Serum anion gap in conditions other than metabolic acidosis
Michael Emmett, MD
Section Editor:
Richard H Sterns, MD
Deputy Editor:
John P Forman, MD, MSc
All topics are updated as new evidence becomes available and our peer review process is complete.
Literature review current through: Feb 2018. | This topic last updated: Jun 03, 2016.

INTRODUCTION — Determination of the serum anion gap (AG) is primarily used in the differential diagnosis of metabolic acidosis [1-4]. (See "Approach to the adult with metabolic acidosis", section on 'Physiologic interpretation of the serum anion gap'.)

However, the serum AG can also become abnormal in other conditions, a finding that may be of diagnostic importance [1-5].

CALCULATION OF THE ANION GAP AND NORMAL VALUES — The serum anion gap (AG) is calculated from the following formula:

 Serum AG  =  Measured cations - measured anions

Since Na is the primary measured cation and Cl and HCO3 are the primary measured anions (calculator 1):

 Serum AG  =  Na - (Cl + HCO3)

Historically, the normal range for the AG has been 12 +/- 4 meq/L. However, the normal range for each of the individual measurements varies depending upon the specific analyzer utilized, and, as new technology has been introduced, the normal range of the AG has decreased over the past several decades. In several reports, for example, the normal range for the AG range was reported as 6 +/- 3 meq/L [6,7].

In addition, many European countries include the serum potassium in the anion gap measurement, resulting in a normal range that is approximately 4 meq/L higher than the number calculated in the United States. If potassium is included in the formula, the AG calculation becomes:

  Serum AG = (Na + K) – (Cl + HCO3)

As a result, we suggest that each laboratory determines its own reference range for the AG [3,8].

The way that anions and cations in serum are interrelated is shown in the figure (figure 1). If all the anions and cations are included in the analysis (and their concentrations are reported in meq/L), then the total anions = the total cations. The figure also shows that the anion gap could be determined with the following equation (figure 1):

 Serum AG  =  All unmeasured anions – all unmeasured cations

For the purposes of this relationship, unmeasured ions are defined as any ion in the serum other than Na, Cl, or HCO3.

This relationship demonstrates that an increased AG can be due to an increase in unmeasured anions or a reduction in unmeasured cations. Conversely, a low or even negative AG can develop if there is a reduction in unmeasured anions or an increase in unmeasured cations [2,4,9].

The major unmeasured anion in serum is albumin. Albumin has many positive and negative charges due to various amino acid side chains that can either release or bind hydrogen ions. The ratio of albumin's positive and negative charges changes with the pH of the solution. At pH 7.4, albumin carries about 20 more negative charges than positive charges (therefore each molecule has a net charge of -20). Other unmeasured anions in serum include phosphate, urate, and sulfate. Unmeasured cations include potassium (in the United States, where potassium is not included in the anion gap calculation), ionized calcium, magnesium, and certain abnormal proteins.

HIGH SERUM ANION GAP — An elevated serum anion gap (AG) is most often due to an increase in unmeasured anions, and this is almost always caused by one of the organic metabolic acidoses (eg, lactic acidosis, ketoacidosis). (See "Approach to the adult with metabolic acidosis".)

Much less commonly, the AG elevation is due to hyperalbuminemia, hyperphosphatemia, or an anionic paraprotein (usually an IgA monoclonal immunoglobulin) [1-4,10-13]. Although a reduction in unmeasured cations could theoretically also raise the AG, this generally does not occur because reductions of potassium, ionized calcium, or magnesium will have a relatively trivial quantitative impact on the AG [3,4].

A small elevation in serum AG also occurs in patients with metabolic alkalosis [13-16]. A study in dogs showed that chronic severe diuretic-induced metabolic alkalosis, which increased the serum bicarbonate to 40 meq/L, was associated with an 8 meq/L increase of the AG [14-16]. At least three factors may contribute to the increase in serum AG in metabolic alkalosis, although these factors may only account for a portion of the increased AG [16]:

An alkaline plasma pH causes albumin molecules to release hydrogen ions (a protein buffering effect), and this increases the net negative charge on each albumin molecule. Thus, for any given albumin concentration, the anionic charge will be greater with alkalemia.

