Brain Aging: Genetics Mapped Region by Region in Landmark Study

For the first time, scientists have created a detailed genetic map showing how individual regions of the brain age – and why some areas are particularly vulnerable to neurodegenerative diseases like Alzheimer’s, and dementia. The landmark study, published in the journal GeroScience, moves beyond treating brain age as a single measure, revealing a complex interplay of genetic factors influencing aging across different brain structures.

Mapping the Polygenic Architecture of Brain Aging

Previous research often assigned a single “brain age” score, estimating how an individual’s brain appears on an MRI scan compared to their chronological age. A discrepancy between the two can signal an increased risk of cognitive decline. However, this approach overlooked the nuanced reality of regional brain aging. The new research, led by Nicholas Kim at the USC Viterbi School of Engineering, asked a more precise question: how do genetic factors contribute to aging in specific brain regions?

“We’ve treated brain age like a single number, almost like a GPA for your brain,” explained Kim. “But just like a GPA, that single score hides a lot of nuances.” The team, advised by Andrei Irimia, associate professor of gerontology, quantitative & computational biology, biomedical engineering and neuroscience at USC, discovered that the brain doesn’t age uniformly. Different regions age at different rates, and these differences are influenced by distributed groups of genetic factors. Irimia emphasized the importance of this finding, stating that brain aging isn’t driven by a single genetic factor, but by a polygenic architecture that varies across brain regions.

The researchers analyzed MRI scans from 41,708 adults participating in the UK Biobank, a large-scale British health database. They divided the brain into 148 distinct regions and measured the extent of excessive or delayed aging in each. Then, they scanned each participant’s DNA, testing over 600,000 genetic variants to identify those linked to aging in specific regions. This analysis revealed 1,212 significant genetic associations, creating a detailed genetic map of brain aging.

Genes Linked to Accelerated and Protected Aging

The study identified genetic factors that both accelerate and protect against brain aging. One gene, KCNK2, which controls potassium channels crucial for neuronal signaling, showed a strong association with accelerated aging in brain regions particularly vulnerable in Alzheimer’s disease. Conversely, variations in the NUAK1 gene, which supports the structural integrity of brain cells, were linked to a younger-appearing brain across large areas of the cortex.

It’s important to note that carrying a genetic variant associated with increased risk doesn’t guarantee a specific outcome. As Kim cautioned, “Carrying a risky genetic variant is like having a slightly heavier backpack. It makes the climb harder, but it doesn’t decide whether you reach the top. Lifestyle, environment, vascular health, cognitive engagement, these all matter enormously.”

Significantly, the study found that the brain regions exhibiting the most excessive aging are also those most severely affected by Alzheimer’s disease and frontotemporal dementia. This suggests a strong link between the genetic predisposition to regional brain aging and the development of these neurodegenerative conditions.

The Power of Artificial Intelligence in Brain Imaging

This research wouldn’t have been possible without the application of artificial intelligence. Each MRI scan generates a three-dimensional image comprised of over two million data points – a volume far beyond human processing capacity. The team developed a custom 3D neural network to detect subtle structural changes associated with aging across all brain regions simultaneously. The project required a computer cluster of four servers running 120 processors for approximately a year and a half.

“AI was essential because some aging signals are very subtle,” Kim explained. “We trained a neural network to learn the structural patterns associated with age, and that gave us the trait we needed to run the genetic study.”

Implications for Future Research and Potential Clinical Applications

Whereas this research is primarily a powerful tool for understanding the fundamental mechanisms of brain aging, it raises the possibility of future clinical applications. Could this knowledge eventually help identify individuals at risk for dementia years before symptoms appear, or guide the development of targeted treatments? While these possibilities are promising, Kim emphasized that the research is not yet ready for clinical use. “This is mostly a powerful research tool right now, not a diagnostic test. There are a lot of barriers to moving to the clinical side. Maybe in decades.”

The study also highlighted the importance of considering regional differences in brain aging when investigating neurodegenerative diseases. Future research will likely focus on exploring the specific biological pathways influenced by the identified genetic variants and investigating how lifestyle factors interact with these genetic predispositions.

Further analysis of the identified SNPs, including rs2707466 located in the WNT16 gene, may provide insights into the underlying mechanisms of cortical bone thickness and bone mineral density, as indicated by research published in PubMed and PMC. While this connection may seem distant, it underscores the complex interplay between genetic factors and various physiological processes throughout the body.

Looking Ahead: Ongoing Research and Surveillance

The findings from this study will likely prompt further investigation into the role of specific genes and brain regions in the aging process. Researchers will continue to analyze large datasets, such as the UK Biobank, to refine the genetic map of brain aging and identify additional risk factors. Ongoing surveillance of cognitive health and neurodegenerative disease incidence will be crucial for tracking the long-term impact of these genetic factors and evaluating the effectiveness of potential interventions.

Nicholas J. Kim et al, Deep neural networks and genome-wide associations reveal the polygenic architecture of local brain aging, GeroScience (2025). DOI: 10.1007/s11357-025-02046-1