There is an hour, usually somewhere after ten, when the scrolling changes. It has not gotten more interesting. If anything it has gone flat. The thumb keeps moving anyway, past things that are not landing, and what used to feel like a small pleasure has quietly become a way of occupying the place where sleep is supposed to go. You are not enjoying it. You are also not stopping. Both are true at once, and that is the part worth looking at.

MoMo hears this described often, almost always with a note of self-reproach attached: the phone as the thing she keeps failing to put down. We wrote recently, in the circuit that runs on sleep, about what a short night does to the morning after it. This piece is about the night itself. Not the deficit the next day, but the loop that produces the deficit, and then feeds on it.

The strong half of the loop

Begin with what the evidence states most plainly: evening device use, and especially in-bed use, is associated with later sleep onset, a shorter night, and a poorer-quality one. A 2026 narrative review makes a useful distinction: it is not total daily screen time that tracks most closely with disrupted sleep, but the screen use that happens at the edge of sleep, in bed, at the hour the body was preparing to let go.[1]

The behaviour underneath that has a name in the research: bedtime procrastination. Voluntarily delaying sleep when nothing external is keeping you up. No deadline, no crying child, no shift. Just the postponement itself. It keeps surfacing as the central hinge between evening screens and lost sleep.[1] And it is not random. A 14-day diary study, tracking people with wearable sleep monitors, found that the more stressful the day had been, the more a person postponed sleep that night, and that postponement was what shortened the sleep.[3] The depleted day did not lead straight to less rest. It led there through the choice to stay up.

That study was in students, so hold it lightly as a universal claim. But the shape is recognisable well beyond a campus. A separate survey of adults, most of them women, mean age in the late thirties, found the same texture: low-arousal and boredom states predicted poorer sleep and, indirectly, more of the phone-in-bed pattern.[2] When there is nothing left to give, the mind reaches for the input that asks the least of it. The scroll asks very little.

The half the research points toward, more carefully

Here is where the loop closes, and where the honesty has to get more careful.

The second arm is the return trip: that the sleep you lose tonight makes it harder, not easier, to stop tomorrow night. There is real support for this direction. A two-wave study following people over six months found the relationship running both ways: problematic phone use predicted later sleep problems, and poor sleep predicted increased phone use down the line.[4] Not a one-way street. A circuit that compounds. And an experiment found that sleep deprivation measurably weakened emotion regulation, and that the weakened regulation was what opened the door to more compulsive use.[5]

I want to name what those two studies are before leaning on them. The six-month one is in university students. The experiment is in students too, in a gaming context specifically. Neither was run in the general population of adult women. So the return arm, the part where lost sleep drives more reaching the next evening, is a direction the research points toward rather than a fact it has nailed down for everyone. It is consistent with everything else here, and with what the regulatory circuitry does under sleep loss, which the circuit piece covered.[7] But consistent-with is not the same as proven-in, and it would be dishonest to blur them.

What we can say without overreaching is this. The reason it is so hard to close the phone at midnight is not, most likely, that some evenings you care less about your own rest. It is that the capacity that would normally step in and stop you is the exact capacity a run of short nights quietly wears down. The reaching is not the cause of the depletion. It is often a readout of it.

What the largest study says about women

Most of the studies here are modest in size or narrow in setting. One is neither. A network analysis of more than twenty-seven thousand community adults mapped how specific patterns of internet use connect to specific sleep problems, and found the connections are not uniform across people.[6]

Two findings are worth pulling out. First, bedtime procrastination shows up as a bridging node across every group: the common thread linking use to lost sleep, whoever you are. Second, and more specifically, in women the central node is not emotional dependence on the device. It is the inability to reduce use. Not craving it, not needing it for company. The difficulty of cutting back once it is underway.

That is a precise, and I think a kind, finding. It locates the hard part exactly where the women MoMo is for tend to locate it themselves: not in wanting the phone, but in the strange difficulty of putting it down once the evening has tipped past a certain hour. The research is describing the thing she already feels, and naming it as a pattern rather than a fault.

Naming the loop without blaming the person in it

It is tempting to file all of this under discipline, and to make the fix a firmer rule about screens after nine. The women MoMo is built for have usually tried the firmer rule. It is rarely the missing piece, because the rule assumes the problem is a decision, and the research keeps describing something closer to a pattern that maintains itself.

A tired evening reaches for the lowest-resistance input it can find. That reaching costs sleep. The lost sleep wears down the very capacity that would have interrupted the reaching. So the next evening arrives a little more depleted, reaching a little earlier, a little harder. Read from inside, that looks like a personal failing that repeats. Read from the outside, it is what it is: a self-sustaining pattern, running on a nervous system doing exactly what a depleted nervous system does. The pattern is self-sustaining; naming it that way is different from reading it as a verdict about who you are.

You do not have to do anything with this. No new rule, no alarm, no audit of your evenings. But the next time you find yourself still scrolling at an hour you did not choose, past things that stopped being interesting a while ago, it might be worth knowing that the difficulty of stopping is documented, and that the research names it as a pattern rather than a verdict.

Noticing that a flat, hard-to-close evening connects back to the short night before it, and the one before that, so the loop becomes something you can see rather than a verdict you keep passing on yourself, is the kind of continuity MoMo was built to hold. But that is a footnote. The essay is the loop, and the way it keeps the light on.

If you're in distress or crisis, please reach out: Mind Sweden 90101 | Samaritans UK 116 123 | your local crisis line.

References

Khan, M. D., Junaidi, S., & Dalbah, M. I. (2026). Digital Engagement and Sleep Dysregulation in Young Adults: A Narrative Review. Cureus, 18(3), e105030. PMID: 41970105.
Fabbri, M., & Martoni, M. (2026). The Relationship Between State Boredom and Sleep-Wake Disruptions: A Mediation Model via Smartphone Use and Bedtime Procrastination. International Journal of Environmental Research and Public Health, 23(6), 728. PMID: 42354259.
Schmidt, L. I., Baetzner, A. S., Dreisbusch, M. I., Mertens, A., & Sieverding, M. (2024). Postponing sleep after a stressful day: Patterns of stress, bedtime procrastination, and sleep outcomes in a daily diary approach. Stress and Health, 40(3), e3330. PMID: 37846558.
Cemei, L., Sriram, S., Holý, O., & Rehman, S. (2024). A Longitudinal Investigation on the Reciprocal Relationship of Problematic Smartphone Use with Bedtime Procrastination, Sleep Quality, and Mental Health Among University Students. Psychology Research and Behavior Management, 17, 3411–3424. PMID: 39359420.
Zhao, Wang, & Liu (2026). Immersive experience of online gaming among Chinese college students: the enhanced effect of sleep deprivation on emotion regulation and problematic gaming within the I-PACE model. BMC Psychology, 14. PMID: 41943088.
Wang, X., Li, Y., Liu, Y., Zhang, N., Ding, Y., Fan, S., Tian, Z., Wu, Y., & Wang, L. (2026). Differential relationships between problematic internet use and sleep problems: a network analysis. Psychology & Health. PMID: 41848289.
Adeniji, K. O., Igado, O. O., Olopade, F. E., & Femi-Akinlosotu, O. M. (2026). Sleep deprivation and memory: A neurobiological perspective. IBRO Neuroscience Reports. PMID: 41631014.