The anion gap is a calculated difference between measured positively-charged ions (sodium, potassium) and negatively-charged ions (chloride, bicarbonate) in blood. It reflects unmeasured acids and helps classify acid-base disorders.
A high anion gap usually signals metabolic acidosis from lactic acid, ketones, or toxins; a normal gap with acidosis points to bicarbonate loss. It's a quick diagnostic shortcut for identifying the cause when bloodwork suggests acid-base imbalance.
Optimal Anion Gap: 7-12 mEq/L.
Typically re-tested every 16 weeks.
Most labs report it automatically alongside a basic or comprehensive metabolic panel, so it needs no extra blood. FitScript's reference range is 3 to 12 mEq/L, and its optimal range is 7 to 12 mEq/L.
The anion gap is the difference between the main measured positive electrolyte in your blood and the two main measured negative ones, calculated as sodium minus the sum of chloride and bicarbonate.
Blood carries no net electrical charge. Every positive ion is balanced by a negative one. Your metabolic panel only measures a handful of them, so the arithmetic leaves a gap, and that gap stands in for everything the panel did not count. On the negative side, the biggest contributor by far is albumin, the main protein in blood, followed by phosphate, sulfate, lactate and ketoacids. On the positive side sit calcium, magnesium and potassium. Because albumin carries most of the uncounted negative charge in a healthy person, the gap is partly a protein measurement and partly an acid measurement, which is the single most useful thing to understand about it.
Two practical details change the number you see. Some laboratories add potassium into the formula, which raises the reported value by roughly the potassium concentration, so a result from one lab does not always line up with another. And modern analysers measure chloride with ion-selective electrodes, a method that reads slightly differently from older approaches, which is why published normal ranges have drifted downward over the decades. Your own laboratory's stated range always outranks a textbook figure.
It is also worth remembering what this value is not. It is not directly measured. It is derived from three separate results, so a problem with any one of them, a difficult blood draw, a delayed sample, an unusually low or high sodium, moves the gap without anything having changed in your metabolism.
It matters because it is a free early signal about acid base balance, and acid base balance sits upstream of kidney function, bone turnover, muscle maintenance and fuel metabolism. A result sitting outside the usual range tells a clinician which mechanisms are worth looking at next, long before a symptom would.
In acute care, this calculation is one of the fastest ways to sort out what kind of acid base disturbance is present, because raised unmeasured anions point towards a different set of causes than a normal gap with low bicarbonate does[1]. The same logic applies in slower, less dramatic situations. A gap that creeps upward across repeated panels is a prompt to look at kidney clearance, fuel use and hydration, not a verdict on any of them.
Research in hospitalised patients suggests that higher values are associated with progression of acute kidney injury and with worse outcomes, which is part of why the gap is tracked rather than glanced at once[2]. For someone optimising health rather than managing illness, the value of the marker is mostly as a cross-check. It helps confirm that bicarbonate, chloride and kidney markers are telling a consistent story, and it quietly flags the occasions when they are not.
A result inside the typical laboratory reference range is considered unremarkable on its own, and FitScript uses a slightly different window as an optimization target.
| Marker | Standard range | FitScript optimal range |
|---|---|---|
| Anion gap (serum, calculated) | 3 to 12 mEq/L | 7 to 12 mEq/L |
Reference ranges vary by laboratory, age, sex and the exact formula used, including whether potassium is part of the calculation. FitScript's optimal range is an optimization target, not a diagnostic cut-off, and no single value defines a condition.
What matters more than a single reading is the direction of travel across repeat panels, and whether bicarbonate, albumin, glucose and kidney markers move with it.
A high result usually means unmeasured acid anions have accumulated in the blood, most often ketones, lactate, or the sulfates and phosphates that healthy kidneys normally clear. It describes a chemical pattern, metabolic acidosis with a high anion gap, and does not by itself name a cause.
Symptoms people report. A mildly raised value typically causes nothing at all and is found on routine bloodwork. When acid accumulation is significant, reported symptoms include deep and rapid breathing, nausea, vomiting, abdominal pain, marked fatigue, confusion, thirst and frequent urination. A raised value together with symptoms like these is worth discussing with a clinician promptly rather than waiting for the next scheduled panel.
