Category: Uncategorized

Answer: A small human experiment found that participants were slightly less accurate when reporting how many briefly displayed objects they had seen after drinking alcohol than when performing the same task sober. In Harvey and Seedhouse’s 2021 study, average exact-count accuracy fell from approximately 67% before drinking to 64% after drinking. The decrement was statistically detectable, but the study did not test actual eyewitnesses, establish an individual alcohol-impairment threshold, or show that a particular witness miscounted people, vehicles, or events. Its design also prevents alcohol from being isolated completely from testing order and practice.

The useful forensic question is narrow: can a brief numerical observation become less accurate under alcohol exposure, even when simple small-number perception looks relatively intact? This experiment provides task-specific evidence consistent with that possibility. It is not a method for translating a breath alcohol reading into a witness-reliability score.

What did the original visual-counting study test?

Alistair J. Harvey and Molly Seedhouse published “Influence of Alcohol and Cognitive Capacity on Visual Number Judgements” in Perception in 2021 (DOI: 10.1177/0301006620984105). The researchers recruited 40 university students aged 18–21 years, including 30 women and 10 men. All had normal or corrected vision and met their alcohol-use eligibility conditions.

The protocol began while participants were sober. They completed a working-memory assessment and then a visual enumeration test. After a vodka-and-tonic beverage, they completed the enumeration test again. Three participants’ enumeration records were excluded because their patterns of responses were considered outlying or inconsistent with the task. The authors’ statistical results for the principal accuracy comparison are consistent with an analyzed group of 37, not a new sample of 47.

In the visual task, a fixation cross preceded a display of one to nine randomly positioned squares for just 200 milliseconds. A visual-noise mask followed for 400 milliseconds. Participants then entered the number of squares they thought had appeared. They were not asked to remember a crime, identify a person, drive, judge distance, or estimate a moving crowd.

This separation matters. A screen-based count of briefly presented squares can reveal something about controlled attention and rapid number judgments. It does not reproduce the viewing conditions, distractions, stress, lighting, divided attention, or later questioning involved in a real eyewitness account.

How was alcohol exposure measured—and was it a blood alcohol concentration?

The investigators administered vodka at 1.5 mL/kg to men and 1.35 mL/kg to women, mixed with tonic water. The beverage was consumed over approximately 15–20 minutes, followed by a short rest, a water mouth-rinse, and a breath test. Alcohol was measured with a Dräger Alcotest 3000 breath instrument, which recorded milligrams of ethanol per liter of breath.

The paper reported alcohol levels as BAC-style percentages derived from breath measurements, applying a fixed 2,300:1 blood-to-breath partition ratio. The reported mean was approximately 0.05%, with participants ranging from roughly 0.03% to 0.08%. These are the paper’s transformed breath-derived values, not independent ethanol measurements from venous whole blood. The study did not validate an individual’s true blood-to-breath ratio or perform a forensic blood gas-chromatography comparison.

Neither the planned vodka dose nor an instrument-derived BAC equivalent predicts identical neurocognitive effects in each subject. The study’s measurements support a comparison between two group testing occasions; they do not estimate the amount of alcohol a particular witness drank at another place and time.

Did alcohol make people worse at counting?

For exact-number answers across the array sizes, the authors reported average accuracy of about 67% sober and 64% after alcohol. The main alcohol-condition result was F(1,35) = 8.31, p = .007. The numerical difference was only about three percentage points in overall accuracy—not a threefold increase in errors, not a 3% BAC shift, and not a measure of how often actual witnesses make inaccurate statements.

The investigators also assessed enumeration precision using a Weber-fraction metric, which reflects the variability of numerical responses relative to perceived quantity. Lower values on that measure indicate greater precision. Mean Weber fraction was 0.115 sober versus 0.133 after alcohol, with a reported alcohol-condition comparison of p = .033. This outcome supported reduced task precision, but it did not provide an individual error probability for a real incident.

The experiment helps distinguish two claims often conflated in forensic discussion: the ability to see a briefly displayed item and the ability to accurately enumerate a set of visible items. An observation may be visually registered without the exact quantity being reported correctly.

Primary results reported in Harvey and Seedhouse (2021); values are group/task observations, not witness-reliability cutoffs.

Study measure	Original finding	What it does not establish
Participants	40 recruited; 3 enumeration records excluded	Representativeness of all witnesses, ages, or drinking patterns
Exposure	Mean reported BAC-equivalent about 0.05% (range 0.03%–0.08%) from a breath instrument	Direct laboratory whole-blood ethanol concentration for any person
Exact-count accuracy	Approximately 67% sober versus 64% after alcohol; F(1,35) = 8.31, p = .007	A 3% reduction for every person, or a case-specific probability of error
Enumeration precision	Mean Weber fraction 0.115 sober versus 0.133 after alcohol; p = .033	Validity of a particular numerical statement or estimate
Small versus larger arrays	Counts of 1–4 remained relatively accurate; errors were more apparent with larger sets	That all basic vision was unaffected, or that only large crowds can be miscounted

Why are four objects different from eight or nine?

The researchers studied a distinction between subitising and more effortful counting or estimation. Subitising refers to the rapid recognition of very small quantities—often up to three or four items—without sequential counting. Above that range, identifying the quantity generally places greater demands on attention and cognitive processing.

In this experiment, exact counts of the small arrays stayed relatively close to ceiling. The clearer errors involved larger arrays. That pattern is consistent with difficulties in controlled numerical judgments rather than a simple inability to register a small number of items. It does not prove where in the brain alcohol acted, or show that vision itself was universally preserved.

For litigation purposes, the distinction suggests asking what numerical operation the witness actually performed. Someone who briefly saw several objects and offered an approximate number is engaged in a different task from someone who slowly and deliberately counted a stable, clearly illuminated line of people.

Did working-memory capacity change the result?

Before drinking, participants completed an operation-span assessment that combined mental arithmetic with remembering letter sequences. The researchers used these scores to examine differences in working-memory capacity (WMC). They initially divided the recruited group into lower- and higher-WMC halves, and then examined numerical-task performance.

Among alcohol-condition responses, higher operation-span scores correlated positively with exact-count accuracy for displays containing six, seven, eight, or nine squares. The reported correlations were r = .398, .426, .432, and .484, respectively, with nominal p-values from .015 to .002. The same paper did not report a significant overall alcohol-by-WMC group interaction for accuracy (p = .619). Those two facts must be interpreted together.

It is therefore too strong to claim that a high working-memory score “protects” an individual witness from intoxication, or that a low score diagnoses vulnerability. These exploratory associations were made in a small convenience sample, and a median split can obscure continuous variation. Working-memory capacity was not itself a measurement of alcohol concentration or a validated forensic risk classifier.

Were responses always slower after alcohol?

No. The authors expected intoxicated participants to take longer to enumerate larger arrays. Instead, their reaction-time pattern showed faster responses for some larger arrays after drinking, especially among participants classified as having lower working-memory capacity. The authors discussed the possibility of reduced deliberation or more impulsive choices, but treated that as a tentative interpretation rather than a demonstrated causal mechanism.

Speed and correctness must be kept separate. A faster entry on a number keypad is not proof of more efficient perception, and a slower entry is not necessarily an error. A person may make a faster but less accurate judgment, or a correct but slower one. The study recorded both outcomes precisely because no single latency value captures the accuracy of a numerical report.

The response-time pattern does not justify calculating a new performance threshold or claiming that alcohol universally speeds reactions; the appropriate interpretation remains task-specific and qualified.

Does this establish that alcohol causes eyewitnesses to miscount?

No. The experiment supplies evidence about numerical judgments made under its particular laboratory sequence. It does not measure eyewitness identification, event reconstruction, line-of-sight limitations, post-event memory, or courtroom testimony. It measured immediate keypad responses to square arrays displayed for 200 milliseconds.

There is also a major design limitation: the sober task always came before the alcohol task. This was not a placebo-controlled, counterbalanced trial in which beverage order varied across participants. Practice, fatigue, expectancy, timing and other session effects could contribute to differences. The authors themselves proposed future counterbalanced beverage designs and more demanding dual-task conditions. Even a statistically significant within-person group comparison does not eliminate these alternative explanations.

Nor did the study supply sensitivity, specificity, a receiver-operating-characteristic curve, a likelihood ratio, or a validated cutoff for classifying an individual as accurate or unreliable. It should not be used to multiply a witness’s account by 0.64, discount a statement by three percentage points, or assert that one identified BAC determines whether a witness was truthful.

What is the difference between numerical accuracy, memory, and credibility?

Numerical accuracy concerns whether a reported quantity matches the objects presented. Memory concerns retention and retrieval after time passes. Credibility is a broader evaluation that may involve perception, circumstances, consistency, bias, sincerity and corroboration. The experimental response was entered almost immediately; the paper did not measure delayed recollection or deception.

A witness could be sincere but mistaken about a rapidly glimpsed quantity. Another could give a precise count based on a long unobstructed view. The study cannot adjudicate either scenario without an event-specific factual record. It also offers no validated conversion from a breath reading or experimental accuracy percentage into the evidentiary weight of an individual’s testimony.

Forensic inference ladder: keeping observations separate from conclusions.

Evidence level	What the study actually offers	Additional evidence needed for an individual claim
Alcohol exposure	Administered vodka and contemporaneous breath measurements reported as BAC equivalents	Actual case drinking timeline; properly documented specimen/instrument, if relevant
Visual display	One to nine squares presented on a computer screen for 200 ms	Scene geometry, duration, lighting, distance, visual obstructions and distractions
Immediate response	Keypad number, accuracy, precision and response time	Witness’s contemporaneous description and how the original quantity was established
Working-memory association	OSPAN score correlations in a small student group	Validated individual prediction model; none supplied by this paper
Witness credibility or impairment	Not directly tested	Case-specific corroboration, circumstances, differential explanations and appropriate legal standards

How should lawyers and experts apply the paper carefully?

For a case involving disputed quantities—a claimed number of people in a crowd, objects handled, vehicles passing, or items briefly seen—the paper can support asking whether the observer had time to count or merely estimated, whether attention was divided, and whether the quantity exceeded a very small instantly recognizable set. But those questions are hypotheses for factual investigation, not conclusions supplied by the experiment.

Identify what was directly observed, how long it was in view, whether it moved, whether the witness had an unobstructed perspective, how soon the quantity was first recorded, and whether another independent record exists. If alcohol is at issue, separate an instrument’s measured breath ethanol from an inferred BAC and from an opinion about cognitive effects. Do not reverse the group result to calculate an earlier drinking level or to declare that an individual account is invalid.

For background on how to inspect source limits rather than treating a citation as an expert conclusion, see the Attorney Scientists research library. A separate brief on two blood samples and alcohol-elimination calculations illustrates a related analytical principle: a correctly reported observation need not justify every proposed extrapolation. The editorial and research source policy explains how these literature-based briefs are prepared and the limits of their use.

Frequently asked questions

Did everyone perform worse after drinking?

The paper reports a group-level reduction in exact-count accuracy, not that every participant became less accurate. Individual responses varied, and its group averages cannot diagnose a particular witness.

Was the mean alcohol exposure really a measured 0.05% whole-blood BAC?

No. The authors measured alcohol in breath with a Dräger instrument and converted the result to a BAC-style percentage using an assumed 2,300:1 partition ratio. They did not report laboratory-measured whole-blood ethanol as the study exposure.

Does a normal ability to identify two or three items mean someone was unimpaired?

No. Near-ceiling performance on small arrays describes that narrow task. It does not exclude deficits on larger number estimates, sustained attention, balance, driving or other functions.

Does faster counting after alcohol prove greater cognitive efficiency?

No. Faster responses coexisted with diminished overall accuracy and precision. The paper discusses speed–accuracy tradeoffs only tentatively, not as a validated individual diagnosis.

Can the study tell whether an eyewitness counted six people accurately?

No. The experiment did not test that witness, the incident setting, viewing duration, attention or later recall. Its task findings can identify relevant questions but cannot determine the truth of an individual count.

Does a 3-percentage-point group difference create a 3% forensic error allowance?

No. The 67%-versus-64% comparison is an average laboratory exact-answer difference. It is not a measurement-uncertainty interval, a probability that a specified testimony is incorrect, or an adjustment to any alcohol test.

Was this a double-blind randomized alcohol-placebo crossover?

No. Sober testing occurred first for participants, followed by vodka and a repeat task. The lack of a counterbalanced placebo condition limits causal interpretation and generalization.

Bottom line: a narrower, defensible forensic conclusion

Harvey and Seedhouse’s original experiment found modestly lower accuracy and precision in a rapid visual number-judgment task after drinking, with more pronounced difficulty for larger sets and a complicated relationship to working-memory scores. It illustrates why seeing items, counting them, remembering them and giving credible testimony are separate questions. It does not establish a universal intoxication threshold, an individual witness’s error probability, real-world event causation, or measured whole-blood BAC. Its fixed sober-then-alcohol design makes cautious interpretation essential.

Primary source: Harvey AJ, Seedhouse M. Influence of Alcohol and Cognitive Capacity on Visual Number Judgements. Perception. 2021;50(1):39–51. doi:10.1177/0301006620984105. AI-assisted literature appraisal of the complete original publication; not a case-specific legal or clinical opinion, and no personal review or adoption by Okorie Okorocha is represented.

Answer: In a controlled study of 24 healthy young adults, drinking alcohol immediately after learning a visuomotor task did not meaningfully improve or diminish the motor memory measured the following day. Participants completed learning while sober, received a placebo or an alcohol beverage, and returned sober 24 hours later for a retention test. Their average retention scores were similar across conditions. This does not mean that alcohol cannot affect motor performance: during a different reaching task performed after drinking, some movement accuracy and speed measures changed. It also does not establish a blood alcohol concentration, an individual’s functional impairment, or causation in a legal case.

That distinction is the central lesson of Hamel, Demers, Lepage, and Bernier (2022), The effects of post-learning alcohol ingestion on human motor memory consolidation, a complete original research report in the European Journal of Neuroscience. The investigators asked a narrow question: might alcohol consumed after acquiring a motor skill alter how well that skill is retained the next day? They separately measured adaptation during alcohol exposure. Those are related experimental outcomes, but they are not interchangeable.

What counts as motor memory in this experiment?

Motor memory here meant retaining a newly learned compensation for altered visual feedback while reaching to a target. In the training portion, participants moved a hand-controlled cursor toward targets while the relationship between movement and visual feedback gradually shifted. Learning to adjust to that distortion is called visuomotor adaptation. The next-day experiment measured whether the adapted reach direction persisted when corrective visual feedback was removed.

This outcome differs from recalling a conversation, accurately identifying a person, remembering an instruction, or performing a previously intended act at a particular time. Those questions involve different kinds of memory, different tests, and sometimes different brain systems. The article did not administer an eyewitness-identification test or ask participants to reconstruct an event.

The principal kinematic variable was hand direction at peak tangential velocity, recorded by the laboratory’s movement-tracking apparatus. This is a measured feature of a reaching movement, not an assay of alcohol in blood and not a clinically validated score of overall impairment. Reach aftereffects give researchers a controlled way to study adaptation retention; they are not a diagnostic shortcut for interpreting a traffic collision or a person’s conduct.

How were the 24 people studied?

The researchers enrolled 24 neurologically healthy, medication-free, nonsmoking adults, including 12 men and 12 women. Mean age was approximately 23 years. Participants reported relatively modest drinking histories and were screened to exclude higher Alcohol Use Disorders Identification Test scores. Their characteristics and the laboratory setting limit extrapolation to older people, people with neurological disease, chronic heavy drinkers, or those taking interacting medication.

Each participant completed all three beverage conditions in a randomized, counterbalanced within-person design: an orange-juice placebo, a medium alcohol dose, and a higher dose. Each condition involved a training visit and a 24-hour retention visit. The three visit pairs were separated by at least seven days, and sessions were conducted at consistent times of day. Thus the design compared a participant against that same participant under different experimental conditions, rather than comparing three unrelated groups.

The initial learning block was performed with the dominant hand before any study alcohol was consumed. Only after acquiring the skill did the participant drink the assigned beverage. About 60 minutes after ingestion, the participant completed a separate adaptation task with the non-dominant hand, this time under the beverage condition’s influence. On the following day, the dominant-hand retention test was performed when the study’s breath testing indicated sobriety. This timing was essential to distinguishing later consolidation from immediate performance under alcohol.

What alcohol was administered, and what was actually measured?

The medium and high beverages were formulated from 94%-by-volume ethanol and orange juice, with dosing adjusted for body weight and sex. For men, the medium and high conditions used 0.5 and 1.0 mL/kg of that ethanol solution; for women, they used 0.425 and 0.85 mL/kg. The study sought peak breath-device readings of approximately 0.05% and 0.095%, respectively. The report describes observed peaks of approximately 0.052% and 0.094%.

The reported values came from a BACtrack electrochemical fuel-cell breath device, with sampling before and after learning and repeatedly for two hours after the beverage. The article refers to these readings as breath alcohol concentration (BrAC), using percent-style device reporting. It did not report forensic laboratory headspace-GC measurements on venous whole-blood specimens. Do not silently rename its device readings as independently measured whole-blood BAC or convert them into legally operative measurements without documenting instrument conventions, matrix, units, calibration, and applicable law.

Alcohol exposure was real and clearly differentiated across the three conditions, but the dose target was not itself a measurement of an individual’s concentration. Nor does a condition average tell us any particular participant’s level at a hypothetical arrest, crash, deposition, or medical encounter. Concentration evidence, performance evidence, and case-specific causal inference answer different questions.

What did the next-day motor-memory test find?

