← Longevity & Supplement Guides
Phone Before Bed? Watch the Clock, Not the Countdown
Most sleep advice about phones comes as a countdown. No screens for the last hour. Put it down thirty minutes before bed. Almost nobody follows it, and almost everybody breaks it the same way: one more video at 11:40, lights out at 11:45, and a small feeling of guilt for the last five minutes.
A study published this month suggests that the guilt is aimed at the wrong minutes. When researchers tracked hundreds of teenagers night after night, the phone use that went with less sleep was not the scroll in the final moments before sleep. It was the time on the phone after a fixed point on the clock. The phone did not seem to poison sleep. It seemed to eat it, one late minute at a time.
Four hundred teenagers, seven thousand nights
The study, by Yihong Zhao, Lauren Hale and Dimitri Christakis, appeared on 19 October in JAMA Pediatrics. It used 425 young people from a large US cohort, average age about 15, half of them girls. Nobody filled in a diary about their screen time. An app on each phone logged when the screen was in use, and a consumer wrist tracker estimated when each person fell asleep, woke up and lay awake. Together that came to 7,591 nights.
The clever part is the comparison. Most screen-and-sleep research compares people: heavy users against light users. That mixes the phone up with everything else that differs between those people, from school pressure to mood to who has a television in the bedroom. This study compared each teenager with themselves. On the nights when you used your phone more than you usually do, what happened to your sleep compared with your own normal nights?
That design removes a lot of noise. Your personality, your family and your bedroom stay the same from Tuesday to Wednesday. Your phone use does not.
Where the extra minute goes
The researchers looked at phone use after a series of cut-off times, from 8 p.m. to midnight. The later the cut-off, the stronger the link. At 10 p.m. the picture was clean. For each extra minute of use after ten, above a person’s own average, they fell asleep about 0.87 minutes later. They also woke up later, but only by about 0.37 minutes. The difference came out of the night: about 0.43 minutes less sleep for every extra minute on the phone.
A minute sounds like nothing. Nights are not made of single minutes, though. On the nights when a teenager’s late phone use was above their own average, they fell asleep about an hour later than usual and slept about twenty minutes less. Morning has a fixed start for most people. Bedtime does not. So a late evening on the phone pushes sleep later at the front and the alarm clock cuts it off at the back.
One extra minute after 10 p.m.
What changed, per extra minute of phone use above a person’s own average
Within-person estimates from Zhao, Hale and Christakis, JAMA Pediatrics 2026. Bar length shows the size of each change on a shared scale where the full track is one minute.
Clock time
Phone use after 10 p.m. went with later sleep and less of it. The later the cut-off, the stronger the link.
The last minutes
Use in the final few minutes went with slightly more time awake early in the night. Across the final twenty minutes, the link was negligible.
Colour filter, app type
Other studies found no sleep difference from a warm night-mode screen, and no clear difference between social media and other screen use.
Three ways to read the finding
The displacement reading: time on the phone is time not asleep. The minutes move the whole night later.
The other-way-round reading: on nights when you are not sleepy, you pick up the phone. The phone is then a sign of a wakeful night, not its cause.
The useful reading: both are probably true. A fixed phone cut-off on the clock helps with the first, and it costs little if the second is the bigger part.
The last twenty minutes were not the problem
The study also looked at the window most advice obsesses over: the final five to thirty minutes before sleep. Here the link was much weaker. Use in the very last few minutes went with a small increase in time awake during the first half hour of sleep. Looked at across the final twenty minutes, the association was negligible.
That fits an older result that surprised many people. In 2023, a team at Masaryk University in the Czech Republic followed about 200 teenagers for two weeks, with phone logs and daily sleep reports, in Computers in Human Behavior. On nights with more phone use before bed than usual, the teenagers fell asleep slightly earlier and slept slightly longer, not less. The links were weak, and the authors said so. But they point the same way: the screen in the last stretch before sleep is not a strong poison.
