Modifiable dementia risk factors

Modifiable Risk Factors for Dementia: What Primary Care Can Measure

Primary care can measure four of the 14 modifiable dementia risk factors in the 2024 Lancet Commission with a number: blood pressure, glucose regulation, LDL cholesterol and adiposity. The others have to be asked about, and some were settled long before the visit.

A population estimate of preventable dementia does not tell a clinician what to order. Sorting the risk factors by how they reach the chart does.

Modifiable risk factors for dementia now number 14 in the 2024 Lancet Commission, which estimates that 45% of dementias worldwide are preventable if all 14 were eliminated. The standard visit measures perhaps four of them with a blood pressure cuff, a glucose and a lipid panel, and files the rest as yes-or-no history. That gap, not the list, is the primary care problem: which factors can be measured, which have to be asked about, and which were settled long before the patient reached the exam room. Measura [Cardiometabolic and Autonomic Health Analysis] addresses the measurable portion, from a standardized cognitive assessment baseline to the metabolic and vascular numbers behind it, and it helps to be exact about where that portion ends.

The 14 factors, sorted by how they reach the chart

The Commission’s list runs across the life course: less education early in life; hearing loss, high LDL cholesterol, depression, traumatic brain injury, physical inactivity, diabetes, smoking, hypertension, obesity and excessive alcohol use in midlife; and social isolation, air pollution and vision loss in late life. High LDL cholesterol and untreated vision loss were the additions in the 2024 report. For a clinic workflow they fall into three groups.

Measured with a number. Blood pressure, glucose regulation, LDL cholesterol and adiposity. This is where measurement can go past the default: fasting insulin beside glucose and hemoglobin A1c on laboratory panels, fat and muscle compartments on bioimpedance body composition instead of body mass index, and vessel wall behavior on arterial stiffness and endothelial function, which records what years of hypertension and hyperglycemia have done rather than today’s reading.
Screened with an instrument. Depression, hearing, vision, alcohol use and social isolation. Hearing and vision testing are done by audiology and eye care, not by Measura; validated questionnaires cover the rest.
Taken as history. Education, head injury and air pollution exposure. They cannot be changed at the visit, but they change how a borderline cognitive score should be read.

What the attributable fractions look like in real cohorts

The global figure pools very different populations, and cohort data narrow it. Norway’s HUNT study collected all 14 factors across adulthood in the same people and made its own dementia diagnoses in adults 70 and older. Among 9,745 participants, the authors concluded that addressing all 14 Lancet risk factors could prevent over half of all dementia cases.

In the United States, the Dementia Risk Prediction Project pooled 37,931 participants across six longitudinal cohorts and split the question by age. Factors present in midlife, 45 to 64, accounted for 22.7% of dementia cases, and factors present in late life, 65 and older, for 16.5%. The largest single contributors were late-life physical inactivity at 10.4%, lower education at 8.1% and midlife obesity at 7.7%. Two of the top three are body and behavior, and one of them is measured poorly by weight.

The vascular timeline is sharper. In the Atherosclerosis Risk in Communities study, with 33 years of follow-up in Black and White adults from four US communities, the share of dementia by age 80 attributable to at least one of hypertension, diabetes or current smoking was 21.8% when those factors were measured at 45 to 54 and 44.0% when they were measured at 65 to 74. After 80, only 2% to 8% of cases were attributable to them. The decade of the annual wellness visit is the decade in which measured vascular risk carries the most attributable dementia, and the window narrows quickly after it.

One factor hides behind a normal eye chart. In a cross-sectional analysis of 2,767 older adults in the National Health and Aging Trends Study, a nationally representative sample of US Medicare beneficiaries, 19.0% of prevalent dementia was attributable to at least one objectively measured vision impairment, and contrast sensitivity impairment carried the largest share at 15.0%. A distance acuity chart does not test contrast sensitivity, so the patient who reads the bottom line can still carry the vision deficit most closely tied to dementia. Depression carries its own attributable share, and why remission does not close it is covered in remitted depression as a screening trigger.

