The study of elite athletes, because of their rarity, renders statistical power difficult to achieve, control groups difficult or impossible to find, and generalizability difficult to demonstrate. Alternative methods of analysis may be better suited to such study.