Too much sleep: when long nights are a warning sign

The Short Version

People who sleep very long do, on average, die sooner than people who sleep seven to nine hours, and both Matt Walker and Peter Attia say the most likely reason is that long sleep is a sign of something else [1][2]. Sick people sleep more; people whose sleep is broken stay in bed longer to make up for it; and people with depression often spend long hours in bed without being asleep for most of them [1][2][3]. Neither man rules out that too much sleep could be harmful in itself, and Attia calls the question "quite a controversial question in the sleep literature" [1][2]. The practical reading is that a regular need for ten or more hours is worth treating as information about your health or your sleep quality, and that for most people the real problem is still too little sleep [1][2]. If long sleep is new, persistent or comes with low mood or daytime exhaustion, see a doctor.

Table of contents

What They Said

Peter Attia: a curve that turns back up

Lewis Howes puts the question to Attia in everyday terms: "And is there such a thing as too much sleep?" If someone is getting 10, 12 or 14 hours every day, does that hurt the body? [1]

Attia's first response is about the state of the evidence. "Really an interesting question, by the way, and quite a controversial question in the sleep literature." [1] Then he states what is not in dispute: "So there is no question that hyper sleep has been associated with poor outcomes." [1]

He describes the shape of the data. Plot mortality against how long people sleep and the line is a curve [1]. At the short end, "people who don't get much sleep have a higher mortality" [1]. Mortality falls as sleep increases, and then it ticks up again among very long sleepers [1].

The question is what the upturn means. Historically, Attia says, it was explained by sick people sleeping more, and he thinks that covers most cases [1]. He adds that there is some emerging evidence of a downside beyond that explanation [1]. And he keeps it in proportion: most people have the opposite problem [1].

Attia's wider account of sleep explains why he does not treat more as automatically better. He describes good sleep as depending on three things moving the right way: "So good sleep is when melatonin and adenosine go up and cortisol comes down." [1] Adenosine is the pressure to sleep that builds through the day's physical and mental work, and napping lowers it [1]. So he is careful about adding sleep where it is not needed. For someone already sleeping seven and a half to eight hours a night, he sees no downside to a 20-minute nap but would be careful about going longer [1]. When he was getting only about five and a half hours a night on a hunting trip, by contrast, he made a midday nap of up to two hours his top priority, to recover one full cycle of about 90 minutes [1]. Extra sleep, in his handling of it, is something you take to repay a shortfall [1].

Matt Walker: three explanations, and a possibility he will not rule out

Walker gives the numbers. Mortality risk rises below seven hours of sleep and curves back up past about 10 to 11 hours [2].

He offers two explanations for the upturn that do not involve sleep doing harm. The first is illness: the long sleep may reflect a sickness that sleep could not overcome [2]. A body fighting disease calls for more sleep, and when the disease wins, the record shows a long sleeper who died [2].

The second is quality. People whose sleep is poor try to offset it with more time in bed [2]. Their long nights are a sign of bad sleep, not of abundant sleep [2].

This fits a point Walker makes throughout the same conversation: quality matters as much as quantity, and you cannot be unimpaired on eight hours of fragmented sleep any more than on four hours of good sleep [2]. He measures quality partly by sleep efficiency, the share of time in bed spent asleep, and puts the healthy level at about 85 percent or higher [2]. Someone in bed for eight and a half hours at that efficiency is awake for about half an hour in total [2]. A person who needs far longer in bed to feel rested may simply be asleep for much less of it [2].

Then he leaves the door open. He does not rule out that too much sleep is possible, in the way that too much food or too much water is possible [2]. In another episode he says the question of whether you can get too much sleep deserves a whole discussion to itself [3].

Walker on depression: long in bed is not long asleep

Walker's fullest treatment of the subject comes when he talks about depression, because that is where the idea of oversleeping is most established. "But one of the places where we see quote-unquote, too much sleep is in the depression literature." [3] The clinical term is hypersomnia, meaning increased or excessive sleep [3].

He then explains why he thinks the label is partly a measurement mistake, crediting his former student Kate Kaplan, now a clinical psychologist, and others who have looked at the data [3]. In the studies behind the conclusion that depressed patients sleep too long, the question patients were asked was "what time do you go to bed and what time do you wake up?" [3] On that question the finding is clear: "people with depression will be in bed for significantly longer periods of time" [3].

