There’s been a rash of studies lately that purport to poke holes in popular notions about using supplements to treat various health conditions.

Here’s the latest: an AP report, headlined “Supplements do little for arthritis, study finds” in the Globe & Mail, which says that glucosamine and chondroitin sulfate don’t seem to have much effect on arthritis symptoms.

The article reports the findings of a study by the National Institutes of Health that was published in today’s New England Journal of Medicine. But when you dig into it a bit, some interesting things come to light. One is that well over half of the subjects did get better–including those who were given a placebo:

Sixty per cent who took the placebos had reduced pain compared with 64 per cent who took glucosamine, 65 per cent who took chondroitin and 67 per cent who took the combo pills.

And get this — of the people who took the prescription drug Celebrex? 70 percent reported less pain. Note that the article sums up the spread between the placebo and supplement results as “These differences were so small that they could have occurred by chance alone.” Presumably we could say the same thing about Celebrex vs. glucosamine and chondroitin, then, right?

(No word on whether the subjects taking glucosamine and chondroitin encountered any Celebrex-style side effects.)

The subtext of the argument that traditional medicine likes to mount against alternatives is that alternatives are witchdoctory. Closer to magic than science. But you know what? Traditional medicine is completely dependent on faith, excuse me, “magic,” as well. Publish a few studies that say Celebrex doesn’t work, and that 70 percent figure would start dipping closer to 60 percent.

Me? I just keep my eye out for that perfect placebo. That would be: something that goes well with a nice red wine.