Magnesium for sleep: which form, what the evidence says, how to test it on yourself
Updated: 2026-09-25
The honest answer to "which magnesium form is best for sleep" is: there's no trial that has ever compared the forms head-to-head for sleep outcomes. What does exist is a small meta-analysis of magnesium in older adults with insomnia (better sleep-onset latency, weak evidence overall) and one RCT of magnesium L-threonate whose authors include employees of the ingredient manufacturer. Below: the actual numbers, what each study can and can't tell you, and a Health OS protocol to find out what works for your own sleep.
What the meta-analysis in older adults actually found
A 2021 systematic review and meta-analysis pooled 3 RCTs of oral magnesium vs. placebo in 151 older adults with insomnia. Magnesium cut sleep-onset latency by 17.36 minutes on average (95% CI -27.27 to -7.44, p=0.0006) — a real, statistically significant effect. Total sleep time improved by 16.06 minutes, but that result was not statistically significant.
The catch: all three trials were rated at moderate-to-high risk of bias, and the authors graded the overall quality of evidence as low to very low. Their own conclusion: the literature isn't solid enough for physicians to make a well-informed recommendation — but since magnesium is cheap and widely available, RCT evidence may support using magnesium supplements (under 1 g, up to three times a day) for insomnia symptoms.
Which form — and the gap nobody fills
No RCT found in this search compared magnesium forms head-to-head on sleep outcomes. A separate 2021 review of 31 sleep-supplement RCTs found clear pooled benefits for amino acids, melatonin and vitamin D — but not enough poolable data on magnesium specifically to run the same analysis; the authors explicitly called for more research on magnesium (alongside zinc, resveratrol and nitrate).
So the form question is mostly unanswered by controlled trials: neither glycinate nor L-threonate has been tested against the other, or against any other form, in a dedicated sleep trial.
The L-threonate trial, read carefully
In 2024, an 80-person, double-blind, placebo-controlled RCT gave adults aged 35-55 with self-reported sleep problems 1 g/day of magnesium L-threonate for 21 days. Compared to placebo, the L-threonate group showed significant (p<0.05) improvements in Oura-measured deep sleep score, REM sleep score and light sleep time, plus better readiness and activity scores. Subjectively, participants also reported better mood, energy, daytime productivity and mental alertness. No safety issues were reported.
Worth knowing before you read too much into it: this is the only trial found in this search, and three of the six authors are employed by the ingredient manufacturer that supplies magnesium L-threonate. That doesn't make the data wrong, but it does mean it hasn't been independently replicated — treat it as a promising first result, not settled science.
Who should be careful
Both Health OS magnesium protocols carry the same core caution: if you have kidney disease, consult a doctor. The L-threonate protocol also flags a possible mild sedative effect and lists GI upset (2+ days) or headache/dizziness as reasons to stop.
As always: talk to your doctor if you're pregnant, on medication, or managing a chronic condition before starting any supplement.
How to test it on yourself with Health OS
Health OS has two ready-made protocols. "Magnesium Glycinate at Night": 400 mg magnesium glycinate about 60 minutes before bed, 14 days baseline vs. 14 days intervention, tracking deep sleep hours and HRV (success = either one improving by 10%+), with sleep score as a secondary check and a minimum of 7 nights of data for confidence.
"Magnesium L-Threonate": 144 mg (2 capsules) of L-threonate 1 hour before bed, 7 days baseline vs. 7 days intervention, tracking deep-sleep percentage and sleep fragmentation, with morning energy rated 1-10 as a secondary measure — closer in design to the actual 2024 trial, just shorter. Run either one and compare your baseline vs. intervention phase averages to see what actually works for you.
FAQ
What's the best magnesium for sleep — glycinate or threonate?
No RCT has compared them directly on sleep outcomes, so there isn't a data-backed answer. The general insomnia meta-analysis in older adults didn't specify which form was used, so it doesn't support glycinate specifically. L-threonate has its own small RCT, whose authors include employees of the ingredient manufacturer, with promising results. Testing both on yourself with Health OS is currently the only way to actually know which works for you.
Does magnesium glycinate actually help you sleep?
A meta-analysis of magnesium generally (not glycinate specifically) in older adults with insomnia found it shortened sleep-onset latency by about 17 minutes, though the underlying trials were low quality. No glycinate-specific RCT on sleep was found in this review.
How strong is the evidence for magnesium and sleep?
The evidence is real but limited: one meta-analysis pooling 3 small trials with low-quality evidence, and one L-threonate RCT with industry-affiliated authors. That's grounds for cautious interest, not certainty.
How long before I'd notice an effect on my sleep?
The RCTs reviewed here ran 21 days (L-threonate) to several weeks. The Health OS protocols run 14 days of intervention for glycinate and 7 for L-threonate — that's a protocol design choice, not a separately proven threshold, but it's a reasonable guide before judging the effect on your own data.
Is it safe to take magnesium every night?
In the trials reviewed here, no serious safety issues were reported. The main documented caution is kidney disease — the Health OS protocols recommend talking to a doctor first if that applies to you, or if you're pregnant or on other medication.