In their placebo-controlled, three-period, crossover trial of quinine vs vitamin E for nocturnal leg cramps, Connolly et al1 provide interesting data on the efficacy of quinine.1 However, we believe that, because the study design fails to adequately negate the likelihood of "period," "carryover," or "sequence" effects,2 one cannot exclude the possibility that bias from these effects is either solely or partially responsible for the therapeutic outcome reported by the authors.
To evaluate the possibility of a period effect, the investigators had to first determine the level of "disease activity," ie, the magnitude and frequency of leg cramps. To accomplish this, study subjects recorded their symptoms during a 4-week "run-in" period prior to the initiation of treatment. The level of activity recorded was considered to be representative of the characteristics of each patient's leg cramps. The difficulty with relying solely on the run-in period to determine disease activity