Foods Highest in Magnesium

Magnesium is one of the few nutrients where I can tell a patient why they are short of it without knowing anything about them. It is the atom sitting at the centre of the chlorophyll molecule. That one fact decides everything else: magnesium arrives in a diet attached to green leaves, seeds, beans and grains that still have their germ and bran. Mill those off and the magnesium goes with them. A diet built on white flour is low in magnesium by construction, and no amount of care with the rest of it will fix that.

Roughly half of American adults take in less than the estimated average requirement [2]. I want to be plain about what that means. This is not a deficiency state I diagnose in clinic. It is an ordinary gap in an ordinary diet, and most of the people who have it will never know.

The foods, by what a portion actually delivers

Ranked by the amount in a portion someone would actually eat, not per 100 g — ranking per 100 g puts seeds and dried herbs on top and tells you very little about a plate. Amounts are USDA FoodData Central values per 100 g converted to the portion shown. “Share of a day” is against the 320 mg reference intake for a woman aged 31-50, which is the higher bar for most of these; a man's target appears in the table below.

Every figure there is a laboratory-analysed USDA value per 100 g, converted to the portion shown [5]. They are the same numbers the calculators on this site work from, so the page and the tools cannot drift apart.

Two things in that table matter. One is how fast it adds up — three or four of those portions in a day covers an adult, with no heroic single food required. The other is that the list is almost all plants, and I did not arrange that. It follows from where the mineral lives. Nuts, seeds, pulses, leafy greens, intact grains. That is the whole map.

Want this checked against a day you really ate, instead of a list? What nutrients am I missing? takes one day of food and reports what it covered and what fell short, magnesium included, against the reference intakes for your own age and sex.

How much you need

NIH Dietary Reference Intakes, the same figures the site's nutrient tools use. Pregnancy and breastfeeding raise several of these; the nutrient gap tool applies the right set for you.

Look at what that table does not do. The requirement never falls. It climbs through adolescence, settles, and stays where it is for the rest of your life. Meanwhile appetite shrinks, medication lists lengthen, and absorption slips a little. Nobody changes what they eat, and the gap opens anyway. That is most of what I see in older patients.

What magnesium actually does

Magnesium is a cofactor in more than three hundred enzyme systems [1]. That number has never once helped me explain anything. Four specific jobs do, because between them they explain why running low feels the way it does.

It makes ATP usable

Every textbook calls ATP the energy currency of the cell. What I was not taught, and had to look up years later, is that ATP does nothing on its own. It is inert until a magnesium ion binds to it. The molecule that actually spends energy is Mg-ATP. So every reaction in every cell that costs energy depends on magnesium at the moment of payment. That is why the first complaints are always the vague ones — tired, weak, off your food. Nothing has broken. Everything is short of small change.

It is the brake on calcium

Calcium makes muscle contract and nerves fire. Magnesium sits across from it, competing for the same channels and holding that excitability down. A muscle cell short of magnesium has a weak brake, and that is the whole story behind the cramps, the eyelid twitch and the restless legs people bring me. At the severe end it is tetany and seizures. The same antagonism relaxes the walls of blood vessels, and it is the reason intravenous magnesium is the treatment for an eclamptic seizure and not merely an addition to it.

Insulin does not work properly without it

The pancreas needs magnesium to release insulin, and the insulin receptor needs it to pass the signal along once insulin arrives. That makes the relationship with type 2 diabetes run both ways. Low magnesium worsens insulin resistance; high blood glucose flushes magnesium out through the kidney and lowers it further. It is a loop, not a line. I have set out the wider metabolic picture in therapeutic lifestyle changes.

It holds potassium inside the cell

The sodium-potassium pump runs on magnesium. That has a consequence which catches people out on the wards as often as in the consulting room: a low potassium that will not correct however much potassium you give is usually a magnesium problem. Replace the magnesium and the potassium follows on its own. It is also why the two fall together in almost anyone on long-term diuretics.

What chronically low intake actually causes

Two very different things both get called magnesium deficiency, and running them together is how the internet ends up recommending magnesium for everything.

Frank deficiency — uncommon, and hard to miss

True hypomagnesaemia, in someone with working kidneys and no drug or bowel cause, is rare. When I have seen it, it was not subtle. Cramps and visible muscle twitching, tremor, numbness, a change in personality, arrhythmias. Alongside those, the two electrolyte signatures I described above: a stubborn low potassium, and a low calcium that also refuses to move, because parathyroid hormone needs magnesium both to be secreted and to work [1]. Torsades de pointes gets intravenous magnesium whatever the measured level says. All of that is a hospital picture. It belongs to your doctor, not to a website.

Chronic low intake — common, and quiet

The situation this page is really about is the other one. Years of intake below requirement, a normal blood test the whole way through, and no symptom anyone would think to mention. Prospective cohort studies keep linking low habitual magnesium intake to higher rates of type 2 diabetes, high blood pressure, cardiovascular disease and osteoporosis [1].

Now let me be careful, because this is where these pages usually go wrong. Those are associations from observation, not results from trials. People who eat more magnesium are also eating more beans, greens, nuts and whole grains, so the diet explains the outcome perfectly well without magnesium being the active part. That confounding is the argument of my supplements article, and it is why the honest conclusion here is eat the foods and not buy the mineral. One exception has real trial evidence behind it: migraine, where the American Academy of Neurology and American Headache Society rate magnesium as probably effective for prevention [6]. That is a genuine indication, at doses above anything food supplies, and worth setting up with a clinician.

