EVIDENCE LIBRARY
Does Looking More Often Mean Better Visual Search After Alcohol?
In a 20-person crossover study, alcohol increased eye fixations and maze completion time. More looking did not demonstrate more effective visual search.
No. More eye movements can accompany less efficient visual search after alcohol. In a double-blind crossover experiment, 20 young adults made more fixations and saccades and took longer to complete a visual maze after alcohol than after placebo. The findings concern a laboratory search task; they do not establish a roadside diagnostic test or an individual crash risk.
For a DUI interpretation, the useful distinction is between how often someone looks and how effectively that person uses visual information. The number of glances alone does not answer the second question.
A maze task, not a driving examination
Silva and colleagues studied 11 men and nine women, aged 18–32, in alcohol and placebo sessions presented in counterbalanced order. Participants viewed a maze while a 300 Hz eye tracker recorded their gaze. They had to follow a route from the center to an exit, including decision points where they selected a path.
The investigators calculated an individualized drinking dose intended to reach approximately 0.08% alcohol concentration. They monitored exposure with a breathalyzer. Although the paper calls these results “blood alcohol concentration,” no blood specimen was analyzed: they were indirect breath-derived estimates.
Testing began around 30 minutes after drinking began. The mean breath-derived estimate was 0.09% at 20 minutes and 0.06% at 40 minutes, so the experiment should not be described as maintaining every participant at an analytically confirmed blood concentration of 0.08% throughout the task.
More fixations, longer completion
The mean number of fixations increased from 26.55 with placebo to 40.60 with alcohol. Mean fixation duration increased from approximately 246.5 to 272.6 milliseconds. Total fixation time also increased.
The mean number of saccades—the rapid movements between gaze locations—increased from 25.45 to 40.15. Total saccade time increased, but mean duration per saccade did not differ significantly between conditions (p = 0.413). That null finding matters: the study did not show that every eye-movement measure changed.
Mean maze completion time was approximately 8.53 seconds with placebo and 11.36 seconds with alcohol. Participants also made significantly more fixations in both predefined decision areas. Together, those findings support less efficient search under the study conditions, rather than the interpretation that extra looking necessarily improved performance.
What the eye tracker can and cannot reveal
The gaze record directly measured where and when the eyes moved. The interpretation that these changes reflected attention, planning or decision-making difficulty is a scientific inference from the task and the pattern of results. The study did not separately isolate each cognitive process or demonstrate a specific neurotransmitter mechanism.
A maze also differs from traffic. It does not reproduce moving vehicles, peripheral hazards, steering demands or the consequences of choosing an unsafe gap. The findings provide a plausible example of altered visual information gathering, not a conversion formula from fixation count to driving impairment.
This question differs from alcohol effects on smooth pursuit and corrective saccades. That experiment examined tracking a moving target. The present experiment examined visual search through a maze. Neither should be treated as validation of a different eye examination merely because all involve eye movements.
Limitations that affect a DUI opinion
The sample was small, young and selected for health and vision criteria. Multiple eye outcomes were tested using a 0.05 significance threshold, without a reported adjustment covering the full set of comparisons. Some weaker findings therefore deserve particular caution.
The paper also contains inconsistent wording and statistics in a sex comparison at 40 minutes: it describes a significant difference alongside a reported p value of 0.46. That inconsistency prevents a confident claim from that comparison. It does not justify silently changing the reported value.
Correlations between drinking-history measures and gaze outcomes were exploratory associations. They do not establish permanent injury, a diagnosis or the effect expected for a particular defendant.
The defensible application is narrower: more frequent looking can coexist with slower laboratory visual search after alcohol. A case-specific opinion still needs actual alcohol measurements, timing and observed behavior. As the site’s laboratory-test calibration analysis explains, sensitivity to alcohol does not itself establish validated prediction of driving.
Primary source
Silva JBS, Cristino ED, Almeida NL, Medeiros PCB, Santos NA. Effects of acute alcohol ingestion on eye movements and cognition: A double-blind, placebo-controlled study. PLOS ONE. 2017;12:e0186061. https://doi.org/10.1371/journal.pone.0186061.