Sleep got longer between 14 and 22 — and fatigue got worse anyway

Sleep advice for young people is built almost entirely around duration: how many hours, and how to get more of them. A cohort study published in Sleep Health on 26 November followed the same people from mid-adolescence into their early twenties and found duration and symptoms moving in opposite directions [s1].

The cohort

The Eating and Activity over Time study surveyed 1559 participants twice — in 2010, at a mean age of 14.3, and again in 2018, at a mean age of 22.0 [s1]. It measured weekday and weekend sleep duration, trouble sleeping, and fatigue, and used mixed linear and logistic regression models to assess change over time and differences by sex and socioeconomic status [s1].

Duration improved

Around 43% of the adolescents were not meeting minimum weekday sleep recommendations [s1]. Average weekday sleep duration was shorter in adolescence, at 8.1 hours, than in young adulthood, at 8.3 hours [s1].

That direction runs against the intuition that sleep deteriorates once school ends and work begins. It is consistent with what the biology predicts: adolescents face a circadian delay that collides with early school start times, and leaving school removes the collision.

The socioeconomic pattern is the sharper finding. During adolescence, the low socioeconomic status group had the shortest weekday sleep duration compared with other socioeconomic status groups — but by young adulthood, all socioeconomic status groups slept for similar durations [s1].

A gap that closes on its own when a structural constraint is removed is informative about where the constraint sat.

Symptoms did not improve

Despite the longer average sleep, a greater proportion of young adults than adolescents reported fatigue and trouble sleeping [s1].

And here the socioeconomic gradient did not close. The low socioeconomic status group and females reported more fatigue and trouble sleeping than their counterparts at both time points [s1].

So two different things are happening. How long people sleep converged across socioeconomic groups between 14 and 22. How rested they feel, and how much difficulty they have sleeping, did not.

The authors' conclusion follows the split: interventions should prioritise weekday sleep duration in adolescence, particularly among low socioeconomic status groups, and address fatigue and trouble sleeping in young adulthood, particularly among females [s1].

A parallel finding in a different population

A study published the following day, 27 November, in the same journal describes sleep health among US rural populations and reaches a compatible conclusion about where disparities concentrate [s2].

It used cross-sectional data from the 2020 and 2022 National Health Interview Survey, covering 1429 children and 8673 adults living in rural counties, and estimated prevalence ratios using Poisson regression with robust variance [s2].

Sleep disturbances were prevalent across all age groups [s2]. Among children, disturbances became more prevalent with increasing age while consistent bed and wake times became less prevalent [s2]. Consistent disparities emerged among 6- to 13-year-olds, including higher sleep disturbances among children of lower versus higher socioeconomic status [s2]. Sex disparities appeared among adolescents aged 14 to 17 [s2]. Among adults, sleep duration and disturbances varied across all sociodemographic characteristics, with generally higher prevalence in groups with adverse social conditions such as being divorced, separated or widowed, or of low socioeconomic status [s2].

Both studies point at the same thing from different angles: socioeconomic disadvantage and sex show up more reliably in sleep disturbance measures than in sleep duration measures.

Limits

Both studies rely on self-report. The Eating and Activity over Time analysis measured sleep by survey at two widely spaced points, which cannot capture what happened between them, and self-reported duration is known to run optimistic in adolescents. The rural analysis is cross-sectional and cannot speak to change over time [s2]; it also reports sleep duration for adults only [s2].

Neither study establishes cause. The convergence of duration across socioeconomic groups by young adulthood [s1] is consistent with school schedules being the binding constraint in adolescence, but the design cannot demonstrate that.

What to take from it

The headline number in most adolescent sleep coverage is hours. These two studies suggest hours are the part that improves on its own, and that the part which persists — fatigue, difficulty sleeping, and the social gradient in both — is the part that is not being measured by the metric everyone quotes.

This article is informational and does not constitute medical advice.

Sources

• [s1] Changes in sleep from adolescence to young adulthood: Findings from the Eating and Activity over Time 2010-2018 Study. Sleep Health, 26 November 2025. https://doi.org/10.1016/j.sleh.2025.10.011
• [s2] Rural sleep health and sleep health disparities among children and adults in the United States. Sleep Health, 27 November 2025. https://doi.org/10.1016/j.sleh.2025.09.006