Brain age

Brain age is an estimate of how old a person’s brain appears to be biologically, derived from neuroimaging features by machine-learning models trained on chronological age in healthy reference cohorts. The clinically meaningful quantity is not the raw brain-age estimate but the brain-age gap (also called brain-PAD, predicted age difference) — the residual between predicted age and the person’s actual chronological age. A positive gap indicates the brain looks older than expected for the person’s chronological age; a negative gap, younger. The gap is interpreted as an integrative biomarker of accumulated brain change across diverse mechanisms.

The framework was formalized by Christian Gaser and Katja Franke and synthesized in their canonical review (Franke & Gaser 2019). Their BrainAGE algorithm uses voxel-based morphometry from structural MRI as input to a relevance-vector-machine regression on age. Subsequent work by James Cole and colleagues showed that brain age predicts mortality (Cole et al. 2018). The metric has been extended to diffusion MRI, functional MRI, and EEG, and consortium-level efforts (ENIGMA-Brain Age) have addressed cross-site harmonization. Sensitivity to disorders including Alzheimer’s, schizophrenia, and HIV-associated neurocognitive disorder is well-documented.

Three points are routinely missed in popular treatments. First, brain age is a biomarker, not a fixed personal property: different ML pipelines applied to the same person can produce systematically different age estimates depending on training data, features, and regression model. Second, the brain-age gap is most useful for group-level epidemiology; individual-level interpretation requires careful validation and is currently more research tool than clinical instrument. Third, brain age is a summary biomarker — a positive gap indicates “something is accumulating” but does not, by itself, identify which specific pathology is driving the signal.

Most people use chronological age as a proxy for cognitive trajectory: at 50, they think about retirement timing differently than at 30; at 70, they think about cognitive reserve differently than at 50. But chronological age is a blunt instrument. Two people of the same chronological age can have brains that differ by a decade or more in biological state. A 2025 analysis of 38,967 UK Biobank participants found brain age gap predicts cognitive decline, neuropsychiatric disorders, and mortality independent of chronological age (Zhang et al., 2025).

Knowing whether a person's brain is aging faster or slower than expected can inform real-life decisions: when to prioritize sleep and exercise, whether to invest in cognitive training, when to undertake major life transitions like a career change or a move, how aggressively to address midlife cardiovascular risk, and how to plan financially for cognitive aging that may arrive earlier or later than expected. Decision-making under this kind of biological uncertainty is exactly where decision hygiene and scenario planning become useful.

Researchers estimate brain age in two complementary ways.

From MRI scans. The seminal approach, introduced by Katja Franke and Christian Gaser and developed further by James Cole and colleagues at King's College London and University College London, trains machine-learning models on tens of thousands of structural MRI scans from healthy individuals across the lifespan. The model learns the typical pattern of brain structure at each age — cortical thickness, grey matter volume, white matter integrity. Applied to a new scan, the model returns a predicted age. The difference between predicted age and the person's actual chronological age is the brain age gap (Papouli & Cole, 2025).

From validated risk factors. A complementary approach asks: given what we know about modifiable factors that accelerate or delay brain aging, what would a person's expected brain age be? The 2024 Lancet Commission on dementia prevention identified fourteen modifiable risk factors that, in aggregate, account for approximately 45% of dementia cases globally (Livingston et al., 2024). Risk-factor-based estimates use answers about these factors to adjust an estimate from chronological age; how many years each factor adds is the estimator's own calibration, because the Commission's population-attributable fractions are population figures, not personal year values. This approach does not require an MRI scan and suits self-report tools.

Both are expressed as a year-equivalent, but they capture different signals. MRI captures structural change that has already occurred. Risk-factor models capture exposure to factors that, on average, drive structural change. The two correlate but are not interchangeable.

The fourteen modifiable factors identified by the Lancet Commission group into five clusters. Each cluster contributes to brain aging through partially distinct biological pathways, which is why a multi-domain approach to brain health tends to outperform single-factor interventions.

• Cardiovascular and metabolic factors. Hypertension, diabetes, obesity, high LDL cholesterol, and physical inactivity. These accelerate brain aging primarily through vascular damage and metabolic dysregulation. Managing them in midlife appears to confer the largest protective effect.
• Sensory and physical factors. Hearing loss, vision loss, and traumatic brain injury. Sensory loss accelerates cognitive decline through reduced environmental engagement; correction with hearing aids and glasses appears to attenuate the effect.
• Cognitive engagement. Education early in life, complex work, and ongoing cognitive challenge. These contribute to cognitive reserve — the brain's resilience to pathology that has already accumulated.
• Social and behavioral factors. Social isolation, depression, smoking, and excessive alcohol consumption. Each has been independently linked to accelerated brain aging in longitudinal cohorts.
• Environmental factors. Air pollution, particularly fine particulate matter (PM2.5). Exposure averaged over decades is associated with measurable structural change.

A 2025 conference abstract that tested the Lancet model in 5,526 participants of the Wisconsin Longitudinal Study, a cohort followed since 1957, found that only a restricted subset of the modifiable risk factors was associated with late-life dementia status; the analysis was also stratified by APOE4 carrier status (Williams et al., 2025).

