Pediatricians say teenagers need eight to ten hours of sleep. In 2007, fewer than a third of American high schoolers got even the minimum on a school night. Sixteen years later it is fewer than a quarter — and the students getting five hours or less have nearly doubled. The eight-hour night is quietly becoming a minority experience of adolescence, with measurable consequences for mood.
The YRBS began asking about school-night sleep in 2007. The share reaching eight hours has fallen in almost every wave since — from 31.1% to 23.2% overall. The decline is steepest exactly where sleep needs are greatest: 9th-graders dropped from 42.3% to 31.0%, tenth graders from 32.4% to 22.0%. Seniors were already sleep-deprived in 2007 (21.8%) and simply stayed there (18.1%). Adolescence has always eroded sleep year by year; what changed is that the erosion now starts from a lower cliff. Girls sleep less than boys (21.7% vs 24.6% reaching eight hours in 2023) — a gap worth remembering two figures from now.
Averages hide where the sleep went. Comparing the full distribution in 2007, 2015, and 2023: the modal teen still sleeps 7 hours (29–30% in every wave, weirdly immovable), but the mass above 8 hours drained into the short-sleep tail. Eight-hour sleepers fell from 23.5% to 17.6%; nine-hour sleepers from 5.9% to 4.1%. Meanwhile 4-hours-or-less rose from 5.9% to 8.8% and five-hour nights from 10.0% to 14.4%. Nearly one teen in four now sleeps what sleep scientists would classify as a clinically short night — on a typical school night, by their own account.
The 2023 wave lets us line sleep up against the survey’s depression-screening item. The gradient is almost perfectly monotonic: persistent sadness afflicts 60.7% of teens sleeping four hours or less, 57.0% at five hours, 43.5% at six, 34.4% at seven, and 25.9% at eight — the risk drops by roughly nine points for every added hour of sleep. In a survey-weighted logistic model controlling for sex, race, grade, and heavy social-media use, an eight-hour night cuts the odds of persistent sadness by more than half (AOR 0.43). Heavy social media use — checking more than once an hour — independently raises those odds (AOR 1.28) and is concentrated among the shortest sleepers: only 21–22% of hourly-plus checkers reach eight hours, versus 31% of teens who use no social media at all.
Causality surely runs both ways — despair keeps people awake, and phones keep people awake, and being awake at 2 a.m. is its own kind of despair. The survey can’t untangle the loop. It can show that the three variables form one tightly wound knot, and that the knot tightened over exactly the years the phones arrived.