Serum vs RBC Magnesium: Which Test Tells You Anything
Your serum magnesium came back normal, and a website told you that means nothing because the real test is RBC magnesium. Half of that is right. Here is what each test measures, where the reference range came from, and which result should actually change what you do.
Evidence strength
Level 2a
Systematic review of cohort studies
Peer-reviewed refs
12
Reading time
10 min
Key Takeaways
- Serum magnesium measures the small fraction of magnesium in blood, which the body defends by pulling from bone. A low value is a definite sign of deficiency. A normal value does not rule it out.
- The usual reference range, 0.75–0.95 mmol/L (1.82–2.31 mg/dL), was derived from the distribution of a 1974 US survey population, not from health outcomes. Two expert groups argue the lower limit should be 0.85 mmol/L.
- RBC magnesium responds to intake, but no outcome-validated reference range exists. 'Target the upper quartile' is a vendor convention, not a finding.
- The reference method is the intravenous loading test: healthy people retain 6.3% on average with a normal range of −19.5% to 27.5%, and retention does not correlate with serum magnesium at all.
- The common causes of a low result are low intake (about half of US adults are below the estimated requirement), proton pump inhibitors (odds ratio 2.0, rising to 7.2 with loop diuretics), diuretics, alcohol, gut losses and diabetes.
- Supplements raise serum magnesium by about 0.05 mmol/L on average, and only in people who start low. Retest at 8–12 weeks, with kidney function, and see a clinician for any value below 0.75 mmol/L or above range.
Serum magnesium is on most routine panels and almost nobody looks at it. When they do, two things happen. The result is normal, which the biohacking internet says is meaningless, or it is low-normal, which a different corner of the internet says is a deficiency that explains everything. The 40-marker panel we publish lists magnesium as a marker worth tracking; this article is the detail behind that line and the correction to it. For what the trials say supplementation then does, see Magnesium for Sleep and Anxiety: What the Trials Show.
What a Serum Magnesium Result Actually Measures
Magnesium sits mostly in bone and inside muscle and other cells. Serum holds a small fraction, and that fraction is regulated: when intake falls, the kidney conserves magnesium and bone releases it, so the serum value can stay inside the range while the body's stores shrink. [3] The German magnesium society's position paper puts the consequence plainly: a low serum magnesium is a definite sign of deficiency, but a value within the reference range does not rule one out. [2]
So the serum test is asymmetric. Low means something. Normal means less than you would like.
That asymmetry does not make the test useless. In a systematic review of 21 supplementation and depletion studies, serum magnesium, red-cell magnesium and urinary magnesium were the three measures that reliably moved when intake changed. [4] It responds. It just responds late and by small amounts.
Reference Range vs the Proposed 0.85 Cut-Off
Most laboratories report a reference range of about 0.75–0.95 mmol/L, which is 1.82–2.31 mg/dL. Those numbers have a history worth knowing. They were derived from the distribution of serum magnesium in the US population surveyed in NHANES I in 1974, which describes where people's values fell rather than which values were healthy. [1]
Two expert groups have argued that the lower bound is too low. A 2016 perspective from magnesium researchers, drawing on cohort and trial data, proposed that values in the lower part of the current range represent subclinical deficiency that contributes to hypertension, diabetes and cardiovascular disease. [1] The 2021 position paper from the German Society for Magnesium Research went further and recommended raising the lower limit to 0.85 mmol/L (2.07 mg/dL), citing associations between values below that and increased health risk. [2]
Read the result in three bands, with the caveat that the middle band is a proposal, not a standard.
| Serum magnesium | How to read it |
|---|---|
| Below 0.75 mmol/L (1.82 mg/dL) | Hypomagnesemia. A cause to find before a supplement to take |
| 0.75–0.85 mmol/L (1.82–2.07 mg/dL) | Inside the laboratory range; low by the proposed evidence-based cut-off. Combined with low intake or a drug that lowers magnesium, a reasonable trigger for a repletion trial |
| 0.85–0.95 mmol/L (2.07–2.31 mg/dL) | Mid-to-upper range. No trial has shown benefit from supplementing here |
| Above range | Usually kidney impairment or supplement excess; needs a clinician |
Units cause errors. Mg/dL, mmol/L and mEq/L are all in use; 1 mmol/L equals 2.43 mg/dL and 2 mEq/L. Check which one your lab reports before comparing with anything you read.
