Earlier Bedtime and Its Effect on Adolescent Sleep Duration

Younger cohort of 77 participants (ages 9.9-16.2) studied annually for 3 years; older cohort of 67 participants (ages 15.0-20.6) studied once

US adolescents and emerging adults aged 10-21

Within-subject experimental manipulation of time in bed (7, 8.5, and 10 hours) for four consecutive nights per condition, with sleep duration measured by polysomnography on the fourth night. Participants kept habitual weekday rise times; time in bed was varied purely by advancing bedtime. Because each participant serves as their own control across conditions, genetic and family confounding is removed by construction. The key limitation, which the authors state, is that this is an experiment-controlled schedule, not a demonstration that adolescents will advance their bedtimes in the real world.

Adolescents CAN substantially increase sleep duration by advancing bedtime, and this ability does not change between ages 10 and 21. PSG sleep duration rose from 402.8 min (SE 1.6) at 7 h time in bed to 470.6 min (SE 2.1) at 8.5 h to 527.5 min (SE 3.0) at 10 h — roughly 68 and 57 additional minutes per 1.5 h of extra opportunity. Sleep onset latency and wake after sleep onset both increased, but not enough to offset the gain. Sleep duration fell with age (1.55 min/year) but the time-in-bed effect did NOT interact with age. This directly challenges the strong version of the circadian argument for later start times — that adolescent bedtime is biologically fixed and therefore only the wake time is manipulable.

None for the internal comparison. This is a within-person experimental design; each adolescent is compared to himself across randomly ordered time-in-bed conditions, so heritable chronotype and sleep need are held constant by construction.