If a fixed nighttime boundary fits your routine, you can download 9to9 from the App Store and choose the apps and websites you want quiet from 9 PM to 9 AM.
This is the article where the honest answer is more complicated than either pole you will find elsewhere. One side says the screens are ruining our sleep — a settled fact, self-evident, proven. The other says the panic is overblown — the studies are weak, the effects are small, the alarm is a moral one dressed as science. The research sits, as research usually does, somewhere more useful than either: a consistent association between bedtime screen use and worse sleep, with real uncertainty about the details. That combination — consistent direction, unsettled mechanism — is worth understanding precisely, because it tells you both that the problem is real and that the popular explanations for it are not quite proven.
Let us take it in three parts: what is solid, what is still uncertain, and what is worth doing anyway.
First, what “before bed” actually means in the research
A clarification that matters more than it looks, because the studies and the headlines mean different things by the same phrase. In the research, “screen use before bed” usually means use in the hour or two before sleep — and sometimes specifically in bed, which is a different and stronger exposure than use earlier in the evening. The headlines tend to flatten this into “screens at night are bad,” which is both vaguer and more alarming than the finding. The more precise version — the scroll in the last stretch, and especially in bed — is the one the association is actually about, and the one the practical advice should target.
What the evidence actually shows
The strongest and most replicated finding is an association: people who use screens in the period before sleep tend to report later bedtimes, shorter sleep, and worse reported sleep quality than people who do not. A large observational study of adults found that daily screen use before sleep was linked to later bedtimes and shorter sleep on weeknights — a real association in a large dataset, reported as an association, not a causal claim.
An expert consensus statement on screen use and sleep across the lifespan reached a similar careful conclusion: screen media use around bedtime is consistently associated with worse sleep-health outcomes, while noting that the effects depend on content, timing, light, age, and context. That is the shape of what the field agrees on — consistent association, context-dependent size.
Sources: Electronic Screen Use and Sleep Duration and Timing in Adults (PMC), The impact of screen use on sleep health across the lifespan: consensus statement (PMC)
It is worth pausing on what “association” means here, because it is the honest crux. The studies show that people who scroll before bed sleep worse. They do not, for the most part, prove that the scrolling causes the worse sleep — the arrow could run the other way (poor sleepers reach for the phone), or both could reflect a third thing (stress, irregular schedules). The direction of the finding is consistent enough to take seriously; the mechanism is not settled enough to treat as proven. That is not a hedge — it is what the evidence actually is.
A note on who is most affected
The association is not evenly distributed, and the honest reading is that it concentrates in the people with the worst patterns. The person who uses a screen calmly for twenty minutes before bed is in a different situation than the person whose feed holds them for two hours past the intended bedtime — the research averages across both, which is part of why the effect sizes vary. The practical implication: if your screen use before bed is the second kind, the association is probably aimed at you; if it is the first kind, the finding is probably describing someone else's problem.
What is still genuinely uncertain
Three things the research has not settled, and which the confident articles usually skip. First, the mechanism — why screen use is associated with worse sleep is not fully established. The two leading candidates are displacement (the screen eats time that would have been sleep) and arousal (the content keeps the mind activated), with light as a third, smaller suspect. Each is plausible; none is conclusively the dominant one.
Second, the magnitude. The associations are real but the effect sizes vary widely between studies — some find meaningful differences in sleep duration, others find small ones. The honest summary is “a real tendency toward worse sleep, of uncertain size, in the people studied” — which is meaningful without being the catastrophe the headline version implies.
Third, the content question — whether what you do on the screen matters as much as when. Some evidence suggests interactive, stimulating, or personally engaging content is more disruptive than passive viewing; a review of media and sleep notes the effects likely depend on what the media is doing to the user, not just that a screen is on. A calm documentary and a doomscroll are probably not equivalent — but the research has not pinned down exactly how different.
Sources: How does modern media use impact sleep? (PMC)
The sleep hygiene guidance that does agree
Whatever the uncertainties about the screen-specific findings, the broader sleep-hygiene evidence is more settled, and it is worth holding separately. Consistent sleep and wake times, a cool dark quiet room, a wind-down routine, limiting caffeine and alcohol late in the day — these are the recommendations that have the strongest and longest-standing support, and none of them are about screens specifically. The screen findings sit inside this larger frame: the late-night scroll is one of the things that breaks the consistency — pushing the bedtime, fragmenting the wind-down — rather than a uniquely modern villain. Which is both reassuring and practical: the fix is not a new behavior, it is protecting the old ones the scroll keeps breaking.
The honest summary of the evidence
For the reader who wants it in one place: the association between pre-sleep screen use and worse sleep outcomes is consistent and reasonably large across studies. The mechanisms are plausible and multiple — displacement, arousal, light — with no single one proven dominant. The causation is not airtight, and the honest framing is “consistent association with strong biological plausibility” rather than “proven.” And the practical implication is the one the field converges on: the pattern is worth changing, the change is cheap, and the direction of the evidence is clear even where the magnitude is not.
Why the blue-light story is too simple
The most popular mechanism — the screen's light suppressing melatonin and delaying sleep — is real but oversold. Light does affect circadian timing; that part is established. But the dose from a phone screen is small compared to the light exposures that produce large circadian effects, and several studies suggest the behavioral factors — the displacement, the engagement, the “one more” — account for more of the harm than the photons. The practical upshot: dimming the screen and turning on night mode are fine, but they address the smaller part of the problem. The light is real; it is just not the main story.
The honest bottom line on causation
It is worth stating plainly what would change the picture: a clear demonstration that reducing pre-sleep screen use improves sleep outcomes in a controlled way would move the finding from association toward causation. Some intervention studies exist and point that direction, but the evidence is not yet the settled, airtight kind. The honest position is the one the field takes: the association is consistent, the mechanisms are plausible and multiple, the intervention is cheap, and acting on it is reasonable even while the proof is still being assembled. That is not a weak case — it is the normal case for a lot of behavioral-health guidance, and it is strong enough to justify the change.
What is worth doing anyway
Here is the part where the uncertainty stops mattering much. Even without a settled mechanism or a precise effect size, the behavioral pattern is clear enough to act on: the scroll before bed is associated with later nights and worse-reported sleep, and the intervention that addresses it costs almost nothing. You do not need the research to be airtight to justify moving the feed out of the last hour — you need the association to be real and the fix to be cheap, and both are true.
The proportionate version: block the feeds during the vulnerable window — a scheduled blocker, a fixed 9-to-9 boundary — keep the sleep-adjacent uses that are calm and finite, dim the screen for what that is worth, and stop expecting a single change to fix a complicated system. The evidence does not promise that cutting the late scroll will transform your sleep; it suggests, consistently, that the late scroll is one of the things making it worse. That is enough to act on, and it is more honest than either the panic or the dismissal.
When the problem is not the phone
The last honesty: the screen is not always the whole story. If you stop the scroll and the sleep is still bad — if you lie awake, wake early, feel wrecked during the day — the phone was probably a symptom rather than the cause, and the real issue deserves its own attention. The research on screens and sleep is about a pattern, not about insomnia; persistent sleep difficulty is a different problem with different help, and a blocked feed will not fix it.
The measured summary, then: the link between the late-night scroll and worse sleep is real and consistent, the explanation for it is still being worked out, and the fix — give the feeds an ending before the bed — is cheap enough that you do not need the science settled to justify it. That is the version of the truth worth acting on.