The evidence base underneath most of this guide was built disproportionately on middle-aged men, and the risk calculators were validated on populations where women were both under-represented and under-diagnosed. That has consequences, and this chapter is about them.
Two in particular. Women's risk is systematically underestimated by the usual tools, partly because a ten-year window flatters a group whose risk arrives later. And a set of events unique to women — pregnancy complications, early menopause, PCOS — carry cardiovascular information that has historically been thrown away, and is only now being formally collected.
1. Pregnancy as a stress test
Pregnancy imposes a large, sustained load: blood volume rises by around 40%, cardiac output rises, insulin resistance rises deliberately, and the vascular system has to remodel to accommodate all of it. Most women handle it. The ones who do not are being shown something.
This is no longer a fringe position. The 2024 AHA/ASA primary stroke prevention guideline named pregnancy complications, premature and early menopause and endometriosis as risk factors, and the 2026 multisociety dyslipidaemia guideline names early menopause, preeclampsia and gestational diabetes among the risk-enhancing factors that should refine risk assessment. Strong
| What happened in pregnancy | What it signals later |
|---|---|
| Preeclampsia | Roughly a two- to four-fold increase in later hypertension and cardiovascular disease, and around an 81% higher risk of stroke in meta-analysis. Now treated as an independent cardiovascular risk factor. Early-onset and severe disease carry more risk than late or mild. |
| Gestational hypertension | The same direction, less strongly. Also a strong predictor of later chronic hypertension. |
| Gestational diabetes | A substantial increase in later type 2 diabetes, and with it the whole cardiometabolic picture. The strongest single predictor most women will ever be given, and one commonly filed and forgotten. |
| Preterm birth | Associated with later cardiovascular disease independently of the other complications. |
| Small-for-gestational-age baby, placental abruption, stillbirth, recurrent loss | All associated with later cardiovascular risk. The common thread is placental vascular function, which is a window onto vascular function generally. |
An caveat about mechanism. It is not settled whether preeclampsia causes later cardiovascular disease or reveals a susceptibility that was always there — the "failed stress test" reading. For what you should do it makes no difference: either way the information is real and predictive. It matters for research, not for the appointment.
What to do with it
Write it down and keep it
Obstetric records do not follow you usefully into middle age. A note of what happened, when, and how severe, is worth more to a GP in twenty years than almost anything else in this guide.
Get blood pressure checked yearly, not opportunistically
After a hypertensive pregnancy this is the single highest-yield follow-up, and it is the one most likely to be missed once the obstetric team discharges you.
Treat gestational diabetes as a standing appointment, not a resolved episode
HbA1c periodically, and take weight and activity more seriously than the numbers alone suggest.
2. Menopause, and what actually changes
The lipid shift across the menopausal transition is real and reasonably consistent: LDL-C and apoB rise, triglycerides rise, HDL function deteriorates even where the HDL number does not, and the pattern shifts toward smaller, more numerous LDL particles. Blood pressure tends to rise, fat distribution moves toward the abdomen, and insulin sensitivity falls. Strong
Two things follow that are worth separating from the general noise about menopause.
A lipid panel taken before the transition is not a lipid panel that describes you afterwards. A reassuring result at 45 does not carry to 55. This is one of the few situations where re-testing is clearly worthwhile even without symptoms, and it is worth reading alongside chapter 2 on why a single panel means less than people assume.
Early menopause is a risk factor in its own right. Menopause before 45, and particularly before 40, is associated with higher cardiovascular risk, and it is now formally named as a risk enhancer in current guidance. If it applies to you it belongs in the conversation about whether a borderline risk score should be acted on.
3. PCOS
Polycystic ovary syndrome is usually presented as a fertility and menstrual problem. It is also, and arguably more importantly over a lifetime, a cardiometabolic condition: it travels with insulin resistance independently of weight, and carries higher rates of type 2 diabetes, dyslipidaemia, hypertension and MASLD.
The practical consequence is a screening one. A PCOS diagnosis at 22 is a reason for periodic HbA1c, blood pressure and lipid assessment through the following decades — and that follow-up frequently does not happen, because the condition was framed as a reproductive one and care ended when the reproductive question did. It is also a reason to read the glucose chapter and the liver chapter, because the same insulin resistance underlies all three.
4. Contraception and clot risk
Combined hormonal contraception — containing oestrogen — raises the risk of venous thromboembolism. The relative increase is real; the absolute risk in a healthy young woman without other risk factors remains low, which is why the method is used at all. Both halves matter, and most coverage gives only one.
What changes the arithmetic materially — noting that the first two are arterial risks rather than venous ones, and are contraindications for that reason rather than because of clot in the legs:
- Smoking, particularly over 35, which is the classic combination and a contraindication rather than a caution.
- Migraine with aura, which raises arterial risk — ischaemic stroke — with combined hormonal contraception, and is a contraindication on that basis.
- A personal history of venous thromboembolism, or a first-degree relative with one, particularly under 45, or a known thrombophilia. This is the venous half of the list and the most important single question.
- Uncontrolled hypertension, obesity, and immobility.
Progestogen-only methods and the intrauterine system do not carry the same venous risk, which is why they are the usual route where any of the above apply. This is a conversation for a GP or sexual health service rather than a page; what it is worth knowing is that the question has a good answer and alternatives exist.
5. Why the risk score reads low
QRISK3 does include some of this — it accounts for migraine, systemic lupus erythematosus, rheumatoid arthritis, corticosteroid use and severe mental illness, all of which affect women disproportionately. What no widely used UK calculator currently ingests is obstetric history.
Combined with the ten-year window, which systematically understates the case in anyone whose risk is loaded later — discussed under time horizons on the clinical page — the result is that a woman with a significant obstetric history can score reassuringly and be reassured wrongly.
The fix is not to distrust the score. It is to treat the risk enhancers as what current guidance now says they are: reasons to act on a borderline number rather than to wait for it to cross a threshold. And where the decision is genuinely finely balanced, apoB, Lp(a) and a CAC score answer the question the calculator could not.
What this chapter is not
It is not about heart disease in pregnancy, which is a specialist area and an urgent one. Chest pain, breathlessness or palpitations during or soon after pregnancy need assessment at the time, not a website. Nor does it cover how heart attacks present differently in women, which belongs to acute care rather than prevention — but two things are worth knowing, and the second is usually left out. Women are somewhat more likely than men to present without classical central crushing chest pain, and more likely to have their symptoms attributed to something else first. Chest pain is nonetheless still the commonest presenting symptom of a heart attack in women, as it is in men. The “women get atypical symptoms” shorthand is a real finding stretched too far, and stretched that way it does harm of its own — a woman with chest pain who has been told her symptoms will probably be unusual has been given a reason to wait.