The short version: the form that costs the most has the least evidence behind it, and the form the marketing calls worthless is the one that actually works for the single indication it's best at. Magnesium citrate is the cheapest per milligram of real magnesium, has the best human absorption data of any form, and is the form named by a clinical guideline. Glycinate's genuine advantage is that it won't give you diarrhoea — which is worth paying for, but it isn't the advantage being advertised.
First, the number nobody shows you
A "1000 mg magnesium glycinate" capsule does not contain 1000 mg of magnesium. It contains 1000 mg of a magnesium salt, and magnesium is a minority of that weight.
| Form | % elemental Mg | mg Mg per 1000 mg of salt |
|---|---|---|
| Magnesium oxide | 60.3% | 603 |
| Magnesium hydroxide | 41.7% | 417 |
| Magnesium carbonate | 28.8% | 288 |
| Magnesium chloride (anhydrous) | 25.5% | 255 |
| Magnesium sulfate (anhydrous) | 20.2% | 202 |
| Trimagnesium dicitrate (anhydrous) | 16.2% | 162 |
| Magnesium malate | 15.5% | 155 |
| Magnesium bisglycinate (anhydrous) | 14.1% | 141 |
| Magnesium glycerophosphate | 12.5% | 125 |
| Magnesium chloride hexahydrate | 12.0% | 120 |
| Magnesium lactate | 12.0% | 120 |
| Magnesium bisglycinate dihydrate | 11.7% | 117 |
| Magnesium citrate nonahydrate | 11.9% | 119 |
| Magnesium sulfate heptahydrate (Epsom) | 9.9% | 99 |
| Magnesium taurate | 8.9% | 89 |
| Magnesium aspartate | 8.4% | 84 |
| Magnesium L-threonate | 8.3% | 83 |
| Magnesium orotate | 7.3% | 73 |
| Magnesium gluconate dihydrate | 5.4% | 54 |
These are stoichiometry, not estimates. Note how much hydration state matters — bisglycinate dihydrate is 11.7% where the anhydrous salt is 14.1%, and the dihydrate is the more common commercial material.
So that 1000 mg glycinate capsule delivers 117–141 mg of actual magnesium, against an Australian RDI of 310–420 mg.
Two real Australian products make the point better than the table does:
- Blackmores Magmin "500 mg" — the label says magnesium aspartate dihydrate 500 mg, magnesium 37.4 mg. That's 7.5%. You would need eleven tablets a day to reach a male RDI, against a recommended dose of one or two.
- Bioglan Active Magnesium Citrate — magnesium citrate 1.2 g, magnesium 184.8 mg. Five times the elemental magnesium of a Magmin tablet, in a cheaper pack.
Both labels are honest. Australian labelling convention requires the elemental figure in brackets. It's just on the back, in small type, while the number on the front is something else entirely.
If you take one thing from this page: ignore the number on the front of the pack. Find the bracketed elemental figure on the back.
Absorption: the marketing hierarchy is backwards
The hierarchy you'll see everywhere is glycinate > citrate > oxide. The human data support exactly one third of that.
The trial that actually compared them
**Walker AF, Marakis G, Christie S, Byng M. Magnesium Research 2003;16(3):183–91.** Randomised, double-blind, placebo-controlled, n=46, 300 mg elemental magnesium per day for 60 days, comparing an amino-acid chelate, citrate and oxide.
- Citrate produced the greatest mean serum magnesium both acutely (p=0.026) and chronically (p=0.006)
- Citrate produced the greatest salivary magnesium chronically (p=0.027)
- Oxide was indistinguishable from placebo
Read that again: citrate beat the amino-acid chelate. This is the closest thing in existence to a glycinate-versus-citrate comparison, and it went the opposite way to the marketing.
Confirmed again on a harder endpoint
**Kappeler D, Heimbeck I, Herpich C, et al. BMC Nutrition 2017;3:7.** Randomised crossover, n=20, 300 mg single dose, citrate vs oxide.
- 24-hour urinary magnesium: citrate 7.2 mmol vs oxide 6.7 mmol; difference 0.565 mmol, 95% CI 0.212–0.918, p=0.0034
- Serum magnesium higher with citrate at 2, 3, 4, 5 and 6 hours
- No significant difference in intracellular magnesium in monocytes, lymphocytes or erythrocytes
Note the honest result buried in a manufacturer-funded trial: citrate won on urine and short-term serum and tied on the cellular measure that arguably matters most. The absolute difference — 0.565 mmol over 24 hours — is about 14 mg of magnesium. Statistically clean, clinically trivial.
