When your brain looks older than expected: combined lifestyle risk and BrainAGE

Nora Bittner,Christiane Jockwitz,Katja Franke,Christian Gaser,Ute J Bayen,Svenja Caspers,Katrin Amunts,Susanne Moebus

doi:10.1007/s00429-020-02184-6

Abstract

Lifestyle may be one source of unexplained variance in the great interindividual variability of the brain in age-related structural differences. While physical and social activity may protect against structural decline, other lifestyle behaviors may be accelerating factors. We examined whether riskier lifestyle correlates with accelerated brain aging using the BrainAGE score in 622 older adults from the 1000BRAINS cohort. Lifestyle was measured using a combined lifestyle risk score, composed of risk (smoking, alcohol intake) and protective variables (social integration and physical activity). We estimated individual BrainAGE from T1-weighted MRI data indicating accelerated brain atrophy by higher values. Then, the effect of combined lifestyle risk and individual lifestyle variables was regressed against BrainAGE. One unit increase in combined lifestyle risk predicted 5.04 months of additional BrainAGE. This prediction was driven by smoking (0.6 additional months of BrainAGE per pack-year) and physical activity (0.55 less months in BrainAGE per metabolic equivalent). Stratification by sex revealed a stronger association between physical activity and BrainAGE in males than females. Overall, our observations may be helpful with regard to lifestyle-related tailored prevention measures that slow changes in brain structure in older adults.

Highlights

Structural brain changes during normal aging comprise decreases in gray matter (GM) and white matter (WM; Fjell and Walhovd 2010)
Even though our study focuses on the association between lifestyle behavior and brain aging in older adults, we chose to train the relevance vector machine (RVM) on the full cohort of 1000BRAINS for two reasons: (i) from a neuropsychological perspective, brain aging is a highly complex phenomenon, which can - as a pattern - better be learned by the algorithm when having access to a larger age range, which in the full cohort spans the entire adult life span (18.5–87 years), i.e., the algorithm has access to phenotypically richer data
Within the older subsample (n = 622), which we used for our main analysis of the association between lifestyle and BrainAGE, mean BrainAGE was 0.23 years (SD = 4.96) with a maximum positive deviation between chronological and estimated age of + 15.92 years and a maximum negative deviation of – 15.67 years

Summary

Introduction

Structural brain changes during normal aging comprise decreases in gray matter (GM) and white matter (WM; Fjell and Walhovd 2010). Some older individuals experience strong and early manifestations (accelerated brain aging), while others of comparable age do not experience changes expected at that age [decelerated brain aging; (Bartrés-Faz and Arenaza-Urquijo 2011; Ziegler et al 2012)]. A protective environment, e.g., healthier lifestyle such as higher physical activity and social integration, may promote a more resilient soil and neuroprotection (Anaturk et al 2018; Arenaza-Urquijo et al 2015; Bittner et al 2019; Fratiglioni et al 2004; McDonough and Allen 2019). Lifespan physical activity has been associated with favorable ratios of brain metabolism markers in magnetic resonance imaging (Engeroff et al 2019) This hints even further at physical activity being a factor for promoting a more resilient neuronal environment. Excessive alcohol consumption can lead to serious neurological diseases, e.g., Korsakow syndrome (de la Monte and Kril 2014) and is associated with reduced GM and WM volume and density (Paul et al 2008; Pfefferbaum et al 1995; Topiwala et al 2017) in alcohol-dependent as well as non-dependent individuals (Mukamal at al. 2001)

Methods

Results

Conclusion