“Why We Sleep” Is Riddled with Scientific and Factual Errors | Hacker News Reader

“Why We Sleep” Is Riddled with Scientific and Factual Errors

guzey.com

Previous discussion: https://news.ycombinator.com/item?id=21546850

While true, it appears to have been extensively edited and added to since then.

Do you have a changelog?

Only informal. The author mentions extensive time spent on this since original publication, mentions questions he's responded to or sought more advice on since publication, and I noticed that that last-updated date, near the top, changes from 2-25 to 2-26. It seems to be actively maintained. While I understand HN's preference for no duplicate posts, this seems to have evolved significantly since original posting.

I've seen this posted all over the internet to great acclaim, and honestly, it just feels feverish and desperate to me, grasping at tiny points as if they refute the central thesis.

For example, the point that you linked to says that the official recommendations on sleep are "7 to 9 hours", which Walker simplified to "an average of 8 hours". I don't see how this simplification is so bad. The point is that these numbers are all much higher than the 5-6 hours many people claim to subsist on. It doesn't even change the numbers that much: slightly more than half of people sleep less than 8 hours, and slightly less than half of people sleep less than 7 hours. This nitpick does not refute the central point that having 5-6 hours of sleep generally impairs you, it only takes aim at phrasing.

I find point #2 (of 5) in this article especially laughable. Here it's claimed that sleep deprivation is actually a good thing, because it can be used as a temporary treatment for severe depression. Under this logic, bulimia is a healthy habit because you should induce vomiting after being poisoned.

I don’t see how this post can be seen to be nitpicking. Many of the points are that there is zero evidence to back up the authors claims, while using his scientific credentials as authority on the topic.

If the claims in this post are true, the author of “Why we Sleep” has behaved appallingly IMO.

I keep seeing this defence of Walker that amounts to, "Yeah, so what he's bloviating. What's the harm in recommending more sleep?"

But there is a lot of harm in recommending more sleep. If someone only needs 6-7 hours but keeps trying to force 8-9, they're going to have all sorts of trouble.

I personally have an issue with a drifting circadian rhythm that makes it impossible to go to sleep at the same time every day if I sleep a full 8-9 hours a night. The cumulative effects of this gives me very poor quality sleep and negatively affects my social life. Changing to sleeping only 7-8 hours addresses this, and I've been testing the cognitive effects of this change to see there aren't any.

I think Walker's overall message is important, but the numerous fabrications and massaging of data he's committed to convey that message is bad science. If you tell a lie (2/3 of people are sleep deprived, if you sleep less than 8 hours you're killing yourself) to dispel another lie (you don't need more than 6 hours sleep, keep hustling and chugging 5 coffees a day), you aren't actually making things better. You're just replacing one problem with another.

> they're going to have all sorts of trouble

1) Like what?

2) Is oversleeping even possible with normal sleeping patterns? When I wake up in the morning well-rested, I am woken up. I could spend the next six hours in bed employing all sorts of mental trickery to squeeze another few minutes of sleep, but it's not going to happen.

Yes, "oversleeping" is definitely possible. I found it to be well correlated in my N=1 to fighting off a cold or other infection.

Typically increases my 6-7h of sleep to some 10h. Unfortunately I do not have the hardware to track sleep quality.

Not sure if restriction would help or harm recovery.

Mostly agreed. I think everyone (including myself) was ready for a huge pop science book to be bad, and boosted the post based just on its title.

In defense of the post, Walker's called-out claim that "Routinely sleeping less than six or seven hours...double[s] your risk of cancer" does seem pretty out there. Even if that correlation were well-established it would be hard to interpret.

Walker wrote that 2/3rds of adults get less than the recommended 8 hours. But 8 hours aren’t recommended, 7-9 are. And that means far less than 2/3rds of people are failing to meet that. That’s a huge deal for a fundamental claim.

That's assuming the distribution within that range is uniform. The edge case would be that 1% of population being recommended 7 hours and 99% recommended 9 hours. But the recommendations of that nature are typically imprecise anyways.

> For example, the point that you linked to says that the official recommendations on sleep are "7 to 9 hours", which Walker simplified to "an average of 8 hours". I don't see how this simplification is so bad

The simplification is no bad, most people need those hours for day to day. But there was a fair point made, within the nonsense, not everyone is the same.

I, hand on heart, know people who can roll on 3-4 hours sleep over the course of 4 months, fully aware and capable, there are those who are trained for it also, so outside the natural feeling to sleep, it can be (to an extent) manipulated.

