EVIDENCE LIBRARY
Can Alcohol Reduce Lottery Risk-Taking While Impairing Working Memory?
A small crossover study found more conservative reward choices alongside poorer demanding working-memory performance. Choosing a sure reward did not establish safer driving or preserved executive function.
Yes. A small alcohol experiment found more conservative choices in a monetary task alongside worse performance on demanding working-memory tests. Choosing a certain reward did not establish preserved cognition or safer driving. “Risk-taking” depends on the task and should not be treated as a single universal alcohol effect.
The comparison involved three dose conditions
Liu and colleagues tested 20 adults, 15 men and five women, in a randomized crossover design. Participants completed higher-dose, lower-dose and placebo sessions one month apart.
The dose calculations used grams of ethanol per liter of estimated body water, with body water calculated as 60% of body weight. The paper’s labels of 1.5 and 1.0 grams per liter describe dose calculations, not directly measured blood-alcohol concentrations.
Breath testing monitored exposure. The printed breath units do not provide a reliable basis for converting the reported numbers into forensic blood values. The results should therefore be discussed by assigned condition rather than used to create a personal BAC threshold.
Conservative choices coexisted with poorer demanding performance
In the reward task, participants chose between a guaranteed monetary amount and lotteries. Some lotteries disclosed the probability of winning; others obscured it. The higher-dose condition reduced the authors’ modeled measures of tolerance for risk and ambiguity relative to placebo. The lower-dose condition did not significantly change those measures.
At the same time, higher-dose alcohol reduced accuracy and sensitivity on the two-back and three-back tasks. These required participants to compare a square’s location with its location two or three presentations earlier. One-back accuracy did not show a significant alcohol effect, and reaction time did not significantly change across the tested working-memory levels.
The lower-dose condition did not show significant working-memory decrements in those comparisons. This is a task-specific null result, not evidence that every relevant ability was intact.
Brain inhibition was an indirect measure
The researchers also used magnetic stimulation over motor cortex and measured the resulting cortical silent period in a hand muscle. Alcohol prolonged this measure at selected times, which the authors interpreted in relation to inhibitory signaling.
This was not a direct measurement of receptor activity or a validated bedside impairment test. Reported correlations between some physiological and behavioral changes cannot establish that one caused the other. The paper did not find significant correlations between silent-period changes and working-memory changes.
Why a sure-money choice is not a safe-driving result
The task involved potential gains, not emergency braking, lane control or decisions made under threat of a collision. The authors themselves cautioned against equating this reward task with dangerous driving.
Only 20 of 23 recruited participants completed the experiment; three withdrew because they could not complete behavioral tasks after the higher-dose condition. That attrition may leave a sample less affected than everyone originally recruited. Blinding effectiveness was not checked, and the small sample limits detailed dose and sex conclusions.
The useful DUI distinction is between a particular preference and the ability to process information. One apparently cautious choice cannot establish normal working memory. The site’s naturalistic unknown-odds lottery study used a different design and reward, illustrating why results across “risk” tasks need their own context.
Primary source
Liu H, Zhang Y, Li D, Wang C, Yuan T-F, Jia Y, Wang F. Acute Alcohol Consumption Reduces Uncertainty Choices. Journal of Addiction Medicine. 2025;19:570–577. https://doi.org/10.1097/ADM.0000000000001456.