How much sleep do people need?

FullSizeRender-7When it comes to time, certain narratives get repeated a lot. One is that we are all increasingly sleep-deprived. Certainly, some people are sleep-deprived; probably a number of readers of this blog! Many people have bad nights here and there. I know I do. Human nature means these bad nights stand out more in the mind than the good nights. However, looking at time diary data from the American Time Use Survey, we find that the average American sleeps 8.8 hours per day.  This comes out to 8.54 hours on weekdays and 9.40 on weekends and holidays. Even employed people with children under 18 years of age in their households slept 8.44 hours/day on average. This breaks down as 8.30 for men and 8.58 for women.*

On one hand, these figures look like they are worth celebrating. Generally people talk about getting 8 hours/day as a good number, and it seems the average person is hitting this.

On the other hand, there is some interesting new research finding that the 8 hours target is not based on solid evidence. Indeed, large studies are finding that people who sleep between 6.5 hours and 7.5 hours tend to do best. Both short sleepers (less than 6.5 hours) and long sleepers tend to do worse on various metrics, but the definition of long sleeping, where the problems start, seems to be right around 8 hours.

The question of correlation vs. causation is of course tricky here. It is unclear if long sleepers would improve outcomes by sleeping less. It is possible that some people need more sleep, and these people also have worse health outcomes, but if they slept less they’d still have bad outcomes and feel miserably sleep-deprived. But this does hint intriguingly at a very different story line. The average American is not sleeping too little. It may instead be true that the average American is sleeping too much for optimal health.

When I spent the past year tracking my time, I learned I spent 51.81 hours/week sleeping. This comes out to 7.40 hours/day. The quality was not as good as I would have liked. I logged 146 interrupted nights, and I defined an interrupted night as being woken up before 4:30 A.M. (after that, it’s just a short night!) So 146 probably understates the non-optimal nature of my toddler’s sleeping, and hence mine.

However, I think the quantity is probably about where it should be. I have logged weeks in the past when I was not caring for a toddler and the totals came up pretty similarly — somewhere around 52 hours/week. In general, I think my body needs just a bit under 7.5 hours, and one way or another, I tend to get it.

These past 16 months have featured a lot of “one way or another.” Right now I am making myself go to bed strictly at 10:30 P.M. because I never know for sure when the day will start, or if I will be woken in the middle of the night. It is happening less than it used to, but it still happens. I do not set an alarm because I know I will be woken up, and also because I want to capture any extra minutes of sleep I can to make up for the shorter nights. As a mom of four, I know from past experience that things will be different in another 12-18 months. And so I have been pondering what I would like my sleep schedule to look like when the littlest is 3 years old.

I think ideally I will go to bed around 10:45 and wake up at 6:00. I will work or read for an hour before starting the day with showers and kid routines. This weekday total of 7.25 hours will be supplemented by a few extra minutes on weekends. That is the key problem now, that I cannot make up the time easily on weekends without arranging for a nap. Having the extra hour in the morning on weekdays will give me some extra time for things I want to do.

It is not that I will be sleeping less or more, it is that the sleep will be more orderly and planned. That makes it possible to use the time in chosen ways.

Interestingly, the idea of sleeping a wee bit less was one of Linda Formichelli’s suggestions in this blog post on how to have an extra hour a day for fun. If you are sleeping 8.8 hours/day on average, cutting this down to 7.8 hours does open up space — if it is not a tip you’ll read many other places. I know from logging my time that under 6 hours is the danger zone for me. I stop being able to function. Between 6-6.5 is functional foggy. 6.5 to 7 is only mildly foggy. Anywhere between 7-7.5 hours is generally great.

* New here? A quick introduction to the ATUS — this is an annual survey done by the Bureau of Labor Statistics. Researchers ask thousands of Americans to report how they spent yesterday (rolling throughout the year, so all days basically become yesterday at some point). This methodology tends to get different answers than asking people how they spent a “typical” day, which is a judgement call. Since researchers aren’t asking about any specific category of time, this methodology also reduces the temptation for people to give socially desirable answers. 

In other news: Linda, who has guest-posted here before, has a new book out called How To Do It All: The Revolutionary Plan to Create a Full, Meaningful Life — While Only Occasionally Wanting to Poke Your Eyes Out with a Sharpie.

17 thoughts on “How much sleep do people need?

