Summary: The aging of the population poses significant challenges in healthcare, necessitating innovative approaches. Advancements in brain imaging and artificial intelligence now allow for characterizing an individual’s state through their brain age,’’ derived from observable brain features. Exploring an individual’s biological age’’ rather than chronological age is becoming crucial to identify relevant clinical indicators and refine risk models for age-related diseases. However, traditional brain age measurement has limitations, focusing solely on brain structure assessment while neglecting functional efficiency.

Our study focuses on developing neurocognitive ages’’ specific to cognitive systems to enhance the precision of decline estimation. Leveraging international (NKI2, ADNI) and Canadian (CIMA- Q, COMPASS-ND) databases with neuroimaging and neuropsychological data from older adults [control subjects with no cognitive impairment (CON): n = 1811; people living with mild cognitive impairment (MCI): n = 1341; with Alzheimer’s disease (AD): n= 513], we predicted individual brain ages within groups. These estimations were enriched with neuropsychological data to generate specific neurocognitive ages. We used longitudinal statistical models to map evolutionary trajectories. Comparing the accuracy of neurocognitive ages to traditional brain ages involved statistical learning techniques and precision measures.

The results demonstrated that neurocognitive age enhances the prediction of individual brain and cognition change trajectories related to aging and dementia. This promising approach could strengthen diagnostic reliability, facilitate early detection of at-risk profiles, and contribute to the emergence of precision gerontology/geriatrics.