I find that a relief. It means the countdown rule, which nobody keeps, may not be the rule that matters.
Light is the smaller part of the story
For years the explanation was blue light. Screens give off short-wavelength light, the body clock is sensitive to it, so the phone delays your clock. Bright light in the evening does shift the body clock, and we covered how in our guide to light exposure and circadian health. A phone held at reading distance is a small light source, though, compared with a lit room.
The colour-filter test shows it. In a 2021 study from Brigham Young University, reported in Sleep Health, 167 young adults spent the hour before bed for a week on a phone with a warm night-mode filter, on a phone without it, or with no phone at all. Wrist sensors measured their sleep. The filter made no difference. Among those who slept about seven hours, the people with no phone at all slept slightly better, whatever the screen colour.
So the colour of the screen is not the main issue. Time on it seems to matter more, and so does what it does to your attention. A feed has no natural end. A book has chapters, and an episode has credits. The phone has a next item, and then another.
What this study cannot tell you
A within-person design removes the differences between people. It does not remove the differences between nights. On a night when a teenager is wired after an exam, or anxious, or simply not tired, two things happen together: they stay on the phone, and they fall asleep late. The arrow could point both ways, and this study cannot separate them. Only a trial that changes phone use on purpose can do that.
There are other limits. Everyone used one type of phone operating system, because the logging app needed it. Sleep was estimated by a consumer wrist tracker, which is good at timing and weaker at the details. And these were teenagers, whose body clocks run later than adults’ for biological reasons. The size of the effect in a 45-year-old is not known from this study.
None of that makes the finding useless. It makes it a strong clue, not a verdict.
Adults are not off the hook
The largest adult data point is a 2025 survey of 45,202 Norwegian university students aged 18 to 28, in Frontiers in Psychiatry. Each extra hour of screen use after getting into bed was associated with 59% higher odds of insomnia symptoms and about 24 minutes less sleep. What people did on the screen made no clear difference. Social media was not worse than films or games. The time was what counted.
That study took a single snapshot and relied on self-reports, so it cannot show cause either. But its shape matches the teenage data. The risk sits in minutes spent, not in the type of app, and not in a special toxic window just before sleep.
A phone curfew that uses the clock
If the clock matters more than the countdown, the practical rule changes. Here is what the evidence supports trying for two weeks:
- Pick a fixed time, not a gap. “Phone down at 10:15” is easier to keep than “one hour before bed”, because bedtime moves and the clock does not. Choose a time that gives you your usual sleep before your alarm.
- Charge it outside reach. Across the room or outside the bedroom. The point is not the light. It is that picking it up becomes a decision instead of a reflex.
- Do not punish the last five minutes. If you check a message right before sleep, the evidence says that is a small thing. The hour before it is the larger one.
- Notice which way the arrow points. If you reach for the phone because you cannot sleep, the phone is a symptom. A regular wake time and morning light, covered in our post on sleep regularity versus duration, are the better tools there.
- Do not trust the filter to save you. Night mode is pleasant on the eyes. It is not a licence for another hour.
To see whether it works for you, watch the timing, not a single score. The Agen Band estimates when you fall asleep and wake up; like any wrist device, it gives estimates, not lab measurements. Two weeks of earlier sleep onset is a clearer signal than one good night. Our guide to reading sleep beyond a single number explains why the trend is worth more than the score.
The bottom line
The new study does not show that phones damage sleep in some mysterious way. It shows something simpler and more useful. Late minutes on the phone went with later sleep, and because mornings are fixed, later sleep meant less of it. The last few minutes before sleep mattered little. The screen colour mattered little. The clock mattered most. If you want one change this week, stop counting backwards from bedtime and pick a time on the clock when the phone goes away. Then let yourself off the hook for the occasional final glance.
This article is for general information and is not medical advice. If you have ongoing trouble falling or staying asleep, or daytime sleepiness that affects your life, talk to your doctor.
