Magnesium for Sleep and Anxiety: What the Trials Show
Magnesium is the most recommended sleep supplement on the internet and one of the least tested. The trials that exist are small, mostly in older adults, and point the same modest way. The anxiety trials everyone quotes share a problem: no placebo group.
Evidence strength
Level 1a
Systematic review of RCTs
Peer-reviewed refs
15
Reading time
10 min
Key Takeaways
- Pooled across three randomized trials in 151 older adults, magnesium shortened sleep-onset latency by about 17 minutes. Total sleep time rose by 16 minutes but not significantly. The reviewers rated the evidence low to very low quality.
- The best single trial gave 500 mg of magnesium as plain oxide to 46 elderly people with insomnia for 8 weeks and improved insomnia scores, sleep efficiency and sleep-onset latency. It is small, subjective and from one centre.
- The two human trials of magnesium L-threonate for sleep were funded by companies that sell it. The larger one, 100 adults for 6 weeks, found a self-reported improvement and no difference on the sleep-tracking ring.
- The anxiety evidence is weaker than it looks. The 264-person stress trial compared magnesium with magnesium plus B6, so both groups improved and nobody knows what placebo would have done. The 126-person depression trial was open-label with no placebo.
- In a 100-person trial, 320 mg of magnesium raised serum magnesium only in people who started below 1.8 mg/dL. Sleep improved in both the magnesium and placebo groups. Status at baseline probably decides who responds.
- Our grade: C for sleep, D for anxiety in otherwise healthy adults. A reasonable eight-week trial at 200–400 mg of elemental magnesium in the evening, with a stop date, if your intake is likely low.
Search any sleep forum and magnesium is the first answer. It is cheap, it is a mineral rather than a drug, and about half of adults in the United States eat less of it than the estimated requirement, which makes a deficiency story easy to believe. The question this review asks is narrower: when researchers gave magnesium to people and compared them with placebo, did sleep or anxiety improve, by how much, and in whom?
The answer is not obvious because the two halves of the literature disagree. Observational studies link higher magnesium intake to better sleep fairly consistently. The randomized trials are few, small, mostly in older adults, and their reviewers describe the quality as low. Below is what each trial measured, where they conflict, and where that leaves someone deciding tonight. For the forms and doses, see Magnesium Glycinate vs Threonate vs Citrate; for whether a blood test can tell you anything, see Serum vs RBC Magnesium.
Sleep: The Human Trials in Order of Quality
The only meta-analysis restricted to randomized trials pooled three studies in 151 older adults with insomnia. Sleep-onset latency was 17.36 minutes shorter after magnesium than after placebo (95% CI −27.27 to −7.44). Total sleep time was 16.06 minutes longer, but the difference was not statistically significant. All three trials were at moderate-to-high risk of bias, and the authors graded the evidence low to very low. [1] Their conclusion was careful: the literature is substandard for a firm recommendation, but magnesium is cheap and the direction is consistent.
| Trial | Who | Treatment | What changed |
|---|---|---|---|
| 2012, Tehran | 46 elderly people with primary insomnia | 500 mg/day as magnesium oxide, 8 weeks | Insomnia Severity Index down (P=0.006), sleep-onset latency down (P=0.02), sleep efficiency up (P=0.03); total sleep time unchanged (P=0.37) |
| 2002, Munich | 12 adults aged 60–80, crossover | Rising dose to 30 mmol/day (about 730 mg), 20 days | Slow-wave sleep 16.5 vs 10.1 minutes; nocturnal cortisol lower |
| 2010, Grand Forks | 100 adults over 51 with poor sleep | 320 mg/day as citrate, 7 weeks | Sleep quality improved in both groups; magnesium raised serum levels only in those starting below 1.8 mg/dL |
| 2011, Pavia | 43 care-home residents | Melatonin 5 mg + magnesium 225 mg + zinc, 8 weeks | Sleep quality score 6.8 points better than placebo; magnesium's share unknowable |
The Tehran trial is the one most articles rest on. Participants were randomized to two tablets of magnesium oxide a day, 250 mg elemental each, or placebo. After eight weeks the magnesium group had lower insomnia scores, shorter sleep-onset latency, higher sleep efficiency and lower serum cortisol, with higher melatonin and renin. [2] Three people dropped out, so 43 finished. The paper reports these as P values; the effect in minutes comes from the pooled analysis above. We break the trial down in The 46-Person Trial Behind the Magnesium-for-Sleep Claim.
The Munich study is the only one with sleep EEG. Twelve older adults took a creeping dose of effervescent magnesium for 20 days in a placebo-controlled crossover. Slow-wave sleep rose from 10.1 to 16.5 minutes, delta and sigma power increased, and night-time cortisol fell. [3] It is a physiological signal in a dozen people, not a treatment trial.
The Grand Forks study is the awkward one. A hundred adults over 51 with poor sleep took 320 mg of magnesium citrate or a sodium citrate placebo for seven weeks. Sleep quality scores improved from 10.4 to 6.6 regardless of group. Serum magnesium rose with supplementation only in the 37 participants who started below 1.8 mg/dL, and C-reactive protein fell only in those who began above 3 mg/L. [4] The authors said plainly that a placebo effect prevented them from telling whether magnesium status affects sleep.
Two newer trials tested magnesium L-threonate, the form marketed for the brain. In 80 adults aged 35–55 with self-reported sleep problems, 1 g a day for 21 days improved deep-sleep and REM scores on a consumer ring and several daytime measures versus placebo. The authors were affiliated with the ingredient's distributor. [6] A larger six-week trial in 100 adults aged 18–45 gave 2 g a day. Self-reported sleep-related impairment improved more than placebo (P=0.043), but sleep disturbance and restorative sleep did not, and the ring recorded no group difference in any sleep outcome. The funder supplied the product and helped design the study. [7] An earlier 12-week trial of the same compound in 44 older adults with cognitive impairment reported that placebo effects on sleep and anxiety were so large the magnesium effect could not be determined. [8]
Behind the trials sit the observational data. In 3,964 adults followed for 20 years, the top quartile of magnesium intake was less likely to sleep under seven hours than the bottom quartile (OR 0.64, 95% CI 0.51–0.81); the association with sleep quality was borderline (OR 1.23, 95% CI 0.999–1.50). [9] A systematic review of nine studies and 7,582 people put it in one sentence: observational studies agree, randomized trials contradict each other. [10]
Anxiety and Stress: Positive Trials With a Missing Placebo Arm
The 2017 systematic review that most anxiety claims cite found 18 studies, all in people already vulnerable to anxiety. Four of eight studies in anxious samples, four of seven in premenstrual syndrome, and one of two in hypertension reported a benefit; the postpartum studies found none. No study used a validated measure of stress, and the authors called the quality of evidence poor. [12]
The two trials that get quoted as the modern upgrade have the same hole.
- The 264-person stress trial enrolled healthy adults with high stress scores and serum magnesium between 0.45 and 0.85 mmol/L. After eight weeks, stress scores fell 42.4% on magnesium alone and 44.9% on magnesium plus vitamin B6. That looks like a large effect until you notice the design: the comparison was magnesium versus magnesium-plus-B6. There was no placebo group. The trial was run by the manufacturer of both products. [13]
- The 126-person depression trial gave 248 mg of elemental magnesium as chloride for six weeks in a crossover against no treatment. Depression scores improved by 6.0 points and anxiety scores by 4.5 points. It was open-label: participants knew when they were taking magnesium, and the control was nothing. [14]
A 2024 review of 15 interventional trials counted five of seven anxiety studies as positive, then listed the reasons for caution: small samples, heterogeneous doses and forms, extra active ingredients, and both negative trials in populations with hormonal drivers. [11]
Where the Evidence Breaks Down
Four problems recur across the sleep and anxiety trials.
