_whatshisname_ :
Magnesium L-threonate produced a small, borderline-significant advantage on one composite cognitive test after six weeks. It did not literally make participants’ brains 7.5 years younger, and the evidence is not strong enough to establish a dependable cognitive-enhancement effect.
What the study actually found
The trial randomly assigned 100 adults aged 18–45 with self-reported unsatisfactory sleep to:
* 2 grams daily of branded magnesium L-threonate, providing 145 mg elemental magnesium
* or rice-flour placebo for six weeks.
On the prespecified primary cognitive score:
* Magnesium group: 112.06 → 120.46, an 8.40-point improvement
* Placebo group: 110.82 → 116.42, a 5.60-point improvement
* Extra improvement attributable to treatment: approximately 2.8 score points
* Statistical result: p = 0.043.
Both groups improved considerably, probably including practice/retesting effects. The meaningful comparison is therefore 8.4 versus 5.6 points, not simply announcing that the magnesium group improved 7.5%.
The “7.5 years younger” claim
This was not measured through brain imaging, biological aging, dementia markers or neurological examination.
Researchers took a 2.24-point difference on an age-unadjusted cognitive score and divided it by an assumed average decline of approximately 0.3 points per year:
2.24 \div 0.3 = 7.47\text{ years}
That created the “7.5-year cognitive age” figure. The paper explicitly labels cognitive age as an exploratory analysis based on population norms.
A more accurate statement would be:
“The groups differed by an amount that the researchers mathematically equated to about 7.5 years of average age-related test-score difference.”
That is very different from reversing seven years of brain aging.
Several results were negative
The post leaves these out:
* Overall fluid cognition: not significantly better than placebo, p = 0.277
* Crystallized cognition: not significant, p = 0.111
* Picture-sequence episodic memory: not significant, p = 0.092
* Attention/inhibitory control: not significant, p = 0.782
* Processing-speed test: magnesium actually improved less numerically, although the difference was not significant
2026-07-27 22:22:55