Brain age is a machine-learning estimate of how old your brain looks based on its structure or function, compared to what is typical for your actual calendar age. The gap between the two, often called the brain-age gap, has emerged over the past decade as a potential biomarker for how well or poorly your brain is aging. A brain that “looks” five years older than it should may signal accelerated wear from disease, lifestyle, or genetics, while a brain that looks younger than expected may reflect protective factors. The concept is generating real excitement in neuroscience, though it comes with important caveats about what it can and cannot tell you today.
How Brain Age Gets Measured
The most common approach uses a structural MRI scan of your head. Researchers feed thousands of brain scans from healthy people into a machine-learning algorithm, training it to recognize the patterns that change as brains age: thinning of the cortex, shrinkage of certain structures, changes in white matter. Once trained, the model can look at a new scan and produce a predicted age. The difference between that prediction and your chronological age is the brain-age gap, sometimes called brain-age delta or predicted age difference (PAD).1PubMed Central. Machine learning for brain age prediction: Introduction to methods and clinical applications A positive gap means your brain looks older than expected; a negative gap means it looks younger.
Models vary in what features they use. Some rely on cortical thickness, cortical volume, and subcortical volume measures. One large study, for example, trained separate models for women and men using these measures from over 22,000 participants in the UK Biobank.2eLife. Biological brain age prediction using machine learning on structural neuroimaging data: Multi-cohort validation against biomarkers of Alzheimer’s disease and neurodegeneration stratified by sex Others incorporate white matter microstructure from diffusion MRI, or even functional connectivity data from resting-state scans. There are also experimental approaches using EEG recordings, though these tend to be less precise, with average prediction errors around seven years compared to roughly two to four years for the best MRI-based tools.3PubMed Central. Predicting Age From Brain EEG Signals—A Machine Learning Approach
Several pre-trained software packages are now publicly available, allowing any research group with MRI data to estimate brain age without building a model from scratch.4PubMed Central. Prediction of brain age using structural magnetic resonance imaging: A comparison of accuracy and test-retest reliability of publicly available software packages That accessibility has fueled a wave of studies connecting brain age to nearly every aspect of health.
Brain Age and the Risk of Dementia
The strongest clinical interest in brain age revolves around dementia, and Alzheimer’s disease in particular. People who already have mild cognitive impairment and a large brain-age gap appear to face a significantly higher chance of progressing to full Alzheimer’s. In one study tracking people with mild cognitive impairment, each additional year of brain-age gap was linked to roughly a 10% greater risk of converting to Alzheimer’s. Those in the highest quartile of brain-age gap had about four times the risk of developing Alzheimer’s compared to those whose brain-age gap was smallest, and their cumulative probability of conversion reached about 92%.5PLoS ONE. BrainAGE in Mild Cognitive Impaired Patients: Predicting the Conversion to Alzheimer’s Disease
A separate study confirmed that the brain-age gap in people with mild cognitive impairment correlated with how quickly they went on to develop Alzheimer’s: a larger gap meant faster conversion.6PubMed Central. NeuropsychBrainAge: A biomarker for conversion from mild cognitive impairment to Alzheimer’s disease When comparing groups of people with normal cognition to those with an Alzheimer’s diagnosis, one analysis found that the Alzheimer’s group had a mean brain-age gap about five years higher, producing a large effect size.7EBioMedicine. Prediction of brain age using structural magnetic resonance imaging: a comparison of clinical utility of publicly available software packages
These are striking numbers, but they come with a caveat. The brain-age gap did not predict the rate of gray matter loss over time in people with normal cognition or mild cognitive impairment. Brain age appears to capture a snapshot of accumulated damage rather than forecast how fast things will deteriorate from that point forward.
Mental Health Conditions and an Older-Looking Brain
Dementia is not the only condition linked to accelerated brain aging. A meta-analysis pooling data across studies of three major psychiatric disorders found that all three were associated with a brain that looks older than expected. Schizophrenia showed the largest average brain-age gap at about three years, followed by bipolar disorder at roughly two years and major depressive disorder at just over one year.8PubMed. Brain age in mood and psychotic disorders: a systematic review and meta-analysis
These gaps are averages across groups, not guarantees for any individual with a diagnosis. But they hint at something important: serious mental illness may not just affect mood and cognition in the moment. It may leave a structural footprint on the brain that resembles premature aging. Whether this is driven by the illness itself, the medications used to treat it, the lifestyle disruptions that often accompany it, or some combination remains an active area of investigation.
