Magnesium L-threonate

Magnesium L-threonate (MgT, also called L-threonic acid magnesium salt or L-TAMS) is a magnesium salt of L-threonic acid, a breakdown product of vitamin C, sold under the brand name Magtein as a supplement for memory and sleep. It was developed by the neuroscientist Guosong Liu and colleagues, whose 2010 study in rats reported that it raised brain magnesium where ordinary magnesium supplements did not, and improved learning and memory. The human evidence consists of a handful of small, short trials, and every placebo-controlled trial of cognition or sleep has been funded by a company that sells the ingredient.

§How it works

Inside neurons, magnesium blocks the channel of the NMDA receptor, a glutamate receptor central to synaptic plasticity, in a voltage-dependent way. Liu's hypothesis is that raising magnesium inside neurons increases the number of functional synapses and makes plasticity more selective. In the founding study, rats given MgT orally performed better on tests of learning, working memory and short- and long-term memory, and aged rats improved on a pattern-completion task. Treated rats had a higher density of presynaptic markers in two hippocampal subregions, and the density correlated with memory performance.¹ Contemporary coverage reported that MgT raised magnesium in the rats' cerebrospinal fluid by 7 to 15 percent after 24 days, while most other oral magnesium compounds tested had no appreciable effect on brain magnesium.²

The rationale for the threonate salt specifically is that magnesium transport from blood into cerebrospinal fluid is tightly regulated, so ordinary salts raise blood levels without changing the brain. The developers' later work in cultured neurons proposed that threonate itself helps carry magnesium into the cell. Chemically, the compound is made by oxidising ascorbic acid and substituting magnesium for calcium, and commercial batches contain roughly 7.5 to 8.8 percent magnesium by weight.³ A 2 g daily dose therefore supplies about 145 mg of elemental magnesium, less than half of the adequate intake that European regulators set for adults.

§Development history

1. 2004
   Plasticity in culture
   Liu's laboratory at MIT reports that adjusting magnesium's block of NMDA receptors durably enhances synaptic plasticity in cultured hippocampal neurons.
2. 2010
   The rat study
   Slutsky, Liu and colleagues report in Neuron that oral MgT raises brain magnesium, hippocampal synapse density and memory in young and aged rats.
3. 2012
   Patents granted
   Two US patents on magnesium compositions for cognitive function and neurological disorders are granted; the ingredient is developed by Magceutics and distributed exclusively by AIDP.
4. 2016
   First randomized trial
   A 12-week trial of 51 adults aged 50 to 70 with cognitive complaints, funded by the developer, reports an improved composite cognitive score.
5. 2021
   ADHD pilot
   An open-label study of 15 adults with ADHD reports symptom response in just under half of participants.
6. 2024
   EU authorisation
   EFSA judges the compound safe as a magnesium source in adult food supplements; the European Commission authorises it as a novel food in October.
7. 2026
   Largest trial of MgT alone
   A six-week trial in 100 younger adults with poor sleep reports a small cognitive advantage and an estimated 7.5-year reduction in cognitive age.

The ingredient has been commercial for most of its history. Neurocentria, which trades as Magceutics, owns the patents; AIDP held exclusive distribution rights by 2012;⁴ and ThreoTech, whose website lists Liu as a part owner, is now the exclusive licensee and global supplier.

§Human trials

Older adults. The first randomized trial enrolled 51 adults aged 50 to 70 with self-reported memory and concentration complaints, sleep difficulty and anxiety, of whom 44 completed it. Over 12 weeks a composite of four cognitive tests improved more on the supplement than on placebo (p = 0.003, Cohen's d = 0.91). Several individual tests did not separate from placebo, and the authors could not determine any effect on sleep or anxiety because of strong placebo responses. The study was funded by Neurocentria, which co-designed it, ran the statistics for the cognitive tests and wrote the paper; the first author was Liu.⁵

Clinical populations. Two open-label pilots, without placebo groups, followed: one in 15 adults with ADHD,⁶ and one in 15 people over 60 with mild to moderate dementia, a population relevant to Alzheimer's disease, using a product that also contained vitamins C and D.³

Healthy and sleep-troubled adults. A 30-day trial in 109 healthy Chinese adults tested a formula combining MgT with phosphatidylserine and vitamins, so its memory result cannot be attributed to MgT alone; it was funded by Wu-xi Magtein Bioscience.⁷ A 21-day trial in 80 adults aged 35 to 55 with self-reported sleep problems, funded by AIDP and with three of its six authors at AIDP, found that both groups improved on the Insomnia Severity Index with no significant difference between them, while several sleep-stage scores from a finger-worn ring favoured MgT.⁸