Why the measurable factors cluster in one patient

Hypertension, diabetes, obesity, high LDL cholesterol and inactivity are usually charted as five problems. Biologically they converge on two. The first is insulin resistance, which starves insulin-dependent neurons of fuel and feeds the metaflammation that stiffens vessels. The second is the cerebral microvasculature itself, where a stiffened aorta delivers pulsatile pressure into small vessels built for steady flow. The third driver is social and economic: less schooling, food environments dominated by acellular carbohydrate and industrial seed oils, and isolation cluster in the same neighborhoods and the same patients. Metabolic health is already the exception: fewer than 12.2% of US adults qualified in NHANES 2009 to 2016, and the stricter post-2021 criteria put the share under 7%. Trials that test one factor at a time exclude this patient. The patient in the exam room carries five at once, which is why measuring the shared terrain beats counting diagnoses.

A less obvious link sits in the traumatic brain injury item. In older adults, head injury is most often a fall, and falls follow balance, vision, orthostatic blood pressure and cognition together. Vestibular and balance testing therefore belongs in dementia prevention as much as in fall prevention. Where fat and muscle sit adds a further measurable layer, discussed in body composition patterns and neurodegenerative risk.

Does changing the factors change cognition?

The largest US test is the POINTER trial, which enrolled 2,111 adults aged 60 to 79 with a sedentary lifestyle and suboptimal diet plus other risk criteria at 5 US clinical sites. Both a structured and a self-guided lifestyle program raised global cognition over two years, and the structured program did so faster by 0.029 SD per year, with benefit in APOE ε4 carriers and noncarriers alike. The effect is modest, and a 2025 review argues that individual-level programs work within limits set by social and commercial determinants. The practice position is that lifestyle and behavior are treatment: exercise improves insulin sensitivity and cerebral blood flow, and structure, accountability and repeated measurement are what distinguished the stronger arm.

Detection before symptoms is moving toward physiology

Research is also pushing detection earlier. In 137 community-dwelling Black Americans aged 60 to 90 recruited through Wayne State University and the Michigan Alzheimer’s Disease Research Center, 84 cognitively healthy and 53 with mild cognitive impairment, a resting-state EEG connectivity model separated the groups with 91.97% leave-one-out cross-validation accuracy and predicted progression trends over 9 to 18 months with 84.61% accuracy in a limited subset. Measura does not perform EEG; that work remains in research settings. What a practice can do now is fix a standardized cognitive assessment as a baseline and repeat it under the same conditions. Before a slipping score is labeled dementia, a reversible liver cause is worth excluding, as described in liver fibrosis screening in patients with dementia.

Workflow: turning the list into orders

Most of the 14 already have a home in the annual wellness visit, which asks about depression, function and cognition. The gap is conversion of the measurable factors into measurements rather than yes-or-no history. Standing orders let staff start the metabolic, vascular and body composition measurements when a trigger is documented, and cognitive assessment and fall prevention describes pairing the cognitive baseline with balance findings. Documenting each factor as measured, screened or historical makes the risk profile legible to the next clinician.

Frequently asked questions

Does the 45% estimate mean that share of a practice’s dementia is preventable?

It does not. The figure is a global population attributable fraction that assumes each association is causal and that exposure could be removed entirely. Cohort estimates range from about a fifth of cases in pooled US data by life stage to over half in Norway. Its clinical value is in pointing to which exposures deserve measurement and follow-up. The patient-facing version of that argument is dementia prevention starts with numbers you can measure now.

Which of the 14 factors can be documented with objective measurement?

Blood pressure, glucose regulation, LDL cholesterol and adiposity can be measured directly, and vascular consequences can be quantified with arterial stiffness testing. Hearing and vision require audiology and eye care, while depression, alcohol use and isolation rely on validated instruments. Recording which factors were measured rather than self-reported strengthens quality documentation, as described in quality measures that cardiometabolic testing supports.

Why add metabolic measurement to a cognitive screen?

Because a cognitive score records an outcome while the metabolic and vascular measurements describe the terrain producing it. A drifting score in a patient with central adiposity and insulin resistance supports a different follow-up plan than the same score in a metabolically healthy patient. Pairing them also sets up a repeat interval tied to findings. The case is developed in cognitive screening in primary care with the metabolic picture.

