Circadian variation in human performance evaluated by the Walter Reed performance assessment battery

As part of a two clock-time (0830 versus 2030) evaluation of administration-time dependent effects of dexedrine (5 mg) and triazolam (0.25 mg) on human cognitive performance, placebo (control) studies were conducted on 12 diurnally active (0700-2300) male adults (23-38 yrs) using a double-blind, randomized crossover design. Testing was conducted hourly during a series of sleepless 13-hr spans commencing in the morning or evening, using the Walter Reed computer controlled and scored multi-task cognitive performance assessment battery. For the placebo condition, Single and Group Cosinor analyses documented circadian rhythms in performance for most tasks (reaction time, logical reasoning, serial add/subtract and spatial orientation) both for individuals and the group. Overall, performance was worse overnight, when sleepiness was greatest, and best between 1830 and 2030. It was most variable around 0600-0700. The day-night variation in performance over all cognitive tests amounted to 21% of the 24-hr mean.