Method-specific Reference Intervals for Serum Anion Gap and Osmolality

¹ Department of Pathology, University of Mississippi, Medical Center, 2500 North State St., Jackson, MS 39216,
² Louisiana State University, Medical Center, P. O. Box 33932, Shreveport, LA 71130
^a Author for correspondence. Fax (601) 984-2885; e-mail wroberts{at}pathology.umsmed.edu.

To the Editor:

We wish to comment on reference intervals for the serum anion gap and osmolality. The serum anion gap, which is most commonly defined as AG = Na-Cl-TCO₂, traditionally has had a reference interval of 8–16 mmol/L with a mean value (i.e., mean of values used to determine its reference interval) of 12 mmol/L (1)(2). In a recent report (3), the stated reference interval for the anion gap was 12–16 mmol/L with potassium included in the calculation, approximately equivalent to 8–12 mmol/L without inclusion of potassium. This range of only 4 mmol/L appears to be too narrow for a healthy human population. Moreover, recent studies have shown that mean values for current analyzers often differ from the traditional value and may vary from 5.9 mmol/L for the Beckman Synchron to 12.4 mmol/L for the Nova analyzer (4)(5)(6). Consequently, depending on the analyzer used, the serum anion gap reference interval may be as low as 2–10 mmol/L or as high as the traditional range of 8–16 mmol/L. We have also found marked interlaboratory differences in measurements by the same type of analyzer (6). Thus, interpretation of the anion gap requires that each clinical laboratory determine its own reference interval.

The reference interval for serum osmolality also deserves study. One recent report (3) included a reference interval of 270–290 mOsm/kg. We believe this interval is inappropriately low, based on our duplicate measurements of serum osmolality by freezing point depression (Advanced Instruments Model 3MO osmometer) in 59 healthy volunteers of ages 20–60 years. Kolmogorov-Smirnov analysis indicated that we could not reject the hypothesis that the results were normally distributed (P = 0.64). The mean serum osmolality was 288 mOsm/kg, with a SD of 4.1 mOsm/kg, yielding a parametric reference interval of 280–298 mOsm/kg. The lowest and highest values measured were 280 and 298 mOsm/kg, respectively. Other published serum osmolality reference intervals based on freezing point depression measurements are 275–290 or 275–295 mOsm/kg in children and adults and 280–301 mOsm/kg in adults >60 years of age (7)(8). Both analytical methodology and characteristics of the reference population, such as age, can contribute to differences in serum osmolality reference intervals. Again, the clinical laboratory must establish or at least verify the serum osmolality reference interval to aid in accurate interpretation.

References

1. Tietz NW eds. Clinical guide to laboratory tests 3rd ed. 1995:56-57 WB Saunders Philadelphia. .
2. Jacobs DS eds. Laboratory test handbook 4th ed. 1996:81-82 Lexi-Comp Hudson, OH. .
3. Eder AF, McGrath CM, Dowdy YG, Tomaszewski JE, Rosenberg FM, Wilson RB, et al. Ethylene glycol poisoning: toxicokinetic and analytical factors affecting laboratory diagnosis. Clin Chem 1998;44:168-177.[Abstract/Free Full Text]
4. Winter SD, Pearson JR, Gabow PA, Schultz AL, Lepoff RB. The fall of the serum anion gap. Arch Intern Med 1990;150:311-313.[Abstract/Free Full Text]
5. Roberts WL, Johnson RD. The serum anion gap: has it really fallen?. Arch Pathol Lab Med 1997;121:568-572.[Medline] [Order article via Infotrieve]
6. Paulson WD, Roberts WL, Lurie AA, Koch DD, Butch AW, Aguanno JJ. Wide variation in serum anion gap measurements by chemistry analyzers. Am J Clin Pathol (in press)..
7. Tietz NW eds. Clinical guide to laboratory tests 3rd ed. 1995:456-457 WB Saunders Philadelphia. .
8. Jacobs DS eds. Laboratory test handbook 4th ed. 1996:172-173 Lexi-Comp Hudson, OH. .