Estrogen, Progesterone, and Cortisol: How Your Hormones Actually Change in Perimenopause

Perimenopause is usually described as “declining estrogen.” That’s incomplete in a way that matters. Estrogen doesn’t decline in a straight line — it fluctuates, often wildly, for years before it settles low. Progesterone, meanwhile, is typically the first hormone to fall, and it falls for a specific mechanical reason that has nothing to do with age itself. Cortisol, the body’s main stress hormone, moves in response to both — and current research doesn’t fully agree on exactly how. Understanding what each of the three is actually doing is the difference between a vague sense that “my hormones are off” and knowing which symptom traces back to which shift.

You’ve probably been told the headline version: estrogen goes down, symptoms happen. It’s not wrong, exactly. It’s just missing the part that explains why perimenopause feels so unpredictable rather than steadily worse — why some weeks feel almost normal and others don’t, sometimes within the same cycle.

This article walks through what the research actually shows about estrogen, progesterone, and cortisol in perimenopause: what changes, in what order, through what mechanism, and what current evidence does and doesn’t establish about how the three interact.

---

Estrogen Doesn’t Decline — It Destabilizes

The word “decline” implies a slope. What the data shows is closer to a system losing its rhythm before it loses its output.

A 2021 longitudinal study — the Swiss Perimenopause Study, which tracked 127 women aged 40–56 over 13 months using more than 1,700 saliva samples — found no significant, progressive decline in estradiol or progesterone across the perimenopausal stages measured. Instead, the authors describe “relatively stable periods of fluctuating hormone levels,” and note that perimenopausal stage did not reliably predict where a given woman’s estradiol would land on a given day (Grub et al., Frontiers in Global Women’s Health, 2021). In plain terms: on any random Tuesday in perimenopause, estradiol could be at a premenopausal level, or well below it, and stage alone won’t tell you which.

That doesn’t mean estrogen never actually falls — it does, and reliably, but the drop is concentrated later and closer to the finish line than most people expect. The Study of Women’s Health Across the Nation (SWAN), which followed over 1,300 women with frequent hormone measurements timed against their actual final menstrual period, found that estradiol’s real decline accelerates starting about two years before the final period and continues for roughly two years after it, before leveling off (Randolph et al., Journal of Clinical Endocrinology & Metabolism, 2011). Put the two studies together and the picture is: years of instability first, then a genuine, steeper decline clustered around the transition’s final stretch.

This is why symptoms so often feel disproportionate to what a single blood test shows. A hormone panel is a snapshot; perimenopause is defined by the variance between snapshots, not any one of them.

Progesterone Is Usually the First to Go — and the Reason Is Mechanical

If estrogen’s story is instability, progesterone’s is much closer to a straightforward decline — and it typically starts earlier than estrogen’s does, for a reason that’s easy to overlook: progesterone is made almost entirely by the corpus luteum, the structure that forms after an egg is released. No ovulation, no real progesterone rise that cycle.

As ovarian reserve declines through your late 30s and 40s, ovulation becomes progressively less reliable, well before periods become visibly irregular. A study that tracked 108 women across five years using daily hormone testing found that the probability of an anovulatory cycle rose steadily through the transition, and that reproductive stage predicted anovulation better than chronological age did. By late perimenopause, more than 60% of cycles in that study were anovulatory — though a meaningful share of longer cycles were still ovulatory, which is the reason contraception still matters later into perimenopause than most people assume (O’Connor et al., Menopause, 2009). Progesterone levels tracked this pattern directly: lower and less consistent as anovulatory cycles became more frequent.

This sequencing — progesterone wobbling years before estrogen’s real decline sets in — is one reason early perimenopause can feel like “off” sleep and mood with a period that still looks normal on the calendar. The cycle length hasn’t changed yet. What’s happening underneath it already has. If your cycles have already started running short, long, or unpredictable, that’s the more visible sign of the same mechanism — covered in more detail in our piece on why perimenopause cycles become irregular.

Progesterone’s other job, separate from ovulation, is sedation. It’s metabolized in the body into allopregnanolone, a compound that acts on the brain’s GABA-A receptors — the same receptor system targeted by anti-anxiety medications and some sleep aids. Less progesterone generally means less of that calming input, which is part of the mechanism behind the sleep disruption so many women report starting well before hot flashes ever show up. We go into that connection specifically in why progesterone decline disrupts your nights.

Where Cortisol Fits In — and Where the Evidence Gets Genuinely Mixed

Cortisol is the part of this trio that gets stated with the most confidence online and has the least settled evidence behind it. Worth being precise here, because the honest picture is more interesting than the tidy one.

What’s reasonably well established: the same allopregnanolone pathway that helps explain progesterone’s effect on sleep also appears to shape how the body regulates its stress response. A 2023 narrative review in Frontiers in Endocrinology — covering HPA-axis (hypothalamic-pituitary-adrenal axis, the body’s central stress-response system) function across the female reproductive lifecycle — describes how falling progesterone-derived neurosteroids alter GABA’s normal restraint on that axis, plausibly “sensitizing” some women to stress during the transition (Hantsoo et al., Frontiers in Endocrinology, 2023). The same review cites longitudinal data from the Seattle Midlife Women’s Health Study showing that basal morning cortisol measurably increases during late perimenopause, alongside rising FSH.

