Is it just me? Are the most popular remedies the least likely to have been studied enough? The things that normal people use to self-treat their pain are rarely tested in good clinical trials. The last good example of this I wrote about was just a couple weeks ago: “spice therapy” (Tiger Balm and such).
Today it’s magnesium supplementation for chronic pain. Can it actually help patients? Is it basically a pain-killer? Shouldn’t we know this for sure by now?!
Of course we should! But of course we don’t.
In this post, I’ll look at some of the scraps of science we do have.
But why? The rationale for magnesium supplementation
It’s a popular idea that magnesium deficiency is a significant factor in many cases of chronic pain. This is one of the main reasons people literally bathe in the stuff (Epsom salts are magnesium sulfate). The magnesium-pain hypothesis has some seeds of truth. For instance, deficiency is somewhat common (even though it’s easy to eat). And it’s clearly a staple molecule in neurology. I discuss this in more detail in my supplements review, and the Epsom salts article really digs into it.
It’s easy to eat magnesium. You just have to be willing to eat salad. Or potatoes!
Testing magnesium with a keyboard and mouse
A 2020 study seemed to go out of its way to try to answer this question without actually talking to a patient with pain — pure data analysis. Tarleton et al mined some old survey data on the health of more than 13,000 American adults (high-volume data, but also poor quality for this purpose, “low resolution”).
They reported a “protective effect” of “increased magnesium intake,” concluding that this “warrants further study.” Which everyone more or less already agreed on, but more data is good… right? Hm, not if it just muddies the waters. Data like this is can be like yelling in a crowded theatre, “There might be a fire! I think I saw some smoke!”
This is a far cry from conclusive evidence that taking magnesium treats any kind of chronic pain. The allegedly protective “increased magnesium intake” they detected is, I believe, inferred from notoriously unreliable self-reporting of diet. The possibilities for confounding factors are endless.
Bonus eye roll: Amusingly, when Tarleton et al explained why they did this research — that “there is some evidence that magnesium supplementation results in improved pain management” — they cite only a single supporting paper. And what is that single paper? A twenty-year-old study showing an inflammatory response to severe deficiency… in rats. 🙄
A Canadian review of not much magnesium science
Banerjee et al concluded in 2016 that “magnesium appears to have an analgesic effect.” Great!
But don’t read the fine print if you want to hang on to that good feeling. Their own summary of findings doesn’t seem to back up the optimism. They clearly state that conclusions were “not possible” for migraine … that the evidence is conflicting for a rather exotic kind of chronic pain (complex regional pain syndrome) … and they mention only a single trial showing a benefit for back pain (with intravenous supplementation, no less). More on that one below.
And that’s just all they wrote about the data. If there is enough evidence to justify “magnesium appears to have an analgesic effect,” it is not reported in this paper.
One of the few good trials of magnesium for pain
Yousef et al did a true trial of magnesium supplementation for back pain — but, crucially, only back pain with a neuropathic component. 40 patients were given a placebo, and their progress over 6 months was compared to 40 more who got intravenous magnesium for two weeks, then oral for another month.
The patients who got magnesium clearly did better in the long run.
Both groups did great at first, more than halving their pain. If stopped there, the study would have shown that magnesium was no better than a placebo. But then the placebo group’s numbers rebounded, while the magnesium folks stayed low… for six months. The improvement wasn’t huge, but it wasn’t small either. I wouldn’t turn it down.
This is a clearly positive result on its face — which is such a rarity in this field that it’s cause for too-good-to-be-true concern. It is “just one study,” with unknown flaws, and unreplicated. Most important, this experiment just cannot tell us anything about the effect of magnesium on the most common kinds of pain, which are mostly not neuropathic. Back pain without neuropathy is a good proxy for many other kinds of pain; back pain with neuropathy is more about neuropathy than “back pain.”
But, for whatever it’s worth, it is indeed a properly positive result — and that’s more than we can say for a great many other trials of any intervention for any kind of back pain. It is noteworthy. If I had neuropathy, I’d probably start eating more salad. Especially potato salad! See dietary sources of magnesium.