EVIDENCE LIBRARY
Can Gastric-Emptying Drugs Change Alcohol Exposure? What an Oral-vs-IV Trial Found
A 16-volunteer alcohol study found that gastric-emptying drugs altered oral ethanol availability; only 7 had matched first-pass comparisons. Read the forensic limits.
Yes. A controlled human experiment found that medicines which sped up or slowed down gastric emptying changed how much orally administered ethanol reached the systemic circulation. In the 1998 study by Oneta and colleagues, faster gastric emptying after metoclopramide was associated with greater oral ethanol availability, whereas delayed emptying after N-butylscopolamine was associated with lower availability and greater calculated first-pass metabolism. These were effects on group concentration–time curves under controlled conditions—not a validated correction to any driver’s blood-alcohol concentration (BAC).
The strongest comparison rests on seven volunteers who had oral and intravenous ethanol challenges with the study medications. Gastric-emptying times were evaluated in nine volunteers. It would be incorrect to describe every endpoint as a 16-person drug comparison, even though 16 volunteers completed the initial oral-versus-intravenous study.
What question did the study answer?
First-pass metabolism describes ethanol eliminated before an orally administered dose reaches the systemic circulation. Once a person drinks alcohol, the dose must leave the stomach, enter the small intestine, be absorbed, and pass through tissues—including the liver—before the observed systemic concentration–time curve is established. Gastric alcohol dehydrogenase may also metabolize ethanol while the dose remains in the stomach. The rate at which the stomach empties could therefore change how much alcohol becomes systemically available even when the administered dose does not change.
The controlled research question was narrower than whether a medication makes someone “more drunk.” It asked whether intentionally changing gastric emptying altered oral ethanol absorption and the calculated fraction of a dose lost on its first passage, using intravenous administration of the same ethanol dose as a comparison. The study did not test police breathalyzer accuracy, alcohol-related crashes, clinical intoxication, field sobriety testing, or any individual defendant’s drinking history.
Who participated, and how was ethanol administered?
Sixteen healthy volunteers—eight men and eight women, aged 24–52 years—received ethanol on separate study days by two routes: orally and by intravenous infusion. The dose was 0.225 grams of ethanol per kilogram of body weight. Each challenge followed a standardized light breakfast by approximately one hour. Oral ethanol was given in a glucose-containing liquid and consumed over about two minutes. The intravenous dose was infused over 30 minutes. Blood sampling extended to four hours, allowing the investigators to compare the resulting ethanol concentration–time curves.
In a subgroup of ten participants, the oral experiment was repeated after intravenously administered metoclopramide (10 mg), intended to accelerate gastric emptying, and after N-butylscopolamine (20 mg), intended to delay emptying. Seven of these volunteers also underwent corresponding intravenous ethanol comparisons after the medications. This distinction is fundamental: the overall enrollment was 16, but the most pertinent medication-specific first-pass comparisons were derived from seven matched oral/intravenous participants.
Ultrasound of the stomach’s antral region was used to follow gastric emptying. Drug-associated gastric-emptying summaries in the primary paper’s Table 2 were based on nine volunteers, not all 16. The experiment controlled ethanol dose and a light meal, but participants were not randomly drawn from drivers, trauma patients, or other forensic populations.
What did the researchers physically measure?
Three sets of measurements served different scientific purposes. Investigators drew serial blood specimens for an enzymatic ethanol assay on a clinical centrifugal analyzer; they monitored stomach emptying by ultrasound; and they obtained gastric biopsies to examine alcohol dehydrogenase (ADH) activity. The original paper alternately refers to serum, plasma, and blood when describing its ethanol curves. Its figures and Table 1 label serum, while its methods also describe serum preparation and a plasma assay. Consequently, the fairest description is blood-derived serum/plasma ethanol measurements by a clinical enzymatic method, with matrix terminology not completely consistent in the source.
The study did not present a paired, independently validated forensic whole-blood headspace gas-chromatography BAC dataset. Nor was ultrasound a direct measurement of ethanol or impairment. The investigators then used the measured concentration–time curves to calculate the area under the curve (AUC), oral systemic availability, and first-pass metabolism. Those are derived pharmacokinetic quantities, not additional assayed analytes.
For comparison, a forensic whole-blood ethanol analysis by headspace GC-FID has distinct matrix, validation, and chain-of-custody requirements. Any claimed correspondence between the Oneta study and a case result has to confront that analytical difference.
How much did the medications change gastric emptying?
The investigators reported the following mean gastric-emptying times, with standard errors, in the nine-person ultrasound analysis. These are averages of the experimental conditions; they do not say how any particular person’s stomach will empty after drinking.
| Condition | Mean gastric-emptying time | Interpretation |
|---|---|---|
| No additional medication | 44.1 minutes (SEM 3.1) | Reference condition |
| Metoclopramide 10 mg IV | 31.6 minutes (SEM 2.8) | Faster emptying than control; reported p = 0.0006 |
| N-butylscopolamine 20 mg IV | 70.7 minutes (SEM 4.5) | Slower emptying; drug-to-drug contrast p = 0.0001 |
Source: Oneta et al., 1998, original Table 2. Values are mean (standard error of the mean), not individual ranges or uncertainty margins for a case.
