Key Takeaways
- A study involving 61,437 UK Biobank participants found that accelerated brain aging, estimated from MRI scans, was associated with higher all-cause mortality and the development of diseases affecting multiple organ systems.
- Researchers used a statistical model to estimate each participant’s brain age and compared it with their chronological age.
- Higher brain-age acceleration was also associated with less favorable measures of physical health, including grip strength and lung function.
- Smoking, greater alcohol consumption, and poorer diet quality were associated with older-looking brains. Sleep duration and body mass index showed more complex relationships.
- Genetic analyses identified 29 genomic regions associated with brain-age acceleration, while blood analyses identified 62 proteins and 133 metabolites associated with the MRI-based measure.
- The study does not establish that accelerated brain aging causes earlier death or that lowering a person’s estimated brain age would extend their life.
Introduction
MRI-based brain-age estimates were associated with mortality risk in a study of 61,437 UK Biobank participants. Image is an illustrative representation, not actual study imaging.
Two people of the same chronological age can have brains that appear quite different on magnetic resonance imaging (MRI). One may show structural characteristics typical of someone younger, while the other may have features more commonly observed in an older person.
Scientists increasingly use these differences to estimate what they call brain age. The difference between MRI-predicted brain age and actual age is known as brain-age acceleration.
A large study published in npj Aging has found that this measure may reveal more than the condition of the brain itself.
Using MRI data from 61,437 participants in the UK Biobank, researchers found that people with greater brain-age acceleration had higher mortality risk and were more likely to develop diseases affecting several organ systems.
The researchers also investigated whether lifestyle habits, inherited genetic differences, and molecules circulating in the blood could help explain why some brains appear to age faster than others.
The study by Zong et al. (2026) provides evidence that MRI-based brain aging is associated with broader physical health. However, the results do not mean that an MRI scan can determine how long an individual will live.
What the Study Examined
The researchers analyzed data from the UK Biobank, a long-running research project that has collected medical, genetic, lifestyle, and imaging information from hundreds of thousands of people in the United Kingdom.
The final MRI analysis included 61,437 participants, comprising 28,795 men and 32,642 women.
The researchers developed a brain-age prediction model using MRI-derived measurements. The model employed a statistical method called least absolute shrinkage and selection operator regression, or LASSO, to identify imaging features useful for predicting chronological age.
After correcting for systematic age-related prediction bias, they calculated brain-age acceleration by subtracting each participant’s actual age from their predicted brain age.
For example, someone aged 60 whose MRI-based brain age was estimated at 65 would have five years of positive brain-age acceleration.
This does not mean that the person’s brain has literally aged five additional years. It means that their MRI measurements resemble those associated with an older brain according to the model.
The investigators then examined associations between brain-age acceleration and several outcomes, including death from any cause, newly diagnosed diseases, physical aging indicators, and lifestyle characteristics.
They also conducted genetic, blood protein, metabolite, and brain-cell analyses to explore biological processes potentially associated with the MRI findings.
For the mortality analyses, follow-up data extended through July 8, 2024. Disease-incidence follow-up extended through October 31, 2022.
What Researchers Found
Older-looking brains were associated with higher mortality risk
The central finding was a statistically significant association between greater brain-age acceleration and higher all-cause mortality.
The researchers used survival models to examine whether participants with older-looking brains were more likely to die during follow-up.
The association followed a graded pattern: mortality risk generally increased as brain-age acceleration increased.
The models accounted for chronological age, sex, education, and ethnicity.
The researchers also estimated differences in remaining life expectancy between participants in the highest and lowest quarters of brain-age acceleration.
Those in the highest quarter had lower modeled remaining life expectancy than those in the lowest quarter.
These estimates were generated from statistical survival models. They were not direct observations that a particular number of years had been lost by each participant.
Brain-age acceleration was associated with mortality risk even after accounting for chronological age and selected demographic characteristics. However, the analysis does not establish how much additional predictive information the MRI-based measure provides beyond established clinical risk factors, such as smoking, cardiovascular disease, and other underlying health conditions.
Further studies are needed to determine whether measuring brain age improves individual risk assessment.
Brain-age acceleration was associated with diseases throughout the body
The researchers examined conditions affecting multiple organ systems, rather than limiting their analysis to dementia and other neurological disorders.
Higher brain-age acceleration was associated with the incidence of diseases involving the cardiovascular, respiratory, digestive, endocrine, neurological, psychiatric, musculoskeletal, and other systems.
The study also identified associations with cancer and several common age-related conditions.
The size of the associations varied by disease category. The results should therefore not be interpreted as showing that brain-age acceleration predicts every disease equally well.
In additional analyses, participants who developed certain diseases at younger ages tended to have higher brain-age acceleration than those diagnosed later or those without the disease.
