What research says about how much sleep you should get per night

Research Summary
Pascal Wallisch
Baby Sleep Study Co-Investigator

How much sleep should you be getting per night?

7-8 hours per night (9 hours if you want to play it safe) for adults (ages 18-60). Read on if you are interested in details of how this impacts different aspects of your health.

Here is what we know from the consensus statement on the recommended amount of sleep to promote optimal health in adults, by the American Academy of Sleep Medicine and Sleep Research Society. This statement is based on the judgment of a panel of 15 experts, who combined information from well over a thousand scientific papers specifically on the relationship between sleep duration and a range of different health outcomes, to determine what constitutes “sufficient sleep”. This was not an easy task, as there is a mix of information that is more or less reliable (self-report vs. more objective measures of sleep duration), observational vs. experimental studies, and studies with drastically different numbers of participants. Yet, a consistent picture emerged.

First, of all, there *is* indeed a sleep crisis. According to the CDC, the number of US adults that are now sleeping less than 6 hours a night – spoiler: this is not good for your health – is now at over 70 million, or about 1 in 3 adult Americans, double the number in 1985. This trend is so serious and that the CDC considers it a public health epidemic.

The panel considered the impact of sleep duration on 9 aspects of human health:

  1. General health: People sleeping less than 7 hours per night report poorer overall health and greater susceptibility to disease, particularly in young adults. There is no consensus about the other end of the range – people sleeping more than 9 hours per night might have poorer overall health, but there is much less evidence for that.  
  2. Cardiovascular health: Shorter sleep durations (particularly, less than 6 hours) are linked to heart disease and high blood pressure. Again, on the other end of the range, sleeping in excess of 8 hours a night, the picture is much less clear with regard to heart disease, as different studies report conflicting results.
  3. Metabolic health: The impact of sleep duration on hormones is fairly well understood and quite dramatic. Sleep deprivation reduces insulin sensitivity, lowers glucose tolerance and increases levels of cortisol, which – over time can lead to the development of type 2 diabetes. In addition, it lowers leptin and increases ghrelin, leading to more pronounced hunger, food intake and –ultimately – increased body weight. Results from cross-sectional and longitudinal studies, as well as experimental studies, are all consistent in this regard. In contrast, there is no consistent evidence for untoward outcomes from longer sleep durations.
  4. Mental health: Short sleep durations – particularly below 6 hours – are associated with a several kinds of mental illness, including depression and suicidal ideation. It is unclear, however, if lack of sleep causes depression or the other way around. Depression and stress often lead to insomnia, but there are also some experimental studies suggesting that short sleep durations directly contribute to depression. As with the other health aspects, there are some studies suggesting increased risk from excessive sleep, but the evidence for this is not consistent.
  5. Immune health: Short sleep durations (below 7 hours) impair killer cell function, render vaccines less effective and increase infection risk.
  6. Cognitive performance: These studies focused on long-term attention, processing speed and memory. Generally speaking, more sleep is better. For example, people sleeping 4 hours outperform those sleeping 3 or less. Even 9 hours of sleep seems to outperform 7 hours of sleep. In addition, cognitive deficits seem to accumulate as short sleep durations persist, although people suffering from these cognitive deficits are largely unaware of it. Driving performance is impaired starting at below 7 hours of sleep.
  7. Cancer: Most studies report no clear associations between sleep duration and cancer risk, although there is some evidence for an increased cancer risk for sleep durations below 5 or 6 hours of sleep.
  8. Pain: People sleeping less than 6 hours – and in particular those sleeping less than 5 hours – report more pain. However, like mental health, it’s unclear if lack of sleep increases pain or the other way around. Perhaps individuals can’t sleep because they are in pain. Again, the picture is unclear at the high end of the range, with little to no evidence for pain implications of longer sleep durations.
  9. Mortality. The lowest mortality rates are found for people who report 7-8 hours of sleep, with increased rates for either more or less than this range. This is a narrower range than any of the other health aspects that the panel looked at. Sleeping too much is clearly associated with greater risk for mortality, but a more in depth analysis of these results suggests that long sleep duration might be a marker of diseases that cause mortality, not long sleep duration per se – healthy individuals do not show a greater risk of dying if they sleep more 8 hours.


To summarize, it is clear that bad things start to happen health-wise, once sleep falls below 7 hours a night for extended periods of time. There is much less certainty about the other end of the range, at sleep durations in excess of 9 hours per night. The panel examined 9 separate aspects of health. They might have reached different conclusions about the recommended amount of sleep for different aspects of health. But instead a consistent picture emerged. This leaves us with a fairly broad plateau of somewhere between 7 and 9 hours of daily sleep that is considered to be optimal for human health (in adults). The fact that this range is so broad suggests either that each individual can adjust to different schedules or that different individuals need different amounts of sleep. The 2nd of these possibilities (that different individuals need different amounts of sleep) is supported by the fact that sleep duration is highly correlated among twins, suggesting a strong hereditary, biological component.

If you have spent much time on the internet , you will realize that pretty much every position will be taken by someone, no matter how outrageous. The topic of sleep is no exception. Everyone has to sleep, so the potential audience of interest is large, attracting no shortage of people who wittingly or unwittingly peddle positions that are patently absurd. Sleeping in spurts – or polyphasic sleeping – is now a “thing”, but would only make sense on a planet with multiple suns. Similarly, the idea that “natural sleep” is segmented into two phases is illogical considering how long a winter night can be in a pre-industrial society without artificial lighting in higher latitudes. Lack of sleep in our society is a public health epidemic, despite claims to the contrary.

Sleep is a biological imperative. We need to – as a society as well as individuals – learn to respect this or suffer the consequences.



NYU Baby Sleep Study needs your help. We are asking parents to track their baby's sleep, meals, and diaper changes on the smartphone app, Baby Connect. With this information, we hope to learn more about infant development and to recognize early markers, or red flags, for developmental disorders. This is science from the comfort of your own home. Sign up, tell your friends, advance science!