The most common thing women say about perimenopause, once they know what to call it, is that they did not know it had started. They knew about menopause, in the way everyone knows about it, as a thing that happens around fifty and involves hot flashes. What they did not know was that the years leading up to it are their own distinct period, that those years can begin in the early forties, that the sleep and the mood and the sudden inability to remember a word are part of the same process, and that nobody was going to mention any of this until they went looking.

The Definitions Are Doing Real Damage

Part of the confusion is built into the vocabulary. Menopause is not a phase. It is a single day, identified only in retrospect: the point twelve months after a final menstrual period. In the United States the median age for that day is fifty-one.

Everything people actually mean when they say menopause happens before it. Perimenopause is the transition, the stretch of years in which hormone levels stop being cyclical and start being erratic, and it commonly begins in the mid-forties, sometimes in the early forties. Its most reliable early sign is not a hot flash. It is a change in the cycle itself, in length or in flow, in a woman who has had the same cycle for twenty-five years.

Getting this wrong has a practical cost. A woman of forty-four with disrupted sleep, a shorter fuse and a cycle that has begun to wander is, on the standard timeline she has absorbed, too young for any of it, so the sleep gets treated as stress, the mood gets treated on its own, and the connecting thread goes unmentioned for years.

How Long It Actually Lasts

Here the data is unusually good, because of one long study. The Study of Women's Health Across the Nation has followed a large, deliberately diverse group of American women through midlife since the 1990s, and it has repeatedly contradicted the received timeline.

In 2015 Nancy Avis and colleagues published the duration finding in JAMA Internal Medicine. Among women with frequent hot flashes and night sweats, the median total duration was 7.4 years. They persisted for a median of 4.5 years after the final period.

Two details inside that number matter more than the headline. Duration varied substantially by group, and Black women in the study reported the longest median duration at 10.1 years. And the strongest predictor of a long course was starting early: women whose symptoms began well before their final period had them longest, while women whose symptoms began after it had the shortest run.

Set that against what most women are told, which is some version of a year or two. A median of more than seven years is not a footnote. It is a decade-shaped fact about the middle of a life, and being handed the wrong figure is what makes so many women conclude that whatever is happening to them must be something else.

The Memory Thing Is Real

The complaint that gets dismissed most often is the cognitive one. The word that will not come. Walking into the room and standing there. The sense, alarming in a way the hot flashes are not, that the machinery is not what it was.

It has been measured. Gail Greendale and colleagues, again within SWAN, tested the same women repeatedly over years rather than asking them how they felt. What they found was a genuine dip during perimenopause in processing speed and verbal memory, and then a return to prior levels afterward.

The shape of it is the interesting part. Take the same cognitive test repeatedly in midlife and you should get slowly better at it, simply through practice. During perimenopause, that expected improvement stopped appearing. The women were not collapsing. They were failing to gain, which is exactly what a person would experience from the inside as everything having become slightly harder.

Two things follow. The first is that this looked time-limited: it resolved in postmenopause, which is worth knowing for anyone currently convinced they are watching the beginning of something permanent. The second is that anxiety and low mood had their own separate unfavorable effect on performance, on top of the transition itself, which is a reason to treat poor sleep and low mood as worth addressing in their own right rather than waiting them out.

The Treatment Conversation Has Moved Twice

Anyone returning to this subject after a few years away will find the ground has shifted, in two different directions, and both are worth understanding before a conversation with a clinician.

The background is the Women's Health Initiative. Its early-2000s results, and the way they were communicated, produced a collapse in hormone therapy use and a generation of clinicians trained to be wary of it. Much of the subsequent reanalysis has focused on the fact that the trial population skewed older and further from the transition than the women who typically seek treatment, and that risk and benefit look different depending on age and on how recently menopause occurred.

In March 2024 The Lancet published a four-paper series arguing a case from the other side: that menopause has been over-medicalized, that a narrative of a hormone deficiency requiring hormone replacement is not well supported for all women, and that better information, better clinical listening and workplace flexibility deserve more weight than they get. The series drew immediate and substantive pushback from menopause clinicians, whose central objection was that a meaningful minority of women, on the order of a quarter or more, have symptoms severe enough to disrupt their working and family lives, and that framing effective treatment as over-medicalization risks making it harder to get.

Then, in November 2025, the FDA announced it would remove the boxed warnings from estrogen-containing menopausal hormone therapy products, on the position that the warnings reflected outdated evidence and had deterred appropriate use, with revised labeling to carry age-specific guidance. It is a significant change in how these products are presented. It is not, and was not claimed to be, new trial evidence, and the underlying individual calculation, involving age, time since menopause, personal and family history, and which symptoms are actually the problem, has not been simplified by it.

What all of this means for a particular woman is genuinely a clinical conversation and not a magazine's to have. What a magazine can usefully say is that the conversation is worth having with someone who knows the current evidence, that a clinician's posture may reflect the era in which they trained, and that being told to wait it out is not, in 2026, the state of the art.

What Else Is on the Table

The non-hormonal options are better than they were, and many women do not know they exist.

Two drugs in a new class have been approved specifically for hot flashes and night sweats, and neither is a hormone. They work on a signalling system in the brain's temperature-regulating region rather than on estrogen receptors, which makes them relevant for women who cannot take estrogen, including many breast cancer survivors. Fezolinetant was approved in May 2023, and elinzanetant followed in October 2025. Both require a prescription and a conversation about monitoring, as any new drug does.

Less expected, and better evidenced than most people assume, is cognitive behavioral therapy. In its 2023 position statement on non-hormone therapies, The Menopause Society gave CBT its highest evidence rating for vasomotor symptoms. The mechanism is not what the skeptical reader expects. It does not mainly reduce how often hot flashes happen. It reduces how much they interfere, and it improves sleep and mood alongside them, which for a great many women is the actual complaint. A symptom that no longer hijacks a meeting is a substantially different symptom.

The ordinary measures are worth taking seriously too, not as a substitute for treatment but because the transition interacts with everything else. Sleep that is already fragile is more easily broken by a night sweat. Strength training in these years does something for bone density that becomes much harder to do later. Alcohol reliably worsens both hot flashes and sleep, which is inconvenient news and consistently reported.

What to Take From It

The single most useful thing is probably the timeline. Perimenopause starts earlier than the culture says, lasts longer than the culture says, and produces a set of symptoms, including cognitive ones, that are documented rather than imagined. A woman who knows that is not going to spend three years wondering what is wrong with her.

The rest is a conversation, held with someone competent, about a decision that belongs to her. There are more options in it than there were two years ago, on both sides of the hormone question, and the honest summary of the evidence is not that one path is correct. It is that the transition is real, that the suffering it causes is measurable, and that nobody in 2026 should be quietly enduring it on the assumption that enduring it is what is expected.