These findings raise the possibility that MRI-based brain aging reflects aspects of overall biological health.
They do not establish whether brain changes preceded the biological processes responsible for those diseases or resulted from them.
Brain-age acceleration was linked to physical aging indicators
The investigators also examined measures of physical function and health.
Greater brain-age acceleration was associated with several unfavorable aging-related indicators, including measures involving grip strength, lung function, and other physiological characteristics.
These associations are important because brain aging is often discussed mainly in relation to memory and dementia.
The results indicate that MRI-derived brain age may also capture changes associated with physical aging elsewhere in the body.
Shared underlying processes could contribute to changes in multiple organs simultaneously.
How Lifestyle Habits Were Related to Brain Aging
The researchers examined diet quality, smoking history, alcohol consumption, sleep duration, and body mass index.
Several patterns emerged.
Higher diet-quality scores were associated with lower brain-age acceleration.
Greater smoking exposure and alcohol consumption were associated with higher brain-age acceleration.
Sleep duration and body mass index showed U-shaped relationships, meaning that both lower and higher values were associated with greater brain-age acceleration than intermediate values.
In the statistical models, the lowest estimated brain-age acceleration occurred at approximately 7.37 hours of sleep per day and a body mass index of 24.68.
These figures describe the lowest points of fitted observational curves. They should not be treated as precise health targets or proof that a particular sleep duration or body mass index will prevent brain aging.
The researchers also performed mediation analyses to investigate whether brain-age acceleration statistically accounted for part of the relationship between lifestyle factors and adverse health outcomes.
For several associations, it did.
However, statistical mediation in observational data does not prove a biological chain in which smoking or diet changes brain age and brain aging subsequently causes disease.
The findings support further investigation into how lifestyle and brain health are connected, rather than establishing that changing a lifestyle factor will reverse an MRI-based brain-age measurement.
Genetic Findings Point to Multiple Biological Pathways
The study went beyond imaging and health outcomes to investigate genetic differences associated with brain-age acceleration.
A genome-wide association analysis involving 59,794 participants identified 36 independent genetic signals across 29 genomic regions.
Many of these regions had not previously been reported in studies of brain-age acceleration.
The researchers then examined genes potentially connected to these signals and assessed whether their activity in different tissues was associated with brain aging.
Several genes, including ORMDL1, STMN4, DCAKD, and STX6, emerged from analyses combining genetic and gene-expression information.
These genes are involved in biological processes such as lipid regulation, neuronal structure, cellular metabolism, and intracellular transport.
The results offer possible explanations for why some individuals may be more susceptible to accelerated brain aging.
However, identifying an associated genetic region does not establish that a particular gene causes brain aging. Additional functional studies are needed to determine which genes are directly involved and how they operate.
The researchers also constructed a polygenic risk score, which combines information from multiple genetic variants.
In a separate UK Biobank group not included in the original genetic association analysis, higher genetic risk scores for brain-age acceleration were associated with mortality and diseases affecting several organ systems.
This provides additional evidence that the biological processes associated with brain aging may overlap with those involved in broader health outcomes.
It does not mean that genetic testing can currently predict an individual’s lifespan from these findings.
Blood Proteins and Metabolites May Offer Additional Clues
One of the study’s more exploratory components examined substances circulating in the blood.
Among 2,923 proteins analyzed, 62 showed statistically significant associations with brain-age acceleration after correction for multiple testing.
The strongest protein association involved growth differentiation factor 15, or GDF15, a molecule previously investigated in relation to cellular stress, inflammation, and aging.
The researchers also analyzed 251 metabolic measurements and found that 133 were significantly associated with brain-age acceleration.
These included measurements related to glucose metabolism and the proportions of different fatty acids.
The analyses suggested that biological processes involving lipid metabolism, cellular maintenance, inflammation, and tissue function may contribute to the relationship between brain aging and systemic health.
The researchers developed blood-based protein and metabolite signatures associated with brain-age acceleration.
These signatures were subsequently associated with mortality and disease incidence in participants who were not used to construct them.
This is an interesting research direction because blood tests are generally more accessible than specialized brain MRI analysis.
However, the signatures remain research tools. The study did not establish that a blood test can accurately diagnose accelerated brain aging in individual patients or improve medical outcomes when used in routine care.
What Brain Cells May Be Involved?
To investigate possible cellular mechanisms, the researchers combined their genetic findings with existing single-cell brain data.
The analysis incorporated 101,924 cell nuclei obtained from brain tissue donated by 10 individuals.
Several cell types were associated with the genetic architecture of brain-age acceleration.
Endothelial cells showed particularly strong associations, followed by pericytes and astrocytes.
