The flashes wake you. Then, months later, the flashes get better and you are still awake at 3am - which is the moment most women conclude nothing works.
Something does work. It is just not aimed at the hormones, because by that point the hormones are no longer the thing keeping you awake.
What the numbers actually show
Sleep gets measurably worse across the menopause transition, and it does not get worse uniformly.
National survey data from the CDC's health statistics center show short sleep - under seven hours - in 32.5% of premenopausal women, 56.0% of perimenopausal women, and 40.5% of postmenopausal women. Perimenopause is the peak, and then it partially recovers.
Trouble staying asleep does not follow that curve. It rises steadily: 23.7%, then 30.8%, then 35.9% - highest after menopause, when the flashes have often settled. That divergence is the whole story of this article in two rows of a table.
The larger longitudinal work agrees on direction: progression through the transition is associated with self-reported sleep disturbance, and more frequent vasomotor symptoms go with higher odds of every kind of sleep difficulty.
The two-problem structure
Here is the sequence almost every woman describes.
A hot flash wakes you at 3am. That is physiological and it is real. You are awake, hot, heart going, and it takes forty minutes to get back down.
This happens for months. And across those months you adapt in ways that make complete sense: you go to bed earlier to bank hours against the ones you will lose. You stay in bed in the morning trying to reclaim some. You start thinking about sleep during the day.
Now the flashes improve - on their own, or on treatment - and the 3am waking stays. It stays because it is no longer being caused by a flash. It is being maintained by nine and a half hours in bed for six hours of sleep, and by a body that has learned that 3am is when we are awake.
That is conditioned arousal, and it is the same machinery described in why you wake at 3am and cannot get back to sleep. Perimenopause is one of the most reliable ways to acquire it, because it supplies months of legitimate nightly wakings for the pattern to build on.
Why the behavioral route was tested here specifically
The obvious objection is that you cannot address insomnia while something is physically waking you up. That objection was tested.
The MsFLASH trial enrolled peri- and postmenopausal women with moderate insomnia and at least two hot flashes a day, and compared telephone-delivered CBT-I against menopause education. At eight weeks the CBT-I group's Insomnia Severity Index scores were about 5.2 points lower than control (95% CI -6.1 to -3.3), and about 4.0 points lower still at twenty-four weeks. Their sleep quality scores improved too.
Those women were still having hot flashes. The flashes were not treated first, and the insomnia improved anyway.
One thing to be careful about, because it gets overstated: what improved alongside sleep was how much the hot flashes interfered with daily life, not how often they occurred. A behavioral sleep program does not reduce vasomotor symptoms and nobody should tell you it does. What it appears to change is how much they cost you.
