Probably helpful for glucose control, constipation, and some broader metabolic health effects. Blood pressure, migraine, inflammation, and several popular use cases remain mixed rather than clearly established, and the sleep claim is still not supported by strong review-level evidence. Effects continue to vary substantially by chemical form.
What the research says
Each row is a different health effect. A means the support looks strongest, B likely helpful, C mixed or modest, and — not enough human evidence yet.
Papers are individual studies. Research reviews summarize evidence from multiple studies, such as systematic reviews and meta-analyses.
B · Likely helpful
B
Diabetes
Improves fasting glucose and insulin sensitivity markers.
32 papers · 10 research reviews
B
Migraine
5 of 6 papers positive. Used in clinical practice for migraine prevention; most trials show benefit.
6 papers
B
Metabolic syndrome
4 of 6 papers show improvements in metabolic parameters. Overlaps with diabetes/BP benefits.
6 papers
B
Constipation
Magnesium oxide or citrate as osmotic laxative.
12 papers
C · Mixed or limited evidence
C
Blood pressure
Modest but consistent blood-pressure-lowering effect.
36 papers · 10 research reviews
C
Pregnancy
Reduces pre-eclampsia and fetal growth restriction risk; effect sizes vary.
65 papers
C
Inflammation (CRP)
Small reductions in inflammation reported in some trials, but the effect is not consistent.
28 papers
C
Asthma
IV magnesium may help acute attacks; oral supplementation less clear.
7 papers
C
Cardiovascular disease
Observational associations with reduced CV events; intervention trials less consistent.
6 papers
C
Exercise performance
6 papers, mixed. Any effect appears small and inconsistent across training types.
6 papers
C
Muscle cramps
Of 8 papers, zero showed clearly positive effects. Despite being one of the most common reasons people take magnesium, the evidence does not support it — except possibly pregnancy-specific cramps.
8 papers
C
Depression
12 papers but no qualifying meta-analysis or systematic review. Can't grade yet.
12 papers
— · Not enough human evidence yet
—
Anxiety
Reviews disagree. Some trials report reduced anxiety scores; others find no effect. Evidence inconsistent.
5 papers
—
Sleep
Only 3 papers in humans. Existing trials are small and heterogeneous. There is not yet enough data to endorse the common "magnesium for sleep" claim.
3 papers
—
Cognition & memory
5 papers but no qualifying review-level synthesis. Some on L-threonate form.
5 papers
—
Arrhythmia
4 papers. Reduces post-surgical atrial fibrillation, but too few trials to grade broader arrhythmia claims.
4 papers
—
Restless legs syndrome
3 papers only.
3 papers
—
PMS
Only 2 papers. Common claim, weak evidence base.
2 papers
—
Bone health
Only 2 papers. Calcium and vitamin D dominate this literature.
2 papers
Quick facts
Dose, form and upper limits based on 14 trials in the database.
Typical dose
300–400 mg/day elemental magnesium
Best taken
Split across the day. Higher doses cause stomach upset when taken at once
Form
Citrate, glycinate, malate, and threonate absorb well. Oxide is cheap but poorly absorbed (~4%)
Common side effects
Loose stools and stomach discomfort, especially with oxide. Resolves with smaller, divided doses
Drug interactions
Reduces absorption of certain antibiotics (tetracyclines, quinolones) and bisphosphonates if taken together
Tolerable upper limit
350 mg/day from supplements (food sources excluded)
Quality-verified products
Products that pass third-party quality testing.
Product quality data coming soon
Product recommendations are being built from USP Verified, NSF Certified, and Labdoor test data. Products will only appear here after the supplement itself clears the evidence bar above — and only if the product passes third-party purity testing.
How we graded Magnesium
Grades come from 96 unique research papers indexed from PubMed — including 38 meta-analyses, 24 systematic reviews, and 151 randomized trials. Form differences (citrate vs. oxide vs. glycinate) introduce variability that pooled meta-analyses rarely separate. Effect sizes vary by form. Each supplement is graded by health effect using a certainty × effect matrix applied across PubMed evidence per health effect. Cochrane reviews and high-quality systematic reviews carry the most weight. From abstract-only data, "high certainty" is rare — most A grades come from moderate certainty plus consistent meaningful-positive evidence. See Methodology for full detail.