- The populations are not the people buying the supplement. The trials with positive results enrolled elderly insomniacs, care-home residents, or stressed adults with low serum magnesium. Nobody has run a well-powered placebo-controlled trial in healthy 30-year-olds with adequate intake, which is most of the market.
- Placebo effects on sleep are large and the outcomes are subjective. The Grand Forks trial improved sleep in both arms. The threonate trial in older adults could not separate drug from placebo. The 100-person threonate trial found a self-report difference and no wearable difference.
- The forms do not line up with the marketing. The trial with the clearest insomnia benefit used magnesium oxide, the form most sites tell you to avoid. The form most heavily marketed for sleep has two supplier-funded trials, one of them negative on objective measures.
- Combination products muddy attribution. The Pavia trial's 6.8-point improvement came with 5 mg of melatonin in the same spoonful of pear pulp. [5]
For balance: magnesium does produce measurable effects in placebo-controlled trials elsewhere. Across 34 double-blind trials and 2,028 participants, a median dose of 368 mg a day for three months lowered systolic blood pressure by 2.00 mm Hg and diastolic by 1.78 mm Hg, with serum magnesium rising by 0.05 mmol/L. [15] The mineral is not inert. The question is whether it does anything for sleep and mood in people who are not short of it.
Why Baseline Status Probably Explains the Split
The Grand Forks trial contains the most useful clue in the literature. Magnesium raised serum levels only in people who started low, and 58% of participants were eating less than the estimated average requirement. [4] The 264-person stress trial enrolled only people with low-normal serum magnesium, and both arms improved. [13] The elderly insomnia trials come from populations in which low intake is common.
Put together, the pattern looks like repletion rather than pharmacology: give magnesium to someone whose intake is poor and something changes; give it to someone replete and you get the placebo arm. That is also what the mechanism predicts. Magnesium blocks the NMDA receptor and supports GABA signalling, but those are ambient roles of a mineral present in every cell, not the actions of a sedative.
Our Read: Grade C for Sleep, D for Anxiety
Sleep: Grade C. Three small randomized trials, pooled, show a 17-minute reduction in time to fall asleep in older adults with insomnia, with low-quality evidence and no significant change in total sleep. The direction is consistent and the cost of trying is low, which is what C means on our scale. It is not the B that many articles imply.
Anxiety and stress: Grade D in otherwise healthy adults. The positive trials either lack a placebo arm or come from populations with low magnesium. The systematic reviews describe the evidence as poor. There may be an effect in people with low intake or low serum magnesium; the trials have not isolated it.
What would move the sleep grade to B: one placebo-controlled trial of 200 or more adults under 60, with an objective sleep measure, a prespecified primary endpoint and no funding from the manufacturer. None is registered that we could find.
What This Means If You Are Deciding Now
If your diet is low in magnesium, which is likely if you eat few legumes, nuts, whole grains and leafy greens, an eight-week trial of 200–400 mg of elemental magnesium in the evening is reasonable. Use a well-absorbed salt, keep a sleep diary from week one, and set a stop date. If nothing has changed by week eight, the trial is over. The Deep Sleep Stack runs magnesium as its second layer, after glycine, for exactly that reason: you add one thing at a time so you know which one is working.
If your intake is already adequate, the trials give you no reason to expect a sleep benefit, and the anxiety trials give you less. Spend the money on the things that do move sleep in replete adults: a fixed wake time, morning light, caffeine before noon only, and alcohol off the evening.
Two questions come up often enough to answer here. Does the form matter for sleep? Not on the evidence: the clearest insomnia benefit came from oxide, and the threonate trials are weaker than their marketing. Should you test first? Serum magnesium is a blunt instrument, but a value below 0.75 mmol/L changes the question from supplementation to diagnosis, so it is worth one test if you are going to supplement long term. Details are in Serum vs RBC Magnesium: Which Test Tells You Anything.