> Here it's claimed that sleep deprivation is actually a good thing, because it can be used as a temporary treatment for severe depression

As an option its a super super quick temporary solution, just because your body and mind is working on other things.

It's not a treatment at all, just a distraction, like drinking, so I agree with you there and I think its bad advise to promote such, same way as drinking..

If you have any emotional issues, talk to a professional, not HN. - And no I'm not talking to OP or anyone directly, just in general (I been there!)

> As an option its a super super quick temporary solution, just because your body and mind is working on other things.

There are no permanent solutions for depression other than long term maintenance on anti-depressants. A couple nights worth of sleep deprivations seems to be miracle cure compared to the effects and lack of efficacy of anti-depressants.

> It's not a treatment at all, just a distraction, like drinking, so I agree with you there and I think its bad advise to promote such, same way as drinking..

I fail to understand the why it is a distraction vs. a cure? Sleep deprivation cures some forms of depression temporarily without use of drugs. Why is that a distraction?

I don't know if it this blog is related but there is a stats podcast from the BBC where they look at the studies and even talk to Matt Walker [0]

at least the dying from too little sleep part is wrong(at very least wrong in the way it's presented ). And no it is not enough for the central thesis to be right to either deliberately or mistakenly lie or make up facts or quotes to undermine your story.

Sure you can do that but then it shouldn't be considered scientific

[0] More or Less: Behind the Stats: WS More or Less: Dozy Science http://www.bbc.co.uk/programmes/p0819trp

Your comparison of bulimia to sleep deprivation makes no sense. Bulimia might actually be a worthwhile cure to research if people often got poisoned without realizing it and this poisoning was a problem that affected them everyday.

Also, I didn't read his point 2 as saying sleep deprivation is good. I see it as refuting the idea that not having "enough" sleep absolutely bad.

It only applies as a temporary treatment for a small subset of people with a severe condition.

Respectfully, who cares? In biology, nothing is an absolute. All good things have bad sides and all bad things have good sides, and that shouldn't stop us from identifying things that are generally good. Nobody giving general advice can or should be expected to cover every edge case.

Disrespectfully, for someone that harps on about the main thesis, you fail to ascertain the blog author's main point:

Matt Walker is lacking integrity. Despite his Ivy League status, despite his PhD and numerous accomplishments, and despite all his prestige and educational attainment, he's still a high-quality conman.

That's the point. He twisted facts to sell a book and abused his position of authority to mislead others for money.

Oops, I didn't mean to link to an anchor! I don't see a way for me to fix that. Apologies.

While I can't speak to the accuracy of the book criticized, science in general is riddled with similar problems. Follow the citation trail and you'll often find that a cited article doesn't say what was claimed or says something similar but not quite the same. Alternatively, you might see that the cited article does say what is claimed, but the evidence is weak.

When researchers talk about all of the "low hanging fruit" being taken, it seems to me that they're blind to all the nonsense that appears once you start following the citation trail. Maybe every topic has been touched, but even something that seems definitive in a review article could have major flaws when examined more closely.

I'm almost done a PhD in engineering, and this has been my experience at least. I try to "debunk" something in roughly half my publications now.

Edit: I don't mean to suggest that identifying many of these problems is easy, just that it's not done frequently enough. For example, if you're doing research in a particular field, you're probably basing it partly on previous review articles. Take a look at some primary sources in addition to that. This applies extra if you're writing a review article. Don't just mirror what some previous review articles say and cite some newer papers. Find some old but good papers that were missed by previous reviews. Check primary sources. Etc. This is the job of a someone writing a review in my view.

It's also quite jarring to come across a paper on a niche topic (e.g a specific hormone's effects on certain biological processes), that completely botches simple, fundamental and accepted facts in your field; facts that the main thesis relies on.

It's absolutely silly and I find myself finding little of use in the author's conclusions, or their observations. After spending way too much time parsing through endless research papers, the only things I pay attention anymore are methodology and data.

This tells me: is the data relevant to my work? And was the data collected "properly" (I swear, half the time the researchers half-ass methodology that the results are fairly worthless)

> the only things I pay attention anymore are methodology and data.

> This tells me: is the data relevant to my work? And was the data collected "properly" (I swear, half the time the researchers half-ass methodology that the results are fairly worthless)

I've come to the same conclusion, and I'm a theorist myself. Theory is hit or miss, mostly miss. I still read it just to find the nugget of truth if there is one. If the data is relevant and was collected properly then it can be a goldmine, regardless of the theoretical explanation given by the researchers. But often the researchers measure the wrong thing, or measure the right thing in the wrong conditions or with some other problem. (One major problem I've found is that the uncertainties on many measurements in my field are enormous, but don't need to be, and very few seems to have noticed or cared.)

