Most women I see arrive having been told one of three things: “you’re too young for menopause,” “your labs are normal,” or “this is just stress, try yoga.” They are 43. They have been gaining weight despite eating less. They wake up at 3am for no reason. Their periods have started doing things they have never done before. And they have been told, by people they trusted, that nothing is wrong.

Something is wrong. Or rather, something is changing, in a way that is normal in the species but profoundly disruptive in the individual woman, and most clinical workups are not built to see it.

This article is about the framework I actually use, in my practice, to diagnose perimenopause. Not a lab list. Not a self-scoring symptom checklist. The thinking. The order of operations. The questions I ask and the patterns I look for, in the order I look for them.

I’m writing it because the absence of this framework in primary care is, in my opinion, the single biggest reason women in their forties spend years feeling worse and getting nowhere.

What perimenopause actually is, in five sentences

The ovary is an endocrine organ with a roughly 35-year working life. Somewhere between 38 and 50, it starts to falter — first irregularly, then progressively. The hormones it produces do not decline in a smooth line. They oscillate, sometimes wildly, before settling into the lower levels of menopause. That oscillation can last anywhere from two to ten years, and it is what we call perimenopause.

Perimenopause is not menopause. Menopause is a single retrospective day defined by 12 consecutive months without a period. Perimenopause is the long, messy descent into menopause, and it is where most of the symptoms women blame on “menopause” actually occur.

This distinction matters for diagnosis. The labs you would order to confirm menopause — low estrogen, high FSH, no period — often look normal in perimenopause. The reproductive system isn’t broken yet. It’s failing inconsistently. A snapshot of one day’s hormones tells you very little about what’s happening across a cycle, let alone across two years.

Why most workups miss it

When a woman in her forties presents with fatigue, weight gain, brain fog, and mood changes to a primary care doctor, the standard workup is some combination of TSH, CBC, fasting glucose, sometimes a B12, sometimes a vitamin D. If those come back within the lab’s reference range, she is told everything is fine.

There are three reasons that workup misses perimenopause.

The labs ordered don’t reveal perimenopause. None of the standard primary-care labs are designed to detect ovarian function. To see what the ovary is actually doing, you need a panel that captures the volatility — multiple data points across a cycle, or specific markers like AMH (which reflects ovarian reserve), or symptom-correlated measurements timed to specific cycle days. None of that happens in a 10-minute visit.

“Within reference range” is the wrong question. Reference ranges are statistical descriptions of the population that walked into a lab — including sick people. They are not optimal ranges for a 43-year-old woman with mitochondria, brain tissue, and bones that all need adequate hormonal signaling. A TSH of 3.8 is “normal” by lab standards. It is also a TSH at which a meaningful percentage of women feel terrible. The same is true for ferritin, B12, free T3, vitamin D, and progesterone.

Perimenopause symptoms overlap with at least eight other conditions, and clinicians are trained to rule out the dramatic before considering the obvious. The differential matters. I’ll get to it in a moment.

The framework, in order

When a woman in her forties walks into my office with the constellation of symptoms that suggest perimenopause, here is the order I actually work through.

Step 1: I listen for the story before I look at any number

Before any lab, before any physical exam, I want to hear the timeline. When did things start changing. What changed first. Has anything new been added — a medication, a stressor, a baby, a death, a job. What is her sleep like, and was it always like this. When was her last regular period.

The story is the diagnosis, ninety percent of the time. Labs are confirmatory. Labs without a story are noise.

I’m specifically listening for two patterns.

The first is the timeline of accumulation — symptoms that have been gradually adding up over 12 to 24 months, often starting with sleep changes or cycle changes, then mood, then weight, then cognitive symptoms. That accumulation pattern is hormonally driven and characteristic of perimenopause.

The second is the pattern of cyclic worsening — symptoms that flare predictably the week before a period, or in a specific window of the cycle. That cyclical signature tells me the system is responding to its own hormonal swings, which is the perimenopausal signature in vivo, not in a lab.

Step 2: I run a structured differential before I assume perimenopause

This is the step I think most workups skip, and it’s the reason I spend so much time on it. Before I conclude that what I’m seeing is perimenopause, I need to rule out — or at least seriously consider — every other condition that produces a similar picture in a woman in her forties.

The eight I always think about:

Thyroid dysfunction in three forms. Clinical hypothyroidism is the easy one. The harder ones are subclinical hypothyroidism — where TSH is “high-normal” and the rest of the panel tells a story TSH alone misses — and euthyroid sick patterns, where the body has down-regulated thyroid in response to chronic stress, illness, or undereating. The euthyroid sick picture often shows TSH-normal with low free T3, and conventional medicine tends to label it as nothing. The thyroid mimics perimenopause better than anything else, but more often, it layers on top of perimenopause and the metabolic shifts I’m about to describe. Many women I see have all three running at once. TSH-only screening misses most of the relevant patterns and misses the layering entirely. This deserves its own article, and I will write it.