The error is in the next step. The inference, which he says could almost be called a conflation, is that "if you're in bed for longer, then you're sleeping for longer" [3]. When researchers asked a different question, what time did you fall asleep and what time did you wake, the picture changed. "That hypersomnia phenomenon is nowhere near as strong as you would have been lent to believe" [3].

His explanation comes from what depression is like. One of its features is not wanting to interact with the world, he says, "And what better place to spend if that's your mentality than this thing called bed?" [3] He gives the thought in the patient's voice: "I just don't want to get out of bed." [3]

Meanwhile the sleep itself is usually worse, not more plentiful. "what we know is that depression will disrupt your sleep and make your sleep shorter," he says [3]. Patients have trouble staying asleep: "They wake up in the middle of the night. They can't get back to sleep." [3] Duration and quality both fall [3].

So a person who reports twelve hours in bed during a depressive spell may be short of sleep [3]. The long night is a marker of the mood, and of broken sleep, more than a surplus [3].

Walker adds a point about timing that bears on long, late mornings. Of the four things he tracks about sleep, quantity, quality, regularity and timing, he says timing may have the strongest evidence in depression, because a mismatch between a person's sleep schedule and their body clock predicts it [3]. Andrew Huberman gives his own case: when he sleeps from about 8:30 or 9 p.m. to 4:30 or 5 a.m., it has a strong antidepressant effect on him, and when he gets the same amount of sleep from midnight or 1 a.m. to 8 or 9 a.m., he carries a low-level depression [3]. The hours were equal; where they fell was not [3].

Huberman: how much most people need

In a separate answer about deep sleep, Huberman supplies the baseline against which long sleep can be judged. Sleep needs vary with stress and physical strain, he says; children and teenagers need more, and needs decline with age [4]. For about 95 percent of people he suggests at least six and ideally seven to eight hours [4]. He personally feels good on six and a half to seven, especially with a nap or non-sleep deep rest [4].

His test of whether you are getting enough is daytime sleepiness. If you keep falling asleep during the day, "That's sort of a duh, but it's an important duh to pay attention to." [4] Wanting a 20 to 30 minute nap in the afternoon, as he does, does not by itself mean you are short [4].

He also cautions that the numbers people use to judge their sleep are rough. Consumer trackers estimate sleep from heart rate, heart rate variability and movement, and no consumer device is perfectly accurate [4]. That applies to totals as much as to stages: a device that counts lying still as sleeping will make the same mistake Walker describes in the depression studies [3][4].

Where They Agree, and Where They Don't

They agree on the data. Both describe a curve with higher mortality at short sleep and again at very long sleep [1][2]. Walker places the upturn past about 10 to 11 hours; Attia does not give a number [1][2].

They agree on the leading explanation. Illness comes first for both: sick people sleep more [1][2]. Walker adds two further explanations that Attia does not mention, poor-quality sleep compensated by longer time in bed, and time in bed mistaken for sleep in depression [2][3].

They agree that the direct question is open. Attia says there is some emerging evidence of a downside to long sleep beyond the illness explanation [1]. Walker says he does not rule it out and compares it to overeating or drinking too much water [2]. Neither claims to know that long sleep in a healthy person does harm, and neither claims that it cannot [1][2].

They differ slightly in emphasis. Attia calls the matter controversial and leans on the historical explanation while noting newer evidence [1]. Walker's instinct is to look at what the long hours consist of: how much is sleep, and how good it is [2][3]. The two views fit together. Attia is describing the debate in the research, and Walker is describing what a long night usually turns out to be when examined [1][2][3].

They agree on which problem is common. Attia says most people have the opposite problem [1]. Walker's working target in the same conversation is seven to nine hours of good-quality sleep, and he says no one should feel guilty for sleeping, calling sleep "a civil right of all human beings." [2]

Why It Works

The sources give four ways a long night can be a warning sign without being the cause of anything [1][2][3].

Illness raises the demand for sleep. This is the historical explanation Attia cites and the first Walker gives [1][2]. In a population study, people who are already unwell show up as long sleepers and then as deaths, which bends the curve upward at the long end [1][2].

Fragmented sleep stretches the night. If sleep is broken, a person needs more hours in bed to collect the same amount [2]. Walker's efficiency figure makes this concrete: at 85 percent, eight and a half hours in bed leaves about half an hour awake, and lower efficiency means proportionally more time awake for the same time in bed [2]. Long time in bed can therefore be the visible sign of poor sleep [2].

Bed is a refuge in depression. People who do not want to face the day stay where they are, so reported bedtimes and rising times overstate sleep [3]. Their actual sleep tends to be short and interrupted [3].