Volume contraction, which usually accompanies the most common forms of metabolic alkalosis, will raise the absolute albumin concentration and its negative charge contribution.

Alkalemia increases the generation and accumulation of organic acids, such as lactic acid, which mitigates the increase in HCO3 and decreases the severity of alkalemia, while increasing the AG.

Lastly, any artifactual increase in sodium concentration or artifactual decrease in the chloride and/or bicarbonate concentration will cause the AG to artifactually increase [17,18].

LOW SERUM ANION GAP — The definition of a "low" serum anion gap (AG) must be determined by each laboratory but is generally a value less than 3 meq/L [3,8]. The most common cause of a low serum AG is laboratory error. For example, in a single-center study of 67,740 consecutive measurements, a low serum AG was noted in 304 patients (0.8 percent) [16]. Of these, 285 (94 percent) could not be confirmed upon repeat measurement and were therefore attributed to laboratory error.

A verifiably low serum AG can occur in the following settings [1,3,4,6]:

A decreased concentration of unmeasured anions, which is most often due to hypoalbuminemia. The serum AG falls by about 2.5 meq/L for every 1 g/dL (10 g/L) reduction in the serum albumin concentration [1,2,6,19]. The expected normal values for the serum AG should be adjusted downward in patients with hypoalbuminemia. Alternatively, the AG can be "corrected" for hypoalbuminemia by adding 2.5 x (4 - [serum albumin concentration]) to the calculated AG.

A severe normal anion gap (hyperchloremic) metabolic acidosis may reduce the AG. This may occur because protons bind to albumin as the pH falls, which reduces albumin's net negative charge. As an example, when chronic metabolic acidosis was produced in dogs by the infusion of hydrochloric acid (reducing and their serum bicarbonate to 10 meq/L), the AG fell by 5 meq/L [14,15]. However, the calculated change in albumin charge only accounted for a portion of the change [15].

An increase in unmeasured cations [3,6,20]. The latter can occur with hyperkalemia, hypercalcemia, hypermagnesemia, or severe lithium intoxication [1,3,21,22]. For this to occur, however, the increase in the unmeasured cation cannot be accompanied by a proportional increase in an "unmeasured anion." As an example, if magnesium sulfate is administered, the increase of magnesium is approximately matched by an increase in serum sulfate, and the AG does not change. However, if MgCl2 is administered, then the chloride concentration increases without a change in sodium or bicarbonate, and the AG will fall.

Another category of unmeasured cations are monoclonal proteins, especially immunoglobulins of the IgG class [3,13,22,23]; thus, patients with IgG multiple myeloma may have a low serum AG. A reduced serum AG has also been reported in some patients with polyclonal IgG gammopathy [13,24,25].

Bromide can interfere with chloride measurement and cause a spurious marked increase of the "chloride" concentration. As a result, bromide ingestion can generate a substantial "pseudohyperchloremia" and thereby produce a low or negative anion gap. An artifactual reduction in sodium or artifactual increased bicarbonate will also reduce the AG.

NEGATIVE SERUM ANION GAP — In rare cases, the serum anion gap (AG) has a negative value [4-6]. This may be due to a sporadic laboratory error (which is not reproducible) or a reproducible laboratory artifact that affects the measurement of the serum sodium, chloride, and/or bicarbonate:

The serum sodium concentration may be underestimated when severe hypernatremia exists (serum sodium above 170 meq/L). Displacement errors will also generate pseudohyponatremia when an accurate quantitative volume of serum or plasma is required for the assay (flame photometry or indirect ion specific electrode methods) [26-30]. (See "Diagnostic evaluation of adults with hyponatremia", section on 'Patients who might have pseudohyponatremia'.)

The serum chloride concentration may be overestimated (producing pseudohyperchloremia) by one of the following abnormalities:

Marked hyperlipidemia can falsely elevate the chloride measurement when certain colorimetric assays are utilized. This lipid effect on light scattering may cause a marked overestimation of the plasma chloride concentration, occasionally to levels above 200 meq/L [31].

