A Field Guide to Menopause

Before you begin

Is this for you?

The transition

What menopause actually is

Menopause is a single point in time, defined retrospectively: twelve consecutive months without a period. Everything before it is perimenopause, everything after is post-menopause. The average age is around 51, though the range is wide and normal. Surgical removal of the ovaries reaches the same endpoint abruptly rather than over years, which is a meaningfully different experience.

The distinction matters because the leverage differs. In perimenopause the ovary still holds follicles, and the erratic gonadotropin dance is still running - the remaining follicles pushed by high follicle-stimulating hormone (FSH), estradiol swinging up and down, and cycle after cycle passing without ovulation and therefore without progesterone. Once the follicle pool is spent, that dance has nothing left to drive it. The question changes from managing fluctuation to deciding what to do about sustained withdrawal.

The signal

Not a reproductive hormone

Estrogen is a systemic signal. It has receptors in tissue that has nothing to do with reproduction - the lining of your arteries, the cells that build and dismantle bone, the regions of cortex that hold memory and mood, and the tissue of the vulva, vagina, and bladder. Those receptors weren't waiting for a pregnancy. They were reading a signal that arrived, in some form, every single month for four decades.

Which makes the transition a whole-body event rather than a reproductive one, and explains why the symptom list is so much longer than the cultural shorthand. Losing the signal is measurable in every one of those systems, and each of them has its own mechanism, its own evidence, and its own timing. That is the territory the rest of this guide covers.

The experience

What it actually feels like

By the time the twelve months have passed and perimenopause is behind you, a great deal of what made the transition miserable is over, and the relief is tangible. The hormonal migraines, the cramps, the bloating, the breast tenderness, the unpredictable flooding cycles, and most of the erratic vasomotor symptoms all belonged to the volatility. When the volatility ends, they most often end with it.

Women describe a body that works but no longer feels like theirs: hair that thins and loses its texture, skin that has lost its fullness, nails that break, vaginal tissue that no longer lubricates, a libido and a general drive that used to arrive spontaneously and now don't, and energy that just sucks all the time rather than rising and falling. Nothing in that list is an emergency, but why does it need to be? It may all be 'normal' with aging, but why just succumb to it all? It's the felt difference between a vibrant system and the shell of one.

When a woman is guided through the transition with hormonal support titrated to her, started in the right window and adjusted as she goes, the peri-to-post shift can pass with remarkably little event. That isn't every woman's choice and it isn't right for every history. The point is that the shell-of-a-body experience is not the only available ending.

The cost of not replacing it compounds

It can be easy to ignore the underlying physiology in the early post-menopausal years because the worst of the noise has ceased, but what accumulates in this 'acute' complete withdrawal phase is slow and more consequential. Bone density falls fastest in the first years after the final period, and fall and fracture risk follow it - a hip fracture in later life is one of the sharpest turning points in all of geriatric medicine, and it feeds the third leading cause of death in the United States: accidents. Cognitive and mood changes tend to settle in rather than resolve. Cardiovascular risk climbs on the schedule the vascular section described. Life satisfaction erodes in a way that's hard to pin on any single symptom because it's constellational in nature.

None of that argues that everyone should be on hormones. It argues that the decision deserves to be made deliberately and early rather than by default and late, because the window in which it's most useful is the same window in which nothing feels urgent.

Measurement

What is worth testing

Bloodwork after menopause isn't for confirming the diagnosis - twelve months without a period does that. It's for characterizing the terrain any decision gets made on.

Bone density is the direct measurement, and it offers better insight the sooner you do it, ideally before loss is significant. Iron studies matter because heavy bleeding in the years before the final period is a common and under-recognized source of depletion that can persist well afterward. Estrone belongs alongside estradiol, since it becomes the dominant estrogen once the ovary stops producing and speaks to the conversion happening in fat tissue. A coronary artery calcium score turns the timing conversation from theory into a picture of your actual arteries - which is precisely what the WHI's cohort problem was about. For women with significant cardiovascular disease risk, a CT angiogram is the gold standard, since with the right analysis (Cleerly, for example) it also identifies soft plaque.

