Brain Health

Up to 45% of dementia risk is attributable to modifiable factors in the 2024 Lancet Commission framework.^[1] The single most important framing shift in modern dementia prevention is that the brain is downstream of cardiovascular and metabolic health — not a separate organ. What's good for the heart is overwhelmingly good for the brain.

The corpus has converged on a small set of high-leverage actions: treat hearing loss early; treat blood pressure aggressively in midlife; protect vision; defend cardiovascular and metabolic health; stay socially engaged and cognitively challenged (not merely cognitively occupied). The interventions that don't work — commercial "brain training" apps, isolated cognitive supplements — have also been clearly mapped.

What the evidence actually supports

Strong (the 14 modifiable factors from the 2024 Lancet Commission, ranked roughly by population attributable fraction):

1. Hearing loss — the largest single modifiable factor, with a population attributable fraction of about 7%. The 2023 ACHIEVE trial showed hearing aids cut the rate of cognitive decline 48% in high-cardiovascular-risk older adults,^[2] and the 2026 ASPREE target-trial emulation found hearing-aid prescription cut 7-year incident dementia from 7.5% to 5.0% (relative risk 0.67), with a clean dose-response by frequency of use.^[3] See Hearing.
2. Hypertension — treat aggressively, especially in midlife. SPRINT-MIND found intensive blood-pressure control (target <120 mmHg systolic) reduced incident mild cognitive impairment and dementia versus standard control.^[4]
3. Smoking — quit at any age; cognitive trajectory improves.
4. Obesity — particularly visceral; metabolic syndrome strongly predicts late-life dementia.
5. Type 2 diabetes — well-controlled diabetes is qualitatively different from uncontrolled for cognitive outcomes.
6. Physical inactivity — exercise is the single strongest behavioural lever; aerobic and resistance both contribute, including measurable hippocampal-volume gains.^[5]
7. Depression — both a risk factor and a prodrome; treat regardless.
8. Social isolation — strong cohort signal; the "loneliness epidemic" is a measurable risk.
9. Lower educational attainment (early life) — cognitive reserve.
10. Excessive alcohol — even moderate use is now considered risk-elevating; see Alcohol.
11. Air pollution (PM2.5) — chronic exposure raises dementia incidence.
12. Traumatic brain injury — repeated subconcussive impacts (contact sports) carry signals.
13. Visual impairment (newly added 2024) — uncorrected visual loss compounds isolation and reduces cognitive stimulation. See Vision.
14. High LDL cholesterol (newly added 2024) — vascular contribution.

Moderate:

• MIND-diet adherence — the observational signal is large (~53% reduction in Alzheimer's at highest tertile in Morris 2015); the MIND-USDA RCT (2023) was less impressive but still favourable.
• Sleep architecture. Slow-wave (deep) sleep clears beta-amyloid via the brain's glymphatic system. A Framingham follow-up found each percentage-point yearly decline in slow-wave sleep was associated with about 27% higher dementia risk over 17 years. See Sleep.
• Cognitive engagement — productive, novel skill acquisition (languages, music, complex crafts, demanding occupational complexity) builds cognitive reserve and delays functional dementia onset. The ACTIVE trial's speed-of-processing arm cut 20-year dementia incidence by about 25% in older adults.^[6] See Cognitive engagement.
• Ultra-processed food. A 2026 BMJ systematic review found ultra-processed-food intake is associated with accelerated cognitive decline and higher dementia risk even after adjusting for adherence to a healthy dietary pattern — a processing-specific effect on the brain on top of nutrient quality.^[7]

Weak / preliminary:

• Specific supplements for cognition (omega-3, B vitamins, ginkgo, curcumin) — most show null or inconsistent results in randomised trials with cognitive endpoints. The exception is the DO-HEALTH signal that omega-3 and vitamin D combined with exercise slow biological-age clocks, but cognitive endpoints there are surrogates, not dementia.
• Commercial "brain training" apps — narrow within-task gains that don't generalise.

Why the brain is downstream of the heart

The single most consequential idea in modern dementia prevention is that the brain is not a separate organ. The same vascular and metabolic biology that drives heart attacks, strokes, and type 2 diabetes drives the slow grind of cognitive decline. The cochlea is one of the most metabolically demanding microvascular structures in the body — which is why hearing loss tracks the same biology as cardiovascular disease and now scores as a biomarker of cochlear microvascular health as much as an isolated sensory deficit.^[8] The retinal microvasculature is the only vascular bed visible non-invasively, which is why deep-learning analyses of fundus photographs now generate Retinal Age Gap scores that rival blood-based aging clocks for predicting all-cause mortality.^[9]

The implication is that the highest-leverage brain interventions are not brain-specific. Treating hypertension, defending cardiorespiratory fitness, sleeping regularly, eating a vegetable-and-fish-heavy dietary pattern, and not smoking dominate any cognitive supplement on offer. The brain-specific actions that do matter — treating hearing loss, treating correctable visual loss, building cognitive reserve through demanding novel skill acquisition — are not substitutes for cardiovascular care; they're additive on top of it. The integrative case is collected under Dementia prevention, built around the FINGER multi-domain trial,^[10] the SPRINT-MIND blood-pressure result, and the Lancet Commission's continuously evolving 14-factor framework.

