Ultradian rhythms in cognitive performance: no evidence for a 1.5-h rhythm.

In the past, Kleitman's notion of a basic rest-activity cycle (BRAC), which states an approximate 1.5-h rhythm of arousal, has received some empirical support in studies investigating physiological or psychological parameters. Our study investigated this ultradian rhythm in human cognitive performance. Sixty subjects were tested every 10 min for 9 h with an elementary cognitive task, the sentence-verification test (SVT), and provided ratings of alertness and mood. Heart period was always measured during performance of the SVT. Using rather conservative statistical methods, spectral analyses of time series revealed no significant 90-min periodicity for any of the variables. Instead, longer periodicities were the major sources of variance. Reanalyses using detrended and/or smoothed time series suggest that emergence of the BRAC might depend also on statistical methods. In conformity with recent studies it was concluded that several positive reports for the BRAC might be due to lack of conservatism in statistical methods.