Common causes.
| Source of unmeasured acid | Typical examples |
|---|---|
| Ketones | Diabetic ketoacidosis, ketoacidosis related to heavy alcohol use, prolonged fasting or very low carbohydrate intake |
| Lactate | Reduced tissue perfusion, severe infection, sustained maximal exercise, some medications in people with reduced kidney clearance |
| Retained kidney acids | Sulfate and phosphate accumulation when kidney filtration falls |
| Ingested substances | Methanol, ethylene glycol and related exposures |
| Other organic acids | Pyroglutamic acid and similar rarer contributors |
Ketoacidosis is the pattern most often behind a markedly raised gap, and research supports measuring ketones directly rather than inferring them, since point-of-care ketone testing has been shown to identify ketoacidosis quickly at the point of first assessment[3]. Glucose is not a reliable gatekeeper here. Studies describe euglycemic diabetic ketoacidosis, in which the gap and ketones rise while blood glucose stays near normal, a presentation reported particularly in people using SGLT2 inhibitor medications[4]. Ketoacidosis associated with heavy alcohol use follows its own pattern, with ketones and a raised gap in the setting of poor intake and vomiting[5]. A case report describes euglycemic diabetic ketoacidosis arising as a complication of severe burns[6].
Lactate is the other major contributor. In critical care, lactate and base deficit are used as endpoints precisely because they reflect whether tissues are actually receiving oxygen rather than whether the monitor numbers look acceptable[7]. That distinction is real. One prospective study found that raising arterial blood pressure with norepinephrine did not improve microcirculatory blood flow, a reminder that a better pressure reading does not guarantee better tissue perfusion[8].
Long-term impact. Short-lived rises that resolve with the cause, a fast, a hard training block, a bout of illness, carry different weight from a value that stays elevated. Persistent elevation alongside falling kidney filtration is associated with worse kidney outcomes[2]. Clinicians also compare how far the gap has risen against how far bicarbonate has fallen, the delta ratio, because a mismatch suggests more than one acid base process is running at once[9].
A low result most often reflects low albumin rather than anything acid related, because albumin supplies most of the uncounted negative charge the calculation is picking up.
Symptoms people report. The value itself produces no symptoms. What people notice belongs to whatever is lowering it, such as the swelling, poor appetite or fatigue that can accompany low protein states, chronic inflammation, liver disease or protein loss through the kidneys or gut.
Common causes. Low albumin heads the list. Because albumin matters so much, clinicians commonly apply a bedside correction that raises the calculated gap when albumin runs below its usual level, and the uncorrected number can otherwise look reassuringly normal.
Long-term impact. The main consequence of a low value is masking. In people with low albumin, a genuine accumulation of ketones or lactate can push the gap from low up into the normal range, where it looks unremarkable. That is one of the clearest arguments for reading this marker next to albumin, bicarbonate and the delta ratio rather than alone, and for repeating a surprising result before drawing conclusions from it.
Quite a lot, which is why trend beats snapshot. Because the value is calculated from three results, anything that moves sodium, chloride or bicarbonate moves it too.
Testing conditions. Hydration status changes all three inputs. A prolonged tourniquet, a difficult draw or a haemolysed sample can distort electrolytes. Samples left sitting before processing can accumulate lactate in the tube. Recent intense exercise raises lactate for a period afterwards. Timing a draw after a rested night and normal hydration gives a more comparable number.
Habits. Extended fasting, very low carbohydrate eating and ketogenic patterns raise circulating ketones, which can nudge the gap upward in someone who is otherwise well. Heavy alcohol intake, especially with poor food intake, is associated with ketone accumulation[5]. Vomiting, diarrhoea and dehydration all shift chloride and bicarbonate.
Medications. Several medication classes influence acid base chemistry. SGLT2 inhibitors have been linked to ketoacidosis occurring at near-normal glucose levels[4]. In cardiac surgery, perioperative metformin use is reviewed carefully because of the risk of lactic acidosis, a rare complication that needs urgent care[10]. Intravenous fluid composition matters as well, and randomised evidence comparing balanced crystalloids with saline in diabetic ketoacidosis has examined exactly this[11], and a randomised trial in children with diabetic ketoacidosis compared normal saline against Ringer's lactate[12].