Does glycinate beat citrate?
We looked specifically for a randomised human trial comparing glycinate or bisglycinate against citrate on any absorption endpoint — serum, urinary, erythrocyte, isotopic or balance. As far as we can establish, no such trial exists.
What does exist:
- Walker 2003, in which the chelate lost to citrate on both endpoints that reached significance.
- **Schuette et al., JPEN 1994**, magnesium diglycinate versus oxide in patients with ileal resection. This is the paper most often cited for glycinate superiority. The comparator was oxide, the population was short-bowel patients, and the sample was small — none of which supports a general claim against citrate in healthy people. (We could not retrieve the full record to verify its figures; treat any number attributed to it as unverified.)
- A mechanistic story: that a chelated amino-acid complex is absorbed intact via dipeptide transporters, bypassing saturable paracellular uptake. Plausible. Also exactly the kind of mechanism that has repeatedly failed to translate — and here, the one trial that tested something like it found the chelate lost.
The honest verdict: citrate > oxide is robust. Citrate ≥ amino-acid chelate on the only comparison available. Glycinate versus citrate on absorption: no evidence either way, and the nearest proxy favours citrate.
What glycinate actually buys you
Tolerability, and it's real. Glycinate is markedly less laxative than citrate or oxide at the same elemental dose. If you need 300–400 mg a day and citrate sends you to the bathroom, glycinate lets you take the dose at all.
That is a comfort claim, not a bioavailability claim, and it should be priced as one. See the price section — you're paying two to four times more per milligram for it.
Magnesium L-threonate: the brain claim
What the rodent study showed
**Slutsky I, Abumaria N, Wu L-J, … Liu G. Neuron 2010;65(2):165–77.**
- 18-month-old male Sprague-Dawley rats, ~50 mg/kg/day elemental Mg as magnesium-L-threonate in drinking water, 24 days
- CSF magnesium rose 7% (p<0.05); corrected for sampling losses, ~15% (p<0.001)
- Magnesium chloride, citrate and gluconate did not significantly raise CSF magnesium
- Declared in the paper: "G.L. declares that he is a cofounder of Magceutics" — the company developing the compound
So the comparative claim was demonstrated, in aged rats, by the compound's commercial originator. The effect size is 7–15% on CSF magnesium. Not a step change.
The human trials
**Liu G, Weinger JG, Lu Z-L, et al. J Alzheimers Dis 2016;49(4):971–90 (MMFS-01). 51 randomised, 44 analysed, 12 weeks, ~25 mg/kg/day magnesium L-threonate. Composite cognitive z-score: week 12 Cohen's d = 0.91, p = 0.003**.
Three things about that result. First, at 8.3% elemental, the 1.5–2.0 g/day dose supplied only 125–166 mg of elemental magnesium — below a single RDI, and less than a cheap citrate tablet. Second, no CSF or brain magnesium was measured; only plasma, urine and erythrocyte. Third, the trial was funded by Neurocentria, designed jointly with Neurocentria, analysed by Neurocentria, and written by Neurocentria — and its first author is the same Guosong Liu who cofounded Magceutics and originated the molecule.
A d of 0.91 on a cognitive composite from n=44 under those conditions is hypothesis-generating. Effect sizes that large in cognition almost never replicate at scale, and this one has not.
Zhang C, et al. 2022, n=109, reported cognitive improvement — but gave Magtein combined with phosphatidylserine and vitamins B6, C and D3. The Alzheimer's Drug Discovery Foundation's review notes it is not possible to determine which component was responsible. This trial cannot support a threonate-specific claim.
**Hausenblas HA, et al. Sleep Medicine: X 2024;8:100121. n=80, 1 g/day threonate (~75 mg elemental) for 21 days, Oura-ring deep sleep p<0.001. Funded by AIDP Inc.** — the ingredient supplier.
The gap, stated plainly
There is no human CSF or brain magnesium data for magnesium L-threonate. None. The entire blood-brain-barrier claim rests on a 7–15% CSF rise in aged rats, measured by the compound's originator in 2010. Every human trial has measured cognition or sleep as a black-box outcome and peripheral magnesium as a surrogate. Not one has measured the thing the mechanism is about.
