Map of life expectancy at birth from Global Education Project.

Wednesday, September 28, 2011

Yet one more disappointment

You may have seen news coverage of this study which finds that saw palmetto extract is utterly useless for benign prostate enlargement. (You only get to read the abstract, but I read the whole thing. Na na na na na.)

I can tell you that this was a very high quality trial, it's got all the right randomization, blinding and assessment procedures, and the results were as negative as you can possibly get. Actually placebo turned out to appear very slightly more effective, but that's probably just chance. If you believe this, saw palmetto absolutely does not work.

What's interesting about this is that saw palmetto for benign prostate enlargement was one of the examples of a traditional herbal remedy that was believed to work, based on previous trials. (Obviously, willow bark works for pain, it's aspirin; poppy resin contains opium, which does what opium does; etc. Many drugs are derived from plants, nothing unusual about that.) However, more recent trials and meta-analyses had called that conclusion into question. One reason they did this was to clear up the confusion, and also to try a higher dose than has been used in the past to make sure they had maximum chance of finding a benefit if the stuff really does work.

So why the fairly steady trajectory from initial positive findings to this disappointing result? Actually, that's pretty common. Early trials of a compound are typically not of the highest quality, in part because there isn't a lot of money for them. People just want to get a first look, and nobody's going to put big bucks into something highly speculative. Lower quality trials are more likely to get positive results because of inadequate blinding and other design flaws, and furthermore their sponsors are usually engaged in selling the stuff. On top of that, trials with positive results are more likely to be published.

So, even with interventions that do ultimately turn out to be effective, it's common for the effect size to diminish as they are better studied. In this case, it went away completely. The universe is against us.

Fortunately, saw palmetto appears to be harmless, so it's just that men have wasted some money. My advice in general however -- don't be eager to take pills that have only been on the market (in the case of pharmaceuticals requiring FDA approval) a short time; or in the case of unregulated supplements, that have only been studied a little bit. Chances are those early encouraging results aren't going to pan out. By the way a good reason to doubt is lack of a known biological mechanism of action. Bayes theorem tells us that with a low prior probability, the positive findings aren't nearly as convincing.

Of course, this won't stop the "supplement" industry from marketing the stuff. Their claims don't have to be true.

No comments: