Alzheimer's Disease: Prevention and Risk Reduction: Modifiable Lifespan Risk Factors, Cardiometabolic Protection, Sleep Clearance, and Cognitive Reserve
A preventive neurology reference on the modifiable risk factors identified by the Lancet Commission, cardiometabolic and vascular control, exercise, diet and sleep, and building cognitive reserve.
Submitted research brief - Submitted to OwieClear; written with an AI assistant - November 2026
Read this first. This document was written outside OwieClear and has not been reviewed by a clinician. We publish it so you can read the source for yourself. It is educational and is not medical advice. Where our own pages disagree with it, follow our pages, which cite their sources, and ask a healthcare professional about anything that affects your care.
What we found that needs correcting
- This brief is built on the right source, the Lancet Commission, and it covers the risk factors accurately. The notes below are mostly about certainty: several of these findings are presented more confidently than the trials support, and in two cases the trial result points the other way.
- The MIND diet is the clearest correction. A randomized trial published in 2023 compared the MIND diet with a mild calorie-restricted control over three years and found no significant difference in cognitive change between the groups. The observational evidence that generated the enthusiasm did not survive its test. The diet remains a reasonable way to eat for cardiovascular health, which does matter for the brain, and it should not be presented as an established dementia-prevention intervention.
- Hearing aids are overstated. The brief says correcting hearing loss significantly reduces cognitive decline. The large trial that tested this found no overall effect across all participants, with a meaningful effect only in a subgroup already at higher risk. Hearing aids are clearly worth having for hearing, communication and mood; the dementia claim is not established.
- The 45 percent figure needs its definition. It is a population attributable fraction, an estimate of how much dementia might not occur if these factors were eliminated across an entire population. It is not an individual risk reduction, and a reader who does everything on the list does not thereby lower their own risk by 45 percent.
- Reverse causation is not acknowledged anywhere, and it is the main weakness of this evidence base. Dementia pathology accumulates for 15 to 20 years before diagnosis, and during that time people become less physically active, less socially engaged and less likely to pursue hearing correction. That makes inactivity and isolation look more causal than they may be. The brief's own opening paragraph about the silent accumulation period sets up this problem without noticing it.
- The exercise figure of 30 to 40 percent risk reduction comes from observational data subject to exactly that problem. Exercise is worth recommending on overwhelming general health grounds; the specific percentage should be presented as an association rather than an effect.
- Type 3 diabetes should be dropped or flagged. It is an informal term that appears in the literature as a metaphor for brain insulin resistance. It is not a recognized diagnosis, and readers encountering it are likely to think it is one.
- The vaccine section is appropriately hedged with the word emerging, and one addition would strengthen it. These are observational findings subject to healthy-user bias: people who keep up with vaccination differ systematically from people who do not. The strongest evidence comes from a natural experiment in Wales where shingles vaccine eligibility depended on date of birth, which is unusually good for this kind of question and still not a trial.
- APOE4 is listed as unchangeable and then left there. Two things readers want are missing: most people carrying APOE4 never develop Alzheimer's, and routine testing is not recommended in people without symptoms because the result does not change what anyone should do. It has become relevant for one clinical reason, which the brief does not mention: APOE4 carriers have a higher risk of brain swelling and bleeding on the new anti-amyloid treatments, so testing is done before those drugs are started.
- The alcohol guidance uses UK units, which US readers will not recognize, and the framing has moved. Current evidence does not support the idea that moderate drinking protects the brain, and the direction of advice is toward less rather than toward a safe threshold.
- No trial has shown that any single intervention prevents dementia, and that belongs in a prevention document. The closest evidence is a large blood pressure trial that found less mild cognitive impairment with intensive control, and multidomain programmes that have shown modest cognitive benefits rather than prevention. This does not undermine the advice; it describes its strength accurately.
- One omission worth adding for older readers: falls. Falls are the leading cause of traumatic brain injury in older adults, they cause hospital admission and delirium, and fall prevention is concrete, effective and immediately actionable in a way that most items on this list are not.
- Everything else in the brief is accurate: the Lancet Commission framing and the 15 to 20 year silent accumulation period; the what-protects-the-heart-protects-the-brain principle; the mechanisms linking midlife hypertension to microvascular damage, blood-brain barrier disruption and white matter lesions; the roles of LDL, insulin resistance and smoking; the glymphatic clearance of amyloid during slow-wave sleep and the importance of identifying and treating obstructive sleep apnoea; cognitive and social reserve and the value of treating depression; traumatic brain injury and air pollution as risks; and the practical measures recommended throughout.
Download the PDF
Our reviewed pages on this
All source documents