posted my wife’s hormone-free perimenopause numbers a while back. she’s been on fezolinetant for a few months now and the thing that keeps jumping out is how few people seem to be tracking WHEN the flashes hit, not just how many. total count per day is kind of a garbage metric on its own. she was averaging 40-something pre-fezolinetant. week 6 it was down to 28. looks like 30% improvement. but the timing shift was more meaningful than the count drop: the late-night clusters (1-4am window, deepest NREM portion of her sleep) dropped faster than the daytime ones. that’s the number that actually matters if you care about sleep architecture. the bidirectionality here is real and I don’t see it discussed enough. hot flash interrupts deep sleep, poor sleep worsens thermoregulatory dysregulation, next night is worse. it’s a loop, not a linear cause. fezolinetant seems to hit the nighttime cluster disproportionately early, which breaks the loop before the daytime count fully resolves. fwiw I track this the same way I track my own protocols: event scale every few days, time-stamped, sleep stage logged separately (she uses a garmin). the journal isn’t for vibes, it’s for catching whether the mechanism is actually load-bearing. if you’re nine weeks in and count is down but you’re still feeling terrible, check whether your nighttime clusters shifted first. if the nocturnal timing improved and daytime count is lagging, that’s probably the loop still unwinding. if both are stalled, that’s a different conversation. genuine question for anyone 6+ weeks in: did the timing shift before the count dropped, or did they move together? trying to understand whether the nocturnal-first pattern holds across different baseline profiles.