The metabolic shift that catches fire in midlife. Most women arrive at perimenopause with decades of metabolic accumulation behind them. Pregnancies. Sleep debt. The cortisol of raising small humans. The workouts that fell off in the 30s. None of it shows up on standard labs while estrogen and progesterone are doing the quiet work of stabilizing the system. The moment those hormones stop steadying the ship, the underlying metabolic dysfunction often catches fire all at once — weight that won’t budge despite eating less, belly fat where there used to be none, fatigue that doesn’t respond to sleep.

The lab work most clinicians order misses this. Fasting glucose alone almost never picks it up; by the time fasting glucose is abnormal, the insulin dysregulation has been brewing for years. The earlier signals show up in fasting insulin, HbA1c, the 2-hour insulin response on a glucose tolerance test, and rising triglycerides. I will sometimes catch a woman where her fasting glucose is 88 and “normal” but her fasting insulin is 18 and her trigs are creeping toward 140. That’s the actual perimenopausal metabolic story, and it’s the one most workups never tell.

Iron deficiency without anemia. Ferritin under 50 is associated with fatigue, hair loss, brain fog, and mood symptoms in menstruating women. CBC will be normal. The standard “you’re not anemic” reassurance fails most perimenopausal women.

B12 deficiency in the functional range — under 400, sometimes under 500, in someone with neurological or cognitive symptoms. The standard cutoff of 200 misses the meaningful clinical picture entirely.

Sleep apnea. Underdiagnosed in women by an order of magnitude. Weight gain in the forties plus daytime fatigue plus cognitive symptoms — I will not commit to perimenopause as the answer until I have at least asked about snoring, witnessed apneas, morning headaches, and unrefreshing sleep.

Adrenal dysregulation from chronic stress. Cortisol patterns that disrupt sleep, alter mood, drive belly weight, and tank thyroid conversion. This often coexists with perimenopause rather than substituting for it, but the proportions matter for treatment.

Autoimmune disease in early presentation. Women in their forties have a peak incidence of autoimmunity. Hashimoto’s, lupus, RA, MS — they can all present with fatigue and mood symptoms before the classic markers show up.

Depression and anxiety, in either direction. This is where I want to be careful. Depression and anxiety can drive every symptom on the perimenopause list, which is why some clinicians try to rule them out first. But depression and anxiety are also direct manifestations of perimenopausal hormonal volatility — estrogen modulates serotonin, GABA, and dopamine signaling, and the swings of perimenopause can produce a depressive picture in a woman who has never been depressed in her life.

I am not interested in adjudicating which came first. The chicken-and-egg debate wastes time. What I want to know is whether depression or anxiety is present right now, and whether it needs its own line of treatment alongside everything else. Sometimes addressing the perimenopause clears the depression. Sometimes addressing the depression clears enough of the symptom load to see the perimenopause clearly. Sometimes both need to be treated simultaneously. The clinical move is to acknowledge it, address it, and stop pretending it’s only one or the other.

I do not run every test for every condition. I order based on the story. If a woman tells me she’s snoring and waking up gasping, I’m sending her for a sleep study before I draw a hormone panel. If her mother has Hashimoto’s and she has new joint stiffness, the antibodies are getting drawn first. The differential drives the testing, not the other way around.

Step 3: I order targeted labs, not panels

When I do test for perimenopause specifically, my approach is opposite of what you’ll see online. Most “perimenopause testing” content tells you to order the entire menu — DUTCH, full hormone panels, six-vial labs. That’s how I would test if I were trying to sell you supplements at the end. It is not how I would test if I were trying to figure out what is actually happening to you.

I think about labs as questions, not as data. Each lab answers one question. I ask the question first, then I order the lab that answers it.

If the question is “is her ovary still functioning predictably” — AMH, day-3 FSH, and estradiol on specific cycle days, with the answer requiring at least two of these read together.

If the question is “does her thyroid need to be ruled in or out” — full thyroid panel: TSH, free T4, free T3, reverse T3, TPO antibodies, TG antibodies. Not just TSH.

If the question is “is her metabolic system contributing to her symptoms” — fasting insulin, HbA1c, fasting glucose, lipid panel with particle size if relevant.

If the question is “is the picture nutrient-driven” — ferritin, B12, vitamin D, magnesium RBC, possibly homocysteine.

