Hacker News new | past | comments | ask | show | jobs | submit login

Re: Why We Sleep by Matthew Walker, this critique of the book is worth a read: https://guzey.com/books/why-we-sleep/



Thanks for sharing. Very informative. Especially this quote gave me pause:

> I wanted to drop you a line to thank you for all the time and effort involved in debunking Matthew Walker’s book. As someone who works with individuals with insomnia on a daily basis, I know from firsthand experience the harm that Walker’s book is causing.

> 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. …


Disappointing... That book has been highly recommended ad infinitum on HN, to the point that it was sounding like gospel. Maybe that should've been a hint, heh.

Has Walker responded to that critique?



This may constitute a response, but it’s not clear if the original author wrote it in response to Guzey’s claims, or if it was the author or just an imitator. The writing style strongly suggests it’s the author, and it uses the author’s twitter handle and website name as the subdomain: https://sleepdiplomat.wordpress.com/2019/12/19/why-we-sleep-...


Glad someone posted this. This book gets so much praise. But like any popsci book, there are huge incentives for the author to exaggerate and manipulate to produce a clear and marketable message. Just consider that if there was no interesting story to tell about sleep, of course there would be no book.


Funnily enough, there's also a completely different book by the very same name, written by a German sleep researcher. And this one's actually really good (imo). [1]

[1] https://service.randomhouse.de/paperback/Why-We-Sleep/Albrec...


What a great critique. I'd like to see a similar treatment of Jason Hickel's "The Divide", which feels both terribly important and very biased and sloppy.


Agreed, I would pay something for a feed of critiques like this, would save me time dodging poorly constructed literature, I like books, they are like the internet but usually a little more refined.


I'd like to give this a more thorough read, but so far this critique strikes me as done in bad faith. I read the book. I'm also in the middle of a biophys PhD, so I like to think my opinion is "extra super special". Here are my comments:

Point 1: the chart bottoms out at 7, which falls within the range the book recommends. I'm fairly sure he recommends 7-9, and that the required amt varies from person to person. Another thought I had: metabolism and longevity go hand and hand. I mean, I just read a brand new paper from George Church's Harvard lab, showing that they reversed several chronic ailments in mice by inserting FGF21, which regulates glucose levels. And sleep absolutely regulates metabolism. Personally, I'm keeping my ears perked up when it comes to metabolism/circadian rhythms/homeostasis, etc.

Point 2: Depression is an incredibly complicated topic. It is a psychological construct, the net result of thousands and thousands of genes, filtered through a modern technological world, and then filtered through inventories, interviews, and assessments. For this reason, I am not at all surprised that Guzey was able to find studies that suggest that sleep deprivation might have some benefit for some people. I would HIGHLY recommend this new, open-access Nature review paper covering the genome wide studies on depression: https://doi.org/10.1038/s41398-019-0450-5. Its almost not worth pitting a complex phenotype like depression against another complex behavior like depression. But what the hell, let's brush with broad strokes: the significant genetic variants associated with depression have to do with regulating homeostasis (eg sirtuins). So I would not be surprised if good sleep is at least correlated with low levels of depression.

Point 3: I know almost nothing about FFI, but we have kept mice awake, and they do eventually die. I'm pretty sure humans would die too, but ethics precludes us from performing such a study.

Points 4&5: This is just fussing about Walker's writing and WHO. I'll have to agree with Guzey that Walker's citations and consistency are often weak. And I honestly couldn't care less about WHO. But I'm pretty confident that sleep quantity has declined with time across the world. I remain curious about the connections between light and circadian rhythms (I think retinal cells go straight to the superchiasmatic nucleus, the circadian rhythm controller). Also, sugar is a thing in the modern world.

