Does Feeling Less Drunk Predict a Lower Measured BAC? A 20-Person Alcohol Challenge

Short answer: Feeling relatively little alcohol effect does not establish that a person’s blood alcohol concentration (BAC) is low, that the person is unimpaired, or that alcohol was eliminated faster. In a controlled oral-alcohol challenge involving 20 healthy young men, those who retrospectively reported needing more drinks to notice effects had lower measured BAC curves after the same body-weight-adjusted dose than men who reported effects after fewer drinks. The difference was associated with early alcohol exposure, while the groups’ estimated systemic elimination rates did not differ significantly. The authors proposed presystemic (first-pass) metabolism as one possible explanation, but did not measure that mechanism directly. This is a group-level association, not a method for predicting a particular person’s BAC, performance, or legal responsibility.

The original research is Bramness and colleagues, The Self-Rated Effects of Alcohol Are Related to Presystemic Metabolism of Alcohol, Alcohol and Alcoholism (2023), DOI: 10.1093/alcalc/agad002. The interpretation below distinguishes what the investigators measured from what they calculated or hypothesized.

What did the researchers actually ask?

The study addressed an unusual forensic problem: why two people receiving the same grams of ethanol per kilogram of body weight might show different BAC trajectories, and whether those differences correlate with a history of feeling alcohol’s effects after relatively few or many drinks. Self-reported response was measured with the Self-Rating of the Effects of Alcohol (SRE) questionnaire, not by asking each person whether he felt drunk at a particular sampled BAC.

The researchers focused on the SRE score for periods of heaviest drinking. A higher SRE score meant that a participant retrospectively reported needing more drinks before noticing specified effects; a lower score represented reported effects after fewer drinks. That direction matters: “higher SRE” is lower self-reported sensitivity, not more subjective intoxication. The study divided the 20 men into relatively higher- and lower-effect groups using a median split; it did not randomly assign anyone to be more or less sensitive.

How were alcohol and BAC measured?

The analysis used the placebo-condition day from a prior double-blind crossover project. Twenty medically screened men, recruited among correctional-officer students in Norway, had prior experience drinking more than five units on at least one occasion. Participants arrived after an overnight fast, received a light breakfast, and then drank vodka diluted in fluid at 1.37 grams of ethanol per kilogram of body weight over 30 minutes. Lunch and dinner were provided later during observation.

Blood alcohol was measured at baseline and nine subsequent scheduled times: 1, 1.5, 2, 3.5, 5, 6.5, 8, 10 and 12 hours. The paper reports BAC expressed as percent alcohol; it does not provide a complete forensic laboratory method-validation package for using its research BAC values as casework measurements. The baseline screening blood, biochemical markers, and retrospective SRE score were different measurements serving different purposes.

From the concentration-versus-time samples, the investigators calculated areas under the BAC curve (AUC) and an average rate of decline. AUC is a mathematical summary of sampled systemic exposure, not a second independent alcohol assay. Nor should the study’s “maximum” at the scheduled two-hour measurement be assumed to prove the precise unobserved instantaneous peak between blood draws.

What were the measured and calculated results?

The contrast was consistent with the study hypothesis: after a fixed weight-adjusted oral dose, the men who reported being more responsive to alcohol had higher sampled BAC levels and greater calculated systemic exposure. The following values are group means from Table 2 of the original study, not personal prediction ranges.

Source: Bramness et al., 2023, Table 2. The first-two-hour AUC p-value is taken from the study’s numeric table; the surrounding narrative prints a different p-value. The report of a measured BAC at two hours does not establish a precisely timed individual peak.

For example, the two-hour group means were 0.141% versus 0.114%, an observed difference of 0.027 percentage points under this study’s controlled dosing and sampling. That contrast is meaningful for interpreting group pharmacokinetics. It cannot be applied as a correction to an accused individual’s BAC simply because the individual reports having a high alcohol tolerance.

