Weeks 9 to 13 With an Anti-Snoring Device: Two Sleeper Trains, One Failing Pad, and a Spreadsheet

I have never seen my brother-in-law wear his anti-snoring device. I have never been in the same house as him while he slept, never heard him snore, and five months ago I had no opinion about whether any of these things do anything at all.

What I have is fourteen weeks of spreadsheets.

Here is how that happened. My sister is the one who lost sleep for years, and she refuses to keep logs — her words, not mine: “I’m not turning my marriage into homework.” He’s the opposite. He likes data. So the arrangement landed on me: every morning he sends me four lines (bedtime, hours worn, what the app recorded, how he felt on waking out of five), and every Sunday night we spend twenty minutes on the phone going through the week. I keep the file. It started as a favour in week one and has quietly become the most interesting document on my laptop.

So this is not a review written by someone who watched it happen. It’s a relay. Every number below came from someone else’s notes, forwarded to me, plus the export he sends from the app every fortnight. Please read it with exactly that much authority and no more.

Weeks 1 to 8, in one paragraph

He was aware of it for about nine or ten nights, then stopped noticing it. It woke him up whenever he rolled over for roughly the first three weeks, and then it didn’t. He never once found it off his neck in the morning. His own summary, sent in week eight, was almost annoyingly plain: “It’s not silent, it’s manageable. That’s the whole thing.” My sister’s summary, sent separately and without being asked: “I sleep through now. I’m not going to say more than that.”

Then month three happened, and nothing about his behaviour changed

This is the part I actually wanted to write down, because it took three weeks to diagnose and the diagnosis was boring.

In week nine the numbers started creeping. His month-two average had been 27 snoring minutes a night. Week nine came in at 41. Week ten was 38, then 44. His morning scores went from a steady 4 to a lot of 3s. No change in bedtime, no change in how long he wore it, no change in where he slept.

He blamed hay fever. Then he blamed a heavier duvet. Then he bought a new pillow because he’d read that pillow height matters, spent a week on it, and the numbers didn’t move. I asked him on the third Sunday whether he’d changed anything at all about the device itself, and there was a pause on the line that I have learned to recognise.

Week eleven he went and found the spare contact pads that had come in the box. He laid the original pad he’d been using since day one flat on the kitchen table next to a new one, and sent me a photo. The old one had gone slightly yellow at the edge, and more importantly it no longer sat flat — it had taken on a curl, like a book cover left open. Three months of nightly wear, nightly wiping, and body heat had given it a shape of its own.

He put a new pad on that Tuesday. Two nights later he was back at 24 and 26 minutes. By the end of that week he had the best five-night average in the entire file: 22.6 minutes a night, against 36 during the three flat weeks before the swap.

I want to be careful here, because one person changing one pad is not a study. But the reason I believe the pad was the cause is that he changed nothing else. The hay fever was still there. The heavier duvet was still there. The new pillow was still on the bed. The only variable that moved between the bad three weeks and the good two was a piece of soft material about the size of a thumb.

Nine days in the middle of it, as they arrived

The window below covers the pad swap and, four days later, an eleven-hour overnight train he takes twice a week for a site job. I’ve kept the entries as short as his messages were.

Day 1 — home, first night on the new pad

Notes at 07:04, four lines as always. 7.6 hours worn. 31 events, 22 minutes. Waking score 4. One extra line, unusual for him: “sits flatter against the neck. in a good way.”

Day 2 — home

7.9 hours. 26 events, 18 minutes. Score 4. Nothing to report, which after three weeks of forty-minute nights is the entire report.

Day 3 — home, worst night of the month, and it doesn’t matter

6.2 hours. 34 events, 21 minutes. Score 3. Cause was obvious and unrelated: a thunderstorm, the dog, and a delivery van at four in the morning. I’m leaving it in the table because I’ve stopped treating any single night as information. One bad night is weather. Three bad weeks is a part that needs replacing.

Day 4 — packing night

He sends a photo of the travel case: device, one spare pad left after Tuesday, the small brush, the short cable, and a folded corner of the manual with a diagram circled. He’s on the 23:40 overnight train the next night, eleven hours, and he’s taking it, which is new. In month one he’d have left it at home rather than wear it in a compartment with strangers.

Day 5 — eleven hours, lower berth

Four-berth compartment, curtain, three people he’d never met. He put it on after the lights went down. 6.8 hours worn. 29 events, 19 minutes. Score 3.