Foundations

The work that applies regardless

Hormone therapy isn't the only intervention in the wheelhouse of menopause support. These are worth doing whether or not a prescription is ever written.

Resistance training is the only intervention that simultaneously addresses bone density, muscle mass, insulin sensitivity, and fall risk. Two to three sessions a week of appropriately loaded work, progressing over time, is the requirement - not just incidental movement or the occasional group class.

The exercise women most often gravitate toward - steady-state cardio, spin, and other low-intensity, moderate-volume work - doesn't substitute for it, and the reason is mechanical before it is hormonal. Bone and muscle remodel in response to load, and non-weight-bearing work supplies almost none. Competitive cyclists are the clearest illustration: despite enormous training volume, they carry markedly lower bone density than runners matched for age and body weight, with several times the odds of spinal osteopenia. Piling on volume without enough protein and recovery adds a catabolic cortisol burden on top of that, and it lands harder after menopause because there's far less anabolic hormonal signal left to offset it, though how much less varies more between women than most expect. None of which makes cardio a mistake - it makes it insufficient for the two tissues most at risk.

On that note, protein intake rises in importance because the anabolic response to using it blunts with age, and again as androgens decline. Adequate protein spread across the day is what lets resistance training build anything.

Sleep deserves direct attention rather than being treated as a symptom that will resolve on its own. A cool sleeping environment, consistent timing, morning light exposure, and limiting alcohol all matter. For a full exploration, see the Sleep Architecture guide.

Alcohol earns a specific mention: it can instigate hot flashes (through a different mechanism than hormonal withdrawal), fragments sleep architecture, and carries its own breast-cancer association - which is worth holding in mind while reading the risk section below.

The evidence

The study that set the conversation back

You can't discuss hormone therapy without addressing the Women's Health Initiative (WHI), the 2002 trial that launched two decades of fear. The story usually told is "hormone therapy causes breast cancer." The story the data actually tells is more specific, and in one crucial respect points the opposite way.

The WHI didn't study "hormone therapy." It studied two specific preparations in two separate randomized arms, and the two arms pointed in opposite directions. The combined arm gave women with a uterus oral conjugated equine estrogen - a horse-derived mixture, not human estradiol - plus medroxyprogesterone acetate, a synthetic progestin rather than progesterone. That's the arm stopped early, and the source of the headline. The estrogen-alone arm gave women who had had a hysterectomy conjugated equine estrogen by itself, and in that arm invasive breast cancer went down, a statistically significant reduction of roughly 23% (hazard ratio around 0.77).

That second arm almost never makes it into the popular summary. It should, because it means the WHI's own data doesn't support a blanket "estrogen causes breast cancer." In the arm without the synthetic progestin, estrogen was associated with less breast cancer.

A closer read

Even the combined arm's signal is routinely mis-stated. The finding was a relative increase of about 26%, and quoting that figure without the absolute risk it's a percentage of is how a small effect becomes a headline. In absolute terms the combined arm produced about 8 additional invasive breast cancers per 10,000 women per year, roughly 41 versus 33 per 10,000 woman-years. That's about 0.08% per year, not the 1% it often gets rounded up to. For scale, that increase sits in the range of the risk associated with a glass of wine a night, or with being sedentary.

Why the cohort mattered

The WHI's participants weren't the women the therapy is now used for. Their average age was 63, more than a decade past the average age of menopause and well outside the window in which hormone therapy is typically started. The baseline population was heavier than average, roughly half had a smoking history (about 40% former, about 10% current), and many carried hypertension or other cardiovascular comorbidities.

Starting an oral estrogen in a 63-year-old who already has a decade of established arterial plaque is an entirely different proposition from starting it in a symptomatic 52-year-old. One of the WHI's own signals reflected exactly this: women who began oral estrogen more than ten years past menopause showed an early increase in coronary events, plausibly because estrogen's effect on an already-atherosclerotic vessel differs from its effect on a healthy one. That's a legitimate caveat, and it points at route and timing rather than at the hormone as such. Thus began the 'timing hypothesis' and the literature everyone quotes when they say you've only got ten years after menopause to start hormone therapy. Yet the same analysis isn't seen with transdermal estradiol. The plot thickens.

