News · Fitness & Exercise
Inherited blood pressure risk tracked dementia in women only
In 9,145 Norwegians aged 70 and over, women in the top fifth for genetic blood pressure risk had higher odds of dementia. Men with the same genetics did not. Among fitter women the association was weaker.
- Genes that push blood pressure up are one of the best-established dementia risks.
- In women, the top fifth of that genetic score had 45% higher odds of dementia (OR 1.45).
- In men the same score did nothing at all (OR 1.03; 95% CI 0.77-1.37).
- Among fit women the association shrank to 1.12; among unfit women it was 1.53.
- Fitness was not randomly assigned, so this shows a pattern rather than a lever.
High blood pressure is one of the firmest entries on the list of things that raise dementia risk. What nobody could easily separate was cause from consequence: blood pressure changes in the years before a diagnosis, so a reading taken at 70 may be reporting the disease rather than predicting it.
Genes sidestep that. They are fixed at conception and cannot have been altered by an illness that arrives sixty years later.
A study in Age and Ageing used that trick. Its authors included 9145 participants in the HUNT4 70+ sub-study of the Norwegian Trondelag Health Study, all aged 70 or older, and found something previously hidden by pooling the sexes together. The inherited blood pressure signal showed up in women and was absent in men.
What a blood pressure polygenic score is
A polygenic risk score is one way by which people can learn about their risk of developing a disease, based on the total number of changes related to the disease. For blood pressure that means adding up hundreds of common variants, each nudging it a fraction of a point.
Two things about the number matter for reading this study. It reveals how a person’s risk compares to others with a different genetic constitution, so it is relative and not a probability. And polygenic scores only show correlations, not causations.
What makes it useful here is when it is set. A score fixed at birth cannot be contaminated by early disease, which is the flaw that has dogged blood pressure research in older adults.
What the HUNT Study found in women and men
Among 5011 females (mean age 78.5 years) and 4134 males (mean age 77.4 years), 812 (16.2%) and 570 (13.8%) dementia cases were identified, respectively.
The split is stark. Females in the highest fifth of the PGS, or polygenic score, had an increased odds of dementia against the rest, by roughly 45%. In men the same score did nothing: this was not observed in males, with an estimate sitting essentially on top of no difference.
That is not a subtle difference in degree. One sex shows the association and the other does not, in the same cohort, measured the same way, using the same score.
Where cardiorespiratory fitness changed the picture
The researchers then split the women by cardiorespiratory fitness and looked again.
The association was somewhat stronger in females with low fitness than in females with high fitness. In the less fit group the genetic signal was clear. In the fitter group it faded to something the data could not distinguish from nothing.
Read that carefully, because the honest version is weaker than the tempting one. The two estimates have wide and heavily overlapping ranges, so the study has not shown that fitness changes the effect of the genes. It has shown that the effect was visible in one group and not the other, which is a weaker claim and the one the authors make.
Their own summary is measured: genetic predisposition to high SBP (systolic blood pressure) was associated with higher odds of dementia in females, particularly if they had low CRF, meaning cardiorespiratory fitness.
Where this sits in the dementia risk picture
Hypertension already sits among the 14 modifiable risk factors named by the 2024 Lancet Commission, alongside physical inactivity. The Commission’s headline estimate is that 45% of future dementia could potentially be prevented if all 14 risk factors were eliminated.
So neither variable is news. What this study adds is that the two interact, and that the inherited half of blood pressure risk does not behave the same way in both sexes.
At the individual level the Commission recommends treating depression, hypertension, high midlife LDL cholesterol, obesity and diabetes, and engagement in physical, social and cognitive activities. Nothing in the Norwegian data argues against any of that. It argues that the benefit may not be evenly distributed.
What a genetic score cannot establish about dementia
Nobody assigned fitness; it was simply measured, and that is the central limit. Women who are fit at 78 differ from women who are not in ways that include how much illness they were already carrying, which is exactly the reverse-causation problem the genetic design was chosen to avoid, reappearing on the other variable.
Fitness here is also estimated from routine measures rather than tested directly, which is practical at this age and blunt as a measurement.
The cohort is Norwegian and the score was built largely in similar populations. The accuracy of polygenic risk scores may only be valid and useful for European ancestry populations, so these numbers should not be carried across ancestries.
And dementia was counted once, at one point, in people aged 70 and over. Anyone who developed dementia and died before the sub-study is missing from it.
What to take from a dementia risk you cannot act on
Nobody should order a polygenic score after reading this.
It reports how your risk compares with other people’s, offers no timeframe, and changes no treatment.
The useful part is a reminder that pooled results can hide the answer. A sex-combined analysis of this cohort would have reported a modest association and told you nothing about who carries it.
The authors ask for the obvious follow-up: future studies should further examine sex-specific effects of genetic and modifiable factors on dementia risk.
Treating high blood pressure and staying fit were already the advice. This is a reason to think the advice is not equally urgent for everyone.
People also ask
What is a blood pressure polygenic score?
A single number summing up the many small common gene variants that nudge blood pressure up or down. A polygenic risk score is one way by which people can learn about their risk of developing a disease, based on the total number of changes related to the disease. It reports relative rather than absolute risk, and it is fixed at conception, which is what makes it useful here: your genes cannot have been changed by early dementia.
What did the study find?
Females in the highest fifth of the polygenic score had an increased odds of dementia (OR 1.45; 95% CI 1.10 to 1.90). This was not observed in males (OR 1.03; 95% CI 0.77 to 1.37). Among 5,011 women (mean age 78.5 years) and 4,134 men (mean age 77.4), 812 (16.2%) and 570 (13.8%) dementia cases were identified respectively.
Where does fitness come in?
The researchers split the women by cardiorespiratory fitness and re-ran the analysis. The association was somewhat stronger in females with low fitness (OR 1.53; 95% CI 1.08 to 2.17) than in females with high fitness (OR 1.12; 95% CI 0.78 to 1.63). Note the intervals overlap heavily, so the difference between those two numbers is not itself established.
Does this mean exercise cancels genetic risk?
No, and the study cannot test that. Fitness was measured, not assigned, so fitter women differ from less fit women in dozens of ways, including how healthy they already were. What the data show is that the genetic signal was visible mainly in the less fit group. Whether raising fitness would move anybody between those groups is a different question requiring a trial.
Why would the result differ between women and men?
The paper does not settle this, and states plainly that future studies should further examine sex-specific effects of genetic and modifiable factors on dementia risk. Candidate explanations include differences in blood pressure trajectories across the lifespan, effects of menopause on vascular health, and the fact that more women than men survive to the ages when dementia is diagnosed. All are hypotheses here.
How was fitness measured in people over 70?
As estimated cardiorespiratory fitness, derived from routinely collected variables rather than a treadmill test to exhaustion, which is impractical at scale in this age group. That is a real limitation. It is also why the fitness split should be read as a rough sorting rather than a precise measurement.
What should I do about it?
Nothing specific, and no one should seek a polygenic score on the strength of this. Scores only show correlations, not causations, and their accuracy may only be valid and useful for European ancestry populations, which this Norwegian cohort is. This is general information rather than medical advice. Treating high blood pressure is already standard advice for dementia risk, and it does not depend on knowing your genotype.
References
- Lerfald, M., Ovretveit, K., Nilsen, T. I. L., et al. Blood pressure polygenic score, cardiorespiratory fitness and odds of dementia: the HUNT Study. Age and Ageing, 2026.
- National Human Genome Research Institute. Polygenic Risk Scores. US National Institutes of Health.
- Livingston, G., et al. Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet, 2024.