mariuscomper.uk RomânăROEnglishENEspañolES

A fact-checked guide, not a marketing chart

Magnesium citrate vs. malate vs. glycinate: which one actually absorbs better

Ten forms of magnesium sit on the shelf, each with a story about absorption. Citrate has four human studies behind it. For bisglycinate, malate, and taurate, this research found no independent absorption trial in healthy adults of comparable size — and that absence is as much the answer a reader is looking for as any number.

Marius Comper · September 2026 · every claim checked against the cited study itself, not a supplement site's summary of it

At a glance: what the evidence shows, by reason and by cost

If you just want a quick answer, here's what the evidence shows for the most common reasons someone looks for a magnesium supplement — not which form is "best," since that's not something the evidence below can settle for everyone at once. The full explanation, study by study, is further down the page.

Everyday, no particular reason

Citrate has four separate human studies behind it, detailed below — more than any other form covered on this page, though one of the four is manufacturer-funded and none compares it head-to-head with bisglycinate or malate. The number of studies alone does not establish that citrate is more effective. At high doses it can loosen your stool.

Sensitive stomach, magnesium has upset you before

In one small trial, bisglycinate was better tolerated than oxide — but in 12 patients with a surgically altered gut, not ordinary sensitive-stomach adults, and not compared head-to-head with citrate. Its "better absorbed" reputation doesn't yet have a trial as large as citrate's behind it.

Occasional constipation

Citrate and sulfate are used clinically as osmotic laxatives, at doses much higher than an ordinary daily supplement — colonoscopy prep, for instance. For an occasional episode, follow that product's own dosing or a pharmacist's guidance rather than a standard supplement dose; see the kidney-disease caution below too.

Want better sleep

Bisglycinate with glycine modestly reduced an insomnia score versus placebo, in a four-week trial in adults reporting poor sleep. Threonate has an interesting animal mechanism — it raises brain magnesium and improves performance on memory tests — but its human evidence for sleep comes from a trial funded by the company that sells it.

Diffuse pain, fibromyalgia-type

Malate was tested for fibromyalgia symptoms, not ordinary muscle cramps. Even there, the 1995 trial's placebo-controlled phase didn't clearly establish a benefit — only a later, open-label phase suggested one. This is the weakest evidence of everything above.

The cost that matters is the price per milligram of elemental magnesium, not the price on the box — and the two can look very different. The calculation: the box's price divided by (elemental magnesium per serving, from the product's own label, times the number of servings in the box) — not the compound's weight, and not the percentage in the table below, which describes the chemical salt only, not a specific product's actual dose. Because manufacturers choose different doses from product to product, two products priced similarly on the shelf can come out very differently once you do that division.

Do you have kidney disease, are you pregnant, or do you take bisphosphonates, tetracycline- or quinolone-class antibiotics, levothyroxine, or an HIV integrase inhibitor? Talk to a doctor or pharmacist before adding any form of magnesium — see the safety section below for why.

From here: the full analysis, study by study.

How much magnesium is actually in the pill

The label says "magnesium citrate 500 mg." The number that matters is the elemental magnesium inside it, not the weight of the whole compound — and that share ranges from 60% down to under 9% across common forms. A 500 mg magnesium oxide tablet contains more than seven times as much actual magnesium as a 500 mg magnesium threonate tablet.

Computed from standard atomic weights (Mg 24.305; O 16.00; Cl 35.45; S 32.07; P 30.97; C 12.01; H 1.008; N 14.01), applied to each salt's own formula — hydrated where that's the common form, as with chloride, lactate, or sulfate.
FormFormulaMolar massElemental magnesium
OxideMgO40.30 g/mol60.3%
CitrateMg3(C6H5O7)2451.12 g/mol16.2%
MalateMgC4H4O5156.38 g/mol15.5%
BisglycinateMg(C2H4NO2)2172.43 g/mol14.1%
GlycerophosphateMgC3H7O6P194.36 g/mol12.5%
Chloride (hexahydrate)MgCl2·6H2O203.30 g/mol12.0%
Lactate (dihydrate)Mg(C3H5O3)2·2H2O238.48 g/mol10.2%
Sulfate (heptahydrate)MgSO4·7H2O246.49 g/mol9.9%
TaurateMg(C2H6NO3S)2272.60 g/mol8.9%
L-threonateMg(C4H7O5)2294.50 g/mol8.3%

The table above shows the arithmetic, not just the result: the number of magnesium atoms in each salt's formula, times magnesium's atomic mass, divided by the whole salt's molecular mass — which is why citrate, with three magnesium atoms in its formula, comes out to 16.2% rather than 5.4%. The NIH's own magnesium fact sheet wasn't used as a source for this table; the numbers are computed directly from the chemical formulas — checkable by anyone with a periodic table. A specific commercial product's label can differ slightly depending on which hydrate the manufacturer used, which is one more reason to read the "elemental magnesium" line on your own label rather than the salt's name alone.

What each study actually shows

Four things matter for every claim here: who was studied, how many, what was measured, and who paid for it. Urinary magnesium excretion is not the same as magnesium retained in tissue; a 12-patient trial in people with surgically altered guts doesn't automatically generalize to a healthy adult; a manufacturer-funded study is still real evidence, but the funding belongs next to the result, not hidden in a footnote.

