Everyone Slept Better in This Sleep App Trial. That’s the Problem

Rain on a tin roof. A soft voice reading a story about a quiet harbour.

Pink noise humming in the dark.

Lots of people now fall asleep to their phones. Sleep sounds and bedtime stories are widely used features in wellbeing apps.

But do they actually help you sleep? Or does it just feel that way?

A large trial in the journal Sleep put them to a fair test. The result is a little awkward for the sleep-app industry.

Everyone slept better. Including the people who got no sleep content at all.

The experiment: four groups, one clever catch

Who took part

The team, from the University of Sussex and the mental health app company Unmind, recruited working adults online.

Everyone had real sleep trouble. On average they slept 6 hours and 14 minutes a night.

About 6 in 10 took more than half an hour to fall asleep. About 7 in 10 had at least mild symptoms of anxiety or depression.

In total, 495 people from the UK and the US took part. Their average age was 33.

What each group got

Each person was randomly placed in one of four groups for four weeks. All of them used the same app.

The sleep groups were asked to use one track a night. They could replay a favourite or try something new.

The clever catch

Many app studies compare users with people on a waiting list. That's an easy bar to clear.

This one was tougher. The control group also used the app, often, with the same player and the same kind of experience.

The only real difference was the "active ingredient": content designed for sleep.

Why the comparison matters so much

Think of it like testing a new painkiller. If you compare it with doing nothing, almost anything looks good.

People who get something expect to feel better, and often do. A waiting list gets none of that boost.

The authors suggest earlier app studies with weaker designs may have overestimated how much sleep content helps. This trial was built to strip that boost out.

What they found

After four weeks, every group reported much less sleep disturbance. The sleep tools just didn't beat the control.

The main measure was a standard sleep questionnaire. On its scale, 50 is roughly the average adult, and higher means worse sleep.

In plain terms, most groups went from noticeably worse-than-average sleep to roughly average. That's a real improvement for everyone.

The sleep groups came out about one point ahead. That gap was small and could easily be chance.

The same pattern held for daytime tiredness, anxiety and low mood, wellbeing, work productivity and a racing mind at bedtime. Everyone improved, and nobody pulled clearly ahead.

The one bright spot: bedtime stories

There was one exception worth mentioning.

People in the bedtime stories group reported sleeping 53 minutes longer by the end. The control group gained 33 minutes.

That extra 20 minutes was statistically significant. Sleep sounds (43 minutes) and sleep skills (25 minutes) didn't differ from the control.

It's one positive result among many measures, and it's self-reported. So treat it as a lead, not a win.

Falling asleep faster, in every group

At the start, about 6 in 10 people needed over 30 minutes to drop off. After four weeks, it was about 3 in 10.

But again, the control group improved just as much.

So why did everyone sleep better?

The trial can't say for sure. But there are a few likely suspects.

Expectations. Signing up for a sleep study and using a wellbeing app every day can make you expect better sleep. Expectations are powerful.

A routine. Spending a few calm minutes with an app may work as a wind-down ritual, whatever the content.

Starting at a low point. People often join a study when their sleep is at its worst. The authors found that those with the worst sleep at the start improved the most, in every group.

Paying attention. Filling in sleep questionnaires and using a wellbeing app for a month may itself nudge your habits.

None of these needs a rain soundtrack. That's the uncomfortable part for sleep apps.

Who ran this, and why it matters

Here's a detail I really respect.

Several authors have ties to Unmind, the company that makes the app. One author's PhD is part-funded by it, and others consult for it or used to work there.

In other words, a company-linked team tested its own sleep features and published a disappointing result. You don't see that every day, and it makes the finding more believable, not less.

Interestingly, people rated the control content as less convincing at the start. Yet their expectations of improvement were similar across groups, and they ended up sleeping about as well.

The authors themselves write that their results don't support recommending these tools on their own for sleep problems.

How much should you trust this?

Promising. That's our label for how solid the evidence is, not for the apps. This is one well-designed trial, so it isn't the final word.

What makes it convincing

• It was randomised, with a large sample of 495 people.
• The control used the same app, so expectations were closely matched.
• It was listed in a public trial registry, and 82% of people completed it.
• People used the app at home, in real life.

What makes me cautious

• It lasted four weeks. Longer use might show different effects.
• Sleep was self-reported. There were no sleep trackers or lab tests.
• Use of the sleep tools was fairly low, and the control group actually used its content more.
• Each person was stuck with one type of audio. In real life, you pick what you like.
• It tested one app's library, which may be smaller than those of dedicated sleep apps.
• Recruitment stopped at 495 instead of the planned 600, for budget reasons.

What this means for you

If you love falling asleep to rain sounds or a story, keep going. The authors don't discourage it, and the tools were generally safe.

About 2 in 100 people had meaningfully worse sleep scores by the end. A few others mentioned downsides like relying on the app, loud volume or boredom.

Just don't expect the audio to fix a real sleep problem on its own.

What about using your phone in bed?

Standard advice is to put screens away before bed. The NHS says to avoid phones right before sleeping, because the light can make you more alert.

This trial adds a small twist. People were asked to use an app at bedtime, and their sleep still improved.

The authors suggest calm, purposeful app use may be less disruptive than other bedtime phone use, like endless scrolling. That's their interpretation, not something the trial tested directly.

If you do use audio, one simple approach is to start a track and put the phone face down, rather than scrolling.

If sleep trouble goes on for months or affects your daily life, the NHS advice on insomnia is a good start. It suggests seeing a GP, who may offer cognitive behavioural therapy (CBT) for insomnia, face to face or online.

The same page lists habits with better evidence behind them:

• Go to bed only when you feel sleepy.
• Get up at the same time every day.
• Relax for at least an hour before bed.
• Keep your bedroom dark and quiet.

For a big-picture look at why sleep matters so much, this talk by sleep scientist Matt Walker is a popular primer:

What we still don't know

1. Does choice matter? People who pick their own sounds or stories might do better.
2. Do longer periods help? Four weeks may be too short.
3. What happens with objective measures? Sleep trackers could show changes that questionnaires miss.
4. Is it real for bedtime stories? The extra 20 minutes needs its own test.
5. What about diagnosed insomnia? This trial studied people with sleep trouble, not a clinical group.
6. Why did everyone improve? Understanding that could help design better tools.
7. Does it hold beyond study volunteers? People were recruited through an online research panel, and most were fairly young and well educated.

My take: the ritual may matter more than the soundtrack

I love this kind of study. It asks a simple question almost nobody bothers to test properly.

The answer isn't "sleep apps are a scam". Everyone in this trial slept better.

It's that the specific sleep content didn't seem to be the magic ingredient.

So if a nightly audio routine helps you switch off, it may be the routine doing the work. And if your sleep problems are serious, the best-tested help is still CBT for insomnia, not a playlist.