Effect of Caffeine and Alcohol on Performance under Stress of Audiofeedback

Abstract

The effect of caffeine, alcohol or both on performance of 8 human subjects (4 females) under stress of audiofeedback was evaluated in a double-blind study. Each subject was observed under 4 conditions: caffeine plus placebo beverage, alcoholic beverage plus placebo caffeine, alcoholic beverage plus caffeine, and placebo beverage plus placebo caffeine. The alcoholic beverage was 100-proof bonded bourbon whisky in quantities calculated to contain 45 ml of alcohol/150 lb. of body weight. The caffeine was offered as 500-mg capsules of caffeine citrate. The subjects were graded on performance on 9 tests, 2 of verbalization, 6 involving mathematics and 1 of color discrimination; an output/error score was calculated for each test as a measure of performance. An analysis of variance and orthogonal contrasts among the four treatments was performed. In all tests, there was a decrease in performance when the subjects'' blood alcohol concentration had reached approximately 50 mg/100 ml. The reduction in performance by alcohol was significant (at the 5% level by Tukey''s criterion) in verbal output, reverse reading and progressive counting. Analysis by orthogonal contrasts showed a significant average effect of impairment on the 2 tests of verbalization and on 3 mathematical tests. There was a trend toward impairment in performance after caffeine but no significant reduction was observed. Improvement of performance after caffeine was suggested on the color discrimination test. No complete picture of antagonism of ethanol-induced impairment of performance by caffeine was demonstrated.