Magnesium Glycinate vs Threonate vs Citrate: Which Form Actually Works for Sleep?

Three forms of magnesium sit on the same shelf, marketed for the same job. They are not the same compound. They are not interchangeable. The label "magnesium for sleep" is doing a lot of work for very different molecules.

If you take "magnesium" at night and could not name the form on the bottle, this article is for you. We are going to look at what each form actually does, which one has the strongest sleep trial behind it, the dose that matches the evidence, the drug interactions worth knowing, and why most "sleep gummies" sold at $40 a bottle contain the wrong magnesium for the job.

Key takeaways

• Form beats milligrams. The number on the front of the bottle means little; the carrier molecule decides how much you actually absorb.
• Glycinate is the form most often recommended for sleep. Its direct human evidence is one 2025 trial showing a small improvement in insomnia scores; the often-quoted sleep-onset figures come from trials of other forms, mostly oxide.
• Dose and timing: 200 to 400mg elemental, 60 to 90 minutes before bed, judged across 14 nights.
• No cycling needed, and it is among the best-tolerated forms.
• It is a cofactor, not a sleep drug. Fix caffeine, alcohol and screens first.

Table of contents

• The basic biochemistry
• What the strongest sleep trial actually says
• Threonate: cognition first, sleep data emerging
• Citrate: cheap, mostly for bowels
• How much magnesium actually absorbs
• The marketing distortion
• The protocol that aligns with evidence
• Who gets the biggest effect
• When magnesium does not work
• A note on drug interactions
• FAQ

The basic biochemistry

Magnesium is not one substance you buy in a bottle. It is an element bound to a carrier molecule. The carrier decides where the magnesium goes, how much your gut absorbs, and what it does once it gets there.

Three carriers dominate the sleep aisle.

Glycinate binds magnesium to glycine, an amino acid that doubles as an inhibitory neurotransmitter in the central nervous system. Glycine itself has documented effects on subjective sleep quality and slow-wave sleep architecture (small trials, Yamadera 2007). The magnesium hitchhikes on a molecule the brain already uses to slow down. Sometimes spelled bisglycinate, which means two glycine molecules per magnesium ion. Same compound, different name.

Threonate binds magnesium to threonic acid, a vitamin C metabolite. Marketed as the form that crosses the blood-brain barrier most efficiently. The claim comes from one rodent paper.

Citrate binds magnesium to citric acid. Cheap to manufacture. Highly soluble. Studied mostly in the context of bowel motility, not sleep.

Three molecules. Three carriers. Three destinations in your body. Pretending they are the same compound because they share the word "magnesium" is like pretending sugar and salt are the same compound because they share the words "white crystal."

What the strongest sleep trial actually says

The Abbasi paper is the one cited most. Worth reading carefully, not just summarized.

Abbasi B et al. 2012 (Journal of Research in Medical Sciences, PMID 23853635), double-blind, placebo-controlled, n=46, elderly subjects with diagnosed insomnia. Treatment: 500mg elemental magnesium per day as magnesium oxide (not glycinate), across eight weeks. Result: improved sleep efficiency, decreased sleep onset latency, increased serum melatonin, decreased serum cortisol. Effect sizes were modest but statistically significant.

Three things to keep in mind.

First, n=46 is small. A real signal, not a finished case.

Second, the population was elderly insomniacs, not a 40-year-old executive after a board-prep week. The mechanism likely generalizes; the effect size is probably smaller in younger, less-deficient people.

Third, some arms used magnesium oxide, which has roughly 4% bioavailability. The fact that an effect appeared at all suggests the dose was high enough to overcome that absorption ceiling. Glycinate at the same elemental dose absorbs far better, which is why practitioner protocols converged on it.

A 2021 systematic review (Mah and Pitre, PMID 33865376), updated with a correction in December 2024, pooled three randomized trials of oral magnesium in 151 older adults with insomnia and found sleep onset about 17 minutes faster than placebo. The authors rated the evidence low to very low quality, and the review did not break results down by glycinate, citrate or threonate. Form-specific evidence comes from newer single trials: a 2025 randomized trial of magnesium bisglycinate (Schuster, PMID 40918053, 155 adults with poor sleep) found a small improvement in insomnia severity, and a 2024 randomized trial of magnesium L-threonate (Hausenblas, PMID 39252819, 80 adults, industry-funded) reported better sleep quality scores. The pooled estimate comes from older adults with insomnia; in younger, less-deficient people the effect is likely smaller.

