Mineral supplement

Magnesium L-Threonate

Magnesium made up 7.5% to 8.8% of the 12 batches in EFSA's batch table and 7.4% to 8.0% of three further 2020 batches; the app lists a 10-hour half-life, and EFSA cites about 2.5 h for l-threonate given as the calcium salt.1

Emerging evidence (the app's grade, not a conclusion of the sources)Get the app

Level after one dose

Magnesium L-Threonate after one dose: 50% at ~10 h, 25% at 20 h, 12.5% at 30 h.

Drawn from the half-life in the app's model, ~10 h. The model's figures

Magnesium L-threonate is sold as the branded ingredient Magtein, and EFSA describes it as an organic salt of the l-threonate anion and the magnesium cation in a 2:1 ratio.1 The app places it among minerals and grades the evidence as emerging. EFSA's Panel on Nutrition, Novel Foods and Food Allergens concluded in 2024 that the ingredient "is safe under the proposed conditions of use" and "is a source from which magnesium is bioavailable".1 The human evidence here is a handful of trials of 15 to 109 people lasting 21 days to 12 weeks, in sleep, memory and cognition, cancer pain, attention-deficit/hyperactivity disorder (ADHD) and dementia. Of the five randomised trials described here, Hausenblas and colleagues, Zhang and colleagues, and Lopresti and Smith name an ingredient company as funder (AIDP, Wu-xi Magtein Bioscience and Threotech), the Miami trial summarised by EFSA came from an application submitted by AIDP Inc., and Wu and colleagues report that no specific funding was disclosed.1,2,3,4,5

The mechanism reports below come from rats and mice; Zhang and colleagues report that brain magnesium was not assessed in their trial, and their discussion cites an earlier double-blind trial in older participants as having demonstrated significant elevation of brain Mg levels, a trial that is not among those summarised here.3 Slutsky and colleagues at the Massachusetts Institute of Technology reported in Neuron in 2010 that "increasing brain magnesium using a newly developed magnesium compound (magnesium-L-threonate, MgT) leads to the enhancement of learning abilities, working memory, and short- and long-term memory in rats".6 The treated rats had "higher density of synaptophysin-/synaptobrevin-positive puncta in DG and CA1 subregions of hippocampus that were correlated with memory improvement", and the record lists a National Institute of Neurological Disorders and Stroke grant.6 EFSA summarised the cerebrospinal fluid (CSF) part of that work: at day 24 the rise in CSF magnesium was "statistically significant by about 7% and 16% as compared to baseline and the control group, respectively", while the control group's loss of magnesium was attributed by the authors to CSF sampling.1 Shen and colleagues gave mice magnesium L-threonate or magnesium sulfate by mouth for 4 weeks and reported, in Neuropsychiatric Disease and Treatment in 2019, that "Mg concentration in the CSF was increased in the mice treated with MgT but not MgSO4".7 In a mouse model of Parkinson's disease induced by MPTP, the same group concluded that "MgT, rather than MgSO4, can significantly attenuate MPTP-induced motor deficits and dopamine (DA) neuron loss", and its authors report no conflicts of interest. The MgT used in that work was supplied by Magceutics Inc., USA.7 Lopresti and Smith, a funded trial team, write that Magtein "has been shown through animal and in vitro studies to increase brain magnesium concentrations", and Zhang and colleagues state that "brain levels of magnesium were not assessed in this study".4,3

3

Half-life and pharmacokinetics

Values from the Milligram app's model.

Half-life
~10 h
Bioavailability
50%
Typical dose
1.5–2 g/day
Route
Oral
Time to peak
~4 h

The second column holds the app's figures, shown beside what the sources report.

Table 3.1Half-life and pharmacokinetics4 rows
ParameterApp's figureReported in the sources
Half-life~10 hAbout 2.5 h mean plasma half-life of l-threonate after single doses of calcium l-threonate in 12 healthy men aged 19 to 40; no accumulation was observed in another 12 volunteers given 2025 mg twice daily for 4 days; this is the threonate moiety given as the calcium salt, not magnesium1
Bioavailability50%EFSA concludes magnesium is bioavailable from the ingredient; in rats, relative absorption of magnesium was reported as about 60% for magnesium L-threonate against about 40%–45% for magnesium chloride, gluconate and citrate; no percentage for people is given1
Time to peak~4 hl-threonate plasma peak at 2 h in fasted men and about 3 h after breakfast, after calcium l-threonate1
Elemental magnesium~144 mg for 1.5–2 g (the app's note)7.5% to 8.8% magnesium in 12 batches reviewed by EFSA; about 124 mg magnesium in 1.5 g and 166 mg in 2 g in the Miami trial; about 75 mg/g in the Magtein used by Hausenblas and colleagues; 108–144 mg/day cited by Zhang and colleagues1,2,3

