MRI data were pre-processed and analyzed using Statistical Parametric Mapping (SPM8, Wellcome Department of Cognitive Neurology, London, UK). For Voxel-Based morphometry (VBM) analysis, T1-weighted data were classified as gray matter, white matter (WM), and cerebrospinal fluid (CSF) using the unified segmentation approach (Ashburner and Friston, 2005). In order to improve registration accuracy, images were non-linearly transformed to the standard Montreal Neurological Institute (MNI) space using the diffeomorphic registration algorithm (DARTEL) (Ashburner, 2007). After registration and bias correction, GM probability maps were subsequently ‘modulated' by the Jacobian determinants of the deformations to account for local compression and expansion during linear and non-linear transformation (Good et al, 2001). Finally, GM probability maps were smoothed with an isotropic Gaussian kernel of 6 mm full width at half maximum (FWHM) to improve Signal-to-Noise Ratio (SNR) and to conform the data more closely to a Gaussian distribution for statistics. Voxel-based inferential statistics were performed on the smoothed modulated gray matter images using a two-sample t-test embedded in the Random Effect framework of SPM8. We included age, total brain volume (TBV, the sum of gray matter
