Magnesium

Mechanism: Magnesium relaxes vascular smooth muscle, supports endothelial function, and modulates parathyroid-driven calcium handling — all upstream of blood pressure regulation.

Across 34 randomized trials, supplemental magnesium reduced systolic BP by ~2.0 mmHg and diastolic by ~1.78 mmHg over 3 months. Effects were larger in hypertensive, diabetic, or low-baseline-magnesium subgroups.

Most pronounced effect in adults already low in magnesium; minimal effect in well-replete normotensives.

• Magnesium and blood pressure (meta-analysis of 34 RCTs)

  positive

  Zhang et al., 2017, Hypertension · meta analysis · n=2028

  Pooled across 34 randomized trials, magnesium supplementation lowered systolic BP by ~2.0 mmHg and diastolic by ~1.78 mmHg. Effects were larger in low-baseline-magnesium, hypertensive, or diabetic subgroups.

  DOI

Mechanism: Magnesium binds GABA-A receptors and antagonizes NMDA — both directionally calming. Deficiency is associated with hyperarousal.

Small trials in older adults with insomnia show modest improvements in subjective sleep quality and reduced sleep-onset latency. Effects are real but modest.

Strongest in older adults and those with subclinical deficiency. Glycinate is the preferred form for sleep due to gentle GI profile.

• Magnesium for sleep — small trials, big claims

  mixed

  Multiple small RCTs in older adults with insomnia (e.g. Abbasi 2012; Rondanelli 2011) · rct

  Small RCTs in older adults show modest improvements in self-reported sleep quality and sleep latency. Effects are real but modest — typically minutes of latency reduction, not transformative.

  Widely shared social-media claims of '160% REM' or '250% deep sleep' do not appear in any peer-reviewed magnesium trial. Treat those numbers as marketing copy, not evidence.

Mechanism: Magnesium is a cofactor for tyrosine kinase activity downstream of the insulin receptor. Hypomagnesemia is overrepresented in type 2 diabetes.

Multiple independent meta-analyses show modest reductions in fasting glucose, HbA1c, and HOMA-IR in adults with T2D or insulin resistance. Effect is real but smaller than first-line glucose-lowering medication.

Meaningful effect in diabetic and pre-diabetic populations. Minimal in metabolically healthy adults.

• Magnesium for glycemic control (pooled meta-analyses)

  positive

  Simental-Mendía et al., 2016, Pharmacological Research; replicated in subsequent meta-analyses · meta analysis

  Independent meta-analyses converge: supplemental magnesium produces modest improvements in fasting glucose, HbA1c, and insulin resistance in adults with T2D or pre-diabetes. The effect is real but smaller than first-line glucose-lowering medication.

  Effects strongest in adults with documented hypomagnesemia; minimal benefit in metabolically healthy adults.

Mechanism: Magnesium dampens cortical spreading depression and stabilizes NMDA-driven neuronal excitability — both implicated in migraine pathophysiology.

300–600 mg/day for 8–12 weeks reduces migraine frequency and intensity in pooled trials. American Academy of Neurology rates the evidence Level B (probably effective).

Effective for episodic and chronic migraine prophylaxis. Not for acute attack abortion.

• Magnesium for migraine prevention (pooled trials)

  positive

  Chiu et al., 2016, Pain Physician (meta-analysis) · meta analysis

  Pooled migraine trials show consistent reductions in attack frequency and intensity with 300–600 mg/day elemental magnesium for 8–12 weeks. Effect size is modest but clinically useful — comparable to several first-line prophylactics.

  The American Academy of Neurology / American Headache Society 2012 guideline rates magnesium Level B (probably effective) for migraine prevention.

Mechanism: Magnesium gates calcium influx at the neuromuscular junction and is required for ATP-bound enzymatic activity. Deficiency genuinely disrupts the contraction-relaxation cycle — the question is whether non-deficient adults benefit from extra.

The 2020 Cochrane review of 11 RCTs (n=735) found no benefit of magnesium over placebo for idiopathic muscle cramps in older adults. Frank deficiency or heavy sweat-loss are different stories — but for the average person reaching for magnesium 'because cramps,' the controlled-trial evidence is null.

Helpful in documented hypomagnesemia or high-sweat-loss endurance athletes. Not supported by RCT evidence in non-deficient adults with idiopathic cramps.

