Key takeaways

  • Without ovulation there is no corpus luteum and no meaningful progesterone, so the lining builds under estrogen and sheds heavily and without organisation.
  • Progesterone falls first, often years before estrogen does, which is why the correct treatment is frequently the one nobody offered.
  • Aromatisation from abdominal fat adds a second estrogen source, and estrogen-responsive fibroids compound the bleeding.
  • Structural and non-structural causes coexist — treating only one leaves a woman still bleeding, which is why severe bleeding warrants imaging.
  • A normal blood count does not exclude iron deficiency; ferritin measures stored iron directly and is the test that answers it.

The change usually arrives without warning. Periods that were predictable for twenty-five years become heavier, longer and more clotted, sometimes within a few cycles. Most women assume it is a one-off, then assume it is stress, then assume it is just what happens. It is none of those. It is a recognisable perimenopausal pattern with a specific hormonal cause and a well-established treatment framework. It goes unaddressed for so long because it is rarely named — not because it is hard to fix.

What the pattern looks like

Cycles shorten, then lengthen unpredictably. Flow becomes heavier and more clotted. Bleeding days extend. Premenstrual symptoms — irritability, breast tenderness, poor sleep, low mood the week before — intensify. Many women describe planning their week around their period for the first time since their teens.

That combination is not random. It is the mechanical consequence of one hormone falling before the other, and once you can see the sequence, everything else follows from it.

The anovulation story

In an ovulatory cycle, the spent follicle becomes a corpus luteum, which produces progesterone for the second half of the cycle. Progesterone does two things to the endometrium: it stops it proliferating, and it organises it — converting a thickening lining into a stable, secretory one that sheds in a controlled way.

In perimenopause, ovulation becomes unreliable. Some cycles you ovulate; some you do not. In an anovulatory cycle there is no corpus luteum, so there is no meaningful progesterone. The lining keeps building under estrogen with nothing to stabilise or organise it, and when it eventually outgrows its own blood supply it sheds — heavily, unpredictably, and with clots, because there was more of it and the shed was not coordinated.

This is why progesterone-first is the correct mental model rather than the "estrogen is dropping" story most women are given. Progesterone declines first, often years before the final period, while estrogen is still fluctuating at or above premenopausal levels (Prior, Front Biosci 2011), and the staging system used to describe the transition places this cycle irregularity squarely within it (Harlow et al., JCEM 2012). Perimenopause is not menopause, and treating it as an early version of the same thing is part of why the treatment offered is often wrong.

What makes it worse

Two things compound the imbalance in this decade.

More estrogen, less progesterone, thicker lining, heavier shed. That is the whole mechanism.

The structural question

Hormonal imbalance is the most common cause, but it is not the only one, and this is where a framework matters more than a hunch. Clinicians divide abnormal uterine bleeding into structural causes — polyps, adenomyosis, fibroids, and rarer endometrial pathology — and non-structural ones, including ovulatory dysfunction, coagulopathy and iatrogenic causes (Munro et al., FIGO classification). The point of the framework is that these are not mutually exclusive: a woman can have both anovulatory cycles and a fibroid, and treating only one leaves her still bleeding.

Fibroids are the common structural finding — up to 70-80% of women develop them by age 50 (Baird et al., Am J Obstet Gynecol 2003). Most are small and silent. Some sit where they distort the cavity and sharply increase bleeding, and because they are estrogen-responsive, the same environment driving anovulatory bleeding also feeds them. A pelvic ultrasound separates these possibilities, and is appropriate whenever bleeding interferes with daily life.

The iron cost, which usually goes unmeasured

Heavy periods deplete iron faster than diet replaces it, silently. The symptoms — fatigue, brain fog, hair shedding, exercise intolerance, restless legs, feeling cold — are the same ones women are told to expect from perimenopause itself, so the cause is routinely missed.

A standard blood count is the wrong test. Haemoglobin falls last; stores empty long before anaemia appears, and a woman can have a normal CBC with almost nothing in reserve. The test that answers the question is ferritin, which measures stored iron directly. Values below about 30 ng/mL indicate deficiency even by conservative criteria (Auerbach et al., JAMA 2025), and many women feeling terrible sit in the teens with a "normal" blood count on file. What ferritin means in both directions and transferrin saturation cover the interpretation, including the fact that ferritin is also an acute-phase reactant and can be falsely reassuring during inflammation.

What to measure, and when

The commonest testing error is drawing hormones on a random day and treating the result as definitive. In perimenopause a one-off estradiol tells you almost nothing; a luteal progesterone tells you something specific.

The treatment framework

  1. Restore the missing progesterone signal. Cyclical bioidentical micronised progesterone through the second half of the cycle replaces what the anovulatory cycles are not producing, which is what stabilises the lining. See what progesterone does beyond the uterus.
  2. Replete iron. A well-absorbed oral form on an empty stomach with vitamin C, away from calcium, tea and coffee. Retest ferritin rather than assuming — repletion takes months, not weeks.
  3. Address the aromatisation load where visceral fat is a factor. Less peripheral estrogen production means less of the imbalance to oppose.
  4. Reduce alcohol. It competes for hepatic clearance of estrogen and worsens the same pattern.
  5. Investigate structurally if bleeding is severe, or if hormonal treatment does not settle it within a few cycles.
  6. Consider a levonorgestrel intrauterine system where bleeding is heavy and structural causes have been assessed — local progestin thins the lining directly.
  7. Treat the whole transition, not one symptom, where several are present — when to start HRT sets out how that decision is made.

What is appropriate for an individual is decided by a physician after evaluation. The 60-second assessment is where that starts.

When this needs a clinician now

The clinical pearl: two things get missed here, repeatedly. The first is that progesterone falls before estrogen, so the correct treatment is often the one nobody offered. The second is ferritin — a normal blood count does not mean iron stores are intact, and a woman treated for bleeding without her iron being replaced will feel exhausted long after the bleeding settles.

Bottom line

Heavier periods in your forties are usually the perimenopausal anovulation pattern: no ovulation means no progesterone, so a lining that builds under unopposed estrogen sheds heavily and without organisation. Aromatisation from abdominal fat and estrogen-responsive fibroids compound it, and the iron cost accumulates quietly underneath. The framework is established — confirm ovulation with a luteal progesterone, exclude structural causes where the bleeding is severe, restore the progesterone signal, replete iron and measure it, and treat the transition rather than the symptom. This is specifically treatable, and it should not be tolerated on the assumption that it is simply what happens.

Educational content, not medical advice. Laboratory interpretation and any treatment decision are made by a licensed physician after individual evaluation. Individual results vary.

Progesterone first
it falls years before estrogen does
70-80%
of women develop fibroids by age 50
Ferritin
the test a normal blood count cannot replace