A summary of published research, not medical advice. See the Disclaimer.
Studies: every indexed trial and review on this topic, scored
Magnesium is a mineral the body uses in hundreds of enzyme reactions, including those for muscle contraction, nerve signalling and energy metabolism. Nuts, seeds, legumes and whole grains supply most dietary intake.
Best-supported effect: Lower blood pressure in adults. Probably helps a little (moderate evidence).
Most-cited weak claim: Better sleep. Unknown (very weak evidence).
How we grade evidence · Last reviewed: October 2026
What you need to know
In adults with and without high blood pressure, magnesium lowered blood pressure a little, with moderate evidence. Claims of better sleep and less anxiety rest on few, small or poor-quality trials, so the evidence is very weak. In older adults with leg cramps it made little or no difference (moderate evidence). Diarrhoea and other minor gut effects were reported, with weak evidence. Case reports describe dangerously high blood magnesium, including deaths, several after magnesium laxatives and some in older people with kidney dysfunction. The blood pressure trials used a median of about 368 mg a day for a median of 3 months, and the US upper level for magnesium from supplements in adults is 350 mg a day.
| Effect | Verdict | What was measured | Studied in |
|---|---|---|---|
| Lower blood pressure | Probably helps a little (moderate evidence) Pooled analysis of 34 trials in adults, two later trials found no change |
Measured | Adults, with and without hypertension |
A meta-analysis of 34 double-blind, placebo-controlled RCTs (2,028 participants) found that magnesium lowered systolic pressure by about 2 mmHg and diastolic pressure by about 1.8 mmHg over a median of 3 months [1]. A pooled analysis of 22 trials (1,173 participants) found reductions of roughly 3–4 mmHg systolic and 2–3 mmHg diastolic, with larger effects at higher doses [2]. An earlier pooling of 20 trials found a smaller overall effect that grew with dose [3]. Two later 24-week double-blind trials in overweight adults found no change in blood pressure: 164 people at 450 mg/day [10] and 52 people at 350 mg/day, on office and 24-hour readings [11]. Blood pressure is the outcome the claim is about, so the lab-marker rule does not apply. Probably helps a little (moderate evidence): downgraded once because trials disagree, with some pooled and single-trial results showing no effect; the largest pooled analysis still found a small reduction whose range excluded no effect [1].
| Effect | Verdict | What was measured | Studied in |
|---|---|---|---|
| Fewer muscle cramps, older adults | Probably makes little or no difference (moderate evidence) Cochrane review of 11 trials, meaningful cramp reduction unlikely in older adults |
Self-reported | Older adults with idiopathic leg cramps |
More on this result under Claimed but unsupported or contested.
| Claim | Verdict | What was measured | Studied in |
|---|---|---|---|
| Better sleep | Unknown (very weak evidence) Few small trials in older adults, review flagged low quality |
Measured | Older adults with insomnia |
| Less anxiety | Unknown (very weak evidence) Review of 18 studies, poor quality, people rated it themselves |
Self-reported | Adults with mild anxiety, premenstrual symptoms or other anxiety vulnerability |
| Lower blood glucose | May help a little (weak evidence) Pooled trials where no one knew who got what varied widely, fasting glucose effect possibly near zero |
Lab marker | Adults with or at high risk of type 2 diabetes |
| Fewer muscle cramps, pregnancy | Unclear: trials disagree (weak evidence) Pregnancy trials conflict and carry high risk of bias |
Self-reported | Pregnant women with leg cramps |
Sleep: a meta-analysis of 3 RCTs (151 older adults) found that magnesium shortened the time to fall asleep by about 17 minutes compared with placebo. The increase in total sleep time was not statistically significant, and the authors rated the evidence low to very low quality [4]. A single trial gave 46 older adults 500 mg/day for 8 weeks and reported better insomnia severity scores and sleep efficiency [5]. Unknown (very weak evidence): the trials are few and small, and the review rated its own evidence low to very low. Younger, healthy people were not studied in these trials.
Anxiety: a systematic review of 18 studies found results leaning toward a benefit for subjective anxiety in anxiety-prone groups and rated the quality of the evidence as poor [6]. Unknown (very weak evidence): the evidence is poor and the outcomes are self-reported.
Blood glucose: a meta-analysis of 18 double-blind RCTs found lower fasting glucose in people with diabetes and better glucose tolerance in people at high risk, with only a non-significant trend toward better insulin resistance scores [7]. A 2021 update reached similar conclusions and also reported improved insulin sensitivity markers [8]. The claim is about glucose itself, so the lab-marker rule does not apply. May help a little (weak evidence): results varied widely between trials and the fasting-glucose effect could be close to zero. No trial here measured diabetes onset or complications.
