An 8-Year Sleep Trial Tests Circadian Change’s Staying Power

A time-limited sleep intervention can shift an adolescent's schedule for at least a year. Whether that shift survives the trip into adulthood is another question. The longest follow-up of a behavioral chronotype trial found that the early advantage had disappeared eight years later—not because the treatment never worked, but because the treatment and control groups had converged.

That result comes from a 2025 randomized controlled trial follow-up by Susman et al. of TranS-C, a transdiagnostic sleep and circadian intervention for young people. It is unusually useful evidence because the comparison group received physical exercise, an active control, rather than a waitlist. It is also unusually easy to overread. The study followed adolescents into young adulthood; it does not show that adult circadian treatment inevitably fades.

The intervention worked before the groups converged

The original trial enrolled 176 participants between 2013 and 2016 in the UC Berkeley catchment area. At long-term follow-up, 106 participants remained: 60% of the original sample and 67% of those eligible. Their mean age was 22.5 years.

TranS-C beat the physical-exercise control at post-treatment, six months, and 12 months on eveningness and the gap between weeknight and weekend wake times. Those are relevant circadian outcomes: stronger evening preference and a larger wake-time gap usually indicate greater delay or social jetlag. The between-group differences were statistically significant at those earlier assessments.

At roughly eight years, the prespecified analyses found no between-group differences on sleep, circadian, or five-domain health outcomes. The publication does not provide a decision-ready effect estimate and confidence interval for those long-term comparisons here, so the defensible conclusion is narrow: the trial detected no lasting advantage of TranS-C over its active control at that assessment. A non-significant result does not prove the two treatments produced identical outcomes, but it does not support a claim of durable superiority either.

This timeline changes the question. TranS-C demonstrated efficacy through 12 months. The eight-year result concerns maintenance, not whether sleep timing can respond to treatment.

Biology may have erased a real treatment difference

Susman and colleagues propose developmental chronotype normalization as the leading explanation. Eveningness tends to peak in mid-to-late adolescence, then shift earlier through the twenties. If both groups moved toward a more typical young-adult schedule as they matured, an advantage produced during adolescence could become harder—or impossible—to detect later.

That explanation fits the population and timing, but the trial cannot prove it. Its follow-up window overlaps the period when participants would be expected to mature biologically. Without a separate account of each participant's untreated developmental trajectory, the study cannot separate natural change from a fading intervention effect.

The active control creates a second plausible explanation. Physical exercise is not inert. It may have improved sleep timing or related behaviors, so convergence could reflect improvement in both groups rather than a return to baseline in the TranS-C group. The trial design rightly protects against mistaking general therapeutic contact for a treatment effect at 12 months. At eight years, however, the active comparator makes the reason for convergence less certain.

Ongoing behavior predicted chronotype better than old assignment

One long-term finding deserves attention, with a large caveat. Across both trial arms, participants who reported more current sleep-health behaviors also reported lower eveningness at the eight-year assessment. Randomized treatment assignment no longer predicted circadian outcomes.

This is an association, not proof that those behaviors caused the lower eveningness. People with an earlier chronotype may find regular sleep habits easier; shared circumstances may drive both. Still, the pattern is consistent with a mundane but important idea: circadian treatment may require continued practice. A past course of therapy cannot keep a schedule regular on your behalf.

That interpretation also fits a 2025 systematic review and meta-analysis of health-behavior habit formation. It found that habit formation varies widely and identified repetition frequency and stable context as important determinants of automaticity. It did not establish a sleep-specific timetable, and it cannot tell us whether TranS-C lasted long enough to make its practices automatic. It does make the idea of maintenance more credible than the fantasy that a brief intervention permanently rewires a calendar.

Adult phase shifting works; adult durability remains unmeasured

The TranS-C follow-up should not be generalized into "circadian interventions do not work in adults." Adult evidence supports shorter-term phase shifting. A systematic review of morning light therapy for adults with delayed sleep-wake phase disorder found that it can advance circadian phase. A systematic review of combined bright light and melatonin likewise found phase-advance efficacy across studied populations.

An adult randomized trial of chronotherapy in people with ADHD and delayed sleep phase syndrome also showed that behavioral and light-based treatment can shift circadian timing. These studies establish short-term biological responsiveness in adults. They do not answer whether the benefit remains after structured protocols end, because they lack multi-year randomized follow-up.

A systematic review of evening blue-blocking glasses reported modest, consistent phase-advancing effects in adults with sleep or mood presentations. Reducing short-wavelength light in the evening offers a plausible entrainment tool, but the review does not turn low-effort use into evidence of long-term adherence or long-term clinical benefit. The same caution applies to multimodal light-plus-melatonin protocols: a stronger initial phase shift does not establish greater durability.

Adults differ from the TranS-C participants in one potentially important way: they have generally passed the adolescent peak-eveningness window. That removes the most obvious developmental explanation for later group convergence. It does not supply the missing long-term trial data.

Why the outcome still matters for health

Residual eveningness and social jetlag are not automatically diseases, and observational links do not establish that an earlier schedule prevents illness. But the associations are worth taking seriously. A 2024 study of adults aged 18 to 40 linked eveningness with higher BMI, alcohol consumption, and cigarette use. A study of hospital employees associated the direction and magnitude of social jetlag with obesity, while research on chronotype, social jetlag, and nicotine use found that later chronotype was strongly associated with smoking prevalence.

Those findings provide context, not a prescription. Health status, work schedules, mental health, socioeconomic conditions, and traits that influence chronotype can confound these relationships. No long-term randomized trial has shown that maintaining a phase advance reduces hard cardiometabolic, psychiatric, or substance-use outcomes.

What the eight-year result can—and cannot—tell you

The study's limitations matter. Forty percent of the original participants did not complete long-term follow-up, and the publication does not report dropout characteristics by treatment arm. Differential dropout could bias the apparent convergence in either direction. The sample came from one geographic area, through clinician referrals and advertisements, which limits generalizability. Exploratory analyses by age, sex, and family income were likely underpowered, so the study cannot reliably identify subgroups with durable benefit.

For adolescents, TranS-C offers evidence of meaningful circadian improvement through 12 months, followed by no detectable advantage over an active control at eight years. For adults, light, chronotherapy, and related approaches can shift phase in the short term, but no comparable long-term randomized evidence tells us how long those changes persist.

The practical conclusion is less glamorous than a permanent reset: use circadian tools as ongoing schedule supports. A consistent wake time, appropriately timed light exposure, and reduced bright evening light make mechanistic sense when they fit a person's life. They are maintenance behaviors, not a one-time software update.