The main study hypothesis predicted that drinking after learning would enhance subsequent consolidation, with greater retention after a larger dose. That prediction was not supported. In the sober next-day retention session, average reach aftereffects were similar after placebo, medium alcohol, and high alcohol. The mixed-model main effect of beverage condition was p = 0.896. The condition-by-retention-phase interaction was p = 0.094.

Original study Table 6: next-day dominant-hand retention, n = 24 in each within-person condition. Mean ± SD of hand direction at peak velocity, in degrees.

Retention phase	Placebo	Medium alcohol	High alcohol
Without corrective visual feedback (NoVision)	5.169 ± 1.214°	5.448 ± 0.956°	5.679 ± 0.947°
With corrective feedback (washout)	3.665 ± 0.765°	3.579 ± 0.625°	3.456 ± 0.729°

These are group laboratory movement angles, not percentages of people who remembered correctly. A higher number does not translate into a corresponding increase in real-world motor competence or driving capability. The investigators used mixed-effects models and additional equivalence analyses. Their paired standardized comparisons for the primary retention phases were small: in the NoVision phase, absolute Cohen’s d_z values comparing conditions were approximately 0.14 to 0.26. The predefined equivalence bound was comparatively large (±0.8). This makes the result stronger than merely noting an insignificant p value, but it does not prove a literally zero effect across every person, dose, movement, or retention interval.

The measured inference is narrow: for this specified visuomotor adaptation task, under these exposure conditions, the experiment did not show a meaningful improvement or decline in next-day retained adaptation. A person could still experience intoxication, show other types of motor or cognitive deficits, or recall unrelated events poorly. None of those separate questions was resolved by the primary retention analysis.

Did drinking affect the different movement task performed while alcohol was present?

Yes, certain secondary movement measurements changed, even though the adaptation measure central to that task remained similar. Sixty minutes after the beverage, participants used the non-dominant hand to learn a second visual distortion. Across conditions, their hand-direction adaptation did not differ significantly; the simple beverage comparisons during its acquisition phase had p = 0.3991. The authors also reported small effects in equivalence analyses for this measure.

But successful adaptation on one variable is not proof of globally normal motor performance. In the same non-dominant-hand phase, the high-dose condition involved faster movement time and greater endpoint error in specific comparisons. Endpoint error was measured as distance, in centimeters, not as a BAC or a categorical pass/fail decision.

Original study’s non-dominant-hand task: selected task-specific observations while participants were exposed to the beverage.

Outcome	Placebo	Medium alcohol	High alcohol	Inference boundary
Acquisition hand direction at peak velocity	9.103 ± 0.547°	9.232 ± 0.539°	9.475 ± 0.570°	No significant simple effect of beverage, p = 0.3991.
Movement time, task aggregate	344 ± 16 ms	333 ± 20 ms	326 ± 18 ms	High vs placebo faster, reported p = 0.0126; speed alone is not safe performance.
Endpoint error, NoVision phase	2.612 ± 0.444 cm	2.862 ± 0.458 cm	3.049 ± 0.442 cm	Worse accuracy for both alcohol conditions vs placebo in reported comparisons.

These observations show why interpreting a negative result requires attention to the exact dependent variable. The researchers could find no meaningful difference in the learning/adaptation angle and still report differences in speed and endpoint accuracy. Such outcomes can coexist without contradiction because they quantify different properties of movement.

The experiment did not use a standardized field sobriety test, a police driving examination, road or simulator driving, a fall-risk evaluation, or a validated determination of when a particular participant was incapable of performing an everyday activity. Its task-specific outcome cannot serve as a universal alcohol-impairment threshold.

Why is “drinking after learning” different from “learning while drunk”?

The distinction is temporal and mechanistic. Acquisition is the period when the person first practices the task. Consolidation refers to changes that help a learned representation persist after practice ends. Retrieval or retention testing examines what remains later. Alcohol can be present during one phase but absent during another, and the research question changes accordingly.

In this study, the dominant-hand skill was acquired while sober. The beverage was administered after that acquisition, and retention was tested sober the following day. Separately, the non-dominant-hand task was undertaken during alcohol exposure, allowing an acute motor-learning comparison. A claim about acute performance therefore should not be mistaken for a claim about delayed consolidation, and vice versa.

Similarly, the literature cited in the original article about alcohol following declarative learning does not turn this study into evidence about the accuracy of an eyewitness’s historical account. A laboratory test of a repeated reach and an interview about a past event differ in encoding, content, demands, and response format. Case reports should specify which memory system and which temporal phase are at issue instead of calling all such findings simply “memory impairment.”

What does the negative main result mean scientifically?

It rejects the authors’ specific prediction that post-practice alcohol would produce a dose-dependent enhancement of retained visuomotor adaptation in their experiment. Their effect-size and equivalence analysis further argues against effects at least as large as the predefined bound under the tested conditions. It is not an evidentiary basis for saying alcohol has no effect on skill learning, no effect on every kind of memory, or no relationship with injuries or collisions.

An inferential problem frequently arises when a negative main finding is exaggerated in either direction. A non-significant comparison by itself does not prove two treatments equivalent. Conversely, a significant change in a secondary movement metric does not establish general intoxication severity. The scientifically defensible description here incorporates both: a specifically tested next-day retention outcome showing little group difference, alongside some contemporaneous movement-speed and endpoint-accuracy changes after drinking.

Nor is a group mean an individual prediction interval. Individuals can vary in metabolism, physiology, adaptation strategy, and task susceptibility. The article was not designed to identify which one person would show a particular effect at a stated concentration, much less why a separate person acted as alleged in litigation.

What are the study’s important methodological limits?

Sample and setting: Twenty-four selected young, healthy adults performed highly structured cursor movements. The sample was balanced by the reported binary sex groups but was not a representative driving, workplace, neurological, or clinical population. Each person received all conditions, strengthening control of between-person differences without removing sample-size or external-validity limits.

Beverage and possible confounding: Every beverage contained orange juice with sugar. The investigators expressly noted that glucose might affect consolidation, so a shared sugar component could make differences harder to detect. The three preparations were not sugar-free controls.

Sleep: Sleep quality between learning and next-day testing was not measured objectively. Alcohol may alter sleep, and sleep can influence consolidation; this possible pathway was not disentangled with sleep physiology in this experiment.

Concentration measurement: Breath-device readings established the study’s exposure pattern and later sobriety under its protocol, not a independently collected forensic whole-blood ethanol curve. Instrument reporting units, source timing, and matrix remain relevant to any comparison with another case.

Scope of motor outcomes: The central task measured hand direction in a visually distorted reaching paradigm. Other movements, fine-motor activities, balance, judgment, decision-making, and divided attention may respond differently. The difference between endpoint error and direction adaptation illustrates this limitation even within the study itself.

Causal interpretation: Randomization supports causal comparison of beverages for outcomes measured under this protocol, but it does not identify the receptor pathway mediating an individual result. The original authors discussed GABA and NMDA systems as possible mechanisms; the experiment did not directly assay each person’s neural receptor activity to establish the precise mechanism.

How should a forensic reviewer apply this evidence?

First define the proposed inference. Is the claim that someone learned a movement while exposed to alcohol, retained a previously learned movement after alcohol, made an error while intoxicated, or remembered a prior occurrence? These are different propositions. A motor-learning retention article may be relevant to the first two; it is not a substitute for eyewitness-memory data or real-world driving evidence.

Next identify the actual proof in the case. A toxicology record may show a measured concentration and specimen collection time. A witness, device, video, or validated assessment may address conduct or task performance. A clinician may document neurologic or other alternative explanations. These sources should not be collapsed into a single asserted equation between concentration and causation.

Finally, consider the strength and direction of the study comparison before using it in advocacy. The most supportable conclusion from Hamel and colleagues is that the tested post-learning alcohol regimens did not produce a meaningful average difference in next-day retained visuomotor adaptation, while certain acute speed and endpoint-accuracy findings changed. A contrary claim that the study proves alcohol improves memory, erases every motor skill, or exonerates a driver from impairment would exceed the research.

For additional context on evaluating source literature, see the site’s forensic research library and source and editorial policy. A separate research brief on alcohol elimination from two blood samples addresses the different question of using measured concentrations and elapsed time, not motor-memory retention.

Frequently asked questions

Did alcohol consumed immediately after practice erase next-day motor memory?

No such effect was found in this 24-person visuomotor adaptation experiment. Group retention measures were similar after placebo and the two alcohol conditions; the study does not establish that every possible motor memory is unaffected.

Did the researchers demonstrate that participants could drive safely after drinking?

No. They did not test vehicle operation or establish an individual driving-impairment threshold. Laboratory reach adaptation is not a validated replacement for direct driving-performance evidence.

Were the reported 0.052% and 0.094% readings measured whole-blood BAC?

No. They were percent-style breath-device results described in the paper as BrAC. The study did not obtain venous whole-blood ethanol results by forensic gas chromatography.

If next-day retention was unchanged, why did accuracy and speed change during exposure?

Because the outcomes measure different things. Retained adaptation was assessed sober on the next day, while speed and endpoint error were measured in a separate non-dominant-hand task during the beverage condition.

Does a non-significant p value establish absolute equivalence?

No. The authors did conduct equivalence analyses against a predefined large effect-size threshold, supporting a bounded inference about meaningful group differences. That is not proof of literally zero effect for each person or every task.

Can this paper resolve whether an intoxicated witness accurately recalls an accident?

No. It tested retention of visuomotor adaptation, not verbal recall, eyewitness identification, prospective intentions, or a witness’s account of an incident.

What else would be needed to address an individual causation question?

Contemporaneous specimen identity and measurement records, timing, exposure history, observed behavior, relevant task demands, plausible alternatives, and validated evidence connecting those facts to the claimed outcome. This single group experiment cannot supply missing case-specific data.

Primary research and interpretation boundary

Primary source: Hamel R, Demers O, Lepage J-F, Bernier P-M. The effects of post-learning alcohol ingestion on human motor memory consolidation. European Journal of Neuroscience. 2022;56(5):4600–4618. doi:10.1111/ejn.15772.

This page discusses the authors’ controlled research results and their limits. It does not provide an individual’s BAC determination, medical diagnosis, legal opinion, or case-specific expert conclusion.

Yes. In a controlled experiment involving 52 adult men, performance on a laboratory test of remembering to carry out future intentions was worse after alcohol than after a placebo beverage. The most persuasive differences concerned remembering to act at a specified time, failing to respond when a cue arrived, and remembering what an intended action was. This is evidence about prospective memory under the study conditions—not proof that a particular witness forgot an event, that a driver was legally impaired, or that any individual’s omission was caused by alcohol.

The original primary study is Smith-Spark, Moss, and Dyer (2016), “Do Baseline Executive Functions Mediate Prospective Memory Performance under a Moderate Dose of Alcohol?” Its distinction between remembering an event and remembering to perform a delayed action can matter when scientific evidence is presented in a civil, criminal, employment, or professional-negligence dispute.

What is prospective memory, and why is it different from event recall?

Prospective memory means remembering to carry out an intended action later: take medication at a specified time, submit a required report, respond when a particular cue appears, or follow up on an agreed task. The person may know the instructions perfectly at one moment and still fail to execute them at the appropriate later moment. A prospective-memory task therefore requires more than retaining information; it also requires noticing a relevant cue, retrieving the intention, and implementing it while occupied with other activity.

Retrospective or episodic memory concerns recalling past information or events. An observer’s ability to describe what happened at a scene does not by itself demonstrate that the observer would reliably remember a separate promise to call someone an hour later. Conversely, failure to perform a delayed action does not automatically prove a person cannot accurately recall a past event. Smith-Spark and colleagues studied the first problem—execution of previously assigned intentions—not eyewitness accuracy, recognition of a perpetrator, or preservation of an incident narrative.

How was the alcohol experiment designed?

The researchers enrolled 52 healthy adult male social drinkers, ages 18–36, in a single-blind, placebo-controlled, within-subject design. Each participant completed one session after an alcoholic beer beverage and another after an alcohol-free beer placebo; beverage order was counterbalanced, with the visits separated by approximately one to two weeks. The published procedures describe a target breath-alcohol level of 0.06 g/kg, and breath testing at intervals during administration. That description is a study target, not a reported direct laboratory whole-blood ethanol measurement for any participant; it should not be silently restated as a legally admissible blood alcohol concentration.

The key outcome was the Memory for Intentions Test (MIST). While completing a word-search activity, participants had to remember eight additional instructions. Some were due after a short or longer time interval, while others depended on noticing an event, such as the examiner making a specified action. Other distinctions concerned whether the response was spoken or behavioral. The six reported MIST subscores reflect overlapping classifications of the same tasks; they are not six independent sets of participants or six independent experiments.

Two parallel test versions were used in a fixed session sequence: MIST-A in the first visit and MIST-B in the second. Counterbalancing the beverage order helps address simple order bias, but fixed test-version order remains relevant when evaluating task equivalence and practice or visit effects. Executive-function measures were collected before beverage administration; they were candidate baseline predictors, not direct measurements of executive impairment caused by alcohol during the session.

What did the directly scored prospective-memory test show?

Across the whole MIST, the alcohol-session mean was 42.56, compared with 45.67 after placebo, on a maximum score of 48. The reported within-person comparison was F(1,50) = 31.81; p < .001. This is a group-level difference of approximately 3.11 score points under controlled testing. A score difference does not tell a forensic examiner which task a particular individual would forget in the outside world or whether a specific omitted action caused any incident.

Selected MIST outcomes, Smith-Spark et al. (2016), original Table 1

Laboratory measure	Placebo mean (SD)	Alcohol mean (SD)	Reported p value	What the comparison addresses
Two-minute delay	7.87 (0.44)	7.52 (0.85)	.007	Remembering a short-delay intention
Fifteen-minute delay	7.35 (1.06)	6.31 (1.57)	< .001	Remembering a longer-delay intention
Time-cued tasks	7.40 (0.98)	6.40 (1.32)	< .001	Acting when a time requirement comes due
Event-cued tasks	7.81 (0.60)	7.42 (1.07)	.018	Acting after an externally occurring cue
Verbal response	7.62 (0.80)	7.02 (0.98)	< .001	Giving the instructed spoken response
Action response	7.60 (0.80)	6.81 (1.40)	< .001	Performing the instructed action

The paper specified a Bonferroni-adjusted significance threshold of p < .008 for the six subscale comparisons. Consequently, the event-cued result at p = .018 did not meet that familywise-adjusted threshold, even though it would satisfy an unadjusted .05 rule. The smaller p values for the other listed subscales did meet the study’s stated adjusted criterion. That difference matters: “a nominal signal was observed” and “the prespecified multiplicity-adjusted test was significant” are not interchangeable statements.

Were failures to act more common than errors about the clock?

The study also classified errors rather than examining only a composite score. Participants made more no-response errors and more errors involving loss of the intended action’s content after alcohol. The authors did not find a statistically significant difference for every error category. These are descriptive groups of test errors, not validated diagnostic markers of alcohol-related omission in a particular real-world case.

Selected MIST error categories, original Table 2

Error classification	Placebo mean (SD)	Alcohol mean (SD)	Reported p value
No response when the prospective task came due	0.08 (0.27)	0.48 (0.73)	< .001
Loss of the content of the intended task	0.23 (0.43)	0.58 (0.70)	.001
Loss of time	0.23 (0.47)	0.31 (0.51)	.316
Substitution of an incorrect response	0.08 (0.27)	0.15 (0.46)	.302
Random response	0.02 (0.14)	0 (SD not reported)	.341

The authors used an adjusted p < .010 threshold for this family of five error comparisons. Thus, the no-response and task-content findings survived the stated correction; the other classifications did not. A person can fail to carry out an intention because a cue was missed, the action was forgotten, competing demands intervened, or an external condition changed. The research does not provide a reliable method for selecting among those explanations in one litigated event.

Does a stronger time-cue effect mean that event cues are unaffected?

No. The longer-delay and time-cue categories produced clear statistically supported differences in this sample. But an event-cue comparison with a p value of .018 is not evidence of equivalence between alcohol and placebo. It means this particular contrast failed to pass the authors’ more demanding familywise significance threshold. The test also had near-ceiling scores in some conditions, and its laboratory event cues may differ from distracting real-world alarms, conversations, or unexpected hazards.

The defensible interpretation is narrower: under these conditions, alcohol-related differences were especially evident in time-based prospective tasks and certain error types. It would be unsupported to turn those group comparisons into a universal claim that an externally cued intention is immune to alcohol, or that every delayed intention is necessarily lost at a given exposure.

Did baseline executive-function scores identify who would fail?

The researchers investigated whether sober baseline measures of executive functions—such as switching, fluency, response control, and working memory—could account for differences in alcohol-session prospective memory. Their full baseline predictor model did not significantly explain the overall change in MIST total score (reported F(6,45) = 1.22; p = .315). Some exploratory models for particular task types were significant in this sample, including a time-cued model (p = .005) and a verbal-response model (p = .032).

Those subgroup-model results are not a forensic forecast for an individual. They do not supply a calibrated classification rule, externally validated probability of forgetting, or clinically established personal threshold. Nor do they demonstrate that alcohol first damaged the measured executive functions and that this damage then caused the prospective-memory effect: the executive-function measures were collected before alcohol consumption. Statistical association among baseline characteristics and later differences should not be described as a directly observed causal chain.

Can good retrospective recognition rule out impaired prospective memory?

Not on these data. The original test included a recognition component, but the critical outcome was whether subjects performed intentions at the appropriate moment, often during a distracting filler activity. Knowing the instruction or recognizing it later is different from executing it when required. A person might recognize all the instructions at the end of an interview while still having missed an earlier time-cued obligation.

