Magnesium: What It Really Does for Sleep and Nerves

Open any supplement forum and magnesium is in every second thread. One person writes that glycinate "gave them their sleep back". Another took it for six months and noticed nothing except an upset stomach. A third insists only threonate counts, because "it gets into the brain". All three are probably right. They just took different substances under the same word.

My own magnesium story started with an eyelid. The right one. It twitched for three weeks straight, then night cramps in the calves joined in, after a hot summer of long walks and a lot of sweat. The first thing I bought was cheap pharmacy "magnesium". Oxide. Two weeks, zero effect, unless you count my gut, which reacted immediately. Then I worked out the forms, switched, and things got interesting.

Below: what magnesium actually does in the body, why a blood test tells you almost nothing, how the forms differ from one another, and who it will not help at all.

What it is and why everyone suddenly cares

Magnesium is the fourth most abundant mineral in the body. An adult carries around 25 grams of it, distributed inconveniently for diagnosis: 50–60 per cent sits in bone, almost all the rest inside cells, and less than one per cent in blood serum. That is why a standard serum test often reads normal in someone with a real shortfall. Blood defends its concentration to the last, pulling magnesium out of bone and muscle to do it.

It takes part in more than 300 enzymatic reactions. The one that matters most: ATP, the cell's energy molecule, does not exist in the body on its own but bound to magnesium. Without it, ATP is biologically inactive. That is not one function among many; it is the foundation. Everything that needs energy — a muscle contracting, a nerve firing, a protein being built — needs magnesium.

Why the noise now. Three things converged. First, food: refined flour loses around 80 per cent of its magnesium in milling, and vegetables grown on depleted soils carry less than they did half a century ago. Large population surveys put roughly half of adults below the recommended intake of 400–420 mg a day for men and 310–320 mg for women. Second, stress: adrenaline and cortisol make the kidneys excrete magnesium faster, so a chronically tense person loses it quicker than they take it in. Third, fatigue with melatonin and sleeping pills: people want something that repairs rather than switches off. Magnesium looks convincing in that role. How justified that is comes next.

How it works: the nervous system and the muscles

There are several mechanisms, and they are well described.

First, a brake on excitation. Nerve cells carry NMDA receptors that pass excitatory signals. A magnesium ion physically sits in the channel of that receptor and lets a signal through only when the stimulus is strong enough. With too little magnesium, the plug weakens, neurons respond to weak triggers, and you get what people describe as "wired", "can't stop my thoughts", "jump at every sound". In parallel, magnesium supports GABA, the main inhibitory neurotransmitter, the same one many sleeping pills act on. Magnesium does not press the pedal, though. It makes sure the pedal works.

Second, muscle. Calcium starts a contraction; magnesium ends it by pushing calcium out of the cell. Short of magnesium, a muscle cannot fully let go: hence night cramps in the calves, a twitching eyelid, shoulders that stay tight even lying down. This is the most visible and most reliable sign of a shortfall, more reliable than any test.

Third, the stress axis. Magnesium helps regulate the hypothalamus-pituitary-adrenal response, meaning how hard and how long the body pours out cortisol. Here is the loop: stress drains magnesium, low magnesium amplifies the stress response. You can break it from either side.

Fourth, sleep directly. Magnesium is a cofactor for the enzyme that turns serotonin into melatonin, and it is involved in regulating circadian genes. In a trial with older adults with insomnia, 500 mg of magnesium for 8 weeks cut time to fall asleep by roughly 17 minutes and increased total sleep time. Modest? Yes. But without sedation and without dependence.

Now the caveat marketing usually skips. All of these effects reproduce in people who are short of magnesium. In someone who gets enough from food, a supplement does very little: meta-analyses on sleep find a small effect and grade the evidence as low quality. Magnesium is not a sleeping pill. It is replacement of what was missing. If it was missing, the effect is noticeable. If it wasn't, you will mostly notice it in your gut.

Forms: why "magnesium" is not one substance

This is where most people get lost. The label says "magnesium 400 mg", and that can mean very different things.

First thing to understand: magnesium in supplements is always bound to something, a salt or an amino acid. The weight on the label sometimes refers to the whole compound, sometimes to pure, elemental magnesium. The difference is severalfold. Look for the word "elemental".

Magnesium oxide. The cheapest form, about 60 per cent elemental magnesium, so the number on the box looks impressive. But absorption is poor, 4–10 per cent by various estimates. The rest travels through the gut and drags water with it. Oxide works well as a laxative, and it is what is most often sold as "marine magnesium" or simply "magnesium" in pharmacies. My first attempt was exactly this.

Magnesium citrate. About 16 per cent elemental, good absorption, moderate price. Mildly loosening, which is a plus for some and not for others. A solid working form for daytime and for anyone prone to constipation. For sleep it is acceptable but not ideal: citrate slightly energises some people.

Glycinate, also called bisglycinate. Magnesium bound to two glycine molecules. About 14 per cent elemental, high absorption, almost no effect on the gut. And a bonus: glycine is itself an inhibitory neurotransmitter, and in trials 3 grams of glycine before bed improved subjective sleep quality and reduced daytime sleepiness. So glycinate works on sleep through two channels. It is the form I settled on, and the one people usually mean when they say "magnesium for sleep".

Malate. Magnesium with malic acid, which takes part in the energy production cycle. A form for mornings and for muscle fatigue, not for evenings.

Threonate. The only form for which animal studies show a rise in brain magnesium concentration. Human data are still limited, the price is 3–5 times that of glycinate, and the elemental magnesium in a standard dose is small, about 100 mg. Interesting for cognitive goals; buying it "for sleep" is overpaying.

Taurate, chloride, lactate are working forms with good absorption, seen less often. Magnesium sulphate is bath salt, and, contrary to popular belief, only negligible amounts of magnesium cross the skin. A warm bath relaxes you on its own.

Simple rule: for sleep and nerves, glycinate; for the gut, citrate; for energy, malate; oxide only if a laxative is the actual goal.

Who this is for — and who magnesium will not help

Signs worth acting on if they persist beyond two or three weeks: calf cramps at night, a twitching eyelid, tension in the shoulders and jaw, irritability without cause, trouble falling asleep with a clearly tired body, a racing heart at rest. Especially against a background of heavy coffee, alcohol, sweaty training, heat, constant stress, or proton-pump inhibitors for heartburn or diuretics, all of which increase magnesium losses.

A separate group is people over 45. Gut absorption of magnesium falls with age, kidney excretion rises, and with them the share of people below the recommended intake. If morning grogginess comes with night cramps, this is worth checking, alongside what I covered in the piece on sleep quality and micro-awakenings.

Now, honestly, who it will not help.

If your sleep is being broken by apnoea, wine, a bedroom at 24 degrees, or anxiety that has lasted for months, magnesium is the wrong tool. It does not fix breathing and does not replace a conversation with a doctor. Leg cramps also come from low potassium, from vein problems, from certain medications; if a month of magnesium changed nothing, the problem is not magnesium.

And the group for whom supplementation is unsafe: people with impaired kidney function. Healthy kidneys clear excess magnesium easily; damaged ones do not, and it accumulates. With any kidney diagnosis, only after talking to a doctor. Caution too with myasthenia gravis and with tetracycline and fluoroquinolone antibiotics, bisphosphonates and levothyroxine, since magnesium reduces their absorption and needs a gap of two to four hours.

How to use it: dose, timing, duration

Dose. 200–400 mg of elemental magnesium a day from supplements; with a diet that includes greens, nuts and whole grains, 200–300 is usually enough. The upper limit for supplements set by nutrition institutes is 350 mg elemental a day; above that most people get diarrhoea. That is not toxicity, it is a signal the gut is no longer taking more. If you want a higher dose, split it in two.

Timing. For sleep and nerves, 1–2 hours before bed, with a small amount of food: glycinate is fine on an empty stomach, citrate not always. If the goal is cramps and energy, half in the morning and half in the evening.

Duration. Intracellular stores refill slowly. Muscle symptoms, cramps and the eyelid, usually go within 1–3 weeks. Sleep and tension levels should not be judged before 4 weeks. If nothing has changed in 6–8 weeks, stop; the cause lies elsewhere. Magnesium is not a supplement to "take forever": with a normal diet, a two- or three-month course closes the gap, and food maintains it afterwards.

Food that does the same job. Pumpkin seeds, around 150 mg of magnesium in 30 grams, are the record holder. Almonds, 80 mg in 30 grams. Buckwheat, about 250 mg in a cup of dry groats. Cooked spinach, 150 mg a serving. Dark chocolate at 70 per cent, 65 mg in 30 grams. Spirulina, roughly 20 mg in 10 grams, modest, but as part of a complete mineral profile. A handful of seeds and a plate of buckwheat cover half the daily requirement without a single capsule.

And how this fits the logic of TCM. Chinese medicine knew nothing of minerals, but its evening advice — a warm dinner without stimulants, greens and seeds, nothing cold and no "revving up" after sunset — turns out to be magnesium-rich by composition. TCM called it "calming the Shen", the spirit that lives in the heart and keeps you awake when agitated. The modern translation is lowering sympathetic tone. In that picture magnesium is no miracle, just one ingredient that dropped off the evening table along with whole foods. Adaptogens belong here too: if evening tension is about stress rather than minerals, I would look at a 30-day course of Lingzhi rather than raising the magnesium dose.

My result, for the record. Bisglycinate, 300 mg elemental, in the evening, since September. The eyelid stopped twitching on day nine, the cramps were gone within two weeks. By the tracker, falling asleep got about ten minutes faster, while the number of awakenings barely changed, because mine were being broken by other things, which I have written about separately. Magnesium did not work a miracle. It removed what was there precisely because of its absence. And that, to me, is an honest outcome for any supplement.

Conclusion

Magnesium is not a sleeping pill and not a "calming agent", but a mineral without which ATP, the brain's inhibitory receptors and muscle relaxation do not work. Half of people fall short, a blood test does not show it, and the most reliable sign is muscle that will not let go: cramps, eyelid, shoulders.

Form decides everything. Oxide is a laxative, citrate is for the gut, glycinate for sleep and nerves, malate for energy, threonate an expensive experiment. 200–400 mg elemental, in the evening, a 4–8 week course, then an honest assessment: did it change or not. If not, the cause is elsewhere and needs finding, not a bigger dose.

It works in those who were short of it. That is the whole truth about magnesium, and it is quite enough.