Where it gets genuinely unsettled: the review also names studies finding no consistent link between perimenopausal depression and altered morning cortisol, and no clear evidence that HPA-axis sensitivity tracks estradiol’s ups and downs the way a tidy narrative would predict. One small trial did find that estrogen supplementation blunted the cortisol, adrenaline and noradrenaline response to an acute stress test in perimenopausal women, compared with placebo — a real finding, but from one small study, not a settled mechanism you can apply to every symptom.

The fair summary: progesterone’s decline plausibly turns down the brain’s built-in stress buffer, and cortisol does appear to shift over the transition — but exactly how estrogen, progesterone, and cortisol interact minute to minute isn’t nailed down, and anyone stating it with total certainty is overselling what two decades of research on this specific question has actually settled.

Why “Fluctuation” Explains What “Decline” Doesn’t

Put the three hormones together and a pattern emerges that matches what women describe far better than a simple decline story does: perimenopause isn’t one dial turning down, it’s three interacting systems moving on different timelines, at different speeds, with different degrees of predictability.

Progesterone starts wobbling first, tied to ovulation rather than age on a calendar. Estrogen stays unstable for years before its real decline concentrates around the final year or two before periods stop. Cortisol’s relationship to both is real but only partly mapped. None of that is fully determined by which day you happen to test on — which is exactly why a single hormone panel so often comes back “normal” for a woman who feels anything but.

This is also the practical reason cycle-phase tracking matters more in perimenopause than a simple calendar does. In a regular cycle, hormone timing is predictable enough that “day 21” means roughly the same thing every month. In perimenopause, with cycles running anywhere from a couple of weeks to well over a month, the calendar date stops being a reliable stand-in for what’s actually happening hormonally — which is the entire reason PeriFlow tracks by phase (Rise, Crest, Root) rather than by day number.

What This Doesn’t Tell You

A few limits worth naming plainly, because overstating what this research shows would be exactly the kind of false certainty this space already has too much of.

None of these studies measured symptoms and hormones in the same women at the scale needed to prove individual cause and effect. The Swiss study measured hormones without a symptom-severity outcome attached; the SWAN data is population-level, describing average timing across thousands of women, not a prediction for any one person’s timeline. Your progesterone may have started dropping earlier or later than “typical,” and there’s no blood test that will tell you precisely where you are on this curve on any given day.

Saliva and urine hormone testing — used in several of these studies — is a research tool, not a diagnostic one. It’s useful for tracking patterns over months in a research setting; it is not the basis for a same-day answer about “your levels today,” and PeriFlow doesn’t build features around single hormone readings for that reason.

The cortisol research in particular should be held loosely. Where studies disagree this much, the honest position is “plausible mechanism, unsettled details” — not “cortisol causes your perimenopause symptoms,” a claim that oversells the evidence in exactly the direction wellness marketing tends to push.

What to Actually Do With This

The most useful thing this research offers isn’t a number to chase — it’s permission to stop expecting your hormones, and therefore your symptoms, to move on a straight schedule. If a week feels harder than the one before it despite “nothing changing,” that’s consistent with what the data actually shows, not a sign that something is wrong with you specifically.

The second is more actionable: because progesterone tends to move first and estrogen’s real decline concentrates later, symptoms that show up years apart can share the same root cause. Sleep trouble and a shorter fuse for stress showing up now, with hot flashes arriving two or three years later, isn’t two separate problems — it’s the same underlying sequence, just at different stages. Naming that link is often the first useful thing your notes give you.

The third is to track by pattern, not by date. Because none of these three hormones move on a fixed calendar, the only way to see your own pattern is to log symptoms against your actual cycle over a few months and look at where the rough stretches cluster — a question a single test can’t answer but a dated record can.

---

Estrogen fluctuating, progesterone falling first, cortisol somewhere in the mix but not fully mapped — none of that is a diagnosis, and none of it replaces a conversation with a clinician if what you’re experiencing is severe or persistent. What it does do is explain why perimenopause so rarely feels like a smooth, predictable slope, and why “wait for it to get consistently bad” is the wrong test for whether something real is happening.

That’s the gap PeriFlow tracks into. It logs your symptoms against your actual cycle phase — Rise, Crest, or Root — rather than a calendar date that stopped meaning much once cycles went irregular, so the pattern underneath three hormones moving on different timelines has somewhere to become visible.

Try PeriFlow free for 7 days See which phase you’re in today, and start building the record that makes sense of the weeks that don’t add up on their own.

---

Suggested internal links: What Is Perimenopause · Why Perimenopause Cycles Become Irregular · Perimenopause Sleep Problems · Perimenopause Exhaustion · Perimenopause Brain Fog