About this passageSummarised for this collection from work on sleep and emotional processing. The 2007 study deprived healthy people of a single night's sleep and measured amygdala response to unpleasant images; the review sets out the broader case that sleep does something to emotional memory rather than merely resting the body. The overnight-therapy framing is the researchers' own and remains argued over.
After one night without sleep, the amygdala — which registers threat — responds about sixty per cent more strongly to unpleasant images. At the same time the connection between it and the prefrontal region that normally moderates it weakens. The alarm gets louder and the hand on the volume comes off.
That is the deprivation finding. The more interesting proposal is about what sleep is doing when you get it: that during certain stages the brain reprocesses the emotional content of the day, keeping what happened while turning down the charge attached to it. The memory survives; the sting is reduced.
Which is why a difficult evening genuinely can look different in the morning, and why it reliably looks worse at two in the morning when you have not slept. That is not weakness of character. It is a measurable state.
How completely sleep does this, and for whom, is still contested.