No pharmacokinetic study of magnesium after magnesium L-threonate was among the 20 PubMed records listed for the compound (55 matched) or in the EFSA opinion. EFSA notes that in the calcium l-threonate study 24-hour urinary excretion of l-threonate "represented only 10.3, 4.2 and 3.3%" of the three single doses, and that "the available information on the metabolic fate of threonic acid in the human body is very limited".1

4

Human studies

Table 4.1Human studies7 rows
Study and yearDesignPeopleDose and lengthMain result
Liu and colleagues, unpublished study report summarised by EFSA, 20241Randomised, double-blind, placebo-controlled, Miami, US; rice protein placebo; study report submitted by the applicant AIDP Inc.51 adults (36 women, 15 men) with self-reported memory and concentration complaints1.5 g or 2 g/day by body weight; 12 weeksUrinary magnesium +41% against +8% at week 12 (p = 0.048); plasma magnesium +4.9% at week 6 (p = 0.026) but +3.2% against +2.5% at week 12, not significant; red blood cell magnesium not significant; 47 adverse events in 28 of 51 people (17 events in 13 of 25 on the ingredient and 30 events in 15 of 26 on placebo), all classified as mild
Zhang and colleagues, Nutrients, 20223Randomised, double-blind, placebo-controlled; formula of magnesium L-threonate with phosphatidylserine and vitamins C and D; funded by Wu-xi Magtein Bioscience109 healthy Chinese adults aged 18-65; 102 finished2 g/day of the formula or starch placebo; 30 daysMemory quotient 81.84 ± 7.18 on the formula against 61.73 ± 10.27 on placebo at day 30 (formula baseline 60.31 ± 11.35); drop-out 6.42%; the EFSA opinion notes the article did not report adverse events
Hausenblas and colleagues, Sleep Medicine: X, 20242Decentralised, randomised, double-blind, placebo-controlled, parallel-arm; funded by AIDP, where three authors are listed80 adults aged 35–55 with self-assessed sleep problems; 40 randomised per group, 38 per group analysed1 g/day or rice protein placebo; 21 daysOura ring deep sleep score favoured magnesium L-threonate (p < 0.001), as did REM sleep score (p = 0.020); sleep latency was slightly increased against placebo; most questionnaire measures did not reach significance against placebo; 4 adverse events with magnesium L-threonate and 13 with placebo
Lopresti and Smith, Frontiers in Nutrition, 20254Randomised, double-blind, placebo-controlled; funded by Threotech100 adults aged 18 to 45 with self-reported dissatisfied sleep; 50 per group2 g/day (1 g twice daily) or placebo; 6 weeksNIH Total Cognition Composite improved more than on placebo (p = 0.043); no group difference on Raven's test (p = 0.953) or sleep disturbances (p = 0.316); self-reported sleep-related impairment improved more on Magtein (p = 0.043); no group differences in ring sleep outcomes; treatment-related adverse events in 10% against 8%; 4 people discontinued
Wu and colleagues, Cancer Medicine, 20235Randomised, double-blind, placebo-controlled; one tertiary hospital in Shanghai; no specific funding disclosed83 enrolled people with advanced cancer taking oral morphine1.5 g or 2.0 g by weight, or starch placebo; up to 12 weeksIncrease in daily morphine dose at day 90 of 21.20 mg/d on magnesium L-threonate against 40.44 mg/d on placebo (p < 0.01); no difference in pain scores (VAS) or breakthrough pain frequency; no serious side effects in either group; 10 people died of tumour progression (6 in the magnesium group, 4 on placebo) and stayed in the analysis
Surman and colleagues, Journal of Dietary Supplements, 2021, with registry NCT025587908,9Open-label single-group pilot at Massachusetts General Hospital; Neurocentria listed as collaborator15 adults with ADHD of moderate severityL-threonic acid magnesium salt formulations MMFS302 and MMFS202, up to 12 weeks47% met the response criteria; the registry lists 20 who signed consent, 15 who began the product and 13 who completed; it lists 10 of 15 with other adverse events, including nausea, vomiting or diarrhea in 5, and no serious events or deaths
Registry entry NCT02210286, Stanford University, 202110Open-label single-group; Stanford University lead sponsor, Magceutics listed as collaborator17 enrolled, with mild to moderate dementia1,800 mg/day of MgT-1219 planned for 60 daysRegistry lists 0 deaths and 1 serious adverse event among 17, a gastrointestinal problem noted as mild gastrointestinal discomfort; 15 completed and two participants discontinued before analysis

