Pregnancy history sharpens 10-year hypertension prediction

Design

Population-based cohort, prediction model development

Population

28,234 women after a first birth in the Netherlands, followed up to 10 years

Primary outcome

10-year incident hypertension

Effect

C-index 0.659 to 0.772 with pregnancy factors (difference 0.113, 0.092-0.132)

A Dutch population cohort (BJOG, 16 October 2026) linked perinatal registry and primary care records for 28,234 women after their first birth and followed them for up to 10 years. 639 developed hypertension, a cumulative incidence of 3.3%.

A model built on sociodemographic and standard cardiovascular factors discriminated modestly (C-index 0.659). Adding pregnancy-related factors raised it to 0.772, a difference of 0.113 (95% CI 0.092-0.132), driven mostly by hypertensive disorders of pregnancy. Reclassification improved for women who went on to develop hypertension (NRI 0.213) without misclassifying those who did not.

The point for practice is less the model itself, which needs external validation, than what it shows: for young women, the pregnancy history carries more predictive information than the conventional risk factors. A risk assessment that ignores it is missing most of the signal. That supports what the pearl describes - making the obstetric diagnosis travel into long-term care.

• Record hypertensive disorders of pregnancy as a cardiovascular risk factor in every woman's notes
• Ask about pregnancy complications when assessing blood pressure risk in women under 50
• Consider earlier and more frequent blood pressure checks after a hypertensive pregnancy
• Treat this model as research-stage until validated outside the Netherlands

Why it matters

Standard risk scores for young women miss the factor that predicted hypertension best: what happened in pregnancy.

Don't overread it

The model has not been validated outside the Dutch population it was built in.

The statistics, in plain English

A C-index is the chance the model ranks a woman who develops hypertension above one who does not: 0.5 is a coin toss, 1.0 perfect. Moving from 0.66 to 0.77 is a substantial gain. The net reclassification of 0.213 means about a fifth more of the women who developed hypertension were moved into a higher risk group, at no cost to those who did not.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.