Mistakes were made

When the full dataset and its long-term follow-up were analyzed for mortality - the 2017 JAMA analysis, 18 years of follow-up - the result wasn't the disaster the early termination implied. There was no significant difference in all-cause, cardiovascular, or cancer mortality in either arm. A therapy demonized as lethal turned out not to move the mortality needle in either direction for the population studied, even using horse-derived estrogen and a synthetic progestin in women who started late.

The correction came too late for the women it was about. Before 2002, roughly 40% of American postmenopausal women were taking hormone therapy. Within months of the press conference that number had fallen by nearly half, and it kept falling. Vasomotor symptoms went untreated. Genitourinary symptoms went unmentioned. The bone conversation was often never had at all. Women were moved onto antidepressants and sleep medication for a problem that was endocrine at its root.

The steepest observable cost landed in the arm where the trial had actually shown benefit. Women who'd had a hysterectomy were candidates for estrogen alone - the arm with less breast cancer and, in the 50 to 59 group, lower mortality - and they avoided it along with everyone else. An analysis in the American Journal of Public Health estimated that between 18,601 and 91,610 hysterectomized women aged 50 to 59 died prematurely in the decade after 2002 because of that avoidance. The WHI's own investigators disputed the modeling (accountability abounds), and the width of that range tells you how much uncertainty it carries. No matter how you look at it, the bottom of the range is still tens of thousands of women. Contrast that with the harm everyone feared, taken entirely at face value: about 8 additional breast cancer diagnoses per 10,000 women per year in the combined arm, which across 18 years of follow-up produced no detectable increase in all-cause mortality.

The hormonal intervention those women gave up lowered breast cancer incidence by roughly a quarter, about 7 fewer cases per 10,000 women per year, and actually did significantly lower breast cancer deaths. They forfeited a therapy that would have reduced the very outcome they were afraid of, because the headline never distinguished their arm from the other one.

The collapse in clinical training compounded the damage. A generation of clinicians finished training in an era when menopause management had loudly left practice, and most obstetrics and gynecology programs still have no dedicated menopause curriculum. That's why a woman today can raise this with a competent, well-meaning physician and get an answer built on a 2002 headline rather than the existing facts, and why I've written more than 100 letters to co-managing primary care physicians, oncologists, and OB/GYNs, explaining the rationale for a mutual client, justifying the decision, and advocating for her autonomy to choose hormone replacement.

Precision

Hormone therapy is not one thing

This is a critical point to understand, and one where public discourse can be discerned as signal or noise: hormone therapy isn't one intervention with one risk profile. The molecule and the route decide which evidence applies. If that degree of nuance isn't in the discussion, you can be pretty confident you're listening to noise.

Progesterone is not a progestin

These two words get used interchangeably, and they aren't the same thing. "Progesterone" - P4, the bioidentical molecule your own ovary made - is available as oral micronized progesterone. "Progestins" are synthetic analogues designed to fit the same receptor, and they're a chemically heterogeneous class whose members differ from progesterone and from each other. Medroxyprogesterone acetate, norethindrone, levonorgestrel, drospirenone, and dydrogesterone are not one drug. Drospirenone is anti-androgenic; others are androgenic. Their receptor binding, metabolic effects, and outcome data differ agent by agent.

This is where understanding a bit of pharmacology is helpful. Progesterone is metabolized to allopregnanolone, a neuroactive compound that modulates the GABA-A receptor - the same mechanism behind a modern approved treatment for postpartum depression (Zulresso), and the same receptor the benzodiazepine class works on. Synthetic progestins aren't all metabolized this way. Strictly speaking, none of them produce allopregnanolone at all, since that molecule is what progesterone itself becomes. A few form their own 5-alpha-reduced relatives with some activity at the same receptor, while medroxyprogesterone acetate - the one used in the WHI - appears to suppress the body's own allopregnanolone. That's why it's a factual error to describe "the progesterone arm" of a trial that actually used a synthetic progestin, and why a woman deserves to know what's actually being offered.

