Summary: Researchers at Karolinska Institutet used an AI-based method to analyze MRI scans from a group of cognitively healthy 70-year-olds and estimate each person’s biological brain age. The team identified clear links between vascular risk factors—such as diabetes, high blood glucose and inflammation—and brains that appear older than chronological age. Conversely, regular physical activity and other healthy lifestyle choices were associated with brains that look younger than expected.
The findings emphasize the role of vascular health and metabolic regulation in preserving cognitive resilience during aging. The research group plans to refine the AI tool for possible clinical applications and will expand future work to examine sex differences and social determinants of brain resilience.
Key findings:
- Older-looking brains: Diabetes, prior stroke, cerebral small vessel disease and systemic inflammation were associated with a brain appearance consistent with greater biological age.
- Younger-looking brains: Regular exercise and several healthy lifestyle factors correlated with brains that appeared biologically younger than chronological age.
- AI potential: The AI algorithm produced consistent estimates of biological brain age and shows promise as a future adjunct in dementia assessment and research.
Source: Karolinska Institute
Study overview
Using an automated AI algorithm, investigators from Karolinska Institutet analyzed MRI scans from 739 cognitively unimpaired 70-year-old participants drawn from the Gothenburg H70 cohort. Of these participants, 389 were female. The algorithm predicted the biological age of each brain image; researchers then compared predicted brain age to chronological age to calculate the brain age gap (BAG), a marker of brain resilience.
Each participant underwent structural MRI and provided blood samples used to measure lipids, glucose markers and inflammatory biomarkers. Cognitive testing and detailed lifestyle and medical-history data were also available, allowing the team to investigate how real-world exposures and biological processes relate to brain appearance.
AI-derived brain age and the brain age gap
On average, the AI model estimated brain age at about 71 years for both men and women in the cohort. The brain age gap (BAG)—calculated as predicted brain age minus chronological age—served as the principal outcome. A larger, positive BAG indicates a brain that appears older than the person’s chronological age, while a negative BAG suggests a younger-looking brain.
Analyses showed that individuals with diabetes, prior stroke, higher burden of cerebral small vessel disease, elevated systemic inflammation and higher blood glucose tended to have larger BAGs. In contrast, participants who reported regular physical activity and overall healthy lifestyles more often had smaller or negative BAGs, reflecting a younger-appearing brain.
“These results reinforce that factors damaging the vascular system are closely linked with an older brain appearance,” says Anna Marseglia, lead author and researcher at the Department of Neurobiology, Care Sciences and Society. “Maintaining vascular and metabolic health—such as managing blood glucose and inflammation—may help protect the brain against age-related decline.”
Importance for dementia and clinical practice
More than 20,000 people in Sweden are diagnosed with some form of dementia each year, with Alzheimer’s disease representing roughly two-thirds of cases. While emerging Alzheimer’s treatments offer new options for some patients, they are not universally effective. Tools that quantify brain resilience and identify modifiable risk factors could help tailor prevention strategies and guide clinical evaluation.
Eric Westman, professor of Neurogeriatrics and principal investigator, notes that the algorithm is accurate and robust in a research setting and that further validation is needed before clinical deployment. The research team intends to continue refining the model and to explore how it could be integrated into diagnostic pathways for cognitive disorders.
Sex differences and future directions
The study found indications that associations between life exposures and BAG may differ between men and women, suggesting sex-specific pathways to building brain resilience. Upcoming studies will examine biological factors—such as sex hormones—and sociocultural elements, including social engagement, connectedness, sleep and stress, with particular attention to women’s health factors.
Funding and disclosures
The study received primary funding from multiple Swedish research agencies and foundations supporting dementia and aging research. No researcher from Karolinska Institutet reported conflicts of interest; one co-author disclosed affiliations with several industry partners.
About this AI and brain aging research news
Author: Press Office
Source: Karolinska Institute
Contact: Press Office – Karolinska Institute
Image: Credit to Neuroscience News
Original Research: Open access. “Biological brain age and resilience in cognitively unimpaired 70-year-old individuals” by Anna Marseglia et al. Alzheimer’s & Dementia
Abstract
Biological brain age and resilience in cognitively unimpaired 70-year-old individuals
INTRODUCTION
This study examined how the brain age gap (BAG)—a measure reflecting the difference between predicted brain age and chronological age—relates to life exposures, neuroimaging findings, biological markers and cognitive performance in septuagenarians without dementia.
METHODS
BAG was derived for 739 participants by subtracting predicted brain age from chronological age. Robust linear regression models tested associations between BAG and life exposures, plasma inflammatory and metabolic biomarkers, MRI measures of brain structure and small vessel disease, cerebrospinal fluid markers, and cognitive test scores.
RESULTS
Greater BAG (older-looking brains) was linked to physical inactivity, diabetes and stroke. Paradoxically, prediabetes was associated with lower BAG in this cohort. Physical activity reduced the association between obesity and BAG. Larger BAG correlated with higher small vessel disease burden, white-matter changes, systemic inflammation, elevated glucose and poorer performance on vascular-related cognitive domains. Several sex-specific associations were observed.
DISCUSSION
The findings indicate that vascular-related lifestyles and health factors shape brain appearance in older age. Inflammation and insulin-related metabolic processes appear to be central to understanding vascular contributions to cognitive disorders and may represent targets for prevention and intervention.