  1. One of the scary functions of my Fitbit is getting data on hours of actual sleep – not time in bed. The two numbers are very different for life with a 7 month old!! I’m not sure how accurate the Fitbit is, but interestingly, it has been reporting a typical 7.5 hours of sleep per night which matches up well with your post. Of course I get that by going to bed soon after the baby does… But I feel rested most days.

  2. The general concern with those studies correlating long sleep time to poor outcomes is that perhaps those with the longest sleep times have underlying medical issues that cause fatigue or require excess sleep (this includes depression).
    Also, time use surveys don’t (can’t) track sleep quality or even actual sleep—I may put “sleep” from 10 PM to 6 AM, but between 10-11:30 maybe I’m trying to fall asleep. Or I wake up at 5 and just lie there for an hour. I’m not doing anything else, but I’m not really “sleeping” either.
    If I’m sleeping regularly most nights, I’ll end up with 7.5 hours; but if there are a few nights of less sleep, I’ll sleep more (by going to bed early!) to make up for it. I wish I needed less sleep. If I try to give up sleep to do other things, after a few days I notice that I’m less productive and motivated, so its not a long term way to “buy an extra hour”

    1. @Ana- definitely true on the underlying medical issues that are causing both fatigue and the poor health outcomes. However the lying-in-bed thing is not completely discounted by time diary studies. Some people write just that: “lie in bed awake” on their logs. When I’ve interviewed people at the ATUS, they talk about distinguishing lying in bed vs. asleep (not that we know exactly when we fall asleep, but people are sometimes cognizant of lying in bed for a while and will mention it as something they did next during the day).
      I agree that cutting sleep from what you need is not a wise way to be productive. However, I think all of this raises the question — it is certainly possible to eat more than you need to (or that is healthy). I imagine it could be possible to sleep more than you need to as well.

      1. ” I imagine it could be possible to sleep more than you need to as well.” I honestly wonder if this is true—I went to bed extra early last night and was wide awake exactly 7.5 hours later, at 5:30 without an alarm (electronic or otherwise!). Unless I’m sleep deprived and need extra, I just CAN’T sleep more.

        1. I have to put myself in the “no” category on this one — in my book, if you can fall asleep, you probably should fall asleep (if you can sleep, you probably need sleep). That’s not to say that there might not be underlying medical conditions that should be treated and treatment of which will reduce the need for sleep (hypothyroidism, anyone?), nor that there aren’t tradeoffs with other ways in which one might use time, but all the same.

          No scientific evidence on this one, but my gut observation based on my own behavior/functioning (at that of those around me) is that unlike eating too much, one cannot sleep too much.

    2. I don’t know how prevalent sleep apnea is, but that seems like another candidate for a condition that would appear to cause excess sleep.

  3. I was just listening to the Happier Podcast, with Gretchen Rubin, and this week’s guest was Arianna Huffington. Gretchen and Arianna were waxing poetic on the need for sleep. I see your time- tracking as a good partner to this, especially if someone believes they are not getting enough sleep. They can track their time and see what changes they may be able to make, i.e. extended phone/laptop/tv time before bedtime coupled with early morning wakeups ( and middle of the night interruptions) by my children made me realize I need to be in bed and logged off by 10, so now I shoot to unwind by 9:30.
    You would be a great guest for the Happier podcast!

  4. Laura I wonder if your thoughts will change when you are older. I’ve begun to have perimenopausal symptoms and one of them is more interrupted sleep. Every little thing wakes me up that I used to sleep through. I use a white noise machine and a sleep mask, which help. I also have an 18 year old who is sometimes coming home late, and I don’t sleep until he’s home.
    On the rare day that my husband takes the baby in the morning and I can sleep late, I feel like a NEW woman. I have read the same thing you have about it being possible to sleep too much, but it’s not my problem. Lol

  5. Completely agree with Ana. People who have underlying sleep disorders (sleep apnea, in particular) will often think they are sleeping when they are frequently rousing to near-consciousness (as often as once a minute.) This causes severe fatigue and a host of other complications. This may be an artifact of my work cohort (I’m a doctor) but most of the people I work with get less than six hours of sleep a night and that is not too much; it’s not even enough. It also will depend on when you wake up in reference to you sleep cycle. If you wake up at the end of a sleep cycle, when you are coming to arousal naturally, you will feel fairly rested no matter how long you slept. If you wake in the middle of a sleep cycle, you will probably feel drowsy and groggy, no matter how long you slept. Sleep cycles vary; they average 2.5-3 hours. My guess is yours are 2.5 hours and that when you sleep 7.5 hours, you wake at the end of the third sleep cycle. The increase in interrupted sleep as we age comes from a reduced ability to smoothly transition from one sleep cycle to the next due to a chance in neurotransmitters.