So yeah, I'm just frustrated by his nitpicking, and his seeming lack of appreciation for sleep as an open biological question. We don't know why we sleep, really. If anything, the title is the worst part of the book. But at least the contents respect the question. This criticism does not. Nonetheless, I am a sucker for obsessive bloggers (eg slatestar, gwern, cowen, etc), so I will definitely be checking out Guzey's other writing. It looks interesting.


A couple of follow-ups here: Andrew Gelman has, I think, the best commentary on Guzey's critique.

https://statmodeling.stat.columbia.edu/2019/11/18/is-matthew...

https://statmodeling.stat.columbia.edu/2019/11/24/why-we-sle...

To my knowledge, Walker hasn't responded to Guzey's criticisms. The other issue that has emerged in the weeks since Guzey published his critique is that Walker seems to be using the erroneous claims in his book in his papers. Relevant exchange here: https://statmodeling.stat.columbia.edu/2019/11/24/why-we-sle...


Hi, I'm the author of that critique.

>Point 1: the chart bottoms out at 7, which falls within the range the book recommends. I'm fairly sure he recommends 7-9, and that the required amt varies from person to person. Another thought I had: metabolism and longevity go hand and hand. I mean, I just read a brand new paper from George Church's Harvard lab, showing that they reversed several chronic ailments in mice by inserting FGF21, which regulates glucose levels. And sleep absolutely regulates metabolism. Personally, I'm keeping my ears perked up when it comes to metabolism/circadian rhythms/homeostasis, etc.

The point you're making has nothing to do with the point I'm making in the relevant section. I take issue with Walker writing "the shorter your sleep, the shorter your life span". You seem to have missed this part entirely. If you say that the book recommends 7-9, feel free to quote the book because it definitely seems that he strongly advocates at least 8 hours. Example: as I showed in section 5, Walker takes 7-9 hours recommendation from the National Sleep Foundation and then falsely claims that they recommend 8 hours of sleep.

>Point 2: Depression is an incredibly complicated topic. It is a psychological construct, the net result of thousands and thousands of genes, filtered through a modern technological world, and then filtered through inventories, interviews, and assessments. For this reason, I am not at all surprised that Guzey was able to find studies that suggest that sleep deprivation might have some benefit for some people. I would HIGHLY recommend this new, open-access Nature review paper covering the genome wide studies on depression: https://doi.org/10.1038/s41398-019-0450-5. Its almost not worth pitting a complex phenotype like depression against another complex behavior like depression. But what the hell, let's brush with broad strokes: the significant genetic variants associated with depression have to do with regulating homeostasis (eg sirtuins). So I would not be surprised if good sleep is at least correlated with low levels of depression.

You missed the point I was making entirely. Walker wrote that there are no biological functions that do not benefit from a good night's sleep. I point out that this is false, as sleep deprivation therapy is a safe, effective, and a very-well studied treatment for depression.

>Point 3: I know almost nothing about FFI, but we have kept mice awake, and they do eventually die. I'm pretty sure humans would die too, but ethics precludes us from performing such a study.

You missed the point I was making entirely, again. Walker wrote that FFI demonstrates that lack of sleep kills people. I pointed out that saying that is completely false. What relation do mice have to the point I was making?

>Points 4&5: This is just fussing about Walker's writing and WHO. I'll have to agree with Guzey that Walker's citations and consistency are often weak. And I honestly couldn't care less about WHO. But I'm pretty confident that sleep quantity has declined with time across the world. I remain curious about the connections between light and circadian rhythms (I think retinal cells go straight to the superchiasmatic nucleus, the circadian rhythm controller). Also, sugar is a thing in the modern world.

Points 4&5: this again has very little relationship with what I was writing. In section 4, I pointed out that Walker has seemingly invented a sleep loss epidemic and attributed it to the WHO. This is not just an issue with citations. In section 5, I pointed out that Walker misrepresents National Sleep Foundations sleep guidelines, saying that they recommend 8 hours of sleep, while in reality they recommend 7-9 hours of sleep.

