Alcohol and Sleep: What a Nightcap Actually Does

The nightcap is the most persistent piece of sleep folklore I know, and unlike most folklore it is half true. Alcohol really does put you to sleep faster. The problem is everything that happens after, and none of it is visible from inside the sleep it ruins. In sleep hygiene I gave alcohol four paragraphs. It earns a full accounting, because this is the one factor where consumer trackers and the clinical literature agree completely.

Why the nightcap feels like it works

Ebrahim’s 2013 review of 27 studies found that alcohol at all doses shortened sleep onset latency. At low doses it also increased slow-wave sleep in the first half of the night. If you stopped measuring at 2 am, alcohol would look like a sleep aid.

This is the honest core of the myth, and it is worth stating plainly rather than pretending drinkers are imagining things. A glass of wine genuinely makes falling asleep easier. So does a sleeping pill, and the comparison is closer than most people would like: sedation is not the same thing as sleep.

What happens in the second half of the night

Your liver clears roughly one standard drink per hour. Two glasses of wine at 9 pm are gone from your blood by about 1 am, and that is when the bill arrives. As alcohol metabolises, the nervous system rebounds in the opposite direction: heart rate up, arousals up, sleep fragmenting into pieces.

This is the mechanism behind the 3 am wake-up that regular drinkers know precisely and rarely connect to the wine. The first half of the night was sedated and deep. The second half is shallow, hot and interrupted. On the clock you were in bed for eight hours; physiologically you had two different nights, and the second one was bad.

The practical consequence follows directly from clearance rates: the later the drink and the larger the dose, the more of the rebound lands inside your sleep instead of before it.

What alcohol does to REM

In Ebrahim’s pooled data, moderate and high doses reduced REM sleep in the first half of the night, dose-dependently. REM rebounds later once alcohol clears, but a fragmented second half is a poor place to make up the deficit.

I hold the REM claims slightly more loosely than the fragmentation claims. REM percentages vary a lot between people and between nights, and “REM matters for memory and mood” is one of those statements that is directionally solid and quantitatively fuzzy. What is not fuzzy: a night of drinking reorders your sleep architecture, and no dose in the review left it untouched.

What your tracker will show

This is where the field data comes in. Pietilä’s team analysed 4,098 Finnish employees wearing heart-rate monitors on ordinary nights, drinking or not, in their own homes. Overnight physiological recovery dropped by 9.3 percentage points after low alcohol intake, 24 after moderate, and 39.2 after high. The effects held for both sexes, and being physically active or young did not soften them. If anything, younger participants took the bigger hit.

A separate polysomnography study by Greenlund confirmed the mechanism: evening binge drinking suppresses cardiovagal tone and baroreflex function through the night. In plain terms, the system that is supposed to idle your heart overnight does not get to idle.

For once, your watch and the literature tell the same story. Two beers show up as elevated resting heart rate and flattened HRV until the early morning. Among everything a consumer tracker measures, evening alcohol is the single cleanest signal.

How much and how late is acceptable?

The uncomfortable honest answer from the Finnish data: there was no visible floor. Low intake, defined there as under about half a gram of alcohol per kilogram of body weight, still cost 9 points of recovery. One large beer for an 80 kg adult sits right around that boundary.

Timing is the lever you actually control. Working from clearance at one drink per hour, a practical rule: finish your last drink at least three hours before bed, four if it was more than two drinks. A glass of wine at 6 pm with dinner is metabolically a different event from the same glass at 10 pm, even though both count identically in a weekly units tally.

I am not going to pretend the optimisation answer is anything other than zero in the evening. But between zero and the 10 pm nightcap there is a lot of ground, and moving the same drink three hours earlier captures most of it.

The same two glasses, three different nights

Because clearance runs at roughly one drink per hour, the identical dose produces different nights depending on when it lands. Two glasses of wine, 11 pm bedtime:

The weekly tally treats these as the same event. Your night does not. The 6:30 pm version costs little; the 10:30 pm version delivers the sedated first half, the 3 am rebound, the elevated heart rate, the whole programme described above.

This is also why “I only had two” and “wine ruins my sleep” are both accurate reports from the same person on different evenings. Before concluding you tolerate evening alcohol, check which row of this table your evenings actually occupy.

Alcohol and snoring

Alcohol relaxes the upper airway muscles, which is why it reliably worsens snoring and breathing pauses. If your partner reports that you snore after wine and not otherwise, treat it as a free diagnostic: it means your airway is close enough to the edge that a depressant tips it over. The screening signals worth taking to a doctor are listed in sleep and longevity.

What we do not know

Whether regular drinkers develop partial tolerance to the sleep effects is genuinely unclear; the rebound mechanics suggest the fragmentation should persist, but good long-term data is scarce. The REM findings come mostly from single-night lab protocols. And the recovery scores in the Finnish study are a manufacturer’s composite, not a clinical endpoint, so I treat the exact percentages as softer than their direction.

Where to start

Run the two-week version: no alcohol within four hours of bed, everything else unchanged, resting heart rate and wake-ups logged each morning. Most people find this test more persuasive than any study, because the effect is large enough to see in a week of your own data. If you are simultaneously chasing a caffeine cutoff, change one thing at a time, in the order laid out in the sleep protocol.