Deeper analysis: the 51-adult Miami trial summarised by EFSA

The trial is an unpublished study report supplied by the applicant, and EFSA lists it as a randomised, double-blind, placebo-controlled trial of 51 adults (36 women, 15 men). The applicant that supplied the report was AIDP Inc., the company that also funded the Hausenblas trial.1,2 Eighteen women and seven men received the ingredient and 18 women and eight men received rice protein, and the efficacy endpoints were anxiety, mood, sleep quality and cognition.1 Adverse events were similar in number of people (13 on the ingredient and 15 on placebo), with more events on placebo (30) than on the ingredient (17); gastrointestinal events affected 5 of 25 and 4 of 26 people.1 EFSA reports that all 47 events were classified as mild, and that the four judged possibly or probably related to the ingredient were loose stools, dry mouth, stomach upset and hypertension. Of the 13 events judged possibly or probably related to the intervention, four were in the ingredient group and nine in the placebo group.1 The Panel's own limitations were "the low number of subjects, the dose, which was only 2 g as compared to the proposed maximum dose of 3 g for the NF and the limited safety endpoints studied".1 EFSA reads the rise in urinary magnesium as the sign of absorption: "The study suggests that magnesium is absorbed, as indicated by the increased urinary magnesium concentration".1 The efficacy results of this trial are not given in the EFSA opinion.

Deeper analysis: Hausenblas and colleagues, sleep, 2024

The trial was run online with participants taking two capsules of 500 mg magnesium L-threonate or identical rice protein capsules 2 hours before bedtime, and 249 people were assessed for eligibility to reach 80 randomised.2 Primary outcomes were weekly sleep questionnaires (Insomnia Severity Index, Leeds Sleep Evaluation Questionnaire, Restorative Sleep Questionnaire) and nightly Oura ring data.2 The authors report ring results of p = 0.010 for activity score and p = 0.003 for readiness activity balance alongside deep sleep score, light sleep (p = 0.006) and REM sleep score, and that magnesium L-threonate slightly increased sleep latency against placebo (p < 0.05).2 On the questionnaires, "it was evident that almost all participants reported their sleep quality and their behavior improved over the three weeks, irrespective of treatment".2 The authors list as a limitation that "Subjective measures had a strong placebo effect" and suggest a longer study duration.2 Funding came from AIDP, three of the six authors list AIDP as their affiliation, and the competing-interest statement says the authors have no known competing financial interests. The Magtein used was produced and supplied by Threotech LLC.2

Deeper analysis: Lopresti and Smith, cognition and sleep, 2025

The trial screened 156 people by telephone and randomised 100 to Magtein or placebo for 6 weeks.4 The primary measure was the NIH Toolbox Total Cognition Composite, on which the Magtein group improved more (p = 0.043), with a 7.5-year reduction in estimated brain cognitive age and a greater improvement in reaction time (p = 0.031).4 The null results were the Raven's test (p = 0.953), sleep disturbances (p = 0.316), restorative sleep (p = 0.439), general wellbeing (p = 0.436) and every sleep outcome from the ring, although resting heart rate fell more (p = 0.030) and heart rate variability rose more (p = 0.036). Self-reported sleep-related impairment improved more on Magtein (p = 0.043), and in a subset of participants with more severe sleep-related problems a group difference in sleep disturbances was identified (p = 0.031).4 Treatment-related adverse events listed by term included increased bowel movements or loose stools in 2 people on Magtein and 1 on placebo, headaches in 2 and 1, and a skin rash in 1 and 0.4 The authors write that "studies of longer duration will be beneficial", that examining older people and those with cognitive impairment will be important, and that the findings carry over only cautiously to other forms of magnesium.4 Threotech funded the trial, supplied the product and "was involved in the conceptualisation of the study design"; the first author is managing director of a contract research organisation that has received funding from nutraceutical companies.4