Endothelial cells line blood vessels, while pericytes help support small blood vessels and regulate their function. Astrocytes perform several essential roles in maintaining the brain’s chemical and structural environment.
These cell types are involved in processes relevant to the blood-brain barrier, cerebral circulation, inflammation, and cellular maintenance.
Their identification raises questions about whether vascular health and the brain’s supporting cells are important components of biological brain aging.
However, the cell analysis was based on a small number of tissue donors and computational integration with genetic data.
It did not directly demonstrate that dysfunction in these cells caused accelerated brain aging among the UK Biobank participants.
How the Study Compares With Earlier Research
The connection between an older-looking brain and increased mortality risk has been reported before. However, earlier investigations differed in sample size, imaging methods, and the health outcomes examined.
A 2018 study linked older-looking brains to earlier death
One of the earlier investigations was conducted by Cole et al. (2018), who analyzed MRI scans from 669 participants in the Lothian Birth Cohort 1936, a Scottish study of aging.
Participants were approximately 73 years old when their brains were scanned. During subsequent follow-up, 73 participants died.
The researchers found that each additional year of MRI-predicted brain age relative to chronological age was associated with a 6.1% higher mortality hazard after adjusting for age and sex.
The association remained statistically significant after further adjustment for factors including smoking, diabetes, cardiovascular disease, education, and childhood cognitive ability. In that model, each additional year was associated with a 5.1% higher mortality hazard.
A higher mortality hazard is a relative measure of the rate at which deaths occurred during follow-up. It does not mean that an individual’s absolute probability of dying increased by 5.1 percentage points.
The study also linked older-looking brains with weaker grip strength, poorer lung function, and slower walking speed.
These results are consistent with Zong et al. (2026), who similarly associated brain-age acceleration with mortality and measures of physical aging.
The earlier investigation involved a relatively small group of older adults, whereas the new study examines a much larger population and a broader range of diseases.
A 2025 study examined brain aging, mortality, and neurological disorders
More recently, Zhang et al. (2025) investigated MRI-derived brain-age differences using data from three research cohorts.
The analysis included 38,967 UK Biobank participants, 1,402 participants from the Alzheimer’s Disease Neuroimaging Initiative, and 1,182 participants from the Parkinson’s Progression Markers Initiative.
Using a deep-learning model to estimate brain age, the investigators found that each additional year of brain-age difference was associated with a 12% higher mortality hazard (hazard ratio 1.12; 95% confidence interval 1.09–1.15) after adjustment for chronological age, sex, body mass index, smoking, and alcohol consumption.
This was a relative association observed during follow-up, not a 12-percentage-point increase in an individual’s absolute probability of dying.
The study also examined associations with Alzheimer’s disease, mild cognitive impairment, and other neurological conditions.
These results provide further evidence that MRI-derived brain-age differences are associated with health outcomes beyond chronological age.
However, the 12% estimate should not be directly compared with the mortality association in the new study. The investigations used different prediction models, populations, statistical adjustments, and definitions of brain-age differences.
Both studies were observational, so neither establishes that accelerated brain aging causes disease or death.
What makes the new study different?
The investigation by Zong et al. (2026) extends this research in three important ways.
First, its MRI analysis included 61,437 participants, providing a large dataset for examining associations across multiple disease categories.
Second, it examined conditions affecting several organ systems rather than concentrating mainly on neurological disorders or mortality.
Third, it combined brain imaging with genetic, blood protein, metabolite, and single-cell analyses to investigate biological processes associated with brain-age acceleration.
The study identified 29 genomic regions, 62 circulating proteins, and 133 metabolic measurements associated with the MRI-based measure.
Together, the studies provide consistent evidence that older-looking brains are associated with poorer health outcomes.
The new research contributes a broader biological picture, but an important question remains unanswered: whether MRI-based brain age provides enough additional information beyond established medical risk factors to improve clinical decisions.
Until that question is addressed through independent validation and studies of clinical usefulness, brain-age estimates remain more appropriate for research than routine screening.
What the Results May Mean for Patients
The findings may eventually contribute to improved methods of identifying people at elevated risk of age-related disease.
An MRI-based brain-age measure could potentially complement other health indicators by capturing structural changes that are not fully reflected in chronological age.
Blood-based measurements may also provide a future avenue for investigating brain aging without requiring an MRI scan.
But the study does not establish that brain-age testing should become part of routine medical care.
It did not test whether providing patients with a brain-age score changes treatment decisions, improves disease prevention, or reduces mortality.
Nor did it demonstrate that any medication or lifestyle intervention can lower an individual’s brain-age acceleration score and thereby extend life.
The researchers identified existing medications with interactions involving genes associated with brain aging. These computational findings are not evidence that those drugs slow brain aging in humans.
For now, the results are most relevant to understanding the biology of aging and developing research tools, rather than making immediate changes to clinical practice.