RBC Magnesium and the Loading Test: Better in Theory
Red-cell magnesium is the test most often recommended as the "real" one. The argument is sound as far as it goes: cells hold far more magnesium than serum, so a cellular measure should reflect stores better. The problem is the next step. RBC magnesium does respond to changes in intake [4], but no reference range for it has been validated against outcomes, and the 2018 review of diagnostic methods lists the reliance on assorted assays as a source of confusion in the literature rather than a solution to it. [3] "Target the upper quartile" is a convention of the labs that sell the test. We have removed that advice from our own blood-panel guide.
The reference method is the magnesium loading test. In 88 healthy adults, an intravenous dose of 30 mmol over eight hours followed by 24-hour urine collection gave a mean retention of 6.3%, with a normal range of −19.5% to 27.5%; a depleted person retains more. Two findings matter for reading the cheaper tests: retention had no correlation with serum magnesium, and the test was reproducible within about 8%. [5] So the best measure of body magnesium and the common blood test do not track each other in healthy people. The loading test is a research tool, but it is why nobody should treat a mid-range serum value as proof of full stores, or an RBC value as a substitute.
Who Runs Low: Intake, Drugs and Conditions That Move It
Intake. Almost half of the US population, 48%, consumed less magnesium than the estimated average requirement from food in 2005–2006, and the calcium-to-magnesium ratio of the diet has risen for 30 years. [6] The 2016 perspective puts the figure at about 50%. [1] The requirement is 400–420 mg a day for men and 310–320 mg for women. [12] Legumes, nuts, seeds, whole grains and leafy greens are the sources; a diet built on refined grains, meat and dairy is short.
Proton pump inhibitors. In 9,818 people from the Rotterdam cohort, PPI users had twice the odds of hypomagnesemia (OR 2.00, 95% CI 1.36–2.93). The risk appeared only with prolonged use (OR 2.99) and jumped in people also taking a loop diuretic (OR 7.22, 95% CI 1.69–30.83). [7] A meta-analysis of nine observational studies with 109,798 patients found a pooled risk ratio of 1.43, rising to 1.63 in the higher-quality studies. [8] The FDA issued a safety communication on this in 2011. If you take omeprazole, esomeprazole or a relative daily for more than a year, serum magnesium belongs on your annual panel.
Diuretics. Loop and thiazide diuretics increase urinary magnesium loss; the interaction with PPIs above is the extreme case.
Alcohol, gut loss and diabetes. Heavy alcohol use, chronic diarrhea, malabsorption and poorly controlled diabetes all raise losses. Anyone in those groups with a low result needs the cause addressed, not just the number.
Vitamin D. The enzymes that activate vitamin D are magnesium-dependent, which is why magnesium appears in The Bone & Arterial Health Stack. The reverse direction, that vitamin D changes serum magnesium, is not established.
What Supplementation Does to the Number
Less than people expect. Across 34 placebo-controlled trials with 2,028 participants, a median dose of 368 mg a day for three months raised serum magnesium by 0.05 mmol/L (95% CI 0.03–0.07) and lowered blood pressure by 2.00 mm Hg systolic and 1.78 diastolic. [10] In a seven-week trial of 320 mg of magnesium citrate in 100 older adults, serum magnesium rose only in the 37 participants who started below 1.8 mg/dL (0.74 mmol/L); in the group as a whole, magnesium and placebo were indistinguishable. [9]
That tells you how to use a retest. A rise of 0.05 mmol/L is within the noise of a single measurement, so a repeat value inside the range after supplementing a low result is success; a value that has not moved from 0.70 after twelve weeks is a reason to look for the cause of loss, not to raise the dose.
Whether supplementing changes anything beyond the number depends on who you are. In people with or at risk of diabetes, magnesium lowered fasting glucose (SMD −0.40) and improved two-hour glucose after a glucose load (SMD −0.35) in a meta-analysis of 18 trials. [11] In adults with adequate intake and mid-range serum magnesium, no trial has shown a benefit over placebo for sleep, mood or anything else we have reviewed.
When to Test, When to Act and When to Retest
Test serum magnesium once if you are about to supplement for more than a few months, and annually if you take a PPI, a diuretic, or have diabetes or a gut condition. Order it with creatinine or eGFR, because the safety of supplementing depends on kidney clearance.
