Mike, 40, Tested Red Wine for 14 Nights. His BRI Doubled.

One glass of cabernet, four nights a week, after dinner. Mike had done this for years. He'd read the research that said alcohol was bad for sleep. He didn't think it was bad for him personally — until he tested it.

This is what he found.

Who Mike is, in one paragraph

Mike (a Lab Member, his name is changed) is a 40-year-old software engineer in the Pacific Northwest. He's not overweight by any clinical measure. His wife says he doesn't snore loudly. He sleeps roughly 7 hours per night. By every conventional measure, his sleep was fine.

What he did notice: he was tired in the mornings about half the time. Not catastrophic — just enough to wonder.

He'd been wearing an Apple Watch for two years. He installed SomniSense after his Watch flagged a possible breathing pattern overnight. The first week of SomniSense data showed his BRI averaging 14 — borderline mild. Some nights it spiked to 18 or 19. Some nights it was 6.

He couldn't see the pattern just by looking. So he ran an experiment.

The setup

Mike used SomniSense's Lifestyle Lab™ feature to tag every night for 14 nights with a single binary variable: did he have a glass of red wine with dinner, yes or no?

The protocol:

• 7 nights with one 200ml glass of cabernet, finished by 7 p.m.
• 7 nights with no alcohol at all
• All other variables held as constant as possible: same bedtime (10:30 p.m.), same wake time (6:30 a.m.), same partner, same bedroom temperature, no other major changes
• Apple Watch tracked sleep duration; SomniSense tracked breathing events

He didn't randomize the order. He alternated. (For a true randomized design he'd have flipped a coin each evening. He didn't, and it would have been better if he had — that's a real limitation.)

The data

After 14 nights, Mike's Lifestyle Lab™ Factor Impact view showed:

That's an absolute difference of 9.3 BRI points. In percentage terms: wine nights were exactly twice as bad as dry nights for him.

The Apple Watch data complemented this:

Two different data sources, same direction. Wine nights worse on both.

The SomniSense audio playback let Mike hear what was happening: on wine nights, his snoring was louder, more sustained, and occasionally punctuated by 11–14 second breathing pauses he'd never heard before.

What he changed

Mike cut wine on weeknights. His wife noticed within a month — "you seem like yourself in the morning again."

His BRI now averages 9–11 on weeknights. On Saturday nights when he still has wine, it's typically 14–17 — the pattern holds. He's accepted that as a trade-off he wants to keep making.

His Apple Watch HRV has also trended upward over the same window. Both data layers move together when he changes the input.

What was actually happening to his body on those wine nights

Mike's data showed the trend. The mechanism behind it isn't mysterious if you know what alcohol does to a sleeping body. I'll be specific because most "alcohol is bad for sleep" articles stop at the platitude:

1. Hours 0–4 after the drink: the body is metabolizing the alcohol in the liver. It's a heavy job. Heart rate stays elevated, you dream more, you don't drop into deep sleep as easily as on a dry night. Most people don't notice this part — they're asleep.
2. Hour 4–6: the liver finishes the job. The sedative effect wears off. The body, mid-night, decides it's awake. People who drink before bed know this wake-up — usually somewhere between 2 and 4 a.m., wide-eyed, hard to fall back asleep.
3. The whole night: alcohol relaxes the muscles in your throat. The same muscles that hold your airway open. So if you're someone whose airway is borderline-stable on a sober night, alcohol pushes it from "narrow but vibrating" toward "narrowed enough to collapse." Snoring goes louder, breathing pauses go more frequent, blood oxygen dips harder.
4. The morning: people wake up foggy and blame the impurities in the drink. Some of that's real. But the bigger driver, in the data we keep seeing, is overnight oxygen drops from a worse breathing pattern. The headache and the fog are downstream of the airway, not the wine itself.

None of which Mike thought about for years. He drank, he slept, he woke tired, he assumed it was age or stress. That's the gap his experiment closed.

What this experiment was — and wasn't

This was an n=1 self-experiment. It is not a clinical trial. The data is observational, not blinded, not randomized in any formal sense. There was no placebo (you can't blind someone to whether they had wine).

Mike's body is not a population. The same wine had no measurable effect on his wife's sleep when she ran a parallel experiment a month later. Hers doesn't snore much, doesn't have breathing pauses, and her sleep efficiency was steady regardless of wine. That's the point of running the experiment on yourself: your body is what matters.

If the same wine had had no effect on Mike's data, that would also have been useful information — and it would have meant his mornings-tired pattern had a different cause worth investigating (caffeine timing? stress? thyroid?). The experiment that doesn't move your number is also a finding.

Should you try this

If you drink alcohol regularly and you're curious whether it affects your sleep, the experiment is reproducible:

1. Install SomniSense, use Lifestyle Lab™ tagging
2. Pick a 14-night window where you can actually control the variable
3. Alternate (or, better, randomize) wine vs no-wine nights
4. Tag each night honestly — including the nights you cheat
5. Read your Factor Impact view at the end

You may discover, as Mike did, that the effect is large and consistent. You may discover the opposite. Either is useful. The point isn't the answer — it's having an answer that's about you, not the average user.