How this instrument works
Only part of the magnesium circulating in blood is free and biologically active; roughly a quarter to a third rides bound to serum albumin, the same carrier protein calcium uses. A standard total magnesium test measures the free and protein-bound portions together, so when albumin drops — a common finding in hospitalized, malnourished, or critically ill patients — the bound share shrinks and the reported total falls even when the free, active magnesium the body actually uses hasn't moved at all.
In 1985, Michael Kroll and Ronald Elin studied the relationship between serum magnesium and serum protein levels and published a regression-based adjustment, expressed in their paper as corrected magnesium (mmol/L) equals measured magnesium (mmol/L) plus 0.005 times (40 minus albumin in g/L). Converting that relationship into the mg/dL and g/dL units on most U.S. lab reports — 0.005 mmol/L times 2.4305 mg/dL per mmol/L of magnesium times 10 g/dL per g/L of albumin — produces the 0.12 coefficient this calculator uses. That 0.12 is this project's own unit conversion, not a number Kroll and Elin printed themselves.
Unlike calcium's well-known Payne correction, magnesium's albumin adjustment never became an accepted clinical standard. A 2022 consensus paper on serum magnesium reference ranges, co-authored by Ronald Elin himself, proposes no albumin-correction formula at all — a notable omission from the very researcher who helped derive one nearly four decades earlier alongside Kroll.
A study comparing directly measured ionized magnesium against total magnesium in patients with intestinal or liver disease found that among cases flagged as hypomagnesemic by total magnesium, roughly two-thirds actually had a normal ionized magnesium level — with the ionized share running lower as albumin ran lower — the fraction that reflects biologically active magnesium and requires its own dedicated blood draw to measure directly. A corrected value that reads low is a reason to consider that test, not a substitute for it.
- Enter Measured total serum magnesium (mg/dL) exactly as reported on the lab panel.
- Enter Serum albumin (g/dL) from the same blood draw — pairing values from different days undermines the adjustment.
- Read Albumin-corrected magnesium (mg/dL): the estimate of what the total would read if albumin sat at the 4.0 g/dL reference point.
- Check Severity tier: 0 marks normal, 1 mild, 2 moderate, and 3 severe, based on the corrected value.
- Treat a low corrected result as a prompt to discuss ionized magnesium testing with the ordering clinician, not as a final diagnosis.
A corrected reading that stays low
A measured magnesium of 0.8 mg/dL with albumin at 1.8 g/dL — well below the 4.0 g/dL reference — starts the correction: 4.0 minus 1.8 leaves a 2.2 g/dL deficit. Multiplying by 0.12 gives 0.264 mg/dL, and adding that to the raw reading, 0.8 plus 0.264, gives 1.064, which rounds to 1.06 mg/dL as the corrected result.
A corrected value of 1.06 mg/dL falls in the moderate band of 0.97 to 1.22 mg/dL, so the calculator reports severity tier 2 (moderate) rather than tier 3 (severe). The correction narrowed the picture but did not erase a genuinely low reading.
That persistence is exactly the pattern worth treating with caution: in a study of patients with intestinal or liver disease, roughly two-thirds of cases flagged as hypomagnesemic by total magnesium alone — the same kind of reading corrected here — turned out to have a normal ionized magnesium level when measured directly. A tier-2 result here is grounds to ask about that direct test, not a reason to stop at the arithmetic.
Questions
Why does low albumin make a magnesium reading look low?
About a quarter to a third of the magnesium in blood rides bound to albumin rather than floating free. When albumin drops, there's less protein available to carry that bound portion, so the lab's total magnesium number falls — even though the free, biologically active magnesium the body actually uses may not have changed at all.
Is the magnesium correction as reliable as calcium's Payne correction?
No, and that difference matters. Payne's calcium correction is a long-standing, widely cited clinical convention. Magnesium's albumin correction comes from a single 1985 regression study and never reached the same status — a 2022 consensus paper on magnesium reference ranges, co-authored by one of the 1985 formula's own authors, does not recommend an albumin correction at all. Treat this calculator's output as a formula some sources use, not an established clinical standard.
Where do the 0.12 and 4.0 in the formula come from?
Kroll and Elin's 1985 study expressed the relationship in metric lab units: corrected Mg (mmol/L) equals measured Mg plus 0.005 times (40 minus albumin in g/L). Converting that into the mg/dL and g/dL units on a typical U.S. lab report produces the 0.12 coefficient shown here — this project's own conversion of their published number, not a figure Kroll and Elin printed themselves.
Can a corrected magnesium result still be wrong?
Yes. A study of patients with intestinal or liver disease found that among cases flagged as hypomagnesemic by total magnesium, roughly two-thirds actually had a normal ionized magnesium level — the direct measurement of biologically active magnesium — and that the ionized share ran lower as albumin ran lower. A low corrected value raises a reasonable question about magnesium status; it doesn't settle it on its own.
What is ionized magnesium, and why does it matter here?
Ionized magnesium is the free, unbound fraction of magnesium in blood — the portion cells actually use — and it requires its own dedicated blood draw and analyzer rather than being derived from a formula. When a result genuinely changes management, especially alongside low albumin, a direct ionized measurement is the more dependable answer; corrected total magnesium is an estimate, not a substitute for it.
What do the severity tiers mean?
Applied to the corrected value, tier 0 is normal at 1.46 mg/dL or above, tier 1 is mild from 1.22 to 1.46 mg/dL, tier 2 is moderate from 0.97 to 1.22 mg/dL, and tier 3 is severe below 0.97 mg/dL, following the classification used in a 2026 clinical review of hypomagnesemia. Serum magnesium reference ranges aren't fully standardized, though: the 2022 consensus paper cited elsewhere on this page proposes a stricter 2.07 mg/dL (0.85 mmol/L) cutoff for hypomagnesemia, so a result this calculator marks tier 0 (normal) could still read low under that more conservative standard.
Can this calculator replace a conversation with a clinician?
No — it reproduces one published arithmetic adjustment and a separate severity classification so the numbers behind a lab report are visible, nothing more. Any magnesium result outside the expected range, corrected or not, belongs in front of the clinician who ordered the test and can weigh the rest of the patient's history and, if needed, order ionized magnesium directly.
References
- Kroll MH, Elin RJ, Clin Chem 1985 — original albumin-magnesium regression (PubMed)
- Rosanoff A, West C, Elin RJ, et al., Eur J Nutr 2022 — magnesium reference-range consensus (PubMed)
- Saha H et al., Clin Chem Lab Med 1998 — ionized vs. total magnesium in intestinal/liver disease (PubMed)
- Papagiannidou A et al., Curr Nutr Rep 2026 — hypomagnesemia clinical update (PMC)
Read this first: This instrument computes a screening figure from population formulas — it is not a diagnosis, and it cannot see the whole picture a clinician can. Use it to inform a conversation, not to replace one.