Most of the synthetic progestins in wide use are built on a testosterone backbone rather than a progesterone one - norethindrone, levonorgestrel, desogestrel and their relatives are 19-nortestosterone derivatives. That structure is what gives them their grip on the receptor and their durability by mouth, and it's why they carry residual androgenic behavior that progesterone doesn't. Excellent properties for a contraceptive. A different question entirely when the goal is replacement.

Oral is not transdermal, but know which risk the route governs

Oral estrogen passes first through the liver, which transdermal estradiol largely bypasses. That first-pass metabolism is why the clotting signal attaches to the oral route specifically. It's a useful distinction, though it's important not to over-extend it. The route difference governs the clotting risk. It doesn't resolve the breast-cancer question either way: the large datasets find no difference in breast-cancer risk between oral and transdermal delivery, and the signal that does appear tracks the progestagen instead. Reaching for first-pass metabolism to settle the breast question is a category error regardless of which way you want the answer to come out.

What the cohort data suggests about the progestagen

Large observational cohorts - the French E3N study followed roughly 80,000 women - found that estrogen combined with oral micronized progesterone carried a lower breast-cancer relative risk, around 1.0 at shorter durations, than estrogen combined with synthetic progestins, around 1.7. That's a meaningful signal in favor of the bioidentical progesterone.

Intellectual honesty requires both halves. That same cohort showed the risk with micronized progesterone rising with duration of use beyond about five years, so the favorable figure isn't permanent. A large 2019 meta-analysis, the Lancet CGHFBC pooling of over 100,000 breast-cancer cases, found micronized progesterone carried a risk closer to combined therapy overall, contradicting the reassuring read. The summary: the distinction is real, observational, duration-bounded, and contested - which is exactly why it belongs in a conversation with a provider who can weigh it against your history rather than in a slogan in either direction.

By hormone

What each one does

The mechanisms below are well-established receptor biology. Where a claim concerns a clinical outcome rather than a mechanism, it's flagged as such, because mechanism and proven outcome aren't the same thing, and the gap between them is where most of the overselling lives.

Estrogen

Where it comes from. Estradiol is built in the ovarian follicle, and it takes two cell types to do it: theca cells make androgens under luteinizing hormone, and granulosa cells aromatize those androgens into estradiol under follicle-stimulating hormone. Across a cycle it climbs through the follicular phase as the dominant follicle matures, peaks just before ovulation - that peak is what trips the luteinizing hormone surge - dips, then rises again more modestly from the corpus luteum. After menopause that machinery is gone, and the dominant estrogen becomes estrone, made by aromatase in fat, skin, and muscle out of adrenal androgens. Estrone is the weaker molecule, and how much you make of it depends substantially on how much adipose tissue you carry. That's why estrone belongs on the panel, and why body composition quietly shapes the post-menopausal hormonal picture.

Cardiovascular. The observational literature associates estrogen therapy started near the time of menopause with lower coronary risk - the Nurses' Health Study suggested roughly a 40% reduction - though observational cohorts carry healthy-user bias. The WHI's coronary signal ran the other way, and it sits in a far narrower place than the headline suggested: oral estrogen, started more than ten years past the final period, in arterial walls already carrying established plaque. It isn't a property of estrogen as such. It doesn't appear with early initiation, and the same analysis isn't seen with transdermal estradiol. Mechanism and outcome part company here entirely on timing and route.

Bone. Bone remodeling runs through estrogen receptor alpha. Estradiol acting on that receptor restrains osteoclast activity and supports osteoblast and osteocyte survival, so withdrawal tips a balanced system toward resorption (bone loss). This dependence isn't a female curiosity: men with aromatase deficiency, or with a mutated estrogen receptor, develop severe osteopenia despite normal or high testosterone. The skeleton reads estradiol, and much of what testosterone does for bone it does only after local conversion to estradiol inside the bone itself.