Cardiovascular and Metabolic Health
Your heart and your metabolism have a surprisingly direct relationship with your brain age. People with metabolic syndrome, the cluster of conditions that includes high blood pressure, high blood sugar, excess abdominal fat, and abnormal cholesterol, show a measurably higher brain-age gap compared to those without it. One large study found the difference was significant even after controlling for confounding factors, and identified specific blood metabolites, including inflammatory markers and certain fatty acids, that partially explained the connection.9PubMed Central. Metabolic syndrome is associated with accelerated brain aging
Type 2 diabetes and high blood pressure each independently contribute to the effect, and there appear to be differences by sex. A multi-ethnic study found that diabetes and hypertension were associated with a higher brain-age gap regardless of sex and ethnicity, with some evidence that the effect was larger in women for certain conditions.10PubMed Central. Different effects of cardiometabolic syndrome on brain age in relation to gender and ethnicity And when cardiovascular and metabolic problems coexist with serious mental illness, the effects on brain structure appear to compound. Both cardiometabolic disorders and serious mental illness independently accelerated brain aging in one analysis, and hypertension had a measurable impact in both groups.11PubMed Central. The additive impact of cardio-metabolic disorders and psychiatric illnesses on accelerated brain aging
The upshot is that managing blood pressure, blood sugar, and cholesterol may protect not just your heart and kidneys but your brain’s structural integrity over time.
Exercise Can Make Your Brain Look Younger
If cardiovascular health affects brain age, it makes sense that exercise might too. A randomized trial tested this directly. Participants who completed a 12-month aerobic exercise program showed a decrease in their brain-age gap of about 0.6 years on average, while a control group drifted slightly upward. The difference between the two groups was about one year after adjusting for other factors. At the start of the study, people with higher cardiorespiratory fitness already had a lower brain-age gap: for every meaningful increase in peak oxygen uptake, brain age dropped by roughly 1.8 years.12PubMed Central. Fitness and exercise effects on brain age: A randomized clinical trial
One year of structured exercise producing roughly a one-year improvement in brain age is a modest but real effect, and it is notable because it came from a randomized trial rather than just observational data. Most brain-age research is cross-sectional, capturing a single snapshot, so this kind of before-and-after evidence is especially valuable.
Sleep, Diet, and Other Lifestyle Factors
Poor sleep quality is consistently linked to a higher brain-age gap. One study using both structural and white matter scans found that worse self-reported sleep quality was associated with a brain that looked about two years older than expected.13PubMed Central. The association between inadequate sleep and accelerated brain ageing Sleep apnea pushes the effect further. In a large population study, the severity of sleep apnea, measured by disrupted breathing events and drops in blood oxygen, was positively associated with a higher brain-age gap.14SLEEP. Associations between sleep apnea and advanced brain aging in a large-scale population study In more severe subtypes of obstructive sleep apnea, the brain-age increase reached over a decade.15PubMed. Obstructive sleep apnea subtyping based on apnea and hypopnea specific hypoxic burden is associated with brain aging and cardiometabolic syndrome
Diet also appears to matter, though the evidence here comes more from structural brain measures than from brain-age models directly. Higher adherence to a Mediterranean-style diet has been associated with larger total brain volume, more gray matter, and more white matter in an older multiethnic cohort. Higher fish intake and lower meat consumption were the components most strongly linked to these differences.16PubMed Central. Mediterranean diet and brain structure in a multiethnic elderly cohort Since brain-age algorithms are largely built on these same volumetric features, it stands to reason that diets protecting brain volume would also produce a more favorable brain-age gap, though direct studies testing this link are still emerging.
Socioeconomic Stress and Brain Aging
Where you fall on the socioeconomic ladder may shape your brain age in ways that go beyond the individual lifestyle choices typically studied. Lower socioeconomic status has been independently associated with premature brain aging, even after accounting for other risk factors.17PubMed Central. Lower socioeconomic status is associated with premature brain aging The mechanisms are likely tangled together: people with fewer resources tend to have less access to healthy food, safe places to exercise, and quality healthcare, and they tend to experience more chronic stress. Chronic stress itself drives inflammation and hormonal changes that can accelerate biological aging across multiple organ systems, including the brain.
This finding matters because it reframes brain age as something shaped not just by personal habits but by systemic conditions. A person with an elevated brain-age gap may be experiencing the cumulative toll of disadvantage rather than any single modifiable behavior.
Genetics and Early-Life Factors
Your brain-age gap is not entirely under your control. Genetic studies have found that brain age is significantly heritable. In a large genome-wide analysis, brain-age gaps derived from gray matter structure showed a heritability of about 47%, and white matter-based brain age was similar at around 46%. Functional connectivity-based brain age was much less heritable, at about 11%.18PubMed Central. The genetic architecture of multimodal human brain age Sixteen genomic regions reached significance across the three measures, with different cell types enriched depending on the modality. Oligodendrocytes were implicated in white matter brain age, and astrocytes in functional connectivity brain age.