That point also runs in the opposite direction. This study is not an experiment on eyewitness identification, recollection of a collision, deposition testimony, false confessions, or memory formation during trauma. Transferring its MIST findings to those domains would require separate evidence that specifically measures the disputed task and the relevant conditions. General statements that “alcohol impairs memory” conceal scientifically important differences among encoding, retrieval, cue detection, task execution, and testimony.

How should breath testing be separated from impairment and causation?

The paper describes a breath-alcohol target for administering a moderate alcohol condition. It does not provide a simultaneous forensic whole-blood gas-chromatography measurement, an individual blood-to-breath partition study, or a validated BAC-based rule for missed intentions. Its measured outcome was laboratory task performance; the breath measurement was an exposure-control procedure. The prospective-memory score was not a measured alcohol concentration, and neither score nor breath target proves a real-world incident’s cause.

In forensic assessment, the distinction matters at each inference stage: the dose and route describe administered exposure; breath testing addresses an exhaled-air measurement; MIST scoring quantifies behavior on assigned laboratory tasks; and a legal conclusion about culpability or causation requires facts and standards beyond the study. For analytical background, see the site’s discussion of blood-alcohol elimination using two samples—a distinct measurement problem that cannot be substituted for evidence about future-intention performance.

What limitations matter when applying this study to a case?

Population. All 52 analyzed participants were healthy male social drinkers from a restricted age range. The article cannot establish a population-specific rule for women, older adults, people with neurological conditions, heavy or dependent drinkers, or people taking sedating medications. Design. Although beverages were counterbalanced, the study was single blind, and test versions were consistently tied to the first and second visit. Expectancy, learning, and version effects may not be entirely removable by the crossover design.

Task ecology. The experiment required short-delay laboratory instructions embedded in a word search, rather than actual medication compliance, emergency response, work safety, jury testimony, or driving. Measurement. The study’s breath target cannot be treated as direct whole-blood BAC, and test scores do not measure clinical impairment or diagnose intoxication. Individual variability. Group means and p values do not identify whether one person’s later failure reflected alcohol, forgetting, distraction, misunderstanding, deliberate noncompliance, or another condition.

A legally useful analysis would align the claim with the correct psychological process: was the disputed behavior an intention to act later, an observation made during an incident, or an action requiring immediate psychomotor control? The answer determines whether this study has meaningful relevance at all.

Frequently asked questions

Does alcohol make people forget to perform tasks they previously agreed to do?

It can reduce performance on some prospective-memory tasks in a controlled setting. This experiment found lower average MIST scores after alcohol than after placebo and more omission-type errors. It cannot prove why a particular person failed a particular obligation outside the laboratory.

Is prospective memory the same as eyewitness memory?

No. Prospective memory concerns performing a future intention after a cue or delay; eyewitness or episodic memory concerns encoding and reporting an earlier event. The ability to recall an incident cannot be directly inferred from MIST scores, and failure to keep an appointment does not establish unreliable eyewitness identification.

Were time-cued tasks more sensitive than event-cued tasks?

The time-cued contrast was statistically significant under the study’s multiplicity correction; the event-cued contrast did not pass that adjusted threshold. This does not prove event-based memory is unaffected. It describes the pattern of this particular test with its specific scores and sample.

Does a participant’s good executive-function score predict that alcohol will not affect them?

No validated individual prediction follows. Baseline executive tests were administered before drinking, and the authors’ full prediction model for the overall MIST difference was not significant. Select task-specific associations should not be mistaken for an independently validated impairment-screening tool.

Does the breath-alcohol target establish the person’s legal blood alcohol concentration?

No. The study described a controlled breath-alcohol target rather than forensic analysis of a whole-blood specimen. Breath testing, laboratory task scores, modeled blood values, and legally operative blood concentrations are different kinds of evidence and must not be interchanged.

Can this paper determine whether an alcohol-exposed person caused a missed safety check?

No. It provides experimental evidence that prospective-memory performance can differ under alcohol exposure. Evaluating one omission requires documented timing, actual duty, instructions, competing tasks, alternative causes, contemporaneous exposure evidence, and the applicable legal framework. No individual causation algorithm was validated.

Where can attorneys find the underlying methods and publication standards?

The complete primary article is available through the original peer-reviewed DOI. The site’s research resources and editorial policy provide further context for evaluating scientific claims. Those resources are not substitutes for examining the full original study and case record.

Bottom line for forensic interpretation

Smith-Spark and colleagues provide controlled evidence of a group-level reduction in remembering to act later, especially for time-cued intentions and omission-type mistakes. They do not demonstrate a validated individual intoxication threshold, a particular witness’s inability to recall past events, actual driving impairment, or the cause of a single disputed omission. In an expert report or courtroom analysis, the scientifically sound sequence is to identify the exact memory process, describe the measured laboratory outcome, disclose uncertainty and study limitations, and resist inferring more than the primary data support.

Short answer: Yes. Gender-affirming hormone treatment can change body composition and the assumptions used in a Widmark-style blood alcohol calculation. A 2022 forensic modeling paper found that choosing different sex-based total-body-water equations could materially change the calculated initial blood alcohol concentration. It did not measure alcohol concentrations in a trial of transgender volunteers, validate an individual correction factor, or establish impairment. The scientifically important issue is whether the report identifies what was measured, which physiological inputs were estimated, and which assumptions drive the final number.

In a forensic case, a numerical calculation is not an independent chemical measurement. Maskell and colleagues used previously measured body-water data and published hormone-treatment body-composition changes to examine an especially important model-input question: how should the analyst represent distribution volume when an individual’s physiology may differ from the population assumed by a standard equation? The complete original paper is Maskell et al., Journal of Forensic Sciences (2022), DOI: 10.1111/1556-4029.15052.

What exactly does a Widmark calculation estimate?

Widmark-style calculations use an assumed ethanol dose and an estimate of the volume into which ethanol distributes. In the formulation analyzed by the authors, the hypothetical initial concentration, designated C₀, depends on the amount of ethanol, total body water (TBW), and the fraction of whole-blood volume consisting of water. Other forensic exercises also require defensible assumptions about when drinking occurred, absorption and elimination. Changing an input changes the output even when there has been no change in the actual evidence.

The paper’s C₀ is a model output, not a blood specimen obtained at “time zero.” It is not the concentration measured by a laboratory, a direct measurement of brain ethanol, or a prediction of a person’s reaction time, judgment, gait, or driving performance. Calling this modeled quantity a “BAC reading” would conceal the central scientific limitation.

An actual forensic whole-blood ethanol result, when properly obtained, is different evidence: it is an analytical concentration reported for a specified specimen and collection time. An instrument’s breath-alcohol result is another measurement type. A retrospective reconstruction or dose-to-concentration estimate must be labeled as an inference, not substituted for either measurement.

Why can hormone treatment matter to alcohol distribution?

Ethanol distributes substantially through body water. Lean mass generally contains more water than adipose tissue; thus two individuals of similar body weight may have different distribution volumes. Gender-affirming treatment can alter fat-free mass, total fat, and hematocrit. Those physiological changes are relevant to the assumptions behind group-derived TBW equations, but they do not supply a direct measure of one person’s current body water.

The researchers drew on previously published longitudinal treatment data in which average lean mass fell by approximately 3% after 12 months among 179 transgender women and rose by approximately 10% among 162 transgender men. Those are group averages from the cited external study, not experimentally established changes in every person. Dose, duration, baseline composition, age, and other differences may matter; the 2022 model did not individually measure all of those variables in a new drinking experiment.

The authors also examined blood-water fraction indirectly. They used published changes in hematocrit and an established hematocrit-to-whole-blood-water relationship to estimate the blood-water component for their calculations. The modeled values were approximately 0.841 for transgender women and 0.826 for transgender men. Neither value was obtained by directly measuring whole-blood water in a newly recruited transgender cohort. Treating them as personally measured results would add certainty the study does not provide.

What was the study design—and what was actually observed?

This was an analytical modeling study built from existing data, not a randomized ethanol-administration trial in transgender participants. Two historical datasets contained directly measured TBW, using a tritiated-water dilution technique, for 582 cisgender men and 884 cisgender women. Published average body-composition changes associated with hormone treatment were then applied to model TBW after treatment. The authors used those modeled TBW figures, modeled blood-water fractions, and specified ethanol doses to calculate hypothetical C₀ values under alternative equation assumptions.

Consequently, the sample sizes of 582 and 884 describe the underlying cisgender body-water datasets. They are not numbers of transgender people prospectively given alcohol, and the outputs are not measured pre/post-treatment BAC values. The key results quantify sensitivity of a mathematical procedure to its selected inputs.

Evidence types in Maskell et al. (2022)

Component	What it represents	What it cannot establish
Historical TBW datasets (582 men; 884 women)	Previously measured total body water in cisgender populations	Direct body-water measurement in a transgender person on treatment
Published 12-month treatment changes	Average changes in lean mass used to adjust the model	An individual patient’s actual percent change
Hematocrit-based blood-water fraction	Estimated water content from an indirect physiological relationship	A new individually measured blood-water fraction or BAC
Modeled C0 at 16 g and 80 g doses	Hypothetical initial BAC obtained under alternate equations	A chemical test, observed peak BAC, impairment score, or event-time BAC

How much could equation choice change the result?

The differences were not merely semantic. Table 5 of the source reports the following group-mean calculated C₀ values for the authors’ hypothetical 80-gram ethanol dose. Units are milligrams per 100 milliliters of blood. These figures are illustrative outputs from alternative equations applied to modeled groups; they are not concentrations measured in persons who drank 80 grams in the study.

Modeled mean initial BAC after a hypothetical 80 g ethanol dose

Modeled group	Sex-at-birth equation	Affirmed-gender equation	Authors’ adjusted model
Transgender women after modeled treatment	154 mg/100 mL (cis-male equation)	188 mg/100 mL (cis-female equation)	156 mg/100 mL
Transgender men after modeled treatment	208 mg/100 mL (cis-female equation)	172 mg/100 mL (cis-male equation)	182 mg/100 mL

For example, in the modeled transgender-women group, the cis-female equation produced a mean C₀ of 188 mg/100 mL versus 156 mg/100 mL with the authors’ adjusted equation. For the modeled transgender-men group, the corresponding cis-female calculation was 208 mg/100 mL versus 182 mg/100 mL after adjustment. The result changes because the assumed distribution terms change, not because the study detected different blood ethanol in the same person.

The authors also modeled a smaller 16-gram ethanol dose. Their overall mean C₀ calculations were 31, 38, and 31 mg/100 mL, respectively, for transgender women using the sex-at-birth, affirmed-gender, and adjusted models. For transgender men, the three corresponding model outputs were 42, 34, and 36 mg/100 mL. These are calculation demonstrations; they are not empirically validated dose-response tables for individual clients.

Is using a gender marker the same as measuring body water?

No. A demographic marker identifies a category for selecting a formula; it does not directly measure total body water, hematocrit, or actual blood ethanol. A choice between two population-derived equations can create markedly different outputs while leaving the person’s true physiology unknown. The correct scientific response is to identify that dependency and describe it, not to pretend that one label resolves all interindividual variation.

Maskell and colleagues suggested a transparent reporting assumption for circumstances in which sex assigned at birth is not known or disclosed. Their proposal concerned how to document the equation choice; it was not validation that a gender marker, birth classification, or treatment history supplies a person-specific distribution factor. The paper itself calls for more representative primary measurements and validation. Nor does a methodological recommendation in a 2022 article determine the governing legal rule in any particular jurisdiction.

When treatment history is relevant and appropriately available for a case, an analyst should distinguish documented facts from assumed population averages. A sensitive personal attribute should not be guessed or presented as an analytic measurement. The objective is to describe the evidentiary basis, alternatives, and uncertainty without drawing a conclusion about identity from a numerical model.

Can these percentages be used as a correction to measured BAC?

They should not be applied mechanically to a laboratory result. The percentages concern model-estimated TBW or hypothetical C₀ under selected equations. For example, an approximately 17.7% modeled TBW underestimation for one comparison does not authorize adding or subtracting 17.7% from a forensic whole-blood GC result. TBW and BAC are different quantities, and the mathematical response is not universally linear across a changing individual, dose, and time.

Likewise, the 2022 paper did not establish a treatment-specific ethanol elimination rate, drinking-time correction, or impairment threshold. The authors noted that their forensic framework used existing elimination-rate ranges for both sexes. A drinking-history calculation can be sensitive to several assumptions at once: ethanol amount, completeness and timing of absorption, dose history, distribution, elimination, and specimen timing. This study isolates a model-input issue; it does not validate all the others.

Comparing a calculated C₀ with a measured blood ethanol concentration also requires care. One may represent an idealized concentration immediately after hypothetical distribution; the other represents ethanol actually assayed at a documented time in a specific specimen. Agreement between numbers would not by itself validate all assumptions, and disagreement does not, by itself, identify which input was wrong.

What uncertainty did the authors acknowledge?

The authors explicitly described their adjusted equations as unvalidated. They modeled average changes following approximately 12 months of hormone treatment, drawing primarily on data from Western populations across a limited adult age range. The historical datasets and treatment cohorts were not a single sample in which actual body water, hormone exposure, and observed ethanol kinetics were prospectively measured together.

Differences between individuals—including baseline body composition, height, weight, treatment duration, and the response of fat-free mass—cannot be represented fully by one average percentage. The hematocrit-to-blood-water relation adds another inferential step. That uncertainty should be characterized, not hidden behind excessive decimal precision or the appearance of a personalized formula.

Importantly, the paper did not provide a validated prediction interval for one person’s post-treatment TBW or an external validation cohort for the proposed equations. Its numerical examples show potential modeling discrepancies, but they cannot determine that a particular person’s true C₀ lies at the adjusted point estimate rather than elsewhere.

Does this paper establish alcohol impairment or crash causation?

No. The investigators did not give transgender participants ethanol and then measure driving, divided attention, field sobriety performance, reaction time, aggression, or crash outcomes. No field-valid diagnostic classifier was trained or tested. Even a well-validated blood ethanol measurement is distinct from a specific conclusion about an individual’s functional performance or legal responsibility at an earlier event time.

The study can support a critique of a model whose input categories were chosen without explaining their physiological assumptions. It cannot identify the alcohol concentration at the time of a disputed incident, establish that a person actually drank the posited number of grams, or connect an estimated number to a claimed act or injury. Those questions require independent case facts and appropriate scientific methods.

How should a lawyer or toxicologist review such an estimate?

Begin by identifying the precise proposition being advanced. Is the witness offering an actual blood test at a known draw time, a breath measurement, a modeled initial distribution concentration, or a retrospective estimate of an earlier concentration? The documentation should not shift between those categories without explanation.

Then identify the equation, inputs, units, body-water method, population assumptions, and uncertainty analysis. If a demographic category selected the equation, ask what physiology the category was meant to approximate and what evidence supports applying it to this individual. When an adjusted TBW estimate is invoked, determine whether it was directly measured or projected from published averages and whether the particular application was validated.

Finally, examine the dose and timing premises independently. Even a well-reasoned treatment of distribution volume cannot rescue an unsupported drinking quantity, an unknown absorption state, or an assumed elimination course. The two-blood-sample elimination research brief explains a separate timing limitation. The Attorney Scientists research library organizes additional analytical and interpretive reading. The site’s editorial and source policy explains how educational summaries differ from adopted expert opinions.

Frequently asked questions

Did the researchers measure blood alcohol in transgender volunteers?

No. They analyzed earlier body-water and treatment-composition datasets and computed hypothetical BAC outputs. This was not a new alcohol dosing trial in transgender volunteers.

Does hormone treatment always produce the study’s average body-water change?

No. The −3% and +10% average changes used in the model are group estimates, not measurements for every treated individual. Duration, baseline characteristics, and individual responses can differ.

Can the modeled 188 mg/100 mL result be described as a measured BAC?

No. It is one group-mean C₀ under a selected hypothetical dose and equation. No blood specimen in this study yielded that number.

Does the paper validate a new Widmark formula for every person receiving treatment?

No. The authors specifically acknowledge that their proposed anthropometric adjustments have not been validated against a representative set of direct body-water measurements in that population.

Should an analyst simply add a fixed percentage to a forensic blood result?

No. The modeled distribution-volume discrepancies are not demonstrated correction factors for already-measured blood ethanol, breath ethanol, or an event-time reconstruction.

Does sex assigned at birth alone determine the person’s total body water?

No. A sex-based equation estimates a population tendency. It does not directly measure the individual’s body composition. The scientific question is what evidence supports a particular assumption and how uncertainty is disclosed.

Does a calculated C₀ establish DUI impairment or responsibility for an event?

No. This paper has no functional impairment or crash-causation endpoint. A dose-to-concentration model cannot substitute for properly interpreted toxicology measurements and separate behavioral or event evidence.

Primary scientific source and interpretive boundary

Maskell PD, Sang KJC, Heymsfield SB, Shapses S, Dekorompay A. Forensic alcohol calculations in transgender individuals undergoing gender-affirming hormonal treatment. Journal of Forensic Sciences. 2022;67(4):1624–1631. DOI: 10.1111/1556-4029.15052. This appraisal addresses the published study’s model assumptions and aggregate calculations. It is educational scientific context, not a validated individual BAC correction, medical advice, or an opinion about a particular case.

No. A positive THC hair result cannot, by itself, establish that cannabis caused a particular adolescent’s memory difficulty. A 2026 longitudinal study found that cannabis-use onset was associated with less favorable age-related changes across several cognitive tasks, and that a smaller group with THC detected in hair showed a different trajectory of episodic-memory performance than hair-negative controls. Those are meaningful group-level findings. They are not a diagnostic rule for an individual child, a measurement of intoxication at a specific event, or proof of causation.

This distinction matters when toxicological records appear in school, injury, juvenile, or other litigation. The defensible question is not simply whether a cannabinoid was detected. It is which specimen and analyte were tested, what timeframe that test can represent, which neurocognitive outcome was measured, and what an observational study can infer about the person in question.