Hearing: the largest single modifiable midlife dementia lever

Hearing loss has been quietly reclassified over the last decade. It is no longer an isolated sensory deficit — it is a systemic biomarker of cochlear microvascular health, a measurable driver of accelerated brain atrophy, and the largest single modifiable midlife dementia risk factor, accounting for an estimated 7% of global dementia cases on its own.^[11] Two large trials in 2023–2026 have converted the observational signal into a near-causal one. The ACHIEVE trial randomised 977 older adults at high cardiovascular risk to hearing intervention versus health education and found a 48% reduction in the rate of cognitive decline at three years in the high-risk arm.^[12] The 2026 ASPREE target-trial emulation in 2,777 adults aged 70+ with moderate hearing loss, dementia-free at baseline, found 7-year dementia risk of 5.0% with hearing aids versus 7.5% without (relative risk 0.67), with a clean dose-response by frequency of use.^[13]

The systemic signal is broader than dementia. A 2025 meta-analysis put the pooled hazard ratio at about 1.21 for all-cause mortality, 1.22 for cardiovascular mortality, and 1.11 for cancer mortality, with audiometrically measured loss (1.28) carrying a stronger signal than self-reported loss — the audiogram is picking up sub-clinical physiology people don't notice.^[14] Hearing-aid use itself is independently associated with lower long-term mortality in a NHANES cohort of 9,885 adults followed for a median 10.4 years.^[15]

The diagnostic shift matters too. Standard pure-tone audiometry caps at 8 kHz, but extended high-frequency (EHF) audiometry up to 16–20 kHz captures damage years before standard audiometry shows it — about 64% of age-related variance in real-world speech-in-noise understanding, versus the standard test's 16%. If you struggle in restaurants but pass a standard audiogram, EHF audiometry is the test to ask for.

Hearing covers the ACHIEVE and ASPREE trial designs, the EHF audiometry case, the noise × air-pollution synergy, why hearing aids are not vanity, the cochlear microvascular biology, and the diabetes-and-hearing crosswalk.

Vision: the eye as a systemic biomarker

Vision has been quietly reclassified along the same arc. The 2024 Lancet Commission added visual impairment to the modifiable dementia list,^[16] and the past three years have produced an oculomics literature in which deep-learning models read fundus photographs to predict cardiovascular events, biological age, and all-cause mortality with a precision that rivals or exceeds blood-based aging clocks.^[17] UK Biobank cohort analysis of 35,913 adults found each one-year increase in the Retinal Age Gap — the difference between AI-predicted retinal age and chronological age — associated with about 2% higher all-cause mortality, with steeper increases in the upper quartiles.^[18] The RetiAGE framework predicts 10-year cause-specific mortality from fundus images alone, with the top-versus-bottom quartile carrying roughly 142% higher cardiovascular mortality and 60% higher cancer mortality.^[19] A separate LAVA framework reports AUROC 0.93 for Alzheimer's diagnosis from a single retinal photograph.

The dementia link is largely mediated by functional vision loss rather than shared pathology — blindness substantially raises Alzheimer's risk while non-exudative age-related macular degeneration (AMD) without severe impairment does not.^[20] The practical implication is identical to hearing: treat the correctable deficit early. Cataract surgery, glasses, low-vision rehabilitation, and AMD management protect the sensory bandwidth that maintains cognitive engagement.

The mechanism extends beyond the retina. Macular pigment — lutein and zeaxanthin concentrated in the fovea — provides photoprotection and improves visual performance; leafy-green-heavy diets raise it. The 670 nm red-light finding from Glen Jeffery's group is interesting (brief deep-red light exposure restored some age-related macular cone-cell function in small trials) but still preliminary at home-device scale.

Vision covers the Retinal Age Gap models in detail, the lutein and zeaxanthin macular-pigment evidence, OCT and OCT-angiography for early detection, the 670 nm red-light data and its limits, the AMD-and-dementia mediation, and why blue-light-blocking glasses do not move sleep or eye-strain endpoints in trials.

Cognitive engagement: build reserve, don't waste it on apps

The most useful framing in this area is brainspan — the years of life during which neural networks remain efficient enough to support agency, autonomy, and coherent behaviour. It's narrower than healthspan: a body can function physically while the brain has already lost the cognitive bandwidth that defines a person.^[21] The mechanism behind the modifiable-factors signal is the cognitive-reserve hypothesis: autopsy studies routinely find brains with extensive Alzheimer's-type pathology in people who showed no clinical symptoms during life, and minimal pathology in people who clearly had dementia. The brain has redundant capacity that mediates between damage and observable decline, and lifelong intellectual challenge builds that redundancy; activities that don't challenge the brain don't.^[22] About 28% of the dementia-protective effect of early-life education is mediated by the complexity of the adult job it enables; the late-life share is just as large as the early-life share — what a 60-year-old does in retirement has effect sizes comparable to whether they went to college.^[23]

The bad news side is firm. In 2016 the US Federal Trade Commission fined Lumosity $2 million for deceptive advertising, finding the company lacked competent evidence for its claims.^[24] A coalition of more than 70 cognitive scientists and neuroscientists, convened by the Stanford Center on Longevity and the Max Planck Institute for Human Development, issued a consensus statement explicitly rejecting the broad claims of the brain-training industry — there is no compelling scientific evidence that software-based brain games alter neural functioning in ways that improve general cognitive performance in everyday life or prevent brain disease.^[25] The underlying problem is the failure of "far transfer" — people reliably improve at the specific game they play, but the gains don't generalise.