I got a strong motivated reasoning/bullshit vibe from Walker in this interview: https://www.npr.org/2018/07/20/630792401/sleep-scientist-war...

Particularly this section:

> “Sleep is not like the bank. So you can't accumulate debt and then try and pay it off at a later point in time. And the reason is this - we know that if I were to deprive you of sleep for an entire night - take away eight hours - and then in the subsequent nights, I give you all of the sleep that you want - however much you wish to consume - you never get back all that you lost. You will sleep longer, but you will never achieve that full eight-hour repayment as it were. So the brain has no capacity to get back that lost sleep...”

I don’t think this follows - seems likely to me that sleep is not some linear time thing and that there’s a standard overhead that doesn’t need to be repeated to extend and make up the time. This feels like a symptom of not understanding the mechanism and making a bad assumption.

I also found the “I won’t mention the cognitive failures I can detect” irritating. If there’s some actual thing to mention, say it - this kind of thing sets off alarms for me.

It doesn’t surprise me that the rest is similarly bad, I’m glad someone dug into it.

Not to choose a side, but sleep deprivation can be related to running an engine with low oil or running it while overheating. Extreme exhaustion is usually a chemically imbalanced period.

If we see 6 hour workdays become the norm during the automation boom, we may see that shifting to 12 hour business days with a well rested workforce was the only way to keep up with the next wave of innovation while keeping 1st world working conditions.

Would you invest an extra 2 hours a day to your sleep?

The author of "Why We Sleep" has written a blog post responding to the OP article and other questions from readers:

https://sleepdiplomat.wordpress.com/2019/12/19/why-we-sleep-...

There's no evidence that this was written by the author of Why We Sleep. This blog never links to Walker or to his social media or email, was never linked by Walker, and is not connected to Walker in any way other than using a "sleepdiplomat" handle (for all we know, I could've written that post).

Random quote from the above: "The book’s misattribution of the CDC statement to the WHO will be corrected in the next edition. "

Walker said this in [1] and I believe that this position is indefensible

At the time at the time of the book’s publication in 2017 (and today), CDC had no documents and no pages on its site that would declare a sleep loss epidemic.

As I noted in section Possible origin of the “sleeplessness epidemic” thing [2],

>Between late 2010/early 2011 and August/September 2015 [3-6], CDC had a page on its site titled “Insufficient Sleep Is a Public Health Epidemic”. More than 2 years before Why We Sleep was published, the page changed the word “epidemic” to “problem”, so that its title became “Insufficient Sleep Is a Public Health Problem”. … the fact that CDC itself changed the wording from “epidemic” to “problem” more than 2 years prior to the book’s publication, indicat[es] that they no longer believed in the presence of an “insufficient sleep epidemic”

Issuing a correction that would just change “WHO” to “CDC” in the next edition of the book – as Walker suggested – is deceptive. The CDC long ago (more than 4 years ago) removed the word “epidemic” from the article, and then removed the article itself.

If Walker believes that that is sufficient to state that CDC has declared a sleep loss epidemic, he might as well say that CDC has declared an epidemic of inhalation anthrax (and forget to note that it was declared in 1958 [3])

Also, Walker never explains why the source for this sleep loss epidemic claim in the book is a random National Geographic documentary.

[1] https://www.bbc.co.uk/programmes/w3csz3s6

[2] https://guzey.com/books/why-we-sleep/#possible-origin-of-the...)

[3] https://perma.cc/72CZ-L9DG

[4] [https://perma.cc/23UM-849W

[5] [https://perma.cc/2E68-ZRL9

[6] [https://perma.cc/UJ4W-JHNY

[7] https://www.cdc.gov/museum/timeline/1940-1970.html

>I have many stories of people who slept well on less than eight hours of sleep, read Walker’s book, tried to get more sleep and this led to more time awake, frustration, worry, sleep-related anxiety, and insomnia.

This is me. I get at most 7 hours, usually 6. I only feel impaired if I drop down to 4.

Reading about Matthew Walker's research years ago caused much anxiety and a loss of sleep, how ironic.

Some of the points feel a little bit pedantic (or even semantic?).

Not sure why this links to section 5 but the entire page is worth reading. It a little ironic to say that this article changed my perspective on sleeping (considering this article is showing how you shouldn't believe the first thing you read) but given how little research I had done prior, I think I at least know now that there's a lot more that I don't know.

I accidentally kept an anchor in the URL. My apologies!

Reply on news.ycombinator.com