Timing can be wrong even when the total is right. Circadian mismatch predicts depression, and the same number of hours taken at the wrong time left Huberman feeling low [3]. A habit of sleeping far into the morning may be a schedule problem more than a quantity problem [3].

On the other side is the possibility both men leave open: that, as with food and water, there is an amount of sleep beyond what the body can use, and that it carries a cost [1][2]. The sources do not say what that cost would be or how it would arise [1][2].

One more mechanism in these conversations explains why trying to bank extra sleep can backfire. Sleep pressure, carried by adenosine, has to build through the day for good sleep that night [1]. Walker's advice after a bad night is to "do nothing": do not sleep in, do not nap, do not go to bed early, because each one undermines the next night [2]. Attia's caution about long naps for people who already sleep enough rests on the same logic [1]. More time asleep or in bed now can mean worse sleep later [1][2].

Where It Applies, and Where It Doesn't

It applies to habitual very long sleep. Walker's upturn begins past about 10 to 11 hours, and Howes' question concerned 10 to 14 hours every day [1][2]. Eight or nine hours is inside the range Walker treats as healthy [2].

It does not apply to catching up. Attia took long naps when his nights were cut short, and described it as replacing a missing sleep cycle [1]. A long sleep after real deprivation is repayment [1].

It does not apply in the same way to the young, or to periods of heavy strain. Huberman says children and teenagers need more sleep, and that need rises with stress and physical load [4].

A modest lie-in is not oversleeping. Walker's best tool for more REM sleep is to sleep 15 to 25 minutes later in the morning [3]. What he warns against is shifting bed and wake times by two or three hours at weekends, which he compares to flying between Los Angeles and New York every week [3].

It is not a reason to cut sleep. Both experts stress that too little sleep is the usual problem, and nothing here supports sleeping less than you need [1][2]. Walker has said he regrets earlier messaging that made anxious sleepers worse, and the same care applies to worrying about sleeping too much [2].

It is an association in populations. The curve describes groups, and the explanations for it are, by Attia's account, contested [1]. It cannot tell an individual that their long sleep is harmless or harmful.

Changes in sleep can be a serious signal. Walker says that in the research on suicide, sleep disruption comes first, which is why he calls it a "canary in the coal mine." [3] He is speaking mainly of short and broken sleep and of nightmares, but his point that a change in someone's sleep pattern deserves attention applies to anyone whose sleep has shifted along with their mood [3]. If that describes you or someone close to you, contact a doctor or a crisis line.

What To Do With It

1. Separate time in bed from time asleep. Ask Walker's second question of yourself: when did I fall asleep and when did I wake, not when did I get into and out of bed [3].

2. Check the quality. If you need many more hours than you used to, look for waking in the night and long stretches awake; about 85 percent of time in bed asleep is the healthy mark Walker uses [2].

3. Use the daytime test. If you sleep long and still keep falling asleep during the day, treat that as a sign that something is off [4].

4. Take new or persistent long sleep to a doctor. Illness is the leading explanation both experts give for the link between long sleep and poor outcomes [1][2].

5. Notice mood. Long hours in bed with low mood and no wish to get up fits Walker's description of depression, where the sleep itself is usually short and broken; that is a reason to seek help [3].

6. Look at timing. If the long sleep runs late into the morning, consider whether your schedule matches your body clock; Walker suggests finding your chronotype and aligning with it as far as life allows [3].

7. Do not chase lost sleep with long mornings. After a bad night, Walker's rule is to keep to your normal schedule, and a lie-in of 15 to 25 minutes is the most he recommends for extra REM [2][3].

8. Keep naps short if your nights are already full. Attia sees no harm in 20 minutes for someone sleeping seven and a half to eight hours and is wary of more [1].

9. Aim for the middle. Seven to nine hours of good-quality sleep is Walker's target; Huberman suggests seven to eight for most people [2][4].

Sources

1. Key Health Habits to Live Over 100 & Prevent Chronic Diseases with Peter Attia · The School of Greatness
2. Dr. Matt Walker on Perfecting Your Sleep: Sleep Stages, Caffeine, Alcohol, Supplements, Naps and Unconventional Tips · Huberman Lab
3. Dr. Matt Walker: Improve Sleep to Boost Mood & Emotional Regulation · Huberman Lab
4. AMA #2: Improve Sleep, Reduce Sugar Cravings, Optimal Protein Intake, Stretching Frequency & More · Huberman Lab