What did the 2026 longitudinal study actually examine?

Wade and colleagues analyzed the Adolescent Brain Cognitive Development (ABCD) Study, following participants who entered the study largely before substantial substance use. The primary longitudinal analysis included 11,036 young people, ages 9–17, using repeated neurocognitive assessments and a time-varying classification of cannabis-use onset. Classification combined participants’ own reports with available urine, oral-fluid, and hair toxicology rather than relying exclusively on self-report.

The researchers modeled age-related change in memory, processing speed, inhibitory control, working memory, language, and visuospatial tasks. Their main models accounted for participant, family, and study-site clustering, along with reported other substance use and several demographic, developmental, and family-related variables. Primary results were assessed using false-discovery-rate correction. These safeguards strengthen the longitudinal comparison; they do not turn a nonrandomized observational cohort into a randomized THC-exposure experiment.

A secondary analysis concentrated on repeated hair testing at study waves 2, 4, and 6. The paper describes 648 classified participants in this hair-testing subset: 546 with no hair cannabinoids detected, 81 classified as THC only, and 21 classified as CBD-positive (whether or not THC was also present). The episodic-memory model reports 645 participants and 1,885 observations. These are related but not identical denominators; neither should be presented as the size of the 11,036-person primary cohort.

Myth: “Teen cannabis users always scored worse from the beginning”

Evidence: The larger cannabis-use group sometimes performed better at younger ages than the comparison group, but generally showed less improvement or flatter trajectories as adolescence progressed. For example, the paper’s working-memory models indicated better estimated performance among young future users during ages 9–14, no difference at 15–16, and worse estimated performance at 17. Other tasks exhibited related but not identical age patterns.

The longitudinal age-by-use interactions therefore matter more than a simplistic statement that every user performed worse at every measurement. The authors discussed alternative explanations for the early advantage and later flattening, including shared vulnerability, developmental factors, possible regression toward the mean, and factors associated with initiating use. Observational associations do not isolate THC as the only mechanism.

Myth: “A hair THC result tells us when a teen was intoxicated”

Evidence: Hair toxicology can contribute information about exposure over a longer retrospective window than many fluid tests. In this study, a proximal hair sample provided an approximate three-month exposure window, and cannabinoid detection was more likely with repeated or moderate-to-heavy use than with isolated experimentation. A hair result did not establish the exact day, amount, dose absorbed, or psychoactive effect at any specific school examination, driving event, injury, or alleged incident.

The investigators grouped hair results as “THC only,” “CBD+,” or hair-negative controls for research. That label is not a clinical diagnosis. The “THC only” group reflected THC or related THC analytes in hair with no CBD detected under the tested conditions; it does not certify that participants exclusively consumed a chemically pure THC product. Similarly, “CBD+” was not a CBD-only exposure group: 13 of the 21 CBD-positive participants also had THC detected.

The paper itself identifies further analytical limits: hair tests were performed by two laboratories with different methods and reporting thresholds; a positive for the parent cannabinoid alone may require attention to possible external contamination; and infrequent exposure can evade detection. The investigators also note that the study’s oral-fluid and urine screens were not confirmatory analyses. These are reasons to examine the underlying analytical report rather than treat any study-group label as a forensic conclusion about one person.

What did the hair-THC analysis actually find about memory?

In secondary models, the researchers reported an interaction between age and the hair THC-only group for episodic memory compared with hair-negative controls. The standardized interaction coefficient was β = −0.60, p = 0.007. Higher episodic-memory scores represented better performance; the negative interaction reflected a smaller age-related improvement in the THC group. At approximately ages 15–17, estimated episodic-memory performance was worse in the THC group than among hair-negative controls.

The paper did not find statistically significant hair-group associations across all other evaluated domains in that secondary analysis. It specifically reported no significant hair-group relationship for receptive language, processing speed, inhibitory control, or oral reading. The smaller secondary analysis also used different covariate controls from the main cohort models. Findings from the full cohort and the hair subgroup therefore should not be described as if the same exposure definition produced identical results on every task.

The CBD-positive group did not differ significantly from hair-negative controls on the reported memory comparison. That is not evidence that CBD restores cognition or prevents THC-related harm: the CBD group contained only 21 participants, some with THC co-detection, and a statistically nonsignificant comparison is not an equivalence or protective-effect trial.

Detection, measurement, developmental association, impairment and causation are different questions

Evidence or outcome	What it can support	What it does not prove on its own
THC-related analyte in hair	Exposure under an assay’s collection, preparation, detection and reporting criteria	Exact use date, administered dose, intoxication at an event, or an individual memory deficit
Urine cannabinoid screen	Detection of a drug-class signal within that screening assay’s limitations	Specific recent use time, blood concentration, or confirmed drug identity without appropriate follow-up
Oral-fluid THC result	Analyte detection in a named specimen over a method-dependent period	A validated universal mapping to memory test performance or driving impairment
Repeated cognitive test scores	Performance on named neurocognitive tasks at assessment times	That THC caused a particular score or explains all real-world functioning
Group-by-age model coefficient	Statistical association between group classification and change over age in the modeled cohort	A causal effect size or diagnostic threshold for a particular adolescent
Individual case opinion	Potentially a carefully bounded interpretation after reviewing complete clinical, analytical and temporal records	Certainty imported from a population association without case-specific support

Why the difference between a longitudinal study and an acute intoxication experiment matters

The ABCD work concerns developmental trajectories over years. It does not assign adolescents to consume THC and then test performance at a precisely controlled post-dose interval. Even though toxicology improved the classification of cannabis exposure, testing did not provide uniform pharmacokinetic sampling after one known administered dose.

For contrast, an acute THC-potency and simulated-driving research brief discusses a controlled trial measuring specific driving outcomes after smoking cannabis. That research asks a different question and has different limitations. Neither article supports treating a hair finding as direct evidence of driving impairment, and neither can replace a clinical assessment of an adolescent’s actual cognitive functioning.

Could the statistical association reflect factors besides cannabis?

Yes. The main models controlled for multiple measured potential confounders, including alcohol, nicotine, other substance use, sociodemographic factors, family substance-use history, prenatal exposure, baseline psychopathology and within-person change. Nevertheless, the researchers expressly acknowledged residual time-varying confounding and the possibility that common vulnerabilities influence both cannabis initiation and cognitive development.

The hair-positive subgroups differed on demographic characteristics and used a reduced covariate set because of their smaller samples. Other limitations included selection of participants with repeated hair samples, small CBD-positive numbers, multiple comparisons, use of different laboratory methods and possible contamination. The authors also noted that motivations for substance or CBD use—such as symptoms they were trying to manage—could affect observed associations.

Those limitations do not erase the observed differences. They constrain the additional inference someone may wish to draw: an association in a repeated-measures model is not proof that the same exposure produced a particular person’s learning or memory problem.

Questions to ask before using this paper in a forensic or educational proceeding

Start with the proposition the evidence is supposed to establish. A laboratory can identify a chemical signal; a psychologist can measure task performance; a longitudinal model can describe associations; and a court may need to decide a separate fact about causation or conduct. Crossing these boundaries requires additional evidence.

For an individual matter, obtain the specimen matrix, collection date, analytes and confirmation methodology, reporting limits, possible contamination and chain-of-custody records, any available independent exposure history, neurocognitive test details, premorbid or earlier scores, medications, sleep and psychiatric factors, developmental history and competing explanations. Ask whether the questioned behavior occurred anywhere near the tested exposure window. Do not substitute a population-average comparison for a person’s observed performance.

Frequently asked questions

Does a positive hair THC test prove memory impairment?

No. The cited study reports group-level age-related associations involving episodic memory. It does not validate a hair-positive/impairment-positive classification rule for an individual.

Can hair testing distinguish one-time experimentation from frequent cannabis use?

Not reliably from a simple positive/negative result alone. The study emphasizes that hair testing is more sensitive to repeated or relatively frequent exposure and may miss infrequent use. Collection, analytes, methods and cutoffs matter.

Did every cognitive task deteriorate in the THC-positive hair group?

No. The secondary hair analysis reported an episodic-memory association, while several other evaluated task domains did not show significant relationships. Primary analyses using a broader lifetime cannabis classification identified more extensive age-related interactions.

Did the study prove CBD protects a teenager’s memory?

No. The CBD-positive group was small, and most importantly its category included some participants with THC detected as well. A nonsignificant result is not a randomized protective-effect finding.

Can these data establish impairment on the day of a crash or alleged event?

No. The paper studied repeated cognitive trajectories over adolescence, not a driving event, a specific use-to-event interval, or a validated individual impairment cutoff.

Is a longitudinal observational association useless in litigation?

No. It can be relevant background about developmental associations and the importance of evaluating both self-report and toxicological measures. Its weight depends on the exact proposition being offered, comparability of the individual and study population, analytical quality, confounding and temporal evidence.

Primary scientific source and reading scope

Wade NE, Sullivan RM, Wallace AL, et al. Longitudinal neurocognitive trajectories in a large cohort of youth who use cannabis: combining self-report and toxicology. Neuropsychopharmacology. 2026;51(9):1546–1555. DOI: 10.1038/s41386-026-02395-1. The complete 22-page user-supplied author-manuscript PDF from THC/CBD Batch 51 was reviewed, including methods, primary and secondary results, discussion and study limitations. The supplied supplementary data were not independently reviewed.

This is an AI-assisted scientific research brief, not a personal expert opinion adopted by Okorie Okorocha, a clinical diagnosis, legal advice or a conclusion about any individual. See the editorial and source policy and the research library for related method-focused reading.

Short answer: Drinking alcohol after witnessing an incident does not, by itself, establish that the witness encoded the incident while intoxicated or that their subsequent memory is inaccurate. In a randomized laboratory study of 60 self-described heavy drinkers, researchers found no statistically significant between-group difference in free-recall completeness or accuracy when participants drank a moderate amount of alcohol after watching a simulated traumatic event. The alcohol group gave more accurate details at a one-week follow-up than at its first interview, but this within-group improvement did not establish that alcohol improved memory relative to placebo. The study is a reason to examine when drinking occurred, not to declare a witness reliable or unreliable automatically. Butterworth et al. (2023), DOI: 10.3389/fpsyg.2023.1007477.

Myth: Any alcohol use after an incident invalidates eyewitness memory

Evidence: Alcohol use before an event, after an event, and during an interview are not interchangeable exposures. The first concerns the conditions under which information was initially encoded. The second occurs after that initial observation. Drinking during a later interview may affect performance at retrieval. These timings can overlap in an actual case, but they are distinct scientific questions. The controlled study considered the post-event sequence specifically: participants watched the same simulated traumatic motor-vehicle incident while sober, then received alcohol or placebo and completed an initial memory test.

The distinction matters in litigation because evidence of drinking after an assault, accident, or other incident is sometimes offered as if it proves intoxication at the moment of observation. It does not. The relevant factual timeline includes the event, any alcohol consumed before it, the first drink afterward, testing and interview times, and what the witness was actually asked to remember.

What did the controlled eyewitness-memory study actually test?

Butterworth, Hand, Lorimer, and Gawrylowicz recruited 63 adult participants who described themselves as heavy drinkers; 60 completed the follow-up. Participants were randomly allocated to alcohol or placebo, watched a film depicting a traumatic car accident, and then consumed study beverages. The participants, but not the experimenter, were blinded to beverage assignment. The researchers sought a breath-alcohol concentration of approximately 0.20 mg/L in the alcohol condition, described in the article as equivalent to about 0.04% blood alcohol concentration. This was a moderate-dose laboratory manipulation, not a test of very high intoxication.

After an approximately 45-minute beverage-and-distractor interval, participants provided free-recall and cued-recall accounts. The investigators repeated testing one week later, when the alcohol condition was no longer acutely alcohol-exposed. The design therefore compared conditions and the change between an initial and a repeated interview. Its outcomes were memory-task measures, not legal determinations of witness credibility.

What did the researchers measure: completeness, accuracy, or confidence?

The study separated several outcomes that should not be collapsed into a single label of “good memory.” Completeness meant the number of details produced in a free account. Accuracy compared reported details against the film and was expressed as the proportion of recalled details coded correct. In the cued-recall task, the investigators also tracked responses labeled “remember,” “know,” or “do not know” and recorded confidence judgments. None of these measures alone is equivalent to credibility in a live court proceeding.

Free-recall measure	Alcohol after film	Placebo after film	Interpretation
First interview: mean total details	25.10	27.10	A descriptive difference; no significant between-group difference was found.
One-week interview: mean total details	28.90	27.93	At follow-up, the groups still did not differ significantly.
First interview: mean proportion accurate	92.32%	91.17%	Group averages, not a probability that a specific witness is truthful.
One-week interview: mean proportion accurate	92.22%	90.71%	Descriptive averages within this particular filmed-event task.

Source: Butterworth et al. (2023), Table 2. These are group means. They are not effect-size estimates, individual predictions, or validated real-world witness-accuracy rates. Full primary study.

Myth: A better second interview proves that alcohol improved memory

Evidence: The alcohol group showed statistically significant within-group gains in its number of recalled details and accurate details from the first to the one-week interview. The placebo group did not show the same change in those outcomes. The authors also reported statistically significant session-by-condition interactions for total and accurate details. Yet there was no statistically significant difference between alcohol and placebo groups at either interview on those free-recall measures.

That distinction prevents an important overstatement. A change over time in one randomized group is not automatically proof that alcohol caused a superior memory outcome compared with not drinking. All participants were asked to recall the incident at the first interview, so practice or retrieval effects from that initial memory test could contribute to later performance. The authors explicitly identified the absence of a condition that omitted immediate recall as a design limitation. Their proposed explanation involving reduced post-event interference is a hypothesis, not a directly established physiological mechanism.

Myth: Saying “I don’t know” after drinking means the memory is false

Evidence: In cued recall, participants in the alcohol condition gave relatively more “do not know” responses than those in the placebo condition; the distribution of answer types differed by condition. However, an inability or unwillingness to supply a detail is not the same as reporting an inaccurate detail. “Remember” referred to conscious recollection with contextual experience, while “know” referred to a sense of familiarity without the same contextual re-experiencing. These are task-specific response categories, not diagnostic tests for whether a witness is lying.

The study also reported that correctly classified “know” responses increased from the first interview to the later interview in the alcohol group. That observation does not make familiarity an automatic substitute for detailed recollection, nor does it establish how accurately people will identify a perpetrator in a real case. The authors acknowledged that participants might not understand the “remember/know” categories identically.

Myth: Confidence and accuracy always move together

Evidence: The alcohol group’s overall confidence decreased between the initial and delayed sessions even as its free-recall detail count increased. The researchers found relationships between accuracy and confidence in their task, but confidence depended on the response type, group, and testing session. The experiment therefore does not support treating a confident statement as automatically correct, or treating reduced confidence as proof that a later account is false.

For an evidentiary review, distinguish the precise proposition: confidence is a subjective judgment; completeness counts details; accuracy requires a reference against which claims can be checked. In the filmed-event experiment the researchers had a known ground truth. In a real dispute, the trier of fact may have an incomplete, contested record, making accuracy much harder to determine directly.

What should lawyers and experts compare in a post-event alcohol timeline?

First identify the encoding interval: what the witness could see and hear, and whether any alcohol had been consumed before the event. Next record post-event consumption, including start time and reported quantity, without assuming that the timing of a later alcohol measurement describes an earlier observation. Finally map the retrieval opportunities: initial informal account, police interview, recorded statement, and later testimony. Note the elapsed intervals, whether the witness was drinking during any account, and what external information or questions may have been introduced in the meantime.

For example, compare a witness who was sober at 9:00 p.m., drank at 9:30 p.m., and was interviewed at 10:15 p.m. with a witness who had been drinking since 8:00 p.m. and was interviewed the next morning. Both may be described casually as “a witness who drank,” yet the potential questions about encoding and retrieval differ. This example illustrates chronology, not an estimate of BAC or individual memory accuracy.

What the study cannot establish for a real witness

The experiment involved an analog trauma film, not a real event with the danger, stress, competing attention demands, or social context of an actual assault or collision. The group was relatively small and selected: adults who self-described as heavy drinkers; participants with certain alcohol-use and trauma-related conditions were excluded. The researchers did not control for every participant’s alcohol-blackout history, did not ensure equal perceived stress across groups, and did not collect every possible subjective intoxication measure. The experimenter was not blinded to drink assignment.

Importantly, the study did not validate memory under very high alcohol concentrations, test whether a particular person fabricated a detail, establish that delayed questioning is always superior, or provide a forensic formula converting breath-alcohol concentration into memory accuracy. A week-later improvement in a controlled film task is not a guarantee that an interview a week after a real incident will produce better evidence. Equally, a person’s post-event alcohol use alone cannot establish that every earlier observation must be discounted.

Frequently asked questions about drinking after an incident

Does alcohol consumed after an event cause an eyewitness to forget what happened?

Not invariably. In this 60-person controlled study, alcohol given after viewing an event did not produce a statistically significant between-group reduction in free-recall accuracy or completeness at either test. The result cannot be generalized to every dose, individual, event, or time interval.

Is drinking before the event the same as drinking afterward?

No. Drinking before an event may affect perception and encoding. The cited study intentionally placed alcohol administration after the film to ask a different question about later recall. The study did not directly test intoxicated observation of the film.

Why did the alcohol group report more details a week later?

The alcohol group improved across the two sessions, while the placebo group did not show the same within-group change. Researchers discussed possible memory and retrieval explanations but did not prove a single mechanism. Both groups had already been tested once, a limitation for interpreting repeated retrieval.

Can a BAC or breath-alcohol number quantify how inaccurate someone’s testimony is?

