It helps you fall asleep and then makes the second half of the night worse, and the evidence for that is more contested than almost any coverage admits. The most cited review reports that at all doses, alcohol reduces the time taken to fall asleep, produces a more consolidated first half of the night, and increases sleep disruption in the second half. Two years later, eight sleep researchers published a rebuttal in the same journal arguing that the review’s design, methods and statistics could not support its conclusions.
So the honest answer is that alcohol appears to do two opposite things in sequence, the pleasant one comes first, and the leading account of it is disputed by specialists.
What the review reported
The reference point is a 2013 review in Alcoholism: Clinical and Experimental Research by Ebrahim and colleagues, which described itself as a qualitative assessment of all known scientific studies on the impact of alcohol on nocturnal sleep in healthy volunteers.
Its core findings: at all dosages, alcohol reduces sleep onset latency, meaning the time taken to fall asleep, produces a more consolidated first half of sleep, and increases sleep disruption in the second half.
The onset of the first REM sleep period is significantly delayed at all doses, which the authors describe as the most recognisable effect of alcohol on REM sleep. Total night REM percentage is decreased in the majority of studies at moderate and high doses, with no clear trend at low doses. In the first half of the night, REM effects appear dose-related: no clear trend at low and moderate doses, significant reduction at high doses.
The majority of studies, across dose, age and gender, confirm an increase in slow-wave sleep, the deepest stage, in the first half of the night compared with baseline. That last claim is the one the rebuttal attacks hardest.
The rebuttal, and what it actually argues
In 2015 the same journal published Alcohol and Sleep Review: Flawed Design, Methods, and Statistics Cannot Support Conclusions, by Pressman and seven co-authors, several of them prominent sleep researchers.
Their opening position: their review of the article found it to be seriously flawed by research design and statistical problems.
The specific charges are worth having, because they are more concrete than the title suggests.
On selection. The 2013 review states in its abstract that it assessed all known scientific studies of the effects of alcohol on the sleep of healthy volunteers. The rebuttal points out that Ebrahim and colleagues selected 20 published articles, and that numerous articles were excluded from consideration. Within those 20 were 38 groups of subjects, split by criteria such as sex or alcohol dose.
On the deep sleep claim specifically. The rebuttal notes that one of the studies inside the review, Feige and colleagues from 2006, reports a non-significant decrease in slow-wave sleep percentage on the first alcohol night, which runs against the direction the review reports.
On method. The rebuttal argues that a meta-analysis approach is necessary to provide reliable conclusions about the relationship between alcohol and slow-wave sleep, rather than the qualitative synthesis the 2013 review performed.
That is a substantive disagreement about whether the headline finding survives its own evidence base, not a quibble. Anyone repeating the deep sleep claim should know it is contested by eight specialists in the journal that published it.
A separate risk that is better established
One effect of alcohol on sleep sits on firmer ground, because it was studied differently.
A 2018 systematic review and meta-analysis in Sleep Medicine examined the association between alcohol consumption and sleep apnoea across comparative epidemiological studies. It concluded that alcohol consumption is associated with a higher risk of sleep apnoea, and that this supports evidence that reducing alcohol intake has potential therapeutic and preventive value in the condition.
Note the shape. This is an association from epidemiological studies, not a causal finding, and the same analysis found mean alcohol intake was two units per week higher in those with sleep apnoea but that this difference was not statistically significant. The review’s own numbers are more equivocal than its conclusion sentence.
Still, for anyone who snores heavily, wakes gasping, or has been told they stop breathing at night, this is the part of the literature most worth taking seriously, and a reason to raise drinking with a doctor rather than adjust it privately.
Why the nightcap survives the evidence
There is a timing mismatch worth naming, because it explains why personal experience and the research disagree so consistently.
The part of the night a drink improves, falling asleep and the first stretch, is the part you are awake for and will remember. The part it degrades happens when you are least able to attribute it to anything. A fragmented four in the morning does not feel like it was caused by a drink at ten.