That receptor is also required for the skeleton's response to mechanical loading. Wolff's law - bone remodeling itself along the lines of force placed on it - runs through the same pathway, which means resistance training and hormonal status are not separate conversations. Part of the receptor's mechanical role appears to work without the hormone occupying it, so this isn't a flat "no estradiol, no adaptation" - that would contradict the evidence for resistance training in women not taking hormone replacement. The combination is superior because the loading response is measurably blunted without estradiol. Training after menopause remains the first thing on the list, and that doesn't mean it displaces the benefit of estradiol.

Brain. The brain is an estrogen-dependent organ in a way that's actually kind of baffling, and that speaks to overall metabolism as well. Estradiol regulates glucose transport into all cells, aerobic glycolysis, and mitochondrial proliferation and energy production - it is, functionally, part of how the brain and the rest of the body pay the energy bill. Given the brain's metabolic demands, that's where the impact is generally experienced first.

In that light, withdrawal can be appreciated as an induced bioenergetic deficit, not only a mood or cognitive one. Neuroimaging through the menopausal transition shows declining brain glucose metabolism, changes in gray matter and connectivity, and increased amyloid-beta deposition - a pattern tracking the endocrine transition rather than chronological age, and it overlaps uncomfortably with the early Alzheimer's phenotype. The brain appears to compensate by shifting toward ketones as a fuel in a metabolically flexible system, which is a workaround rather than a repair.

Despite the evolutionary and mechanistic support, the clinical picture remains unsettled. Part of the reason is the same one that governs the arteries. The other is tied up with age and the decline of the human system - the brain doesn't have the regenerative capacity most other tissues do, which is my argument for maintaining the premenopausal hormone profile rather than giving entropy the opening. Observational data associates early estrogen initiation with lower dementia risk, while the WHI's own memory sub-study found increased dementia risk when combined therapy started in women over 65. A proposed explanation is the healthy cell bias of estrogen action: estradiol supports neurons whose mitochondria are still working, and offers little or worse to neurons already metabolically compromised. That is the same logic as estrogen on a healthy endothelium versus an established plaque - one hormone, opposite outcomes, decided by the state of the tissue it arrives in.

Genitourinary. The vulva, vagina, urethra, bladder trigone, and pelvic floor all arise from the same embryological tissue, which is why they carry estrogen receptors together and why they change together. Estradiol maintains the epithelium, the collagen and elastin in the tissue beneath it, and the vascular plexus whose transudate is what lubrication actually is. Without it the epithelium thins and loses its rugal folds, the tissue beneath gives up collagen, and what remains is paler, more fragile, and less elastic - vulvovaginal atrophy, the older name this cluster carried before it widened into genitourinary syndrome of menopause. The felt version is dryness, burning, and dyspareunia: pain with sex. Another reason to loathe your partner that has nothing - well, maybe something - to do with them. In all seriousness, it's the symptom that most often drives the relational dissatisfaction, the shame, the avoidance, and a loss of libido that has nothing to do with actual desire. It fucking hurts, so no. Lubricants help, and that support matters - but the actual problem is addressable.

The cascade that follows withdrawal explains something women are rarely advised on and oh too often experience. Estradiol drives glycogen accumulation in vaginal epithelial cells. Glycogen feeds lactobacilli. Lactobacilli produce lactic acid, and that acid holds vaginal pH down near 4. Remove estradiol and the chain unwinds in order: less glycogen, fewer lactobacilli, pH climbing toward neutral, and a flora that shifts toward enteric organisms. Recurrent urinary tract infection in older women is largely this - not hygiene, not bad luck, just no environmental balance - which is why vaginal estrogen reduces recurrence when antibiotics only clear the current episode. In a population where urosepsis is a cause of hospitalization and death, it makes sense to have the awareness.

Unlike the hot flashes, this physiologic cluster doesn't burn itself out. Vasomotor symptoms typically abate over years; genitourinary syndrome of menopause is progressive and worsens untreated. It is simultaneously the most directly treatable consequence of the entire transition and one where women either aren't clued in or accept it as a part of aging. Vaginal estrogen is usually discussed as a local therapy with a simpler risk conversation than systemic hormone therapy - broadly true at standard doses, though "local" oversells it, since vaginal tissue absorbs efficiently and serum levels do move (see the clinical note above).

