Your Screens Are Stealing Your Sleep. Here Is the Fix That Works

Blue light has taken the blame for a decade, and it has been a convenient villain because it comes with a product attached. Buy the glasses, enable the filter, and the problem is solved without changing anything you do. The evidence does not support that arrangement. It supports something duller and harder: the phone is mostly costing you sleep by taking the time you were going to sleep in.

Both mechanisms are real. One of them is much bigger than the other in ordinary life, and it is not the one being sold to you.

What screen light demonstrably does

The strongest causal evidence comes from a controlled in-lab study by Chang and colleagues, published in PNAS in 2015. Twelve participants read for four hours before bed on five consecutive nights, alternating between a light-emitting tablet and a printed book, in otherwise dim conditions. The tablet delivered 31.73 photopic lux at the eye; the printed book, 0.91.

The results were not subtle. The tablet condition “suppressed evening levels of melatonin by 55.12 percent.” Dim light melatonin onset came more than an hour and a half later. Participants “averaged nearly 10 min longer to fall asleep” and had significantly less REM sleep.

The mechanism behind it is well established. The NIH’s National Institute of General Medical Sciences explains that “the SCN controls production of the hormone melatonin based on the amount of light the eyes receive. In the evening, a person’s master clock tells their brain to make more melatonin, causing sleepiness.” The NHLBI adds that “exposure to bright artificial light in the late evening can disrupt this process and prevent your brain from releasing melatonin,” naming “the light from a TV screen, a smartphone, or a very bright alarm clock.”

And the clock is more sensitive than people assume. A landmark dose-response study by Zeitzer and colleagues found that “half of the maximal phase-delaying response achieved in response to a single episode of evening bright light (approximately 9000 lux) can be obtained with just over 1 percent of this light (dim room light of approximately 100 lux).”

Now the caveat that matters. The PNAS protocol was four hours of continuous close-held tablet reading in a dim laboratory. That is not checking your messages for ten minutes. The study proves light can shift the clock; it does not establish that a brief glance does much.

What the glasses do

Here the evidence is unusually clear, and it points the other way. A Cochrane systematic review published in 2023 pooled 17 randomised controlled trials of blue-light filtering spectacle lenses. Its conclusion: such lenses “may not attenuate symptoms of eye strain,” with “no difference in subjective visual fatigue scores” and probably “little or no effect on visual acuity.” On sleep specifically, the review found the evidence insufficient to support a benefit.

The American Academy of Ophthalmology goes further on its own account: “The American Academy of Ophthalmology does not recommend blue light-blocking glasses because of the lack of scientific evidence that blue light is damaging to the eyes.” The Academy is softer on night mode, with a clinical spokesperson suggesting warm display colours make it easier to fall asleep, but that is expert opinion on a consumer page rather than a graded evidence review, and it sits awkwardly beside the Cochrane finding. The NHLBI describes filters neutrally, noting only that “some people use physical filters or software to filter out some of the blue light from these devices.” It does not endorse them.

Night mode is free and costs you nothing. The glasses are a purchase with no demonstrated sleep benefit behind them.

The bigger effect nobody is selling

In 2025, Hjetland and colleagues published an analysis of 45,202 Norwegian university students in Frontiers in Psychiatry. It found that “a one-hour increase of screen time after going to bed was associated with 59 percent higher odds of having symptoms of insomnia” and “a reduction in sleep duration of 24 minutes.”

The decisive finding was a null one. “The associations between screen time and sleep outcomes did not differ for social media use versus other activities.” The authors’ own reading: “Our results can be interpreted as supporting the displacement hypothesis, i.e., that the time spent on screens displaces sleep. If increased arousal was an important contributor to poorer sleep, we would expect to see different associations between screen time and sleep for the different screen activities.”

This is a cross-sectional, self-reported study of students in one country. It cannot prove causation and it did not measure light. But it is enormous, and its logic is hard to argue with: if the content mattered, doomscrolling and reading a novel on the same device would not look the same. They did.

Take the 24 minute figure at face value for a moment, as an illustration rather than a prediction. An extra hour of in-bed phone use on five weeknights costs two hours of sleep a week, roughly 104 hours a year, or about 15 full nights of seven-hour sleep. That arithmetic is ours, and it extrapolates an association rather than a demonstrated causal effect. Even discounted heavily, it dwarfs anything a pair of tinted lenses could recover.

How short are people actually sleeping?

The CDC recommends “7 or more hours” for adults aged 18 to 60, matching the American Academy of Sleep Medicine’s 2015 consensus statement that “adults should sleep 7 or more hours per night on a regular basis to promote optimal health.” Teenagers aged 13 to 17 need 8 to 10 hours.

The gap between that and reality is wide. The National Center for Health Statistics reported that “in 2024, 30.5 percent of adults slept less than 7 hours on average in a 24-hour period,” and that 15.4 percent “had trouble falling asleep most days or every day.” Among teenagers the picture is worse: the CDC’s 2025 Youth Risk Behavior Survey found that “in 2025, 24 percent of high school students got at least 8 hours of sleep on an average school night.” Three quarters do not.

What the evidence actually supports, ranked

Sleep advice is delivered as an undifferentiated list, which hides the fact that the items rest on very different evidence. Here is the same list sorted by how well supported each item is, and by which body says so.

The rule that does the work

Both the CDC and the AASM land on the same instruction. The CDC’s sleep page lists “turning off electronic devices at least 30 minutes before bedtime.” The AASM’s patient education arm says the same, and a 2025 AASM article extends it to “power down electronics at least 30 to 60 minutes before bed,” adding “keep it in another room if possible.”

Be aware of what these figures are. There is no formal, graded AASM position statement setting a screen curfew. The 30 minute rule is consumer guidance from two authoritative bodies, which is a perfectly good reason to follow it and not the same thing as a clinical trial result.

The AASM’s own 2025 survey of 2,007 US adults found that “one-half of adults (50 percent) report using a screen (TV, smartphone, computer, tablet, e-reader) while in bed every day,” and that “more than one-quarter (26 percent) of adults prioritize screen time with their phone over getting the recommended amount of sleep.” That last figure is the honest description of the problem. It is not a lighting problem. It is a choice people are making and mostly not enjoying.

There is also evidence that the phone’s presence matters independently of its screen. An analysis of 10,280 adolescents in the NIH-funded ABCD study found elevated risk of trouble falling or staying asleep associated with having devices in the bedroom (adjusted risk ratio 1.27), leaving the phone’s ringer on (1.23), and using a device after waking in the night (1.34). A silent phone in another room removes all three.

The rest of the evening

The NHLBI’s healthy sleep habits page is short and specific. Go to bed and wake up at the same time every day, varying by “no more than about an hour” at weekends. “Keep your bedroom quiet, cool, and dark.” “Use the hour before bed for quiet time.” Avoid heavy meals and alcohol close to bedtime. “Adults should nap for no more than 20 minutes.” The National Institute on Aging adds a useful specific: “try to get regular exercise, but not within three hours of your bedtime.”

One thing you will not find in any of them is a bedroom temperature. We looked across the CDC, NHLBI, NIA and AASM and none gives a figure in degrees. They all say “cool.” The widely circulated numbers do not come from these sources, and we are not going to invent one.

On alcohol, the specifics are worth knowing because the drink genuinely does make you fall asleep faster. A review in the NIAAA’s Alcohol Research: Current Reviews describes the trade: a dose-dependent increase in slow wave sleep in the first half of the night, “followed by a decrease in slow wave sleep in the second half,” and “a significant decrease in the percentage of REM sleep.” Later in the night, “sleep becomes more disrupted and awakenings are more frequent.” You are borrowing from the second half to pay for the first.

If it has gone on for months

At that point this stops being a screens article. The NHLBI describes cognitive behavioural therapy for insomnia as “usually recommended as the first treatment option for long-term insomnia,” delivered as “a 6- to 8-week treatment plan,” available in person, by telephone or online. In April 2026 the AASM published a guideline on combination treatment whose lead author stated that “our analysis suggests that CBT-I by itself is the most efficacious first-line treatment for insomnia.”

Melatonin is not that. The NIH’s National Center for Complementary and Integrative Health states that “there’s not enough strong evidence on the effectiveness or safety of melatonin supplementation for chronic insomnia to recommend its use.” It also flags a quality problem that should give anyone pause: in one analysis, “22 out of 25 over-the-counter melatonin gummy products were inaccurately labeled,” with actual melatonin levels ranging from 74 to 347 percent of what the label claimed.

Why any of this matters beyond feeling tired

NHTSA’s overview of 2024 traffic crashes records that “the number of fatalities involving drowsy drivers was 644 or 1.6 percent of total traffic fatalities in 2024.” It is a small share of a large number, and it is the end of a chain that begins with an extra hour in bed with a phone.

The part that is actually hard

The honest version of this advice is not a technique. It is that the phone is more interesting than lying in the dark, and it will be more interesting tomorrow too. Nothing in the research suggests that willpower at 11pm is reliable, and the interventions that work are the ones that remove the decision rather than winning it: the charger in the kitchen, the alarm clock that is not a phone, the fixed wake time that eventually makes the body do the work for you.

None of that is satisfying to read, and the market has correctly identified that people would rather buy glasses. The glasses do not work. The charger in the kitchen does, and it costs nothing, which is presumably why nobody is advertising it.

General information, current as of September 2026, and not personalised advice. If sleep problems persist, speak to a qualified health professional.