A parallelism between human body temperature and performance independent of the endogenous circadian pacemaker.

A battery of performance tests involving manual dexterity, serial search, and verbal reasoning was given about seven times per day to 2 healthy young male subjects (22 and 25 years of age) involved in separate forced desynchrony studies, each involving several months of temporal isolation. In these studies, the period lengths (denoted T) of the imposed day lengths (sleep/wake and light/dark cycles) were 25.8 and 26.0 h for the 2 subjects. For each subject, the endogenous circadian pacemaker (ECP) failed to entrain to a period of T and instead free ran at a period length denoted tau (24.2 and 24.5 h). By educing performance rhythms (and rectal temperature rhythms) separately at tau and at T (after three complete beating cycles for the first subject and two complete beating cycles for the second subject), the hypothesis could be tested as to whether performance and temperature were parallel, both when educed at tau (indicating ECP influence) and when educed at T (indicating sleep/wake cycle influences). The hypothesis was consistently confirmed at tau and mostly confirmed at T. For most variables, when educed at T, both performance speed and body temperature showed an inverted V-shaped function, with a peak about 9 to 12 h after waking.