Blog
In older adults with insomnia, taking plain magnesium can take roughly 17 minutes off how long it takes to nod off. That’s the headline. What it won’t reliably do is keep you asleep longer, or sort out a healthy adult who already sleeps fine, despite what the back of the tub would have you believe.
If you do want to try it, the form I point clients to first is magnesium glycinate, 200 to 400 mg of elemental magnesium, taken about an hour before bed. No trial has ever pitted the forms against each other and named a winner, so treat anyone selling you “the best magnesium for sleep” with a raised eyebrow.
That’s the honest version. The detail, the caveats, and how to effectively use magnesium to help sleeping can be read below.
Magnesium is the fourth most abundant mineral in your body, and it’s busy. It has a hand in over 300 enzyme reactions, plenty of them tied to how the nervous system fires.
Here’s a quirk worth knowing. Less than 1% of your magnesium is in your blood. The rest is locked away in bone, muscle and soft tissue, which is exactly why a normal blood test can come back fine while your overall stores run low (NIH Office of Dietary Supplements, 2022).
On the labelling, the law is strict. Under EU and UK authorised health claims, magnesium “contributes to normal functioning of the nervous system” and “to the reduction of tiredness and fatigue” (EU Regulation 432/2012). What it can’t legally claim, and what I won’t claim for it, is to “cure insomnia”.
Some of us, yes. Around 19% of UK adults in their twenties take in less than the Lower Reference Nutrient Intake for magnesium, going by NDNS data (Derbyshire, 2018). Call it one in five.
But low intake and clinical deficiency are not the same thing, and the supplement aisle leans hard on blurring them. True hypomagnesaemia, where serum magnesium drops below 0.75 mmol/L, only hits about 2 to 3% of people. So a fifth of UK adults are eating below the recommended floor; far fewer are genuinely deficient. Keep those two figures separate in your head.
For the targets: the NHS puts the reference intake at 300 mg a day for men aged 19 to 64 and 270 mg for women. The European Food Safety Authority sits a touch higher, with an Adequate Intake, note, not an RDA, the evidence wasn’t strong enough to set a full population requirement, of 350 mg for men and 300 mg for women (EFSA, 2015).
Start with the best evidence we’ve got. A 2021 meta-analysis by Mah and Pitre in BMC Complementary Medicine and Therapies pooled three randomised controlled trials, 151 older adults with insomnia between them. What’s telling is the gap between its two results. On getting to sleep, magnesium delivered: sleep onset latency fell by 17.36 minutes (95% CI −27.27 to −7.44, p=0.0006), a real effect and a statistically sound one. On staying asleep, it didn’t quite. Total sleep time edged up 16.06 minutes but missed significance (p=0.15). Put plainly, you’ll likely nod off quicker, with no promise of a longer night. A 2024 correction ran in the same journal and left those numbers where they were.
Now the part nobody mentions on Instagram. Every person in those trials was an older adult with insomnia, not a student, not a nightshift nurse, not the stressed thirty-something who saw a reel about it. And the magnesium they swallowed was oxide or citrate, never glycinate, the form half the internet now swears by. Oxide in Abbasi (2012) and Held (2002); citrate in Nielsen (2010). Go looking for a glycinate sleep-onset trial and you’ll come up empty.
So here’s the awkward bit. When a brand says “studies show glycinate helps you sleep,” it’s leaning on results that used other forms entirely. Oxide and citrate are what got older insomniacs off to sleep faster. Glycinate earns its place on tolerability and how kindly it treats your gut, not on a sleep trial, because there isn’t one to point to.
Three mechanisms get wheeled out to explain the calm, and they don’t all carry the same weight.
The cortisol one has the most behind it. Abbasi’s 2012 trial recorded a genuine fall in serum cortisol (p=0.008) in elderly insomniacs taking 500 mg of elemental magnesium as oxide daily, in two 250mg doses, across eight weeks. There’s mouse work hinting magnesium also tweaks the CRH/ACTH signalling upstream (Sartori et al., 2012), although, to be fair, corticosterone itself didn’t drop in that one. The catch is that the human effect mostly shows up in people who were short or sleeping badly to begin with, not in a well-fed thirty-year-old.
Then the NMDA angle. Magnesium plugs the NMDA glutamate receptor and quietens excitatory signalling, which is about the cleanest mechanical reason for that switched-off feeling at bedtime.
And the glycine. On its own it nudges sleep along at 3 g before bed in a handful of small Japanese trials (Bannai et al., 2012; Yamadera et al., 2007; Inagawa et al., 2006). Trouble is, a 300 mg dose of magnesium glycinate only carries about 1.6 to 2 g of it, short of what those trials used, so whatever it adds is small.
You’ll also see “magnesium boosts GABA” repeated like gospel. That link is indirect at best. I wouldn’t bank a night’s sleep on it.
No head-to-head trial ranks the forms for sleep, so what follows is practitioner consensus, weighted to how well each one absorbs and how kind it is to your gut.
| Form | What it is | Tolerability | Where it fits for sleep |
|---|---|---|---|
| Magnesium glycinate (also sold as bisglycinate) | Magnesium bound to two glycine molecules | The gentlest of the lot, least likely to loosen stools | My first choice for sleep, on tolerability grounds |
| Magnesium citrate | Well absorbed | Laxative pull at higher doses | The form used in Nielsen et al. (2010); fine if your gut copes |
| Magnesium oxide | Cheap, poorly absorbed | Likely to loosen things up | Oddly, the form behind most of the positive sleep trials in older insomniacs |
| Magnesium L-threonate | Pitched for getting magnesium into the brain, on the back of Slutsky et al. (2010) in rodents | Generally well tolerated | No solid independent human sleep RCT. Early days, and pricey |
| Magnesium taurate | Magnesium paired with taurine | Generally well tolerated | Plausible on paper; human sleep trial evidence is basically absent |
If you take one line from that table, take this: pick glycinate for sleep, and don’t waste money on oxide. The NIH ODS is blunt that small studies have found oxide and sulfate to be less bioavailable than the aspartate, citrate, lactate and chloride forms (NIH ODS, 2022).
So, where to start: 200 to 400 mg of elemental magnesium as glycinate, 30 to 60 minutes before bed. Begin at the low end. If your stools loosen, drop the dose.
The US Food and Nutrition Board sets the tolerable upper intake for supplemental magnesium at 350 mg a day, and the NHS reckons up to 400 mg from supplements is unlikely to do harm (NIH ODS, 2022; NHS, 2020). That ceiling is for supplements. Magnesium from food doesn’t count against it, so eat your greens freely.
Some people genuinely should run it past their GP first. Don’t start a supplement on your own if you:
Overshoot the dose and the body lets you know the same way every time: loose stools, nausea, cramps. It usually settles on its own and it’s your cue to ease off.
Before the supplement aisle, raid the fridge. Magnesium turns up across a lot of everyday food, and a plate of it brings fibre and other minerals a tablet leaves behind. The figures below come straight from the NIH ODS magnesium fact sheet (Table 2).
|
Food |
Serving | Magnesium |
|---|---|---|
| Pumpkin seeds, roasted | 1 oz (28 g) | 156 mg |
| Chia seeds | 1 oz | 111 mg |
| Almonds, dry-roasted | 1 oz | 80 mg |
| Spinach, boiled | ½ cup | 78 mg |
| Cashews, dry-roasted | 1 oz | 74 mg |
| Peanuts, oil-roasted | ¼ cup | 63 mg |
| Black beans, cooked | ½ cup | 60 mg |
| Edamame, cooked | ½ cup | 50 mg |
| Salmon, cooked | 3 oz | 26 mg |
| Avocado, cubed | ½ cup | 22 mg |
A handful of pumpkin seeds, a portion of spinach and a square or two of dark chocolate, and most adults clear 200 mg before a supplement gets a look-in.
Here’s the loop I see in clients, though. The magnesium-rich foods are the ones that need cooking, and broken sleep is exactly when you can’t be bothered to cook. So you eat worse, and you miss the magnesium that might have helped in the first place. We don’t make magnesium supplements, but our GSN signature meals range is built around the same oily fish and vegetables in that table, a firecrackersalmon among them, and they go from freezer to plate in minutes.
