Doing Science (2) Sports supplement index.

In my last post I said I have science to do. That may have been a little misleading, because this is about to be another example of why some science simply hasn’t been done. I’ve been into my fitness for a while now, including on-and-off use of protein supplements (and creatine, but never for long). Nearly all of the premium (i.e. expensive) supplements claim to be scientifically demonstrate to improve performance or muscle size. However, none of them claim to be better than others. One of them must be best, and the best one must be interested in proving it. And there are ways to do that.

The necessary step is to have a standard experiment that you run with a numerical result. Then those numbers are comparable so comparisons can be made; there would be a comparative index. Then, as a customer I can decide between cost and efficiency (as measured by this index).

In theory, the ideal way to measure the this would be to have a reasonable number of people with a controlled diet and exercise over 6 weeks. Confine 300 people to a complex and give them a strict regime of food and exercise. 150 of those people would would take the supplement on top and the other 150 do not. Then, all 300 people are measured for their metric (strength gains, athletic performance or something). Then, the supplement-takers can be said to have made gains in percentage terms, relative to the control group. In an experiment like this, 300 people will probably be enough because so many variables are being carefully controlled. But, for ethical and practical reasons you probably can’t do this experiment or make these controls.

You can’t really confine people to a complex for 6 weeks. And because of that, you can’t really control people’s exercise regime or diet. If you can’t control those things, your sample size needs to do something very impressive: it has to be big enough to statistically balance variation in diet, exercise, sleep and other variables. Now we are talking about needing thousands of people in each group attending pre- and post-trial athletic performances. Not just once, either. This experiment would have to happen for every product for every brand before it has any meaning.

And, because this experiment has never been done, before anyone invests in this study they are going to question whether they really come off well… But this is a thing that is essentially knowable. Some supplements are better than others to the population at large. (This experiment doesn’t look at whether different supplements are better for different individuals.) As this investigation is unlikely ever to be done, it does raise an interesting philosophical question (in my next post).