Hacker News Reader

Comment by sillysaurus2 | Hacker News Reader

A "toxin" is something which is toxic to a system. But our knowledge of the brain is so primitive that we can't reasonably claim to know which chemicals are toxic at tiny, long-term dosage levels, unless it leads to death. There is no evidence that it's possible to die directly from sleep deprivation. Therefore this seems a dubious headline.

View on HN

The word "toxin" has become a signal for pseudoscientific medicine, so I also furrowed my brow at that. But the original use of the word is much more clear in context. From the abstract which tokenadult kindly linked to:

Thus, the restorative function of sleep may be a consequence of the enhanced removal of potentially neurotoxic waste products that accumulate in the awake central nervous system.

Also, regarding the dangers of sleep deprivation, the first paragraph of the paper are relevant (http://www.sciencemag.org/content/342/6156/373.full):

Sleep deprivation reduces learning, impairs performance in cognitive tests, prolongs reaction time, and is a common cause of seizures (3, 4). In the most extreme case, continuous sleep deprivation kills rodents and flies within a period of days to weeks (5, 6). In humans, fatal familial or sporadic insomnia is a progressively worsening state of sleeplessness that leads to dementia and death within months or years (7).

Reference 7 is http://www.thelancet.com/journals/laneur/article/PIIS1474-44..., but I found the Wikipedia article more consumable for someone who is not up on medical research: http://en.wikipedia.org/wiki/Fatal_familial_insomnia

Direct link to reference 7, "Familial and sporadic fatal insomnia": http://a.pomf.se/7Uu7.pdf

From the paper, "Phenotypic variability is another issue. In some patients, insomnia, one of the cardinal symptoms, was not reported."

The other half of that sentence is obviously "... but Dementia and death still occurred."

Therefore, this research demonstrates that fatal familial insomnia results in Dementia and death, but with insomnia as a secondary effect. The insomnia is a symptom, not the cause.

I'm only trying to show how tricky it is to base an argument on scientific research. There are all sorts of corner cases and gotchas. In this case, the fact that insomnia is unrelated to the Dementia and death is buried on page 8 of a 10-page article, embedded in a gigantic paragraph. A rather important point for such little treatment by the article!

Rats die "within a matter of weeks" if they do not sleep: http://www.sciencedirect.com/science/article/pii/01664328950...

Curses, you've made me break my "No HN on my work computer" anti-procrastination method. At least it's been only 15 minutes though.

Direct link to the research: http://www.gwern.net/docs/algernon/1995-rechtschaffen.pdf

The paper is titled "Sleep deprivation in the rat by the disk-over-water method." As the name implies, rats are suspended on a disk over water. They can't fall asleep without starting to drown. Thus they are kept awake.

The paper begins with, "Because short-term sleep deprivation (SD) may stimulate only sleep-promoting mechanisms, chronic SD may be required to elicit function-revealing deficits. However, the enforcement of chronic SD requires repeated, intrusive stimulation which can blur the interpretation of effects. Do they result from sleep loss or from the strong stimulation used to enforce SD? To simplify communication, we speak of the 'effects' of SD, but strictly speaking, SD studies are correlational. We apply stimuli to enforce SD and report the relationship between the ensuing sleep loss and changes in performance or physiology. However, the changes and the sleep loss could be independent responses to the stimulation. The interpretation that the changes result from sleep loss hinges on minimizing the contribu- tion of the deprivation-enforcing stimulation, which can be especially difficult when strong stimulation is used to en- force chronic SD."

There are two possibilities. Either rats being repeatedly almost-drowned doesn't contribute to death after 45 days, or it does. The paper goes on to demonstrate the steps they took to minimize the chance that almost-drowning rats resulted in death:

"Hypothermia had been suspected as a proximal cause of death [3,31,34] because all SD rats showed an eventual decline in intraperitoneal temperature (Tip); a decline to more than 1 °C below baseline in otherwise untreated SD rats has been a reliable indicator of impending death within a day or two. However, TSD rats kept warm by exogen- ous heating died nevertheless"

"A second possible cause of death is breakdown of body tissues due to catabolism, secondary to the high metabolic rate in TSD rats [3,14,31]. [...] Evidence against catabolism as a mediator of preterminal effects includes a lack of preterminal serum albumin de- cline in PSD rats [22] and the deaths of all rats in two TSD groups protected against catabolic effects, hypothy- roid rats and high-calorie diet rats. Thus tissue breakdown secondary to catabolism was not a necessary cause of death."

