Linking Brain Age Gap to Mental and Physical Health in the Berlin Aging Study II

Abstract

From a biological perspective, humans differ in the speed they age, and this may manifest in both mental and physical health disparities. The discrepancy between an individual's biological and chronological age of the brain ("brain age gap") can be assessed by applying machine learning techniques to Magnetic Resonance Imaging (MRI) data. Here, we examined the links between brain age gap and a broad range of cognitive, affective, socioeconomic, lifestyle, and physical health variables in up to 335 adults of the Berlin Aging Study II. Brain age gap was assessed using a validated prediction model that we previously trained on MRI scans of 32,634 UK Biobank individuals. Our statistical analyses revealed overall stronger evidence for a link between higher brain age gap and less favorable health characteristics than expected under the null hypothesis of no effect, with 80% of the tested associations showing hypothesis-consistent effect directions and 23% reaching nominal significance. The most compelling support was observed for a cluster covering both cognitive performance variables (episodic memory, working memory, fluid intelligence, digit symbol substitution test) and socioeconomic variables (years of education and household income). Furthermore, we observed higher brain age gap to be associated with heavy episodic drinking, higher blood pressure, and higher blood glucose. In sum, our results point toward multifaceted links between brain age gap and human health. Understanding differences in biological brain aging may therefore have broad implications for future informed interventions to preserve mental and physical health in old age.

Keywords: Berlin Aging Study II (BASE-II); aging; brain age gap; cognition; mental health.

FIGURE 1

Brain-predicted (“brain age”) vs. chronological age stratified by sample and tissue class. Blue dots reflect the estimates of the BASE-II sample (N = 335), with their fitted linear regression line shown in blue. Gray dots reflect the estimates of the UK Biobank imaging cohort (N = 32,634), among whom age estimation models were trained and applied in a 10-fold cross-validation manner. The linear regression line fitted on the UK Biobank data is shown in gray. The identity line (y = x line) is shown in black. At this stage, brain-predicted age estimates have not been corrected for regression dilution, that is, the overestimation of younger participant’s ages and vice versa. Prediction accuracy (blue: BASE-II, black: UK Biobank) was quantified by MAE (mean absolute error between brain-predicted and chronological age), wMAE (weighted MAE defined as ratio between MAE and age range) and rho (Pearson’s correlation coefficient between brain-predicted and chronological age).

FIGURE 2

Partial Pearson correlations between the 27 criterion variables and gray matter, white matter, and combined gray and white matter brain age gap. Effects of sex, age, age², and total intracranial volume were partialled out. Only cells containing associations with p < 0.05 (one-tailed) have been assigned with colors of the blue and red color palette. Green cells show associations not reaching nominal significance (p ≥ 0.05) but with hypothesis-consistent effect directions. Note that the number of observations varied across the criterion measures so that stronger associations do not necessarily reflect results with lower p-values. We provide an interactive version of this plot at https://github.com/pjawinski/base2/. TH, time horizon.

FIGURE 3

Permutation-based quantile-quantile plots showing the distribution of observed p-values from the association analyses (blue circles) sorted from largest to smallest and plotted against the expected p-values under the null hypothesis (1 million permutations; one-tailed testing). The solid diagonal line reflects the mean expected p-values (-log10 scale). The lower and upper bound of the gray shaded area represent the 5th and 95th percentile of the expected p-values. The plots show the association results between the 27 criterion variables and gray matter, white matter, and combined gray and white matter brain age gap, respectively. Overall, quantile-quantile plots suggest that association analyses revealed stronger evidence than expected under the null.