  6. This is one of those areas that I’m not sure the ATUS is meaningful. Sure, it can answer the question how much does the average American sleep….but that has very little to do with how much sleep an individual gets or needs to be healthy.

    Thanks to seeing a sleep specialist (and iron and melatonin supplements for my KID and prescription sleep meds for me) I’m sleeping better than I have since he was born. But I’m still averaging less than the average American, and I’m still tired. As others say above, there are other variables…i.e. for me, sleeping 9-4 or 10-5 is not the same as sleeping 11-6.

    I also listened to Huffington on Rubin’s podcast…planning on reading her new book.

  7. I average just under 8 hours. If it drops much below 8, I am insanely fatigued. That said, I do have some health issues, so that may be a contributing factor, either that I need more sleep than average or my medications make me tired/hard to wake up. But if I don’t set an alarm, I sleep 8 hours before I wake on my own. I also agree with gwinne that there’s a component to WHEN you go to sleep that makes a difference. If I go to sleep too late (after 11), I’m dragging the next day, no matter how much I sleep.

    I track my sleep via the Sleep As Android app (and a Pebble watch), btw.

  8. I dunno, I am pretty obsessive about my sleep, an while yes I can function on less than 8, WHY????? I feel so much worse than with 8 to 8.5. So what if i’d be more productive! Who wants to go through life feeling like crap?

  9. Well, the ones interested in the subject are the militaries. For obvious reasons. So they found that the best for the physical health is … just 6h48m. But under 7h30m the friend-foe identification error within 1s interval is growing dramatically. Also this seems to be the best for the factual memory. And for the one to be on best form for mathematical and geometrical problems’ solution it is more than 8h15m. And there are even further limit, about 9h, very imprecise one when the people are able to connect previously uncorrelated issues in novel and fruitful way, so to be novelty creative. But take care, that the sleep above 8h has its drawbacks.

    Well as we see — we are not very well adjusted to have just one optimum.

  10. As others have suggested, quality of sleep affects my quantity of sleep needed. Just as you discuss in your books Laura- spending 40hrs per week with your child running around doing housework/errands is different than spending 30hrs per week with your child doing a mix of outings/playdates plus the necessary housework/errands at a minimum. The latter is more likely to leave you feeling like you spend adequate time with your family. Sleep is the same for me- pre-kid I needed 7.5hrs, on the dot. I think it is about the same now, but I am pretty inconsistent about getting to that average and spent the first 18months of her life with interrupted sleep. I spent 11hrs in bed a few nights ago- the first 7hrs were with my restless, ill toddler who woke me up briefly 4-5times wanting to snuggle, the last 4hrs were spent blissfully alone in the guest room. The next night I slept 7.5hrs alone and woke up on my own before my alarm. Going to bed before 10pm means that I am not woken up prematurely by the birds or my husband’s alarm clock. I have also noticed that stress management works wonders for my sleep, but it is challenging to execute. Turns out that I need a bedtime routine/transition time as much as my kid does, and what I need depends on what type of stress I had that day (end-of-life discussions at work call for a philosophical discussion with my husband before bed, alone-with-toddler day calls for spiced cider and the Mindy Project).

  11. This is very interesting. I have never had a chance to experiment with my sleep length yet. We a toddler who just finally started sleeping through the night, I couldn’t afford to experiment, because I need to make sure I am rested. However, seeing how I feel and perform after a certain number of hours of sleep would be very helpful. But I wonder if testing each pattern would be best for around 28-30, as the hormones might be greatly affecting how I feel throughout the month and a week might just not represent it very well.

    1. @Yana – I, too, feel like I couldn’t afford to experiment for a long time. I still can’t – which is why I’m going to bed at 10:30 whenever I can. However I do think this will be getting better soon, and 10:30-5:45 could work really well for me during the week if I could pull it off. We shall see…

Leave a Reply

Your email address will not be published.