However, this third point is a key conflating factor:

"A third major candidate for proximal cause of death is organ failure secondary to systemic infection, as suggested by bacteremia, which Everson has observed in five of six TSD rats obviously near death [11]. We subsequently confirmed bacteremia in two additional preterminal TSD rats [20]. Very recently, we treated six TSD rats with antibiotic cocktails; five progressed to an apparently ter- minal condition nevertheless and were then killed (after 10-16 days of TSD). Neither heart blood samples, livers, kidneys, nor mesenteric lymph nodes showed aerobic bac- terial or fungal infection. The sixth rat died after 19 days; blood could not be drawn, but the other tissues were har- vested shortly thereafter and were also free of aerobic bacteria and fungi. These results indicate that microbial invasion is not a necessary cause of death in TSD rats."

If you read carefully, they begin by saying "almost-drowning rats eventually results in organ failure." They end with "microbial invasion is not a necessary cause of death."

I think it's possible that almost-drowning rats causes organ failure for reasons other than microbial infection. The fact that the rats progressed to a terminal condition even after antibiotics seems to support this hypothesis. Their organs failed anyway. I wonder if dunking them in water for a month had something to do with it, and not the sleep deprivation?

The paper seems to tiptoe around this point.

Thanks for all the info!

It seems possible that torturing animals for days on end could be fatal to the victim. I hesitate to suggest to try adding doses of sleep to this diabolic regimen, for fear it would be attempted.

>It seems possible that torturing animals for days on end could be fatal to the victim.

yes, all the things we humans call science...

> A "toxin" is something which is toxic to a system. But our knowledge of the brain is so primitive that we can't reasonably claim to know which chemicals are toxic at tiny, long-term dosage levels, unless it leads to death.

If it leads to harm or death of individual neurons, we have lots of knowledge about such things. For example, we know that glutamate is toxic in large doses to neurons,

http://en.wikipedia.org/wiki/Excitotoxicity

If you don't sleep for long enough (about 7 days), you'll start to have visual and auditory hallucinations and also delusional thinking. You'll also experience severely reduced effectiveness of just about every system in your body. You might not die if you don't sleep for a year (say), but you'll probably get permanent brain damage.

The point of science is to never fall back on claims like "you'll probably get permanent brain damage" unless it's known what brain damage you may get, along with bounds on a confidence interval. Otherwise we're just philosophizing.

Science is slowly getting to those conclusions:

> Taken together, these changes in brain and body are further evidence that sleep deprivation is a chronic stressor and that the resulting allostatic load can contribute to cognitive problems, which can, in turn, further exacerbate pathways that lead to disease.

http://www.sciencedirect.com/science/article/pii/S0026049506...

Well, it's a testable hypothesis. We will probably never test it in humans or higher mammals due to ethical concerns.

may be we willn't as we already did. Google "sleep deprivation experiment".

Many people experience hallucinations if they stay awake for periods longer than 48 or even 36 hours.

I think i believe this. A friend of mine is a virgin to late nights and all nighters. We were studying one night trying to cram for a test and communicating every so often on WhatsApp. At around 12AM, he told me he thought he was hallucinating. I laughed as he was only up for like 14 hours. He fell asleep shortly thereafter.

I can relate. I've only had two "all nighters" in my life, and felt utterly miserable each time.

Towards the end of both, I would develop minor visual hallucinations (seeing little things in the corner of my eye, etc.). 14 hours is a bit silly, though. 24 hours is probably the minimum for most people.

In clinical settings people experienced hallucinations with less than 48 hours of sleep deprivation.

http://www.amsciepub.com/doi/abs/10.2466/pms.1989.68.3.787?j...

I have no idea where you get 7 days from, I'm not sure its possible to stay awake for 7 days.

On reflection, maybe I got confused with the number of days you can go without water (also 7, at least according to my memory). I stayed awake for 5 days once and there was only a bare minimum of hallucination at the end. I could have stayed awake longer, it wasn't really hard by that point. The people I've known who stayed up forever on crystal meth typically started hallucinating after a week. I'm surprised that the limit is so much shorter, but I'll believe it, thanks for the link.

There is no evidence that it's possible to die directly from sleep deprivation.

Perhaps not directly, as the actual cause of death in FFI is unknown, but:

https://en.wikipedia.org/wiki/Fatal_familial_insomnia

Actually, the insomnia isn't the cause of death. FFI is just a bad name for the disease. https://news.ycombinator.com/item?id=6568961

Reply on news.ycombinator.com