Why a normal blood test does not settle it

This part surprises patients, and a fair number of doctors. Under one per cent of your magnesium is circulating in blood. The rest is in bone and inside cells, and the body will strip magnesium out of bone to keep the circulating number looking normal. A normal serum magnesium therefore sits quite comfortably alongside depleted stores [1].

I am not telling you to ignore the test. A low serum magnesium is a real finding and needs chasing. I am telling you not to read a normal one as an all-clear. For most people the more informative question is what they habitually eat, and that is a question they can answer without anyone's permission.

Who tends to run short

Beyond the general pattern of eating, a few groups are genuinely more likely to be depleted:

• Anyone on a proton pump inhibitor for years. The FDA put out a safety communication in 2011 after cases of low magnesium in people taking PPIs for a year or more [3]. If you have been on omeprazole since you can't remember when, raise it.
• Loop and thiazide diuretics, which push magnesium out in the urine.
• Type 2 diabetes, for the reason in the loop above.
• Crohn's disease, coeliac disease, or chronic diarrhoea, where the limiting step is absorption and not intake.
• Heavy drinking, which manages poor intake, urinary loss and impaired absorption all at once.

Food or a supplement?

This is the question people are really asking when they land on a page like this one. For magnesium the answer is clearer than it is for most nutrients.

Why food comes first

Three reasons, in the order I think they matter.

• There is no upper limit on magnesium from food. A working kidney clears any dietary excess without trouble, so overdoing it on beans and greens is not a thing that happens. The limit that does exist applies only to supplements, and I have put it in the box below.
• The food brings the rest of the parcel. Go back up to that table: everything near the top also carries potassium, fibre and folate, and the potassium shortfall in a typical Western diet is worse than the magnesium one. A tablet gives you one mineral. A cup of black beans gives you four things you are probably short of.
• It is what the evidence supports. The associations above came from studies of what people ate. Reproducing a dietary finding with a supplement is a step that fails again and again across nutrition, often enough that it is the spine of that supplements article.

When a supplement is the right answer

Food first has never meant food only, and I prescribe magnesium in these situations without hesitation:

• A documented low level on a blood test, which needs a cause found as well as a number corrected.
• Ongoing loss rather than low intake — long-term PPIs, a loop or thiazide diuretic, malabsorption. Eating more may simply not keep up.
• Migraine prevention, at the doses used in the trials behind that guideline rating [6].
• Pregnancy conditions treated with magnesium sulfate. That is obstetric care, not shopping.

What is not on my list: taking it because an advert mentioned sleep, or because it helped your sister. If you cannot say which of those four applies to you, the first step is finding out what you are eating.

If you do take one, the form matters more than the dose

Magnesium oxide is the cheapest salt, the one most likely to be on the shelf, and the worst absorbed. Much of an oxide dose stays in the bowel and works as a laxative instead. The organic salts — citrate, lactate, glycinate, aspartate — absorb better and sit better with people [1]. Two things get missed. Magnesium blocks the absorption of tetracycline and quinolone antibiotics, bisphosphonates and levothyroxine, so those need a couple of hours' separation. And if your kidney function is reduced, magnesium can build up to genuinely dangerous levels; that decision belongs to your own doctor [1].

So the practical answer is dull, which in nutrition is usually the sign it is right. A handful of pumpkin seeds. Beans in something twice a week. Greens most days, and grains that still have their bran. That will move this number further than a bottle will, and it improves three or four other numbers on the way past.

Related, and free: top foods by nutrient ranks the whole USDA database for six nutrients against your own daily values, and the calorie and nutrient counter looks up any single food.

Questions people actually ask about this

Which foods are highest in magnesium?

By a portion someone would actually eat: pumpkin seeds, cooked spinach, Swiss chard, sunflower seeds, tempeh, black beans and quinoa all deliver roughly 120 to 165 mg — between a third and a half of a day in one serving. Ranked per 100 g instead, seeds and dried herbs dominate and the answer stops being useful, because nobody eats 100 g of pumpkin seeds.

How much magnesium do I need a day?

The NIH Dietary Reference Intake is 420 mg a day for men aged 31 and over and 320 mg for women, having risen from 400 and 310 mg in the 19 to 30 band. Requirements do not fall with age, which is part of why shortfall is common in older adults.

Why is magnesium so often low in modern diets?

Magnesium sits at the centre of the chlorophyll molecule, so it arrives with green leaves, seeds, beans and whole grains. Refining a grain strips the germ and bran where most of it lives, and a diet built on refined starch is therefore low in magnesium almost by construction. Analyses of US national dietary survey data find roughly half of adults taking in less than the estimated average requirement.

Can a blood test tell me if I am low in magnesium?

Not reliably. Less than one per cent of the body's magnesium is in the blood, and the body defends that number by pulling magnesium out of bone and cells. A normal serum magnesium is therefore compatible with genuinely depleted stores, which is why what you habitually eat is the more informative question.

Should I take a magnesium supplement?

Food first, for a specific reason: the tolerable upper intake level of 350 mg a day applies to supplemental magnesium only, not to magnesium from food, and the commonest side effect of exceeding it is diarrhoea. If you take a proton pump inhibitor long term, a loop or thiazide diuretic, or you have type 2 diabetes or a malabsorption condition, that is a conversation to have with your own doctor rather than a decision to make in a supplement aisle.

Does cooking destroy magnesium?

Magnesium is a mineral, so it cannot be destroyed by heat, but it does leach into cooking water. Boiling vegetables and discarding the water loses some; steaming, roasting, or using the liquid in a soup or stew keeps it. Cooked spinach appears high on this list partly because cooking collapses the volume, so a cup holds far more leaf than a raw cup does.