A blood test may tell a different story depending on genetics
Blood tests are becoming increasingly important in research on Alzheimer’s disease. But the meaning of a biomarker result may depend partly on a person’s age and genetic background.
A new study in Nature Communications analysed blood samples from 5,913 adults in the Netherlands Twin Register, ranging in age from 30 to 80.
The researchers found that several Alzheimer’s-related blood markers changed more steeply with age in people carrying the APOE ε4 genetic variant.
Among carriers, the amyloid-beta 42/40 ratio declined more sharply after approximately age 50, while phosphorylated tau 217 rose more steeply after approximately age 55.
Another marker, GFAP, showed a steeper age-related rise after about age 60. These are group-level patterns, not personal diagnostic thresholds.
The findings could improve how scientists interpret emerging blood tests, particularly when evaluating people who do not yet have symptoms.
Why this is a timely research question
Interest in Alzheimer’s blood biomarkers has grown because blood collection is less invasive than some established methods used to investigate disease-related brain changes.
Researchers hope blood measurements can help identify underlying pathology earlier and support more targeted follow-up assessments.
However, a blood result does not exist in isolation. Age, biological variation, health conditions and genetics can influence the concentration of measured proteins.
If two people have the same laboratory result but different genetic backgrounds, that result may not carry identical implications.
The new study addresses this challenge by examining the interaction between genetic risk, age, sex and several blood biomarkers.
It does not introduce a consumer screening test or establish that everyone should undergo genetic testing.
What APOE ε4 actually means
APOE is a gene involved in lipid transport and other biological processes. One of its common variants, called ε4, is associated with increased susceptibility to Alzheimer’s disease.
Carrying the variant changes risk at a population level, but it does not mean that a person will inevitably develop dementia.
Likewise, people without APOE ε4 can develop Alzheimer’s disease, so the variant is not a complete explanation of who becomes ill.
In the study cohort, 29.8 percent of participants carried ε4. The average patterns for carriers and non-carriers could therefore be compared.
The researchers also assessed a broader polygenic risk score that excluded the APOE region.
That additional score summarises genetic susceptibility across multiple variants rather than focusing on one well-known gene.
The researchers examined nearly six thousand adults
The primary dataset contained 5,913 participants from the Netherlands Twin Register. Their ages ranged from 30 to 80 years.
Women accounted for 64.1 percent of the sample, an important detail when interpreting the findings.
The study measured four plasma biomarkers associated with neurological or Alzheimer’s-related processes.
These were phosphorylated tau 217, the amyloid-beta 42/40 ratio, neurofilament light and glial fibrillary acidic protein.
The analysis investigated whether the relationship between age and each biomarker differed by APOE ε4 status, broader genetic risk or sex.
An additional longitudinal cohort, EMIF-AD, was used to examine whether a key biomarker pattern was consistent with Alzheimer’s pathology.
The large primary sample strengthens the study’s ability to identify group-level associations, although its population composition remains relevant.
Four blood markers, four different signals
Phosphorylated tau 217, usually written p-tau217, is a blood biomarker of particular interest in Alzheimer’s research.
The amyloid-beta 42/40 ratio compares two forms of amyloid-beta peptide. Changes in that ratio can be associated with amyloid-related pathology.
Neurofilament light, or NfL, is a marker associated with neuronal injury, but it is not specific to Alzheimer’s disease.
GFAP is a protein associated with astrocytes, cells that play important supporting roles in the nervous system.
These measurements provide different windows into biological processes, and they should not be treated as interchangeable.
The study investigated how their concentrations vary across age and genetic groups, rather than assuming that one marker captures the entire disease process.
That distinction matters because clinical interpretation often depends on combining several kinds of evidence.
The age of fifty emerged as an important point
Among APOE ε4 carriers, the decline in the amyloid-beta 42/40 ratio became steeper after approximately age 50 compared with non-carriers.
This does not mean that a person suddenly develops Alzheimer’s disease on their fiftieth birthday.
Instead, the result describes a change in the statistical relationship between age and biomarker concentrations in the study population.
Individual results can vary substantially around group averages, and not everyone follows the same trajectory.
The finding may nevertheless help researchers understand why fixed reference values can be difficult to apply across genetically diverse groups.
It also highlights why interpreting a single result without considering age could be misleading.
Tau changes became clearer later in life
For p-tau217, the researchers observed a steeper age-related increase in ε4 carriers after approximately age 55.
That finding is especially relevant because p-tau217 has attracted attention as a potential marker of Alzheimer’s-related pathology.
The investigators reported that the APOE ε4-associated effect on p-tau217 was stronger in males than females.
That sex-related difference is a research finding requiring careful interpretation, not a reason to assume that one sex can safely ignore testing.
The longitudinal validation cohort supported the interpretation that changes in p-tau217 among carriers reflected Alzheimer’s pathology.
However, the study does not provide a personalised forecast of when any participant will develop symptoms.
Another signal appeared around sixty
The age-related increase in GFAP became steeper among APOE ε4 carriers after approximately age 60.
GFAP can reflect changes involving astrocytes, but its concentration may also be affected by biological processes outside Alzheimer’s disease.
The result adds to the picture of several markers changing at different stages of adult ageing.
It would be an oversimplification to interpret these age points as a fixed sequence that every person follows.
Researchers are interested in these patterns because they may improve the timing and interpretation of future investigations.