The androgen side of the same tissue. Estradiol gets the whole conversation here, and it isn't the only signal that tissue reads. Androgen receptors sit throughout the vaginal mucosa, submucosa, stroma, smooth muscle, and vascular endothelium, and they're densest in exactly the place that hurts - the germinal layer of the vestibular epithelium and the minor vestibular glands. The vestibule, meaning the ring of tissue at the entrance rather than the canal behind it, is androgen-dependent, and so are the Bartholin's, Skene's, and minor vestibular glands that lubricate it. Testosterone maintains the smooth muscle thickness and contractility, the compactness of the collagen, and the neurovascular machinery behind arousal and lubrication, and it modulates nociception and mucin secretion on top of that.

That anatomy explains a pattern I saw repeatedly. A woman started vaginal estrogen, the canal improved, the dryness improved - and it still hurt - it took many women complaining before one highlighted "at the entrance only". Estrogen restored the epithelium and the flora. It did not restore the androgen-dependent vestibule, and the vestibule is where dyspareunia is usually generated. Reading that as a failed trial of vaginal estrogen gets the situation backwards: it's a partial answer to a two-hormone problem.

Vaginal dehydroepiandrosterone (DHEA) is the elegant answer to that, and it has been approved for it since 2016. Prasterone, sold as Intrarosa, is a 6.5 mg insert, and the mechanism is intracrine. DHEA arrives inert, and the vaginal cells themselves carry the enzymes - aromatase, the hydroxysteroid dehydrogenases, 5-alpha-reductase - to convert it locally into both estradiol and testosterone, inside the cell, where it acts and is then inactivated. Serum estrogens and androgens stay within postmenopausal range on daily use. One insert supplies both arms of what the tissue is asking for without a systemic decision attached to it, which makes it the most elegant tool in the genitourinary kit.

Vaginal testosterone is the other route, compounded rather than approved, and it earns its place in a specific situation: women on aromatase inhibitors, where the estrogen door is closed and the usual answer is a shrug and a lubricant. A four-week course improved the signs and symptoms of aromatase-inhibitor-related atrophy without moving serum estradiol or testosterone. It's a small literature rather than a settled one, but for a woman whose oncologist has shut the estrogen conversation down, it is the difference between an option and, no option.

The practical version. If vaginal estrogen has been running and penetration still hurts, the missing signal is more likely androgenic than more estrogenic, so ask about vaginal DHEA by name. A provider who hasn't heard of it isn't a bad provider - it's a young approval in a field with no dedicated curriculum - their curiosity to find out is what you want.

Progesterone

Where it comes from. Progesterone is made by the corpus luteum - what the follicle becomes after it releases an egg. That is the dependency that matters: no ovulation, no corpus luteum, no progesterone. Across a cycle it sits near zero through the follicular phase, climbs sharply after ovulation, peaks mid-luteal, and falls away when no pregnancy arrives, and that withdrawal is what triggers a period. The adrenal glands contribute a trace; in pregnancy the placenta takes the job over entirely. This is why progesterone is the first thing to go in perimenopause, as cycles begin passing without releasing an egg, and why after menopause there is essentially none left.

Sleep and anxiety. This is where progesterone earns its place for most women who take it. Oral micronized progesterone is dosed at bedtime for a reason: the allopregnanolone it generates is a positive modulator of the GABA-A receptor, and it is remarkably sedating. Women commonly report deeper sleep and a quieter nervous system within the first few nights. Given how much of the post-menopausal picture runs downstream of sleep that stopped restoring, that is a godsend for most women. I say most because a very small percentage can't tolerate the oral micronized version at all - in their own words, near uniformly, they went batshit. A compounded sublingual helped some. Others had to settle for low-dose topical, which for a woman with a uterus means the endometrial question has to be solved another way, since topical progesterone doesn't reliably cover it (I prescribed low-dose drospirenone).

