Effects of Physical Activity on Brain White Matter Integrity Are Region-, Dose-, and Age-Dependent: Findings From the Population-Based Rhineland Study

Abstract

White matter integrity is a sensitive marker of cerebrovascular health and aging. Physical activity may promote white matter health, but previous studies were limited by self-reported activity and small sample sizes. We examined the association between accelerometer-based physical activity and novel indices of white matter integrity while identifying potential mediators. We analyzed cross-sectional baseline data from 4188 participants (58% women, mean age: 55.4 years) of the Rhineland Study, a prospective cohort study in Bonn, Germany. Physical activity measures (independent variables) were derived from activPAL3 micro-accelerometers. Fractional anisotropy, mean diffusivity, free water fraction, orientation dispersion, and neurite density (outcomes) were obtained from 3 T MR images for the whole brain and across 27 white matter tracts. We performed multivariable polynomial regression and mediation analyses for cardiometabolic and inflammatory factors. Higher energy expenditure was related to lower orientation dispersion, with the association leveling off at higher amounts (standardized β_linear = −0.153 [−0.183; −0.122]; β_quadratic = 0.034 [0.017; 0.051]; both p < 0.001). Physical activity was linked to lower neurite density in motor tracts (βs between −0.133 and −0.056, all p < 0.001). However, with advancing age, physical activity became more strongly associated with higher neurite density and lower mean diffusivity of supratentorial fibers. Body weight, cardiometabolic, and inflammatory factors partly mediated effects of physical activity on white matter microstructure. Physical activity is related to white matter integrity in a nonlinear, age- and region-specific way. Physical activity's effects are greatest in sedentary and older individuals, likely mediated through its impact on body composition, cardiometabolic health, and systemic inflammation.