Clinical decisions still require appropriate context, validated tests and professional assessment.
What the broader genetic score did not show
One of the study’s important negative findings concerned the Alzheimer’s polygenic risk score excluding APOE.
The researchers did not find an association between that score and the age-related concentrations of the measured biomarkers.
Negative findings can be scientifically valuable because they help distinguish promising signals from relationships that were not supported by the data.
This result does not mean that other genetic variants are irrelevant to Alzheimer’s disease risk.
It means that the particular broader score did not show the investigated biomarker association in this analysis.
Genetic risk prediction and blood biomarker interpretation remain related but distinct scientific tasks.
Why a separate cohort mattered
The researchers used a longitudinal cohort known as EMIF-AD to examine whether the p-tau217 pattern aligned with Alzheimer’s pathology.
Longitudinal data follow participants over time and can provide information beyond a single cross-sectional comparison.
The additional analysis supported the interpretation that changes in p-tau217 among APOE ε4 carriers reflect disease-related processes.
That strengthens confidence in the biological relevance of the finding without converting it into proof of an individual’s future diagnosis.
The main population-based cohort and the validation cohort serve different analytical purposes.
Independent confirmation in further populations would still be useful, particularly in settings with different ancestry and healthcare characteristics.
Why blood tests are not the same as diagnosis
A biomarker can provide information about a biological process without being sufficient to diagnose a condition by itself.
Alzheimer’s disease is complex, and clinical evaluation can involve symptoms, cognitive assessment, medical history and other investigations.
Blood tests may eventually make parts of that process more accessible, but their interpretation must be validated for the intended use.
An abnormal result in someone without symptoms does not automatically mean that dementia is imminent.
Similarly, a reassuring result does not guarantee that a person will never experience cognitive decline.
The new research concerns statistical patterns in biomarkers, not a recommendation for population-wide self-testing.
Anyone worried about memory changes should seek an appropriate clinical assessment rather than relying on a single genetic or laboratory measurement.
Why the sample has limitations
The Netherlands Twin Register provided a substantial dataset, but it is not a random sample of every population in the world.
Patterns may differ across ancestry groups, age distributions and health settings.
The primary cohort also included a higher proportion of women than men, which matters when examining sex-related effects.
Observational biomarker research can identify associations but cannot by itself establish every underlying biological mechanism.
Changes in blood protein levels can reflect several influences, and measurement methods also require careful standardisation.
These limitations do not negate the results; they define where additional work is needed before broad clinical application.
What this could mean for future testing
If age and APOE ε4 status systematically influence biomarker trajectories, future diagnostic models may need to account for those factors.
That could improve how researchers compare an individual’s result with an appropriate reference population.
However, adding more information to a model does not automatically improve care. The model must be tested for accuracy, fairness and practical benefit.
Genetic information also raises questions about consent, privacy and counselling.
The study does not show that routine APOE genotyping alongside every blood test would improve patient outcomes.
That question requires separate clinical evaluation and careful consideration of potential harms as well as benefits.
How this connects to other Alzheimer’s research
Other research has investigated how Alzheimer’s-related proteins behave in the living human brain and why some people remain cognitively resilient despite disease-associated pathology.
Research Today previously reported on differences between tau turnover and amyloid plaque stability in living brain tissue.
Another Research Today article examined why some brains showed resilience despite elevated Alzheimer’s disease markers.
These findings explore different questions and should not be treated as direct confirmation of the new blood-biomarker analysis.
Together, they illustrate why researchers examine pathology, biological markers and cognitive outcomes as separate but connected dimensions.
That broader context is useful for understanding why no single marker tells the whole story.
The most important unanswered questions
Researchers still need to establish how well genetically informed biomarker models perform in diverse clinical populations.
Prospective studies could investigate whether adjusting for age, sex and APOE status improves prediction of clinically meaningful outcomes.
Additional work could compare the benefits and risks of different testing strategies in people with and without symptoms.
Health systems would also need to consider access, cost, counselling and the consequences of uncertain results.
The new findings offer a stronger basis for those investigations, but they do not settle how blood testing should be used in routine care.
The bottom line
In 5,913 adults aged 30 to 80, APOE ε4 carrier status was associated with steeper age-related changes in several Alzheimer’s-related blood biomarkers.
The amyloid-beta 42/40 ratio changed more sharply after about age 50, p-tau217 after about 55, and GFAP after about 60.
The broader Alzheimer’s polygenic risk score excluding APOE did not show the same association with biomarker concentrations over age.
A separate longitudinal cohort supported the relevance of the p-tau217 finding to Alzheimer’s pathology.
The study strengthens the case for considering genetic background and age when interpreting emerging blood markers.
It does not provide a stand-alone diagnostic test, establish that carriers will develop dementia, or justify self-screening without clinical guidance.
The value of the research lies in making future early-detection tools more precise, while preserving the distinction between risk and diagnosis.
Source Information
Study: APOE ε4 carriership status enhances age-related changes in Alzheimer’s disease plasma biomarkers concentrations.
Authors: Rebecca Z. Rousset, Anouk den Braber, Bram Bongers and colleagues.
Journal: Nature Communications.
Published: 17 October 2026.
Methodology: Population-based analysis of 5,913 Netherlands Twin Register participants aged 30–80, measuring p-tau217, Aβ42/40, NfL and GFAP, with additional longitudinal cohort analysis.