The largest trial of MgT alone, by Lopresti and Smith, randomized 100 adults aged 18 to 45 with self-reported dissatisfied sleep to 2 g of Magtein or placebo for six weeks. The primary outcome, the NIH Toolbox total cognition composite, improved more on the supplement (p = 0.043), with the clearest differences on working- and episodic-memory tests; the fluid and crystallised composites and Raven's matrices did not differ. One of three self-reported sleep scales improved, ring-measured sleep showed no benefit, and sleeping heart rate fell slightly while heart-rate variability rose. The study measured no magnesium levels. It was funded by Threotech, which supplied the product, and run by a contract research organisation, led by the first author, that has received funding from nutraceutical companies.⁹

What "cognitive age" means here

Neither "brain age" figure is a scan or a biological measure. Liu's 9-year figure came from one test, trail-making speed, set against published age norms.⁵ Lopresti and Smith's 7.5 years, labelled an exploratory analysis, divides a 2.24-point group difference on the cognition composite by an assumed decline of 0.3 points per year after age 20; a figure caption in the same paper gives 7.8 years. Both are restatements of a test-score difference, not measures of Biological age.

§Magnesium and cognitive aging

The wider interest in magnesium and the aging brain rests on observational data. A 2024 meta-analysis of cohort studies found a consistent U-shaped association: both low and high serum magnesium were linked to a higher risk of dementia and cognitive impairment, with the lowest risk near the middle of the reference range. Associations with dietary magnesium intake were inconsistent, and the three randomized trials of magnesium supplements that met its criteria were too few to support any conclusion.¹⁰ None of this distinguishes a cause from a marker of kidney function, diet or general health, and none of it is specific to the threonate salt.

§Limitations

The weakest part of the case

Every placebo-controlled trial of cognition or sleep was paid for by a company with rights to the ingredient: Neurocentria, Wu-xi Magtein Bioscience, AIDP or Threotech. No independent group has published a replication of the cognitive effect, and the brain-uptake mechanism that distinguishes MgT from cheaper salts has been shown only in rats. The largest trial's primary result sits just under the significance threshold, which its methods set as one-tailed, with no correction reported for the many outcomes tested.

The trials are short, between three and twelve weeks, and small enough that a single result in either direction would be weak evidence. Outcomes are scores on laboratory test batteries, which improve with practice in both arms; none has measured a real-world outcome or a diagnosis. Human doses are also far below what the rat work implies: one outside commentator in 2010 estimated a starting human-equivalent dose of about 7 g a day, against the 1 to 2 g used in trials.² The same tension runs through Nootropics generally, where self-reported benefit routinely outruns measured performance.

§Risks and regulation

In the trials, MgT was well tolerated, with mild gastrointestinal complaints and headache reported at rates similar to placebo. EFSA concluded that up to 3 g a day, about 250 mg of magnesium, is safe for adults other than pregnant and lactating women, and noted that this means more threonate exposure than vitamin C metabolism produces, with limited information on how the body handles it. Its opinion assessed safety and magnesium bioavailability, not any cognitive claim.³ In the United States, MgT is sold as one of the Dietary supplements, which require no proof of efficacy before sale.

§Outlook

The questions are straightforward and unanswered as of October 2026: whether oral MgT raises magnesium in the human brain, whether its cognitive effect survives a trial its makers do not fund, and whether it does anything for older adults that matters outside a test battery. Trial registries list further studies with universities as lead sponsors, including in collegiate athletes, a group for whom the anti-doping rules described in Enhancement in sport make supplement purity a practical concern. Until those report, MgT shares the position of NAD+ precursors: a plausible mechanism in animals, a consumer market, and human data too thin to separate a real effect from noise. Its newer trials borrow their sleep outcomes from the Wearable health sensors popular in Biohacking and grinders and Quantified self circles, which makes them cheap to run and hard to compare with clinical sleep research. No trial has examined Healthspan or any outcome the Geroscience hypothesis treats as central, and "cognitive age" joins the other unvalidated age estimates used to market interventions. For cognition in later life, exercise and adequate sleep have far more support. Whether MgT belongs in discussions of Human enhancement or Intelligence amplification at all depends on a replication that has not yet been published.

§See also

• Nootropics
• Dietary supplements
• NAD+ precursors
• Alzheimer's disease
• Biological age
• Sleep and longevity
• Biohacking and grinders

§References