Magnesium

oxide · citrate · glycinate · malate · taurate · threonate

An essential mineral involved in hundreds of enzyme reactions. The interesting questions are about the form — elemental content ranges from ~60% (oxide) down to ~7% (threonate) by weight — and about who is actually deficient.

last reviewed 2026-08-30

by draft — unverified

GRADE

CLAIM

SPECIES

EFFECT

STUDIES

blood pressure reduction — grade B, small effect (human-rct)

Zhang et al. (2016)

species humandesign meta-analysis of 34 RCTsn 2028duration median 3 monthsdose median 368 mg/dayfunding publicoutcome systolic & diastolic blood pressure

About −2.0 mmHg systolic and −1.8 mmHg diastolic. Real, replicated — and small. A textbook case of grade B with a small effect.

sleep quality in older adults with low magnesium — grade C, small effect (human-rct)

Abbasi et al. (2012)

species humandesign RCT, double-blindn 46duration 8 weeksdose 500 mg oxide/dayfunding publicoutcome insomnia severity index, sleep efficiency, serum markers

Better subjective sleep in elderly insomniacs — but n=46 and a population likely to be deficient at baseline.

Mah & Pitre (2021)

species humandesign systematic review of 3 RCTsn 151duration up to 8 weeksdose mixedfunding publicoutcome sleep onset latency, total sleep time

Small improvement in sleep-onset latency; authors rate the evidence quality as low. This is why the claim sits at C, not B.

migraine prophylaxis — grade B, small effect (human-rct)

Peikert et al. (1996)

species humandesign RCT, double-blindn 81duration 12 weeksdose 600 mg citrate/dayfunding not-declaredoutcome migraine attack frequency

Attack frequency fell ~42% vs ~16% on placebo. Enough for several neurology guidelines to list magnesium as probably effective.

glycemic control in type 2 diabetes — grade C, small effect (human-rct)

Simental-Mendía et al. (2016)

species humandesign meta-analysis of 18 RCTsn 670duration 4–24 weeksdose mixedfunding not-declaredoutcome fasting glucose, insulin sensitivity indices

Improvements in fasting glucose in deficient or diabetic participants — surrogate markers only, and heterogeneous trials.

Grade timeline

2026-01

sleep quality in older adults: B → C

Re-review against the 2021 systematic review's low quality rating; the original B leaned too heavily on one small trial.

Dosing & safety

TYPICAL SUPPLEMENT DOSE200–400 mg elemental/day

DOSES USED IN STUDIES300–600 mg elemental/day in most trials

UPPER LIMIT350 mg/day elemental from supplements (food doesn't count)

SAFETY NOTE Doses above the supplemental UL commonly cause diarrhea. People with impaired kidney function should not supplement without medical supervision.

COST SNAPSHOT

€0.36/g cheapest per gram of active · Magnesium oxide

OPEN CALCULATOR →