If the question is “is there an inflammatory or autoimmune driver” — hsCRP, ANA, sometimes a comprehensive metabolic panel.

I have a complete reference for which labs I order in which clinical situation, and I’ve put it together as a downloadable resource at Bespoke — that’s where the full diagnostic-decision tree for perimenopause labs lives, organized by chief complaint. Check it out here

The point of this article is not the lab list. It’s the thinking.

Step 4: I interpret labs in context, not in isolation

A lab result without context is a number. A lab result inside a clinical picture is information.

A TSH of 2.8 in a woman with no symptoms is not concerning. A TSH of 2.8 in a woman with fatigue, weight gain, hair loss, cold intolerance, and a family history of Hashimoto’s — that’s a different number. The lab hasn’t changed. The interpretation has.

This is why I run labs after the story, not before. If you run labs first, you end up trying to retrofit the symptoms to whatever the labs show. If you build the clinical picture first, the labs become tools for confirming or refining what you already suspect.

It’s also why I’m cautious about patients self-ordering enormous lab panels and showing up with twelve “abnormal” markers. Most of those are within statistical normal variation that means nothing in a person without symptoms — or they are pointing at something secondary while the actual driver sits unexamined. Labs without a clinician interpreting them in context are not better information. They are just more noise.

Step 5: I treat the pattern, not the lab number

Once I have the picture — perimenopause, with whatever overlapping conditions are also true — treatment becomes a question of priorities.

If a woman has perimenopause AND untreated thyroid dysfunction AND iron deficiency, treating the iron and thyroid often resolves 60-70% of her symptoms before we even discuss hormone-specific intervention. The hormones become a smaller, cleaner conversation.

If she has perimenopause AND insulin resistance, fixing the insulin resistance frequently reduces hot flashes, improves sleep, and stabilizes mood — without touching estrogen. The metabolic work is the perimenopause work.

If she has primary perimenopausal symptoms with everything else clean, then the conversation is about whether and when to introduce hormone therapy, lifestyle intervention, and what role specific supplements actually have evidence for. Most don’t. The supplement industry has done significant damage to women’s perception of what works in perimenopause.

The point: perimenopause is rarely a standalone problem to solve. It’s a stage of life that exposes every other dysfunction in the body. The clinician’s job is to figure out which dysfunctions are present, in what proportions, and treat in the order that gives the woman the most relief for the least intervention.

Common pitfalls I see, and what to do instead

A few patterns I see often enough to call out:

The “let’s just check your hormones” workup. A serum estradiol and progesterone drawn on a random day, in isolation. This will almost always be either normal or unhelpfully out of range, and it will not tell you what your hormones are doing across your cycle. If a clinician orders only this and tells you your hormones are fine, they have not actually evaluated your hormones. They’ve taken a single photograph of a film.

The “you’re too young” dismissal. Perimenopause can begin in the late thirties. Women have presented to me with classic patterns at 39, 40, 41 and been told for two years they were too young to be perimenopausal. They were not.

The “everything is normal, it must be stress” deflection. Stress is real, and it does drive symptoms, but it is rarely the only thing driving symptoms in a woman with the perimenopausal symptom cluster. The clinical move is to investigate further, not to send the woman home with sleep hygiene advice.

The supplement-stacking trap. A woman starts on five supplements based on a wellness influencer’s recommendation. Her symptoms don’t change. She concludes nothing works. In reality, she never had a diagnosis to begin with — the supplements were guesses, and guesses don’t work. Diagnosis first, intervention second.

The “DUTCH test fixes everything” myth. The DUTCH is a useful tool in specific situations. It is not a diagnosis, and most of the time it tells me what the clinical picture already showed. I order it when there’s a specific question it can answer, not as a default.

When this framework helps you most

If you are a woman in her forties who has been told her labs are normal but feels increasingly unrecognizable to herself — this framework is for you to hand to your doctor, or to use to find a different one, or to use to frame your own thinking before your next appointment. The diagnostic logic is reproducible. The questions are askable. The labs are orderable.

If you are a clinician who treats perimenopausal patients — this is roughly the structure I teach in Foundations, my course for women and clinicians who want to learn this framework in practical depth. The article is a sketch. Foundations is the full curriculum, with case examples, decision trees, and the lab interpretation library I use in clinical practice. Check it out here

If you are in Ontario and want to work with me directly, let’s chat.

If you are outside of Ontario, you can learn this approach in my guided program, Metabolic Foundations, which includes the complete Perimenopause Lab Testing Guide

For everyone else — read the next article in this series, on the five symptoms I take most seriously in women 40-50, which goes deeper into how I weight the chief complaint when I’m structuring a workup. The whole point of the series is to give you the thinking, not just the conclusion.