No. The target dose in this study characterized its experimental alcohol exposure; the research did not derive a mathematical conversion from a concentration to an individual’s truthfulness, error rate, or admissibility. Blood or breath concentrations and memory-task results measure different constructs.

Does a second interview always improve the evidence?

No. A second retrieval opportunity was part of the laboratory design and may yield additional details in some circumstances. The study did not evaluate every interviewing technique, suggestion, intervening information source, or risk of memory contamination in real investigations.

Can this paper establish whether a particular witness is credible?

No. It informs the scientific questions counsel may ask about timing, alcohol exposure, recall testing, and claimed inference limits. A case-specific credibility assessment requires the actual statements, circumstances, corroborating records, and applicable rules; this single experiment does not decide it.

Primary source and related reading

Primary evidence: Butterworth B, Hand CJ, Lorimer K, Gawrylowicz J. “The impact of post-encoding alcohol consumption on episodic memory recall and remember-know responses in heavy drinkers.” Frontiers in Psychology. 2023;14:1007477. DOI: 10.3389/fpsyg.2023.1007477. PMID 36960000. The complete source is open access. This article relies on that study for its scientific findings; no unrelated references are represented as independently verified.

For the site’s approach to source appraisal, see the research library and interpretation reading paths and editorial and research source policy. A separate blood-alcohol sampling brief illustrates why a laboratory concentration question must not be confused with a behavioral or memory question.

Scope: Educational discussion of a published experimental study, not legal advice, a patient assessment, or an opinion about an identified witness.

Short answer: In a within-person study of 41 prescribed medicinal-cannabis users, a dose of their own THC-containing oil was not associated with a statistically significant change in response time on a video hazard-perception test. The same participants, however, selected slower hypothetical speeds and greater comfortable following distances after dosing. Their decisions about when to accept a gap in traffic did not change significantly. These are video-based responses and choices, not observed vehicle movements, a blood-THC impairment threshold, or proof of safe driving.

The distinction matters whenever a forensic opinion uses one passing cognitive test, a cautious answer on a driving questionnaire, or the presence of THC to draw a conclusion about a particular driver. Different measures ask different questions. In this experiment, no statistically detectable difference in hazard recognition coexisted with measurable differences in two driving-related choices. Neither finding established that the participants’ overall driving skill was normal or impaired.

This appraisal is grounded in the complete original investigation by Taren Mieran, Andrew Hill, Mark S. Horswill, Mathew J. Summers, and Kayla B. Stefanidis: The effects of orally ingested Delta-9-Tetrahydrocannabinol on drivers’ hazard perception and risk-taking behaviours: A within-subjects study of medicinal cannabis users, published online in 2025 and in Psychopharmacology 243:651–670 (2026).

What question did the researchers actually test?

The investigators asked whether established medicinal-cannabis consumers would respond differently to traffic videos after using their prescribed THC oil than at a separate baseline appointment. They looked at four behavioral outcomes: recognizing developing traffic conflicts, selecting a preferred speed relative to the camera vehicle, selecting a minimum comfortable following gap, and indicating when they would turn into a gap in oncoming traffic. They also asked participants how well they believed they performed and how safely they believed they would drive.

This design can identify within-person differences between two laboratory sessions. It is not a randomized, blinded THC-versus-placebo drug challenge. Participants knew whether they had consumed their medication, product THC and CBD compositions differed, and symptoms or other conditions might have changed between appointments. A change between sessions therefore should not automatically be treated as a pure pharmacological effect of THC.

Who participated, and how were the sessions organized?

Forty-three adult patients were recruited in Queensland, Australia; two withdrew before the second session, leaving 41 participants for the main analyses. The final group included 23 men and 18 women, with a mean age of 46 years and a range of 21–67. Each held an unrestricted driver’s license and drove at least once a week. Participants had been prescribed THC oil for a mean of 10 months and reported an average of 15 years of cannabis-use history. Twenty-eight reported chronic pain among the conditions for which cannabis was prescribed; sleep and mental-health conditions were also common. The treating conditions were self-reported, not verified from complete medical records.

One appointment was designated baseline, with no cannabis taken during that session. At the other appointment, participants took a single observed dose of their prescribed oil orally or sublingually. Sessions were approximately one week apart, order was counterbalanced, and testing occurred at approximately the same clock time for each participant. Everyone was asked to abstain from cannabis for at least 11.5 hours before either appointment. These controls are useful, but they do not make the baseline biologically THC-free.

What THC dose was given, and was blood THC measured?

Unlike a fixed-dose trial, this study allowed individual prescribed products and doses. The administered THC amount averaged 10.80 mg (standard deviation 11.95 mg; observed range 0.06–50 mg). The authors also reported a mean of 0.12 mg/kg THC, but this was calculated from reported administered amounts and body weight; it was not a laboratory determination of the THC entering circulation. Mean coadministered CBD was 16.05 mg, with wide variation and an observed range of 0–350 mg. Thirty-eight of the 41 users placed oil under the tongue, while three used the oral route.

No venous whole-blood THC concentration, serum THC, plasma THC, quantitative oral-fluid THC, or serial absorbed-dose profile was measured. Dräger DrugCheck 3000 oral-fluid screening was performed at the start of each session, but it is a qualitative screen for recent drug evidence, not a blood concentration or validated numerical impairment metric. THC screens were positive in 19/41 baseline appointments and 20/41 post-consumption appointments before the monitored dose. Fourteen participants screened positive at both sessions. The pre-dose screen cannot establish the concentration or functional effect at the later video assessment. Nor does a negative screening result guarantee no THC molecule was present.

What did each traffic-video task measure?

The hazard-perception task used 30 driving-view clips per appointment. Participants clicked the other road user associated with a developing traffic conflict as soon as they recognized it. Different, counterbalanced sets of clips were used across sessions. The primary score was an average standardized response-time measure in seconds, not brake force, steering control, distance traveled, or a vehicle collision.

The Video Speed Test displayed 16 road clips and asked how much faster or slower, in km/h, a person would choose to drive than the camera vehicle. It did not give a speedometer reading, require actual speed regulation, or record speed on a road. The Following Distance Test asked for a comfortable minimum following distance across 20 traffic clips; responses were converted into time gaps in seconds. The Gap Acceptance Test asked when a waiting driver would turn across oncoming traffic in 23 clips and scored time until the selected gap. These tasks are meaningful controlled proxies but not full driving examinations.

Table 1. What was observed, and what was not directly measured

Study component	Actual observation or measurement	Boundary on inference
Prescribed THC/CBD oil	Observed oral or sublingual use; reported product dose and composition; average THC 10.80 mg	Administered label-derived dose, not absorbed THC or venous blood concentration
Pre-session oral-fluid testing	DrugCheck 3000 screening: THC positive 19/41 baseline and 20/41 post-consumption, before dosing	Qualitative evidence, not quantitative impairment or session-at-testing blood THC
Hazard recognition	Computer mouse response time to developing conflicts in traffic videos, n=41	Not actual braking, collision avoidance, or complete on-road skill
Chosen speed and following gap	Responses to video scenarios in km/h relative to a camera car and preferred following-time gaps, n=40	Expressed choices, not instrumented vehicle speed or distance
Gap acceptance	Time to select a gap for a proposed turn in traffic videos, n=41	Not an observed physical turn or intersection outcome
Self-assessment	Visual analogue self-ratings and a separate drug-effects questionnaire	Subjective appraisal, not chemical analysis or proof of individual driving safety

Did oral THC oil slow the hazard-perception test?

No statistically significant group change was detected. Mean hazard-response time was 6.233 seconds at baseline (SD 1.717) and 6.144 seconds after oil consumption (SD 1.735), with a paired-test p=.644 in 41 participants. These are the study’s standardized scores. The slight numerical decrease is descriptive and not reliable evidence that THC improved hazard recognition. The authors’ bias-corrected bootstrap interval for the within-person contrast encompassed no difference.

The investigators also divided participants by a median split of administered THC per kilogram. That exploratory model found no significant condition effect (F[1,39]=0.188, p=.667), dose-group main effect (F[1,39]=1.784, p=.189), or condition-by-dose-group interaction (F[1,39]=0.027, p=.871). It does not follow that every person at every THC dose has unchanged hazard recognition. The sample was modest, doses were heterogeneous, and there was only one post-consumption measurement for this outcome.

Did people select slower speeds after taking their medication?

Yes, on the video-choice measure. Average preferred-speed response shifted from −1.058 km/h relative to the depicted camera vehicle at baseline (SD 4.447) to −3.302 km/h post-consumption (SD 5.106), a descriptive difference of −2.244 km/h. The paired result was p<.001 in 40 complete datasets and remained significant after the study’s familywise multiple-comparison correction. The reported Cohen’s d was 0.73.

The participants therefore indicated that they would choose to drive more slowly than the camera vehicle under the post-consumption condition. The study did not show that their actual road speed fell by precisely 2.244 km/h, that everyone made the same choice, or that taking THC reduced crash probability. A choice to slow down could reflect caution, expectations, symptom changes, familiarity with the assessment, or some combination; the design cannot isolate the motive.

Did participants choose more following distance?

Yes, for the video-based preferred gap. Mean minimum comfortable following distance rose from 1.537 seconds (SD 0.188) at baseline to 1.645 seconds (SD 0.325) post-consumption, a mean difference of 0.108 seconds. The paired test gave p=.019 in 40 complete participants, with reported Cohen’s d=0.50; the result survived the authors’ Holm adjustment for the related video-choice comparisons.

These were selected gaps on a screen, not time headways calculated from a participant’s instrumented vehicle. The observed increase also does not show that other performance dimensions were compensated to a level of adequate driving safety. In actual traffic, appropriate distance depends on roadway, speed, visibility, reaction capability, vehicle and circumstances; this experiment was not designed to prescribe an individualized safe time gap.

Was turning into a traffic gap also affected?

The study found no statistically significant change on its gap-acceptance task. Mean response or waiting time was 14.935 seconds (SD 6.837) at baseline and 14.580 seconds (SD 5.780) after oil administration, with p=.534 across 41 observations. This result is distinct from the speed and following-distance choices. It also provides no proof that a particular left-turn decision was safe, unsafe, or unchanged after cannabis use. A non-significant result is not an equivalence test.

Table 2. Original paired task findings in the 41-patient study

Outcome	Baseline mean (SD)	Post-oil mean (SD)	n	Paired p and interpretation
Hazard-perception response time, study seconds	6.233 (1.717)	6.144 (1.735)	41	.644; no significant difference
Preferred speed relative to camera car, km/h	−1.058 (4.447)	−3.302 (5.106)	40	<.001; slower selected speed
Preferred minimum following gap, seconds	1.537 (0.188)	1.645 (0.325)	40	.019; longer selected gap
Gap-acceptance response time, seconds	14.935 (6.837)	14.580 (5.780)	41	.534; no significant difference
Self-rated performance on hazard-perception task, 0–100	76.93 (14.01)	71.45 (15.18)	40	.031; lower self-rating

The self-rating outcome belongs to a separate appraisal question; it should not be treated as another objectively measured driving outcome. The significant speed and following-distance results survived the specified correction for the related risk-choice tests. The parenthesized spreads are standard deviations, not standard errors.

How did objective hazard detection compare with perceived performance?

After completing the hazard-perception videos, participants judged their own performance against other drivers on a 0–100 scale. Their mean rating fell from 76.93 (SD 14.01) at baseline to 71.45 (SD 15.18) after oil, a significant paired difference at p=.031 in 40 participants. Yet the group hazard-detection response time itself did not significantly worsen. This contrast demonstrates that subjective confidence and measured performance can move differently within one experiment.

Neither session demonstrated a statistically significant Pearson correlation between the self-rating and the measured hazard-perception response time: baseline r=.21, p=.190 and post-consumption r=.17, p=.296, as reported in the paper. A nonsignificant correlation is not proof that everyone misjudged their ability or that the true population correlation is exactly zero. It does limit attempts to substitute one person’s confidence for an independently validated objective measurement.

Did participants believe they were safer or less safe on public roads?

Four additional ratings asked how participants imagined their current ability to predict traffic conflicts, drive skillfully, drive safely, and manage crash risk relative to other local drivers. Some unadjusted comparisons produced nominal p values below .05, including perceived traffic-conflict prediction (.022) and perceived driving safety (.042). None of these four comparisons remained significant after the authors applied the Holm correction, so they should not be promoted to four independent positive effects.

The authors also collected a separate 17-item subjective drug-effects questionnaire. At approximately 85 minutes after dosing, mean rated drug effect was 27.66/100 (SD 25.53) and self-rated high was 20.17/100 (SD 25.68). Those numbers describe reported experiences. They are not a blood concentration, a known impairment percentage, or an observed hazard-response deficit. A person may acknowledge feeling drug effects without any one test settling overall functional status.

Why do the positive and null findings need to be interpreted together?

The study did not produce a single all-purpose ‘driving impairment’ score. It observed a null difference in hazard recognition, two positive changes in expressed speed and gap preferences, a null gap-acceptance result, and lower self-appraisal of hazard recognition. These outcomes are not logically contradictory: detecting a conflict and deciding how fast to go are distinguishable processes, with separate operational definitions and response methods.

One positive result cannot automatically be carried across other dimensions, and a null result cannot automatically erase a different positive finding. Nor do slower stated speeds and wider stated gaps prove that every underlying THC-related deficit, if present, has been fully corrected. The word compensation describes one possible interpretation of the choices; it is not the experimentally identified psychological cause or a validated measure of risk offset.

What are the major confounders and generalizability limits?

These 41 individuals were established, licensed medicinal-cannabis patients, not randomized occasional users or a representative roadside population. The repeated-session design controls some stable between-person characteristics, but the sessions differed in medication exposure, symptoms, expectations, potentially sleep and other day-specific factors. There was no blinded placebo oil and no fixed pure-THC dose across patients. A participant’s medication could relieve pain or anxiety while introducing other effects, and this study did not isolate those pathways.

Almost half screened oral-fluid THC-positive before each session despite the requested abstinence interval. Some participants had concurrent medical conditions or prescribed medications, and one showed positive screens for methamphetamine and cocaine; excluding that participant in a sensitivity analysis did not change the study’s t-test conclusions. Self-reported THC use, oil content and medication history are not equivalent to verified blood concentration and measured pharmacokinetics. The chosen 90-minute wait, together with approximately 20 minutes of preliminary neuropsychological testing, yielded only one principal post-dose video-test occasion. A time course, peak-effect verification, dose-response curve and individualized impairment threshold were not established.

How should lawyers and toxicologists use this evidence?

Begin by identifying the proposed evidentiary proposition. If the proposition concerns chemical exposure, this paper reports oral-fluid screening and observed prescribed-oil administration, not numerical blood THC. If it concerns how quickly someone recognizes a hazard in videos, use the hazard-perception result and its uncertainty. If it concerns stated speed and following choices, use the separate video proxy measurements. If it concerns actual lane keeping, a collision or a particular driver’s fitness, acknowledge that those endpoints were not measured in this trial.

A case-specific analysis would require independent records appropriate to the actual question: the product and dosing history, relevant specimen and collection time, screening versus confirmation methods, concurrent substances, observed driving or task performance, and plausible non-drug explanations. No inference about a defendant’s driving intent, a witness’s legal capacity, or an individual’s accident causation should be attributed to this group study alone.

For related but separate study designs, the published THC potency and driving research brief discusses a randomized smoked-cannabis trial with simulator and blood measurements; its design and measured outcomes must not be conflated with this oral-oil/video-choice study. The published research library organizes specimen, analytical and inference questions, while the source and editorial policy explains the limits of literature-based briefs.

Frequently asked questions

1. Did the study prove that prescribed THC oil does not impair driving?

No. It found no statistically significant difference on one video hazard-perception response-time task. It did not conduct an actual on-road safety trial, measure collisions, or establish equivalence of overall driving performance before and after medication.

2. Were the measured speed changes actual reductions in vehicle speed?

No. The −1.058 to −3.302 km/h shift was a change in participants’ hypothetical preferred speed relative to a vehicle shown in 16 videos. No participant had to operate a real vehicle in this measure.

3. Was 10.80 mg the THC concentration found in participants’ blood?

No. It was the mean administered amount of THC in prescribed oil, in milligrams. The study did not measure blood THC concentration, which would be reported in a concentration unit such as ng/mL, or quantify absorbed THC.

4. Why were oral-fluid THC screens positive during the no-dose session?

Participants were regular users who reported at least 11.5 hours of abstinence before testing. A qualitative pre-session THC-positive screen does not identify an exact last-use time or the quantitative drug concentration during later task performance.

5. Does a longer preferred following gap mean THC made these drivers safer?

Not by itself. A preferred time gap increased by 0.108 seconds on average in the study’s video-choice task. Whether this choice translated to actual safe headway or counteracted any other impairment was not evaluated.

6. Does the nonsignificant hazard response result show that every participant was unimpaired?

No. Statistical nonsignificance at the group level does not validate individual normal performance, clinical equivalence, or impairment cutoffs. It also does not rule out effects at other doses, times or traffic demands.

7. Can self-rated sobriety, confidence, or perceived skill replace objective testing?

No. Participants lowered their average self-assessment despite no significant group change in hazard-response time, and self-assessment was not significantly correlated with the task in either session. These are population findings about particular scales, not a validated individual credibility or impairment test.

Primary research and interpretive boundary

Original study: Mieran T, Hill A, Horswill MS, Summers MJ, Stefanidis KB. The effects of orally ingested Delta-9-Tetrahydrocannabinol on drivers’ hazard perception and risk-taking behaviours: A within-subjects study of medicinal cannabis users. Psychopharmacology. 2026;243:651–670. DOI: 10.1007/s00213-025-06869-w. The interpretation above is limited to the original Methods, Results, tables and Discussion and should not be mistaken for an individualized legal, toxicological or medical finding.