For readers of this exchange: if you're still unsure how serious my concerns with the book are, the clearest example is provided in section 18, where I show deliberate data manipulation by Walker. He simply edited out the part of the graph that contradicted his argument in the book: https://guzey.com/books/why-we-sleep/#appendix-what-do-you-d...


Hey Guzey, thanks for hopping in. I like your blog.

Walker's book is sloppy. The graph you mention last is egregious. No disagreements here. Honestly I agree with you at least partly on all of your points.

So why did I miss your points? I was trying to give better evidence/info relevant to the topic of each point. My hope was people would read it and continue to respect sleep as an open and very interesting biological question. Maybe it came off as ignoring the points. I understand that. But honestly I don't care about Walker that much - I care about sleep. Its really weird, and we don't understand it.

Out of curiosity, what is your current stance on sleep?

EDIT: after re-reading this comment, I realized it might seem a bit disingenuous, because I did criticize your arguments as bad faith. Clearly I care a little about your take-down of Walker.

I said this because you found claims in Walker's book, found errors in his evidence, and then denied the general claims in your bold-font headings. It seemed like you were dismissing the putative importance of sleep along with dismissing Walker the author. I felt this was uncharitable to the topic, so I gave you the "bad faith" charge.


Got it. I apologize if my tone was too harsh. I do stand by all of the bold-font headings, but yes, primarily the post is directed at Walker.

My current stance on sleep is that we don't know much about it and we don't know much about how much exactly we need to sleep. I'm collecting some notes on sleep (very preliminary so far) here: https://guzey.com/sleep/


Oh its alright, I didn't feel attacked.

If you're curious, here's another GWAS open-access paper on sleep: https://doi.org/10.1038/s41467-019-08917-4. Its self reporting, but its tricky to do EEG or some other technique on half a million people. Also this associated portal is cool: http://sleepdisordergenetics.org/home/portalHome


Thank you for all your work, which must largely be a thankless task.


Thank you.


'Point 2: Depression is an incredibly complicated topic. It is a psychological construct, the net result of thousands and thousands of genes, filtered through a modern technological world, and then filtered through inventories, interviews, and assessments. For this reason, I am not at all surprised that Guzey was able to find studies that suggest that sleep deprivation might have some benefit for some people. I would HIGHLY recommend this new, open-access Nature.....'

All of this is completely irrelevant to Guzey's point, which is that sleep deprivation is a known treatment for depression. You are going off on a tangent about the genetics of depression, while failing to engage with the topic under discussion.


This is also a very strange claim to make. I am pretty sure that you should be surprised - the anti-depressant/sleep thing is well known in part because it's so unique: I've never heard of even a single study finding sleep deprivation to be a striking temporary treatment for schizophrenia, autism, bipolar disorder, psychopathy, etc, all of which are influenced by 'thousands and thousands of genes, filtered through [etc]', the way it is for depression to the point of being meta-analyzable. (And if they exist, I don't recall them being mentioned anywhere in the sleep/depression papers.) It's a really strange finding!


If I completely buy Guzeys citation, sleep deprivation works for half of people temporarily. So, kinda useful but not sustainable. It also ignores memory and other effects sleep has on mood.

The point I am making is we can keep throwing new treatments at depressed people, or we can finally recognize the immense heterogeneity of depression, in both its genetic & environmental causes, as well as its many phenotypes (there are lots of ways to hit the inventory threshold). This fact is why this argument over sleeping or not is even happening. It’s not a tangent, in my opinion. This is the central problem.


>If I completely buy Guzeys citation, sleep deprivation works for half of people temporarily. So, kinda useful but not sustainable. It also ignores memory and other effects sleep has on mood.

You'll be interested in reading this study: "How to preserve the antidepressive effect of sleep deprivation: A comparison of sleep phase advance and sleep phase delay" https://link.springer.com/article/10.1007/s004060050092 and googling the keywords from that study

Also, you again ignored the point I was responding to in that section.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: