Some older adults maintain exceptional memory and thinking skills well into late life, and a new study provides evidence that their brains may biologically age at a slower pace. Published in GeroScience this research from a team at Ewha Womans University in South Korea used machine learning to predict the “biological age” of the brain based on brain scans. The findings suggest that people known as “superagers” — older adults with memory performance similar to people decades younger — tend to have brains that appear structurally younger and age more slowly over time compared to their peers.
Aging is commonly associated with declines in memory and thinking abilities. However, some older adults continue to perform at high levels on memory tests, even into their 70s and 80s. These individuals are often called “superagers.”
Scientists have been curious about whether the brains of these individuals are structurally different or whether they simply follow a more favorable aging path. While past research has shown that superagers tend to retain more brain volume in areas tied to memory, it has not been clear whether this preservation reflects a younger biological age or just a normal but healthier version of aging.
The current study aimed to go beyond simply measuring brain volume. Instead, it used a sophisticated machine learning model to estimate the biological age of a person’s brain, a method that looks at overall patterns of brain structure rather than specific regions. The researchers wanted to find out whether superagers’ brains appear younger than their actual age and whether their brains age more slowly over time. The team also hoped to understand whether having a younger-appearing brain was tied to better memory and thinking abilities.
To answer these questions, the researchers recruited 153 adults between the ages of 61 and 93 from the community. Each participant underwent brain scans and completed a battery of cognitive tests. Based on their memory performance, 63 individuals were classified as superagers. These participants performed as well as or better than middle-aged adults on memory tests, particularly on measures of delayed recall. The other 90 participants, whose memory scores were average for their age, were categorized as typical older adults. The researchers followed up with 64 participants two years later to assess how their brain aging had changed over time.
To estimate brain age, the team built a machine learning model using brain scans from nearly 900 healthy people between the ages of 31 and 100. This model was trained to predict a person’s age based on structural patterns in their brain. It used information from different types of brain tissue, including gray matter and white matter, and considered sex-based differences in brain aging. The model was then adapted specifically for use in older adults using a technique known as transfer learning.
The key measurement used in the study was the “brain age gap.” This represents the difference between a person’s predicted brain age and their actual age. A negative gap means the brain looks younger than expected, while a positive gap suggests the brain looks older.
At the first time point, superagers had an average brain age gap of about minus 1.4 years. In other words, their brains appeared younger than their chronological age. In contrast, typical older adults had a slightly positive brain age gap of about 1 year. The difference between these two groups was statistically significant, suggesting that superagers have structurally younger brains than their peers.
Two years later, the researchers repeated the brain scans on a subset of participants. At this point, the difference between superagers and typical older adults became even more pronounced. Superagers’ brain age gap had grown more negative, averaging almost minus 2.8 years, while typical older adults showed a positive brain age gap of more than 2.4 years. This meant that over time, superagers’ brains not only remained younger but were aging more slowly than those of their peers.
The study also tracked how the brain age gap changed over the two-year period. This allowed the researchers to estimate the rate of brain aging in each group. Participants who stayed in the superager category across both time points showed almost no increase in their brain age gap, suggesting minimal biological aging.
On the other hand, participants who remained in the typical group showed a significant increase in their brain age gap, indicating faster aging. Interestingly, some individuals who improved their memory scores over time and became superagers also showed a decrease in brain age gap, meaning their brains appeared to get younger. In contrast, those who declined from superager status showed a steep increase in brain age gap.
Beyond group differences, the study also examined whether the brain age gap was linked to memory and general thinking abilities. Across the entire sample, individuals with a younger brain age tended to perform better on memory tasks and scored higher on a common measure of overall cognitive function. This suggests that a younger-appearing brain is associated with better mental performance in late life.
But there are some caveats to consider. One of the biggest constraints was the small number of participants who completed the two-year follow-up, which makes it harder to draw conclusions about long-term patterns. In addition, the superagers in this study were defined using criteria that included people as young as 60, whereas other research sometimes focuses on those over 80. This could affect how the results apply to different groups of older adults.
Another limitation relates to the brain age model itself. It was based only on structural brain features visible in MRI scans. Brain aging is a complex process that involves many factors beyond structure, including blood flow, metabolism, and the buildup of proteins like beta-amyloid. The authors note that future studies could improve brain age prediction by incorporating more types of brain data.
Finally, the model used a linear approach to estimating change over time, which may not fully reflect the more irregular, non-linear patterns of brain aging that can occur in real life. New methods that capture these complex patterns might provide a clearer picture of how brain aging unfolds.
The study, “Preserved brain youthfulness: longitudinal evidence of slower brain aging in superagers,” was authored by Chang‑hyun Park, Bori R. Kim, Soo Mee Lim, Eun‑Hee Kim, Jee Hyang Jeong, and Geon Ha Kim.
