Serum anion gap: reevaluation of normal values

Abstract

As it is well known, the serum anion gap (SAG) equals the difference between the serum concentrations of the major cation (sodium) and the major measured anions (chloride + bicarbonate), and it is a useful tool for identifying the cause of a metabolic acidosis.1,2 Additionally, a low SAG has been found to be of value when evaluating a variety of conditions, such as multiple myeloma, hypoalbuminaemia, po1yc1onal hypergammaglobu1inaemia and bromism.3 The normal values of SAG are considered to be 12 ± 4 mmol/l (range 8-16 mmol/l). However, it has repeatedly been reported that because of the widespread use of ion-selective electrodes, the range for the SAG should be significantly lowered.4,5 We have prospectively studied 92 individuals (46 male, 46 female) with a mean age of 41 years (range. 18-64 years) admitted to our hospital for elective procedures. Subjects with diabetes mellitus. liver or renal failure, hypoalbuminaemia (serum albumin <40 g/l), as well as subjects receiving drugs affecting acid-base and electrolyte parameters were excluded. In all individuals the SAG was calculated. Serum sodium and chloride levels were determined by ion-selective electrodes, while serum bicarbonate levels were calculated from arterial blood pH and arterial blood carbon dioxide tension according to the Henderson-Hasselbalch equation. The results of our study are as follows: serum sodium level, 139 ± 3 mmol/l (range 136-143 mmol/l); serum chloride level, 104.5 ±4 mmol/l (range 99-110 mmol/l); serum bicarbonate level, 24.5 ± 2 mmol/l (range 21-26 mmol/l); and SAG, 8 ± 2 mmol/l (range 5-11 mmol/l). The SAG level measured was slightly higher than that recently reported.4,5 but significantly lower than the previously regarded normal value. In 36 cases, serum chloride levels were also measured by the mercuric nitrate-thiocyanate colorimetric reaction. By this method, the mean (SD) value of serum chloride levels was significantly lower than that obtained by ion-selective electrode (100.8 ± 3 vs 104 ± 3.5 mmol/l, respectively), while the mean (SD) value of SAG was 12 ± 3 mmol/l, almost identical to the traditionally normal value. We suggest that with ion-selective electrode technology, values of SAG should be re-evaluated. Additionally, the normal values of the serum sodium/chloride concentration ratio, commonly used for the assessment of mixed acid-base disorders, should be modified, as a previously normal value of 1.4 seems to be inappropriately high in the light of the higher serum chloride concentration measured by the newer autoanalysers.