Short answer: In a 2026 observational comparison of 154 adults, frequent cannabis users tested 45 minutes after smoking scored lower on a task requiring them to reproduce a sequence of spatial locations in reverse order. The difference from same-sex nonusing controls was larger among men than women. Three other span tests did not show significant group differences. But the investigators did not assay THC in blood, breath, oral fluid, or urine, did not test the same participants before and after cannabis, and did not measure driving. The results describe group differences in a particular cognitive task; they do not prove the pharmacologic dose received, establish an individual’s impairment, or demonstrate why a particular incident happened.

The source is Ryder and colleagues’ 2026 original study, “Same dose, different impact: acute cannabis intoxication impairs visuospatial working memory in both sexes, with disproportionate male vulnerability”, published in Frontiers in Behavioral Neuroscience. The question for forensic interpretation is not whether a statistically significant effect exists: it does. It is what comparison produced that effect, how exposure was characterized, what was actually measured, and what additional evidence would be necessary to apply the result to one person.

What was actually tested?

The researchers recruited 154 community participants: 77 who reported using cannabis regularly and 77 nonusing comparison participants. Regular use was defined for eligibility as at least five days a week over at least a year; nonusers reported fewer than five lifetime uses and none during the preceding year. The users smoked approximately one gram of their usual medical-grade cannabis product and began testing 45 minutes later. Controls were tested without a corresponding cannabis or placebo administration. This was a between-subjects, cross-sectional comparison, not a randomized trial assigning regular users to THC or placebo, and not a within-person evaluation of change from a sober baseline.

The four tasks crossed two distinctions: auditory-verbal versus visuospatial material, and forward repetition versus backward manipulation. Repeating a series in its original order chiefly measures short-term maintenance. Reversing it adds mental manipulation. The researchers recorded the number of correctly recalled sequences; they did not measure response latency. Thus, an unchanged forward score is not proof that every aspect of attention or reaction speed was unaffected.

Who was in each comparison?

The cannabis group consisted of 46 men and 31 women; controls consisted of 32 men and 45 women. The sex distributions were therefore not identical across exposure groups, even though the total sample comprised 78 men and 76 women. Mean ages in the four groups were similar, and the investigators reported no statistically significant four-group age difference, F(3,150) = 0.45, p = 0.715. Most participants were young adults; the overall reported mean age was 24.58 years (SD 5.73), with eligibility extending from 18 to 50 years.

The study relied on community recruitment through convenience and snowball sampling. That is important for generalizability: people reached through those channels, with frequent cannabis histories and willingness to participate in cognitive testing, cannot be assumed to represent all occasional users, patients, workers, drivers, or people undergoing a forensic investigation. Nor was sex itself randomized or experimentally manipulated.

Was “the same dose” a measured blood THC exposure?

No. Participants were asked to smoke one approximately one-gram cigarette from a product reported to contain 16.1% delta-9 THC and less than 1% CBD. This identifies product mass and labeled/laboratory-tested composition, not each participant’s measured absorbed THC dose. The protocol did not standardize inhalation depth, puff duration, breath-holding, smoke loss, or individual systemic bioavailability. Testing at a common 45-minute interval likewise does not make internal exposures equal.

Most importantly, the authors expressly state that no biological samples were collected to verify drug exposure or nonuse. There was no venous whole-blood THC concentration, plasma THC curve, oral-fluid THC result, urine screen, or breath THC analysis. Consequently the paper contains no concentration-performance regression, no blood-THC threshold, no analytic limit of detection or quantitation for THC, and no specimen-based individual timing estimate. Referring to its results as proof of a particular blood THC level would add measurements the investigators never obtained.

In toxicology, the distinction is basic but consequential: the THC percentage of plant material is a product characteristic; smoked mass is an administration-related estimate; circulating THC is an analytical result from a named specimen; and cognitive performance is a separate functional outcome. None can be silently substituted for another.

Which of the four memory tasks differed?

The statistically significant pattern centered on Spatial Span Backward, the task of reproducing a series of locations in reverse order. Digit Span Forward, Digit Span Backward, and Spatial Span Forward did not show significant main effects or group-by-sex interactions in the reported primary tests. These three null findings should be preserved, rather than presenting the paper as evidence of generalized memory failure after cannabis.

The following table reproduces the reported descriptive means and standard deviations from the original Table 2. Values are task scores, not percentages of persons impaired and not THC concentrations. Higher values mean more correct sequences.

Observed memory-task scores by group and sex (mean [SD]; original Table 2)

Task	Men using cannabis (n=46)	Male controls (n=32)	Women using cannabis (n=31)	Female controls (n=45)
Digit Span Forward	8.80 (2.05)	9.16 (2.55)	8.71 (2.33)	8.07 (2.12)
Digit Span Backward	5.80 (2.21)	6.75 (2.52)	5.65 (2.47)	5.53 (2.44)
Spatial Span Forward	6.76 (1.73)	7.41 (1.95)	7.52 (1.67)	7.67 (1.81)
Spatial Span Backward	5.60 (1.77)	7.15 (1.76)	6.32 (1.40)	7.10 (1.95)

It would also be incorrect to call the three null tasks proven unaffected. A nonsignificant result says the study did not detect a difference under the particular test, sample, and analysis. It does not establish equivalence at every effect size, in every population, or in unmeasured facets such as processing speed. These tests scored accuracy rather than the time needed to respond.

What did the reported group-by-sex interaction mean?

The primary Spatial Span Backward analysis yielded a group main effect, F(1,150) = 18.84, p < 0.001, partial eta-squared 0.112, and a group-by-sex interaction, F(1,150) = 9.74, p < 0.01, partial eta-squared 0.061. The reported main sex effect was F(1,150) = 4.13, p < 0.05, partial eta-squared 0.027. The authors applied a Bonferroni-adjusted threshold of 0.0125 across the four primary ANOVAs; they report that the group effect and group-by-sex interaction remained significant under that threshold.

Compared with male controls, the men in the cannabis group averaged 5.60 rather than 7.15, a 1.55-sequence descriptive gap. Compared with female controls, women in the cannabis group averaged 6.32 rather than 7.10, a 0.78-sequence descriptive gap. The published standardized between-group effects were Cohen’s d = −0.87 for men (95% CI −1.34 to −0.41) and d = −0.48 for women (95% CI −0.94 to −0.02). The male and female comparisons were both reported statistically significant, with the male comparison p < 0.001 and the female comparison p < 0.05.

This interaction addresses the difference between user and nonuser groups within each sex. It does not mean every man had a larger deficit than every woman. It does not establish what the same man or woman would have scored before smoking. And it does not show that being male biologically caused susceptibility: the comparison is observational, with exposure history and other potential differences tied to group membership.

What did adjustment for alcohol and nicotine change?

The investigators also modeled Spatial Span Backward with self-reported weekly alcohol and nicotine use frequency as covariates. Neither frequency was a statistically significant predictor in that reported sensitivity analysis: alcohol F(1,148) = 1.12, p = 0.291; nicotine F(1,148) = 0.85, p = 0.358. The group effect remained F(1,148) = 17.56, p < 0.001, and the group-by-sex interaction F(1,148) = 8.23, p = 0.005, partial eta-squared 0.053.

These adjusted findings make it harder to attribute the association solely to the two recorded frequency variables. They do not remove the principal design limitations. Frequency of alcohol and nicotine use is self-report, not a contemporaneous blood or breath toxicology test. The researchers did not directly measure THC dose, confirm control-group abstinence, assess pre-exposure cognitive scores, or comprehensively adjust for unmeasured baseline intelligence, education-related performance, sleep, motivation, psychiatric screening error, or other correlates of habitual use.

What did the study measure directly—and what did it not?

A forensic report benefits from a measurement inventory before any legal characterization. Here the direct study endpoints were the correct-sequence counts on standardized memory tasks. The product’s nominal THC percentage and the approximate cigarette mass characterized the administered material, while participant questionnaires characterized use history. Those are different evidentiary categories.

Evidence inventory: measured task, exposure description, and unsupported inference

Item	What the study actually supplied	What it does not establish
Spatial working-memory performance	Number of correct reversed spatial sequences, compared across four sex-by-use groups	Individual driving fitness, workplace failure, crash responsibility, or universal performance cutoff
Other memory performance	Three additional forward/backward span score distributions, no significant group effects	Equivalence of all mental abilities or absence of any THC-related behavioral effect
Cannabis product	Approximately 1 g smoked material; product tested at 16.1% THC and <1% CBD	Mass of THC absorbed by each participant or concentration at the brain
Testing time	Tasks started 45 minutes after reported smoking in the cannabis group	Individual peak blood concentration, time-to-recovery, or earlier-event concentration
Biological toxicology	No THC blood, plasma, saliva, or urine samples	A measured THC concentration, validated cutoff, analytic uncertainty, or drug-test interpretation
Sex-related association	Group-by-sex interaction and comparisons against separate same-sex controls	Person-specific biologic susceptibility, hormone mechanism, or causal sex-specific dose-response

Why a statistically significant result is not an individual impairment test

A conventional statistical test asks whether an observed association is unlikely under a specified null model. It does not by itself classify an individual. The original researchers did not derive or externally validate a threshold separating impaired from unimpaired drivers. There is no sensitivity, specificity, false-positive rate, false-negative rate, positive predictive value, or individual prediction interval for a proposed forensic determination.

Two people with the same task score may differ in premorbid ability, education, sleep, task familiarity, underlying illness, medications, use history, anxiety, or effort. In a real case, the relevant question is not merely whether the group mean changed; it is whether the scientific method has been validated to connect a person’s observations to the claimed conclusion with known limitations. A sex-stratified Cohen’s d cannot legitimately be converted into a probability that an identified person drove impaired.

Why the lack of a pre-smoking baseline matters

The study tested frequent users after cannabis and different individuals who did not use cannabis. It did not obtain each user’s sober Spatial Span Backward score on a separate matched day, and it did not use a placebo-controlled crossover. The lower average score among users could include an acute drug component, longstanding differences associated with frequent use, pre-existing group differences, or combinations of these. The authors explicitly acknowledge that acute and chronic effects could not be disentangled.

This problem persists even when a drug-related hypothesis is biologically plausible and a statistical interaction survives adjustment for two self-reported covariates. A controlled within-person design can remove many stable between-person differences; this investigation did not provide that counterfactual. Its positive finding is an observational association under recent smoking, not a clean estimate of the effect of administering THC to the same individual.

Does a 16.1% cannabis product define a comparable dose for everyone?

Not in the pharmacokinetic sense. A product labeled or assayed at 16.1% THC expresses the proportion of THC in the plant material, while approximately one gram describes material selected for smoking. Systemic delivery depends on what actually burned or was lost, the smoke inhaled, smoking technique, absorption, distribution, metabolism, and timing. None of those individual exposure variables was quantified. One cigarette does not yield one uniform venous whole-blood THC result.

The paper’s phrase “same dose” therefore should be read in the narrow procedural sense of using approximately the same amount and product specification. It is not confirmation that men and women absorbed equivalent THC amounts. Nor can the absence of a blood assay be corrected by substituting the product percentage or applying a theoretical blood-THC conversion. Such a calculation would require assumptions unsupported by these individual data.

Do the negative memory tests show tolerance?

The study’s discussion considers whether experienced users might develop different degrees of tolerance across cognitive systems. That is a potential explanation, not a measured physiological outcome. There was no repeated individual drug challenge over time to quantify tolerance, no first-time-user comparison arm, and no receptor imaging. Alternatively, the three null tasks may not have been sensitive enough to detect smaller effects. Both possibilities are consistent with the reported observations.

Likewise, the paper proposes fronto-parietal and hormonal mechanisms to explain the spatial and sex-related findings. No participant underwent neuroimaging or had estradiol, CB1 receptor availability, or relevant enzyme activity assayed for the study’s principal comparison. These mechanisms belong to interpretation and hypotheses for future testing, not direct measured findings from this sample.

What is relevant to DUI and other litigation?

Visuospatial working memory has plausible relevance to navigation, maintaining spatial relationships, and mentally reversing or updating a sequence. Yet this study did not ask participants to drive on a road or simulator, avoid hazards, respond under time pressure, perform standardized field sobriety tests, or recreate a collision. An association in one board-based span task is not a calibrated estimate of lane deviation, reaction time, collision risk, intent, capacity to consent, or truthfulness.

A careful analysis of a driving or workplace matter would begin with the person’s actual acts and recorded times, any contemporaneous observations, the specific specimen and confirmed analytical method if a sample exists, reported dose or product evidence, co-exposures, clinical conditions, and alternative explanations. This paper can provide context for why one cognitive domain deserves attention. It cannot supply missing evidence from the event or support a case-specific expert conclusion solely from group membership, product label, or sex.

For a contrasting randomized THC potency and driving-simulator experiment, consult the separate same-site research brief; its controlled design, whole-blood assays, and driving endpoints differ from the current observational memory comparison. Neither study alone creates a universal THC impairment threshold.

How should the four-task pattern be reported fairly?

A scientifically balanced statement names both the positive result and the three negative results, then qualifies the population and design. One defensible summary is that, among 77 frequent cannabis users tested after reported smoking and 77 separate nonusing controls, backward spatial span was significantly lower for the users, with a larger user-versus-control difference in men, while three other span scores were not significantly different. It must also state that no toxicology concentration or driving performance was measured.

Overstatements run in both directions. “Cannabis did not affect memory” suppresses the positive spatial result. “Cannabis impairs every cognitive function, especially in every man” ignores the three null domains, distributions, observational confounding, and variability between individuals. “The same THC dose impaired men more” mistakes standardized material for measured absorbed exposure and a group interaction for a controlled pharmacokinetic comparison.

Frequently asked questions

Did the researchers measure THC in blood?

No. The original article states that biological drug verification was not performed. There are no individual or group venous whole-blood THC concentrations to compare with the memory scores, and no blood concentration cutoff can be inferred from this study.

Were participants randomized to THC or placebo?

No. Frequent users smoked their usual standardized-source product before testing; different people formed the nonusing comparison group. The data therefore combine recent reported smoking with differences associated with the histories of the two groups.

Did men always perform worse than women?

No. The statistically significant interaction describes larger male-user versus male-control differences than female-user versus female-control differences. It does not compare every individual man with every individual woman, or demonstrate a universally higher male susceptibility.

Which cognitive task showed the clearest difference?

Spatial Span Backward, which requires recalling a tapped sequence of locations in reverse order. The investigators did not report significant group effects on Digit Span Forward, Digit Span Backward, or Spatial Span Forward.

Does one gram of 16.1% THC cannabis mean everyone absorbed the same THC dose?

No. Those figures characterize the material and approximate smoking amount. The absorbed dose and blood concentration were not measured, and smoking behaviors were not tightly standardized.

Can the findings prove a defendant or driver was impaired at a specific time?

No. The paper provides no validated individual classification threshold, event-time blood measurement, driving test, or causal reconstruction. It is evidence about average test performance under a selected study design, not a direct finding about an identified person.

Do three nonsignificant span tests prove those skills were unaffected?

No. They indicate that the study did not detect statistically significant differences with those measures, sample sizes, and models. Smaller or different impairments, reaction-time changes, or effects in other populations were not excluded.

Source and further reading

Original primary research: Ryder CH, Gal C, Levy E, et al. Same dose, different impact: acute cannabis intoxication impairs visuospatial working memory in both sexes, with disproportionate male vulnerability. Frontiers in Behavioral Neuroscience. 2026;20:1785335. doi:10.3389/fnbeh.2026.1785335. The study’s Methods, original Tables 1–3, Results, and Discussion govern the numerical interpretation above.

For methods and other research questions, use the Attorney Scientists research library. For information on sourcing, AI-assisted editorial preparation, and review boundaries, consult the editorial and research source policy. This material is an educational appraisal of one original study, not an individual medical assessment, legal opinion, or certification that a particular driver was impaired.

Answer: No. In a 2014 validation program, a computerized cognitive instrument called MedDrive detected a statistically significant mean response to an alcohol challenge in 20 younger adults. Yet in a separate sample of 182 older licensed drivers, its composite score accounted for only 5.3% of observed differences in on-road driving ratings. A test can be responsive to an experimentally administered substance without being accurate enough to decide whether one particular person was unsafe to drive. Neither result is a forensic blood-alcohol measurement, an individual diagnostic threshold, or evidence that alcohol caused a particular collision.

The full original primary study is Vaucher and colleagues, A neuropsychological instrument measuring age-related cerebral decline in older drivers: development, reliability, and validity of MedDrive (2014), DOI: 10.3389/fnhum.2014.00772. Its unusually useful feature for evidence evaluation is that it investigated several different measurement questions rather than treating a single significant laboratory result as proof of real-world performance.

What did MedDrive actually measure?

MedDrive was developed as a computer-based research instrument for aspects of visual and cognitive processing associated with aging. Four tasks assessed visual recognition and processing, responses to an orientation cue, detection of movement accompanied by attention shifts, and spatial working memory. The researchers transformed multiple task measures into a composite score on a nominal zero-to-100 scale, with a higher score representing better performance on those selected tasks.

The output was a performance score, not a physical assay of alcohol. It was also not a direct electrical recording of neuronal firing or an imaging measure of brain tissue damage. Calling the proposed latent construct “cerebral decline” did not convert a statistical factor derived from task behavior into direct proof of cerebral pathology in an individual. The instrument measured how quickly and accurately people completed specified computer activities under specified viewing conditions.

The distinction is legally consequential. A person’s detection time on a screen may be associated with driving behavior, but real driving additionally involves motor habits, judgment, situational awareness, compensation, fatigue, traffic complexity, vehicle conditions, and environmental hazards. One score cannot silently stand in for those unmeasured elements.

