Cognitive Fitness in Ageing (COFITAGE): A Multimodal and Longitudinal Neuroimaging Dataset
Purpose: Brain aging involves interrelated changes in molecular processes, neuroinflammatory mechanisms, brain macro- and microstructure, sleep physiology, and cognition. The 50 to 70 years age range represents a critical transition period, in which these subtle alterations may precede measurable cognitive decline and the onset of clinical neurodegenerative disease. To allow systematic investigation of these early alterations and the subsequent progression in brain aging, we provide an open-access data resource from a multidisciplinary longitudinal study integrating neuroimaging, genetics, sleep, and neuropsychological phenotyping with assessments at baseline and at 2-year follow-up. Acquisition and Validation Methods: The baseline cohort comprises 101 community-dwelling participants (50-69 years old) who underwent magnetic resonance imaging (MRI) using a 3T protocol that included high-resolution structural imaging (T1- and T2-weighted), quantitative multi-parametric acquisitions with B1 mapping, and multi-shell diffusion-weighted imaging. Moreover, positron emission tomography (PET) imaging was performed using [18F]Flutemetamol or [18F]Florbetapir (amyloid-beta tracers) in all participants, with a subset also undergoing [18F]THK-5351 PET (tau-related/neuroinflammation). The dataset was complemented by extensive phenotypic data, including sleep and neuropsychological assessments, and by genotype data through genetic analysis. 66 participants underwent a 2-year cognitive follow-up, enabling longitudinal analyses of cognitive trajectories. Data acquisition and curation were performed using standardized procedures, with systematic quality control to support reliable cross-sectional and longitudinal analyses. Data Format and Usage Notes: All data are distributed in a BIDS-compliant format, and released in open-access (EBRAINS). Potential Applications: This dataset supports multimodal analyses, allowing the identification of interpretable patterns characterizing brain aging from multiple perspectives. It enables the comparison of different models to derive (semi)quantitative MRI parameters, the discovery of imaging biomarkers associated with early cognitive decline, and the monitoring or prediction of brain aging progression. In addition, it offers focused coverage of adults aged 50-70 years, which is often underrepresented in existing healthy subjects public datasets.