Urine anion and osmolal gaps in metabolic acidosis

INTRODUCTION — Measurement of the urine anion gap (UAG) and/or urine osmolal gap (UOG) may be helpful in the evaluation of patients with a normal anion gap (hyperchloremic) metabolic acidosis by providing an estimate of urinary ammonium (NH₄⁺) excretion (table 1) [1-5]. The normal renal response to metabolic acidosis is to increase urinary ammonium excretion. (See "Approach to the adult with metabolic acidosis".)

The clinical use of the urine anion and osmolal gaps will be reviewed here. Issues related to use of the serum anion and osmolal gaps are discussed separately:

●(See "The delta anion gap/delta HCO3 ratio in patients with a high anion gap metabolic acidosis".)

●(See "Serum osmolal gap".)

BRIEF OVERVIEW OF RENAL ACID EXCRETION — Ingestion of a typical Western diet generates approximately 50 to 100 mEq of nonvolatile acid (ie, acids other than carbon dioxide [CO₂]) per day. To maintain acid balance, these nonvolatile acids must be excreted in the urine. The hydrogen ions are excreted via a process that includes the following:

●First, filtered bicarbonate must be reabsorbed since sodium or potassium bicarbonate loss is equivalent to the generation of hydrogen ions.

●Next, the hydrogen ions associated with the nonvolatile acid that is generated must be excreted. Only trivial amounts of free hydrogen ions can be excreted since, at a urine pH of 4.5, the free hydrogen ion concentration is less than 0.04 mEq/L. Thus, urine hydrogen ions must be bound to buffers such as phosphate (converting HPO₄ to H₂PO₄), urate, citrate and creatinine, or ammonia (NH₃; converting NH₃ to the ammonium ion [NH₄⁺]).

●The main adaptive renal response to chronic metabolic acidosis is to increase hydrogen ion excretion in the form of ammonium. Patients who develop persistent (greater than several days) and severe metabolic acidosis may increase their ammonium excretion rates from the normal value of 30 to 40 mEq/day (for those on a typical Western diet) to as much as 200 to 300 mEq/day [6,7]. Ammonium is often excreted with chloride but may also be excreted with other anions such as ketoacid anions, sulfate, hippurate, citrate, or bicarbonate (although the excretion of ammonium with bicarbonate or citrate does not remove nonvolatile acid from the body, since bicarbonate or potential bicarbonate loss is equivalent to hydrogen ion generation).

USE OF URINE ANION AND OSMOLAL GAPS IN THE DIAGNOSIS OF DISTAL RTA — The usual renal response to acute metabolic acidosis with acidemia is to reduce the urine pH to 5.3 or less. Distal renal tubular acidosis (RTA) is characterized by a urine pH that is persistently above 5.5 despite metabolic acidosis and acidemia. However, chronic metabolic acidosis, such as the acidosis of chronic diarrhea, can be associated with a relatively high urine pH despite a normal kidney response to acidosis. That is because with chronic metabolic acidosis, especially when associated with hypokalemia, increased ammoniagenesis in the kidney raises urine ammonium (NH₄⁺) levels that, in turn, raise the urine pH. RTAs, conversely, are associated with low urine levels of ammonium.

Both the UAG and UOG, which are potential analogs of the urine ammonium concentration, may be useful in distinguishing distal RTA from certain nonRTA causes of metabolic acidosis. Because of potential shortcomings of the UAG described below, the UOG may be a better analog of urine ammonium. However, reliable and widely available assays to directly measure the urine ammonium concentration will make both of these indirect estimates obsolete [8]. (See 'Limitations of the UAG' below.)

Patients with metabolic acidosis due to severe, watery diarrhea may have a urine pH above 5.5 despite normal kidneys and an appropriate renal response to the disorder [1]. This occurs because both metabolic acidosis and hypokalemia combine to generate a potent stimulus to renal ammoniagenesis and the secretion of ammonia into the urine. Hypokalemia does this, at least in part, by promoting the exit of potassium from cells in exchange for hydrogen entry into cells. The resulting intracellular acidification stimulates ammoniagenesis and renal tubular ammonium secretion. Hypokalemia also has other effects on renal ammoniagenesis and ammonia transport that simultaneously enhance ammonium excretion and potassium conservation [9-13]; expression of major ammonia transporters are increased, such as Rhesus glycoproteins in alpha-intercalated cells and principal cells of the collecting duct. Most of the ammonia (NH₃) that is secreted into the tubular lumen binds to hydrogen ions (H⁺) to form ammonium:

Binding H to NH₃ reduces the urine hydrogen ion concentration, and this can raise the urine pH above 5.5. When this occurs, the relatively high urine pH in a patient with hyperchloremic, hypokalemic metabolic acidosis may incorrectly suggest that the patient has RTA. However, the high urine pH in patients with distal RTA reflects a defect in renal hydrogen ion secretion and is associated with a reduced urine ammonium excretion rate. By contrast, diarrhea-associated metabolic acidosis may generate a relatively high urine pH because of the potent renal ammoniagenesis and ammonium excretory response.