The good news side is real. The ACTIVE trial randomised 2,802 cognitively healthy adults aged 65+ to one of four arms (memory, reasoning, visual speed-of-processing, or no-contact control), with just 10 sessions of 60–75 minutes over 5–6 weeks. Twenty years later, the speed-of-processing arm showed about 25% lower dementia incidence than control.^[26] The signal that works is demanding, novel skill acquisition — language, music, complex crafts, demanding occupational complexity, real-world analogue puzzles. Analog crosswords beat purpose-built computer games head-to-head; bilingualism delays functional dementia onset by up to 5 years.

Cognitive engagement covers the cognitive-reserve mechanism, the early/mid/late-life staging, the ACTIVE trial 20-year follow-up in detail, the brain-training-app evidence and consensus, the productive-versus-receptive engagement distinction, and the case for occupational complexity as a midlife lever.

Dementia prevention: the integrative pillar

Putting the 14 modifiable factors into practice is the job of Dementia prevention. The strongest single-trial signal for multi-domain lifestyle intervention came from FINGER — a Finnish randomised trial that combined diet, exercise, cognitive training, and vascular-risk monitoring in older adults at risk of cognitive impairment, with measurable cognitive benefit over two years.^[27] The follow-up programme U.S. POINTER and its international counterparts have extended that approach. The integrated message: cardiovascular risk reduction is the engine of cognitive protection, sensory and cognitive levers stack additively on top, and the actions are concrete enough to put on a checklist.

Dementia prevention covers the FINGER and U.S. POINTER protocols, the MIND-diet evidence in detail (the gap between Morris 2015's observational signal and the MIND-USDA RCT), the SPRINT-MIND blood-pressure result, an honest read on what cognitive supplements do and don't deliver, and the practical action list across all 14 modifiable factors.

Practical brain-health checklist

1. Treat hearing loss. If you suspect any hearing decline, get audiometry. Hearing aids may be one of the single most impactful interventions for cognitive aging. If you struggle in noisy environments but pass a standard audiogram, ask for extended high-frequency (EHF) audiometry. See Hearing.
2. Get a comprehensive dilated eye exam at midlife and treat any correctable visual loss. Cataract surgery, glasses, low-vision rehabilitation, and AMD management protect cognitive engagement. Ask about OCT and OCT-angiography. Eat leafy greens daily for macular lutein and zeaxanthin. See Vision.
3. Treat hypertension aggressively in midlife. Systolic <130 mmHg if tolerated; lower under clinical supervision. See Blood pressure.
4. Stay metabolically healthy. Avoid type 2 diabetes; treat insulin resistance.
5. Train both ways. Aerobic and resistance exercise both have independent effects on hippocampal volume and cognitive function. See Exercise.
6. Sleep well. 7–8 hours, regular schedule, screen for obstructive sleep apnea. See Sleep.
7. Eat MIND-pattern. Leafy greens daily, berries 2+ times a week, fish, nuts, olive oil, whole grains, beans, poultry. Limit red meat, butter and stick margarine, cheese, fried foods, sweets, and ultra-processed foods.
8. Stay socially connected. Real-world relationships, regular interaction. The strongest single behavioural predictor in cohort data.
9. Stay cognitively engaged — productively. Learning new languages, music, demanding crafts, deep semantic work. Passive consumption and casual socialising don't build cognitive reserve; novel skill acquisition does. See Cognitive engagement.
10. Don't smoke; drink minimally; reduce air-pollution exposure where feasible (HEPA filtration in homes near major roads is a small, plausible win).
11. Wear seatbelts and helmets; avoid repeated head impacts (contact sports).

What's overrated

• Commercial "brain training" apps (Lumosity, BrainHQ for general use, etc.) as substitutes for productive cognitive engagement. Narrow within-task gains; no far transfer.
• Cognitive supplements as standalone interventions (omega-3, B vitamins, ginkgo, curcumin). Most are null or inconsistent in trials with cognitive endpoints; the omega-3 epigenetic-age signal is interesting but uses surrogate outcomes.
• Blue-light-blocking glasses for sleep or eye strain. No movement on the relevant trial endpoints.
• "Detox" interventions for cognitive health. Cognitive function does not respond to detox protocols.
• Treating cognitive decline as inevitable above a certain age. Up to 45% of dementia risk is modifiable, and the late-life share is just as large as the early-life share.
• Single-supplement "anti-Alzheimer's" stacks. No combination tested has approached the effect size of treating hearing loss, hypertension, and physical inactivity simultaneously.