Oxide, chloride, lactate — and four studies on citrateWell documented

  • Firoz & Graber, 2001 — healthy volunteers, ~250 mg elemental magnesium/day, four commercial forms. Fractional absorption estimated from urinary excretion: oxide ≈4%, chloride/lactate/aspartate ≈9–11%. The study's abstract states no funding source.
  • Lindberg et al., 1990 — in vitro solubility comparison (oxide nearly insoluble in water; citrate soluble across every acidity tested) plus human absorption; citrate more soluble and more bioavailable than oxide.
  • Kappeler et al., 2017 — 20 healthy men, single dose, crossover design; citrate produced significantly higher 24-hour urinary excretion and serum levels than oxide. Funded by Verla-Pharm Arzneimittel GmbH & Co. KG, a pharmaceutical manufacturer; the sponsor was involved in the study's design, not in analyzing the data.
  • Walker et al., 2003 — 46 healthy adults, 300 mg elemental magnesium/day for 60 days, citrate vs. oxide vs. "an amino-acid chelate" (its exact identity couldn't be confirmed as bisglycinate). Citrate and the chelate showed higher urinary excretion than oxide; oxide didn't differ from placebo on any measure.
  • Schutten et al., 2022 — 164 overweight adults, 450 mg/day for 24 weeks, citrate/oxide/sulfate. None of the three forms showed a statistically significant effect on arterial stiffness or blood pressure versus placebo — a negative result on that surrogate measure (arterial stiffness estimates cardiovascular risk, but isn't itself a clinical event like a heart attack or stroke). The trial also measured plasma and urinary magnesium as secondary outcomes, but the comparative result between forms on those measures couldn't be confirmed from primary text during this research.

Bisglycinate: better tolerated, not necessarily better absorbedLimited evidence

  • Schuette, Lashner & Janghorbani, 1994 — 12 patients with prior ileal resection (a surgically altered gut, not healthy adults), an isotope-labeled dose, crossover design. Across the whole group, no statistically significant difference from oxide was found (23.5% vs. 22.8%); in the four patients with the worst oxide absorption, bisglycinate was clearly better (23.5% vs. 11.8%) and better tolerated by all 12.
  • Schuster et al., 2025 — 155 adults reporting poor sleep, 250 mg elemental magnesium plus 1,523 mg glycine/day, four weeks. A small but statistically significant drop in insomnia score versus placebo (p=0.049). This is a symptom trial, not an absorption trial — it doesn't establish superior absorption. The exact funding source couldn't be confirmed from primary text during this research.
  • Beyond these two trials, this research found no independent, manufacturer-free absorption trial in healthy adults of comparable size to citrate's. The studies identified here do not establish superior absorption in healthy adults.

Malate: symptom data, not absorption dataVery limited evidence

  • Abraham & Flechas, 1992 — small open-label study (~15 fibromyalgia patients), on a combined magnesium-and-malic-acid product.
  • Russell et al., 1995 ("Super Malic") — ~24 fibromyalgia patients, double-blind, placebo-controlled crossover, followed by an open-label phase. Measured tender-point index and myalgia score — not magnesium absorption.
  • Neither study measured urinary or serum magnesium. The claim "malate absorbs well" rests on it being another organic-acid salt, by analogy with citrate — not on a direct absorption trial of malate itself.

Taurate: no identified human trial of the combined saltNo identified human trial of the combined salt

  • Animal studies have tested magnesium taurate as the combined salt. In humans, the available evidence comes from elsewhere: epidemiological studies that treat magnesium and taurine as separate nutrients, not the combined salt (cross-country correlations between taurine and magnesium intake and cardiovascular risk, for instance).
  • Combining these two literatures as proof of a demonstrated cardiovascular effect of the combined salt in humans isn't supported by the current data.

L-threonate: an interesting animal mechanism, thin human evidence with a declared financial interestPromising mechanism, thin human evidence

  • Slutsky et al., Neuron, 2010 — rats and mice; oral threonate raised cerebrospinal fluid and hippocampal magnesium and improved performance on learning and memory tasks. Senior author Guosong Liu holds the foundational patent, with co-inventor Fei Mao, and founded the company that commercializes it.
  • Liu et al., 2016 — 51 randomized (25/26), 44 completed (23/21), adults aged 50–70 with subjective memory complaints, sleep disturbance and anxiety (not a clinical diagnosis of mild cognitive impairment), 12 weeks. Measured plasma, urinary, and red-blood-cell magnesium alongside cognitive tests; the threonate group showed cognitive improvements reported as statistically significant versus placebo. The same financial-interest note applies, since the author is the same.
  • Frontiers in Nutrition study, published 12 January 2026 — 100 randomized, 50 per arm, only four discontinuations, six weeks. The dose provided 145 mg elemental magnesium/day. It measured cognition and sleep quality, not any magnesium biomarker (serum, urinary, intracellular) — this isn't an absorption study. Funded by Threotech Inc., which also supplied the intellectual property and was involved in conceiving the study design.
  • No independent human trial not involving the compound's inventor or manufacturer was found.

