Place · Level 3 · Debunking
Magnesium L-Threonate (Magtein)
唯一能进大脑的镁营销学传奇 · 机制基石是 Slutsky 2010 单篇大鼠 / 小鼠论文 · 唯一人体支持是 Liu 2016 N=44 公司资助 RCT · 价格是普通柠檬酸镁的 5-7 倍 · 与 Boyle 2017 系统综述 + NIH ODS 给出的任何形式镁补足缺口都有效一比, 营销叙事 ≫ 证据 · 大多数人该买便宜的镁
Story path
- 1What is Mg-L-threonate · the Magtein storyWhat is Mg-L-threonate · the Magtein story
- 2Marketing claim · three layersMarketing claim · three layers
- 3Slutsky 2010 · the animal paper as foundationSlutsky 2010 · the animal paper as foundation
- 4Liu 2016 · N=44 · company-funded · sole RCTLiu 2016 · N=44 · company-funded · sole RCT
- 5Decision tree · most people should buy other MgDecision tree · most people should buy other Mg
Chapter 1
What is Mg-L-threonate · the Magtein story
What is Mg-L-threonate · the Magtein story
Magnesium L-Threonate (Mg-L-T) is a magnesium salt — elemental magnesium chelated with L-threonic acid. L-threonic acid is itself a metabolite of ascorbic acid (vitamin C), produced via oxidation/degradation in vivo. So at the molecular level, magnesium L-threonate = 'magnesium + a vitamin C metabolite' — which by itself carries no special pharmacological meaning. It is just a salt.
There is no real chemical or regulatory controversy here: it is GRAS, has a clean safety profile, no hepatorenal toxicity and no notable drug interactions. What we are debunking is not 'whether this is magnesium' — it is. What we are debunking is 'whether it is worth 5× the price of other magnesium forms', which is a marketing claim. The next four scenes peel that claim back layer by layer.
There is no real chemical or regulatory controversy here: it is GRAS, has a clean safety profile, no hepatorenal toxicity and no notable drug interactions. What we are debunking is not 'whether this is magnesium' — it is. What we are debunking is 'whether it is worth 5× the price of other magnesium forms', which is a marketing claim. The next four scenes peel that claim back layer by layer.
Brand history · MIT lab to Magtein™
The reason this compound became a supplement-market star is almost entirely the marketing narrative:In 2010, MIT neuroscientist Guosong Liu (also affiliated with Tsinghua), working with Susumu Tonegawa's lab, published the Slutsky 2010 paper in Neuron suggesting that this particular magnesium salt could raise rat/mouse cerebrospinal-fluid (CSF) magnesium and improve learning/memory task performance.That same year Liu and his team founded Magceutics, with U.S. distribution later picked up by AIDP (a California ingredient company), and the patented molecule was registered as the Magtein™ brand.From 2012 onward, AIDP licensed Magtein™ to dozens of supplement brands in the U.S., Europe and China (Life Extension, Pure Encapsulations, Jarrow, Doctor's Best, Now Foods and many others all carry a Magtein version).During 2020–2024, riding the Huberman / Joe Rogan / TikTok 'sleep / anxiety / memory' wave, 50+ Magtein products appeared in the U.S. market, priced at USD 30–60 per month — roughly 5–7× the cost of an equivalent elemental dose from magnesium citrate or glycinate.
Chapter 2
Marketing claim · three layers
Marketing claim · three layers
Peel the magnesium L-threonate marketing into three layers, each requiring its own validation.
Layer 1 (consumer-facing slogan): 'the only magnesium that gets into the brain' / 'the magnesium for ADHD, anxiety, Alzheimer, insomnia' / '+25% memory in 12 weeks' / 'synapse density enhancer'. This kind of language appears on the Magtein brand site, on dozens of supplement product pages, and is the version chanted in KOL videos.
Layer 2 (evidence the brand cites): nearly every slogan traces back to two papers — the animal study Slutsky 2010 (Neuron) and the single human RCT Liu 2016 (Journal of Alzheimer's Disease, also known as the MMFS-01 trial). Liu Guosong is the corresponding author on both, and is also the Magtein patent holder and Magceutics founder. The brand additionally lists a long bibliography of 'related' in vitro experiments and aged-rat-model papers, but those do not constitute direct support for human claims.
Layer 1 (consumer-facing slogan): 'the only magnesium that gets into the brain' / 'the magnesium for ADHD, anxiety, Alzheimer, insomnia' / '+25% memory in 12 weeks' / 'synapse density enhancer'. This kind of language appears on the Magtein brand site, on dozens of supplement product pages, and is the version chanted in KOL videos.