In other words, the motility interventions operated in the intended opposite directions. Gastric emptying averaged approximately 12.5 minutes faster with metoclopramide than without added medication and approximately 26.6 minutes slower with N-butylscopolamine than without added medication. These are arithmetic differences between group means; neither is a pharmacokinetic constant suitable for inserting into an individual’s BAC reconstruction.
Did oral ethanol availability change as well?
Yes, in the study’s within-person experimental comparisons. Oral systemic ethanol bioavailability was highest with metoclopramide and lowest with N-butylscopolamine. Reported oral-availability comparisons were statistically significant for untreated versus metoclopramide (p = 0.0158), untreated versus N-butylscopolamine (p = 0.0425), and metoclopramide versus N-butylscopolamine (p = 0.0097). The medication-specific oral plots involved ten volunteers, while the matched intravenous comparisons needed to estimate first-pass loss were available in seven.
This evidence supports an association between altered gastric motility and systemic exposure after oral dosing at the specified dose and meal conditions. However, increased oral bioavailability is not the same endpoint as a faster rate of ethanol elimination from blood, nor is it proof of a particular level of psychomotor impairment. Absorption, first-pass removal, systemic elimination, and observed functional effects answer different questions.
One further distinction matters: because the intravenous infusion itself took 30 minutes, the oral and intravenous concentration curves do not share the same administration-time profile. The pharmacokinetic computation compares total systemic exposure, not simply the height of two peaks at a chosen instant. A single peak concentration cannot replace the integrated AUC analysis.
What was actually significant for calculated first-pass metabolism?
Using matched oral and intravenous exposure measurements, the investigators found greater calculated first-pass metabolism after N-butylscopolamine than after metoclopramide (p = 0.0004). That was the clearest first-pass medication comparison. Two other contrasts did not cross the authors’ conventional significance criterion:
| Comparison or analysis | Study result | Supported interpretation |
|---|---|---|
| First-pass metabolism, N-butylscopolamine versus metoclopramide | p = 0.0004; matched route subgroup n = 7 | Statistically significant difference in derived first-pass loss |
| First-pass metabolism, N-butylscopolamine versus no medication | p = 0.0525; matched route subgroup n = 7 | Not significant at p < 0.05; do not label this pairwise test positive |
| First-pass metabolism, metoclopramide versus no medication | p = 0.4154; matched route subgroup n = 7 | No demonstrated significant difference in this contrast |
| Gastric-emptying time versus calculated first-pass metabolism | Pearson r = 0.43; p = 0.0407; 23 pooled observations | Modest positive study-level association; 23 observations are not 23 independent volunteers |
Source: Oneta et al., original Results and Figure 9. A p value is neither an effect size nor the probability that a conclusion is true; multiple contrasts and the small matched subgroup warrant restraint.
The numerical contrast between p = 0.0525 and p = 0.0004 is particularly important for scientific testimony. An article may reasonably describe the overall physiological pattern, but one should not silently convert the untreated-versus-N-butylscopolamine first-pass comparison into a statistically significant individual result. Likewise, a correlation calculated over repeated experimental observations must not be presented as if 23 independent people supplied a generalizable prediction rule.
Is this evidence that the stomach—not the liver—removed the alcohol?
No. The measured data cannot independently partition first-pass ethanol metabolism between those organs. The stomach contains ADH enzymes, and holding alcohol in the stomach longer may permit more gastric metabolism. But delayed emptying also slows the delivery of ethanol to the small intestine, from which much ethanol is absorbed. Slower delivery to the liver could change the share removed during hepatic first passage. Both pathways could contribute to the resulting lower systemic exposure.
The paper tested gastric biopsy ADH activity in subsets of participants. The reported correlations between ADH activity and estimated first-pass metabolism were not statistically significant: antrum r = 0.538, p = 0.0577 (n = 13); corpus r = 0.448, p = 0.1439 (n = 12); and the mean of those sites r = 0.531, p = 0.0928 (n = 11). These results neither prove that gastric ADH is irrelevant nor establish that gastric ADH alone caused the motility findings.
Separate in vitro experiments did not show direct effects of the studied medicines on gastric ADH activity at the tested concentrations. That finding is consistent with a motility-associated explanation, but it is not a complete clinical mechanistic separation of gastric metabolism, hepatic metabolism, and absorption. The authors themselves recognized that disentangling the organs can be difficult without considering gastric-emptying time.
Why does this matter when interpreting an alcohol concentration?
It illustrates why dose alone does not uniquely determine the entire concentration–time curve. Even at a fixed administered dose, conditions affecting gastric emptying may alter the temporal arrival of ethanol and the fraction that reaches systemic circulation. A forensic reconstruction based on only an assumed drink amount, a reported drinking time, or a generic absorption interval may fail to represent those processes accurately.