Deeper analysis: Zhang and colleagues, memory, 2022

The product tested was Magtein PS, a formula of magnesium L-threonate with phosphatidylserine and vitamins C and D, so the result is for the formula and not for magnesium L-threonate alone.3 Of 109 enrolled (54 and 55), 102 finished, and the Clinical Memory Test was given at baseline and at 30 days.3 The authors report that the formula group scored higher on all five subtests and on the memory quotient, that the placebo group showed no significant improvement on any test, and that on portrait-features memory the placebo group scored significantly worse after the trial (p = 0.017).3 The authors list their limitations as "a small sample size, just 51 people in the treatment group", no measurement of brain or serum magnesium, and that "all study subjects were Chinese".3 Wu-xi Magtein Bioscience funded the trial and the formula was supplied by Magceutics; the authors state no known competing financial interests, and one author is listed at Nutrasource, Canada.3

Deeper analysis: Wu and colleagues, cancer pain, 2023

The primary outcome was "the increase in the daily oral dose of morphine in each of the two groups", measured over 90 days in people already taking oral morphine.5 Daily morphine doses rose in both groups: by 9.85 against 20.49 mg/d at day 30, 15.96 against 29.06 mg/d at day 60 and 21.20 against 40.44 mg/d at day 90, favouring the magnesium group.5 Pain scores and breakthrough pain frequency did not differ between groups.5 The most common side effect after constipation was nausea, and the authors report no significant difference between groups in side effects. Constipation, a secondary outcome measured with the Wexner score, was significantly relieved in the magnesium group compared with placebo on day 7, and the authors found no significant differences in other adverse events except for constipation.5 Ten patients died of tumour progression (6 in the magnesium group and 4 on placebo) and their data stayed in the analysis; anxiety scores did not differ between the groups at any time, although the average GAD-7 score in the magnesium group on day 90 was higher than at baseline (12.22 against 11.45).5 The authors name as limitations that "we did not measure patients' blood magnesium levels", that there were "no exact criteria for dose escalation", and that "our sample size was small, and due to the withdrawal of several patients, the target sample capacity was not reached".5 The preparation came from Wuxi Brain Magnesia Biomedical Technology, no specific funding was disclosed, and the authors declared no potential conflicts of interest.5

5

Doses reported in studies and on labels

Table 5.1Doses reported in studies and on labels9 rows
SettingDose reportedPopulation and lengthReference
EFSA proposed conditions of useUp to 3000 mg/day of the ingredient, equivalent to about 250 mg magnesiumAdults, excluding pregnant and lactating women; food supplement1
Miami trial1.5 g (50 to 70 kg) or 2 g (70 to 100 kg) per day, providing about 124 mg or 166 mg magnesium51 adults, 12 weeks, against rice protein placebo1
Hausenblas and colleagues1 g/day as two 500 mg capsules, about 75 mg magnesium80 adults, 21 days, against rice protein placebo2
Zhang and colleagues2 g/day of a formula with 400 mg magnesium L-threonate per capsule, two capsules morning and two evening109 adults, 30 days, against starch placebo3
Lopresti and Smith2 g/day as 1 g twice daily100 adults, 6 weeks, against placebo4
Wu and colleagues1.5 g or 2.0 g by weight, once daily83 people with advanced cancer taking morphine, up to 12 weeks, against starch placebo5
Surman and colleagues, as summarised by EFSA1.5–2 g/day of an anhydrous form15 adults with ADHD, 12 weeks, open-label1
Stanford dementia registry entry1,800 mg/day of MgT-121917 enrolled, 60 days planned, open-label10
Rat study cited by EFSAElemental magnesium targeted at 50 mg/kg/day in drinking water, about 604 mg magnesium L-threonate/kgRats, 24 days (animal data)1

No dosing label for a single-ingredient magnesium L-threonate medicine was identified: DailyMed returned 4 labels for the term, each a multi-ingredient prescription vitamin product listing magnesium L-threonate among many ingredients.