“Magnesium may help with sleep problems, especially if they are related to scenarios which are caused by a deficiency of magnesium, such as leg cramps.” Dr Abhinav Singh, MD, MPH, FAASM, Medical Director, Indiana Sleep Center; Clinical Assistant Professor, Marian University.
Read that “if they are related to” again. It’s carrying the whole sentence, and it’s the fairest summary of where magnesium actually fits.
Magnesium can’t out-muscle bad sleep habits. Before you spend on a supplement, sort the basics.
Caffeine first. Its half-life runs anywhere from 2 to 10 hours depending on the person (O’Callaghan et al., 2018), and in that study even a dose six hours before bed measurably messed with sleep (Drake et al., 2013). Stop earlier than feels necessary. If you want the timing in detail, we’ve covered how long caffeine actually lasts separately.
Then alcohol, the great pretender. It “causes a reduction in sleep onset latency, a more consolidated first half sleep, and an increase in sleep disruption in the second half of sleep” (Ebrahim et al., 2013). The nightcap knocks you out and then wrecks the back half of the night.
And don’t overlook sleep apnoea. It’s estimated to affect 936 million adults aged 30 to 69 worldwide (Benjafield et al., 2019, Lancet Respiratory Medicine). If you snore heavily and feel shattered all day, that’s a GP referral, not a supplement. No mineral fixes a breathing problem.
Light, room temperature, a consistent bedtime. Dull, I know. Still beats any pill in the published literature.
Magnesium earns its place as a low-risk add-on if you’re an older adult with insomnia or your intake is genuinely low. It is not a sleeping pill. A modest head start on falling asleep, no guarantee on sleeping longer, and no substitute for fixing the caffeine, the wine, the screens and the bedtime first.
Want to give it a fair go? 200 to 400 mg of elemental magnesium as glycinate, 30 to 60 minutes before bed, with food. Two to four weeks. Stop if your stomach turns or your stools loosen. And check with your GP if you take regular medication or have any kidney concerns.
For the bigger picture on evening eating, here’s what to eat before bed for better sleep. And if cooking is your sticking point, the salmon and vegetable dishes in GSN signature meals come straight out of the freezer in about eight minutes.
Does magnesium actually help you sleep?
It can, but don’t wait for a knockout blow. The clear benefit shows up in older adults with insomnia who were low to start with, and even then it’s about 17 minutes faster to sleep. Total sleep time didn’t improve significantly. Worth a try. Not a miracle.
What’s the best magnesium for sleep?
No trial has crowned one. Glycinate is the practitioner pick on absorption and a gentle gut. Citrate works too if your stomach tolerates it. Oxide is poorly absorbed, even though, oddly, it’s the form most of the positive trials used.
What’s the right magnesium glycinate dosage for sleep?
Most people are fine on 200 to 400 mg of elemental magnesium. Start low. Keep supplements at or below 350 mg a day, the tolerable upper intake, though the NHS says up to 400 mg is unlikely to harm.
When should I take magnesium for sleep?
Roughly 30 to 60 minutes before bed, with food, same time each night. Give it two to four weeks before you decide.
Can I get enough magnesium from food alone?
For most people, yes. Pumpkin seeds, greens, nuts, a little dark chocolate across the day gets you close to the UK target, with fibre and minerals a tablet skips.
Is it safe to take magnesium every night?
For most healthy adults, yes, as long as you stay near 350 mg or below from supplements. Kidney problems or regular medication, including some antibiotics, bone-density drugs or proton pump inhibitors? Check with your GP first.
Can you take too much magnesium?
From food, no, your gut and kidneys deal with the surplus. From supplements, yes. Past roughly 350 mg a day you risk loose stools, nausea and cramps, and very high doses are dangerous for anyone with poor kidney function.
Health Director at Health Labs HQ
Sam Gaunt is Health Director at Health Labs HQ, a Fellow of the Royal Society for Public Health (FRSPH), a member of the International Stress Management Association (MISMA), and holds the Advanced Nutrition Diploma (Level 5, Future Fit Training). She spends most of her week helping shift workers, new parents and frazzled professionals sleep better. She’s also a long-standing GSN customer. Read her full credentials on her author profile.