Muscle cramps: a Cochrane review of 11 trials found it unlikely that magnesium produces a clinically meaningful reduction in cramps in older adults, rated as moderate-certainty evidence [9]. For cramps in pregnancy, the trials in the same review disagreed with each other and carried a high risk of bias. Unclear: trials disagree (weak evidence) [9]. The claim rests mainly on magnesium's role in muscle contraction. None of the reviews cited here measured long-term outcomes such as cardiovascular events, diabetes incidence or clinically diagnosed anxiety disorders.
Shown in humans
Shown only in animals or cells
Theory only
| Side effect | Verdict | How often vs placebo | Source | Studied in |
|---|---|---|---|---|
| Diarrhoea and other minor gut effects | May happen (weak evidence) 7 of 10 studies found no difference, some people dropped out over diarrhoea |
Not reported; 7 of 10 studies (128–1,200 mg/day) found no significant difference in diarrhoea, and 3 of 13 trials in one meta-analysis reported withdrawals for diarrhoea | [9][17] | Adults |
| Dangerously high blood magnesium (hypermagnesaemia), including deaths | Unknown (very weak evidence) Reports of single patients plus one group of hospital patients, no comparison group |
Case reports only; 23% of 320 hospital patients prescribed magnesium oxide in one cohort, with no comparison group | [18][19][20][21][22] | Hospital patients |
| Serious adverse events | May not increase it (weak evidence) Two trials of 52 and 164 people, one unrelated serious event |
No adverse events in a 52-person trial; one serious event, judged unrelated, in a 164-person trial | [10][11] | 164 people |
Minor adverse events could be assessed in only some of the cramp trials [9]. The US upper level of 350 mg/day from supplements was set with diarrhoea as the limiting effect; a 2023 review argues newer trial data support raising it [17]. Most hypermagnesaemia cases involved magnesium laxatives or high-dose products, not standard supplement doses [18][19][20][21].
This lists what studies and labels report. A pharmacist or prescriber can say whether it applies to a specific person.
This lists what studies and labels report. A pharmacist or prescriber can say whether it applies to a specific person.
| Combined with | What happened | Verdict | Source | Studied in |
|---|---|---|---|---|
| Ciprofloxacin and other fluoroquinolone antibiotics | A magnesium-aluminium antacid given 5–10 minutes before ciprofloxacin cut its absorption to 15% of normal, 23% at 2 hours and 70% at 4 hours. Given 6 hours before or 2 hours after, it had no effect. Reviews report reduced absorption of all quinolones with magnesium antacids, through chelation | May happen (weak evidence) Studies where healthy volunteers tried each option, magnesium-aluminium antacid not magnesium alone |
[23][24] | Healthy volunteers |
| Tetracycline antibiotics | Antacids cut tetracycline absorption by more than 90% through chelation. Data come from antacids with mixed cations, not magnesium alone | May happen (weak evidence) Absorption data from antacids with mixed minerals, not magnesium alone |
[25][28] | |
| Gabapentin | 1 g magnesium oxide lowered gabapentin exposure (AUC) by 43% and bioavailability by 32% in 13 healthy adults | May happen (weak evidence) One study in 13 healthy adults where each tried both, 1 g magnesium oxide |
[26] | 13 healthy adults |
| Levothyroxine | Magnesium aspartate lowered thyroxine exposure (AUC) by 12% in 15 healthy adults; magnesium citrate lowered it by 7%, not significant | May happen (weak evidence) One study in 15 healthy adults who tried each option, effect differed by form |
[27] | 15 healthy adults |
| Bisphosphonates | Most bisphosphonate prescribing information cautions about antacids with polyvalent cations; no quantitative data exist | Unknown (very weak evidence) Prescribing information cautions only, no measured data |
[28] |
These are the amounts used in the studies. They are not recommendations.
The blood pressure trials used a median of about 368 mg/day of magnesium for a median of 3 months [1]. Effects were larger at doses above roughly 370 mg/day [2]. A sleep trial in older adults used 500 mg/day of magnesium for 8 weeks [5]. The cited abstracts for these outcomes do not report which forms were used. One 24-week trial compared citrate, oxide and sulfate at 450 mg/day: plasma magnesium rose similarly with all three, urinary magnesium rose more with citrate, and none changed blood pressure [10].