Mid-life brain traits genetically linked to late-life memory map to AD-vulnerable regions, suggesting biologically relevant risk pathways and potential drug targets for cognitive decline.
Key Findings
Results
Diffusion and structural imaging-derived phenotypes in medial temporal and frontal regions showed the strongest genetic covariance with memory performance and decline.
Analysis leveraged GeNetic cOVariance Analyzer (GNOVA) to estimate genetic covariance between GWASs of memory performance (MEM) and decline (memslopes) and 3935 UK Biobank imaging-derived phenotypes (IDPs)
IDPs spanned diffusion, structural, and functional modalities
Memory GWASs included 24,216 non-Hispanic White older adults (mean age = 74.46 years)
IDP GWASs included 33,224 European-ancestry mid-life participants (age range 45.1 to 81.8 years, mean = 64.28 years)
Medial temporal and frontal regions were identified as showing strongest genetic covariance
Results
Default mode network functional connectivity showed shared genetic architecture with memory performance and decline.
Functional connectivity IDPs from the default mode network demonstrated genetic covariance with memory traits in addition to diffusion and structural IDPs
Analysis covered imaging phenotypes across diffusion, structural, and functional modalities
Results were stratified by cognitive status and apolipoprotein E (APOE) inclusion
Results
Mid-life brain imaging traits that are genetically linked to late-life memory map to Alzheimer's disease-vulnerable regions.
The study design connected mid-life brain traits (mean age ~64 years) with late-life memory outcomes (mean age ~74 years)
Regions identified correspond to those known to be vulnerable to AD pathology
Authors conclude findings suggest 'biologically relevant risk pathways and potential drug targets for cognitive decline'
Memory is described as 'a strong endophenotype for Alzheimer's disease' but 'typically detectable only after substantial brain change'
Methods
The study applied a genetic covariance approach to link memory GWASs with brain imaging GWASs, enabling cross-trait genetic analysis without requiring the same individuals.
GNOVA (GeNetic cOVariance Analyzer) was used to estimate genetic covariance between memory and imaging phenotypes
Memory GWASs and IDP GWASs were derived from separate cohorts of different age ranges
3935 imaging-derived phenotypes were tested across multiple modalities
Analyses were stratified by cognitive status and APOE inclusion to examine subgroup effects
Background
The study's rationale was that genetically linking late-life memory with mid-life brain traits may identify early markers of AD-related cognitive decline.
Memory decline is typically detectable only after substantial brain change has already occurred
Using genetic covariance across GWASs allows identification of shared biology between mid-life imaging and late-life cognitive outcomes
The approach aimed to identify potential early, pre-symptomatic brain markers
Sample included non-Hispanic White and European-ancestry participants, limiting generalizability
What This Means
This research suggests that certain brain structural and connectivity features measurable in middle age share a common genetic basis with memory performance and memory decline measured in later life. Using a statistical technique called genetic covariance analysis, the researchers linked large datasets of brain scans from roughly 33,000 middle-aged adults with memory test data from over 24,000 older adults, without needing the same people to be in both studies. The brain regions most strongly connected to memory through shared genetics were the medial temporal lobe and frontal regions — areas known to be among the first affected by Alzheimer's disease — as well as a brain network called the default mode network.
This matters because memory problems in Alzheimer's disease are usually only detectable after significant brain damage has already occurred. By identifying brain imaging features in midlife that share genetic underpinnings with late-life memory, this research suggests it may be possible to identify biological risk pathways for Alzheimer's-related cognitive decline much earlier than current methods allow. The authors propose these genetically linked brain markers could point toward potential targets for drug development aimed at preventing or slowing cognitive decline.
It is important to note that the study focused on participants of European ancestry, which limits how broadly these findings can be applied to other populations. The use of genetic covariance rather than direct longitudinal tracking of the same individuals also means the findings reflect shared biology rather than direct causation. Nevertheless, this approach offers a promising framework for identifying early, biologically grounded markers of Alzheimer's disease risk.
Yang Y, Lorenz A, Sathe A, Schilling K, Gaynor L, Choi S, et al.. (2026). Genetically linked brain imaging markers of memory decline in aging and Alzheimer's disease.. Alzheimer's & dementia : the journal of the Alzheimer's Association. https://doi.org/10.1002/alz.71663