Your FSH Test Is Probably Not the Answer You Need

FSH — follicle stimulating hormone — is the signal your brain sends to your ovaries telling them to mature an egg. It's produced by the pituitary gland, and its level is governed by feedback: when your ovaries produce plenty of estrogen and a hormone called inhibin B, the pituitary gets the message and keeps FSH low. When ovarian output falls, the pituitary shouts louder and FSH climbs.

That logic is why FSH became the traditional menopause test. In a woman who is definitively postmenopausal, FSH is reliably high — usually above 30 IU/L, often much higher — because there's essentially no ovarian feedback left to suppress it. The test works beautifully at that endpoint.

The problem is everything before the endpoint. Perimenopause is not a gradual slide; it's a chaotic oscillation. Your ovaries in your late forties don't quietly wind down. They sputter. Some cycles produce a robust follicle and a surge of estrogen far higher than you ever made at 30. Others produce nothing at all. FSH follows that chaos, which means a single blood draw is a photograph of a hurricane — technically accurate, informationally useless.

This is the single most common source of frustration women bring to us: symptoms that are obviously hormonal, a blood test that says 'normal', and a clinician who concludes nothing is happening.

Because in perimenopause your FSH is genuinely normal on some days and genuinely elevated on others — and you were tested on a normal day. Studies that measure FSH repeatedly across a single cycle in perimenopausal women find values crossing back and forth over the 'menopausal' threshold within weeks. A woman can have an FSH of 45 IU/L in March and 8 IU/L in April, with no error in either measurement.

The STRAW+10 staging system — the international reference framework for reproductive aging, published by Harlow and colleagues in 2012 — is explicit about this. Its authors note that because of the significant variation in hormone levels, measurements of FSH and estradiol are not routinely indicated to diagnose menopause. STRAW+10 stages the transition primarily on menstrual cycle pattern: early perimenopause is marked by persistent cycle-length variation of 7 days or more, and late perimenopause by an interval of 60 days or more between periods.

That's why guidelines converge on the same advice. NICE guideline NG23 recommends not using FSH to identify perimenopause or menopause in people aged 45 or over, and its associated quality standard states that women over 45 presenting with menopausal symptoms should be diagnosed on symptoms alone. The Menopause Society, ACOG, and the International Menopause Society all advise against relying on a single FSH value in perimenopause.

There's a second, subtler problem with the test that rarely gets mentioned: reference ranges assume a stable state. Laboratory cut-offs for FSH were derived from clearly premenopausal and clearly postmenopausal populations, where the number is stable and meaningful. Perimenopause is, by definition, the unstable middle — the one physiological state the reference range was never designed to describe. Applying a binary threshold to a variable that swings across it repeatedly produces a result that is technically valid and clinically empty.

A normal FSH is not evidence that your symptoms are stress, or aging, or in your head. It is evidence that you had a blood test on a Tuesday.

There are three situations where FSH earns its place, and they're all about age or about missing information.

If you're under 40 and your periods have changed or stopped, FSH is essential. Premature ovarian insufficiency affects roughly 1 in 100 women under 40, and diagnosing it changes everything — these women are usually advised to take hormone therapy until at least the average age of natural menopause, because the long-term bone and cardiovascular consequences of years without estrogen are serious. In this setting, guidelines call for two elevated FSH measurements taken four to six weeks apart, not one.

If you're between 40 and 45 with menopausal symptoms and a change in your cycle, NICE says FSH should be considered. It won't be definitive, but it can support a picture that's otherwise ambiguous at that age.

If you've had a hysterectomy but kept your ovaries, you have no menstrual pattern to read, so FSH becomes one of the few objective signals available. The same is true if you have a hormonal IUD suppressing your bleeding, or if you're on certain treatments that mask cycles.

One important exception in the other direction: combined hormonal contraception makes FSH uninterpretable. The pill suppresses FSH by design. If you're on it, the number tells you the pill is working, nothing more. This comes up constantly for women navigating [birth control in perimenopause](/blog/birth-control-in-perimenopause-what-to-use-after-40), where the contraception is simultaneously managing symptoms and hiding the transition underneath.

AMH — anti-Müllerian hormone — is more stable across the cycle than FSH, which makes it a better ovarian reserve marker. Research from the SWAN cohort has shown that very low AMH combined with age can help estimate the likely timing of the final menstrual period at a population level. But 'helps predict at a population level' and 'tells you when your last period will be' are different claims. Individual prediction remains imprecise, often by years, and AMH does nothing to explain your symptoms or guide treatment. It is a fertility test that got recruited into menopause marketing.

Estradiol has the same volatility problem as FSH, with an added twist: in early perimenopause, estradiol is often *higher* than in your thirties, not lower. Those surges are part of why breast tenderness, heavy bleeding, and migraine can worsen before anything resembling deficiency appears. A high estradiol result does not mean you aren't perimenopausal.

At-home menopause test kits almost universally measure FSH from urine or a finger-prick. They inherit every limitation above and add sampling variability on top. A positive result tells you FSH was high that morning. A negative result tells you nothing at all. They are not a substitute for a conversation with a clinician.

If you want blood work in midlife, the more valuable panel is usually different: TSH and free T4 for thyroid, ferritin for iron, vitamin D, a lipid panel, and HbA1c. These identify treatable conditions that overlap heavily with menopause symptoms — which is exactly the problem covered in [thyroid or perimenopause](/blog/thyroid-or-perimenopause-overlapping-symptoms-what-to-test).

By bringing a pattern rather than a symptom. Clinicians respond to structure, and the STRAW+10 framework gives you exactly the structure they're trained to recognize. Before your appointment, track three things for two to three months: cycle length (first day of bleeding to first day of the next), the specific symptoms and their timing within the cycle, and the impact — sleep lost, days affected, work missed.

The two most diagnostically powerful sentences you can say are about cycle variation. 'My cycles used to be 28 days and now they range from 22 to 40' places you in early perimenopause under STRAW+10. 'I've gone 70 days without a period' places you in late perimenopause. Neither requires a blood test.

Bring the symptom cluster rather than the headline. Hot flashes get recognized instantly, but they're often not the first or worst thing. Sleep fragmentation at 3 a.m., a rage response that feels alien to you, joint stiffness in the mornings, brain fog that makes familiar words vanish mid-sentence, and new anxiety are all commonly earlier. Many women spend years being treated for the pieces separately — an antidepressant here, a sleep aid there — before anyone assembles them. That fragmentation is why [perimenopause is so often mistaken for ADHD](/blog/perimenopause-or-adhd-why-midlife-blurs-the-line) or [for depression](/blog/menopause-depression-why-risk-peaks-in-perimenopause).

If a clinician tells you that a normal FSH rules out perimenopause, it is entirely reasonable to say: 'My understanding is that guidelines recommend diagnosing this clinically over 45 because FSH fluctuates too much to be reliable. Can we approach it that way?' That is not being difficult. That is quoting the guideline.

And if you're also dealing with weight changes that arrived alongside everything else, [perimenopause weight gain](/blog/perimenopause-weight-gain-why-it-happens-swan-data) has its own physiology worth understanding separately.