Associations between accelerometer‐based physical activity and brain white matter microstructure: Findings from the population‐based Rhineland Study

Abstract

AbstractBackgroundWhite matter (WM) integrity declines with age and is affected by neurodegenerative diseases. Physical activity promotes neuronal health in animals, and is associated with WM microstructure metrics in humans. However, previous diffusion tensor imaging (DTI) studies were limited by small sample sizes and subjective questionnaire‐based physical activity assessments. Moreover, recently introduced neurite orientation dispersion and density imaging (NODDI) may offer complementary insights. Therefore, we examined the association between accelerometer‐based physical activity and DTI and NODDI measures of WM microstructure in a large population‐based cohort.MethodWe analyzed data of 2,455 participants (57.9% women, mean age: 54.7 years) with valid physical activity and MRI measurements from the Rhineland Study, a prospective cohort study in Bonn, Germany. We recorded physical activity using accelerometers and derived metabolic equivalent (MET)‐hours, step counts and time spent performing sedentary, light‐intensity and moderate‐to‐vigorous (MVPA) activities. DTI (fractional anisotropy (FA), mean diffusivity) and NODDI (orientation dispersion index, neurite density index) measures were obtained from 3T MR images using the microstructure diffusion toolbox. The association of physical activity with WM microstructure was assessed using multivariate polynomial regression, while adjusting for age, sex, education, body mass index and smoking status.ResultPhysical activity was associated with several global and tract‐specific DTI and NODDI measures. Effects were strongest at low physical activity quantities and weakened at higher levels. While higher MET‐hours, step counts and light and moderate‐to‐vigorous intensity activities were related to greater global FA, sedentary time was associated with reduced FA. We observed the strongest tract‐specific effects for moderate‐to‐vigorous intensity activities on the corpus callosum and corona radiata (standardized effects: MVPA on FA corpus callosum, ßlin = 0.0009 [95% CI: 0.0001; 0.0018] and ßqu = ‐0.0006 [95% CI: ‐0.0011; ‐0.0001]; MVPA on FA corona radiata, ßlin = 0.0011 [95% CI: 0.0001; 0.0021] and ßqu = ‐0.0009 [95% CI: ‐0.0014; ‐0.0003]). We obtained similar results for NODDI measures.ConclusionOur findings suggest that WM microstructure and particularly tracts connecting motor regions profit from physical activity. Sedentary individuals may benefit more from additional physical activity than already active individuals. NODDI metrics provide valuable complementary insights into the effects of physical activity on WM microstructure.