Should You Get an MRI to Check Your Brain Age?
For most people, this study does not provide a reason to request an MRI solely to estimate brain age.
Although researchers can use specialized imaging models to identify patterns associated with biological aging, these measurements are not established as routine screening tools for predicting an individual’s lifespan or future disease risk.
The findings also do not establish that a particular brain-age score should determine treatment decisions.
People concerned about maintaining brain health can instead focus on established preventive measures, including avoiding smoking, staying physically active, managing blood pressure and other cardiovascular risk factors, getting adequate sleep, and following a balanced diet.
The National Institute on Aging’s guidance on cognitive health supports these general health measures. They have broader evidence supporting their benefits, although this particular study did not demonstrate that they reverse MRI-measured brain-age acceleration.
Anyone experiencing new or worsening problems with memory, thinking, movement, or other neurological functions should consult a healthcare professional. A research-based brain-age estimate cannot diagnose the cause of these symptoms or replace a medical evaluation.
Limitations
Several limitations affect how the results should be interpreted.
First, the research was observational. Although the mortality and disease analyses included follow-up data, the study cannot establish that accelerated brain aging causes disease or earlier death.
Second, the researchers developed their MRI-based brain-age prediction model using UK Biobank participants and evaluated its performance using additional imaging data from the same research resource. Although the model continued to show associations with chronological age and aging-related biomarkers, these findings do not establish that it would perform equally well in a completely independent population. Validation across different healthcare settings, demographic groups, and MRI scanning protocols is necessary before the approach can be considered broadly applicable.
Third, UK Biobank participants are predominantly of European ancestry and tend to be healthier and more socioeconomically advantaged than the general population.
People who undergo brain imaging may be an especially selected group. These characteristics limit how confidently the findings can be generalized to other populations.
Fourth, the blood protein and metabolite analyses involved subsets of participants, and the available tests measured only a portion of the molecules circulating in human blood.
Fifth, many analyses were conducted across genetic variants, proteins, metabolites, diseases, and biological pathways.
Although the researchers applied statistical corrections in several analyses, some exploratory findings may reflect residual confounding or chance and require independent replication.
Finally, the mortality models adjusted for selected demographic factors, but unmeasured or incompletely measured health differences may still explain part of the observed associations.
Final Thoughts
This large MRI study adds evidence that the apparent biological age of the brain is associated with health outcomes extending well beyond neurological disease.
Among more than 61,000 UK Biobank participants, greater brain-age acceleration was linked to higher mortality risk and diseases affecting multiple organ systems.
The researchers also identified associations involving lifestyle habits, genetic differences, blood proteins, metabolites, and brain-supporting cells.
Earlier research had already connected older-looking brains with mortality and neurological conditions. The new study strengthens that evidence while providing a more detailed examination of the biological processes that may contribute to brain aging.
However, it does not show that accelerated brain aging causes earlier death, that brain-age testing improves patient outcomes, or that reversing an MRI-based age estimate would extend lifespan.
The next challenge is to determine whether these measures provide clinically useful information beyond established risk factors and whether interventions that influence brain aging can meaningfully improve health.
Funding and Disclosures
The researchers acknowledged support from the National Natural Science Foundation of China, the Guangdong Basic and Applied Basic Research Foundation, science and technology projects in Guangzhou, and a Guangxi Natural Science Foundation project.
However, the manuscript’s separate funding declaration states that the authors received no specific funding for the work. The paper does not explain the distinction between these statements.
The authors reported no competing interests.
References
Cole, J. H., Ritchie, S. J., Bastin, M. E., Valdés Hernández, M. C., Muñoz Maniega, S., Royle, N., Corley, J., Pattie, A., Harris, S. E., Zhang, Q., Wray, N. R., Redmond, P., Marioni, R. E., Starr, J. M., Cox, S. R., Wardlaw, J. M., Sharp, D. J., & Deary, I. J. (2018). Brain age predicts mortality. Molecular Psychiatry, 23(5), 1385–1392. https://doi.org/10.1038/mp.2017.62
Zhang, R., Yi, F., Mao, H., Huang, Z., Wang, K., & Zhang, J. (2025). Brain age gap as a predictive biomarker that links aging, lifestyle, and neuropsychiatric health. Communications Medicine, 5, Article 441. https://doi.org/10.1038/s43856-025-01100-5
Zong, X., Wen, S., Xu, S., Yang, C., Zhu, Z., Cen, H., Zhu, J., Du, Q., Li, Y., Lv, J., Zhang, Y., Ding, C., & Ruan, G. (2026). Integrative analysis of brain-age acceleration: Health associations and multi-omics profiles. npj Aging. Advance online publication. https://doi.org/10.1038/s41514-026-00557-5
