Same Checklist, Different Rules: Dementia Risk Factors Split by Gene in Wisconsin Study

Applying the 2024 Lancet Commission's 14-factor dementia checklist to a seven-decade Wisconsin cohort, researchers found the list works differently depending on a single gene: midlife diabetes tracked with dementia mainly in carriers of the APOE4 variant, while midlife hearing loss tracked with dementia mainly in people without it.

Researchers at the University of Wisconsin-Madison, applying the Lancet Commission's 2024 list of 14 modifiable dementia risk factors to 5,526 participants in the Wisconsin Longitudinal Study (WLS), report that the factors do not carry the same weight for everyone: among carriers of the APOE4 gene variant, midlife diabetes was the risk factor most strongly tied to higher dementia odds, while among non-carriers, midlife hearing loss was the factor most strongly tied to higher odds. The findings appear in two abstracts from the same research team, published in Innovation in Aging on December 31, 2025, and in Alzheimer's and Dementia on December 23, 2025, both linked to national gerontology and Alzheimer's conferences.

The 2024 Lancet Commission report, published in The Lancet on July 31, 2024, estimated that addressing 14 modifiable factors across the life course, including hearing loss, hypertension, diabetes, smoking, and air pollution, could theoretically prevent or delay close to 45 percent of dementia cases worldwide, based on a synthesis of many separate studies. The Wisconsin team set out to test how that list performs inside one long-running cohort. The WLS has followed a random sample of 1957 Wisconsin high school graduates and one sibling of each since 1957, drawing on nearly 70 years of prospectively collected data; it is a sibling-based design, not a study of identical twin pairs. Dementia status was determined using a validated telephone-based cognitive screening protocol rather than in-person clinical or neuropathological diagnosis. The researchers used logistic regression with multiple imputation, controlling for age and sex, then split the analysis by APOE4 carrier status, the gene variant that carries the strongest known genetic risk for Alzheimer's disease.

When all 14 factors were modeled together, only a limited subset was associated with dementia status at all, and the pattern diverged by genotype. Among APOE4 carriers, midlife diabetes was linked to higher dementia odds, while midlife hypertension was linked to lower odds, a result the researchers themselves flag as needing explanation, since it runs counter to hypertension's usual midlife risk profile and could reflect treatment effects or other confounding not resolved in this analysis. Among non-carriers, midlife hearing loss and midlife diabetes were both linked to higher odds, with hearing loss emerging as the stronger signal. In an earlier, unstratified look at the same data, less early-life education, late-life social isolation, and late-life vision loss also showed univariate associations with dementia, factors that dropped out or shifted once the researchers adjusted for age, sex, and genotype together.

The team frames its work as a test of whether a population-wide checklist performs the same way inside a single cohort, and their answer is that it does not, at least not uniformly. Neither abstract has yet appeared as a full peer-reviewed paper with complete statistical detail; families and clinicians weighing which midlife conditions deserve the most attention for a given person still have that conversation to have directly, informed by that person's own genetic and health history rather than a generic list.