Brain-predicted age difference score is related to specific cognitive functions: a multi-site replication analysis

Abstract

Brain-predicted age difference scores are calculated by subtracting chronological age from 'brain' age, which is estimated using neuroimaging data. Positive scores reflect accelerated ageing and are associated with increased mortality risk and poorer physical function. To date, however, the relationship between brain-predicted age difference scores and specific cognitive functions has not been systematically examined using appropriate statistical methods. First, applying machine learning to 1359 T1-weighted MRI scans, we predicted the relationship between chronological age and voxel-wise grey matter data. This model was then applied to MRI data from three independent datasets, significantly predicting chronological age in each dataset: Dokuz Eylül University (n = 175), the Cognitive Reserve/Reference Ability Neural Network study (n = 380), and The Irish Longitudinal Study on Ageing (n = 487). Each independent dataset had rich neuropsychological data. Brain-predicted age difference scores were significantly negatively correlated with performance on measures of general cognitive status (two datasets); processing speed, visual attention, and cognitive flexibility (three datasets); visual attention and cognitive flexibility (two datasets); and semantic verbal fluency (two datasets). As such, there is firm evidence of correlations between increased brain-predicted age differences and reduced cognitive function in some domains that are implicated in cognitive ageing.

Keywords: Biomarkers; Brain ageing; Cognitive ageing; Cognitive function; MRI; Machine learning.

Fig 1.

Violin plot comparing distributions of brainPADs between sexes across all datasets

Fig 2.

Violin plots comparing distributions of brainPADs between sexes within datasets

Fig 3.

Binarised regression coefficients (positive coefficients shown in pink, negative coefficients shown in yellow) overlaid on 5 coronal slices. A: No threshold applied; B: thresholded at 25^th percentile of absolute value of regression coefficients; C: thresholded at 50^th percentile of absolute value of regression coefficients; D: thresholded at 75^th percentile of absolute value of regression coefficients; E: thresholded at 95^th percentile of absolute value of regression coefficients.

Fig. 4

Scatterplots of replicated correlations between the residuals of brainPAD and cognitive measures after regressing brainPAD on age and sex, and each cognitive measure on age and sex. A: General cognitive status; B: Semantic verbal fluency; C: Processing speed, visual attention, and cognitive flexibility; D: Visual attention and cognitive flexibility. For scatterplots of non-replicated correlations, see Supplementary Info, Figure S.4.