Direct measurement of urine ammonium should readily differentiate these conditions. Unfortunately, most hospital clinical laboratories do not offer this measurement. Thus, two alternative urine chemistry measurements/calculations have been proposed as indirect, semiquantitative analogs of urine ammonium. They are the UAG and the UOG. These indirect indicators of urine ammonium can help differentiate certain nonRTA-induced hyperchloremic metabolic acidoses, such as that generated by chronic diarrhea, from RTA [1,2].

URINE ANION GAP — The UAG, sometimes called the "urine cation gap" or the "urine net charge", is calculated as the difference between the sum of the urine sodium (Na) plus potassium (K) concentrations and the urine chloride (Cl) concentration (calculator 1).

It is important to understand that the formula for the UAG is different from the formula used to calculate the serum anion gap. The serum anion gap is the difference between the serum sodium and the sum of the serum chloride and bicarbonate concentrations. (See "Approach to the adult with metabolic acidosis", section on 'Physiologic interpretation of the serum anion gap'.)

When individuals ingest a typical Western diet, the quantity of sodium and potassium absorbed by the gastrointestinal tract usually exceeds the quantity of absorbed chloride. Thus, under normal circumstances, net balance considerations mandate that the sum of sodium and potassium excreted into the urine is greater than the amount of excreted urine chloride. The UAG therefore has a positive value (between 20 and 90) in healthy individuals ingesting such diets [4,14].

If large volume and/or chronic, watery diarrhea develops, hyperchloremic metabolic acidosis can develop due to the loss of sodium and potassium with bicarbonate and with potential bicarbonate anions (such as butyrate, citrate, and lactate) in the stool. Loss of these ions causes renal sodium and potassium excretion to fall without a parallel fall in chloride excretion. The major cation excreted in the urine with the chloride under these conditions is ammonium (NH₄⁺). The excretion of ammonium chloride (NH₄Cl) is, from an acid-base perspective, functionally like the excretion of hydrochloric acid (HCl). As a result, the urine Cl concentration usually exceeds the sum of urine sodium plus potassium concentrations, and the UAG becomes negative [1,2,4,5]. By contrast, patients with hyperchloremic metabolic acidosis generated by renal tubular acidosis (RTA) have reduced urinary ammonium excretion. Also, they are not losing large amounts of sodium and potassium in the stool. Thus, their urine sodium, potassium, and chloride excretion reflects their dietary intake, and these patients generally have a positive UAG. The UAG can therefore provide an indirect, or surrogate, estimate of urinary ammonium excretion.

Although the UAG has been proposed as a gross indicator of the urine ammonium concentration, the relationship is too loose for this calculated number to be considered a quantitative or exact measurement of urine ammonium concentration [1,3-5,14,15]. In addition, under certain conditions (described below), the UAG calculation will definitely not reflect urine ammonium excretion.

Usual clinical application of the UAG — In patients with a hyperchloremic (ie, nonanion gap) metabolic acidosis, the UAG is typically used in the following way:

●A positive UAG (generally between 20 and 90 mEq/L) is usually indicative of low or normal ammonium excretion. Thus, patients with metabolic acidosis due to impaired renal ammonium excretion (such as a distal RTA) will have a positive UAG. The UAG is also typically positive in patients with a chronic respiratory alkalosis; metabolic compensation in such patients is associated with an appropriate reduction in ammonium excretion and a decrease in the serum bicarbonate (HCO₃) concentration [16].

●A negative UAG (generally between -20 and -50 mEq/L) is usually indicative of increased ammonium excretion (ie, greater than 80 mEq/L). Such values of the UAG occur in patients with metabolic acidosis generated by diarrhea.

●Results near zero (ie, between +20 and -20) cannot be reliably interpreted.

Limitations of the UAG — It is important to understand that several conditions can disrupt the typical relationship described above between the urine ammonium concentration and the UAG. When these conditions exist, the UAG may no longer correctly indicate the level of urine ammonium excretion and therefore the etiology of the metabolic acidosis [1,2,17,18]:

●The UAG becomes a less reliable predictor of the urine ammonium concentration in patients with acute or chronic kidney disease [19,20]. Its performance in these patients could be improved by incorporating the urine phosphate and sulfate concentrations, but that added complexity makes this an impractical analysis.

