Alcohol effects on human risk taking

Rationale. Despite a well-established relationship between alcohol and risky behavior in the natural environment, laboratory investigations have not reliably shown acute alcohol effects on human risk-taking. Objectives. The present study was designed to demonstrate a dose-response relationship between acute alcohol administration and human risk taking. Further, this investigation sought to delineate behavioral mechanisms that may be involved in alcohol-induced changes in the probability of risky behavior. Methods. Using a laboratory measure of risk taking designed to address acute drug effects, 16 adults were administered placebo, 0.2, 0.4, and 0.8 g/kg alcohol in a within-subject repeated measures experimental design. The risk-taking task presented subjects with a choice between two response options operationally defined as risky and non-risky. Data analyses examined: breath alcohol level (BAL), subjective effects, response rates, distribution of choices between the risky and non-risky option, and trial-by-trial probabilities of making losing and winning risky responses. Results. The alcohol administration produced the expected changes in BAL, subjective effects, and response rate. Alcohol dose-dependently increased selection of the risky response option, and at the 0.8 g/kg dose, increased the probability of making consecutive losing risky responses following a gain on the risky response option. Conclusions. Acute alcohol administration can produce measurable changes in human risk-taking under laboratory conditions. Shifts in trial-by-trial response probabilities suggest insensitivity to past rewards and more recent losses when intoxicated, an outcome consistent with previous studies. This shift in sensitivity to consequences is a possible mechanism in alcohol-induced changes in risk taking.