Dietary preference, biological aging, and incident dementia: a prospective cohort study.
Key claim: A sweet-tooth dietary preference profile was associated with higher all-cause, Alzheimer’s, and vascular dementia risk and with accelerated biological aging (KDM-BA, PhenoAge, shorter LTL), particularly among people at high genetic risk.
Abstract
Dietary patterns are modifiable determinants of dementia risk, but the interactions between individual dietary preference profiles, biological aging, and dementia risk remain unclear. We aimed to investigate associations of dietary preference profiles with biological aging and incident dementia, as well as the potential modifying effect of genetic susceptibility. Dietary preference profiles were identified using latent profile analysis on food preferences questionnaire data. Cox models estimated the association between dietary preferences and dementia risk and their interaction with genetic risk. Associations with biological aging acceleration (Klemera-Doubal method biological age [KDM-BA] acceleration, phenotypic age (PhenoAge) acceleration, leucocyte telomere length [LTL]) were assessed with linear regression. Mediation analysis explored potential mediators. Three dietary preference profiles were identified: health-aligned, broad-preference, and sweet-tooth. The sweet-tooth group showed a 79% higher risk of all-cause dementia, 62% higher risk of Alzheimer’s disease, and 202% higher risk of vascular dementia compared to the health-aligned group. The sweet-tooth profile was associated with accelerated KDM-BA, PhenoAge, and shorter LTL. A significant additive interaction between sweet-tooth profile and high genetic risk amplified dementia risk. Exploratory mediation analyses suggested that biological aging markers statistically accounted for 2.28-4.82% of the observed association. Sweet-tooth dietary preference was associated with higher dementia risk and accelerated biological aging, particularly among individuals with high genetic susceptibility. These findings support targeted dietary interventions for dementia prevention, especially in genetically at-risk populations.
Why this matters
Most diet–dementia evidence scores nutrients or named patterns; this cohort instead derives latent dietary preference profiles and finds the ‘sweet-tooth’ class tracks both incident dementia and three biological-aging surrogates (KDM-BA, PhenoAge, LTL) simultaneously, with a gene–diet interaction that concentrates risk in high–polygenic-risk individuals. For this area it extends the sugar/UPF-dementia signal from intake measurement to behavioral preference phenotyping, and offers a rare example of surrogate aging clocks moving in the same direction as a hard incidence endpoint within one cohort.