> So the ancient mystery of why we need sleep might have just been answered.
There are layers to this, some of which are definitely not ancient mysteries. We sleep because the environment has a day-night cycle. If any task an organism must perform is better done during the day, then evolution has a very clear gradient towards only doing that thing during the day. That leads to doing other things at night, since it would be comparatively advantageous to do them at night, given whatever task is most benefited from being done during the day.
If there wasn't a day-night cycle it's unlikely that the brain would have evolved to crucially depend on approximately a night's worth of time of not using the body.
The question isn't the timing but why it happens at all. Even at night, being unaware of one's surroundings during sleep is a huge disadvantage that requires lots of effort and adaptation to work around. It needs to produce commensurate benefits, but we're not sure what they are.
Exactly! Going offline for hours every day in an adversarial world is positively nuts! The reason can't be idiosyncratic. No gentle gradient of comparative advantage can rationalize it. It must be something severe and nigh impossible to do any other way.
Furthermore, sleep is very specifically about the taking the brain offline: that's what deteriorates first in the absence of sleep and the tortured workarounds for animals that absolutely must avoid sleep (e.g. migratory birds) involve sleeping part of the brain at a time. Any explanation that isn't highly specific to the brain's responsibilities has the immediate hurdle of explaining this away, and for that reason I don't buy the mitochondrial explanation. Mitochondria are too universal and sleep is too specific to the brain. Energy is fungible, so I don't buy that nature wouldn't figure out the "trick" of having a subset of the mitochondrial population sleep at a time.
My money is on the "brain algorithm" requiring an online/offline phase, whether that's contrastive learning or memory consolidation or something else. There are lots of candidates for fundamental brain algorithms with the "feature" that they require an offline phase that cannot be incrementally worked around, and these fit the observations much better.
> Mitochondria are too universal and sleep is too specific to the brain
The brain has uniquely high specific power requirements per gram of dry weight. Not even the heart is this power-hungry. This surely places a lot of uniquely high metabolic stress on the neural cells.
And neural cells are long-living, so they can't be easily replaced throughout the lifetime. So their housekeeping has to be very thorough, carefully cleaning up all the waste products.
Mitochondrial dysfunction literally leads to Alzheimers, dementia, etc. The link is clear as day - don't sleep, lose your mind. Put a different way, life rusts (oxidizes) your brain, and sleep de-rusts it. And unfortunately I'm still someone who regularly pulls all nighters because of a combination of disorders, ADHD and sleep cycle issues. I'm killing myself rather prematurely. But then, all addictions tend to do that and tend to be things the addict has trouble controlling. :(
Also, the theory would better be expressed as "all mitochondria require rest, neuronal rest in the brain looks like sleep (but many cells in the body also get quite a bit of rest during this time)" - so many people here seem to be getting this backwards thinking sleep is the special thing - it's one way large scale mitochondrial rest (as well as lots of other important co-occuring processes) presents in the brain.
The really interesting question is... how do heart cells do it? Because they're a clear exception to this theory... Lactate?
Maybe the heart can just handle the stress longer before something visible breaks? It has a simpler function than the brain, which probably helps. If people tend to die of cancer first (or whatever), the heart only needs to withstand about 80 years of gunk buildup to not be the weakest link.
> Furthermore, sleep is very specifically about the taking the brain offline
Animals that lack a nervous system, like sponges appear to sleep. It may be that becoming unconscious is a symptom of sleep, not a cause of it. Not much point being awake if your body is effectively shut down.
I mean, it still can be idiosyncratic if the local maximum is steep enough. Identifying and signalling subgroups of mitochondria in a cell to put on pause might be prohibitive, for instance, and would still reduce the power available to that cell.
Or maybe going all the way on and mostly-off with your mitochondria, even specifically with your brain mitochondria, really is that much more efficient than having half of them offline (but still consuming energy for upkeep) at any time. The brain is a big ol energy hog, after all.
I will admit I'm mostly ignorant on these subjects, but just using rational/logic
> If any task an organism must perform is better done during the day, then evolution has a very clear gradient towards only doing that thing during the day.
But wouldn't remaining conscious and aware be the optimal state? So you don't get eaten by predators or attacked by other humans etc? It seems to me your sentence points to an ultra low energy but conscious state, not one in which you're very vulnerable...
But maybe the vulnerability is just too little, maybe cooperative tribal/family type arrangements covered this sufficiently to not be selected?
What you say is true and fairly obvious, but the interesting mystery is the mechanism of that dependency, not its evolutionary advantage.
Knowing the mechanism opens the door to medical interventions. Analogously, no one is confused as to why the human body stores fat and gets hungry, but knowing the mechanism allows weight-loss treatment like Ozempic.
It's interesting that sleep is controlled by mitochondria, but sleep is clearly involved in learning, and whatever algorithm for intelligence the brain does. Do those algorithms still work if you intervene at the level of the mitochondria? Or are the mitochondria just a good way of measuring elapsed time through energy expenditure? e.g. The algorithm needs a sleep phase to run roughly every x neural firings, or performance degrades and mitochondria were available as measuring devices when nature needed a way to guess how long the wake phase had been running.
Maybe you could intervene to prevent anyone from feeling tired, but would the learning algorithm still work? That part is still a mystery.
That's a good point. Maybe we found the mechanism to stay awake, but if that doesn't also translate to normalizing everything else that happens while sleeping, then who knows. Maybe people turn into wide awake zombies after a few days.
There are also other reasons for sleep, like cleaning up neurotransmitters and stocking them up in advance. I would guess it's a more immediate trigger?
If the brain fundamentally needs sleep then we'd sleep regardless, just not aligned to the day-night cycle. There's quite a bit of variation in sleep patterns and amounts between different animals. Chinstrap Penguins only sleep a few seconds at a time, but still manage to rack up ~11hr sleep in a 24hr period! Elephants only sleep for ~2hr/day, horses for 3hr/day.
There are layers to this, some of which are definitely not ancient mysteries. We sleep because the environment has a day-night cycle. If any task an organism must perform is better done during the day, then evolution has a very clear gradient towards only doing that thing during the day. That leads to doing other things at night, since it would be comparatively advantageous to do them at night, given whatever task is most benefited from being done during the day.
If there wasn't a day-night cycle it's unlikely that the brain would have evolved to crucially depend on approximately a night's worth of time of not using the body.