Why does the paper contain several different participant numbers?

The work comprised a development program and several validation studies. Study 1 derived scoring from 106 older drivers attending a driving-refresher program. Study 2 evaluated a separate sample of 182 older drivers against a structured on-road assessment. Study 3 examined test–retest reliability in 17 younger volunteers. Study 4 investigated psychophysical relationships in 47 older drivers. Study 5 compared alcohol and placebo conditions in 20 younger volunteers on MedDrive and a driving simulator.

Those denominators answer different questions. The 20 people exposed to ethanol were not the 182 older people whose ordinary on-road performance was assessed. The 17-person reliability sample also overlapped with volunteers in the later alcohol study; those participants should not be added to the 20 as though they represented another independent alcohol trial. The development and cross-sectional samples likewise do not establish an independent external validation of alcohol-related impairment among older drivers.

Study within the original article	Participants and setting	Principal measurement question
Study 1: derive the composite	106 older licensed drivers	Which computerized processing tasks contribute to a candidate score?
Study 2: validate on roads	182 older volunteer drivers	How closely does the score track a blinded instructor’s real-road assessment?
Study 3: reliability	17 younger volunteers, five repeat sessions	How reproducible is the instrument and how much does practice matter?
Study 4: psychophysics	47 older drivers	Which separate sensory and attention tasks correlate with its components?
Study 5: alcohol challenge	20 younger people, repeated placebo/alcohol conditions	Does acute ethanol alter MedDrive scores and artificial driving tasks?

Combining samples into one large pooled “alcohol and driving” population would falsely inflate the strength of the evidence. The older-driver observations concern age and ordinary driving; the younger-driver crossover concerns acute alcohol, computerized performance and simulated driving.

How well did MedDrive correspond to actual on-road performance in older drivers?

In Study 2, 182 older licensed drivers participated in a research program linked to voluntary driver-refresher courses in Switzerland. A driving instructor, unaware of the participant’s cognitive-test result, evaluated each driver during an approximately 45-minute road route. The assessment covered such matters as road rules, lane position, speed adaptation, observation, vehicle handling, anticipation and behavior toward other users. It was a real on-road assessment, not a crash-history comparison or a test of drunk driving.

MedDrive’s score was statistically related to the instructor-rated road-performance categories: the reported coefficient of determination was R² = 0.053, p = 0.002. Numerically, that is roughly 5.3% of shared variation in the study’s statistical model, not a 5.3% probability of impairment or a 94.7% error rate. It signifies a detectable but weak relationship. The same instrument showed R² = 0.061 with the older participants’ self-reported driving self-restrictions. A meaningful association is not the same as useful classification for a particular defendant or licensing decision.

Selection also matters. These participants were community-dwelling, generally licensed drivers who volunteered through a refresher course; people with important known neurological diagnoses were excluded. It would be unsound to extend the observed model, without further validation, to drivers recovering from trauma, affected by other drugs, or evaluated in a police roadside encounter.

What happened when researchers tried to classify older drivers using a cutoff?

A MedDrive composite threshold below 25 points provided a practical stress test of whether a group association could be used for an individual decision. In the authors’ analysis, a low score corresponded to a threefold estimated relative risk of poor on-road performance, but the confidence interval was broad (0.90–10.2), and the result did not reach conventional statistical significance (p = 0.070).

More importantly, the investigators reported that such a cutoff would falsely label about five of every six flagged drivers as unfit and detect only one of every four of the poorly performing drivers. Their analysis of a shorter movement-detection-only alternative did not solve the classification problem: roughly nine of ten drivers flagged by its suggested threshold would have been false positives in that sample.

These are study-specific operating characteristics rather than universal sensitivity or specificity for a forensic test. Nevertheless, they illustrate the difference between a continuous group association and a tool suitable for depriving a particular person of a driving privilege. A seemingly strong relative-risk number is inadequate when the uncertainty is wide and the practical false-positive burden is high.

How was alcohol administered in the 20-person experimental study?

Study 5 enrolled 20 healthy adults, aged 23–40 years, including 11 women, in a randomized, double-blind, four-condition crossover experiment. Each person completed a placebo condition and three alcohol conditions designed to achieve nominal concentrations of 0.5, 0.65 and 0.8 grams of ethanol per liter. Participants consumed alcohol mixed with cranberry juice, with initial amounts calculated from the Widmark formula. During the sessions, a second investigator used calibrated breath-testing devices and additional drinks at approximately 15-minute intervals to adjust and sustain the target condition.

Measurement boundary: The original authors use the term “blood alcohol concentration” throughout, but their described session monitoring employed exhaled-breath instruments to generate the reported concentration equivalents. The study did not report independently collecting each participant’s forensic venous whole-blood specimens and measuring ethanol by headspace gas chromatography. Its reported 0.5 g/L condition should therefore not be recast as a directly assayed whole-blood result for a legal case.

Actual reported session means were approximately 0.49, 0.64 and 0.78 g/L for the three nominal alcohol targets, respectively. The authors gave an observed standard deviation of 0.04 g/L from intended levels. Achieving those levels required, on average, 69% more administered ethanol than the initial Widmark calculation anticipated. That finding is a caution about a dose-to-concentration calculation, not a rule that any particular person’s Widmark estimate should always be raised by 69%.

Blinding was imperfect: participants identified their assigned condition better than chance (reported κ = 0.383, 95% CI 0.255–0.511). Taste and perceived effects were among the cues. Randomized conditions strengthen interpretation, but better-than-chance condition guessing remains a potential source of expectancy effects.

What change in the computer score was detected at the lowest alcohol condition?

The overall MedDrive score showed a dose-dependent response to alcohol (reported R² = 0.366, p < 0.001). At the condition targeting 0.5 g/L, the mean difference from placebo was a reduction of 5.3 MedDrive points (p = 0.008). Fifteen of the 20 participants had a lower score in that condition. This supports a group-average laboratory response and also documents individual variation: not all 20 people showed the same direction of change.

At the lower exposure, visual processing and orienting-related response tasks showed statistically detectable effects, whereas movement detection and spatial memory did not. At the highest intended condition, 0.8 g/L, the source reported significant changes across all four task categories. Consequently, it would be incorrect to describe a universal deficit in every cognitive process at 0.5 g/L merely because the composite moved.

The difference is not itself a driving-error count. A five-point composite change is a scale-specific difference on a derived score; it cannot be converted into a fixed number of lane departures, missed hazards, or seconds of braking delay unless an independently validated model establishes those conversions for the relevant population and context.

What did the same 20 people do in the driving simulator?

The young volunteers completed a computer-driven vehicle simulation that included a seven-minute straight-road tracking segment, a variable-speed lead-car-following task, and an added peripheral-signal-response condition. The simulation used a vehicle mock-up with normal controls and three screens covering roughly a 200-degree view. In the tracking condition, the participants were instructed to maintain 120 km/h; following conditions involved lead-car speeds varying roughly from 90 to 110 km/h.

The researchers analyzed the standard deviation of lateral position (SDLP), a measure of how widely the simulated vehicle wandered within its lane. They reported an increase in mean SDLP of 8.7 centimeters per g/L in a statistical model of the alcohol conditions (p < 0.001). Alcohol exposure explained 52.9% of differences in mean SDLP between experimental sessions under that model. These numbers do not mean that 52.9% of crashes were caused by alcohol or that a person with a particular MedDrive score has a 52.9% chance of departing a lane.

The investigators also found that the MedDrive composite could statistically model 36.6% of the alcohol-associated variation in their simulator endpoint (p < 0.001). A movement-detection subtest yielded a smaller reported relationship, R² = 0.178 (p = 0.027). The simulated driving measurements are observed functional outputs within an artificial environment. They are closer to driving than a reaction-time questionnaire, but they remain different from natural traffic and do not identify the cause of any real collision.

Two validation questions, two very different evidentiary answers

Original research comparison	Source-grounded result	What the result cannot establish
Older-driver on-road validation, n = 182	MedDrive versus rated road performance R² = 0.053; p = 0.002	No reliable yes/no decision that a named older driver is fit or unfit
Older-driver cutoff below 25	Estimated relative risk 3.0; 95% CI 0.90–10.2; p = 0.070; about five of six flagged were false positives	No validated universal legal screening cutoff
Younger experimental alcohol crossover, n = 20	At nominal 0.5 g/L, mean MedDrive difference −5.3 points; p = 0.008; decrease in 15/20	No directly assayed forensic blood concentration or universal person-level effect
Same younger participants’ simulator	SDLP slope +8.7 cm per g/L; p < 0.001; MedDrive-associated model 36.6%	No observed on-road DUI violation, collision cause or crash probability

These comparisons are not competing estimates of the same endpoint. The older cohort supplied on-road ratings without experimental alcohol; the younger cohort supplied alcohol-sensitive computerized and simulated outcomes without observed ordinary-road driving in traffic. Neither sample alone closes the inferential gap between an alcohol concentration, an observed task deficit and the causal circumstances of a particular event.

Why does test reliability not settle driving validity?

Study 3 asked whether MedDrive provided relatively stable differences between individuals on repeated administrations. Seventeen younger volunteers completed the instrument five times. The researchers reported a composite intraclass correlation of about 0.852, and about 0.879 for its movement-detection component. Those were encouraging repeatability findings in a small healthy sample.

Repeatability is necessary for many measurement purposes, but it is not the same property as prediction of actual driving ability. A bathroom scale can give a consistent reading without measuring blood alcohol; likewise a cognitive test may be consistent without diagnosing unsafe driving. In this case, the original publication itself found a relatively strong repeated-score metric alongside a very weak older-driver on-road performance association.

Practice is another source of systematic change. From the second administration onward, mean MedDrive scores improved by about 1.3 points per subsequent session (95% CI 0.8–1.8). The research team therefore pre-trained participants before the alcohol crossover. When interpreting a one-time field or clinical score, a reader must ask whether the individual was familiar with the computer interface, how the instructions were provided, and whether sensory or language barriers affected performance.

What did the psychophysical study add?

Forty-seven older drivers underwent additional visual, attentional and executive-function testing. MedDrive’s visual components were associated with several psychophysical measures, and the overall composite had a significant association with visual search (R² = 0.256, p = 0.001). The investigators also observed a relationship between spatial memory and visual search (R² = 0.143, p = 0.009), but not between that same spatial task and a separate backward digit-span measure (R² = 0.023, p = 0.313).

Such findings help clarify what a task may be responding to. They do not prove that the program measures every relevant cortical function or that the computational score is a direct biological marker of neural network integrity. Some observed variation can reflect the task chosen, sensory processing, time pressure and response strategy. An inference about neuronal causation would require independent neurophysiological evidence.

How should the reported dose levels and statistical percentages be read?

The three experimental concentration targets were expressed in g/L, the convention used by the original investigators. They were not quantities of beer, wine or vodka consumed; the alcohol dose was adjusted during each session to reach the targeted device-reported concentration. Nor were they contemporaneous serum, plasma or confirmed venous whole-blood laboratory measurements. The paper is therefore useful for studying how computerized performance changed across the researchers’ experimental breath-monitored conditions, but it is not an independent validation of evidential blood sampling or blood-to-breath conversion in legal casework.

R² measures the fraction of variation accounted for in a specified model and set of observations. An R² of 0.366 in the controlled younger experiment does not automatically transfer to a 75-year-old driver; R² = 0.053 against older drivers’ actual road ratings is not an error probability or a direct measure of the likelihood of intoxication. Similarly, the SDLP coefficient in centimeters per g/L is not a conversion formula for individual crash risk. The endpoints, populations and comparators must accompany every numerical result.

The null and weak findings matter at least as much as the positive ones. A useful scientific report should not quote only the p = 0.008 computer response, omit the modest natural-road relationship, and then use the selective result as a categorical impairment opinion. Those are different claims with different evidentiary burdens.

What records would be needed before an individual forensic opinion?

For a real DUI or driver-capacity case, establish the event chronology, physical specimen and time, method of ethanol testing, direct measurements versus conversions, any screening or confirmation history, and uncertainty. If a breath instrument was used, identify the instrument, observation procedures, calibration records, actual displayed unit and whether the result represents breath concentration or a device conversion to an estimated blood equivalent. A model target derived from Widmark’s equation is not a laboratory result.

Next, identify what impairment question is being asked. A cognitive test result relates to one set of task demands; a driving simulator evaluates selected responses in a synthetic road environment; a witness account, vehicle data, and physical evidence describe an event in the real world. No single item supplies all of the others. Where possible, obtain the individual person’s complete performance record, baseline or repeat tests, sensory corrections, medications, fatigue and learning history, rather than trying to infer an event-level conclusion from a paper’s mean effect.

For introductory context on sample-based alcohol interpretation, see the site’s two-blood-sample alcohol-elimination research brief. It concerns a different analytical question and should not be combined with this instrument-validation result as if the two studies measured a common endpoint. Additional primary-source reading paths appear in the Attorney Scientists research library.

What are the paper’s decisive limitations?

The older cohort was a selected volunteer sample, not a representative census of licensed drivers or an impaired-driving arrest population. Only 112 of the 182 older participants completed all four cognitive tasks; the research model used imputation for missing measures to generate composite scores for all 182. The researchers described their older-group assessment as cross-sectional, so change with chronological age was not directly followed within the same individuals over decades. Their road rating was a structured instructor assessment, not a count of actual crashes.

The acute-ethanol arm enrolled just 20 relatively young volunteers, with prior practice on the same tests. A randomized crossover can improve within-person experimental inference, but it does not establish population-wide individual sensitivity or a clinically actionable cutoff. Participant guesses about treatment conditions exceeded chance. Simulator sickness and learning can complicate some vehicle simulation endpoints. Older drivers’ ability to compensate for slower processing may differ from the response of a younger participant abruptly exposed to ethanol.

The article itself noted that risk avoidance and sustained attention were not incorporated as separate MedDrive battery domains, even though they can matter to real-world decisions. It also identified precision limitations in the central-visual-processing component. These are substantive construct-validity restrictions, not mere footnotes.

Most important, the investigators expressly cautioned against relying on a MedDrive score alone to prohibit an older driver from driving. That is a defensible reading of their results: the instrument detected measurable cognitive variation and group responses to alcohol, but its individual road-safety classification remained inadequate.

Frequently asked forensic questions

Did the paper directly measure venous whole-blood BAC in its alcohol arm?

No independent forensic venous whole-blood gas-chromatography result was described for each session. The reported experimental concentrations were controlled using breath testing and dose adjustments. The author’s “BAC” shorthand should be read in light of that method.

Does a statistically significant MedDrive decline mean a driver was impaired?

It demonstrates a statistically detected group difference in specified computer tasks under the research protocol. It does not alone determine whether a named driver was impaired at a different time or in ordinary traffic.

Why is the on-road R² = 0.053 result important?

It shows that the composite was only weakly associated with real on-road driving ratings in the 182-person older cohort, notwithstanding its responsiveness to alcohol in a different, younger group.

Can the below-25 MedDrive cutoff be used to declare an older person unfit?

Not on the evidence of this study alone. The authors reported many false-positive classifications and missed poor performers; the cutoff’s relative-risk confidence interval included one.

Did the researchers test actual crash risk after alcohol exposure?

No. They measured computerized cognition and laboratory simulator outcomes in the young alcohol group. They did not observe traffic crashes or establish event-specific collision causation.

Is simulator SDLP the same as alcohol concentration?

No. SDLP is a measured property of simulated vehicle lateral movement. Alcohol concentration was separately approximated through the session’s breath-monitoring protocol. The statistical relationship is neither an ethanol assay nor a personal diagnosis.

Are repeated MedDrive scores independent observations from different people?

No. The 17-person reliability study contained repeated tests in the same volunteers, and those young volunteers were related to the pool used in the 20-person alcohol arm. The repeated observations cannot be counted as independent recruits.

Answer first: In a controlled before-and-after visual-testing study, 37 adults consumed the same 450 mL volume of red wine, and several measures of visual performance deteriorated. Women showed greater average deterioration on selected measures and a composite visual-deterioration score. But they also reached higher measured breath alcohol concentrations. The investigators’ adjusted analysis associated breath concentration with the composite outcome and did not establish a separate, reliably independent sex effect. The results concern group testing under specific conditions; they do not establish what any individual driver saw, did, or caused.

This account examines the original primary study by Casares-López and colleagues, Changes in Visual Performance under the Effects of Moderate–High Alcohol Consumption: The Influence of Biological Sex, published in 2021 in the International Journal of Environmental Research and Public Health. Read the complete original paper (DOI: 10.3390/ijerph18136790).

The forensic question is not simply whether alcohol affects eyesight

The more specific issue is whether a person who sees well on an ordinary high-contrast acuity chart must also perform normally when detecting faint objects near a bright light at night. The study examined a different set of visual functions: detection of peripheral light stimuli around a bright source, contrast sensitivity, binocular depth discrimination, and optical image quality. These functions are related to vision, but none is identical to ordinary distance acuity, and none by itself is a measurement of actual driving behavior.

The investigators also asked whether biological sex explained variation in the changes after alcohol exposure. The crucial design feature was that each participant consumed the same volume of wine, rather than receiving a dose adjusted to produce the same breath alcohol concentration in every person. Consequently, a difference between sex groups might reflect achieved exposure, visual physiology, body characteristics, experimental variability, or several factors together. It should not be converted into a statement that all women are more impaired than all men at a matched alcohol concentration.

Who participated and how was the wine challenge conducted?

