Substrate ·evidence-based
interventions / Magnesium

Magnesium: what the evidence actually shows

settled  magnesium — evidence and attention roughly aligned

Reviewed by Substrate editorial, Independent research collective ·

⚠ Safety in context

Evidence by outcome

heart healthgrade Bmoderate certainty— Meta-analysis of 38 RCTs (2709 participants): magnesium supplementation modestly lowered blood pressure — systolic by about 2.8 mm Hg and diastolic by about 2.0 mm Hg versus control. A small but consistent blood-pressure effect; not shown to reduce cardiovascular events directly.
inflammationgrade Bmoderate certainty— Meta-analysis of RCTs: magnesium supplementation significantly reduced several inflammatory markers, most consistently serum C-reactive protein (and nitric oxide), versus control. The effect on downstream clinical outcomes is not established — this is a biomarker signal.
sleepgrade Clow certainty— may modestly reduce sleep-onset latency in older adults with insomnia, but the evidence is limited and low quality; benefit for general adults or magnesium glycinate specifically is uncertain
stressgrade Clow certainty— may modestly reduce subjective anxiety/stress, esp. in deficiency
blood sugargrade Cmoderate certainty— modest improvement in insulin sensitivity / fasting glucose, mainly in deficiency or diabetes
migraine frequencygrade Clow certainty— 10 oral-magnesium RCTs (n=789): reduced migraine frequency and intensity vs. placebo (odds ratios 0.20 and 0.27, no CIs given); some trials had inadequate randomization per the review authors. A separate IV arm covered acute attacks, not prevention.
bone healthgrade Clow certainty— Systematic review in older adults: higher magnesium status/intake is associated with greater bone mineral density and, in some studies, lower fracture risk. Evidence is largely observational, so the association does not establish that supplementation prevents fractures.
cognitiongrade Clow certainty— Systematic review of magnesium and cognitive health in adults: the evidence linking magnesium status or intake to cognition is limited and inconsistent, drawn largely from observational data, and the authors call for better-designed studies with repeated magnesium biomarkers before any conclusion. No demonstrated cognitive benefit from supplementation.
moodstudied, no effectmoderate certainty— Not recommended for unipolar depression

People say · Evidence shows

“Magnesium glycinate is better for sleep because it crosses the blood-brain barrier”

aheadlargemechanistic speculation

What it probably doesn't do

Grade history

Re-review cadence: every 12 months (Grade A/B) · next scheduled by · methodology v1.

By outcome — the money pages

Compare

More

sources: PMID:41000008 · PMID:35277037 · PMID:35184264 · PMID:33865376 · PMID:28445426 · PMID:35045911 · PMID:26752497 · PMID:34666201 · PMID:39009081 · PMID:35311615

Cite this page

Reuse under CC-BY 4.0 with attribution to evidencebased.info.

Plain-text citation
Substrate. Magnesium: what the evidence actually shows. https://evidencebased.info/interventions/magnesium. Updated 2026-07-19.
BibTeX
@misc{substrate_terventionsmagnesium,
  author = {Substrate editorial},
  title  = {Magnesium: what the evidence actually shows},
  year   = {2026},
  url    = {https://evidencebased.info/interventions/magnesium},
  note   = {evidencebased.info}
}