●Increased urinary excretion of various unmeasured anions (ie, nonchloride anions) that are not normally present in the urine in high concentrations can disrupt the relationship described above between the urine ammonium and UAG. Some examples of these unmeasured urine anions include beta-hydroxybutyrate and acetoacetate in patients with ketoacidosis, hippurate following toluene inhalation (glue sniffing), bicarbonate when patients with proximal RTA are treated with alkali therapy, D-lactate in D-lactic acidosis, and 5-oxoproline associated with chronic acetaminophen ingestion (especially in malnourished women). These anions may be excreted with sodium or potassium or with ammonium. The excretion of ammonium with such unmeasured anions will not reduce the UAG as occurs when ammonium is excreted with chloride. In addition, the excretion of sodium and/or potassium with such unmeasured anions contributes to a positive UAG. Thus, the excretion of such anions will disrupt the relationship between the UAG and urine ammonium [21,22]. (See "The delta anion gap/delta HCO3 ratio in patients with a high anion gap metabolic acidosis".)

Consequently, the relationship between the UAG and urine ammonium excretion may be disrupted in many patients with high anion gap metabolic acidoses. When these conditions exist, the UOG (discussed below) is a superior indicator of urine ammonium excretion.

●If the urine pH is >6.5 then bicarbonate in the urine can also affect the calculation of the UAG. The urine bicarbonate would need to be subtracted from the UAG result for it to remain an index of urine ammonium.

●Neonates excrete other unmeasured anions at relatively high rates; therefore, the UAG is an unreliable marker of the urine ammonium concentration in these patients [18].

The metabolic acidosis generated by toluene inhalation is a good example of a condition that disrupts the association between the UAG and urine ammonium excretion [17]. Toluene is metabolized to benzoic acid and then rapidly to hippuric acid. Accumulation of these relatively strong acids in the body fluids generates metabolic acidosis. To the extent that the benzoate and hippurate anions are retained in the extracellular fluid, the metabolic acidosis is of the high-anion-gap type (caused by a reciprocal reduction in bicarbonate and increase in benzoate and hippurate). However, in patients with normal or near-normal kidney function, the hippurate and benzoate anions are rapidly excreted in the urine by both glomerular filtration and tubular secretion. These anions are initially excreted together with sodium and potassium (in this phase, the UAG will become more positive). With time, urine ammonium excretion progressively increases in response to both acidemia and hypokalemia. The ammonium is largely excreted with hippurate. Under these conditions, the UAG will likely remain positive and erroneously suggests a relatively low ammonium excretion (since the UAG only detects ammonium excreted with chloride).

Marked hypokalemia often develops in these patients because the early loss of sodium hippurate generates volume contraction and secondary hyperaldosteronism. High aldosterone levels, together with high distal tubule sodium delivery (as sodium benzoate and sodium hippurate), causes substantial renal potassium wasting. Hypokalemia further increases renal ammoniagenesis and ammonium excretion. As a result, more hydrogen ions become bound to NH₃ to form ammonium, and the urine pH will often rise above 5.5. The combination of a hyperchloremic, hypokalemic metabolic acidosis, a high urine pH, and a positive UAG mimics the findings expected with distal RTA.

Additional concerns — In all steady state conditions, the intake and excretion of electrolytes must be equal. Thus, the daily excretion of Na, K, and Cl must equal their daily intake (and absorption), and the UAG is therefore a function of the absorption of these electrolytes [23]. Changes in dietary composition due, for example, to food additives in which sodium is added and excreted with a non-chloride anion such as bicarbonate, phosphate, nitrate, or benzoate can increase the UAG independent of changes in ammonium excretion. Similarly, among individuals eating a meat-based diet, urinary ammonium can be increased due to the acid load, yet the UAG may also increase as a result of increased potassium ingestion, which is then excreted with non-chloride anions including phosphate and sulfate.

Considering these concerns and the limitations described above, we discourage the use of the UAG as a surrogate for the urine NH₄⁺ concentration.  

These diagnostic difficulties associated with the UAG led to the introduction of another surrogate indicator of urine ammonium: the UOG. (See 'Urine osmolal gap' below.)

URINE OSMOLAL GAP — An alternative method to qualitatively estimate the urine ammonium (NH₄⁺) concentration is the UOG [2,17]. This surrogate marker of urine ammonium generally works well whether the ammonium is excreted together with chloride and/or with any other anions. The relationship between the UOG and ammonium excretion is generally not disrupted by several of the conditions that disrupt the diagnostic utility of the UAG. (See 'Limitations of the UAG' above.)

Under most conditions, the urine osmolality is generated by sodium salts, potassium salts, urea, glucose (if present), and ammonium salts. Thus, if the urine osmolality is actually measured (using an osmometer) and then calculated using the measured concentrations of sodium, potassium, urea, and glucose (if present), the difference should represent the concentration of ammonium salts. The calculated urine osmolality (not including the ammonium salts) is determined with the following formula (calculator 2 and calculator 3) [15]:

The sodium and potassium concentrations are multiplied by 2 to account for the osmotic contributions of their dissociated anions, while the divisors 2.8 and 18 are required to convert the urea nitrogen and glucose concentration units from mg/dL to mmol/L or mosmol/kg. However, if the urea and glucose concentrations are reported in standard units (ie, mmol/L), the following, simpler formula should be used:

The UOG is the mathematical difference between the measured urine osmolality and the urine osmolality calculated with one of the above equations. Since, under most conditions, ammonium salts are the only other major urinary solutes that contribute importantly to the urine osmolality, this result should be proportional to urinary ammonium. The ammonium salts include ammonium chloride (NH₄Cl) as well as ammonium excreted with any "unmeasured" anions such as beta-hydroxybutyrate or hippurate. Thus, the UOG will increase when the urinary excretion of any ammonium salt increases. In contrast to the UAG, the relationship between the UOG and urine ammonium is not disrupted by a high concentration of "unmeasured" urine anions.

Under normal conditions, the UOG falls between 10 to 100 mosmol/kg [14,15,24,25]. If all of the above assumptions were correct, then ammonium excretion would be approximately one-half of this value (due to accompanying anions, assuming they are univalent) and would normally be 5 to 50 mEq/L. However, as was discussed with regard to the UAG, it is best to use the UOG as a gross qualitative, rather than quantitative, measure of the urine ammonium concentration.

Usual clinical application of the UOG — The major renal response to chronic metabolic acidosis is increased ammonium excretion, and this may exceed 200 to 300 mEq/day in patients with chronic severe metabolic acidosis [6,26]. Conversely, a value below 75 mEq/L in a patient with chronic metabolic acidosis suggests impairment in ammonium excretion [14]. Thus:

●A UOG of less than 150 mosmol/kg in a patient with chronic metabolic acidosis suggests that ammonium excretion is impaired [14,15]. Because ammonium excretion is reduced in patients with distal renal tubular acidosis (RTA), a low UOG is consistent with (but not diagnostic of) this diagnosis. Reduced urine ammonium concentrations are also consistent with type 4 RTA.

●When the UOG exceeds 400 mosmol/kg, it is likely that the urine ammonium concentration is 200 mEq/L or greater [15]. This would be expected with hyperchloremic metabolic acidosis generated by chronic diarrhea, by inhalation of toluene, and by other disorders in which the renal response to acidemia remains intact.

Limitations of the UOG — The concerns that have been raised in reference to the UAG were discussed above, and they may also apply to the UOG calculation [27]. (See 'Additional concerns' above.)

Urinary tract infections caused by bacteria that produce urease will disrupt the relationship between the UOG and renal ammonium excretion. Urease catalyzes the formation of ammonium and bicarbonate from urea and water. The UOG will increase, reflecting the high ammonium bicarbonate (NH₄HCO₃) concentration. However, this ammonium is formed outside of the kidney (in the bladder or the collection container) and therefore does not represent renal ammonium or acid excretion.

Urinary excretion of osmotically active, non-ammonium solutes that are not included in the above formulas (eg, alcohols [methanol, ethylene glycol] and mannitol) will increase the UOG even when renal ammonium excretion is not high. Although ethanol is metabolized quite rapidly, toxic alcohols are significantly excreted in the urine, especially during treatment with ethanol or fomepizole. Thus, measurement of the UOG as well as the plasma osmolal gap can be helpful in the diagnosis of toxic alcohol ingestion.

Conversely, the UOG will decrease if urine sodium and potassium do not completely dissociate from their respective anions [27]. Dissociation of sodium and potassium salts is inversely related to the urine osmolality; therefore, the UOG may underestimate renal ammonium excretion when the urine is highly concentrated.