And because threonate is only 8.3% magnesium, the human trials delivered less elemental magnesium than a supermarket tablet. If those effects are real, either threonate is doing something magnesium-independent — the 2010 paper itself raises that possibility — or they are detecting the correction of a dietary shortfall that any form would have fixed more cheaply.
What magnesium actually does
Migraine prophylaxis — the strongest indication, with guideline status
This is where magnesium earns its place.
- American Academy of Neurology / American Headache Society: Level B — probably effective, should be considered for migraine prevention.
- Canadian Headache Society (Pringsheim et al. 2012): strong recommendation. Specified dose: 600 mg elemental magnesium daily as magnesium citrate. The guideline names citrate explicitly, "because positive results were only obtained with magnesium citrate in the reviewed trials."
Note what that guideline did: it named the form, and the form was citrate — not glycinate, not threonate. Someone buying glycinate for migraine is buying a form no guideline specifies, at two to four times the cost.
Separately, the Canadian Headache Society recommends against intravenous magnesium sulfate for acute migraine. Prophylaxis and acute treatment are different questions with different answers.
Blood pressure — real, small, dose-dependent
**Zhang X, Li Y, Del Gobbo LC, et al. Hypertension 2016;68(2):324–333.** 34 RCTs, n=2,028, median dose 368 mg/day, median 3 months.
- Systolic −2.00 mmHg (95% CI 0.43–3.58; p=0.01)
- Diastolic −1.78 mmHg (95% CI 0.73–2.82; p=0.001)
- Larger in people already on antihypertensives (systolic −5.69 mmHg)
- Institutionally funded, not industry
Two millimetres of mercury is genuine and is not a treatment. Reasonable adjunct; not a substitute for antihypertensive therapy.
Sleep — weak, and the field says so
**Mah J, Pitre T. BMC Complement Med Ther 2021;21:125. Only 3 RCTs, n=151 total. Sleep onset latency −17.36 min (95% CI −27.27 to −7.44; p=0.0006); total sleep time +16.06 min, p=0.15 — not significant. The authors' own verdict: "low to very low quality of evidence"**, all three trials at moderate-to-high risk of bias, and "the quality of literature is substandard for physicians to make well-informed recommendations."
**Schuster J, et al. Nature and Science of Sleep 2025 (PMID 40918053) is the best trial available: n=155, 250 mg/day elemental as bisglycinate, double-blind. Insomnia Severity Index change at week 4: magnesium −3.9 vs placebo −2.3, p = 0.049, Cohen's d = 0.2 — which the authors themselves call "small."** No significant differences on the other psychological measures. The effect was larger in people with lower baseline dietary magnesium.
A p of 0.049 and a d of 0.2 in the best available trial, concentrated in people with low intake. The defensible reading is that magnesium helps sleep a little, mostly by fixing a shortfall. "Magnesium for sleep" as marketed is well ahead of that.
Muscle cramps — mostly null, and this is what most of it is sold for
Garrison SR, Korownyk CS, Kolber MR, et al. Cochrane Database of Systematic Reviews, September 2020 (CD009402.pub3). 11 trials, 735 participants.
- Older adults, cramp frequency: −9.59%, 95% CI −23.14% to +3.97% — not significant
- No RCTs at all for exercise-associated cramps
- Pregnancy: inconsistent, "important limitations to their reliability"
- Minor adverse effects in 11–37% of magnesium recipients, mainly gastrointestinal
- Conclusion: it is "unlikely that magnesium supplementation provides clinically meaningful cramp prophylaxis to older adults"
This is the indication most magnesium in Australia is sold for, and it is the indication the evidence most clearly does not support.
Constipation — the one place oxide wins, and it wins because it's poorly absorbed
**Mori S, Tomita T, Fujimura K, et al. J Neurogastroenterol Motil 2019.** n=33, magnesium oxide 1.5 g/day for 28 days.
- Overall symptom improvement: 70.6% vs 25.0% (p=0.015)
- Spontaneous bowel movements per week: +6.07 vs +2.86 (p=0.002)
The mechanism is the point. Unabsorbed magnesium sits in the lumen and holds water osmotically. Oxide's failure as a nutritional supplement — indistinguishable from placebo on serum magnesium in Walker 2003 — is exactly what makes it work as a laxative. Note the dose: 1.5 g of oxide is ~904 mg of elemental magnesium, far above the supplemental upper limit. It's being used as a drug, not a nutrient.
