Since it's paywalled, here is the summary from the (recommended) BrainPost newsletter:
<a href="https://www.brainpost.co/weekly-brainpost/2024/1/23/the-purpose-of-sleep-is-to-restore-our-brain-to-an-optimized-state-called-criticality" rel="nofollow">https://www.brainpost.co/weekly-brainpost/2024/1/23/the-purp...</a><p>This bit was particularly interesting:
"First, the authors found that more time in a waking state correlated with higher deviation from criticality (i.e., higher DCC scores) and more time in sleep correlated with lower DCC scores, consistent with their theory. Interestingly, the effect was greater when the rats spent more time moving during wake and the effect was absent when the rats were awake but in the dark. This suggested that not all wake experiences are the same and that more stimulation during a waking state can result in a greater deviation from criticality."
I have to admit I'm a little confused by this. I understand network criticality in terms of autocorrelation and crosscorrelation, but it seems like they define deviation from criticality in terms of duration of neural ensemble rest periods (they link to another paper for this DCC definition). So isn't this in a sense saying (very very roughly) something like "mice sleep more when their neurons are resting more than expected based on a power law distribution"? If so seems kind of tautological.<p>It's an interesting idea and kind of at the boundary of my understanding so I'm probably missing something.
What does this mean in the context of depression? Sleep deprivation is a method to treat depression for a short time until medication kicks in. In addition in the former melancholic depression there is a recognized pattern of mood improvement over the course of the day.
"Criticality is a concept borrowed from physics that describes when a system of individual parts will be most effective at responding to an input."