SUMMARY

●Diagnosis of distal renal tubular acidosis (RTA) can usually be made from the history, serum chemistries, and urine pH. Distal RTA is characterized by a urine pH that is persistently above 5.5 despite metabolic acidosis and acidemia. The normal renal response to metabolic acidosis with acidemia is to increase acid secretion and reduce the urine pH to 5.3 or less. However, some patients with normal renal acidification mechanisms who develop metabolic acidosis and hypokalemia may have a urine pH above 5.5. This is due to high levels of urine ammonium (NH₄⁺). If patients with hyperchloremic metabolic acidosis and acidemia have a urine pH above 5.5, measurement of urine ammonium can differentiate those with distal RTA from those with normal renal acidification mechanisms. However, if urine ammonium cannot be readily measured, then the urine anion gap (UAG) or urine osmolal gap (UOG) may be useful in distinguishing RTA from nonRTA causes of metabolic acidosis (such as diarrhea). (See 'Use of urine anion and osmolal gaps in the diagnosis of distal RTA' above.)

●The UAG (sometimes also called the "urine cation gap" or the "urine net charge") is calculated as the difference between the sum of the urine sodium plus potassium concentrations and the urine chloride concentration (calculator 1). The UAG can provide an indirect, or surrogate, estimate of urinary ammonium excretion. However, under certain conditions the UAG may provide misleading information. (See 'Urine anion gap' above.)

●In patients with certain forms of metabolic acidosis (for example, due to diarrhea), renal sodium and potassium excretion falls while urine chloride excretion remains higher. The cation excreted with the chloride in such cases is ammonium. Thus, the urine chloride usually exceeds the sum of urine sodium plus potassium, and the UAG becomes negative (generally less than -20). By contrast, patients with hyperchloremic metabolic acidosis due to RTA have reduced urine ammonium excretion, and the UAG is generally positive (usually greater than 20). In patients with a hyperchloremic (ie, nonanion gap) metabolic acidosis, the UAG is typically used in the following way (see 'Usual clinical application of the UAG' above):

•A positive UAG that is 20 or greater is usually indicative of a low or normal ammonium excretion. Thus, patients with metabolic acidosis due to impaired renal ammonium excretion (such as a distal RTA) will have a positive UAG.

•A negative UAG that is less than -20 is usually indicative of increased ammonium excretion (ie, greater than 80 mEq/L). Thus, the value of the UAG is generally between -20 and -50 mEq/L in patients with metabolic acidosis generated by diarrhea.

●Certain conditions, such as ketoacidosis or toluene inhalation, can result in the urinary excretion of a high concentration of "unmeasured" anions (eg, beta-hydroxybutyrate and acetoacetate or hippurate). In these cases, ammonium is excreted with these anions and not with chloride. The excretion of ammonium with such unmeasured anions will not reduce the UAG as occurs when ammonium is excreted with chloride. When these conditions exist, the UAG may become a misleading indicator of urine ammonium excretion and no longer correctly indicate the etiology of the metabolic acidosis. (See 'Limitations of the UAG' above.)

●An alternative method to qualitatively estimate the urine ammonium concentration is the UOG. This calculation correlates well with urine ammonium whether it is excreted with chloride or any other anion. Thus, the relationship between the UOG and ammonium excretion is generally not disrupted by those conditions that will disrupt the relationship between urine ammonium and the UAG. The UOG is calculated as the mathematical difference between the directly measured urine osmolality and the calculated urine osmolality derived from the urine concentrations of sodium, potassium, urea nitrogen (or urea), and, if the dipstick is glucose positive, glucose concentrations (calculator 2 and calculator 3) (see 'Urine osmolal gap' above and 'Usual clinical application of the UOG' above):

•A UOG of less than 150 mosmol/kg in a patient with chronic metabolic acidosis suggests that ammonium excretion is impaired. Because ammonium excretion is reduced in patients with distal RTA, a low UOG is consistent with this diagnosis.

•When the UOG exceeds 400 mosmol/kg, it is likely that the urine ammonium concentration is 200 mEq/L or greater. This would be expected with hyperchloremic metabolic acidosis generated by chronic diarrhea and other metabolic acidosis that are not due to renal tubular defects.

●If a urinary tract infection caused by urease-producing bacteria exists, then the urine ammonium concentration will increase markedly in the bladder or the collection container. The urine ammonium concentration will no longer reflect renal ammonium excretion. Both direct measurement of urine ammonium and the UOG will then fail as indicators of renal ammonium excretion. The UOG will also become misleading as an index of urine ammonium if osmotically active, non-ammonium solutes are excreted that are not included in the calculated urine osmolality (eg, alcohols [ethanol, methanol, ethylene glycol] and mannitol). (See 'Limitations of the UOG' above.)