Magnesium

The usual blood tests, which measure magnesium in serum, do not reflect the body's real status well: less than 1% of the body's magnesium circulates in the blood. Half lives in the bones and the rest in muscles and other tissues. We will come back to this.

The primary source of magnesium should not be a supplement but food: pumpkin seeds, almonds, green leaves, legumes, dark chocolate... Supplementation makes sense in specific scenarios we will see shortly.

The name has a geographical history: Magnesia was a region of Greece that owes its name to an ancient tribe known as the magnetes or magnesians. This region was rich in magnesium compounds, a root that also named manganese and magnetism.

Magnesium is the obligate counter-ion of ATP, the body's energy currency: every enzyme that uses or produces energy — those that move the muscles, those that maintain the voltage of each cell — needs magnesium to work. It also modulates the brain's NMDA receptors (key in memory and learning), antagonizes calcium in muscular and vascular channels, and is a cofactor of the enzymes that activate vitamin D — without enough magnesium, vitamin D performs worse. Once this biochemical centrality is understood, it stops being surprising that its deficiency is associated with such disparate symptoms.

Magnesium is one of the few supplements with several uses backed by solid evidence, and at the same time one of those that carries the most inflated expectations: where it really works and where the effect is modest or nil.

The evidence is solid on four fronts: prevention of chronic migraines, modest reduction of blood pressure in mild hypertension, a mild laxative effect, and irreplaceable hospital uses (eclampsia, certain arrhythmias, severe asthma) given by injection under supervision.

Special mention goes to sleep, probably the no. 1 reason many people come to magnesium. It helps older people with insomnia or with a deficiency, but it is not a sedative: do not expect to fall asleep better from a one-off dose.

The most rigorous studies suggest that supplementing magnesium without a prior deficiency does not improve aerobic performance, and may even slightly worsen it. For hypertrophy or injury recovery there is no solid evidence. The headline “more magnesium = better athlete” does not hold.

Magnesium in the form of L-threonate has been studied for deep sleep and for mild memory complaints, with positive results that have not yet been confirmed by independent teams.

The indication with the best balance between evidence and clinical use is the prevention of chronic migraines. The American Headache Society and the American Migraine Foundation still recommend it at doses of 400–500 mg a day (the classic guidelines went up to 600 mg, a dose still used in trials). The pivotal studies — Peikert 1996 with trimagnesium dicitrate and Köseoglu 2008 with citrate — documented reductions of between 33% and 42% in attack frequency. These doses exceed the recommended limit for supplements (350 mg/day), so they are prescribed with medical follow-up.

For mild hypertension the large meta-analyses calculate mean reductions of 2 to 3 mmHg in systolic pressure. The magnitude rises to about 7–8 mmHg in treated hypertensives or hypomagnesemic patients, where supplementation takes on a clear therapeutic role. In 2022 the FDA issued a qualified statement rating the overall evidence as “inconsistent and inconclusive”.

Sleep is the terrain where marketing has pushed hardest. The classic RCT by Abbasi (2012), with magnesium oxide in older people with insomnia, showed significant improvements in latency, efficiency and sleep time, plus biological changes (more melatonin, less cortisol). The modern RCT by Hausenblas (2024) with L-threonate found improvements in the insomnia index and in deep sleep measured with an Oura ring — but the study is funded by the distributor of Magtein® and three of its authors are employees of the manufacturer. The derived systematic review (Mah & Pitre 2021) rates the overall quality of the evidence as low. In short: it can help deficient people and older people with insomnia; in healthy people without a deficiency the documented effect is marginal.

In anxiety and depressive symptoms there is a single strict positive RCT (Rajizadeh 2017) in depressed and deficient people, which suggests the effect depends on baseline status. For type 2 diabetes with low magnesium, Rodríguez-Morán and Guerrero-Romero (2003) documented clear improvements in glycated hemoglobin, although the ADA does not recommend routine supplementation.

As an osmotic laxative, magnesium salts are drugs of predictable and well-established efficacy. In the hospital setting, injected magnesium sulfate is the worldwide standard of care for preventing eclampsia (Magpie Trial 2002, n>10,000) and first line for serious arrhythmias such as torsades de pointes and for refractory asthma crises.

The most rigorous recent trial — Bomar and colleagues (2025) — put 15 healthy trained adults through a triple-blind crossover with magnesium chloride for 9 days. Result: VO₂max dropped significantly and mean sprint power fell by 5%. The most plausible explanation is that in people already replete with magnesium, overloading it interferes with mitochondrial respiration.

For muscle hypertrophy and recovery from musculoskeletal injuries, the search in PubMed, Cochrane and ClinicalTrials.gov returns no double-blind placebo RCTs that meet the criteria. It is not that the evidence is weak: it is that it does not exist.

The only positive ground is recovery after eccentric exercise: Steward (2019) documented in runners with a low baseline magnesium diet less delayed muscle soreness and a lower inflammatory peak (IL-6) after a downhill run. Useful, but contextualized: n=9, men only, low baseline diet.