Depression and anxiety
**Tarleton EK, et al. PLoS ONE 2017;12(6):e0180067 is the trial everyone cites. n=112 analysed, 248 mg/day magnesium chloride, PHQ-9 improvement −6.0 points (95% CI −7.9 to −4.2; p<0.001)**.
A 6-point PHQ-9 improvement is larger than many antidepressant trials report. It is also exactly the effect an open-label, unblinded, placebo-free, self-report crossover is built to produce — which is what this was, by the authors' own stated limitation: "No placebo arm and randomization was not blinded for either the study team or volunteer." Serum magnesium wasn't measured.
For anxiety specifically, the better-designed Schuster 2025 bisglycinate trial (n=155, double-blind) found no significant difference on psychological measures. That's the result to weight.
Exercise performance — the recent data point the wrong way
**Bell DR, et al. Nutrients 2025;17(5):915.** Randomised, double-blind, placebo-controlled crossover, n=15, magnesium chloride 600 mg/day for 9 days.
- VO₂max decreased: 41.3 vs 44.4 mL/kg/min (p=0.005)
- 30-second sprint mean power decreased: 415 vs 439 W (p=0.03)
- 10 km time trial: no difference
- Authors' conclusion: regular exercisers free from hypomagnesaemia "should not supplement their diet with magnesium"
- Funded by a magnesium product manufacturer — which makes a negative finding more credible, not less
n=15 is small and needs replication. But it's the most recent controlled evidence in the area.
Why your blood test won't tell you
Less than 1% of total body magnesium is in blood serum. Fifty to sixty percent is in bone, most of the rest in soft tissue. The NIH Office of Dietary Supplements puts it directly: serum levels "do not accurately reflect total body magnesium levels or concentrations in specific tissues."
The physiology matters here. Serum magnesium is homeostatically defended — the kidney and bone buffer it. You can be meaningfully depleted at tissue level with a serum magnesium squarely mid-range, because the body will strip bone and muscle to hold serum constant. A normal result does not exclude depletion; a low one indicates depletion that is already substantial.
RBC magnesium is conceptually better because it samples an intracellular compartment, but it reflects status at the time the cells were made (red cells live ~120 days), varies with reticulocyte fraction, and has never been shown to predict who responds to supplementation. It is heavily promoted by supplement and functional-testing businesses, which is a warning sign rather than a recommendation.
The magnesium loading test — give a parenteral load, measure 24-hour urinary retention — is a research tool, not routinely available in Australia. The ODS notes experts disagree on its utility and that "no single method is considered satisfactory."
The practical upshot: there is no good test. A clinician cannot reliably tell you whether you're magnesium-depleted. That's why the field runs on intake estimates — and also why the supplement industry can make claims about widespread deficiency that nobody can falsify.
Who in Australia is actually short
Australian RDI (NHMRC Nutrient Reference Values): men 19–30, 400 mg; men 31+, 420 mg; women 19–30, 310 mg; women 31+, 320 mg; pregnancy 350–400 mg depending on age.
Proportion below the Estimated Average Requirement (ABS Australian Health Survey, 2011–12):
| Age | Males | Females |
|---|---|---|
| 4–8 | 0% | 0.3% |
| 9–13 | 11.8% | 24.6% |
| 14–18 | 61.3% | 71.6% |
| 19–30 | 33.5% | 37.2% |
| 31–50 | 36.4% | 32.8% |
| 51–70 | 46.5% | 30.6% |
| 71+ | 63.9% | 48.5% |
| All ages | 37% | 34% |
The "one in three Australians don't get enough magnesium" line in supplement marketing is, unusually, accurate — it traces to these figures. But look at where the problem actually sits: teenagers (61–72% below EAR) and the over-71s (49–64%), not the thirty-something wellness market that buys most of the product.
(The ABS released updated figures in March 2026 reporting that inadequacy has decreased since 2011–12. The specific 2023 percentages sit in data tables we could not retrieve; the direction of change is down.)
Groups genuinely at risk: gastrointestinal disease (Crohn's, coeliac), type 2 diabetes, chronic alcohol dependence, older adults, long-term PPI users — and, on the ABS data, Australian adolescents.
Safety, and the upper limit people get wrong
The Australian Upper Level of Intake for magnesium is 350 mg/day for adults — and it applies to supplemental magnesium only, not total intake.
This is the point most consumer articles miss. Magnesium in food has never been shown to cause adverse effects in healthy people with normal kidneys; the osmotic diarrhoea that defines the limit is a property of a bolus of soluble magnesium salt hitting the gut, not of magnesium in a meal.
So you can eat 500 mg of dietary magnesium and be nowhere near the limit, and take a single 400 mg tablet and exceed it. Several mainstream Australian products deliver a daily dose at or above it.
Diarrhoea threshold by form, roughly in order: oxide and citrate are the most laxative per mg (oxide because it stays in the lumen, citrate because it's highly soluble and osmotically active); bisglycinate, glycerophosphate and malate are better tolerated; chloride sits in between. (No form-stratified dose-response study exists — this ordering is mechanistic and consistent with the Cochrane adverse-event data, not a measured threshold.)
Kidney disease is the one place not to hedge. Magnesium is cleared renally. Hypermagnesaemia has been reported with high-dose magnesium-containing laxatives and antacids including fatal cases, and toxicity runs through hypotension, nausea, muscle weakness, respiratory difficulty, irregular heartbeat and cardiac arrest. If you have chronic kidney disease, don't take magnesium supplements without your doctor involved.
Interactions worth knowing
| Drug | What happens | What to do |
|---|---|---|
| Proton pump inhibitors (omeprazole, esomeprazole, pantoprazole) | PPIs cause hypomagnesaemia with prolonged use, typically beyond a year. Pooled odds ratio 1.775 (95% CI 1.077–2.924), p<0.001 across 9 studies and 115,455 patients (Park CH, et al. PLoS ONE 2014). Caveat: I² = 98% — extreme heterogeneity, so the association is real but the magnitude isn't reliably estimated. | The FDA recommends baseline and periodic serum magnesium on long-term use. This is the under-recognised one. |
| Loop and thiazide diuretics (frusemide, hydrochlorothiazide, indapamide) | Increase urinary magnesium loss | A common cause of depletion — and extremely often prescribed alongside a PPI in older Australians, where the risk compounds |
| Potassium-sparing diuretics (spironolactone, amiloride) | Reduce magnesium excretion | Opposite direction; adds hypermagnesaemia risk |
| Quinolones (ciprofloxacin, norfloxacin) | Magnesium chelates the drug in the gut and blocks absorption | Antibiotic 2 hours before, or 4–6 hours after magnesium |
| Tetracyclines (doxycycline, minocycline) | Same chelation | Same spacing |
| Bisphosphonates (alendronate, risedronate) | Reduced absorption | At least 2 hours apart |
Doxycycline is one of Australia's most-prescribed antibiotics and alendronate one of its most-prescribed osteoporosis drugs. These spacing rules matter in practice and are almost never on the supplement label.
(A digoxin interaction is widely described — hypomagnesaemia increasing myocardial sensitivity to digoxin, and magnesium antacids reducing its absorption. We could not verify it from an authoritative source and have left it out rather than assert it.)
Food beats all of it
Per healthdirect Australia:
| Food | Serve | Magnesium |
|---|---|---|
| Pumpkin seeds (pepitas) | 30 g | 159 mg |
| Chia seeds | 30 g | 114 mg |
| Almonds | 30 g | 80 mg |
| Cashews | 30 g | 75 mg |
| Brown rice, cooked | ½ cup | 72 mg |
| Spinach, boiled | ½ cup | 57 mg |
| Peanuts | 30 g | 48 mg |
| Salmon | 100 g | 30 mg |
| Milk | 1 cup | 28 mg |
| Wholemeal bread | 1 slice | 25 mg |
| Avocado | 75 g | 20 mg |
About 30–40% of dietary magnesium is absorbed — and the same range applies to supplements, which is the quiet answer to the whole form debate. No supplement form has been shown to beat food.
The arithmetic you can act on: 30 g of pepitas (159 mg) plus half a cup of brown rice (72 mg) plus half a cup of boiled spinach (57 mg) is 288 mg — most of a woman's RDI, for about the price of two glycinate tablets, with fibre, potassium and folate attached.
