This test is most useful if any of these apply to you.
Most people who order a basic metabolic panel never look at the anion gap. It is a small number tucked into the corner of the report, and it does something the rest of the panel cannot: it tells you whether unmeasured acids are quietly building up in your body. That signal can show up well before the kidneys, lungs, or metabolism start failing in obvious ways.
Higher anion gap (AG) numbers have been linked to worse outcomes across a wide range of conditions, from kidney disease and diabetes to liver disease and heart failure. In one national study of US adults, rising albumin-corrected anion gap tracked with higher rates of death from any cause, from heart disease, and from cancer. Knowing this number, and watching how it shifts over time, gives you an early read on a process most people never check.
Your blood is electrically neutral, meaning the positive and negative particles in it balance out. A standard chemistry panel measures the most common positive particle (sodium) and two of the most common negative ones (chloride and bicarbonate). The anion gap is the leftover space between them, representing all the negatively charged particles that are not routinely measured, including proteins, phosphates, and various organic acids.
When that gap widens, it usually means an acid that is not measured directly is accumulating. The classic offenders are lactate (from poor oxygen delivery to tissues), ketones (from diabetes or starvation), and acids from kidney failure or certain poisonings. A wider gap is a clue that something is producing acid faster than the body can clear it.
Because the anion gap is calculated rather than measured directly, it depends on three other lab values being accurate. Small shifts in the chloride or bicarbonate measurement, changes in lab equipment, and variations in albumin (a major blood protein) can all push the gap up or down without any real change in your underlying acid-base status.
Albumin is a negatively charged protein that contributes a meaningful share of the gap. When albumin runs low (common in liver disease, kidney disease, malnutrition, or chronic illness), the calculated gap drops even when acid is genuinely building up. This is how a dangerous acid buildup can hide behind a normal-looking number.
An adjusted version, called the albumin-corrected anion gap (ACAG), accounts for this by adding back what the low albumin subtracted. In hypoalbuminemic intensive care patients, ACAG markedly improves the detection of hidden acid buildup that the uncorrected gap misses. If you have low albumin, the corrected version is the more honest read.
Kidney function and the anion gap are tightly linked. As filtration declines, acids that normally get excreted accumulate in the blood, and the gap widens. The trend matters more than any single value.
In a cohort of 1,168 adults with advanced chronic kidney disease, those with the highest anion gap (about 9.2 mmol/L or above on a sensitive assay) had roughly three times the risk of progressing to kidney failure and about five times the risk of dying compared with those at lower levels. In critically ill people with diabetic kidney disease, higher anion gap levels tracked with higher all-cause mortality. Among 1,260 people with acute ischemic stroke, both the anion gap and its albumin-corrected version predicted who would develop acute kidney injury, with the corrected version performing better.
Heart problems often show up in the anion gap because the heart and kidneys are deeply linked, and reduced perfusion produces acid byproducts. A meta-analysis of patients with heart failure found that higher anion gap consistently tracked with higher mortality, making it a simple way to flag who needs closer monitoring.
In 7,787 critically ill people with congestive heart failure, higher albumin-corrected anion gap independently predicted death over the following year, with the link strongest in those with weaker hearts (lower ejection fractions). In 1,248 people with cardiogenic shock, higher anion gap was associated with worse 30-day, 90-day, and one-year survival. After cardiac surgery, those with anion gap above 13 mmol/L had roughly 4.5 times the in-hospital death rate of those at or below 11 mmol/L.
Diabetes raises acid load in two ways: through ketones during poorly controlled disease and through lactate when blood sugar control medications interact with kidney function. In a national cohort of 8,161 adults with type 2 diabetes, higher albumin-corrected anion gap was linked to greater risk of dying from any cause, from cardiovascular disease, and from cancer.
The signal extends to people without diagnosed disease. In a cohort of 94,448 adults, even modest excess weight tracked with higher anion gap and a greater chance of developing high-anion-gap metabolic acidosis. In 44,023 US adults without chronic kidney disease, metabolic syndrome and abdominal obesity were strongly linked to lower bicarbonate, higher anion gap, and greater odds of acid-gap acidosis. This is acid retention happening quietly in people who would not otherwise be flagged as sick.
The liver helps clear lactate and other organic acids. When it falters, acids accumulate. In 3,084 critically ill people with cirrhosis, higher anion gap on admission strongly predicted death over 30 days to a year. In 1,340 people with sepsis layered on top of cirrhosis, those with elevated albumin-corrected anion gap had higher 28-day death rates.
In 23,333 critically ill patients with gastrointestinal bleeding, higher anion gap independently predicted ICU death and 90-day mortality. In acute pancreatitis, an albumin-corrected anion gap above about 21.5 mmol/L was associated with worse short- and long-term survival. The pattern is consistent: across organ systems, the wider the gap, the worse the outlook.
There is no single universal reference range for the anion gap. Reported upper limits across labs vary from 10 to 19 mmol/L, and the width of the normal interval ranges from 2 to 11 mmol/L depending on the equipment and methods used. A study of 222 adults with normal albumin and creatinine using modern ion-selective electrodes found a mean of 6.6 mmol/L with a range of 2.6 to 10.6, considerably narrower than the textbook 8 to 16. Compare your number to the reference range printed on your own report, not to a single textbook value.
These are research-derived orientation ranges drawn from large prognostic studies. They are not clinical cutoffs in the strict sense, and your lab will likely report different numbers.
| Tier | Approximate Range | What It Suggests |
|---|---|---|
| Optimal | Below 11 mmol/L (uncorrected) or below 16 mmol/L (albumin-corrected) | Lower end of typical reference ranges; minimal acid retention seen in healthy adults |
| Watchful | 11 to 16 mmol/L (uncorrected) or 16 to 19 mmol/L (corrected) | Within most reference ranges but at the higher end; worth checking trend and looking at companion values |
| Elevated | Above 16 mmol/L (uncorrected) or above 19 to 21 mmol/L (corrected) | Range associated with higher mortality and complications across many ICU and chronic disease cohorts; warrants investigation |
Source: Sahebjami 1995 (modern ISE range); Alter and Zha 2025 (range variability across labs); MIMIC-IV cohort studies in COPD, heart failure, AKI, and other conditions for prognostic cutpoints. Always compare your results within the same lab over time for the most meaningful trend.
A single anion gap reading can move several mmol/L between draws on the same person, due to normal fluctuation in chloride, bicarbonate, and albumin. One study found individual baseline values can swing by about 4 mmol/L from routine variation alone. Acting on a single reading is risky. Watching the trend across multiple draws is far more informative.
Get a baseline now. If your number is at the higher end of your lab's range, retest in 4 to 8 weeks alongside a full chemistry panel and albumin. If you are making changes that affect kidney or metabolic health, retest in 3 to 6 months. Once stable, at least annual rechecks are appropriate, and more often if you have diabetes, kidney disease, liver disease, heart failure, or are using medications that can shift acid balance.
A high anion gap is a signal, not a diagnosis. The next step is to figure out what is widening it. The standard workup includes serum lactate, ketones (especially beta-hydroxybutyrate), creatinine and eGFR for kidney function, glucose and HbA1c for diabetes, albumin (so you can calculate the corrected gap), and a venous or arterial blood gas if you want to confirm acidosis.
If the elevation is mild and isolated, retest with a fresh draw and check albumin. If the elevation is consistent and you have known kidney, liver, or metabolic disease, work with a nephrologist or endocrinologist on the underlying driver. If the gap is markedly elevated and you have symptoms (fatigue, rapid breathing, confusion, abdominal pain), this is an emergency, not a wait-and-watch situation.
The anion gap depends on three other lab values, so anything that distorts those distorts the gap.
Evidence-backed interventions that affect your Anion Gap level
Anion Gap is best interpreted alongside these tests.