The only form with specific human evidence of improving cognitive function is L-threonate: Liu (2016) documented in n=44 adults aged 50–70 with subjective cognitive complaints, after 12 weeks, improvements in memory, attention and processing speed. The study is serious, but virtually all the human evidence for this form comes from teams linked to the patent holder (Magceutics, Neurocentria, AIDP).

It is worth asking whether our diet covers the minimum intake before supplementing. The early symptoms of deficiency are nonspecific and shared with many other situations: fatigue, muscle weakness, sleep disturbances, irritability and headache. On their own they diagnose nothing, but in people with poor diets or chronic medication that predisposes to losing magnesium, they are signals not to ignore.

Supplementation makes sense above all in four groups where deficiency is likely or the indication is backed by clinical trials: older people (they eat less and absorb worse); patients who take omeprazole or other stomach protectors chronically (the FDA issued an alert about this in 2011); people with type 2 diabetes (because of urinary losses); and those with specific indications such as chronic migraine or cramps during pregnancy. For a healthy person who eats well, the benefit of supplementing is marginal.

Older people: intake falls with age, intestinal absorption drops and renal excretion rises — a triple loss that explains why the deficiency accumulates without obvious symptoms. Chronic PPI users (omeprazole, esomeprazole, lansoprazole): the mechanism involves an alteration of the TRPM6 transporter in the intestine, and in some cases documented by the FDA in 2011 the deficiency was so severe that it required withdrawing the drug. Diuretics, loop or thiazide, cause sustained renal losses. Type 2 diabetes: glucosuria drags magnesium out in the urine, a sustained loss associated with more prevalent deficiency. Chronic alcoholism: deficiency documented in 30–60% of hospitalized patients. Intestinal malabsorption: Crohn's disease, celiac disease, resection, gastric bypass.

The Recommended Dietary Allowance (RDA) is about 400 mg in adult men and about 320 mg in women, counting everything you ingest from food, water and supplements. A varied diet with seeds, nuts, green leaves and legumes usually covers those figures effortlessly.

When supplementing is needed, the usual useful dose is between 200 and 400 mg a day of elemental magnesium — not of magnesium salt. In a supplement of “500 mg of magnesium oxide” you are not taking 500 mg of magnesium, but a much smaller fraction. Serious labeling states the elemental magnesium separately.

Three practical rules:

Higher doses — such as the 600 mg a day used in some migraine prophylaxis trials — are backed by the evidence, but must be prescribed by a professional because they exceed that limit.

Vitamin D and magnesium need each other: magnesium activates the enzymes that convert vitamin D into its useful form, and vitamin D improves the intestinal absorption of magnesium. And the next section explains in detail why splitting the dose works better than taking it all at once.

Quinolone antibiotics (ciprofloxacin, levofloxacin) and tetracyclines (doxycycline) form non-absorbable chelates with magnesium: the prudent thing is to consult a professional; in general it is recommended to separate the intake by 2 to 4 hours. Oral bisphosphonates for osteoporosis (alendronate, risedronate) require the same separation. Levothyroxine: leave at least 4 hours between taking the hormone and the magnesium so as not to lose absorption. Proton pump inhibitors (omeprazole, esomeprazole, lansoprazole) taken for more than a year can cause significant depletion through alteration of the intestinal TRPM6 transporter — FDA alert 2011. Diuretics, thiazide and loop, cause sustained renal losses. SGLT2 inhibitors (dapagliflozin, empagliflozin) slightly raise serum magnesium, a favorable side effect. Cisplatin and anti-EGFR antibodies cause marked tubular losses that often require scheduled replacement.

Of the magnesium we ingest, only a fraction is absorbed — between 30% and 40% on usual diets. Absorption happens by two routes: the paracellular, passive, predominant when intake is high; and the transcellular, active and saturable through the TRPM6 and TRPM7 transporters, predominant when intake is low.

The most useful detail shown by the classic Fine (1991) study is that the absorbed fraction falls the higher the dose: from 65% at low intakes to 11% at very high intakes. This explains why splitting magnesium into two doses with meals absorbs appreciably more than a single large dose, one of the few practical optimizations with clear support.

The short answer is no. There are more than a dozen forms of magnesium on the market and, although the differences between them are smaller than marketing suggests, they do matter in two things: how much magnesium really enters the body and how well the stomach tolerates it.

For most people who want to cover their day-to-day magnesium, magnesium citrate is the option with the best cost-absorption-tolerance balance. If your stomach is sensitive or you take it at night and do not want any laxative effect, bisglycinate is the “gentle” form par excellence. It has a very salty, bitter taste, hard to disguise. The powder form may be cheaper, but it is worth considering taking it in capsules. Magnesium oxide, paradoxically the best-selling in pharmacies, is the worst absorbed — perhaps useful as an occasional laxative or antacid, but not the best for replenishing magnesium.