What it costs
Prices from Chemist Warehouse and iHerb AU, 1 October 2026. Chemist Warehouse figures are promotional prices and move.
| Product | Form | Elemental Mg/tab | $ per 100 mg Mg |
|---|---|---|---|
| Bioglan Active Magnesium Citrate (200) | Citrate | 184.8 mg | $0.054 |
| Ethical Nutrients Mega Magnesium Glycinate (120) | Glycinate blend | 200 mg | $0.114 |
| Blackmores Magnesium Glycinate (100) | Glycinate | 150 mg | $0.200 |
| Blackmores Magmin "500 mg" (100) | Aspartate | 37.4 mg | $0.588 |
| NOW Magnesium Citrate (iHerb, 120) | Citrate | 133 mg | $0.149 |
| Doctor's Best Lysinate Glycinate (iHerb, 240) | Glycinate | 100 mg | $0.182 |
| California Gold Bisglycinate (iHerb, 240) | Bisglycinate | 100 mg | $0.189 |
| Life Extension Magnesium L-Threonate (iHerb, 90) | L-threonate | 48 mg | $1.325 |
The glycinate premium is 2.1× to 3.7× per milligram of elemental magnesium against citrate, for a claim the available evidence contradicts. What it does buy is tolerability — roughly 6 to 15 cents a day extra to be able to take a dose you otherwise couldn't. That's a defensible reason to pay it. "Better absorbed" isn't.
The threonate premium is indefensible. At $1.325 per 100 mg of elemental magnesium it costs 24.5× the citrate and 5.5× the most expensive glycinate here. A 90-capsule bottle contains 4.3 g of elemental magnesium in total — about eleven days' RDI — for $57.
The aspartate "500 mg" tablet is the worst value on the list at eleven times the citrate, purely because 92.5% of its headline weight isn't magnesium.
What to actually do
Start with food. Three servings from the table above clear 288 mg and cost less than any supplement here.
If you supplement, buy citrate. Cheapest per milligram of real magnesium, the best human absorption data of any form, and the form a clinical guideline names for migraine prophylaxis.
Move to glycinate only if citrate's laxative effect stops you taking the dose you need — and pay the premium knowingly, for comfort, not absorption.
Buy oxide only if what you want is a laxative, in which case it's both effective and cheap.
Skip threonate until someone measures magnesium in a human brain.
And check the back of the pack. The difference between a good product and a bad one in this category is mostly the difference between 185 mg and 37 mg of actual magnesium per tablet, and both of those tablets say "magnesium" on the front.
Frequently asked
Which form of magnesium is best absorbed? Citrate, on the available human evidence. In the only randomised trial comparing citrate, an amino-acid chelate and oxide head to head (Walker et al. 2003, n=46), citrate produced the highest serum magnesium both acutely and chronically, and oxide was indistinguishable from placebo. There is no human trial showing glycinate is better absorbed than citrate.
Is magnesium glycinate worth the extra money? For tolerability, often yes. For absorption, no — the claim isn't supported. Glycinate costs two to four times more per milligram of elemental magnesium than citrate in Australian pharmacies. If citrate causes loose stools at the dose you need, that premium buys you a usable dose. Otherwise it buys nothing the evidence can demonstrate.
Does magnesium help you sleep? A little, and mostly if your dietary intake is low. The best trial (n=155, double-blind, bisglycinate 250 mg) found an insomnia-score improvement with p = 0.049 and an effect size the authors themselves describe as small. The Cochrane-style review of older-adult insomnia found only three trials totalling 151 people, at low to very low quality.
Does magnesium stop muscle cramps? Probably not. The 2020 Cochrane review of 11 trials and 735 participants found no significant reduction in cramp frequency in older adults, and no randomised trials at all for exercise-associated cramps. It's the most common reason magnesium is sold in Australia and the least supported.
How much magnesium should I take? The Australian upper limit for supplemental magnesium is 350 mg/day for adults — dietary magnesium isn't counted. Most trials showing an effect on blood pressure or migraine used 300–600 mg of elemental magnesium daily, which means migraine-prophylaxis doses sit above the supplemental upper limit and belong under medical supervision.
Can I just get it from food? Yes, for most people. Thirty grams of pumpkin seeds, half a cup of brown rice and half a cup of spinach is 288 mg — most of an adult woman's RDI. Absorption from food is the same 30–40% range as from supplements.
Last reviewed 1 October 2026. Elemental magnesium percentages are calculated from standard atomic weights. Prices are Australian retail at the date of review and change frequently. This page is general information, not medical advice — if you have kidney disease, take a diuretic or a proton pump inhibitor, or are on antibiotics or bisphosphonates, speak to your doctor or pharmacist before supplementing.