Abstract:BACKGROUND:Frailty is a geriatric condition characterized by a decreased reserve. The Edmonton frailty scale (EFS) has been widely used as an assessment tool in clinical practice. However, the brain's underlying pathophysiological changes in frailty and their associations with the EFS remain unclear. This study aimed to explore the associations between brain volumetry and relaxometry signatures and the EFS (and each domain score of the EFS) in frailty. METHODS:A total of 40 non-demented subjects were enrolled in this prospective study. Frailty assessment was performed for each subject according to the EFS. All subjects underwent synthetic magnetic resonance imaging (MRI) (MAGnetic resonance image Compilation, MAGiC) and three-dimensional fast spoiled gradient-recalled echo (3D-FSPGR) T1-weighted structural image acquisitions on a 3.0 T MR scanner. Brain segmentation was performed based on quantitative values obtained from the MAGiC and 3D-FSPGR images. Volumetry and relaxometry of the global brain and regional gray matter (GM) were also obtained. The associations between the total EFS score (and the score of each domain) and the brain's volumetry and relaxometry were investigated by partial correlation while eliminating the effects of age. Multiple comparisons of regional GM volumetry and relaxometry analyses were controlled by false discovery rate (FDR) correction. All data were analyzed using the SPSS 13.0 statistical package (IBM, Armonk, NY, USA) and MATLAB (MathWorks, Natick, MA, USA). RESULTS:For global volumetry, significant correlations were found between multiple global volumetry parameters and the EFS, as well as the cognition score, functional independence score, nutrition score, and functional performance score (P<0.05). For global relaxometry, notable positive correlations were found between the T2 values of gray and white matter (WM) and the EFS (r=0.357, P=0.026; r=0.357, P=0.026, respectively). Significant correlations were also identified between the T2 value of GM, the T1, T2, and PD values of WM, and the cognition score (r=0.426, P=0.007; r=0.456, P=0.003; r=0.377, P=0.018; r=0.424, P=0.007, respectively), functional independence score (r=-0.392, P=0.014; r=-0.611, P<0.001; r=-0.367, P=0.022; r=-0.569, P<0.001, respectively), and functional performance score (r=0.337, P=0.036; r=0.472, P=0.002; r=0.354, P=0.027; r=0.376, P=0.018, respectively). For regional GM volumetry, multiple regions showed significant negative correlations with the EFS (P<0.05). Notable negative correlations were found between multiple regional GM volume and the functional independence score (P<0.05). For regional GM relaxometry, the T1 and T2 values of several regions showed significant negative correlations with the functional independence score (T1 value of caudate, r=-0.617, P<0.001; T2 value of insula, r=-0.510, P=0.015; T2 value of caudate, r=-0.633, P<0.001, respectively). No significant correlation was found between the domain scores of the EFS and regional GM PD values (P>0.05). CONCLUSIONS:In conclusion, brain volumetry and relaxometry signatures showed strong associations with the EFS and some EFS domain scores in frailty. These associations may reveal the possible underlying pathophysiology of the EFS and different domains of the EFS.

Associations Between Brain Volumetry and Relaxometry Signatures and the Edmonton Frail Scale in Frailty.