The percentage that appears on the label (for example “500 mg of magnesium oxide”) is not elemental magnesium — it is the weight of the whole salt. The elemental magnesium fraction varies a lot between forms, and together with bioavailability and digestive tolerance, that percentage decides how much magnesium really enters the body per capsule.

Magnesium oxide is the market paradox: the highest percentage on the label and the lowest real fraction. Its value is in the predictable diarrhea it produces — useful as a laxative or antacid, not as systemic replenishment. Citrate is the workhorse: good absorption, good tolerance and a reasonable price. Bisglycinate is absorbed by a different dipeptide transporter, which gives it especially good digestive tolerance; it is the “gentle” form when there is a sensitive stomach or you want the calming effect without laxative risk.

L-threonate is the form with the best-documented evidence for reaching the brain by crossing the blood-brain barrier, studied for deep sleep (Hausenblas 2024) and for mild memory complaints (Liu 2016) with positive results.

The conclusion is less spectacular than marketing implies: the differences between forms exist but are minor. What most determines whether supplementation works is the dose and the baseline status, not the chosen salt.

Beyond the seven forms in the table, the market offers other salts with more specific profiles and, in general, scant clinical evidence. The aspartate has bioavailability comparable to chloride but carries a minor controversy about excitotoxicity. The lactate (~12% elemental) is the basis of the European Magnesium Diasporal®, with good absorption. The taurate is sold for the theoretical taurine-magnesium synergy on cardiovascular tone, but that synergy comes from a 1996 hypothesis article; there are no specific RCTs of taurate as a salt. The orotate has a single positive RCT in advanced heart failure (Stepura 2009, MACH study) — single-center, funded by Wörwag and not replicated. The pidolate (PCA-Mg) is popular in France under the Mag2® brand, with good bioavailability. The carbonateis used mainly as an antacid, with absorption dependent on gastric acid. The cross-sectional review by Schuchardt and Hahn (2017) sums up the landscape well: the differences between forms are smaller than marketing suggests, and the factors that weigh most on the clinical outcome remain the dose and the patient's baseline status.

Magnesium is usually extracted from mineral or marine sources, susceptible to contamination with heavy metals, especially lead. ConsumerLab.com has documented over the years products with lead above California's Proposition 65 limit or with real magnesium content below what is declared.

The most recognized certifications are USP Verified, NSF Certified for Sport, Informed Sport and GMP (Good Manufacturing Practices). Each certifies compliance with specific standards: composition matching the label, absence of specific contaminants, or screening for substances banned in sport. In Spain it is uncommon to find them on the supplements for sale.

A selection of the most cited studies with the highest methodological quality (randomized, double-blind, placebo-controlled), prioritizing those published since 2020. Human trials only. At the advanced level you will find, for each study, the full methodological card with population, intervention, results, limitations, conflicts of interest and the link to the original.

The organization by blocks follows the main indications: sleep, anxiety and mood, blood pressure and cardiovascular health, migraine, sport and performance, cognition and type 2 diabetes. When an indication lacks sound RCTs — hypertrophy, injury recovery, fibromyalgia — it is declared explicitly as a gap rather than filled with weaker evidence.

Sleep

Anxiety and mood

Blood pressure and cardiovascular health

Migraine

Sport and performance

Cognition

Type 2 diabetes

Declared gaps

Muscle hypertrophy. The search in PubMed, Cochrane and ClinicalTrials.gov returns no double-blind placebo RCTs evaluating magnesium on muscle hypertrophy measured by DXA, MRI, ultrasound or biopsy in subjects without a baseline deficiency. The only historical candidate (Brilla & Haley 1992, n=26 novice subjects, approximately 30 years ago) does not measure hypertrophy directly and carries serious limitations. The evidence is scant, old and suggestive rather than confirmatory.

Recovery from musculoskeletal injuries. No sound RCTs. The evidence is indirect (the role of Mg as a cofactor in protein synthesis, modulation of IL-6, competition with Ca²⁺ in the sarcoplasmic reticulum). The Cochrane Garrison 2020 documents that not even cramps respond consistently.

Three more myths worth nuancing for the reader who wants to go all the way to the bottom.

There is no simple test to know whether you are deficient: the blood magnesium test is done but reflects the body's real status very poorly. It is not clear how much magnesium each person needs: the optimal dose depends on diet, age, kidney function and medications. And it is not clear whether supplementing someone without a deficiency provides any benefit: the FDA itself issued a statement in 2022 rating the general evidence as “inconsistent and inconclusive”.

No sound RCT has been found for magnesium in muscle hypertrophy, in recovery from musculoskeletal injuries, or in fracture healing in athletes. The status of magnesium in fibromyalgia rests on a single 1995 trial with weak, unreplicated results. In bone quality there are studies on mineral density and remodeling markers, but not on fracture incidence as a primary outcome.

Most RCTs use different chemical forms, which complicates comparison between them. Knowing which form works best for which specific indication remains, surprisingly, an open question.