Background: Dementia arises from the interplay between inherited susceptibility and modifiable lifestyle factors. Yet it remains unclear how these determinants vary across early- and late-onset forms, whether genetic background modifies response to pr...
Background: Dementia arises from the interplay between inherited susceptibility and modifiable lifestyle factors. Yet it remains unclear how these determinants vary across early- and late-onset forms, whether genetic background modifies response to preventive interventions, and through which biological pathways lifestyle factors exert their effects. This dissertation addresses these gaps through three complementary but interlinked investigations, each focused on a different point along the spectrum from disease heterogeneity to intervention efficacy to mechanistic pathways.
Methods: Three complementary investigations were undertaken. (1) Multi-omics profiling of early-onset (EOAD) and late-onset (LOAD) Alzheimer’s disease incorporated genomic, transcriptomic, proteomic, and metabolomic data from European and Korean cohorts. (2) Reanalysis of the SUPERBRAIN-MEET randomized trial evaluated whether polygenic risk scores (PRS) modified the efficacy of a 24-week multidomain intervention in mild cognitive impairment. (3) Dietary factors, which are one of the main lifestyle risk factors, were integrated with proteomic and metabolomic profiling, mediation analysis, and Mendelian randomization (MR) to elucidate pathways linking 27 dietary factors to dementia risk.
Results: Multi-omics profiling demonstrated that EOAD and LOAD, while both influenced by polygenic risk, diverge in downstream signatures: EOAD primarily reflected neuroaxonal injury, whereas LOAD involved systemic metabolic and immune-vascular dysregulation. In the intervention trial, genetic stratification further revealed that individuals at highest inherited risk, particularly APOE ε4 carriers with elevated PRS, derived disproportionate benefit from multidomain lifestyle modification. Dietary analyses then identified specific dietary factors, which is one of the main lifestyle risk factors, that are associated with dementia risk, with adverse diets linked to biomarkers of axonal and astroglial injury and protective diets associated with favorable lipid and glucose metabolism. Mediation and MR analyses supported causal roles for SNAP25 and VLDL-related metabolic pathways, reinforcing neuroaxonal and systemic metabolic mechanisms as unifying biological axes.
Conclusions: EOAD and LOAD represent biologically distinct subtypes. Genetic profiling with PRS and APOE can identify individuals most likely to benefit from lifestyle interventions. Dietary patterns influence dementia risk through reproducible neuroaxonal and metabolic pathways, nominating proteomic and metabolic biomarkers for precision prevention. Together, these findings demonstrate the utility of integrative multi-omics in refining dementia risk stratification and informing personalized intervention strategies.