Sleep advice circulates widely and much of it rests on single studies or on extrapolation. A few things are well established.
What is solid
Chronic short sleep is associated with substantially worse outcomes across cardiovascular, metabolic and cognitive measures. This is consistent across large populations and long follow-up.
Regularity matters, possibly as much as duration. Going to sleep and waking at consistent times produces better outcomes than the same total hours taken irregularly.
Light exposure sets the internal clock, and morning light is the strongest signal for keeping it aligned.
And sleep debt does not fully repay. Weekend recovery sleep improves how you feel and does not reverse the measured effects of weekday restriction.
What is weaker than it sounds
The eight-hour figure, which is an average of a distribution, not a requirement. Individual need varies meaningfully and some people genuinely function well on less.
Specific claims about sleep stages, which consumer devices report with a confidence their accuracy does not support. Compared against laboratory measurement, stage detection in wearables is approximate.
Most supplements, where evidence is thin and dose-response poorly characterised.
What works, unglamorously
Consistent timing. A cool, dark room. Limiting caffeine after the early afternoon, since its half-life is around five hours. Getting out of bed when unable to sleep rather than lying there, which is a core element of the behavioural treatment for insomnia.
The thing worth knowing
Behavioural therapy for insomnia outperforms medication for long-term outcomes and is far less prescribed, largely because it takes more effort to deliver.
Anyone with persistent sleep problems should be asking a doctor about it by name.