Hypertension in pregnancy tracked with later dementia — until the full adjustment

Hypertensive disorders of pregnancy — gestational hypertension, preeclampsia, HELLP syndrome and eclampsia — are among the best-established early markers of a woman's later cardiovascular risk [s1]. Whether that risk extends to the brain, in the form of dementia decades later, has been less certain. A nationwide cohort study published in JAMA Network Open on September 24 set out to test it, and its most important result is not the headline association but what happened to that association when the analysis was pushed further [s1].

The cohort

The researchers drew on the Korean National Health Insurance Service database, identifying women aged 16 or older who delivered between January 1, 2002, and December 31, 2005, and following them through December 31, 2024 — more than two decades [s1]. The analysis itself was conducted between October 2025 and May 2026 [s1].

Among 1,370,485 women, with a mean age of 29.2 years at baseline, 28,747 (2.1%) had experienced a hypertensive disorder of pregnancy [s1]. The primary outcome was incident dementia, including Alzheimer disease and vascular dementia, estimated with Cox proportional hazards models [s1].

What they found

The crude signal was there. The cumulative incidence of dementia was higher among women with a history of hypertensive pregnancy than among those without — 7.2 versus 2.8 per 10,000 women [s1].

The association persisted after adjustment for baseline maternal, socioeconomic and obstetric factors [s1]. But it did not survive the next step. After additional adjustment for post-delivery vascular and metabolic conditions — the hypertension, diabetes and related disease that commonly follow a hypertensive pregnancy — the association was no longer statistically significant [s1].

That is the finding that matters, and it points in a specific direction. It suggests the extra dementia seen after a hypertensive pregnancy is largely carried by the cardiovascular and metabolic disease that develops in the years afterward, rather than by something the pregnancy itself stamps directly onto the brain. The hypertensive pregnancy looks, on this evidence, less like an independent cause of dementia than an early flag for a vascular trajectory that plays out over decades.

Why caution runs in both directions

The authors are explicit that the number of dementia events among women with hypertensive pregnancies was small, and that the findings should be interpreted cautiously [s1]. A cohort that starts in women delivering in the early 2000s is, by 2024, only beginning to reach the ages when dementia becomes common. Follow-up that ends in middle age captures early-onset cases and little else, so both the association and its disappearance rest on relatively few events.

It also means the two analyses should not be over-read against each other. "No longer statistically significant after further adjustment" is not the same as "no effect." Adjusting for post-delivery vascular and metabolic conditions can remove a genuine causal pathway as easily as it removes confounding, because those conditions may be the very mechanism through which a hypertensive pregnancy raises dementia risk. Statisticians call this conditioning on a mediator, and it can make a real effect vanish.

How it sits with earlier work

An earlier and much-cited Danish national cohort of 1,178,005 women reported that a history of preeclampsia was associated with more than three times the rate of vascular dementia later in life (hazard ratio 3.46, 95% CI 1.97-6.10) [s2]. The Korean study does not contradict that so much as locate it: a strong crude and partially adjusted link, concentrated in the vascular subtype, that may be mediated by the cardiovascular disease a hypertensive pregnancy heralds [s1][s2].

What to watch

Whether longer follow-up, as this cohort ages into the eighth decade, strengthens or erodes the signal; whether the vascular-dementia subtype carries the association, as the Danish data imply [s2]; and whether researchers treat post-pregnancy vascular disease as a confounder or as a mediator, because that single choice decides whether the link appears or disappears.

This article describes observational cohort studies and is informational only. It is not medical advice.

Sources

• [s1] Hypertensive Disorders of Pregnancy and Incident Dementia, JAMA Network Open, 2026;9:e2635819, published online 2026-09-24. PMID 42783371.
• [s2] Pre-eclampsia and risk of dementia later in life: nationwide cohort study, BMJ, 2018;363:k4109, published 2018-10-17.