But this is not a license to reverse-engineer a person’s stomach function from one concentration. The study did not validate metoclopramide or N-butylscopolamine as a forensic BAC calculator input, and it did not estimate the probability that any driver was impaired. For a specific case, the actual medication history, dose, formulation, timing, food intake, absorption state, specimen collection details, tested matrix, and laboratory method would all have to be established independently. Further, a person’s medication exposure does not itself establish which direction, or how much, their observed BAC differed from a hypothetical untreated condition.
A pharmacokinetic interpretation should therefore keep the observed analytical result separate from any model of earlier absorption or elimination. When considering whether a measured value can be translated into a historical concentration, the assumptions must be expressly stated and evaluated against the available record rather than treated as measured facts.
Does faster stomach emptying necessarily produce higher BAC in everyone?
No. The study establishes a response under one small ethanol dose, a standardized light meal, two pharmacological motility manipulations, and a limited healthy-adult sample. It does not show a universal quantitative rule for all doses, fasting states, beverages, gastrointestinal conditions, medications, or drinking patterns. Even the authors’ untreated first-pass estimate, mean 6.5% with SEM 6.7% across 16 people, was imprecise; it should not be applied as a fixed percentage correction to an individual’s drink or test result.
The paper also found no significant male–female difference in its particular oral systemic bioavailability comparison (p = 0.6194) or its first-pass comparison (p = 0.7080). These non-significant tests in a small sample do not establish equivalence across sexes or rule out effects in broader populations. Absence of detected significance is not evidence that all individuals absorb ethanol identically.
What kinds of forensic conclusions are outside this experiment?
This research did not establish a legal blood-alcohol threshold, demonstrate that a specific person’s blood BAC was “wrong,” validate a field impairment test, or identify a cause of a collision. It did not measure driving behavior, divided attention, reaction time, nystagmus, witness memory, or clinical infectious disease. Its gastric ultrasound measure is a motility observation, and its model-calculated first-pass percentage is an estimated biological process. Neither is a direct assay of impairment, intent, or causation.
It also provides no scientific basis for recommending that someone take either studied medication before or after consuming alcohol. Those agents were research interventions under defined clinical supervision. Their use, risks, and appropriate treatment indications cannot be inferred from ethanol bioavailability data.
For an individual specimen, collection time, specimen identity, and preanalytical handling still govern what the laboratory measurement means. A population trial cannot replace authentic case records, validated testing, appropriate uncertainty assessment, or incident-specific evidence.
Frequently asked questions
Does delayed gastric emptying lower alcohol exposure?
It was associated with lower oral systemic ethanol availability in this controlled study at 0.225 g/kg after a light meal. The size and even applicability of an effect in another person cannot be inferred without evidence.
Did metoclopramide produce a higher blood alcohol level?
Metoclopramide accelerated gastric emptying and was associated with greater oral ethanol bioavailability than the other experimental conditions. The primary study analyzed clinical serum/plasma-labeled ethanol curves, not a forensic whole-blood case result or an individual’s driving impairment.
Did N-butylscopolamine significantly increase first-pass metabolism relative to no drug?
Not under the reported pairwise test’s 0.05 criterion: the p value was 0.0525. The N-butylscopolamine-versus-metoclopramide first-pass contrast was significant (p = 0.0004). Confusing those comparisons would overstate the evidence.
How many participants supplied the main first-pass medication comparison?
Seven had corresponding oral and intravenous ethanol challenges under the medication conditions. Sixteen volunteered for the initial oral/intravenous experiment; ten repeated the oral challenges, and nine supplied the tabulated gastric-emptying analysis.
Does the study show that gastric alcohol dehydrogenase alone explains the effect?
No. Delayed gastric emptying could modify gastric exposure and the rate of ethanol delivery to the intestine and liver. Biopsy ADH correlations were not significant; the trial did not separately quantify gastric and hepatic first-pass removal.
Can these results establish a defendant’s earlier BAC from a later blood sample?
No. It supplies mechanistic reasons to examine absorption assumptions, but no validated individual retrograde-calculation equation, case-specific gastric-emptying measurement, or correction factor. Actual specimen and time evidence remain indispensable.
Are stomach ultrasound, a serum alcohol assay, and a forensic BAC the same kind of result?
No. Ultrasound assesses gastric motility; a clinical enzymatic assay measures ethanol in a laboratory specimen whose matrix must be identified; a forensic whole-blood BAC requires its own analytical and interpretive validation. Modeled first-pass metabolism is calculated from concentration curves, not independently measured in a vial.
Bottom line: a measurable pharmacokinetic effect, not an individual verdict
The Oneta experiment shows that controlled acceleration or delay of gastric emptying can change oral ethanol systemic availability and calculated first-pass metabolism. Its most defensible forensic contribution is a warning against assuming a fixed absorption pattern from dose alone. The work does not justify declaring that a particular person had a specific unmeasured BAC, was impaired, was unimpaired, or caused an event. Each conclusion must be tied to the actual specimen, method, physiological context, and independent case evidence.
Primary study: Oneta CM, Simanowski UA, Martinez M, et al. “First pass metabolism of ethanol is strikingly influenced by the speed of gastric emptying.” Gut. 1998;43(5):612–619. doi:10.1136/gut.43.5.612.