What it can tell you. Brain age is best understood as a summary risk indicator: a single number that condenses many small contributions to brain health into something interpretable. It is associated, in research populations, with future risk of cognitive decline, dementia diagnosis, and mortality. Longitudinal change in brain age — the trajectory over time — appears to be more informative than any single measurement (Lin et al., 2025). Crucially, the metric appears modifiable: in one 2025 analysis, smoking cessation, moderate alcohol consumption and physical activity were linked to slower brain-age-gap progression in people with advanced neurodegeneration (Zhang et al., 2025).

What it can't tell you. Brain age is not a diagnosis. A higher-than-expected brain age does not mean a person currently has a cognitive disorder, and a lower-than-expected brain age does not guarantee cognitive resilience in the future. The metric is a population-level statistical association applied to an individual, which means uncertainty bands are wide. Different estimation methods can produce different brain ages for the same person. And the metric does not capture every relevant factor — genetics, accumulated trauma, and conditions that have not yet been validated as risk factors are not necessarily reflected.

"My brain age determines my cognitive ability." It does not. Brain age is a risk indicator, not a measure of current cognitive performance. Many people with higher-than-expected brain ages perform cognitively well; many with lower-than-expected brain ages will still develop cognitive decline.

"Brain age is fixed." It is not. Research suggests that lifestyle and clinical factors can slow the trajectory. Neuroplasticity persists across the lifespan in modified form, which is the biological substrate of this modifiability.

"Self-report brain age is the same as MRI-based brain age." They are related but not interchangeable. They measure overlapping constructs through different signals and have different error profiles. Neither replaces the other.

"A small brain age gap is meaningless." At the individual level, gaps of less than ±3 years sit within typical estimator error and should be interpreted cautiously. At the population level, even small shifts can predict meaningful differences in long-run risk.

Consider a 55-year-old considering whether to take an early retirement package or work another decade. Her chronological age tells her she is at the typical pre-retirement decision point; her brain age, estimated at 49, suggests her cognitive trajectory is favorable. This does not by itself answer the retirement question — financial, family, and meaning considerations dominate — but it does change the weighting on a specific scenario: she is, on average, well-positioned for cognitively demanding work in her sixties, which makes "stay and pursue advancement" more viable than chronological age alone would suggest.

Now suppose her brain age is estimated at 62. The same financial and family considerations apply, but the weighting shifts: pursuing a less cognitively demanding role becomes more attractive, as does prioritizing sleep, exercise, and cardiovascular health regardless of which path she chooses. Brain age does not make the decision; it adjusts the probability of different outcomes under each option.

• Cognitive reserve — the brain's capacity to maintain function despite accumulated pathology. Brain age captures the pathology; cognitive reserve captures the resilience to it.
• Neuroplasticity — the brain's capacity to reorganize structure and function. The biological mechanism that makes brain age modifiable.
• Biological age — a broader concept covering the biological state of the body as a whole, of which brain age is one component. Other components include epigenetic age, telomere length, and inflammatory markers.
• Executive function — the cognitive processes that govern planning, decision-making, and self-regulation. Often the first cognitive domain affected as brain age accelerates.
• Risk prediction — the broader statistical practice of using biomarkers to estimate future health outcomes. Brain age is one of many candidate biomarkers in this space.
• Effect size — the magnitude of brain age's association with outcomes. Larger effect sizes mean stronger predictive value at the individual level.
• Chronotype — circadian preference, one of several lifestyle factors that interact with brain age trajectories.
• Neurogenesis — adult hippocampal neuron generation; one of the substrate processes that may slow brain aging in healthy adults, though the magnitude and persistence are contested (Boldrini 2018 vs Sorrells 2018).
• Default mode network — the canonical resting-state network; DMN integrity and dynamics are among the candidate biomarkers in brain-age estimation models.

The LifeByLogic Brain Age Index implements a self-report risk-factor-weighted estimate built around the 2024 Lancet Commission's potentially modifiable risk factors. Its 25 questions, 23 of them grouped into six domains, add or subtract a conservatively scaled number of years from chronological age; those year values are the tool's own calibration, not the Commission's population-attributable fractions. The result is shown in whole years and is best read as approximate. The full methodology — variable list, scoring algorithm, and limitations — is documented on the Brain Age Index methodology page.

Calculate your Brain Age Index →

This entry is intended as a citable scholarly reference. Choose the format that matches your context. The retrieval date should reflect when you accessed the page, which may differ from the entry's last-reviewed date shown above.

LifeByLogic. (2026). Brain Age: Definition and How MRI Estimates It. https://lifebylogic.com/glossary/brain-age/

LifeByLogic. "Brain Age: Definition and How MRI Estimates It." LifeByLogic, 25 Oct. 2026, https://lifebylogic.com/glossary/brain-age/.

LifeByLogic. 2026. "Brain Age: Definition and How MRI Estimates It." October 25. https://lifebylogic.com/glossary/brain-age/.

@misc{lblbrainage2026,
  author = {{LifeByLogic}},
  title = {Brain Age: Definition and How MRI Estimates It},
  year = {2026},
  month = {oct},
  publisher = {LifeByLogic},
  url = {https://lifebylogic.com/glossary/brain-age/},
  note = {Accessed: 2026-10-25}
}