Perhaps even more provocatively, a longitudinal study found that brain age in adulthood was not associated with the rate of actual brain change measured over time. Instead, it was more strongly associated with factors present from birth, including birth weight and genetic predisposition. In other words, much of what makes your brain look “older” or “younger” than expected may be baked in early, reflecting how your brain developed rather than how fast it is declining.19eLife. Individual variations in ‘brain age’ relate to early-life factors more than to longitudinal brain change This is an important nuance: a high brain-age gap does not necessarily mean your brain is deteriorating faster right now. It may mean your brain started from a structurally different baseline.
How Brain Age Relates to Epigenetic Aging
Scientists have other biological clocks beyond brain imaging. Epigenetic clocks estimate biological age from chemical marks on your DNA, typically measured from a blood sample. A natural question is whether these two clocks agree. The answer is: only partially. One study found that a specific epigenetic clock, DNAmPhenoAge, mediated the relationship between calendar age and brain age, and that epigenetic aging was particularly linked to accelerated aging in higher-order brain regions involved in complex thought.20PubMed Central. Epigenetic age is associated with regional brain aging along the sensorimotor-to-association axis of cortical organization
But another study using a different epigenetic clock found no association with brain age at all. Instead, one smoking-related epigenetic marker was linked to volume loss in the frontal and temporal lobes but not in the hippocampus or other areas typically hit earliest by Alzheimer’s.21PubMed Central. Association of epigenetic age acceleration with MRI biomarkers of aging and Alzheimer’s disease neurodegeneration The takeaway is that brain age and epigenetic age appear to capture different aspects of biological aging. They are complementary signals, not interchangeable ones. If both are elevated in the same person, that may carry more clinical weight than either alone, though this idea has not yet been rigorously tested.
Why Your Doctor Cannot Order a Brain-Age Test Yet
For all the promising associations, brain age is not ready for clinical use at the individual level. The biggest problem is generalizability. Different software packages for estimating brain age correlate with each other only moderately, and the predictions are sensitive to which MRI scanner was used, how the scan was acquired, and the demographics of the population the model was trained on.22PubMed Central. Benchmarking the generalizability of brain age models: Challenges posed by scanner variance and prediction bias
There is also a well-documented statistical bias. Current models systematically overestimate brain age in younger people and underestimate it in older people, pulling predictions toward the middle of the age range. This means that an older adult with genuine neurodegeneration can still get a brain-age estimate that looks “normal” or even young for their age, simply because the model is dragging their prediction downward. A recent multi-cohort evaluation concluded bluntly that current models are not suitable for interpreting a single individual’s brain-age gap in isolation.23PubMed Central. Bias and generalizability of brain age prediction models: A multi-cohort evaluation with anatomical and interpretability insights Another analysis reinforced this, finding limited effectiveness of the brain-age gap as an individual risk marker and noting that the degree of disease-related brain aging varied across the age continuum.24PubMed. Distribution Bias in Brain Age Research: Toward Age-Specific Interpretation of Brain Age Gaps
In group-level research, these biases can be statistically adjusted. In a clinical setting, where a doctor is looking at one patient’s scan and trying to decide if it warrants concern, they cannot yet be reliably corrected. Brain age works well as a research tool for identifying population-level associations. It does not yet work as a diagnostic tool for an individual sitting in the office.
Senolytics and the Future of Intervention
While lifestyle factors like exercise and sleep management represent current actionable strategies, pharmaceutical research is beginning to explore whether drugs could directly slow or reverse brain aging at the cellular level. One emerging class of compounds is senolytics, which target and destroy senescent cells. These are cells that have stopped dividing and accumulate with age, secreting inflammatory molecules that damage surrounding tissue. In the brain, senescent cells have been linked to the buildup of toxic protein aggregates and oxidative stress. Compounds like dasatinib, fisetin, and quercetin have shown the ability to clear senescent cells in preclinical work, reducing these harmful accumulations.25Molecular Neurobiology. Senolytics as Modulators of Critical Signaling Pathways: a Promising Strategy to Combat Brain Aging and Neurodegenerative Disorders Whether this translates into measurable changes on a brain-age scan in humans remains to be seen. Clinical trials are underway, but we are years away from knowing whether popping a senolytic pill could genuinely make your brain look younger on a scan, let alone function better in daily life.