Anion gap from sodium, chloride and bicarbonate
Enter sodium, chloride and bicarbonate from your basic or comprehensive metabolic panel. Bicarbonate is often printed as CO2 or total CO2. Add albumin, if your panel has it, to see the corrected gap.
Educational tool, not a diagnosis. The anion gap is a calculated value, and its reference range depends on the analyzer: modern ion-selective electrode methods commonly give about 3–10 or 4–12 mEq/L, while older references quote 8–16. The bands shown are broad orientation only; compare against the range printed on your own report. Albumin correction adapts Figge et al. (1998), about 2.5 mEq/L of gap per 1 g/dL of albumin: corrected gap = gap + 2.5 × (4.0 − albumin in g/dL). Figge's own reference albumin was 4.4 g/dL, which would add about 1 more. An abnormal anion gap is interpreted by a clinician together with pH, bicarbonate, kidney function and the clinical picture.
How the anion gap is calculated
The anion gap is not measured by the lab. It is calculated from electrolytes already on a basic or comprehensive metabolic panel (BMP or CMP):
Anion gap = Na − (Cl + HCO3)
All three are in mEq/L, which for these ions is the same number as mmol/L, so the formula works on a US or a European report without conversion. Bicarbonate is usually printed as CO2 or total CO2. Blood is electrically neutral; the gap exists because the routine panel measures the main positive ion, sodium, but leaves out several negative ones, above all albumin, along with phosphate, sulfate and organic acids. The gap is a rough measure of those unmeasured anions.
Some laboratories include potassium: Na + K − (Cl + HCO3). Because potassium is normally around 3.5 to 5 mEq/L, that version reads roughly 4 points higher and needs its own reference range. The toggle above switches between the two.
Anion gap calculation: worked examples
| Panel values | Calculation | Anion gap |
|---|---|---|
| Na 140, Cl 104, HCO3 24 | 140 − (104 + 24) | 12 (top of the usual modern range) |
| Na 138, Cl 103, HCO3 27 | 138 − (103 + 27) | 8 (within the usual range) |
| Na 137, Cl 108, HCO3 26 | 137 − (108 + 26) | 3 (low end; check albumin) |
| Na 140, Cl 100, HCO3 14 | 140 − (100 + 14) | 26 (clearly high) |
| Na 139, Cl 105, HCO3 26, albumin 2.4 g/dL | 8 + 2.5 × (4.0 − 2.4) | 12 corrected, from 8 uncorrected |
| Na 140, K 4.0, Cl 104, HCO3 24 | 140 + 4.0 − (104 + 24) | 16 (with potassium; compare to a K-inclusive range) |
Anion gap normal range: why it depends on the lab
There is no single normal anion gap. The value depends on how the lab measures chloride and sodium, and the range has drifted down as analyzers changed. Older references, from the flame-photometry era, quoted roughly 8 to 16 mEq/L. Modern ion-selective electrode analyzers read chloride slightly higher, so their ranges are lower, commonly cited as about 3 to 10 or 4 to 12 mEq/L. A review in the Clinical Journal of the American Society of Nephrology notes that the normal value varies widely, both between methods and between individuals (Kraut and Madias, 2007). In practice, this means a gap of 14 can be flagged high by one lab and pass without comment at another. The range printed next to your result is the one that applies.
| Anion gap | How it is usually read |
|---|---|
| Under 3 mEq/L | Low. Most often low albumin or a lab artifact; rarely paraproteins or lithium or bromide exposure. |
| 3–12 mEq/L | Within the range most modern analyzers report. Check albumin before reading it as reassuring. |
| 12–16 mEq/L | Above many modern ranges but inside the older 8–16 reference. Your lab's range decides which applies. |
| Above 16 mEq/L | High on almost any reference range. Associated with metabolic acidosis from unmeasured acids. |
What a high anion gap suggests
A clearly raised gap suggests that acids the routine panel does not measure have accumulated, using up bicarbonate. This is the pattern called high anion gap metabolic acidosis, and it is associated with diabetic or alcoholic ketoacidosis, lactic acidosis (from severe infection, shock or poor circulation), advanced kidney failure, and certain poisonings, including methanol, ethylene glycol and large doses of aspirin. These are conditions people are usually unwell with, and they are diagnosed by a clinician with pH, blood gases and other tests, not from the gap alone.
A mildly raised gap on a routine panel from someone who feels fine is a different situation. Mild dehydration, lab variation and an older reference range explain many of these, and the lab itself can be the source: Kraut and Madias list laboratory error alongside metabolic causes. A repeat panel, ideally drawn the same way, is the usual next step. Kidney function belongs in the same reading, so the creatinine and eGFR lines on the same report matter, and the BUN/creatinine ratio calculator uses two more numbers from it.
What a low anion gap suggests
Low values are less common and usually less dramatic. Albumin carries a large share of the negative charge the panel does not measure, so when albumin falls, the gap falls with it. Two causes dominate a low anion gap: lab error and low albumin. Rarer causes include high levels of positively charged abnormal antibodies (paraproteins, as in multiple myeloma) and exposure to lithium, bromide or iodide.
Why albumin correction matters
The same albumin effect works in the other direction: a low albumin can hide a raised gap by pulling it back into the normal band. Figge and colleagues measured this in intensive care patients and found that each 1 g/dL drop in albumin lowers the gap by about 2.5 mEq/L (Figge et al., Crit Care Med 1998). The correction this calculator applies is:
Corrected gap = gap + 2.5 × (4.0 − albumin in g/dL)
Figge's own reference albumin was 4.4 g/dL; many calculators round to 4.0, and 4.4 would raise the corrected value by about 1 mEq/L. The correction matters most when albumin is well below normal, as in serious illness, liver disease or significant protein loss. With an albumin around 4, it changes almost nothing.
Reading it on a routine panel
For most people, the anion gap shows up as one more flagged line on an annual CMP. It is interpreted, not acted on alone, and a single value tells you less than the same value on two panels drawn the same way. Preparation affects electrolytes too; the essentials are in how to prepare for a blood test, and how often to get blood work covers how often a metabolic panel is worth repeating.
Already have your bloodwork? Merios reads it next to your watch data, so a value like this sits in the context of your other markers and how it has moved between panels.