Methodological Issues in Studies of Rhythms in Human Performance

In this review various sources of measurement error are considered in the context of investigating rhythms in human performance. The reproducibility of performance in any exercise task is an important factor if rhythmic variations are to be detectable. Available physical performance tests range from simple efforts lasting only a couple of seconds to sustained endurance exercise. When measuring muscle strength, the options include static or dynamic actions, slow or fast movements, voluntary or electrically stimulated contractions and maximal force or maximal power output. For studies of circadian rhythms the researcher has to choose between using nychthemeral or controlled conditions, and the number of times a day to be used for observations, and to decide how to control for loss of sleep, diet and prior activity. The need to recover energy between tests has led research groups to employ diurnal rhythm models in preference to cosinor analysis of circadian designs. The use of male subjects has also been favoured due to difficulties of controlling for menstrual cycle phase. Nevertheless recent attention has been given to research models for investigating the effects of rhythmic variations in female steroid hormones and on interactions between circadian and circamensal rhythms. There are real challenges in exploring seasonal variations in human performance and in examining how the body adjusts after desynchronisation as occurs during nocturnal shift-work and travelling across multiple time zones. The methods adopted must accommodate flexibility between laboratory-based and field-studies depending on the context of the research questions being pursued.