Two things worth keeping. First, the pillow problem came back — a train pillow plus a folded jacket, which is exactly the chin-down angle that cost him sleep in week three. The first ninety minutes were the noisiest of the whole window. Second, at six in the morning the man across from him asked what it was, and the conversation lasted as long as it takes to boil water. Nine weeks ago that question would have been a whole thing. On day five it was a thirty-second answer.

Day 6 — arrived, own pillow, back to normal

7.1 hours. 24 events, 15 minutes. Score 4. His one-line message: “own pillow. normal. the train wasn’t the device.”

Day 7 — eleven hours back, upper berth

Upper berth this time: closer ceiling, warmer, worse airflow, and he says he’d have guessed a worse night. 7.0 hours. 26 events, 17 minutes. Score 4.

The only difference in method: this time he put it on twenty minutes early, before the compartment filled up, and read for a while. He described it as the difference between doing something in front of people and doing something you’ve already finished. Same object, same twenty seconds. Different night, entirely because of the order he did it in.

Day 8 — home, the longest night in the file

8.1 hours. 22 events, 13 minutes. Score 5. My sister’s comment, forwarded to me verbatim without context: “two nights on a train and he comes home quieter than he left.” I have no idea whether she meant the snoring or his personality.

Day 9 — close of the window

7.4 hours. 25 events, 16 minutes. Score 4. That makes the five nights since the pad swap average 22.6 minutes, against 36 minutes a night for the three weeks before it.

The numbers, with the bad rows left in

Every figure below is from his app export or his own morning note. I’ve rounded nothing that would flatter him.

One honest caveat on the two train rows: the minutes look good partly because he slept less overall. Nineteen minutes inside six-point-eight hours isn’t the same performance as eighteen minutes inside seven-point-nine, and I’m not going to pretend it is. What those two rows do show is that the thing survived a strange berth, a stranger’s pillow, and a curtain, and still landed roughly where his own bed lands.

The maintenance schedule he now actually keeps

This is the only part of the file I’d send to someone who already owns one. It came out of month three the hard way.

• Every night (about 15 seconds): wipe the contact surface before it goes on, and press the pad flat with a thumb once. If it feels like it’s rolling at the edge, that’s the beginning of the problem.
• Every week: rinse the pad, air-dry it away from heat — never on a radiator or in a car in the sun — and check the strap against a small mark he drew on it in week one. Tension creep is real and invisible.
• Every month: lay the pad flat on a table edge. If it curls or lifts instead of lying down, it’s done. This is the thirty-second test that would have saved him three bad weeks.
• Every three months: replace the pad on a schedule rather than on symptoms. Keep the old one in a bag until the new one has five nights on it, so you’ve got something to compare against.
• Before any trip: one spare pad, not two. He took three on the train and used none, but the case is small enough that one costs nothing.

What I’d pass on, as the person who only reads the notes

• The most likely reason “it stopped working” is that something soft wore out. Not the battery, not the app, not the person. A pad that has taken on a set shape stops sitting the way it used to, and the decline is slow enough — roughly a third worse over three weeks in his case — that you’ll blame the weather, your weight, or your pillow first.
• Give a new pad three nights before you judge it. Night one was good, night two was better, night three was a thunderstorm. Only the five-night average meant anything.
• On a train, put it on before the compartment fills up. The object is the same at 23:30 and at 23:50. The feeling is not.
• Take your own pillow setup if you possibly can. It’s the one variable that followed him across eleven hours, two berths, and a hotel, and it was worth more than any setting he’s ever changed.
• Keep the log, even a lazy one. Four lines a day is what turned three confusing weeks into one thirty-second explanation. Without it, he’d still be blaming the duvet.

The bit I keep coming back to

My sister spent somewhere around eleven years being the only person in that house who knew how bad the nights were. Now her entire contribution to this file is one forwarded sentence every few weeks, and the sentence on day eight was that he came home quieter than he left.

I can’t tell you what any of this would look like for you. I’ve never worn one, never watched anyone wear one, and my whole view of the thing is built out of someone else’s four-line messages and a spreadsheet I maintain for free. But if you’re at month three and the numbers are drifting, the answer is probably smaller and cheaper than you think, and it’s probably soft, and you can test it in thirty seconds on a kitchen table.

(Names and details changed; the numbers are one person’s, self-reported, from a consumer app — not a sleep study. Snoring and sleep quality swing wildly night to night for reasons that have nothing to do with any device, and outcomes differ from person to person. Nothing here is medical advice. If you or someone in your household snores heavily, pauses in breathing during sleep, or is exhausted during the day, the right next step is a doctor, not a gadget.)