Layer 2 (evidence the brand cites): nearly every slogan traces back to two papers — the animal study Slutsky 2010 (Neuron) and the single human RCT Liu 2016 (Journal of Alzheimer's Disease, also known as the MMFS-01 trial). Liu Guosong is the corresponding author on both, and is also the Magtein patent holder and Magceutics founder. The brand additionally lists a long bibliography of 'related' in vitro experiments and aged-rat-model papers, but those do not constitute direct support for human claims.
Layer 3 · actual physiology
Layer 3 (the actual scientific background): this is where it matters. Adult plasma magnesium sits around 0.75–0.95 mmol/L, cerebrospinal fluid magnesium around 1.0–1.2 mmol/L. All forms of oral magnesium — oxide, citrate, glycinate, chloride — when they raise plasma magnesium after correcting a deficit, also raise CSF magnesium via the TRPM7 / MagT1 transporters at the blood-brain barrier (the de Baaij 2015 Physiological Reviews paper lays out the full physiology). There is no form of magnesium that 'rises in plasma but stays out of the brain', and no form that 'bypasses the blood-brain barrier and goes straight to the brain'. The marketing concept 'the magnesium that gets into the brain' is not a meaningful physiological distinction — the implied contrast ('other magnesium can't reach the brain') is simply false.This is the core problem with Magtein marketing: between Layer 1 (what lay readers hear) and Layer 3 (the actual physiology) sits a very thin shell built on two papers. The next two scenes pressure-test exactly those two papers.
Chapter 3
Slutsky 2010 · the animal paper as foundation
Slutsky 2010 · the animal paper as foundation
Slutsky I, Abumaria N, Wu LJ, Huang C, Zhang L, Li B, Zhao X, Govindarajan A, Zhao MG, Zhuo M, Tonegawa S, Liu G. (2010). *Enhancement of learning and memory by elevating brain magnesium.* Neuron 65(2): 165–177. This is the 'mechanistic foundation' of the entire Magtein marketing story — and it is a textbook early-stage animal study, nowhere near able to hold up the human claims the brand stacks on top of it.
What the paper did:
Administered various magnesium salts (oxide, citrate, L-threonate) orally or by injection to rats and mice, then measured plasma and cerebrospinal-fluid (CSF) magnesium changes.Reported that animals on Mg-L-threonate showed larger CSF magnesium increases than animals on equivalent elemental doses of magnesium oxide or citrate.Hippocampal slices showed increased NR2B-subunit NMDA-receptor expression and synaptic density.Behavioral assays (Morris water maze, object recognition) showed faster learning in the Mg-L-T group.
What the paper did:
Administered various magnesium salts (oxide, citrate, L-threonate) orally or by injection to rats and mice, then measured plasma and cerebrospinal-fluid (CSF) magnesium changes.Reported that animals on Mg-L-threonate showed larger CSF magnesium increases than animals on equivalent elemental doses of magnesium oxide or citrate.Hippocampal slices showed increased NR2B-subunit NMDA-receptor expression and synaptic density.Behavioral assays (Morris water maze, object recognition) showed faster learning in the Mg-L-T group.
Methodological gaps · six standard red flags
Standard methodological limitations any biomedical-methods course would flag:Species: rodents ≠ humans, with differences in magnesium metabolism, intestinal absorption, and blood-brain-barrier transporter expression.Sample size: the typical small N of an animal study — exploratory, not designed to underwrite human conclusions.Controls: dose-equivalent comparators for high-absorption forms like magnesium citrate or glycinate were not fully matched; marketing language has spun this into 'other magnesium absorbs poorly', which overreaches what the paper shows.Replication: 14 years on, there is no independent replication from labs unaffiliated with Liu's team and not funded by Magceutics.Endpoints: hippocampal NR2B upregulation + better water-maze performance are easy-to-hit endpoints in rodents; they are not human 'memory improvement' and certainly not human 'Alzheimer prevention'.Conflicts of interest: the corresponding author Liu Guosong is also the Magtein patent holder and later the founder of Magceutics/Magtein. The disclosure is in the paper, but readers should treat it as the 'founding paper' of a commercial product line.
What it supports · what it doesn't
What it can support: in rodents, magnesium L-threonate is an interesting tool compound that, under specific conditions, can raise CSF magnesium higher than magnesium oxide does (in animals).What it cannot support: 'the only magnesium that gets into the human brain' / 'improves memory in humans' / 'prevents Alzheimer'. Those extrapolations cross species, clinical endpoints and disease categories, and a single 14-year-old animal paper cannot carry that weight. The next scene examines the one human RCT Liu's group itself ran.