How should LDL cholesterol be handled in a patient worried about statins and memory?

High LDL cholesterol is now on the Commission’s list, and many patients have also read that statins affect memory. A documented cognitive baseline before or early in lipid therapy turns a later complaint into a comparison rather than a guess, and keeps the lipid decision with the treating physician. The measurement protocol for patients on statins is described in statins and dementia: measuring cognition in patients on statins.

Should EEG-based detection change practice now?

Not yet at the point of care. The resting-state EEG model performed well in a modest sample with a limited progression subset, and it needs replication before it guides individual decisions. The transferable lesson is the value of a physiological measurement repeated over time, which a standardized cognitive screen already provides in a primary care setting. Why the first administration matters most is explained in what a cognitive baseline is for.

Put the measurable risk factors on paper

See how the Measura protocol turns the metabolic, vascular, body composition and cognitive items on the dementia risk list into documented measurements inside your existing visits.

Send a message

4477 Woodson Rd, Suite 201, St. Louis, MO 63134. Monday to Friday, 9:00 a.m. to 5:00 p.m. Please do not send symptoms, diagnoses or images through a web form — a website form is not a secure medical channel. Send your name and number and we will call you back.

References

Livingston, G., Huntley, J., Liu, K. Y., Costafreda, S. G., Selbæk, G., Alladi, S., et al., & Mukadam, N. (2024). Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet, 404(10452), 572-628. https://doi.org/10.1016/S0140-6736(24)01296-0
Ma’u, E., Cullum, S., Röhr, S., & Brayne, E. C. (2025). Are modifiable risk factors for dementia really modifiable? Current Opinion in Psychiatry, 38(5), 348-354. https://doi.org/10.1097/YCO.0000000000001018
Ellingjord-Dale, M., Strand, B. H., Skirbekk, V., Bratsberg, B., Mekonnen, T., Zotcheva, E., et al., & Engdahl, B. (2025). Potentially modifiable risk factors for dementia in Norway (HUNT4 70+): a retrospective cohort study. The Lancet Healthy Longevity, 6(12), 100802. https://doi.org/10.1016/j.lanhl.2025.100802
Li, J. M., Gauen, A. M., Zmora, R., Stephen, J. J., Petito, L. C., Scholtens, D., et al., & Allen, N. B. (2026). Midlife and late-life population attributable fractions of risk factors for dementia in the United States: The Dementia Risk Prediction Project. Alzheimer’s & Dementia, 22(1), e71065. https://doi.org/10.1002/alz.71065
Smith, J. R., Pike, J. R., Gottesman, R. F., Knopman, D. S., Lutsey, P. L., Palta, P., et al., & Deal, J. A. (2025). Contribution of Modifiable Midlife and Late-Life Vascular Risk Factors to Incident Dementia. JAMA Neurology, 82(7), 644-654. https://doi.org/10.1001/jamaneurol.2025.1495
Smith, J. R., Huang, A. R., Zhou, Y., Varadaraj, V., Swenor, B. K., Whitson, H. E., et al., & Ehrlich, J. R. (2024). Vision Impairment and the Population Attributable Fraction of Dementia in Older Adults. JAMA Ophthalmology, 142(10), 900-908. https://doi.org/10.1001/jamaophthalmol.2024.3131
Baker, L. D., Espeland, M. A., Whitmer, R. A., Snyder, H. M., Leng, X., Lovato, L., et al., & Carrillo, M. C. (2025). Structured vs Self-Guided Multidomain Lifestyle Interventions for Global Cognitive Function: The US POINTER Randomized Clinical Trial. JAMA, 334(8), 681-691. https://doi.org/10.1001/jama.2025.12923
Deng, J., Sun, B., Kavcic, V., Liu, M., Giordani, B., & Li, T. (2024). Novel methodology for detection and prediction of mild cognitive impairment using resting-state EEG. Alzheimer’s & Dementia, 20(1), 145-158. https://doi.org/10.1002/alz.13411