A drink before a long-haul sleep does not buy you better rest. In a controlled altitude chamber, moderate alcohol combined with cabin-level pressure suppressed deep sleep, pushed median blood oxygen down to 85.32%, and raised median heart rate to 87.73 bpm.
The catch: those subjects slept flat on their backs, were all healthy and aged 18 to 40, and drank a fixed two-drink dose. The finding is real. The numbers are not a seat-by-seat forecast.
Alcohol is a sedative. That is pharmacology, not marketing, and it is why a glass of wine before a night flight feels like sensible trip planning — it shortens the time it takes to fall asleep, and on a plane that is worth something. What it does not do is produce the sleep your body is trying to get.
A study in Thorax put healthy volunteers to sleep in a controlled altitude chamber set to the pressure of a cabin at cruising altitude, gave half of them a moderate dose of alcohol, and tracked their sleep stages, blood oxygen, and heart rate. The results are stark enough that most coverage stops there.
The more useful question is where the evidence ends — the study measured something real, but it measured it in a posture, a cohort, and a dose that do not match the person in seat 42C.
What the chamber study actually measured
Forty-eight healthy adults aged 18 to 40 were randomized into two arms — a sleep laboratory and an altitude chamber. Twenty-three from the lab and 17 from the chamber produced data that made the final analysis.
Each participant slept under four conditions: normal pressure without alcohol, normal pressure with alcohol, simulated cabin pressure without alcohol, and simulated cabin pressure with alcohol. The dose was fixed, matching two 5% beers or two 175 ml glasses of 12% wine.
The combined condition is where the numbers turn uncomfortable.
| Condition | Median SpO2 | Median heart rate | Median N3 sleep | Median REM sleep |
|---|---|---|---|---|
| Normobaric, no alcohol | 95.88% | 63.74 bpm | 67.50 min | Not reported in retrieved primary text |
| Normobaric, alcohol | 94.97% | 76.97 bpm | 84.00 min | Not reported in retrieved primary text |
| Hypobaric, no alcohol | 88.07% | 72.90 bpm | Not reported in retrieved primary text | Not reported in retrieved primary text |
| Hypobaric, alcohol | 85.32% | 87.73 bpm | 46.50 min | Not reported in retrieved primary text |
| Source: Thorax | ||||
Read the two hypobaric rows against each other. Cabin pressure alone already drags median SpO2 down to 88.07%. Add alcohol and it falls to 85.32%, while median heart rate climbs from 72.90 to 87.73 bpm — a cardiovascular system working noticeably harder while its owner is unconscious.
Deep sleep tells the same story in reverse: under normal pressure, alcohol raised median N3 sleep to 84.00 minutes, but under cabin pressure it collapsed to 46.50 minutes.
Superdeals cover departures from North America, Europe and Australia.
Flight deals most people never see
Air Traveler Club monitors thousands of routes for pricing anomalies that traditional search engines miss. With ATC Superdeals members save up to 80% on each flight vs. regular fares.
Too good to be true? Check it out yourself. It's free.
Delivered weekly. Unsubscribe anytime.
The three gaps between the chamber and seat 42C
Start with posture, because it is the one almost nobody mentions. The chamber participants slept supine, flat on their backs, while economy passengers mostly do not.
Posture is the variable the paper never resolves. Its participants lay flat, and whether sitting upright would amplify or mute the oxygen and heart-rate changes is left open — which means the chamber medians are not a prediction for a passenger in a seat. That is a real unknown, not a rounding error.
Then there is who was in the study. Every volunteer whose data made the final analysis was healthy and between 18 and 40, which the authors treat as a limit on how far the results travel. They also warn that alcohol at altitude could aggravate symptoms in people with heart or lung conditions, and that older travellers may react more strongly.
Neither of those groups was actually enrolled, though, so that caution is the researchers’ inference rather than a measured result. It is a reasonable one — but treat it as a floor, not a finding.
Finally, dose. The vodka stood in for the drinks, scaled to match two 5% beers or two 175 ml glasses of 12% wine. The study tested one quantity and one kind of drink, and it plotted no dose-response curve.