The analyzed sample comprised 37 adults: 20 women and 17 men. The investigators initially assessed 39 people and excluded two—one because of abnormal binocular-vision testing and one because of simulator sickness. Participants had corrected near and distance acuity meeting the study’s criterion, no reported disorder that would limit binocular visual performance, no prior refractive surgery, and no medication expected to affect the tests or alcohol absorption. They were characterized as social drinkers, with Alcohol Use Disorders Identification Test scores of eight or lower.

Each participant attended two experimental sessions, separated by two weeks and held at the same time of day. The order of the sessions was randomized. In one session the participant was tested under baseline conditions; in the other, the participant consumed 450 mL of red wine containing 13.5% alcohol by volume over approximately 30–40 minutes. The wine was consumed about two hours after the last meal. Training was supplied before the baseline testing so subjects understood the visual tasks.

There was no alcohol-free, taste-matched beverage placebo challenge. The baseline condition was not a separate randomized sham-drinking experience. Randomizing the order of the two sessions helps mitigate some order effects but does not blind the participant or experimenters to all differences in the experience of drinking wine.

What was actually measured: exhaled breath, not venous whole-blood ethanol

The investigators used a Dräger Alcotest 6820 to measure ethanol in exhaled breath. The original measurement unit was milligrams of ethanol per liter of exhaled air (mg/L). An initial breath measurement was taken 30 minutes after wine consumption; additional measurements followed at approximately 20-minute intervals, providing four readings averaged for each participant. This protocol measures breath alcohol during the test period. It does not supply four gas-chromatographic measurements of venous whole-blood ethanol.

For all 37 participants, the mean measured breath concentration was 0.34 ± 0.01 mg/L; the male subgroup averaged 0.30 ± 0.02 mg/L and the female subgroup 0.38 ± 0.02 mg/L. The paper expressed 0.34 mg/L as an approximately 0.07% BAC-equivalent, relying on a breath-to-blood convention. That author-provided equivalence is not a separate laboratory analysis of a blood specimen and should not be entered into a forensic record as though a venous whole-blood ethanol assay had been performed.

The researchers used a modified Widmark calculation before the challenge, assigning distribution factors of 0.60 for women and 0.70 for men. Those factors helped plan exposure, but a calculated target or expected BAC is not interchangeable with a breath-instrument reading. Both are different again from an actual chromatographic whole-blood measurement. For general distinctions between measured samples and later concentration calculations, see the two-blood-sample alcohol elimination research brief.

How did the researchers test visual disturbance around a bright light?

One outcome was the visual disturbance index (VDI), derived from a halometer examination. Participants looked toward a central bright stimulus under dim-room conditions and attempted to identify peripheral targets. The central stimulus had a reported luminance of 176 cd/m²; peripheral stimuli were 61 cd/m². An increased VDI meant that more peripheral stimuli were not detected in the presence of the bright central source—an operational measure of visual disturbance or halo perception.

The protocol tested both one-eye viewing and binocular viewing. It did not recreate an individual driver’s exact windshield, headlamps, vehicle speed, road contrast, weather, visual field, gaze behavior, or hazard timing. A halometer result can reveal poorer laboratory detection under glare-like conditions without measuring whether a particular roadside object was visible to a particular driver.

The original results showed monocular VDI increasing from 0.21 ± 0.02 at baseline to 0.31 ± 0.03 after alcohol (p < .001), and binocular VDI increasing from 0.14 ± 0.01 to 0.19 ± 0.02 (p = .005). Participants’ average measured pupil size also increased from 5.6 ± 0.2 mm to 5.9 ± 0.2 mm (p < .001). These are sample averages and source-reported standard errors, not a clinical prediction interval for an individual.

Which visual measurements changed after the same wine volume?

The original study assessed more than glare. It measured stereopsis—the ability to discriminate depth from binocular disparity—at near and far distances. Stereoacuity is expressed in arc seconds, and a higher threshold means poorer performance on that task. It also assessed contrast sensitivity using patterned targets, where a lower score indicated worse sensitivity on the authors’ test. Optical image quality was characterized using double-pass retinal imaging, including the objective scatter index (OSI), the Strehl ratio and modulation-transfer-function (MTF) cut-off.

Selected directly reported baseline and post-wine group measurements from the original primary study. Values are mean ± reported standard error; units and score direction vary by instrument.

Visual measure	Baseline	After 450 mL wine	Source test result	What a change means
Monocular VDI	0.21 ± 0.02	0.31 ± 0.03	p < .001	More peripheral stimuli missed near the central light
Binocular VDI	0.14 ± 0.01	0.19 ± 0.02	p = .005	More binocular visual disturbance under that test
Pupil diameter	5.6 ± 0.2 mm	5.9 ± 0.2 mm	p < .001	Wider mean pupils in dim surroundings
Distance stereoacuity	52.7 ± 6.66 arc sec	128.3 ± 14.14 arc sec	p < .001	Higher depth-discrimination threshold
Near stereoacuity	18.2 ± 2.76 arc sec	38.6 ± 6.38 arc sec	p < .001	Higher near-depth-discrimination threshold
Monocular contrast sensitivity	124.9 ± 3.39	99.9 ± 3.62	p < .001	Lower performance on the source’s contrast test
Objective scatter index, 4-mm pupil	0.56 ± 0.06	0.82 ± 0.10	p < .001	More scatter on the double-pass optical measurement

These are outcomes in the same 37 participants before and after the alcohol condition, not measurements from 74 independent people. The study compared different physical and psychophysical constructs. A difference of 0.10 in VDI cannot be equated numerically with a 0.10 change in breath alcohol, and a change in arc seconds cannot be treated as a percentage change in crash risk.

Does decreased contrast sensitivity tell us what a driver could detect?

Contrast sensitivity is the ability to detect patterns that differ only modestly from their surroundings. It is not synonymous with resolving high-contrast black letters against a white background. The researchers used monocular and binocular assessments across spatial frequencies from 0.75 to 18 cycles per degree. Their results indicated worse contrast sensitivity after alcohol on the tested spatial frequencies, with the pattern of deterioration depending on frequency and whether viewing was monocular or binocular.

The group mean monocular summary score declined from 124.9 to 99.9; the source reports a statistically significant paired difference. The binocular summary score also fell, from 149.8 to 130.1. Those numerical scores belong to this testing arrangement and should not be presented as universal contrast-detection probabilities. Environmental glare, angular size, the contrast of the actual object, corrective lenses, adaptation to darkness and attention all affect practical detectability.

The investigators reported that higher visual disturbance was correlated with worse contrast sensitivity: VDI versus monocular contrast sensitivity had Spearman ρ = −0.461 and VDI versus binocular contrast sensitivity had ρ = −0.558, both with p < .001. These correlations do not prove that one test causes the result in the other or that either score can reconstruct an individual driver’s missed hazard.

What did the retinal optical-quality measurements add?

The double-pass system introduced a point of infrared light into the eye and collected the reflected retinal image. This gives optical-quality information that is not obtained simply by asking a person whether they felt intoxicated. An increased OSI indicates greater light scattering under the device’s conditions; decreases in Strehl ratio or MTF cut-off indicate diminished optical image quality by their respective definitions.

At a fixed 4-mm pupil, the mean OSI rose from 0.56 ± 0.06 to 0.82 ± 0.10 (p < .001). At a 5-mm artificial pupil, mean MTF cut-off decreased from 36.31 ± 1.82 to 31.28 ± 1.65 cycles per degree (p < .001); mean Strehl ratio decreased from 0.21 ± 0.01 to 0.17 ± 0.01 (p < .001). The researchers observed a modest association between OSI and VDI (Spearman ρ = 0.288, p = .042).

These optical measurements reflect changes in ocular imaging within the study protocol, not directly measured brain ethanol, eye-movement coordination, mental workload, road tracking or collision probability. Optical quality, behavioral detection and driving performance are related concepts, but each requires its own evidence.

Did women and men have the same measured breath exposure?

No. Although everyone received the same volume of wine, the mean breath alcohol readings differed: 0.38 ± 0.02 mg/L in women versus 0.30 ± 0.02 mg/L in men. The study found that women also had lower average body mass index. This matters because fixed-volume administration is not a matched-concentration experiment. At one end, individuals can achieve different breath concentrations from the same beverage volume; at the other, even matching concentrations would not automatically guarantee identical visual responses.

The investigators constructed a composite visual deterioration score (VDS) using standardized changes across their visual variables. On their coding scheme, more negative VDS values indicated greater deterioration. Women had an average VDS of −0.22 ± 0.12 and men 0.27 ± 0.12; the unadjusted difference was reported at p = .005. This means the direction of the score cannot be inferred from its name alone: a value moving down was worse under the paper’s coding convention.

Across individuals, the article reported a negative correlation between breath concentration and VDS (Spearman ρ approximately −0.47, p < .01). That is directionally consistent with greater reported deterioration at higher achieved breath concentrations. A correlation is not a dose-response calibration that converts a new person’s breath result into a defensible individual VDS.

Which sex differences remained significant for individual visual tests?

The authors compared deterioration by sex on multiple measures. The results were not a uniform finding that each endpoint differed significantly by sex. The source’s Table 3 identified significant unadjusted sex-group differences in monocular contrast sensitivity and two optical-quality measures at the 5-mm pupil. Several other sex contrasts—including the monocular and binocular halo indices, near and distance stereoacuity and the 4-mm objective scatter index—were not statistically significant in those unadjusted comparisons.

Selected between-sex deterioration tests and exposure distinction, as reported in the original paper. These are group tests, not validated sex-specific legal impairment thresholds.

Comparison	Source result	Interpretation boundary
Breath alcohol concentration after same 450 mL volume	Women 0.38 ± 0.02 mg/L; men 0.30 ± 0.02 mg/L; group difference reported significant	Sex groups had different achieved exposure; not a matched-BrAC comparison
Monocular contrast-sensitivity deterioration	Unadjusted sex test p = .004	A group contrast on one testing scale, not every person’s response
5-mm pupil Strehl-ratio deterioration	Unadjusted sex test p = .034	One optical-quality endpoint, sensitive to pupil/testing conditions
5-mm pupil MTF cut-off deterioration	Unadjusted sex test p = .036	Another optical-quality group difference; not a driving score
Monocular halo disturbance index, sex difference	p = .497 (not significant)	Both groups could change after wine without a proven between-group difference
Binocular halo disturbance index, sex difference	p = .407 (not significant)	No significant unadjusted sex difference on this specific halo measure
Near and distance stereoacuity, sex differences	p = .220 and .390 (both not significant)	Within-person deterioration and sex-group contrasts answer different questions

The key distinction is between a within-person baseline-versus-alcohol effect and a between-sex effect on the size of that change. They are different statistical comparisons. A visual function can deteriorate after alcohol for the combined sample while its observed deterioration does not differ significantly by sex. Likewise, a significant composite difference does not make all of its contributing measurements significant.

What happens when breath alcohol, age and BMI are considered together?

The researchers performed an analysis of covariance with the composite VDS as the outcome, biological sex as a group factor, and breath alcohol concentration, age and BMI as covariates. Their stated interpretation was that breath concentration remained associated with composite visual deterioration (reported p = .041), while a separate sex difference was not significant after adjustment. BMI and age were not reported as independently significant covariates in that analysis.

This is a more limited conclusion than saying “sex is irrelevant.” There were only 37 participants, the same wine volume yielded different average breath exposure by sex, and adjustment is a statistical modeling technique rather than a new experiment that holds dose, breath concentration and all biological variables constant. The study does not supply a sex-specific individual prediction interval, a validated forensic impairment threshold or proof about another age distribution. Forensic interpretation should preserve the distinction between association with achieved breath exposure and proof of an inherent sex-dependent susceptibility.

Are halo perception, optical scatter and stereopsis all measures of the same impairment?

No. The halo test required subjects to detect peripheral targets around a light; contrast sensitivity required recognizing faint spatial patterns; stereoacuity required distinguishing depth from binocular disparity; the double-pass optical instrument measured image quality. Participants might change on several outcomes for different reasons. The paper documented group changes across those domains, but it did not establish one latent “driving impairment” score with a known false-positive rate, diagnostic specificity or cross-population calibration.

The composite VDS was assembled by standardizing each study variable and averaging the changes with equal weights, reversing some signs so more-negative values consistently represented worse outcomes. Equal weighting was a methodological choice, not proof that each component has equal importance in a crash reconstruction. A score assembled for an experiment may summarize a group pattern yet still lack validation for a new clinical, police or expert-witness purpose.

Nor does a fall in contrast sensitivity explain all observed visual disturbance. Pupil diameter, intraocular scatter, the tear film, optical aberrations, binocular alignment, attention and test strategy are distinct possible contributors. The study measured some of these domains, discussed others from published literature, and did not independently measure every proposed biological mechanism.

How far can these data travel into DUI or collision testimony?

They provide a scientifically relevant caution against assuming that ordinary visual-acuity performance rules out impairment on other visual functions. They also show how averaging a breath measurement across a testing session is different from obtaining an instantaneous value at a specified driving time. But no participant drove a real vehicle in the analyzed experiment, and the researchers did not measure crash involvement, detection distance for a real roadway obstacle, steering variability, emergency-braking latency, night-adaptation at a collision location or causal responsibility for an event.

When the question concerns one person at one time, the relevant evidence must be considered separately: the original breath instrument and its readings; observation and specimen times; actual reported blood specimen type if any; roadway illumination and object contrast; vision correction; sightline and headlamp geometry; medical eye history; and contemporaneous behavior. The original study cannot retrospectively determine these facts. A person may have experienced a visual change without the evidence proving that the change caused a specific failure to see; conversely, a normal acuity test alone cannot exclude the type of laboratory change studied here.

The study also cannot establish a conversion from a person’s BrAC in mg/L to a sex-specific halo score. A breath measurement is analytically different from a whole-blood ethanol assay, a Widmark estimate is a model rather than a specimen, and a correlated task score is not a direct indicator of driving causation. For broader reading about what research material is and is not being used in a forensic appraisal, consult the editorial and research source policy and the scientific research library.

What design limitations should remain visible in the interpretation?

First, 37 people provide limited precision for subgroup interactions and do not represent every driver, age range, visual condition or alcohol exposure history. All participants met specific baseline ocular criteria, and the mean ages reported for the two sex groups were in the late twenties or early thirties. The source did not validate the composite score in older drivers or people with ocular disease. The study was not designed as a population-level screening trial.

Second, everyone consumed the same 450 mL of one 13.5% red wine rather than an independently titrated, blood-confirmed equal-concentration dose. Measured breath concentrations varied. This is important because the central comparative claim depends on separating group membership from achieved exposure. The investigators addressed these factors statistically, but no analysis removes all possible confounding in a small experimental sample.

Third, participants necessarily knew whether they had consumed a substantial amount of wine, and there was no parallel taste-matched sham-drinking control. Although the two sessions were randomized in order and separated by two weeks, expectation, fatigue, learning and differences between sessions may still affect some psychophysical tests. Finally, multiple vision endpoints and exploratory sex comparisons were analyzed. Individual p values should not be read as independent replications of a single broad claim or as evidence of an individual’s legal impairment.

Seven practical questions about the original study

1. Does this paper say every woman has worse night vision than every man after drinking?

No. It compares group averages under one fixed-volume red-wine challenge. Female participants had a higher mean BrAC and a more negative average composite score, while many individual between-sex visual tests were not statistically significant. The study did not establish a rule about every individual at the same achieved alcohol concentration.

2. Did the investigators measure blood alcohol by gas chromatography?

No. They recorded exhaled-breath ethanol with a Dräger Alcotest 6820 in mg/L, then discussed conventional BAC equivalents and used a planning calculation. No separate forensic venous whole-blood gas-chromatography concentration was measured in this experiment.

3. Did all 37 people receive the same amount of alcohol per kilogram?

No. They received the same 450 mL wine volume, not the same weight-adjusted grams per kilogram. Body size and ethanol distribution differences help explain why achieved breath readings differed. The paper’s Widmark estimates do not convert the protocol into a matched-BrAC experiment.

4. Did alcohol increase halos around a bright light?

The group’s monocular and binocular visual disturbance indices were significantly higher after wine consumption in a laboratory halometer test. That finding supports a change in the test’s target-detection measure; it does not calculate glare disability for a particular oncoming headlamp or establish what someone saw before a collision.

5. Was every visual outcome significantly different between women and men?

No. Selected unadjusted contrast-sensitivity and 5-mm optical-quality measures differed by sex, while the two halo-indices, stereopsis measures and several other contrasts did not. This is different from the separate finding that many visual measures deteriorated for the sample as a whole after wine.

6. Does a higher breath measurement prove exactly how poor a person’s vision was?

No. The composite visual score and measured BrAC were associated across the sample, but no validated individual prediction equation, sensitivity/specificity threshold or interval for an individual driver was supplied. A measured breath exposure is not itself an observed visual-task result.

7. Can the results determine whether alcohol caused an accident or a witness failed to observe something?

No. The study did not test road driving, real incident reconstruction, eyewitness identification or collision causation. It is relevant background evidence about distinct visual functions that may change after alcohol; applying it to an individual requires independent case evidence and careful attention to the specific inferential question.

Primary source and narrow conclusion

Casares-López M, Castro-Torres JJ, Ortiz-Peregrina S, Martino F, Ortiz C. Changes in Visual Performance under the Effects of Moderate–High Alcohol Consumption: The Influence of Biological Sex. International Journal of Environmental Research and Public Health. 2021;18(13):6790. https://doi.org/10.3390/ijerph18136790.

Bottom line: A single fixed wine volume was followed by group deterioration on halo-related visual detection, contrast sensitivity, stereoacuity and optical image quality. Apparent sex-group differences were entangled with different achieved breath alcohol exposures. The study supports careful separation of visual-test performance, the measured breath specimen, a modeled blood-equivalent concentration and the entirely different question of whether an identified person was functionally impaired or caused a specific event.