One number that stands out

+8.4 cognition points, versus +5.6 on placebo

In the Frontiers in Nutrition trial on L-threonate, the NIH Toolbox cognition composite score rose 8.4 points in the threonate group versus 5.6 points in the placebo group (p=0.043); working memory improved separately, with its own statistical significance (p=0.033).

Threonate+8.4
Placebo+5.6

Worth knowing before drawing a conclusion

  • Funded by Threotech Inc., the company that sells threonate, and designed with the sponsor involved.
  • Run on healthy adults reporting poor sleep — not people with a diagnosed memory problem.
  • Measured no magnesium biomarker (serum, urinary, intracellular), so it doesn't confirm the mechanism proposed from animal studies.
  • A single trial, not yet confirmed by an independent replication.

What the reviews say

  • Pardo et al., 2021 — systematic review, 14 studies included out of 433 screened; the abstract states no funding source. Conclusion: organic forms tend to be more bioavailable than inorganic ones, and absorption declines as dose rises; effectiveness in elderly or already-deficient people isn't assured. It's not a ranking among organic forms — too few of the 14 studies compare organic forms against each other rather than each against oxide.

Why some forms cause diarrhea and others don't

Unabsorbed magnesium left in the gut draws water into the colon — that's the mechanism behind the laxative effect. At high doses, citrate and sulfate are used deliberately as osmotic laxatives (colonoscopy prep, for instance). That's not the same as solubility: oxide is poorly soluble and poorly absorbed, but citrate is highly soluble and can still act as a laxative at high doses — the two effects shouldn't be folded into one explanation.

Bisglycinate is marketed as gentler on the premise that magnesium bound to glycine may be absorbed through an amino-acid transport route rather than as a free ion in the gut. The mechanism is plausible and has some direct support — Schuette's 1994 trial found better tolerability in all 12 patients tested — but a head-to-head trial at matched elemental doses, using a standardized digestive-symptom score, wasn't found in this research. Treat the tolerability claim as plausible, not proven.

Official thresholds differ by authority. The US National Institutes of Health set the upper limit for supplemental magnesium at 350 mg/day, based on the diarrhea threshold. The European Food Safety Authority uses a stricter figure: 250 mg/day, derived from a no-observed-adverse-effect level of 250 mg and a lowest-observed-adverse-effect level of 360–365 mg, where a minority of adults experienced mild diarrhea. Both limits apply only to magnesium from supplements or fortified sources, not to dietary magnesium from food, which healthy kidneys clear efficiently.

When magnesium becomes a real risk

The kidney is the main route of magnesium clearance. In chronic kidney disease or acute kidney injury, magnesium from any form — including magnesium-containing laxatives and antacids, often a larger and less obvious source than a supplement pill — can cause hypermagnesemia: weakness, low blood pressure, arrhythmia, and at extreme levels, respiratory depression. That's the real-world risk, not a generic "ask your doctor" line. Bowel obstruction or significantly impaired gut motility adds a separate risk: it extends contact time between unabsorbed magnesium salt and the gut wall, which can raise how much is absorbed from a single dose.

Magnesium forms insoluble complexes in the gut with oral bisphosphonates, tetracycline- and quinolone-class antibiotics, levothyroxine, and some HIV integrase inhibitors (dolutegravir, for instance). Each class has its own timing rule — for antibiotics, roughly 2 hours before or 4–6 hours after magnesium — but bisphosphonates, levothyroxine, and integrase inhibitors each carry their own separate dosing instructions. Check the specific medication's label or ask a pharmacist rather than applying one interval to all of them.

Long-term use of proton pump inhibitors is associated with reduced intestinal absorption of magnesium, with the kidney trying to compensate by conserving it — not a primary defect in renal reabsorption. Supplementation alone sometimes fails to correct the deficit, and reviewing the medication, occasionally stopping or switching the PPI, needs clinical supervision.

This page doesn't recommend one form. It lays out the evidence for each on the axes that matter to a decision — tolerance, elemental density, strength of evidence — and leaves the reader to match it against their own kidney function (with a clinician), medication list, and budget.

How each claim was checked

Every comparison on this page is tied to a named study, its design, and exactly what it measured — urinary excretion, serum magnesium, intracellular magnesium, or a clinical outcome. Its number of participants and funding source appear too, where they could be confirmed from primary text; a manufacturer-funded study isn't excluded, but the funding sits next to the result, not in a separate note. Where an independent human trial for a frequently repeated claim wasn't found — bisglycinate's or malate's supposed absorption edge, for instance — the page says so directly instead of repeating the claim as settled fact.

A few figures commonly quoted elsewhere don't appear here because they couldn't be confirmed from primary text during this research: the exact identity of the amino-acid chelate Walker tested in 2003, the secondary urinary-magnesium comparison from Schutten's 2022 trial, and the exact funding source of Schuster's 2025 trial. Where a specific figure couldn't be verified, the page either omits it or marks it explicitly as unconfirmed.

The sources, directly

Every study cited above, linked to its primary source.