So a smaller pour, a larger one, or a different drink — none of it has been measured for its effect on oxygen saturation, heart rate, or sleep stages.
Why alcohol and cabin pressure work against each other
The paper, published in Thorax under DOI 10.1136/thorax-2023-220998, takes as its subject the way hypobaric hypoxia and a moderate drink interact for long-haul flyers’ sleep, blood oxygen, and heart rate.
The mechanism is worth understanding, because it explains why the sedation and the recovery pull in opposite directions. At cruising altitude the cabin is already a mildly hypoxic environment. Alcohol compounds that in two ways: it dulls the breathing reflex, so you take shallower breaths when you most need deeper ones, and it disrupts the body’s normal overnight cardiovascular adjustment.
So the night feels like sleep. It runs on a system that is working harder than it should be at rest.
What this means before your next long-haul
The two-drink dose was a study condition, not a personal safety line. Treat it as the only quantity anyone has actually measured here, and nothing more.
For travellers with heart or lung disease, for older passengers, or for anyone unsure about their fitness to fly, the sensible move is a word with a clinician before the trip — not a drink as a sleep aid.
Carriers and regulators already publish the practical version of this advice. On its long-haul services, All Nippon Airways advises drinking water, working the feet and legs every two to three hours, and cutting back on both alcohol and caffeine. The UK Civil Aviation Authority reaches the same conclusion from the regulator’s side: a drink may send you off faster, but the sleep that follows is poorer — which is why it does little to blunt jet lag.
Neither is a response to the Thorax paper — no carrier or regulator has moved to ban or restrict alcohol on the strength of it. And none of it adds up to a posture-specific risk for the passenger in any given seat.
Key terms
- SpO2
- SpO2, or peripheral oxygen saturation, is the share of hemoglobin in the blood that is carrying oxygen, read non-invasively with a pulse oximeter. It is a routine vital in both clinical and in-flight settings, and the Thorax study tracked it alongside heart rate and sleep stage across every condition. In this study, alcohol at simulated cabin pressure pulled the median down to 85.32% — a reading a doctor would normally want to investigate in a resting adult.
- Hypobaric hypoxia
- Hypobaric hypoxia is the pairing of low air pressure with reduced oxygen availability — the condition a passenger breathes inside a pressurised cabin at cruising altitude. The Thorax study reproduced it in a chamber so that sleep, blood oxygen, and heart rate could be measured under something close to real flight conditions. Because the cabin arm produced lower oxygen saturation and higher heart rates than the ground arm even before alcohol was added, it is the half of the experiment that speaks most directly to long-haul travel.
- N3 sleep
- N3 sleep, also called slow-wave or deep sleep, is the deepest stage of non-REM sleep and the period when the body does most of its physical recovery. It is measured in minutes, and the Thorax study recorded a median of 84.00 minutes under normal pressure with alcohol against 67.50 minutes without it. The same alcohol under cabin pressure cut N3 to 46.50 minutes, which is the stage-level evidence behind the article’s point that a nightcap does not buy better rest.
Questions? Answers.
How much alcohol did the study give participants, and did it test any other amount?
The exposure matched two 5% beers or two 175 ml glasses of 12% wine, with vodka used as the drink. No other quantity or drink type was tested, so the study establishes no safe personal threshold and no predictable effect for larger or differently composed drinks.
Did the study include older passengers or people with heart or lung conditions?
No. Older passengers and people with cardiac or pulmonary disease were not directly tested as separate groups — the cohort was healthy volunteers aged 18 to 40. The researchers’ suggestion that those groups might see stronger effects is an inference, not a measured result.
How many people took part in the study, and how many were in the final analysis?
Forty-eight healthy adults aged 18 to 40 were randomized. The final analysis drew on 23 sleep-laboratory participants and 17 altitude-chamber participants, for 40 in total.
Has any real flight been documented as diverted because a passenger drank and slept?
Not in the material retrieved for this piece. No verified named flight, date, and outcome was found linking alcohol, sleeping, and cabin-pressure hypoxia to a specific inflight medical diversion.