Magnesium L-threonate improved cognition by 7.5 cognitive years in 6-week trial

A randomised, double-blind, placebo-controlled trial in Frontiers in Nutrition just gave magnesium L-threonate, the form sold as Magtein, its tightest human data set yet. Two grams a day for six weeks. One hundred adults, all unhappy with their sleep. The result: overall cognition improved, and estimated brain cognitive age dropped by 7.5 years compared to placebo. Adrian Lopresti and Stephen Smith ran the study out of Clinical Research Australia in Perth. The paper landed in January 2026.

Here is why the form matters. Magnesium glycinate and citrate, the stuff on most pharmacy shelves, raise blood magnesium. They do not reliably raise brain magnesium. The L-threonate ligand crosses the blood-brain barrier by hitching a ride on glucose transporters. Sun and colleagues worked out the mechanism in 2016. If you care about cognition rather than muscle cramps, the form you pick is most of the battle.

How the study was designed

One hundred participants, 18 to 45 years old, all self-reporting crap sleep for at least a month. Fifty got 2 grams of Magtein a day, split across morning and evening doses. Fifty got rice flour in identical capsules. Compliance was 92 percent.

The testing battery went further than most supplement trials bother to go. Cognition was measured on the NIH Toolbox Cognitive Battery, a validated computer-based instrument, plus Raven’s Progressive Matrices for abstract reasoning. Sleep got three self-report measures (two PROMIS scales, one restorative sleep questionnaire) and an Oura Ring that tracked sleep stages, resting heart rate, and heart rate variability. The researchers also ran a 3D Aim Trainer test for hand-eye coordination. Nobody had tried that in a magnesium study before.

A detail that matters for interpreting what comes next: this cohort was cognitively above average at baseline. NIH Toolbox scores sat roughly one standard deviation above the population norm. These were healthy, sharp people who happened to sleep badly. You are not looking at a deficit-correction story.

What the results showed

The primary endpoint, NIH Toolbox Total Cognition Composite, moved (p = 0.043). Magtein added 8.40 points. Placebo added 5.60 points. Both arms improved, which happens when people know they are in a study. The gap was 2.80 points, and it was real.

Then the cognitive age number. Using the normative data inside the NIH Toolbox, Lopresti and Smith estimated that the Magtein group’s improvement worked out to a 7.5-year reduction in brain cognitive age relative to placebo (p = 0.041). A separate 12-week Magtein trial in older adults had found roughly 9 years. That is now two trials, different age groups, both showing large cognitive-age shifts. Harder to wave off as noise.

Memory carried the signal. List Sorting Working Memory hit p = 0.033 between groups. Picture Sequence Memory, episodic recall, trended at p = 0.092. The authors describe this as “consistent with the established biological role of magnesium in supporting synaptic density and neural connectivity, particularly within brain regions subserving memory and executive control.”

Not everything moved. Raven’s Progressive Matrices, abstract reasoning largely independent of memory, showed nothing (p = 0.953). That null result has information in it. Magnesium L-threonate does not seem to boost fluid intelligence. It looks domain-specific: hippocampal and prefrontal.

The 3D Aim Trainer test, the hand-eye coordination measure nobody had tried before, showed a signal (p = 0.031). The authors think this could point toward faster driving reactions or sharper visuo-motor performance. Exploratory endpoint, single trial. Salt recommended.

Sleep effects were mixed, but heart rate shifted

The sleep story is less clean. The PROMIS Sleep-Related Impairment scale, which measures how much bad sleep disrupts your daytime functioning, favoured Magtein (p = 0.043). But sleep disturbances, restorative sleep, and general wellbeing showed no group differences in the full sample. The supplement did not make everyone sleep better.

Unless you look at the people who actually had bad sleep. In the subgroup whose baseline PROMIS disturbance scores sat in the top quartile, Magtein dropped both sleep disturbance (p = 0.031) and sleep-related impairment (p = 0.012) more than placebo. The effect appeared where there was an actual problem to solve.

The Oura Ring data contain the study’s most interesting finding, and it is not about sleep architecture. Total sleep time, efficiency, latency, REM, deep sleep. None of it differed between groups. Zero across the board. But average sleeping heart rate fell 1.32 bpm in the Magtein group and did not move in placebo (p = 0.030 between groups). RMSSD, a heart rate variability measure that tracks parasympathetic nervous system activity, rose in the Magtein arm relative to placebo (p = 0.036).

Lopresti and Smith call this “the first study to demonstrate an effect of dietary magnesium supplementation on HRV and resting heart rate during sleep.” Their read: magnesium L-threonate may be pushing the autonomic nervous system toward parasympathetic dominance, the “rest and digest” state. A post-hoc analysis showed lower sleeping heart rate tracked with better subjective sleep scores, which gives the physiological shift some real-world plausibility even if the standard sleep-stage measures did not cooperate.

The caveats

Six weeks. That is all the data we have. Whether the cognitive effects last, fade, or plateau is unknown. The cohort was young and cognitively above average. We do not know how these numbers translate to older adults or people with actual cognitive impairment. The earlier trial in 50-to-70-year-olds helps, but it is one data point.

No blood magnesium levels were drawn. Without serum or erythrocyte measures, we cannot say whether Magtein corrected a subclinical deficiency or pushed brain magnesium into a supraphysiological range. The authors want magnesium status tracked at baseline in future trials. Fair request.

Follow the money. Threotech Inc. owns the Magtein patent. They funded the trial. They helped design it. Lopresti runs Clinical Research Australia, a contract research organisation that takes money from nutraceutical companies. Smith works there. Industry funding in supplement research is normal and does not automatically mean the results are wrong, but this is not independent replication. It is an industry-funded study by authors with financial ties to the product category. That does not make the data invalid. It does mean you want to see the protocol run by an unaffiliated group before treating the findings as settled.

Side effects were minor. Ninety-eight percent rated tolerability good or excellent. GI symptoms and headaches showed up in roughly 4 percent of both groups. Nobody in the active arm dropped out because of side effects.

Who this matters for

This trial moves the human evidence for magnesium L-threonate into a new demographic, adults under 45 with subjective sleep complaints, and gives us the first data showing the supplement shifts HRV and resting heart rate during sleep. The cognitive age finding has now appeared in two independent trials with different age brackets. That is the difference between “interesting single study” and “probably real.”

If you are thinking about taking it: the cognitive effects look domain-specific. Memory and reaction time, yes. Abstract reasoning, no. Whether you notice a sleep difference may depend on how bad your sleep actually is. Two grams a day for six weeks was safe in this cohort. Longer-term safety data at this dose does not yet exist.

Magnesium L-threonate is not a sleep hygiene stand-in. It does not replace exercise or a diet that already contains magnesium from actual food. The authors call for longer trials, magnesium-deficient populations, and experimental stress models. Until those arrive, what sits on the table is a well-run, industry-funded RCT that adds a solid piece of evidence to a supplement category that had been running on mechanism and anecdotes.

Consult your doctor before starting any supplement.

References

1. 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. Frontiers in Nutrition. 2026;12. https://doi.org/10.3389/fnut.2025.1729164
2. Slutsky I, Abumaria N, Wu LJ, et al. Enhancement of learning and memory by elevating brain magnesium. Neuron. 2010;65(2):165-177. https://doi.org/10.1016/j.neuron.2009.12.026