In the Miami trial EFSA describes 47 adverse events in 28 of 51 people, spread across the ingredient and placebo groups, with loose stools, dry mouth, stomach upset and hypertension among those judged possibly or probably related to the ingredient.1 In the 100-person trial of Lopresti and Smith, 98% of those on Magtein reported good or excellent tolerability on the Global Assessment of Tolerability to Therapy, one person reported moderate tolerability, and one person on placebo stopped because of gastrointestinal symptoms.4 The authors of that trial write that "the most common adverse reaction associated with magnesium supplementation is gastrointestinal disturbances such as diarrhoea or loose stools, particularly at higher doses", which they report was not observed in their study, although their table of possibly or probably related adverse events lists increased bowel movements or loose stools in 2 people on Magtein and 1 on placebo.4 Hausenblas and colleagues recorded 4 adverse events on magnesium L-threonate (nocturnal awakening once, leg cramps twice, sunken eyes once) and 13 on placebo, and write that "MgT appeared to be safe and well-tolerated".2 In the open-label ADHD pilot the registry lists nausea, vomiting or diarrhea in 5 of 15 people, 10 of 15 with other events, and no serious events or deaths.9 In rats, the dose that caused diarrhoea in half the animals was 131.5 mg magnesium/kg for magnesium L-threonate, against 89.9 mg/kg for magnesium chloride and 99.7 mg/kg for magnesium gluconate, and no rat given 50 mg magnesium/kg as magnesium L-threonate had diarrhoea.1 EFSA reports no genotoxicity concern and, for the open-label dementia and ADHD studies, that "no serious adverse events were observed in these trials", while judging that those two studies "do not provide evidence regarding the safety of the NF".1

7

Bloodwork and markers

In the Miami trial EFSA lists plasma, red blood cell and urinary magnesium, blood counts, metabolic panels, body weight, blood pressure and heart rate as measured, with "no clinically significant changes in safety laboratory results" reported in the study report and no statistically significant changes in body weight, blood pressure or heart rate at any of the four visits.1 Zhang and colleagues report blood counts, liver enzymes, urea, creatinine, glucose and lipids before and after 30 days, with the experimental and control groups not significantly different in their table at either time point, but state that serum magnesium was not measured.3,1 The app lists no markers to monitor for this compound.

8

Regulatory and anti-doping status

Sources in this part

  1. 1EFSA Panel on Nutrition, 2024

EFSA's Panel adopted its opinion on magnesium L-threonate as a novel food on 30 January 2024, after an application submitted by AIDP Inc., and concluded it is safe under the proposed conditions of use.1 The proposed use was as a food supplement for the general adult population, excluding pregnant and lactating women, at up to 3000 mg/day.1 EFSA notes that the roughly 250 mg magnesium at that intake "corresponds to the UL for supplemental magnesium on top of the magnesium intake from the background diet".1 The Panel also states that it "could not have reached the conclusion on the safety of the NF" without data the applicant claimed as proprietary, including the randomised, placebo-controlled human study.1 No single-ingredient DailyMed drug label was identified (4 multi-ingredient labels matched).

No pharmacokinetic study of magnesium after magnesium L-threonate was identified in the records searched, so the app's half-life, bioavailability and time-to-peak figures have no matching source figure for magnesium itself. EFSA describes the available information on the metabolic fate of threonic acid in people as very limited.1 The trials described here lasted 21 days to 12 weeks and were mostly industry-funded, and the authors of the Zhang and Wu trials list small sample sizes among their limitations.3,5 Lopresti and Smith write that "there have been no studies directly comparing the sleep-related effects of different magnesium forms".4 Brain magnesium was not assessed in the trial by Zhang and colleagues and blood magnesium was not measured in the trial by Wu and colleagues; Zhang and colleagues cite an earlier double-blind trial as having demonstrated significant elevation of brain Mg levels, and the brain evidence cited by Lopresti and Smith is from animal and in vitro studies.3,5,4 The app's note that this is the only magnesium form proven in animal studies to cross the blood-brain barrier is not repeated here; Lopresti and Smith state only that raised brain magnesium "has not been demonstrated convincingly with other forms of magnesium".4

What is the half-life of magnesium L-threonate?

The app lists about 10 hours, and none of the sources read reports a half-life for magnesium after magnesium L-threonate.1 EFSA cites a mean plasma half-life of about 2.5 h for l-threonate after calcium l-threonate in 12 men.1

How much magnesium is in magnesium L-threonate?

EFSA's batch table gives a magnesium content of 7.5% to 8.8% across 12 batches, and its certificates for three further 2020 batches give 7.4% to 8.0%.1 The app's note gives about 144 mg of elemental magnesium for 1.5 to 2 g, EFSA's Miami trial summary gives about 124 mg for 1.5 g and 166 mg for 2 g, and Hausenblas and colleagues give about 75 mg/g.1,2

Is magnesium L-threonate bioavailable?

EFSA concluded that the ingredient is a source from which magnesium is bioavailable, and the app lists 50%.1 The rat absorption figure EFSA reports is about 60%, and the human evidence it cites is one 51-person trial in which urinary magnesium rose more than on placebo at week 12 (+41% against +8%), while plasma magnesium at week 12 and red blood cell magnesium were not significantly different from placebo.1

Has EFSA assessed magnesium L-threonate?

