Brain-Age Scores, APOE and the Limits of Personalised Brain-Health Advice

Text written with the help of AI.

Brain-age scores and genetic findings can make brain-health advice feel personal. The harder question is whether using that information to make a different choice improves an outcome you care about. A change in a score, an association within a genetic group and a demonstrated health benefit are three different findings.

What a younger brain-age score measures

BrainYears estimates age from electrical brain signals and task-related measurements. A March 2026 preprint reported an average reduction of 5.18 years in predicted brain age after a neuromodulation programme; a minimal-exposure comparison group showed almost no average change. Lore and colleagues, 2026.

The result is emerging evidence about this model’s output. The intervention comparison was not randomised, and the authors acknowledged uncontrolled factors including sleep, medication changes and placebo effects. Two authors were affiliated with the device developer. The study did not establish that participants gained five years of independent living, reduced their dementia risk or reversed Alzheimer’s disease. Testing whether the score predicts durable changes in independent cognitive assessments and everyday function would make it more useful for decisions.

What an APOE association permits you to conclude

A 2026 study followed 2,157 older Swedish adults and examined meat consumption alongside APOE genotype. Among participants with APOE ε3/ε4 or ε4/ε4, those in the highest fifth of total meat consumption had a dementia subdistribution hazard ratio of 0.45 compared with those in the lowest fifth. That relative measure is not an individual prediction, and the reported outcome was dementia, not a demonstrated reduction specifically in Alzheimer’s disease. Norgren and colleagues, 2026.

Participants were observed, not assigned diets. The comparison concerned total meat intake, rather than a tested serving of one food. There is a further limit to the personalised interpretation: the test of whether the association differed by genotype was significant for cognitive decline, but did not meet the conventional threshold for dementia incidence (P = 0.10). Participants averaged about 71 years old, which also limits direct application to someone in their forties.

The finding may help researchers frame a dietary trial. It does not establish that people with these genotypes should eat more meat or that meat offsets inherited risk.

What intervention trials add

Randomised trials can test whether an intervention changes an outcome, though their conclusions still depend on who took part and what they were compared with. In a 2023 trial, 604 older adults received either a MIND diet with mild calorie restriction or a control diet with mild calorie restriction. Both groups received support. After three years, cognitive and brain MRI changes did not differ significantly between groups. That finding concerns the additional benefit of this programme over an active comparison; it does not show that food is irrelevant to brain health. Barnes and colleagues, 2023.

The 2025 US POINTER trial enrolled 2,111 adults aged 60 to 79 at increased risk of cognitive decline. A structured programme combining exercise, dietary change, cognitive and social activity, and cardiovascular monitoring produced greater improvement in global cognitive scores than a less intensive, self-guided programme. The difference was 0.029 standard deviations per year over two years. It was a cognitive test result, not proof of dementia prevention, and the trial cannot identify which component produced the difference. US POINTER investigators, 2025.

Blood-pressure treatment offers another example of why the outcome matters. SPRINT enrolled adults aged 50 and older with hypertension and elevated cardiovascular risk, excluding people with diabetes or a previous stroke. Its extended follow-up found fewer combined cases of mild cognitive impairment or probable dementia after intensive treatment, while the result for probable dementia alone remained inconclusive. This supports a discussion about blood-pressure management with a clinician; it does not provide a universal target or a reason to adjust medication independently. SPRINT investigators, 2025.

These trials provide evidence about the interventions and outcomes they tested. None establishes that choosing a diet or treatment according to a BrainYears score or APOE result improves brain health.

Decide what the information would change

Before paying for a personalised brain-health result, identify the decision it would alter and the benefit you expect from that decision. Then look for research that tests the action and measures that benefit. If the evidence ends at a model score or an observational subgroup, waiting for stronger evidence is reasonable.

Personal observations can help you judge whether a change fits your life or seems to affect your day-to-day experience. They cannot establish that your brain is ageing more slowly or that your future dementia risk has fallen.

Noticeable changes in memory or daily function call for assessment rather than interpretation through a consumer score. The US National Institute on Aging identifies repeated questions, getting lost in familiar places and difficulty following familiar instructions as reasons to seek help. Sleep problems, medication effects and other conditions can also affect memory. National Institute on Aging.

Sources