Abstract
Chronobiological rhythms gain increasing relevance in experimental biology and clinical medicine. To detect underlying rhythms and to illustrate the result are important issues of rhythmometry. A new statistical approach based on bioequivalence testing (crossover analysis) in clinical pharmacology is proposed for single time series. With regard to the mean of a reference time (μr) and means of values (μ1, μ2, etc.) of subsequent times (t1, t2, etc.) ratios and 90% confidence limits are calculated. Log-transformed data are used in case of log-normal distribution. A chronobiological rhythm is supposed, if the 90% confidence limits lie outside the acceptance limits that are generally supposed to be ± 20%.