Salicylate intoxication can falsely elevate chloride because the high salicylate levels alter the permeability of certain ion-selective electrodes used for chloride measurement [32-34].

Bromide ingestion and accumulation can cause marked pseudohyperchloremia. Bromide ions are measured as chloride by most clinical chloride analyzers. Furthermore, each bromide ion reacts like three or more chloride ions. Therefore, a serum bromide of 5 meq/L may be measured as an additional 15 or 20 meq/L of "chloride" [35-38].

This phenomenon was seen more commonly in the past when bromide salts were used as sedatives. These included over-the-counter medications, such as Bromo-Seltzer and Miles Nervine. Bromide has been removed from most medications, with the exception of pyridostigmine bromide, which is used for the treatment of myasthenia gravis, and some herbal medications [3,4,35-38]. A study of patients taking pyridostigmine bromide, for example, found that their AG was 8 meq/L lower than in normal subjects [35]. Bromide, as potassium bromide, is also occasionally used to treat refractory seizure disorders in children [39].


The serum anion gap (AG) (calculator 1) can be increased in conditions unrelated to an organic metabolic acidosis, a finding that may be of diagnostic importance. The clinical conditions that cause this include hyperalbuminemia, hyperphosphatemia, anionic paraproteins (usually IgA monoclonal immunoglobulins), metabolic alkalosis, a sporadic laboratory error, or a reproducible artifactual increase in the sodium concentration or decrease in the chloride and/or bicarbonate concentrations. (See 'High serum anion gap' above.)

A low serum AG is usually caused by hypoalbuminemia but may also be produced by a severe normal anion gap (hyperchloremic) metabolic acidosis or an increase in the concentration of unmeasured cations (eg, hyperkalemia, hypermagnesemia, or IgG monoclonal proteins in someone with multiple myeloma). (See 'Low serum anion gap' above.)

A negative serum AG is usually due to a sporadic laboratory error. A reproducible negative serum AG may be generated by marked hyperlipidemia, salicylate intoxication, or bromide ingestion. (See 'Negative serum anion gap' above.)

Use of UpToDate is subject to the Subscription and License Agreement.