Where the route logic inverts. With estrogen, first-pass metabolism through the liver is the liability - it drives the clotting signal. With progesterone it is the feature. Oral micronized progesterone is extensively metabolized on that first pass, and allopregnanolone is largely what that pass produces. Transdermal progesterone creams generate far less of it, and, more importantly, they do not reliably achieve serum levels adequate to protect the endometrium. A woman with a uterus taking estrogen alongside a compounded progesterone cream may believe she is covered when she is not. Same molecule, opposite conclusion about route.

A note on neuroprotection. Progesterone has a mechanistic case for neuroprotection, and it is often stated as established fact. It isn't. Two large phase-3 trials settled the acute question in the same 2014 journal issue. ProTECT III gave intravenous progesterone within four hours of moderate-to-severe traumatic brain injury and was stopped early for futility, with favorable outcomes at six months running 51% on progesterone against 55% on placebo. SYNAPSE asked the same question internationally in severe injury and found no difference in outcome or mortality. Both followed roughly two hundred animal studies and encouraging phase-2 data, which makes the pair one of the cleanest examples in modern medicine of a mechanism failing to translate.

One detail is easy to miss and cuts against a clean reading. Both trials gave progesterone intravenously, which they had to - these were unconscious, intubated patients who couldn't swallow, in a window measured in hours, needing a defined blood level held for days. An intravenous drip also bypasses the liver's first pass, and the first pass is exactly where allopregnanolone comes from. If the neurosteroid is doing the work, those trials delivered plenty of the parent hormone and comparatively little of the metabolite that acts on the receptor in question. The acute-rescue result stands. Whether it was ever a fair test of the mechanism is a separate matter.

The field went on to test the metabolite directly, and there the answer is different. Brexanolone - allopregnanolone itself, given intravenously - became the first drug ever approved specifically for postpartum depression in 2019. Zuranolone, an oral analogue, followed in 2023 for the same indication, though regulators declined to approve it for ordinary major depression after a phase-3 trial missed its endpoint. Ganaxolone, another analogue, is approved for a rare childhood epilepsy. Three approvals say the neurosteroid pathway is legitimate, drug-worthy, and does something clinically measurable when the active molecule actually reaches the receptor.

As such, what has been approved is mood and seizure control - receptor modulation, the thing GABA-A does. Tissue rescue is a different claim, and it remains unproven; work testing allopregnanolone for regeneration in early Alzheimer's is underway and has not reported. We shall see.

Endometrial protection. In a woman with a uterus taking estrogen, progesterone and the progestins protect the uterine lining from estrogen-driven overgrowth. This isn't optional; it's why the combined regimen exists, and it's the one place where the choice of preparation is a safety question rather than a preference. Unopposed estrogen in a woman with a uterus raises endometrial cancer risk substantially, which is the specific harm the combined regimen was built to prevent.

After hysterectomy. Losing the uterus removes the red flag indication for progesterone. It doesn't remove progesterone from the physiology, and there are two arguments I'd make for keeping it in the conversation. The first is the allopregnanolone one above: the sedating, anxiolytic neurosteroid the brain had monthly access to for decades doesn't stop mattering because the endometrium is gone, and sleep and anxiety are frequently the symptoms a woman most wants addressed. The second is that progesterone appears to oppose estrogen at breast tissue, which some clinicians read as a reason to keep it on board rather than a reason to drop it. Both are mechanistic arguments rather than outcome-proven ones, and the conventional position is that a woman without a uterus needs no progesterone at all. Clinicians land in different places on this, which makes it a specific thing to put to a prescriber rather than assume in either direction.

Testosterone

Where it comes from. In women testosterone arrives from two places in roughly equal measure - the ovary and the adrenal cortex - plus peripheral conversion of DHEA and androstenedione in fat, skin, and liver. It rises modestly around ovulation, which isn't a coincidence. The important difference from the other two: testosterone doesn't fall off a cliff at menopause. The ovarian stroma keeps making androgens for years after the follicles are spent, and the decline that does happen tracks age and falling adrenal DHEA more than it tracks the final period. That's part of why some women arrive in post-menopause with a better anabolic profile than others, and part of why testosterone gets missed - the fall is gradual enough to pass for ordinary aging.