Huberman Lab Episode 84 (the magnesium and sleep episode) walks the same literature. Huberman personally favors threonate for its brain-penetration hypothesis, but the human sleep evidence for threonate is still limited to a few small, industry-funded trials, and his glycinate dosing matches this guide: 200 to 400mg elemental, 60 to 90 minutes before bed. Do not expect any of them to behave like a prescription sedative.

Threonate: cognition first, sleep data emerging

Magnesium threonate has a different story.

The foundational paper is Slutsky et al. 2010 (Neuron), a rodent study showing that magnesium-L-threonate elevated brain magnesium concentration and improved learning and memory metrics in mice and rats. Solid for what it is: a basic-science cognition paper.

Human trials are still few. Early ones were small and aimed at cognition. Two newer randomized, placebo-controlled trials did target sleep: Hausenblas 2024 (PMID 39252819, 80 adults, 21 days) reported better sleep quality scores and ring-measured deep and REM sleep scores, and Lopresti 2025 (PMID 41601871, 100 adults, 6 weeks) found less self-reported sleep-related impairment but no difference in ring-measured sleep. Both were funded by or linked to the ingredient supplier, and no meta-analysis has pooled them yet. Threonate also claims superior brain entry across the blood-brain barrier, but human bioavailability data is thin; most estimates are extrapolated from rodent or in-vitro work.

Use case for threonate as it currently stands: a hypothesis for focus and cognition support, with promising but thin human evidence. It is the most expensive form on the shelf, which does a lot of the perceived work. If you want it for daytime cognition, take it in the morning. For sleep, the threonate evidence is early, small and industry-funded.

Citrate: cheap, mostly for bowels

Magnesium citrate is the form your gastroenterologist prescribes before a colonoscopy. That is not a metaphor. It is what it does at high doses.

At lower supplemental doses (200 to 400mg elemental), citrate has decent bioavailability and works fine as a general magnesium repletion strategy. For sleep specifically, the evidence is thinner than for glycinate but not zero.

The catch: if your gut is sensitive, you will know by the second night. Citrate's osmotic effect at moderate doses can produce loose stools. Many "sleep magnesium" gummies are citrate because it is cheap to formulate. You then take four gummies because the dose per gummy is small, plus 12g of sugar before bed. The product is fighting itself.

How much magnesium actually absorbs

This is the single most useful thing to understand, so look at the math, not the front label.

"Extra strength, 1000mg magnesium" in oxide form, at about 4% bioavailability, delivers roughly 40mg of usable magnesium. A single 250mg glycinate capsule, at roughly 35% absorption, delivers about 88mg. The smaller, cheaper-sounding capsule puts more than twice as much magnesium into you. Total dose printed on the bottle is meaningless without the form.

Here is how the common forms stack up on absorption:

• Oxide: roughly 4% absorbed. Cheap, poor uptake, strong laxative effect.
• Citrate: roughly 30% absorbed. Good uptake, but osmotic, so it loosens stools at higher doses.
• Glycinate: roughly 35% absorbed. The strongest sleep data, and the gentlest on the gut.
• Threonate: human absorption figures are largely extrapolated from animal data; the brain-penetration claim is its selling point, not its sleep evidence.

The marketing distortion

Walk into any supplement store, search "magnesium sleep" on Amazon, scroll the wellness section of Whole Foods. You will see three patterns.

Pattern one: the proprietary blend curtain. "Sleep Magnesium Complex 500mg." Of what? The breakdown will say "Magnesium Blend" with no per-form mg. That means the formulator is hiding the ratio, usually cheap oxide or citrate padding with a sprinkle of glycinate listed first.

Pattern two: the gummy problem. Gummies require sugar, gelatin, citric acid and binders. The active dose per gummy is small. You then take four gummies and swallow 12g of sugar before bed, in the same hour you are trying to protect your sleep architecture.

Pattern three: the "extra strength" trap. "1000mg Magnesium" in oxide form at 4% bioavailability delivers about 40mg absorbed. A 250mg glycinate capsule delivers more usable magnesium. Total dose is meaningless without form.

