Are U.S. Magnesium Reference Ranges 50 Years Out of Date?
Researchers just built the first contemporary serum-magnesium reference intervals for Americans since the Nixon era. Here is what that quietly changes about your lab results — and your supplement habit.
The reference range printed next to your magnesium result is, in all likelihood, older than your primary care doctor. It was built from blood drawn between 1971 and 1974, when the average American ate differently, weighed less, and had never heard of a magnesium glycinate gummy. A new analysis of contemporary national survey data argues it is time — finally — to redraw the lines.
In a study published this year in The Journal of Nutrition, researchers led by Keyi Jiao used serum samples from the 2021–2023 cycles of the National Health and Nutrition Examination Survey (NHANES) to construct the first updated U.S. population reference intervals for serum magnesium in roughly half a century. The analytic sample included 787 children and 5,474 adults, and the authors followed the methodology recommended by the International Federation of Clinical Chemistry to derive the new bounds. The headline finding: a contemporary reference interval of 1.70–2.19 mg/dL for boys and 1.64–2.18 mg/dL for girls, with adult intervals stratified by sex, age, and metabolic health.
That may sound like the most inside-baseball revision imaginable. It is not. Serum magnesium is the cheapest, most widely available biomarker for one of the most heavily marketed minerals on the supplement shelf, and the cutoffs your lab uses determine whether your result comes back flagged or shrugged off.
- The old norms are old. U.S. reference intervals for serum magnesium have largely reflected NHANES I data from 1971–1974.
- New intervals exist. A 2026 analysis used 2021–2023 NHANES data to derive contemporary intervals for children and adults.
- Metabolic health matters. Adults with diabetes showed significantly lower mean serum magnesium than metabolically healthy peers.
- Sex differences are real but small. Girls had lower mean serum magnesium than boys; the practical gap is modest.
- Serum is a blunt instrument. It is practical and cheap, but it is not the same as whole-body magnesium status.
- Talk to a clinician. A 'low normal' result is a conversation, not a prescription for pills.
Why a 50-year-old yardstick is a problem
Reference intervals are not laws of nature. They are statistical descriptions of where the middle 95% of an ostensibly healthy population falls on a given test. When the underlying population shifts — in diet, body composition, prevalence of chronic disease, even assay technology — the yardstick drifts out of calibration. The authors note that intervals commonly used in U.S. clinical settings largely reflect data from NHANES I (1971–1974). Fifty years is a long time to keep using the same ruler.
The practical risk runs in both directions. A range that is too generous can wave through people whose magnesium is genuinely depleted. A range built on a healthier past population can also mislabel today's typical reading as suspect. Either way, the lab printout shapes the conversation in the exam room — and increasingly, the algorithm-driven recommendation in your supplement app.
Reference intervals describe where most ostensibly healthy people fall — not what is optimal.
What the new numbers actually say
For adolescents, the updated interval landed at 1.70–2.19 mg/dL for boys and 1.64–2.18 mg/dL for girls (0.70–0.90 and 0.68–0.90 mmol/L, respectively). The sex difference in means was statistically significant but small. For adults, the authors stratified by sex, age, and metabolic health status — total population, metabolically healthy, hypertension, diabetes, and chronic kidney disease — which is itself a meaningful methodological choice. A single 'normal' range pretending that a 25-year-old with no conditions and a 70-year-old with type 2 diabetes should fit the same bell curve has always been a polite fiction.
The more clinically pointed finding sits in the subgroup analysis. Mean serum magnesium concentrations were significantly lower in adult males and females with diabetes compared with metabolically healthy adults. That association does not, on its own, tell us whether low magnesium contributes to diabetes, whether diabetes (and its medications) deplete magnesium, or both. It does tell us that a one-size-fits-all reference interval is doing real interpretive work it was never built to do.
A range built on a healthier past population can quietly mislabel today's typical reading.
The serum-magnesium caveat nobody puts on the bottle
Here is the part the supplement aisle would prefer you skip. Serum magnesium is a practical biomarker for assessing nutritional status in clinical settings — emphasis on practical. Roughly speaking, only a small fraction of the body's magnesium circulates in blood; the bulk sits in bone and soft tissue, and the body works hard to keep serum levels stable even when intake is poor. That means a 'normal' serum result can coexist with suboptimal tissue stores, and a 'low normal' result is not automatically a deficiency diagnosis.
None of that makes the new reference intervals less valuable. It makes them more honest. A contemporary, stratified range gives clinicians a better baseline to ask the next question — about diet, medications, kidney function, and metabolic context — rather than a worse one.
Dietary sources still matter. Serum levels are stabilized partly at the expense of tissue stores.
The grade
The evidence here is moderate, and the claim being graded is narrow: that the U.S. now has a contemporary, methodologically rigorous serum-magnesium reference interval drawn from a large, nationally representative 2021–2023 sample, and that adults with diabetes show lower mean concentrations than metabolically healthy peers. That is what the data support. What they do not support — yet — is a leap from updated lab cutoffs to specific supplementation strategies, dosing, or disease prevention claims. The new ruler is sharper. What you build with it is still up to you and your clinician.