Act on a low result by finding the cause first. Below 0.75 mmol/L, the list above is where to look, and a clinician should be involved. Between 0.75 and 0.85 with a poor diet or a magnesium-lowering drug, a trial of 200–400 mg of elemental magnesium a day for 8–12 weeks is reasonable; the forms are compared in Magnesium Glycinate vs Threonate vs Citrate. Above 0.85, supplementing to "optimise" has no trial support.
Retest at 8–12 weeks. Expect a small rise if you started low and none if you did not. Stop the supplement if the result was mid-range to begin with and nothing you were tracking has changed.
Skip RBC magnesium unless a clinician wants it for a specific reason. It will not tell you whether to supplement, and its reference ranges are not tied to outcomes.
When to See a Clinician
Serum magnesium below 0.75 mmol/L on two occasions, any value above the reference range, a low value while taking a PPI or diuretic, muscle cramps or palpitations with a low result, or any kidney disease at all: these are reasons for a clinician to interpret the result and decide on treatment, which may be intravenous rather than oral. Serum magnesium also moves with calcium and potassium, and a low magnesium that will not correct often has a low potassium or calcium behind it that needs its own workup.
Scientific References
- [1]Costello RB, Elin RJ, Rosanoff A, et al.. Perspective: The Case for an Evidence-Based Reference Interval for Serum Magnesium: The Time Has Come — Advances in Nutrition (2016)Oxford 5PMID 28140318
- [2]Micke O, Vormann J, Kraus A, Kisters K. Serum magnesium: time for a standardized and evidence-based reference range — Magnesium Research (2021)Oxford 5PMID 34463286
- [3]Workinger JL, Doyle RP, Bortz J. Challenges in the Diagnosis of Magnesium Status — Nutrients (2018)Oxford 5PMID 30200431
- [4]Witkowski M, Hubert J, Mazur A. Methods of assessment of magnesium status in humans: a systematic review — Magnesium Research (2011)Oxford 2aPMID 22064327
- [5]Gullestad L, Midtvedt K, Dolva LO, Norseth J, Kjekshus J. The magnesium loading test: reference values in healthy subjects — Scandinavian Journal of Clinical and Laboratory Investigation (1994)Oxford 2bPMID 8171268
- [6]Rosanoff A, Weaver CM, Rude RK. Suboptimal magnesium status in the United States: are the health consequences underestimated? — Nutrition Reviews (2012)Oxford 5PMID 22364157
- [7]Kieboom BC, Kiefte-de Jong JC, Eijgelsheim M, et al.. Proton pump inhibitors and hypomagnesemia in the general population: a population-based cohort study — American Journal of Kidney Diseases (2015)Oxford 2bPMID 26123862
- [8]Cheungpasitporn W, Thongprayoon C, Kittanamongkolchai W, et al.. Proton pump inhibitors linked to hypomagnesemia: a systematic review and meta-analysis of observational studies — Renal Failure (2015)Oxford 2aPMID 26108134
- [9]Nielsen FH, Johnson LK, Zeng H. Magnesium supplementation improves indicators of low magnesium status and inflammatory stress in adults older than 51 years with poor quality sleep — Magnesium Research (2010)Oxford 1bPMID 21199787
- [10]Zhang X, Li Y, Del Gobbo LC, et al.. Effects of Magnesium Supplementation on Blood Pressure: A Meta-Analysis of Randomized Double-Blind Placebo-Controlled Trials — Hypertension (2016)Oxford 1aPMID 27402922
- [11]Veronese N, Watutantrige-Fernando S, Luchini C, et al.. Effect of magnesium supplementation on glucose metabolism in people with or at risk of diabetes: a systematic review and meta-analysis of double-blind randomized controlled trials — European Journal of Clinical Nutrition (2016)Oxford 1aPMID 27530471
- [12]Office of Dietary Supplements, National Institutes of Health. Magnesium — Fact Sheet for Health Professionals — NIH ODS (2022)Oxford 5View source
Follow us on Google Search
Add BiohackingHub as a preferred source to see more of our research in Google results.
Related protocols
Partner with BiohackingHub
Editorial placements, sponsorships, and product features for evidence-minded brands.