Not a male hormone. Testosterone is one of the highest-concentration hormones in a woman's body - present at several times the concentration of estradiol through most of the reproductive years. It isn't a male hormone that women happen to have; it is a female hormone, present and needed in both sexes, because "testosterone for women" still gets heard as something borrowed rather than something replaced.

What it does. It is anabolic, which is the through-line: it supports the muscle mass and bone density that the transition erodes, and it does part of the bone work by being aromatized to estradiol inside bone itself, as the bone entry above describes. It improves endothelial function. It contributes to drive, motivation, and the sense of being up for things, which women often describe before they mention libido at all.

Libido, and why it isn't the whole story. Testosterone appears necessary but not sufficient for sexual function. That distinction earns its keep clinically, because low desire in a post-menopausal woman is frequently not an androgen problem at all - it is the pain problem described earlier. A body that has learned that sex hurts will stop wanting it, and no amount of testosterone rewrites that lesson while the tissue is still unaddressed. Not to mention, address libido with testosterone before the vaginal tissue and you end up with an extra angry client who actually wants sex and can't tolerate it - always be up front with your clinician about the environment. Treating the genitourinary syndrome first, and then asking what remains, is the order that answers the question.

The interaction nobody mentions. Oral estrogen raises sex hormone-binding globulin, and SHBG binds testosterone. Raise SHBG and the free fraction - the part that actually reaches receptors - falls. A woman started on oral estrogen can end up with less usable testosterone than she had before, while her total testosterone looks unchanged on paper. That is one of the better arguments for measuring free testosterone and SHBG rather than the total alone, and one of the quieter arguments about route.

Where the evidence actually sits. There is no testosterone product approved for women in the United States. Prescribing means an off-label male formulation at a fraction of the dose, or a compounded preparation, which is a real barrier and a real source of variability. The international consensus position is narrow: testosterone is supported for post-menopausal women with distressing low sexual desire, and the evidence is considered insufficient to recommend it for bone, mood, cognition, or general vitality. Mechanistically it plainly does more than libido. Whether it does enough more to justify treatment is a question the trials have not answered, and you should be told that rather than sold past it.

How to supplement

The core, and what to route elsewhere

Nothing here replaces the hormone conversation, and nothing here replaces the training and the protein. What a stack can do is cover the specific gaps this transition opens, and there are fewer of those than the supplement aisle suggests. The core below is short on purpose. Everything that isn't menopause-specific - sleep, mood, anxiety, muscle, glucose - already has its own plan, and pointing you there beats padding this one.

The rest of what a woman needs in this window is already covered by a plan built for it, and those are better maintained than a menopause-plus-everything stack would be. That said, these plans are not built with menopause in mind and include duplicate-ish things you will not need - so make sure you read the accompanying guides to figure out exactly what you may benefit from.

• Insomnia that persists once vasomotor symptoms settle - the Comprehensive Sleep plan, alongside the Sleep Architecture guide
• Low mood - the Negative Mood plan, alongside the Depression guide
• Anxiety and rumination - the Anxiousness plan, alongside the Anxiousness guide
• Muscle and bone, alongside the training - the Female Anabolic Optimization plan
• Insulin resistance and the metabolic shift - the Glucose Metabolism plan, alongside the Metabolic Derailment guide
• Cardiovascular risk, which is where the timing conversation lands - the Atherosclerotic Risk plan, alongside the Atherosclerotic Risk guide

Three months is the minimum window for any of it, although many symptoms have the potential to resolve quite quickly. Retest what the earlier section listed rather than judging by symptoms alone - the cardiometabolic panel, bone density on the schedule you and your provider set, and hormone levels if you're using anything that moves them. If nothing has shifted by then, that's information rather than failure, and it usually points at the parts no supplement addresses: the training, the protein, the sleep, and the prescriptive hormone conversation itself.