• Cochrane review — magnesium for muscle cramps

  null

  Garrison et al., 2020, Cochrane Database of Systematic Reviews · systematic review · n=735

  Cochrane review pooled 11 RCTs (n=735) of magnesium for idiopathic muscle cramps in older adults. No benefit over placebo for frequency, intensity, or duration. The popular intuition that 'magnesium fixes cramps' does not survive controlled trials in this population.

  Pregnancy-related cramps and exercise-induced cramps in heavy sweat-loss athletes are different populations and were not well represented.

  DOI

Mechanism: A large fraction of body magnesium is stored in bone, where it influences hydroxyapatite crystal size and parathyroid signaling.

Observational cohorts associate higher magnesium intake with better BMD. The honest read: there is no RCT showing supplemental magnesium reduces fractures, and the intervention studies are small with mixed results. Bone benefits are mechanistically plausible but not RCT-proven.

Magnesium is supportive at best. Calcium, vitamin D, K2, and weight-bearing activity drive measurable bone outcomes.

• Magnesium and bone mineral density (pooled evidence)

  positive

  Rondanelli et al., 2021, Nutrients (review of cohort + intervention data) · systematic review

  Higher magnesium intake associates with better bone mineral density across observational cohorts. Most of the body's magnesium reservoir lives in bone, which makes the mechanistic story plausible, but no large RCT has shown supplemental magnesium reduces fracture risk.

  Magnesium supplementation alone is not osteoporosis treatment. Calcium, vitamin D, vitamin K2, and weight-bearing activity remain primary.

Mechanism: Magnesium modulates HPA-axis tone and dampens NMDA-driven excitatory signaling. Low intracellular magnesium correlates with anxiety and depression scores in cohort data.

The only RCT of magnesium for depression (Tarleton 2017) was open-label crossover — patients knew they were getting magnesium, which inflates measured effect sizes. The 6-point PHQ-9 drop is real but largely uninterpretable as efficacy. There is no rigorous double-blind replication.

Reasonable to try in documented deficiency or alongside standard care. Not a standalone depression treatment.

• Magnesium for mild-to-moderate depression

  positive

  Tarleton et al., 2017, PLoS ONE · rct · n=126

  126 outpatients with PHQ-9 mild-to-moderate depression saw a 6.0-point PHQ-9 reduction over 6 weeks (p<0.001). Effect appeared within 2 weeks and persisted.

  Open-label crossover design — patients knew they were receiving magnesium, which inflates measured effect sizes.

Mechanism: Magnesium is required for ATP to bind enzymes — every ATP-dependent reaction is technically Mg-ATP-dependent. Deficiency impairs cellular energy economy directly.

Mechanism is unimpeachable. The clinical-trial evidence is not: there is no clean RCT showing supplemental magnesium improves energy or reduces fatigue in non-deficient adults. Repletion in genuinely deficient individuals reliably helps; supra-physiological dosing in the replete does not.

Helps people who are deficient. Does nothing extra for those who aren't.

Mechanism: Magnesium L-threonate (Magtein®) crosses the blood-brain barrier more readily than other forms, raising brain magnesium concentration — the proposed mechanism for cognitive benefit.

One small (n=44) industry-funded RCT showed modest cognitive improvements at 1.5–2 g/day over 12 weeks. The mechanism is plausible; the trial is not yet replicated by independent labs.

Worth a try in older adults with subjective cognitive decline, with the understanding that the evidence base is one small manufacturer-funded trial.

• Magnesium L-threonate for cognition

  positive

  Liu et al., 2016, Journal of Alzheimer's Disease · rct · n=44 · industry-funded

  Small RCT in 44 older adults using 1.5–2 g/day of magnesium L-threonate (Magtein®). Modest cognitive improvements vs placebo on a composite score over 12 weeks.

  Manufacturer-funded; small sample; dose was supplemental L-threonate, not generic magnesium. Replication in independent labs is limited.

Surfacing failed trials alongside the positive evidence. Leaving them out would be marketing, not science.

• IMAGES — IV magnesium in acute stroke

  negative

  Muir et al., 2004, Lancet · rct · n=2589

  Large randomized trial of intravenous magnesium sulfate in 2,589 acute ischemic stroke patients. No improvement in 90-day death or disability vs placebo. Magnesium is not a stroke neuroprotectant.

  One of several large negative trials of magnesium as a neuroprotectant. The hypothesis was reasonable; the data was not.

• FAST-MAG — pre-hospital magnesium for stroke

  negative

  Saver et al., 2015, NEJM · rct · n=1700

  Paramedics initiated IV magnesium within 2 hours of stroke onset — the fastest neuroprotectant trial ever attempted. 1,700 patients. No functional-outcome benefit at 90 days vs placebo.

