Elmsword
← Sleep

Sleep · Research · 5 min read

Screens in the hour before bed: later bedtimes, a little less sleep

A 2025 study of 122,058 US adults linked daily phone, laptop or tablet use in the hour before sleep to more poor sleep quality, later bedtimes and approximately 50 minutes less sleep each week.

A large 2025 study of US adults found that people who used a smartphone, laptop or tablet in the hour before sleep every night were more likely to report poor sleep quality than people who never did. They also reported going to bed about 19 minutes later and sleeping about 5 to 8 minutes less per night. The links were larger in people with an evening chronotype. This is an association from a survey, not proof that screens cause poorer sleep. [1]

What the researchers did

The data come from the American Cancer Society Cancer Prevention Study-3 (CPS-3), a cohort of adults from 35 US states and Puerto Rico. This cross-sectional analysis used the 2018 follow-up questionnaire and included 122,058 adults. Median age was 56 (range 27–85), 80.0% were women, and people who reported shift work were excluded. [1]

Screen use was a single measure: how many days a week people watched or read a smartphone, laptop or tablet in the hour before sleep. TV was not included. In the sample, 41.2% did this every night and 17.4% never did. [1]

All outcomes were self-reported: sleep duration and bedtime on workdays and nonworkdays, the single sleep-quality item from the Pittsburgh Sleep Quality Index, and chronotype. The paper describes chronotype as the body's natural internal clock; here it came from one item of the Horne-Östberg questionnaire, ranging from morning to evening types. [1]

The main results come from fully adjusted models, which accounted for age, sex, race, ethnicity, education, bedroom light and sleep-mask use. A prevalence ratio (PR) compares how common something is in one group versus another: a PR of 1.33 means poor sleep quality was 33% more common among daily users than among people with no screen use. [1]

“The prevalence of poor sleep quality was 33% (PR, 1.33; 95% CI, 1.27-1.39) higher among daily users compared with those reporting no screen use.”

Zhong et al. · Results · full text

Key findings

Compared with no screen use in the hour before sleep. Overall and frequency rows: fully adjusted models, Table 2. Chronotype rows: daily users on workdays, Table 4. [1]
OutcomeDifference
Poor sleep quality, daily usePR 1.33 [1.27 to 1.39]
Bedtime, workdays, daily use18.82 min later
Bedtime, nonworkdays, daily use19.69 min later
Sleep duration, workdays, daily use7.64 min less
Sleep duration, nonworkdays, daily use5.04 min less
Workday sleep, 1–3 nights a week3.40 min less
Workday sleep, 4–6 nights a week6.35 min less
Workday sleep, every night7.64 min less
Evening chronotype, daily useBedtime 15.62 min later; sleep 8.36 min less
Morning chronotype, daily useBedtime 9.33 min later; sleep 5.64 min less
Screens in the hour before bed and sleepWorkday sleep was 3.40, 6.35 and 7.64 minutes shorter with 1-3, 4-6 and 7 nights a week of screen use. Among daily users, evening chronotypes went to bed 15.62 minutes later and slept 8.36 minutes less; morning chronotypes 9.33 minutes later and 5.64 minutes less.Screens in the hour before bed and sleepDifferences vs. adults who reported no screen use (fully adjusted models)A. Workday sleep duration, by nights per week of screen use1–3 nights a week3.40 min less sleep4–6 nights a week6.35 min less sleepEvery night7.64 min less sleepB. Daily users on workdays, by chronotypeLater bedtimeLess sleepMorning chronotype9.33 min later bedtime5.64 min less sleepEvening chronotype15.62 min later bedtime8.36 min less sleepAssociations from one cross-sectional survey; they do not show cause and effect.Source: Zhong et al., JAMA Network Open 2025, Tables 2 and 4.
Source: Zhong et al., JAMA Network Open 2025, Tables 2 and 4. Bars show minutes compared with no screen use. [1]

“Daily screen use was associated with later bedtimes and approximately 50 minutes less sleep each week.”

Zhong et al. · Abstract, Conclusions and Relevance · full text

That weekly figure adds up small nightly gaps (the Discussion puts it at 48 minutes): in Table 2, daily users slept 7.64 fewer minutes on workdays and 5.04 fewer minutes on nonworkdays. [1]

Differences for people with an evening chronotype looked bigger. On workdays, daily users with an evening chronotype went to bed 15.62 minutes later and slept 8.36 fewer minutes; daily users with a morning chronotype went to bed 9.33 minutes later and slept 5.64 fewer minutes. [1]

“The associations between screen use and sleep appeared to differ on the basis of chronotype, with more disruptions to sleep occurring in those reporting a later chronotype.”

Zhong et al. · Discussion · full text

The difference also grew with the number of nights. Compared with no use, workday sleep was 3.40 minutes shorter with 1–3 nights a week of use, 6.35 minutes shorter with 4–6 nights and 7.64 minutes shorter with daily use. [1]

“We do not have additional information beyond weekly frequency of screen use but did observe a trend with increasing days of use reported.”

Zhong et al. · Strengths and Limitations · full text

What this doesn't tell us

Screen use and sleep were measured at the same single point in time. That means the study can't tell whether screens push sleep later or whether people who already sleep late use screens more. [1]

“Limitations of our study include the cross-sectional nature of our assessment of sleep and screen use.”

Zhong et al. · Strengths and Limitations · full text

Everything was self-reported, although the authors say their sleep-duration measure was validated against actigraphy and sleep diaries within CPS-3. Screen use was counted only as nights per week, with no information on minutes, content such as social media, reading or video, or TV. How long it took to fall asleep was not measured. [1]

“We also did not have information on other aspects of sleep that may be affected by screen use, such as sleep latency.”

Zhong et al. · Strengths and Limitations · full text

The sample was 80.0% women, 89.2% White and 58.0% college-educated, so the results may not apply equally to other groups. The differences are also small: about 5 to 8 minutes of sleep per night. [1]

Finally, the reason for the link is not known. This study did not separate the light from a screen and what people were doing on it, so it can't say which one matters. The authors cite a National Sleep Foundation expert panel that reached no consensus for adults. [1]

“However, for adults, there was no consensus on light exposure (56%), content (63%), or the overall association between screen use and sleep (75%).”

Zhong et al. · Discussion · full text

What to take away

Opinion, not a study result: If I were in your spot and a night owl, I would try keeping the phone out of the last hour before bed and see whether bedtime moves earlier. The study did not test this, and results may differ from person to person. If sleep problems persist, talk to a doctor.

Reference

  1. JAMA Network Open 2025[1] Zhong C, Masters M, Donzella SM, Diver WR, Patel AV. Electronic Screen Use and Sleep Duration and Timing in Adults. JAMA Network Open. 2025;8(3):e252493.
    DOI: 10.1001/jamanetworkopen.2025.2493 · Full text

This article is for education only and is not medical advice. If sleep problems persist, talk to a doctor or a trained CBT-I provider about your options.