The MIT study shows CSF waves—normally a sleep-only process that flushes metabolic waste—intruding into wakefulness when you're sleep-deprived. Your brain is apparently so desperate for the cleanup that it forces the process to happen anyway. Cost: attention lapses.
From what I've read, delirium tremens during alcohol withdrawal seems to follow a similar pattern, except it's REM sleep intruding into waking consciousness instead of CSF flushing.
[Polysomnographic studies from the 1960s-80s](https://pubmed.ncbi.nlm.nih.gov/7318677/) documented this. Patients in alcohol withdrawal exhibit what researchers call ["Stage 1-REM"](https://www.sciencedirect.com/topics/neuroscience/delirium-t...)—a hybrid state where wakefulness and REM sleep characteristics get mixed together. Right before full-blown DTs, [some patients hit 100% Stage 1-REM](https://link.springer.com/chapter/10.1007/978-1-4757-0632-1_...). The hallucinations appear to be [literally enacted dreams](https://www.sciencedirect.com/science/article/abs/pii/S01651...) occurring while technically awake. The sleep-wake boundary just completely breaks down.
What strikes me is the system-level similarity here. Sleep normally maintains clean states: you're either awake (alert, reality-testing intact, no CSF flushing) or asleep (offline, dreams permitted, maintenance running). But when the system gets stressed enough—whether through sleep deprivation or the neurochemical chaos of alcohol withdrawal—it seems to start making desperate tradeoffs.
The brain apparently needs certain processes to run. Period. Total no-brainer! CSF flushing can't wait indefinitely. Neither can REM sleep, which serves its own critical functions. So when normal sleep architecture fails, the system appears to force these processes anyway, even though the conditions are completely wrong for them.
Maybe that's why the costs are so specific. CSF intrusion during wakefulness costs you attention. REM intrusion costs you reality testing, because REM is the state where your brain accepts impossible narratives without question. Same compensatory mechanism, different critical process forced into the wrong state.
What I find interesting is how the brain knows what lever it needs to pull and how it pulls it. Sleep deprivation forces waste removal. REM deprivation forces wakeful dream states; which might be a side effect not the actual goal. The brain seems to know what maintenance is overdue and attempts the repair, consequences be damned.