Longitudinal Between- and Within-Person Associations Among Screen Time, Bedtime, and Daytime Sleepiness Among Adolescents: Three-Wave Prospective Panel Study.
Tkaczyk M, Ksinan A, Smahel D • Journal of medical Internet research • 2026
Screen time and bedtime show a complex, bidirectional, and mutually reinforcing interplay over time among adolescents, whereas daytime sleepiness does not appear to be directly affected by this dynamic at the within-person level.
Key Findings
Results
At the between-person level, adolescents with habitually higher screen time reported later bedtimes and greater daytime sleepiness.
Between-person correlation between screen time and bedtime: r=0.23, 95% CI 0.15–0.31; P<.001
Between-person correlation between screen time and daytime sleepiness: r=0.25, 95% CI 0.16–0.34; P<.001
These associations reflect stable individual differences rather than within-person fluctuations
Sample: 2500 Czech adolescents, mean age 13.43 (SD 1.70) years, 50% girls, quota-based
Results
No direct within-person effects were found between screen time and daytime sleepiness across any wave transition.
W1→W2: β=.02, 95% CI −0.11 to 0.16; P=.71
W2→W3: β=.02, 95% CI −0.10 to 0.14; P=.66
This indicates that when individual differences are controlled, temporary increases in screen time do not predict subsequent changes in daytime sleepiness
Analyses used random intercept cross-lagged panel models (RI-CLPM) to separate between- and within-person variance
Results
Within-person increases in screen time at Wave 1 predicted later bedtime at Wave 2, which in turn predicted higher screen time at Wave 3, indicating a bidirectional reinforcing cycle.
Screen time W1 → bedtime W2: β=.14, 95% CI 0.01–0.27; P<.05
Bedtime W2 → screen time W3: β=.24, 95% CI 0.11–0.36; P<.001
This suggests a mutually reinforcing dynamic between screen time and delayed bedtime over a 12-month period
Waves were separated by 6-month intervals
Results
Temporary within-person spikes in screen time coincided with delayed bedtimes at the same wave (concurrent within-person correlations).
W1: r=0.16, 95% CI 0.04–0.27; P<.01
W2: r=0.23, 95% CI 0.09–0.36; P<.001
W3: r=0.09, 95% CI 0.00–0.20; P=.049
Authors interpreted the negative direction of this association as indicating these behaviors are 'partly mutually exclusive and likely shaped by shared contextual influences'
Results
Baseline screen time restriction within one hour before sleep did not moderate the within-person effects of screen time on bedtime or daytime sleepiness.
Moderation test: Δχ²(6)=5.3; P=.51
Screen time restriction before sleep was assessed as a binary moderator at baseline
This null moderation finding held across all within-person cross-lagged paths tested
Authors concluded that 'interventions should prioritize consistent sleep schedules rather than focusing solely on reducing screen use'
Methods
The study used random intercept cross-lagged panel models (RI-CLPM) across three waves to separate between- and within-person associations, addressing a methodological limitation of prior cross-lagged panel studies.
Three-wave prospective panel design with 6-month intervals between waves
Sample of 2500 Czech adolescents collected via online survey
Screen time assessed with 3 items covering computers, smartphones, tablets, and television on a typical school day
Daytime sleepiness assessed with 4 items from the Pediatric Daytime Sleepiness Scale
Age and sex were included as covariates
Discussion
The authors identified this as the first study to test reciprocal longitudinal associations among adolescents' screen time, bedtime, and daytime sleepiness while separating between- and within-person associations.
Prior research had not disentangled stable individual differences from dynamic within-person change in this context
The RI-CLPM approach addresses 'potential bias common in prior cross-lagged panel studies'
Findings 'refine theoretical understanding by indicating a complex, bidirectional, and mutually reinforcing interplay between screen time and bedtime'
What This Means
This research suggests that the relationship between screen time and sleep in teenagers is more nuanced than previously thought. When comparing teenagers to each other, those who generally spend more time on screens tend to go to bed later and feel sleepier during the day. However, when tracking individual teenagers over time, a temporary increase in a teen's own screen use does not directly cause them to feel sleepier later — instead, the connection runs through bedtime: using screens more than usual is linked to going to bed later, and going to bed later is then linked to using screens even more in the future, creating a reinforcing cycle between screen habits and sleep timing.
Notably, this research suggests that daytime sleepiness in teenagers is not directly driven by short-term changes in their own screen time, once stable personal traits are accounted for. This challenges simpler narratives that more screen time straightforwardly causes more sleepiness on an individual level. The study also found that teens who already restricted their screen use in the hour before bed at the start of the study did not have better outcomes than those who did not, suggesting that this particular habit did not buffer against the screen-bedtime cycle.
From a practical standpoint, this research suggests that public health and parenting strategies may be more effective if they focus on maintaining consistent, earlier bedtimes rather than concentrating solely on cutting screen time before sleep. The bidirectional nature of the screen-bedtime relationship means that interventions targeting either behavior could potentially disrupt the reinforcing cycle, but scheduling and sleep routine consistency appears to be a key leverage point.
Tkaczyk M, Ksinan A, Smahel D. (2026). Longitudinal Between- and Within-Person Associations Among Screen Time, Bedtime, and Daytime Sleepiness Among Adolescents: Three-Wave Prospective Panel Study.. Journal of medical Internet research. https://doi.org/10.2196/78972