Each claim gets one of four grades based on human trials. Strong means several good placebo-controlled trials or a solid meta-analysis agree. Moderate means more than one placebo-controlled trial points the same way but the trials are small or mixed. Weak means few, tiny or uncontrolled studies. None means no supportive placebo-controlled trial.
On this page
What do the four grades mean?
| Grade | What it means | What would move it up |
|---|---|---|
| Strong | Several good-quality placebo-controlled trials, or a solid meta-analysis of them, agree | It is already the top grade |
| Moderate | More than one placebo-controlled trial points the same way, but they are small, short or uneven in quality | A large, well-run independent trial confirming the result |
| Weak | Studies are few, very small, mixed, or have no proper placebo group | Consistent placebo-controlled results |
| None | No placebo-controlled human trial supports it, or good trials found no effect | A well-run human trial showing a benefit |
The bars beside each grade fill with it: three for strong, two for moderate, one for weak and none for no good evidence.
What do we check in each study?
- A placebo and blinding. Sleep, mood, energy and pain all respond strongly to expectation, so trials without a placebo group count for much less.
- Size and length. A handful of people for a few weeks can show a big effect by chance.
- The main outcome. A benefit on the result the trial was designed to measure counts for more than one found among many secondary measures.
- Who took part. A result in people who were low in magnesium, or in older adults, may not apply to a healthy 30-year-old.
- The form and dose. We record how much elemental magnesium people took, not just the compound weight.
- Who paid. We note when the trial was funded or run by the product’s maker.
Why does magnesium need extra care?
Two reasons. First, many studies show that people with a condition have lower magnesium levels. That is interesting but does not prove that taking magnesium helps, because illness, diet and medicines can all lower magnesium. We grade on trials of supplements, not on these associations.
Second, people who start with low magnesium are the most likely to benefit, and most trials do not measure it. So a weak or absent effect in a trial of well-fed volunteers does not rule out a benefit for someone who is genuinely short. We say so where it matters.
First published 4 November 2026. Spotted an error? Tell us and we'll correct it. Read our editorial policy.