Chapter 4
Liu 2016 · N=44 · company-funded · sole RCT
Liu 2016 · N=44 · company-funded · sole RCT
Liu G, Weinger JG, Lu ZL, Xue F, Sadeghpour S. (2016). *Efficacy and safety of MMFS-01, a synapse density enhancer, for treating cognitive impairment in older adults: a randomized, double-blind, placebo-controlled trial.* Journal of Alzheimer's Disease 49(4): 971–990. This is the human evidence the brand cites most often, and so far the only RCT testing Magtein on cognitive endpoints.
Trial design:
Subjects: adults aged 50–70 with subjective memory complaints but not meeting dementia criteria, recruited in Southern California.Sample: 44 randomized (about 51 enrolled with some dropout), allocated to MMFS-01 (i.e. magnesium L-threonate) at 1.5–2 g/day or placebo.Duration: 12 weeks.Endpoints: a composite cognitive score combining four domains (attention / working memory / executive / episodic memory), plus several secondary tests.
Reported main results: the treatment group showed larger gains on the composite score than placebo; on a 'cognitive age' estimate the group appeared to walk back about 9 years. These numbers are reproduced on Magtein product pages over and over.
Trial design:
Subjects: adults aged 50–70 with subjective memory complaints but not meeting dementia criteria, recruited in Southern California.Sample: 44 randomized (about 51 enrolled with some dropout), allocated to MMFS-01 (i.e. magnesium L-threonate) at 1.5–2 g/day or placebo.Duration: 12 weeks.Endpoints: a composite cognitive score combining four domains (attention / working memory / executive / episodic memory), plus several secondary tests.
Reported main results: the treatment group showed larger gains on the composite score than placebo; on a 'cognitive age' estimate the group appeared to walk back about 9 years. These numbers are reproduced on Magtein product pages over and over.
Six methodological flaws
Serious limitations of this RCT:Sample: N = 44 in a cognitive intervention study is exploratory pilot territory, not a confirmatory trial. By comparison, the FINGER trial of older-adult cognition had N = 1260; PREDIMED on the Mediterranean diet had N = 7447. The size gap on that one axis alone is 30–150×.Duration: 12 weeks is extremely short relative to the timescale of cognitive decline itself (5–20 years).Funding: the trial was funded by Magceutics. Corresponding author Liu Guosong is the founder of Magceutics, the patent holder for Magtein, and the inventor of MMFS-01, the very test article. In a drug RCT, that level of overlap would normally trigger strict independent oversight or third-party analysis; the supplement market imposes no such constraint.Endpoint design: the primary endpoint is a weighted composite of four subtests. Composite primary endpoints leave a lot of room for the analyst — if any sub-component nudges past significance, the composite can drag the overall result across the line. Modern clinical trials increasingly require preregistration with pre-specified sub-endpoint analyses; this paper is not transparent on that front.The four measured sub-domains: attention improved significantly, but executive / working memory / episodic memory did not all independently cross significance. Slogans like '+25% memory' and 'cognitive age −9 years' depend almost entirely on post hoc selection of the favorable cut.Replication: 8 years on, no independent (non-Magceutics-funded, non-Liu-led) RCT has reproduced these numbers. A single small-sample trial is universally regarded in biomedicine as 'worth a larger follow-up study', not as 'fit for product marketing'.
Cross-comparison · narrative ≫ evidence
Cross-comparison makes the scale gap clearer:Boyle 2017 (Nutrients) systematic review on magnesium supplementation for subjective anxiety and stress pools multiple RCTs (hundreds of participants) at B-tier evidence — any form of magnesium has a mild positive effect on anxiety and subjective stress when intake is deficient, and no included study showed L-threonate outperforming citrate or glycinate.de Baaij 2015 (Physiological Reviews), the authoritative magnesium-physiology review, attributes every known cognitive and mood signal of magnesium to 'correcting a deficit', not to a specific salt form.NIH ODS 2022 magnesium professional fact sheet likewise makes no recommendation favoring Mg-L-T over any other form.
Bottom line: Liu 2016 is not the apex of the Magtein evidence story — it is the entire evidence story. N = 44 + company funding + founder as corresponding author + composite endpoint + no independent replication: any single item is survivable, but stacked together this paper cannot bear the scientific weight of an entire supplement subcategory.
