = (2 x Serum [Na]) + [Glucose]/18 + [BUN]/2.8 + [Ethanol]/4.6
As noted above, the serum sodium is multiplied by two to account for accompanying anions (normally chloride and bicarbonate) and the divisors 18, 2.8 and 4.6 convert units of mg/dL to mosmol/kg (calculator 1). (See ‘Ethanol ingestion’ above.)
When international units (mmol/L) are used (calculator 2) [4,5]:
Calculated Sosm = (2 x Serum [Na]) + [Glucose] + [Urea] + (1.25 x [Ethanol])
Metabolism of ingested methanol or ethylene glycol causes a progressive decline in the osmolal gap. When metabolism converts the parent molecule to acid metabolites, such as formic acid (in the case of methanol) or glyoxylic and oxalic acid (in the case of ethylene glycol), their osmotic contribution disappears because each molecule of organic acid produced generates an equimolar disappearance of bicarbonate. Thus, this increase in organic anion osmoles is matched by an identical fall in bicarbonate osmoles. Consequently, the serum osmolal gap estimates the molar quantity of the uncharged parent alcohol molecules and any uncharged metabolites but not their strong acid metabolites [19].
The result of these metabolic effects is that the osmolal gap reflects the concentration of the ingested alcohol or glycol and usually falls with its metabolism, while the anion gap acidosis is a consequence of metabolism and increases as the alcohol or glycol is converted to acidic products. Most often when a patient presents to the emergency department after one of these ingestions, elements of both conditions are present; that is, both an osmolal gap and an anion gap metabolic acidosis coexist. However, only the osmolal gap may be found in patients who present very early after ingestion, and only the anion gap metabolic acidosis may exist in patients who present very late after ingestion. In addition, co-ingestion of ethanol (which slows metabolism of these toxic alcohols and glycols) may reduce or even eliminate the development of the anion gap metabolic acidosis. (See “Methanol and ethylene glycol poisoning: Pharmacology, clinical manifestations, and diagnosis”, section on ‘Plasma osmolal gap’.)
These principles can be illustrated by the following example. The addition of 20 mmol of methanol to each liter of extracellular fluid will raise the serum osmolality by approximately 20 mosmol/kg. If all of the methanol is converted to formic acid, then 20 mosmol of methanol has been converted to 40 mosmol of solute (20 hydrogen ions and 20 formate anions). However, the hydrogen ions will be largely buffered by bicarbonate. This causes bicarbonate to be converted to H2CO3 and then to H2O and CO2, which results in disappearance of both the hydrogen ion and the HCO3. The net effect is that, in each liter of extracellular fluid, 20 formate anions “replace” 20 bicarbonate ions, the serum osmolality then returns to its baseline level prior to the ingestion, and there is no serum osmolal gap.
Other alcohols — Diethylene glycol is another toxic glycol that can produce both an osmolal gap and potentially fatal anion gap metabolic acidosis. This glycol caused many epidemic poisonings when it was used as a solvent for various medications. The most famous epidemic occurred in 1937 (the “Massengill Incident”) when the S.E. Massengill Company used diethylene glycol in a sulfanilamide preparation [20]. Other epidemics and sporadic cases have resulted from incorrectly formulated cough syrups (in Panama), toothpaste (which was produced in China), teething mixtures (Nigeria) [21,22], wallpaper stripping solution [23], and antifreeze and brake fluids [24,25]. In addition, diethylene glycol intoxication is often associated with counterfeit drugs, and a growing illegal internet drug market makes it a potential public health threat, particularly in Africa and Asia [22,25]. One of the principal metabolites of diethylene glycol is 2-hydroxyethoxyacetic acid, a potent neurotoxin that also produces a high anion gap metabolic acidosis and diglycolic acid, which is nephrotoxic [22,24,25]. Patients with diethylene glycol poisoning often present with gastrointestinal symptoms, hepatitis, and pancreatitis. Later, they may develop encephalopathy, cranial nerve and peripheral neuropathies with fulminant ascending paralysis, and respiratory failure. Bifacial plegia occurs in up to 50 percent of affected patients [26]. Oliguric or anuric acute kidney injury can occur due to necrosis of the proximal tubular cells. Unlike ethylene glycol toxicity, there is no deposition of calcium oxalate crystals, since oxalic acid is not generated [22]. Measurement of diethylene glycol in the serum establishes the diagnosis; however, it is not routinely performed, and the diagnosis often relies upon the history and clinical presentation.
