On screen time and mental health, randomised trials show that cutting phone use produces small to medium improvements in stress, depressive symptoms, wellbeing and sleep quality. The effects are real and modest, and in the one trial that followed up, usage returned towards baseline once the limit ended.
Most writing about screen time and mental health is either alarmist or dismissive, and both sides tend to rely on observational data that cannot tell you which way the arrow points.
There are two randomised trials worth knowing about. Here is what they measured and what they found.
Trial one: cutting phone use to two hours a day
Pieh and colleagues published this in BMC Medicine in 2025. It was a non blinded, parallel group randomised controlled trial with 111 of 125 healthy students randomised, 58 to the intervention and 53 to control. Mean age 22.68, 70 participants female, baseline screen time averaging 276 minutes a day.1
The intervention group reduced phone use to two hours or less a day for three weeks.
Post intervention group differences, with the measures they used:
| Measure | Mean difference | 95% confidence interval |
|---|---|---|
| Stress (PSQ) | 6.91 | 0.01 to 13.81 |
| Wellbeing (WHO-5) | 1.54 | 0.18 to 2.89 |
| Depressive symptoms (PHQ-9) | 2.11 | 0.87 to 3.36 |
| Sleep (ISI) | 2.59 | 0.66 to 4.51 |
The authors describe the effects as small to medium in size.1 Look at the stress confidence interval: 0.01 at the lower bound, which is about as close to nothing as a significant result gets. That one is genuinely borderline and should not be quoted as if it were solid.
The depression and sleep intervals are more comfortably clear of zero.
A note on reading that table, because these scales are not intuitive. PHQ-9 runs from 0 to 27 and a change of around 5 points is usually treated as clinically meaningful, so a difference of 2.11 is a real signal rather than a dramatic one. The ISI runs 0 to 28 on similar logic. In other words, the direction of all four results is consistent and favourable, and the magnitudes are the kind you would notice over weeks rather than overnight. That is a less exciting summary than most articles on this subject offer, and it is what the numbers actually say.
It is also worth noting what the design does and does not control. The trial was not blinded, which is close to unavoidable here, since you cannot hide from someone whether they are being asked to use their phone less. That means expectancy effects are in play: people who know they are in the intervention group may report feeling better partly because they expect to. The sleep result is somewhat less vulnerable to that than the mood results, which is one reason to weight it slightly more heavily.
Trial two: deleting Facebook for four weeks
Allcott, Braghieri, Eichmeyer and Gentzkow measured what 2,844 Facebook users would need to be paid to deactivate for four weeks, then randomly assigned a subset to actually do it, with deactivation verified rather than self reported.2
They found small but significant increases in wellbeing, more time on offline activities including socialising with family and friends, and reduced political polarisation.2
They also found a cost: reduced factual news knowledge.2 Worth knowing that the trade is not entirely one way.
The finding that matters most for screen time and mental health
In the BMC Medicine trial, screen time rose quickly after the intervention ended, and by the six week follow up it was approaching baseline.1
That is the single most practically useful result in this literature, and it is almost never quoted, because it undercuts everything sold as a reset or a detox.
The improvement was conditional on the behaviour change, and the behaviour change did not survive the removal of the constraint. So if you are doing this for the mental health benefit, the relevant question is not what you can manage for three weeks. It is what you can keep.
The Facebook study is more encouraging on exactly this point: there, the reduction in use persisted after the experiment.2 Deleting something seems to stick better than limiting it.
Four things these trials do not show
- They do not show your brain is being rewired. Nothing here measured brain structure. Claims about dopamine receptors and rewiring are not coming from this evidence.
- They do not show phones cause depression. They show that reducing use improved depression scores in a specific group over a few weeks. That is a narrower claim.
- They were not run on you. Trial one used 111 students with a mean age of 22. Trial two used Facebook users around a US election. Generalise carefully.
- They are small. One hundred and eleven people is a reasonable trial, not a settled science. The effects being described as small to medium is the authors being appropriately modest.
What to actually take from it
Three things, stated at the confidence the evidence supports.
Reducing phone use probably helps a bit. Across sleep, mood and stress, modestly, for most people. That is worth having and it is not life changing.
Permanent beats intense. The arrangement has to outlast your motivation, because the benefit disappears with the constraint.
Removing beats limiting, where you can. Deletion produced lasting change. Limits did not.
Where SETL fits
SETL is built on the permanence finding rather than on any claim about your brain. It is an iPhone app blocker that takes your chosen apps away automatically at your bedtime, indefinitely, with no decision to make each night. It will not treat anything. If you want the related evidence, read whether screen time affects sleep and whether deleting social media helps.
This article is general information, not medical advice. Speak to your doctor about ongoing sleep, attention or mental health concerns.
Sources
- Pieh C, Humer E, Hoenigl A, Schwab J, Mayerhofer D, Dale R, Haider K, Smartphone screen time reduction improves mental health: a randomized controlled trial, BMC Medicine, 2025, 23(1):107
- Allcott H, Braghieri L, Eichmeyer S and Gentzkow M, The Welfare Effects of Social Media, American Economic Review, 2020