Scientific References
- [1]Mah J, Pitre T. Oral magnesium supplementation for insomnia in older adults: a Systematic Review & Meta-Analysis — BMC Complementary Medicine and Therapies (2021)Oxford 1aPMID 33865376
- [2]Abbasi B, Kimiagar M, Sadeghniiat K, et al.. The effect of magnesium supplementation on primary insomnia in elderly: A double-blind placebo-controlled clinical trial — Journal of Research in Medical Sciences (2012)Oxford 1bPMID 23853635
- [3]Held K, Antonijevic IA, Künzel H, et al.. Oral Mg(2+) supplementation reverses age-related neuroendocrine and sleep EEG changes in humans — Pharmacopsychiatry (2002)Oxford 2bPMID 12163983
- [4]Nielsen FH, Johnson LK, Zeng H. Magnesium supplementation improves indicators of low magnesium status and inflammatory stress in adults older than 51 years with poor quality sleep — Magnesium Research (2010)Oxford 1bPMID 21199787
- [5]Rondanelli M, Opizzi A, Monteferrario F, et al.. The effect of melatonin, magnesium, and zinc on primary insomnia in long-term care facility residents in Italy: a double-blind, placebo-controlled clinical trial — Journal of the American Geriatrics Society (2011)Oxford 1bPMID 21226679
- [6]Hausenblas HA, Lynch T, Hooper S, et al.. Magnesium-L-threonate improves sleep quality and daytime functioning in adults with self-reported sleep problems: A randomized controlled trial — Sleep Medicine: X (2024)Oxford 1bPMID 39252819
- [7]Lopresti AL, Smith SJ. The effects of magnesium L-threonate (Magtein) on cognitive performance and sleep quality in adults: a randomised, double-blind, placebo-controlled trial — Frontiers in Nutrition (2026)Oxford 1bPMID 41601871
- [8]Liu G, Weinger JG, Lu ZL, et al.. Efficacy and Safety of MMFS-01, a Synapse Density Enhancer, for Treating Cognitive Impairment in Older Adults: A Randomized, Double-Blind, Placebo-Controlled Trial — Journal of Alzheimer's Disease (2016)Oxford 1bPMID 26519439
- [9]Zhang Y, Chen C, Lu L, et al.. Association of magnesium intake with sleep duration and sleep quality: findings from the CARDIA study — Sleep (2022)Oxford 2bPMID 34883514
- [10]Arab A, Rafie N, Amani R, Shirani F. The Role of Magnesium in Sleep Health: a Systematic Review of Available Literature — Biological Trace Element Research (2023)Oxford 2aPMID 35184264
- [11]Rawji A, Peltier MR, Mourtzanakis K, et al.. Examining the Effects of Supplemental Magnesium on Self-Reported Anxiety and Sleep Quality: A Systematic Review — Cureus (2024)Oxford 2aPMID 38817505
- [12]Boyle NB, Lawton C, Dye L. The Effects of Magnesium Supplementation on Subjective Anxiety and Stress-A Systematic Review — Nutrients (2017)Oxford 2aPMID 28445426
- [13]Pouteau E, Kabir-Ahmadi M, Noah L, et al.. Superiority of magnesium and vitamin B6 over magnesium alone on severe stress in healthy adults with low magnesemia: A randomized, single-blind clinical trial — PLoS One (2018)Oxford 2bPMID 30562392
- [14]Tarleton EK, Littenberg B, MacLean CD, Kennedy AG, Daley C. Role of magnesium supplementation in the treatment of depression: A randomized clinical trial — PLoS One (2017)Oxford 2bPMID 28654669
- [15]Zhang X, Li Y, Del Gobbo LC, et al.. Effects of Magnesium Supplementation on Blood Pressure: A Meta-Analysis of Randomized Double-Blind Placebo-Controlled Trials — Hypertension (2016)Oxford 1aPMID 27402922
Follow us on Google Search
Add BiohackingHub as a preferred source to see more of our research in Google results.
Related protocols
Partner with BiohackingHub
Editorial placements, sponsorships, and product features for evidence-minded brands.