Answer first: A person can report feeling lightheaded after a drink they believe contains alcohol even when the breath instrument reads 0.00. A controlled 2017 study of 64 social drinkers found greater self-reported lightheadedness and alcohol craving after an alcohol-cued placebo than after a clearly nonalcoholic control. The placebo also increased the proportion of alcohol-free beer selected in a laboratory taste test. It did not establish that the volunteers were ethanol-impaired or that subjective lightheadedness diagnoses intoxication. The experimental drink was prepared with vodka applied to the glass rim and misted on its surface, a detail that prevents a categorical assertion of zero ethanol exposure even though the reported breath readings were 0.00.
This distinction matters when a witness or examiner equates an individual’s statement that they “feel drunk” with measured blood alcohol concentration (BAC), objectively demonstrated functional impairment, or the cause of a particular event. Feelings, beliefs about a drink, measurable ethanol, and performance are different evidentiary questions. The controlled experiment illuminates one of those boundaries rather than deciding an individual case.
What did Christiansen and colleagues actually test?
Christiansen, Townsend, Knibb, and Field published the original study in Psychopharmacology in 2017. They enrolled 64 university social drinkers, 53 of them women, aged 18 to 37. Every participant attended two sessions approximately one week apart: one involving a drink presented as alcoholic, and another involving an openly nonalcoholic control. Session order was counterbalanced. Participants were allocated by blocks to either a laboratory configured like a pub or a conventional psychology laboratory. The contextual setting was a comparison between different participants, while drink condition was compared within the same participant.
In the placebo session, participants received 500 mL of chilled lemonade. The researchers added vodka to the rim of the glass and sprayed a vodka mist onto the drink’s surface to create alcohol-related sensory cues. The control session used 500 mL of water; participants were told it contained no alcohol. Researchers recorded alcohol craving and six self-rated feelings before and after each drink. They also asked participants how much vodka they thought the purported alcoholic drink contained. This was an expectancy manipulation, not a study comparing an ordinary intoxicating ethanol dose with sobriety.
All participants provided a zero breath alcohol reading before the sessions, and the investigators report that the breath samples taken after the drinks also read 0.00. The publication does not present paired whole-blood or serum ethanol assays, a direct quantification of the small vodka application, or a validated objective driving or neurocognitive impairment examination. A breath display of 0.00 is a device result at a time of sampling, not proof that no ethanol molecule entered the body.
What changed after the drink presented as alcoholic?
The study’s most useful result is specific rather than sweeping: subjective lightheadedness increased after the placebo, while the control condition did not show a significant change. Craving also increased following the placebo but not the control. In the later taste test, participants consumed a greater proportion of the supplied alcohol-free beer after the placebo than after the control. The authors reported no comparable placebo effect on the five other subjective-intoxication ratings they measured.
| Outcome | Reported comparison | Proper evidentiary limit |
|---|---|---|
| Breath alcohol | Post-drink breath samples reported as 0.00 in both conditions | Device readings only; no whole-blood BAC, serum concentration, or quantified ethanol on the rim |
| Self-rated lightheadedness | Increased after placebo (p < .001); not significantly changed after control | Subjective feeling, not a validated psychomotor-impairment measure |
| Self-rated alcohol craving | Increased after placebo (p < .001); not significantly changed after control | Motivation questionnaire, not proof of actual ingestion or blood concentration |
| Other five subjective ratings | No significant placebo effect reported for other measured intoxication descriptors | Do not generalize the lightheadedness result to every feeling or symptom |
| Preference for alcohol-free beer | Greater percentage of total test liquid consumed as beer after placebo (p = .039) | The beer was nonalcoholic; this outcome was not ethanol consumption or driving performance |
| Bar versus plain laboratory | Did not significantly change the main placebo effects; bar setting increased estimates of vodka amount (p = .035) | Context affected a quantity estimate, not measured BAC or a validated impairment outcome |
Why is the taste-test result easy to misunderstand?
The researchers offered 275 mL of Beck’s nonalcoholic beer and 275 mL of an orange-flavored nonalcoholic drink. Participants were allowed to drink as much as needed to rate flavors, and investigators recorded the proportion of beer out of all test liquid consumed. The paper’s phrase “ad-lib consumption” describes how participants consumed the test drink under these controlled conditions; it does not mean the study demonstrated that a placebo led them to consume an intoxicating amount of alcohol.
The distinction is important in expert reports. An experiment can measure a real behavioral choice without measuring impairment, intoxication, or a real-world harmful act. Calling the test outcome “increased drinking” without identifying the liquid and the denominator would distort the paper. The authors measured a proportional preference in a brief laboratory exercise, not diagnosed alcohol-use disorder, later alcohol use, or crash risk.
Does a zero breath reading refute someone who says they feel drunk?
No. A properly described 0.00 breath reading and a statement of lightheadedness can coexist. The instrument addresses a sampled breath-alcohol signal; the statement addresses a person’s subjective state. The two do not measure the same thing, so they should not be treated as competing readings of one variable.
Nor does the experiment establish that every person reporting drunkenness with a zero result experienced a placebo response. Alternative explanations are case-specific and require an actual history, observations, time-ordered records, instrumentation details and, when relevant, medical evidence. Christiansen and colleagues did not evaluate dizziness in patients, falls, field sobriety tests, memory reliability, or impairment from other drugs. Applying this study to any particular person therefore requires a separate evidentiary foundation.
What did the pub-like setting add?
Researchers predicted that familiar bar cues would strengthen the placebo effect. Their data did not show that interaction for craving, lightheadedness or the percentage of test beer consumed. The bar-like environment was, however, associated with higher estimates of how much vodka had supposedly been in the placebo. That is a narrower result: context was related to the belief about dose, not an objectively administered ethanol dose.
The environmental groups were allocated in counterbalanced blocks rather than by individual randomization, and the research setting was deliberately controlled. Therefore, it would overstate the results to declare that bars never affect expectancy, or that a context effect of a particular magnitude applies to a witness in an actual bar, accident scene or police interview.
Separate the forensic questions before drawing a conclusion
Forensic toxicology requires a clear chain from the object measured to the claim being made. The study measured an instrument display and collected questionnaire responses. It also measured a brief consumption choice. It did not measure a participant’s venous whole-blood ethanol and it did not assess standardized driving, divided attention, balance, reaction time, or clinical causation.
| Question | Relevant evidence | What this experiment can establish |
|---|---|---|
| Was ethanol analytically detected? | Validated breath, blood or other specimen method with timing and reporting rules | All recorded breath readings were 0.00; blood was not tested |
| How much ethanol was in the body? | Quantified specimen concentration with matrix and uncertainty | No participant-level venous BAC or ethanol dose quantification |
| Did a participant feel intoxicated? | Specific self-report at a known time | Placebo increased lightheadedness as one subjective dimension |
| Did task performance worsen? | Validated task-specific performance data | Not measured in this study |
| Did alcohol cause a collision, injury or disputed act? | Case evidence, alternative causes, temporal sequence and relevant performance data | Not investigated |
What makes the study informative, and what limits it?
Strength: a participant serves as their own control
The within-participant placebo versus control comparison reduces some between-person variation in baseline characteristics. Separate pre- and post-drink ratings further clarify what changed over a session. Objective breath checks help distinguish the expectancy manipulation from an ordinary pharmacological ethanol dosing study.
Limitation: the comparison was not chemically ethanol-free in an absolute sense
The vodka applied to the glass rim and misted onto the surface matters. The authors reported 0.00 breath readings, but did not quantify trace exposure or establish analytically that the administered placebo contained no ethanol at all. Saying “zero BAC” as a shorthand for a displayed zero requires retaining this qualification, particularly in litigation.
Limitation: young, selected social drinkers rather than a forensic case population
Most volunteers were women and the sample was recruited from an undergraduate population of beer-drinking social drinkers. The reported mean AUDIT score was 12.78 and mean self-reported weekly consumption was 22.74 UK units. The findings should not be generalized automatically to all ages, abstainers, patients, alcohol-dependent persons, motorists, or people using other drugs.
Limitation: no direct performance or legal-threshold validation
Questionnaires and a taste-test choice are legitimate outcomes for the authors’ psychological research question. Neither is a psychomotor battery. The study contains no sensitivity or specificity for identifying ethanol exposure or functional impairment from lightheadedness; it cannot supply a cutoff, a BAC conversion, a safety interval, or a probability that an individual was intoxicated.
Frequently asked questions
Can someone feel lightheaded while a breathalyzer displays 0.00?
Yes. In this experiment lightheadedness ratings increased after the placebo while reported breath readings were 0.00. The finding concerns subjective lightheadedness, not measured driving or motor impairment.
Did the researchers administer absolutely no ethanol?
The principal drink was lemonade made to seem alcoholic, but vodka was applied to the rim and misted onto its surface. The study reported 0.00 breath readings and did not quantify any trace ethanol exposure; describing it as an unequivocally ethanol-free dose would be too strong.
Does this prove that a person’s alcohol symptoms are imaginary?
No. Subjective experience is itself a reportable phenomenon. The study shows how expectation can influence selected reports under controlled conditions. It does not determine the source of any individual’s actual symptoms.
Did placebo produce genuine performance impairment?
Not established here. Researchers did not measure driving, psychomotor testing, standardized sobriety performance or injury. The effect on lightheadedness cannot substitute for those outcomes.
Was the beverage consumed in the taste test alcoholic?
No. The investigators used a nonalcoholic beer alongside a nonalcoholic orange drink. The increased percentage of test beer consumed should not be reported as a measured increase in grams of ethanol consumed.
Did a bar environment intensify the placebo effect?
Not for the main craving, lightheadedness and test-consumption results. Participants in the bar-like setting did, however, give higher estimates of how much vodka they believed the placebo contained.
Can this paper estimate BAC or determine fault in a DUI or injury case?
No. It supplies neither a quantified blood concentration nor tested driving performance or event causation. A case-specific analysis would need appropriate specimen records, timing, observations and alternative explanations.
Source and research pathway
Primary study: Christiansen P, Townsend G, Knibb G, Field M. Bibi ergo sum: the effects of a placebo and contextual alcohol cues on motivation to drink alcohol. Psychopharmacology. 2017;234:827–835. DOI: 10.1007/s00213-016-4518-0. The analysis above is limited to this complete primary research article from the supplied Ethanol Papers Batch 16 collection.
For related same-site reading, use the research library and its question-led evidence paths, the brief distinguishing measured alcohol change from estimated earlier BAC, and the site’s source and editorial policy. These pages provide orientation, not independent confirmation of the study’s experimental results.
Editorial disclosure: AI-assisted appraisal of the complete primary study. No personal review or adoption of this article by Okorie Okorocha is represented. Educational research context only; not medical advice, legal advice, or an opinion about an individual case.