Is it true that sleep before midnight is better than sleep after midnight?
Falling asleep late is associated with adverse health effects, but 'before midnight' is not a magic cut-off: those who fall asleep early but sleep briefly also sleep poorly. A consistent rhythm and sufficient total sleep duration count at least as much.
Direct evidence for the idea that 'sleeping before midnight is better' is entirely absent from the available studies. What does exist are indications that a late sleep-onset time is associated with poorer health outcomes, but those indications also show that the picture is not as straightforward as the folk wisdom suggests.
In British teenagers (aged 13-15, over 10,000 participants), falling asleep later was clearly associated with higher rates of overweight and obesity. Boys who fell asleep after midnight had a 76% greater chance of being overweight than peers who fell asleep before 22:00. This is an observational association: cause and effect have not been proven, but the signal is remarkably strong.
Interesting is the study among female university students. During the academic year, a larger proportion fell asleep before midnight than during the summer holidays. Yet sleep quality during that period was actually worse, and sleep satisfaction plummeted from 25% to less than 10%. The reason: total sleep duration dropped from nearly nine hours to seven hours. Falling asleep early did not help when early rising was also required. This makes clear that timing alone is not the whole answer: how much total sleep you get counts at least as heavily.
What is well supported is the importance of a stable day-night rhythm. Research among people who overwinter in Antarctica shows that disrupted circadian rhythms lead to greater sleep fragmentation, less deep sleep, and a poorer sense of feeling rested. This supports the idea that sleeping regularly and at a fixed time is beneficial, but it does not draw a hard line at midnight.
The conclusion is therefore more nuanced than the folk wisdom. A structurally late sleep-onset time, certainly beyond midnight, is associated with adverse outcomes in young people. But if falling asleep early comes at the cost of less total sleep, it defeats its own purpose. Your body needs sufficient sleep and a stable rhythm, more than it needs to meet a specific deadline at midnight.
Four studies used: one large observational study in adolescents (n=10,619, PMID 34291853), one quality-improvement project in an ICU (PMID 40168014), one observational study in university students (PMID 39379085), and one field study in Antarctica (PMID 28460798). No RCTs are available on this specific timing question. The evidence is associational or drawn from unusual circumstances; direct causality has not been demonstrated.