In the "no f'ing kidding" department

[drwex]

This is me ranting about modern snake oil. Skip if easily offended.

Story in the NY Times today: Study Sees Little Benefit in Chondroitin for Arthritis. Detail: a review of 20 earlier studies had found that the benefit of the supplement was “minimal or nonexistent.”

Chondroitin is one of the popular remedies found in "dietary supplements" and favored by an alarming number of my acquaintances as miracle cures. Right up there with zinc, emergen-C, garlic and other such foolishness. Yes, healthful supplements, yum yum. Miracle cures for joint pain, making colds vanish? No.

Repeat after me: "There is no biologically plausible way that [chondroitin] can work to repair joints damaged by arthritis." OK, that quote is from Dr. Eric Matteson, chair of the division of rheumatology at the Mayo Clinic. Repeat after him, if it works for you.

Promotion of this product is twenty-first-century snake oil. That is, it's an item with plausible value in some situations that gets touted as the miracle cure-all, hey can't-hurt-to-take-it, my-friend-Joe-swears-by-it, whatever. And people shell out money time after time. Willing dupes at best.

Let me try to be plain: if you have a cold, pounding large doses of Vit C will not make the cold go away any faster, even if it is wrapped up in "32 different mineral complexes." It'll probably just make you pee funny. If you have injuries or arthritis or other joint pains, eating sugar and shark cartilage will not cure them. If you believe otherwise, demonstrate a biologically plausible way for it to happen.

It's unclear to my why so many people who are otherwise intelligent and rational, with fully functional critical thinking facilities, can fall prey to this kind of scamming. Is it because we desperately WANT there to be miracle cures? Sure, being sick sucks and arthritis hurts like hell. But that experience doesn't make me blind to my knowledge of biology, chemistry, and medicine. It's beyond me why people believe in this stuff.

It doesn't help me feel better at all when the Natural Products Association (there's an unbiased organization, yeah) responds that "Clearly, research has demonstrated..." Uh, no. No, it hasn't. That's the POINT, you charlatans.

For the record, I'm not speaking about herbal medicine. If you really want snake oil, take the actual stuff for your arthritis. The original snake oil (which I'm told you can still get in Chinese medicine shops) really does help join pain because it's an anti-inflammatory. Many years ago, I worked with a guy who took it regularly for bursitis. Said person wasn't miraculously cured of his condition but he did get relief with fewer side effects than western-style NSAIDS gave him.

There. I feel better now.

Comments

[drwex]

Sorry, but "YMMV" is precisely the kind of thinking I get irked by. In fact, my mileage doesn't change the biological bases of human metabolism. The precise point is that this is not something subject to whim; it's science. Which includes placebo effects and suchlike. But it's certainly not a matter of personal opinion whether or not something has statistical validity.

[imvfd.livejournal.com]

Well... I've got a stats professor that might argue with you on the "it's certainly not a matter of personal opinion whether or not something has statistical validity"

[drwex]

*sporfle*

Too true.

[sebastian-tombs.livejournal.com]

See, here is where you are completely wrong.

When I take a beaker of one chemical, say hydrochloric, and mix it with another chemical, say sodium hydroxide, junior high school chemistry says:

HCL + NaOH -> HoH + NaCL

But that is not what happens. Real science tells you that you will have varying combinations of both sides of the equation (admittedly, more of the latter than the former). And then you also have the added complexity of the ionization of the salt in the water...

So, science. Right. So, is a low carb diet better for you than a low fat diet, or vice versa? I can throw at you a goodly number of studies that will say low carb, and a goodly number of studies that say low fat. First off, we apparently don't understand enough science to actually know which is better (in theory). Secondly, the statistics actually say "it depends". So, what's with the studies that give us categorical answers? Excellent question. The you get newspaper articles that quote these studies, saying "studies show" when, in fact, they should say "studies disagree, but the better sounding story is, 'studies show', and that is how I can sell more column space / newspapers / advertising (hmmm...perhaps there is a more direct link to the advertising sales....).

Statistical validity means you have to include all the tests, and it means that all the tests had to be valid in the first place. This is why the "science" of Cold Fusion (as reported in the media) lost the right to the word when no one could reproduce the results.

Another thing...would you say that aspirin is a pain-killer? I assume, yes. So, would you say that it works on *everyone*? I mean, science would seem to indicate that, wouldn't it? But it doesn't have any affect on me. Whereas wren13 loves ibuprofin and the like, but gets no help from other painkillers, such as vicodin or novocaine. Basic science says they should help, there is (as far as I know) no understanding of how tolerances like that work, esp. after only very limited exposure, and the reality is that they do nothing for her.

"But it's certainly not a matter of personal opinion whether or not something has statistical validity." Then we could discuss the use of statistics, a conversation for which my training leaves me poorly equipped. However, I do know that there are a great number of ways to play with your statistics before you release the study. I also will add that there are studies (I just lent the book with the references, so I can't provide them, but check the reference appendix of Michael Crichton's State of Fear) that show that studies are statistically likely to prove the point desired by the studiers (or, more to the point, those funding the study).

As someone who has taken a fair amount of chemistry, I will go back to the simple point that statistics and percentages is a huge deal in science - in fact, it is exactly why chemical engineers exist in the first place. "YMMV" is really a statistically relevant comment.

[drwex]

Your discussion misses the point, which is that "your mileage may vary" is not a valid scientific method. If you're trying to make a different point then I'm not getting it.

[sebastian-tombs.livejournal.com]

Actually, you're missing the point, which is that in the case of very complex systems, YMMV actually is a valid scientific result.