  1. Gabow PA. Disorders associated with an altered anion gap. Kidney Int 1985; 27:472.
  2. Rose BD, Post TW. Clinical Physiology of Acid-Base and Electrolyte Disorders, 5th ed, McGraw-Hill, New York 2001. p.583.
  3. Kraut JA, Madias NE. Serum anion gap: its uses and limitations in clinical medicine. Clin J Am Soc Nephrol 2007; 2:162.
  4. Emmett M, Narins RG. Clinical use of the anion gap. Medicine (Baltimore) 1977; 56:38.
  5. Emmett M. Approach to the Patient With a Negative Anion Gap. Am J Kidney Dis 2016; 67:143.
  6. Jurado RL, del Rio C, Nassar G, et al. Low anion gap. South Med J 1998; 91:624.
  7. Winter SD, Pearson JR, Gabow PA, et al. The fall of the serum anion gap. Arch Intern Med 1990; 150:311.
  8. Paulson WD, Roberts WL, Lurie AA, et al. Wide variation in serum anion gap measurements by chemistry analyzers. Am J Clin Pathol 1998; 110:735.
  9. Oh MS, Carroll HJ. The anion gap. N Engl J Med 1977; 297:814.
  10. Murray T, Long W, Narins RG. Multiple myeloma and the anion gap. N Engl J Med 1975; 292:574.
  11. Thatte L, Oster JR, Singer I, et al. Review of the literature: severe hyperphosphatemia. Am J Med Sci 1995; 310:167.
  12. Kirschbaum B. The acidosis of exogenous phosphate intoxication. Arch Intern Med 1998; 158:405.
  13. van Hoeven KH, Joseph RE, Gaughan WJ, et al. The anion gap and routine serum protein measurements in monoclonal gammopathies. Clin J Am Soc Nephrol 2011; 6:2814.
  14. Madias NE, Ayus JC, Adrogué HJ. Increased anion gap in metabolic alkalosis: the role of plasma-protein equivalency. N Engl J Med 1979; 300:1421.
  15. Paulson WD. Effect of acute pH change on serum anion gap. J Am Soc Nephrol 1996; 7:357.
  16. Adrogué HJ, Brensilver J, Madias NE. Changes in the plasma anion gap during chronic metabolic acid-base disturbances. Am J Physiol 1978; 235:F291.
  18. Goldwasser P, Manjappa NG, Luhrs CA, Barth RH. Pseudohypobicarbonatemia caused by an endogenous assay interferent: a new entity. Am J Kidney Dis 2011; 58:617.
  19. Feldman M, Soni N, Dickson B. Influence of hypoalbuminemia or hyperalbuminemia on the serum anion gap. J Lab Clin Med 2005; 146:317.
  20. Goldstein RJ, Lichtenstein NS, Souder D. The myth of the low anion gap. JAMA 1980; 243:1737.
  21. Figge J, Jabor A, Kazda A, Fencl V. Anion gap and hypoalbuminemia. Crit Care Med 1998; 26:1807.
  22. Kelleher SP, Raciti A, Arbeit LA. Reduced or absent serum anion gap as a marker of severe lithium carbonate intoxication. Arch Intern Med 1986; 146:1839.
  23. De Troyer A, Stolarczyk A, De Beyl DZ, Stryckmans P. Value of anion-gap determination in multiple myeloma. N Engl J Med 1977; 296:858.
  24. Qujeq D, Mohiti J. Decreased anion gap in polyclonal hypergammaglobulinemia. Clin Biochem 2002; 35:73.
  25. Keshgegian AA. Decreased anion gap in diffuse polyclonal hypergammaglobulinemia. N Engl J Med 1978; 299:99.
  26. Turchin A, Seifter JL, Seely EW. Clinical problem-solving. Mind the gap. N Engl J Med 2003; 349:1465.
  27. Krish P, Jhaveri KD. The Case ∣ Hyperbicarbonatemia in a patient with Waldenstrom's macroglobulinemia. Pseudohyperbicarbonatemia due to paraproteinemia. Kidney Int 2012; 81:603.
  28. Bern M. Clinically significant pseudohyponatremia. Am J Hematol 2006; 81:558.
  29. Weisberg LS. Pseudohyponatremia: a reappraisal. Am J Med 1989; 86:315.
  30. Aw TC, Kiechle FL. Pseudohyponatremia. Am J Emerg Med 1985; 3:236.
  31. Graber ML, Quigg RJ, Stempsey WE, Weis S. Spurious hyperchloremia and decreased anion gap in hyperlipidemia. Ann Intern Med 1983; 98:607.
  32. Zimmer BW, Marcus RJ, Sawyer K, Harchelroad F. Salicylate intoxication as a cause of pseudohyperchloremia. Am J Kidney Dis 2008; 51:346.
  33. Mori L, Waldhuber S. Salicylate interference with the Roche Cobas Integra chloride electrode. Clin Chem 1997; 43:1249.
  34. Jacob J, Lavonas EJ. Falsely normal anion gap in severe salicylate poisoning caused by laboratory interference. Ann Emerg Med 2011; 58:280.
  35. Faradji-Hazan V, Oster JR, Fedeman DG, et al. Effect of pyridostigmine bromide on serum bromide concentration and the anion gap. J Am Soc Nephrol 1991; 1:1123.
  36. Wacks I, Oster JR, Pérez GO, Kett DH. Spurious hyperchloremia and hyperbicarbonatemia in a patient receiving pyridostigmine bromide therapy for myasthenia gravis. Am J Kidney Dis 1990; 16:76.
  37. Danel VC, Saviuc PF, Hardy GA, et al. Bromide intoxication and pseudohyperchloremia. Ann Pharmacother 2001; 35:386.
  38. Bowers GN Jr, Onoroski M. Hyperchloremia and the incidence of bromism in 1990. Clin Chem 1990; 36:1399.
  39. McTague A, Cross JH. Treatment of epileptic encephalopathies. CNS Drugs 2013; 27:175.
Topic 2347 Version 13.0