Studies of larger populations yield more typical results. Consequently, studies of smaller populations yield more extreme results.

That's not to say that these results might not be significant -- what you propose may be the case -- but I'd want to see an actual mechanism of action before buying something like this.