Never start, stop or change a prescribed medication based on a lab value. A licensed provider evaluates whether any prescription option is appropriate for a given person.
Method differences. A laboratory that includes potassium in its formula will report a higher number than one that does not. Compare results from the same lab wherever possible.
The gap is calculated from the electrolyte results on a standard metabolic panel, so it appears as part of routine comprehensive bloodwork rather than as a separate test. FitScript members have 150+ biomarkers tested per year, which means the inputs behind this calculation are measured repeatedly rather than once.
Atlas, FitScript's AI health advisor, reads the result in context, next to albumin, bicarbonate, chloride, sodium, kidney markers and glucose, and uses that context to build your FitScript from your own biology. Context is the whole point with this marker. A value of 13 mEq/L in someone with normal albumin, normal bicarbonate and stable kidney markers reads very differently from the same value in someone whose bicarbonate is falling.
Where a result sits outside the expected range, or where a trend moves consistently in one direction, the appropriate next step is clinical review. A licensed provider evaluates the full picture and decides whether any treatment, prescription or otherwise, is appropriate.
This calculation only makes sense next to its inputs and its near neighbours. The following markers are the ones most often reviewed with it:
The anion gap is arithmetic, not a measurement. A high result points towards unmeasured acids such as ketones, lactate or retained kidney acids, and a low result usually points towards low albumin or a sampling issue. Neither names a condition on its own, and both are most informative as a trend read beside albumin, bicarbonate and kidney markers.
This article is educational and is not medical advice. Discuss your own results, and any treatment decisions, with a clinician or licensed provider, and seek care promptly for a markedly abnormal result accompanied by symptoms.
Calculators that use Anion Gap:
It means the calculation from your sodium, chloride and bicarbonate results left a larger than usual difference, which suggests unmeasured acid anions such as ketones, lactate or retained kidney acids have accumulated. It describes a chemical pattern rather than naming a condition, so it is interpreted alongside albumin, bicarbonate, glucose and kidney markers. A raised value with symptoms such as rapid deep breathing, vomiting or confusion is worth discussing with a clinician promptly.
FitScript's reference range is 3 to 12 mEq/L, and its optimal range is 7 to 12 mEq/L. Ranges vary by laboratory, age, sex and the formula used, since some labs include potassium in the calculation and report a higher number as a result. FitScript's optimal range is an optimization target, not a diagnostic cut-off.
Yes. Ketones from prolonged fasting, very low carbohydrate eating or heavy alcohol use can raise it, as can lactate from poor tissue perfusion, severe infection or sustained maximal exercise. Reduced kidney clearance of sulfate and phosphate is another common route, and certain ingested substances raise it sharply. Ketoacidosis can also occur while blood glucose sits near normal, so glucose alone does not rule the pattern in or out.
A low result is most often explained by low albumin, since albumin supplies most of the uncounted negative charge in the calculation. Laboratory artefact, certain blood protein disorders, raised calcium or magnesium, and severe hyponatremia are other possibilities. The practical concern is masking, because low albumin can pull a genuinely elevated gap down into the normal range. It is worth reviewing with a clinician alongside albumin.
The value comes from a standard metabolic panel, and fasting requirements depend on what else is being measured on the same draw. Follow the instructions given for your panel. What does help comparability is normal hydration, avoiding a hard training session beforehand, and using the same laboratory each time so the formula and analyser method stay consistent.
Several classes can. Intravenous fluid composition in hospital also matters. Never change a prescribed medication based on a lab value, since a licensed provider evaluates what is appropriate.
The most common reason is the formula. Some laboratories calculate it as sodium minus chloride and bicarbonate, while others add potassium, which raises the reported figure by roughly the potassium concentration. Analyser methods for chloride also differ slightly between labs, which is part of why published normal ranges have shifted over the decades. Compare results from the same lab where possible.
That depends on why it was flagged and what the surrounding results show, and a clinician makes that call. In general, an isolated borderline value in someone feeling well is usually reviewed on the next scheduled panel, while a markedly abnormal value or one accompanied by symptoms is reviewed sooner. Repeat testing is what turns a single number into a trend, which is far more informative.