Did the study find that some participants eliminated alcohol faster?

No statistically significant between-group difference in the calculated systemic elimination rates was established. The investigators subtracted the BAC at 6.5 hours from the BAC at 3.5 hours and divided by three hours. This yielded mean rates of 0.019 and 0.022 percentage points per hour, with p = 0.247. The result is not affirmative proof that the rates were identical; a sample of 20 also has limited precision for detecting smaller differences.

The early BAC gap and larger total AUC therefore must not be translated into “the low-response group burned off alcohol faster.” An alcohol curve reflects more than systemic elimination: administered dose, entry to circulation, distribution, sampling schedule, and pharmacokinetic variability matter. See the separately verified forensic pharmacokinetics overview for the distinction between an observed concentration and a kinetic reconstruction.

Was first-pass or gastric alcohol metabolism directly demonstrated?

No. Presystemic or first-pass metabolism refers to ethanol that is metabolized before entering the systemic bloodstream after an oral dose. Bramness and colleagues interpreted the early differences as compatible with variation in this process. The study did not quantify each participant’s gastric alcohol dehydrogenase activity, compare oral and intravenous exposure in the same individuals, or independently measure a person-specific first-pass fraction.

Alternative and overlapping explanations remain relevant, including differences in gastric emptying, absorption timing, distribution, and other interindividual factors. Participants received a standardized breakfast, but their gastric-emptying rate was not directly controlled or measured. The authors expressly flagged that omission. Gastric-versus-hepatic partitioning was a proposed mechanism, not a demonstrated locus of metabolism. A reader must not transform “could be related to presystemic metabolism” into “proved faster gastric clearance.”

Why is the SRE questionnaire not a BAC meter?

The SRE is retrospective. It asks about how many drinks an individual recalls consuming before experiencing specified subjective effects in earlier periods of life, including effects such as dizziness, slurred speech, stumbling, and unintentional sleep. The questionnaire is not a concurrent blood or breath alcohol assay, not a validated estimate of BAC at the time of an incident, and not a psychomotor-performance examination.

That distinction is particularly important when a witness says “I felt sober,” “I need more drinks than other people,” or “I can handle alcohol.” Such statements concern recalled or perceived response. They cannot retrospectively establish the concentration in a biological specimen, show that the brain was unaffected, or rebut observed driving or balance evidence by themselves. Conversely, a measured BAC does not by itself recreate a complete record of what a person thought, perceived, or did during a particular event.

The study’s baseline alcohol-related indicators—AUDIT questionnaire scores, gamma-glutamyl transferase, carbohydrate-deficient transferrin, and blood-cell indices—were also not real-time intoxication measurements. The authors found no meaningful between-group differences in the tested baseline alcohol-use indicators. That negative comparison was not an independent measurement of drinking histories and cannot exclude every exposure difference.

What does this mean for a courtroom BAC interpretation?

A defensible analysis begins with the actual evidence type. An analytically validated blood alcohol result from a particular specimen is not equivalent to a self-reported history, and neither is equivalent to a person-specific impairment opinion. The controlled oral dose in this research reduces some sources of variability, yet the group mean after that dose cannot be substituted for an unsampled litigant’s concentration at an unknown time.

In litigation, request the original specimen identification, matrix, collection time, assay method and reporting uncertainty before treating BAC as a fact. The laboratory method remains a separate question from subjective effect; consult blood ethanol chromatography and quality controls. If the case instead rests on a breath instrument, its physiology and sampling conditions are separate and should be considered under breath alcohol measurement.

How much can these 20 participants tell us about other populations?

The sample was small and deliberately homogeneous: 20 screened men, all in the study’s recorded white/Caucasian eligibility group, with prior binge-drinking experience. The authors acknowledged that this limits generalization to women, older adults, other demographic groups, people with alcohol use disorder, different diets, and varied medical or metabolic conditions. The 1.37 g/kg challenge was unusually high for an experimental study. Results under that protocol do not define a dose-independent adjustment.