  Negative even with optimal time-to-treatment. Closes the door on magnesium for acute neuroprotection.

• Magnesium glycinate (bisglycinate)Well absorbed
  Best for: Sleep, calm, stress, magnesium repletion with low GI side effects
  Glycine is itself mildly calming. The glycinate form is the default recommendation when GI tolerance matters.
• Magnesium taurateWell absorbed
  Best for: Cardiovascular and blood-pressure support
  Taurine itself supports cardiovascular function — the pairing is intentional.
• Magnesium L-threonateMagtein®Modest elemental dose per capsule, but threonate appears to cross the blood-brain barrier
  Best for: Cognition, brain magnesium status
  The cognition trial (Liu 2016) used 1.5–2 g/day of the Magtein® compound — higher than typical Mg supplements. Manufacturer-funded; replication remains limited.
  brain: 1500–2000 mg
• Magnesium malateGood
  Best for: Energy, fatigue, fibromyalgia symptom support
  Malate is a Krebs-cycle intermediate — a plausible energy pairing, modest direct evidence.
• Magnesium citrateGood
  Best for: General repletion, mild laxative effect for constipation
  The mild osmotic-laxative effect can be either a feature or a side effect depending on use case.
• Magnesium chlorideGood
  Best for: General repletion; the depression RCT used this form
  Salty taste in solution; commonly sold as flakes for transdermal use (transdermal absorption is poorly evidenced).
• Magnesium aspartateModerate
  Best for: Older formulation; used in some athletic blends
  Aspartate is a non-essential amino acid; no clear advantage over glycinate or citrate for most uses.
• Magnesium oxidePoor — roughly 4% absorption
  Best for: Cheap. Acts as an osmotic laxative more than a magnesium source.
  Common in budget gummies and multivitamins. If repletion is the goal, this is not the form to use.
• Magnesium stearateNot a usable magnesium source
  Best for: Tablet excipient (flow agent), not a supplement form
  If a label lists magnesium stearate as the magnesium source, the product is functionally un-dosed.

• !Avoid magnesium supplementation in chronic kidney disease without nephrology guidance. Reduced renal clearance can drive dangerous hypermagnesemia at otherwise routine doses.
• !Social-media claims of '160% REM sleep' or '250% deep sleep' from magnesium have no peer-reviewed source. Real magnesium sleep effects in trials are modest — minutes of latency reduction, not transformative.
• !Intravenous magnesium has been tested as a stroke neuroprotectant in two large, well-designed trials (IMAGES 2004, FAST-MAG 2015) and failed both times. Magnesium is not a stroke treatment.

• What's the best form of magnesium to take?▾

  For most people: glycinate (bisglycinate). It absorbs well, is gentle on the GI tract, and the glycine itself is mildly calming. Use citrate if you want the mild laxative effect, taurate for cardiovascular support, or L-threonate (Magtein®) for cognition specifically — at the higher 1.5–2 g/day dose used in the cognition trial. Avoid magnesium oxide for repletion; absorption is around 4%.
• How much magnesium should I take?▾

  The RDA is 310–420 mg/day depending on age and sex; the FDA upper limit on supplemental magnesium is 350 mg/day (this does not include magnesium from food). Most clinical trials use 200–400 mg/day of elemental magnesium for general health endpoints. For migraine prevention, 300–600 mg/day; for the cognition L-threonate use case, 1.5–2 g/day of Magtein®.
• Will magnesium help me sleep?▾

  Modestly, especially if you're older or magnesium-insufficient. Trials show small improvements in self-reported sleep quality and sleep-onset latency. The viral '160% REM' and '250% deep sleep' claims do not appear in any peer-reviewed trial — those numbers are marketing copy.
• Can I take too much magnesium?▾

  Diarrhea is the practical limit for most people — supplemental doses above 400–600 mg/day commonly cause GI side effects, particularly with poorly-absorbed forms like oxide or citrate. True magnesium toxicity is rare and almost exclusive to renal impairment. If you have kidney disease, do not supplement without medical guidance.
• Does magnesium interact with my medications?▾

  Yes, in a few specific ways. Magnesium binds bisphosphonates, tetracyclines, fluoroquinolones, and levothyroxine in the gut — separate dosing by 2–4 hours. Long-term PPI use depletes magnesium; supplementation is often warranted. If you take digoxin, discuss with your prescriber before starting magnesium.