That asymmetry is our reading rather than a finding from the sources, and it survives the rebuttal, since nothing in the dispute concerns the sleep onset finding. It also connects to what a bad night actually costs you, where the damage falls hardest on sustained attention rather than on complex thinking, and to whether trackers help or hurt when you try to measure any of this yourself.
When this stops being a sleep question
If you are drinking in order to get to sleep, and particularly if you now need it or sleep badly without it, that is a different situation from the one this article describes.
Alcohol dependence and insomnia commonly occur together and each makes the other harder to treat. A doctor is the right place for that, and it is a common enough conversation that it will not surprise anyone. If persistent insomnia is the underlying problem, there are treatments for it with considerably better evidence than anything discussed here.
Common questions
Why do I fall asleep faster after a drink?
Because that part is real and not disputed. A 2013 review of studies on alcohol and nocturnal sleep in healthy volunteers found that at all dosages, alcohol reduces sleep onset latency and produces a more consolidated first half of sleep, followed by increased sleep disruption in the second half. The onset of the first REM period is significantly delayed at all doses, which the authors call the most recognisable effect of alcohol on REM sleep. Note that this review drew a formal rebuttal in 2015 from eight sleep researchers who found it seriously flawed by research design and statistical problems, though their objections centre on the deep sleep findings rather than on sleep onset.
Does alcohol really increase deep sleep?
This is the contested claim. The 2013 review reported that the majority of studies confirm an increase in slow-wave sleep in the first half of the night. A 2015 rebuttal in the same journal, by Pressman and seven co-authors, argues the review was seriously flawed by research design and statistical problems. Specifically: the review claimed to assess all known scientific studies but selected 20 published articles with numerous articles excluded from consideration, and one study inside it reported a non-significant decrease in slow-wave sleep percentage on the first alcohol night. The rebuttal argues a meta-analysis is necessary to reach reliable conclusions on this. Treat the deep sleep claim as disputed rather than established.
How much does it take to affect sleep?
The 2013 review reports effects at all doses for some measures and dose-dependent effects for others. Reduced sleep onset latency, a more consolidated first half, increased second-half disruption, and delayed first REM onset were reported at all dosages. Reduced total-night REM appeared at moderate and high doses, with no clear trend at low doses, and first-half REM reduction was significant only at high doses. This article gives no threshold or unit guidance on purpose, because the review’s dose categories do not translate cleanly into drinks, and because that is a clinical conversation if it matters to you. The review’s methods were formally contested in 2015.
Is alcohol worse for sleep if I snore?
This is the part with the most direct clinical relevance and the least dispute. A 2018 systematic review and meta-analysis found alcohol consumption is associated with a higher risk of sleep apnoea, and concluded this supports evidence that reducing intake has potential therapeutic and preventive value in the condition. Read the qualifier: this is an association from epidemiological studies, not a causal result, and within the same analysis the difference in mean alcohol intake between those with and without sleep apnoea was not statistically significant. If you snore heavily, wake gasping, or have been told you stop breathing during sleep, that warrants a doctor’s appointment regardless of drinking, since untreated sleep apnoea has consequences well beyond tiredness.
The half of the night you do not remember
Two things make this topic generate such confident and contradictory advice.
The first is structural. The nightcap works, for about three hours, and then reverses in the stretch of the night you are least equipped to notice or attribute. The evidence you gather personally is systematically weighted towards the good half. That is not a failure of attention on your part; it is a feature of when the effects land, and the same asymmetry is why a daytime nap is easier to evaluate honestly than a nightcap.
The second is that the science itself is less settled than its popular version. The review everyone cites was formally challenged by eight sleep researchers who said its design, methods and statistics could not support its conclusions, and who noted it claimed to cover all known studies while selecting twenty. That exchange sits in the same journal and almost never travels with the finding.
None of which tells you whether to have the drink. It tells you that both the feeling of it working, and the confident article saying it does not, are weaker evidence than they appear.
Leave a Reply