EFSA's NDA Panel published an opinion in 2024 concluding that magnesium L-threonate is safe under the proposed conditions of use.1 Those conditions were up to 3000 mg/day in food supplements for adults other than pregnant or lactating women.1

Does magnesium L-threonate raise magnesium in the brain?

In rats and mice, cerebrospinal fluid magnesium rose with magnesium L-threonate, by about 7% against baseline and 16% against controls at day 24 in the rat work EFSA summarises.1,7 Zhang and colleagues state that brain magnesium was not assessed in their human trial.3

Did magnesium L-threonate improve sleep in trials?

In the 80-adult trial by Hausenblas and colleagues, the ring deep sleep score favoured magnesium L-threonate, while most questionnaire measures did not reach significance against placebo.2 The same trial reported that magnesium L-threonate slightly increased sleep latency compared with placebo (p < 0.05), its authors list a strong placebo effect on the subjective measures as a limitation, and AIDP funded it.2 In the 100-adult trial by Lopresti and Smith, there were no group differences in sleep outcomes from the ring and none in sleep disturbances (p = 0.316), while self-reported sleep-related impairment improved more on Magtein (p = 0.043).4

Did magnesium L-threonate improve memory in trials?

The NIH Total Cognition Composite improved more with Magtein than placebo in the 100-adult trial (p = 0.043), and Raven's test did not differ (p = 0.953).4 The Chinese trial of 109 adults tested a formula that also contained phosphatidylserine and vitamins C and D, so its memory result is for the formula.3

What side effects were reported in trials of magnesium L-threonate?

In the 100-adult trial, treatment-related adverse events were reported by 10% on Magtein and 8% on placebo.4 In the Miami trial summarised by EFSA, 47 adverse events were reported in 28 of 51 people (17 events in 13 of 25 on the ingredient and 30 events in 15 of 26 on placebo), all classified as mild, with loose stools, dry mouth, stomach upset and hypertension among those judged possibly or probably related to the ingredient.1 In the 80-adult trial by Hausenblas and colleagues, 4 adverse events were recorded with magnesium L-threonate and 13 with placebo.2

12

References

  1. EFSA Panel on Nutrition, Novel Foods and Food Allergens (NDA). Safety of magnesium l-threonate as a novel food pursuant to regulation (EU) 2015/2283 and bioavailability of magnesium from this source in the context of Directive 2002/46/EC. EFSA Journal 2024. EFSA opinion
  2. Hausenblas HA, et al. Magnesium-L-threonate improves sleep quality and daytime functioning in adults with self-reported sleep problems: A randomized controlled trial. Sleep Med X 2024. PMID 39252819
  3. Zhang C, et al. A Magtein, Magnesium L-Threonate, -Based Formula Improves Brain Cognitive Functions in Healthy Chinese Adults. Nutrients 2022. PMID 36558392
  4. Lopresti AL, Smith SJ. The effects of magnesium L-threonate (Magtein) on cognitive performance and sleep quality in adults: a randomised, double-blind, placebo-controlled trial. Front Nutr 2025. PMID 41601871
  5. Wu S, et al. Oral application of magnesium-L-threonate enhances analgesia and reduces the dosage of opioids needed in advanced cancer patients: a randomized, double-blind, placebo-controlled trial. Cancer Med 2023. PMID 36703238
  6. Slutsky I, et al. Enhancement of learning and memory by elevating brain magnesium. Neuron 2010. PMID 20152124
  7. Shen Y, et al. Treatment of magnesium-L-threonate elevates the magnesium level in the cerebrospinal fluid and attenuates motor deficits and dopamine neuron loss in a mouse model of Parkinson's disease. Neuropsychiatr Dis Treat 2019. PMID 31806980
  8. Surman C, et al. L-Threonic Acid Magnesium Salt Supplementation in ADHD: An Open-Label Pilot Study. J Diet Suppl 2021. PMID 32162987
  9. ClinicalTrials.gov. NCT02558790. L-Threonic Acid Magnesium Salt (L-TAMS) in ADHD. Massachusetts General Hospital. NCT02558790
  10. ClinicalTrials.gov. NCT02210286. Magnesium L-Threonate for the Enhancement of Learning and Memory in People With Dementia. Stanford University. NCT02210286