Com- pounding this problem, the patterns of steroid use by regular users, includ- ing doses and frequency of administration, are highly variable, which makes properly-controlled studies in an experimental state of affairs difficult at greatest and, per- h