The osmolal and anion gaps may also be elevated in patients who are treated with intravenous propylene glycol [27-32]. This pharmaceutical solvent is used to dissolve a variety of medications and becomes toxic when administered in large doses. Propylene glycol is metabolized to several products including L-lactic acid, D-lactic acid, and glyoxal. The combination of a large osmolal gap and a high anion gap acidosis in patients treated with medications dissolved in propylene glycol are hallmarks of this disorder. Most reported cases of propylene glycol poisoning have occurred in the intensive care unit setting when high doses of lorazepam or phenobarbital are used to treat alcohol withdrawal or to induce a therapeutic coma [28,29]. Other drugs dissolved in propylene glycol include diazepam, phenytoin, trimethoprim-sulfamethoxazole, etomidate, nitroglycerin, and esmolol [31,32].
Isopropyl alcohol is metabolized to acetone, which is also a nonionized molecule. Acetone is not an acid and does not produce metabolic acidosis or elevate the anion gap. Thus, isopropyl alcohol ingestion generates a large osmolal gap (both the isopropyl alcohol and the acetone contribute to the osmolality) but will not directly cause a metabolic acidosis. (See “Isopropyl alcohol poisoning”.)
In contrast to methanol, ethylene glycol, and isopropyl alcohol, ethanol is rapidly metabolized to carbon dioxide and water. Thus, although its osmotic contribution wanes over time, metabolic acidosis does not occur.
Ketoacidosis and lactic acidosis — The addition of ketoacids or lactic acid to the serum will not directly raise the serum osmolality or generate a serum osmolal gap. When lactic acid or a ketoacid accumulates in the extracellular fluid the hydrogen ions from the acid combine with bicarbonate, creating carbonic acid, which is then expired as carbon dioxide. Thus, multiplying the sodium concentration by 2 includes all the osmoles derived from these acids.
Nevertheless, the serum osmolal gap can increase to a small degree in ketoacidosis and in lactic acidosis. There are several explanations for the elevated osmolal gap in patients with these disorders:
●In lactic acidosis, small, osmotically active, noncharged molecules and positively charged molecules (such as amino acids) are released from dead and dying cells [33].
●In ketoacidosis, increased plasma acetone and glycerol levels raise the osmolal gap [34].
Advanced chronic kidney disease — The serum osmolal gap is increased in advanced chronic kidney disease due to the retention of unidentified small solutes [6,9]. Dialysis removes these molecules and eliminates the serum osmolal gap in these patients. This effect is less pronounced with acute kidney injury.
Low osmolality intravenous contrast infusion — The intravenous infusion of low osmolality contrast computed tomography (CT) scan can increase the osmolal gap by 3 to 4 mosmol/kg [35].
Pseudohyponatremia — Any measurement device that is affected by alterations in the water content of serum can generate pseudohyponatremia. This can occur whenever marked hyperproteinemia or hyperlipidemia exists. It is therefore generally recommended that serum osmolality be measured in the setting of otherwise unexplained or unexpected hyponatremia. If the calculated osmolality is substantially lower than the measured osmolality, then pseudohyponatremia may be present. Inappropriate treatment to raise the serum sodium in patients with pseudohyponatremia (ie, when hyponatremia is a laboratory artifact) can result in serious morbidity. (See “Causes of hypotonic hyponatremia in adults”.)
Other — An elevated serum osmolal gap can occur in a variety of other clinical settings. These include (table 1):
●The absorption or accidental systemic infusion of nonconductive glycine, sorbitol, or mannitol irrigation solutions during transurethral resection of the prostate or bladder or hysteroscopy, or the intravenous infusion of hypertonic mannitol to, for example, treat cerebral edema. (See “Hyponatremia following transurethral resection, hysteroscopy, or other procedures involving electrolyte-free irrigation” and “Complications of mannitol therapy”.)
●The administration of intravenous immune globulin in a 10 percent maltose solution to patients with kidney failure who are unable to metabolize the maltose. (See “Overview of intravenous immune globulin (IVIG) therapy”.)
●The “sick cell syndrome” in patients with multiorgan failure [33,36].
SUMMARY AND RECOMMENDATIONS
●The serum (or plasma) osmolality is determined by the concentrations (in mmol/L) of the different solutes in the plasma. In normal subjects, sodium salts (mainly chloride and bicarbonate), glucose, and urea are the primary circulating solutes. If no other solutes are present in serum at high concentrations, then these three solute concentrations can be used to predict the measured osmolality. A variety of formulas have been evaluated for this purpose (calculator 1 and calculator 2). (See ‘Determinants of serum osmolality’ above.)
●The osmolality of serum (or any other aqueous solution) is directly measured by an osmometer. The most commonly used clinical osmometers measure the temperature at which the solution freezes (freezing point depression). Osmometers based upon vapor pressure depression are much less frequently utilized by clinical laboratories. These devices are less useful in certain circumstances because volatile solutes such as ethanol, methanol, ethylene glycol, and isopropyl alcohol do not reduce vapor pressure as much as equimolar concentrations of nonvolatile solutes. (See ‘Measurement of serum osmolality’ above.)
●The measured osmolality and the calculated osmolality should be similar (generally within approximately 6 mosmol/kg). An elevated serum osmolal gap exists if the measured osmolality is more than 10 mosmol/kg greater than the calculated osmolality. The cause of the elevated osmolal gap will often be apparent from the clinical context (table 1). (See ‘Clinical utility of the serum osmolal gap’ above.)
●There are two mechanisms, other than sporadic laboratory errors, which cause the measured serum osmolality to be significantly higher than the calculated value (see ‘Clinical utility of the serum osmolal gap’ above):
•There may be an additional solute or solutes (such as ethanol or ethylene glycol) other than a sodium salt, glucose, or urea that is present at a concentration high enough to raise the osmolality.
•The measured sodium concentration may be spuriously reduced (called pseudohyponatremia) as a result of marked hyperlipidemia or hyperproteinemia, when measured with a device susceptible to such artifacts.
●Major causes of an elevated serum osmolal gap include (table 1) (see ‘Major causes of an elevated serum osmolal gap’ above):
•Ethanol ingestion, which is the most common cause of an elevated serum osmolal gap. (See ‘Ethanol ingestion’ above.)
•Toxic alcohols and glycols (eg, methanol and ethylene glycol), which may also generate a high anion gap metabolic acidosis. (See ‘Toxic alcohols and glycols’ above.)
•Other alcohols such as diethylene glycol, propylene glycol, and isopropyl alcohol. (See ‘Other alcohols’ above.)
•Ketoacidosis and lactic acidosis, which may produce small elevations in the serum osmolal gap. (See ‘Ketoacidosis and lactic acidosis’ above.)
•Advanced chronic kidney disease, due to the retention of unidentified small solutes, in a patient not receiving dialysis. (See ‘Advanced chronic kidney disease’ above.)
•Pseudohyponatremia, which can occur when certain electrolyte analyzers (for example, those utilizing flame photometers or indirect ion-selective electrode methodology) are utilized in the presence of marked hyperproteinemia or hyperlipidemia. (See ‘Pseudohyponatremia’ above.)