Several further uncertainties are case-critical. There was no independently assigned high-versus-low SRE treatment. Median dichotomization loses information compared with a continuous score. The reported SRE reflects recalled earlier drinking, while the BAC came from one test day. Gastric emptying was not directly measured; the original experimental study was not preregistered. The secondary analysis was not a validation study for forensic back-calculation or driver classification.

There is also a disclosed funding relationship: two coauthors owned a sports-drink firm that funded aspects of the earlier experiment, and another author reported a design advisory honorarium. Funding is not evidence that results are false, but the disclosure belongs in a complete source appraisal. Independent replication with larger, more varied samples and direct mechanistic measurements would be needed before strong general causal claims.

Does “tolerance” always refer to the same phenomenon?

No. Subjective sensitivity concerns the effects someone notices or recalls. Behavioral or functional tolerance concerns performance at exposure; this paper did not measure it with validated driving or cognitive tasks. Metabolic tolerance refers to altered alcohol handling, especially elimination, and was not significantly demonstrated between this study’s groups. First-pass metabolism concerns presystemic disposition before alcohol reaches circulating blood, not simply the descending-limb elimination slope.

Using one label—“tolerance”—for all four mechanisms invites overstatement. The study suggests a connection between recalled sensitivity and systemic exposure under one controlled oral-dose protocol. It does not supply a rule that tolerance defeats BAC evidence, immunizes a person from psychomotor impairment, or predicts an individual’s susceptibility to a collision.

Frequently asked questions

Does a person who says they need more drinks always have a lower BAC?

No. A group association in 20 men does not predict an individual’s BAC after a particular drinking episode. Timing, dose, food, formulation, physiology, and actual analytical measurements still matter.

Were the participants asked if they felt drunk at the time of the blood draw?

The key grouping variable was a retrospective SRE questionnaire about earlier drinking experiences. The research did not validate a contemporaneous “felt sober” answer against task impairment during the challenge.

Was alcohol concentration directly measured or inferred?

The study reports serial blood BAC measurements at specified times. AUC, interval-average elimination rates, and the proposed first-pass explanation are derived quantities or interpretations rather than separate directly measured analytes or enzyme activity.

Did the study establish faster systemic alcohol elimination in the low-response group?

No. The two group mean decline rates differed numerically, but their comparison was not statistically significant (p = 0.247). The result does not establish equality of all individual rates either.

Did the researchers prove that the stomach metabolized more alcohol?

No. Gastric/presystemic metabolism was a proposed mechanism. The study did not administer paired oral and intravenous doses to quantify a first-pass fraction, and gastric-emptying measurements were not available.

Can these findings establish impaired driving, crash causation or criminal intent?

No. The paper did not test simulated driving, real driving, divided attention or collision outcomes. A forensic opinion would need the relevant case facts and independently supported links between evidence and claimed effect.

Why might the sampled two-hour BAC differ despite the same dose per kilogram?

A common weight-adjusted dose does not guarantee identical circulating exposure. Absorption timing, presystemic processes, distribution and other individual variability may influence the curve. This trial establishes an association, not the amount contributed by each mechanism.

Primary evidence and interpretive conclusion

Primary study: Bramness JG, Skulberg KR, Skulberg A, Moe JS, Mørland J. The Self-Rated Effects of Alcohol Are Related to Presystemic Metabolism of Alcohol. Alcohol and Alcoholism. 2023;58(2):203–208. https://doi.org/10.1093/alcalc/agad002. Complete original open-access research article.

Forensic bottom line: Retrospective self-reports of alcohol sensitivity were associated with different BAC profiles following a standardized oral challenge. The measured group BAC separation is real within the reported study, but the mechanism is inferential and personal impairment was not an endpoint. A self-rating, a sampled biological concentration, a modeled concentration-time quantity, and evidence of task impairment each answer different questions; none should be silently substituted for another.