The protocol that aligns with evidence

If you want a protocol that mirrors what the trials actually showed:

• Form: glycinate (bisglycinate), named on the label, no proprietary blend.
• Elemental dose: 200 to 400mg.
• Timing: 60 to 90 minutes before bed. Glycine peaks in plasma around 45 to 60 minutes.
• Duration: judge it across 14 nights, not one. Mean across two weeks.
• Cycling: not needed. Excess is cleared in urine; continuous repletion is the goal.

What we do at Apexzen: Magnesium Drift is magnesium glycinate, 275 mg elemental magnesium per 3-capsule serving, single ingredient, capsule format. No gummy sugar, no proprietary curtain, no oxide padding. The label states the elemental amount.

Who gets the biggest effect

The honest answer changes your expectations. The largest effects show up in older adults and in people with a documented magnesium deficiency. Abbasi's subjects were elderly insomniacs, and their cortisol drop was steeper than a younger, healthier person should expect.

If you are 30, eat leafy greens daily, and already sleep reasonably well, 200mg of glycinate at night is a 5-to-10-minute acceleration and a slightly calmer baseline, not a transformation. If you are over 40, chronically stressed, under-sleeping, and running low on dietary magnesium, you sit closer to the population where the trials found their signal. Magnesium rewards the deficient more than the already-replete.

When magnesium does not work

Honest section. Magnesium is not a sleep drug. It is a cofactor for the systems that make sleep possible.

If your sleep problem is fundamentally behavioral (caffeine at 4 PM, alcohol at 9 PM, phone in the bedroom), magnesium will not save the protocol. Fix the behaviors first.

If your sleep problem is clinical (diagnosed insomnia, sleep apnea, depression-driven insomnia), magnesium is an adjunct at best. See a sleep physician. This is a structure-function discussion, not medical advice.

If stress is what wrecks your nights, the bigger lever is usually a different ingredient. Abbasi's trial showed magnesium nudged cortisol down, but if "I cannot switch my brain off" describes your 2 AM, read our companion guide on ashwagandha and cortisol. And if you are coming off a high melatonin dose, see why 2mg melatonin works when 10mg does not before you switch.

A note on drug interactions

Magnesium can bind to certain medications in the gut and reduce how much of each gets absorbed, so timing matters. The common ones: bisphosphonates for bone density (such as Fosamax), tetracycline and quinolone antibiotics (such as doxycycline or ciprofloxacin), levothyroxine for thyroid, and some heart medications (such as digoxin). The fix is simple: separate magnesium from these by at least two hours. If you take prescription medication, it is a 30-second conversation with your pharmacist worth having before you start.

FAQ

Is bisglycinate the same as glycinate?

Yes. Two glycine molecules per magnesium ion, same compound, two names on different labels.

How much magnesium glycinate should I take for sleep?

200 to 400mg of elemental magnesium, 60 to 90 minutes before bed. Check the elemental number, not the compound weight.

How do I find the elemental magnesium on a label?

Look for "elemental magnesium" or "magnesium (as bisglycinate)" on the supplement facts panel. If the label only shows the compound weight (for example "500mg magnesium bisglycinate"), the elemental amount is roughly a quarter of that, so about 125mg. If a brand hides the elemental figure, that is a signal in itself.

Can I take magnesium glycinate every night?

Yes. There is no need to cycle it; excess is excreted in urine and continuous repletion is the point.

Does magnesium glycinate have side effects?

It is among the best-tolerated forms, gentle on the gut precisely because it avoids the osmotic effect that makes oxide and citrate loosen stools.

Will it make me drowsy?

Some people feel calm within 45 to 60 minutes; others only notice better sleep efficiency across two weeks. Both are normal.

Should I take glycinate and threonate together?

You can; there is no significant interaction. Most people take threonate in the morning (cognition) and glycinate at night (sleep).

About Apexzen Magnesium Drift

Our Magnesium Drift is bisglycinate, 275 mg elemental magnesium per 3-capsule serving, single ingredient, named on the label, no blend and no sugar. The first 100 customers, our Founding 100, keep 10% off for life.

These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.

Related reading on Apexzen Learn

• Why 2mg Melatonin Works When 10mg Does Not: The U-Shaped Dose Curve
• Ashwagandha for Cortisol: The 8-on/2-off Protocol Most Brands Skip