Chapter 5
Decision tree · most people should buy other Mg
Decision tree · most people should buy other Mg
Reduce the whole thing to 'how should I spend $30 from my wallet'.
Background first: U.S. NHANES data estimate ~50% of adults fall below the magnesium RDA. Chinese CHNS data put it close to half as well. So most people topping up magnesium has real meaning — but that is a different question from 'which magnesium'.
RDA (NIH ODS): men 19–30 400 mg / men 31+ 420 mg / women 19–30 310 mg / women 31+ 320 mg / pregnancy +40 mg.
Priority 1 (food): leafy greens (1 cup cooked spinach = 78 mg) / pumpkin seeds (1 oz = 156 mg) / almonds (1 oz = 80 mg) / black beans (1 cup cooked = 120 mg) / whole grains (1 cup brown rice = 85 mg) / dark chocolate 70%+ (1 oz = 64 mg). Someone who eats greens, nuts and whole grains can meet the daily RDA from food alone — in which case this box is closed and no supplement is required.
Background first: U.S. NHANES data estimate ~50% of adults fall below the magnesium RDA. Chinese CHNS data put it close to half as well. So most people topping up magnesium has real meaning — but that is a different question from 'which magnesium'.
RDA (NIH ODS): men 19–30 400 mg / men 31+ 420 mg / women 19–30 310 mg / women 31+ 320 mg / pregnancy +40 mg.
Priority 1 (food): leafy greens (1 cup cooked spinach = 78 mg) / pumpkin seeds (1 oz = 156 mg) / almonds (1 oz = 80 mg) / black beans (1 cup cooked = 120 mg) / whole grains (1 cup brown rice = 85 mg) / dark chocolate 70%+ (1 oz = 64 mg). Someone who eats greens, nuts and whole grains can meet the daily RDA from food alone — in which case this box is closed and no supplement is required.
Supplement choice · cheap Mg vs Magtein premium
Priority 2 (ordinary magnesium supplement, when food doesn't fill the gap): not complicated.Magnesium citrate: high absorption, GI-friendly, cheap (about USD 5–10 per month for 200 mg elemental Mg/day). The pragmatic first pick for most people.Magnesium glycinate / bisglycinate: slightly better observational signal on sleep and anxiety, very little GI upset, suits evening dosing, USD 8–15 per month.Magnesium chloride / lactate: also fine, well absorbed.Avoid magnesium oxide as a standalone (poor absorption — fine as one component of a blended formula).
Priority 3 (magnesium L-threonate / Magtein): USD 30–60 per month. 5–7× more expensive than Priority 2. It is genuinely magnesium, it is safe, and there is no ethical concern about it — we are not calling it harmful. What we are saying is that no A- or B-tier evidence shows it outperforms magnesium citrate at the actual job of correcting a magnesium deficit. Slutsky 2010 is an animal paper, Liu 2016 is an N=44 company-funded single RCT — both are C-tier evidence.
Are you willing to pay an extra USD 25–50 per month for 'maybe a small extra cognitive boost that has not been independently replicated in 8 years'? This is a real wallet question with no universal answer — but understand that what you are buying is a 'C-tier-evidence premium', not 'a smarter magnesium'.
Actually protect cognition · A-tier spending
When you genuinely want to invest in cognitive health, spending the same budget and time in any of these directions beats Magtein:150 minutes/week of moderate aerobic exercise + 2 strength sessions (A-tier, FINGER and multiple meta-analyses).7–9 hours of sleep per night (A-/B-tier, multiple prospective cohorts).Mediterranean / DASH / MIND eating pattern (A-tier, PREDIMED, FINGER).Weekly deep contact with friends / family (A-tier, UK Biobank and Harvard Adult Development Study).Hearing screening and using a hearing aid when indicated (one of the largest modifiable factors in the Lancet 2024 dementia commission report).Controlling blood pressure / glucose, and not smoking (A-tier).
This is not a scam alert: Magtein is real magnesium, safe, and in a person who has a magnesium deficit it does close that deficit — just as magnesium citrate does. The problem is the distance between the marketing narrative and the evidence: 'the only magnesium that gets into the brain' is wrong, and '5× smarter than other magnesium' has no evidence.
One-sentence summary:
> Magnesium L-threonate is magnesium, not magic. Most people have a magnesium gap — close it first with food, then cheap forms (citrate or glycinate); paying 5× for Magtein buys a premium on the marketing claim, not a better brain.
>
> If you really want to protect cognition, take those USD 25–50 of monthly difference and spend them on gym, better food, or sleep-environment upgrades. Those are A-tier evidence; Magtein is C-tier.