Coding and vigilance under alcohol.

Abstract

Social drinkers (10 men, mean age 24) were tested on a task which required them to perform complex coding while remaining vigilant for irregularly occurring light signals placed beyond the subject''s [Ss] peripheral vision. Drug-free trials were administered until CV [coding-vigilance]-performance had stabilized. Then 4 alcohol test sessions were held over an 8-wk period. At each session the Ss received 1.98 ml/kg of 40% alcohol (mixed with an equal amount of a carbonated beverage) and performed the CV task before alcohol administration and during rising, peak (mean 0.064%) and falling blood alcohol concentrations (BAC). Drug-free performance on the CV task remained stable over the 4 alcohol test sessions (average of 168 items coded/trial). Coding under alcohol, though initially impaired, improved over the 4 alcohol sessions so that the average number of items correctly coded increased from 159 during session 1-180 during session 4. During all 4 sessions, coding efficiency diminished as BAC rose and abruptly improved when BAC began to fall. Vigilance scores remained stable over all 4 alcohol test sessions and were unaffected by rising and falling BAC. The improvement in coding under alcohol over the 4 test sessions could be the result of drug-state dependent learning or of increased effort to compensate for the expected impairment of performance under alcohol.