MCI from anterior and mid-ventricular brain ratios, corti-cal atrophy measures, hippocampal volumes, volumes of amygdala and white matter hyperintensities after adjusting for age, gender, years of education, depression and physical disability. None of the neuroanatomical substrates appeared as predictors of MCI. The only pre-dictors were higher depression scores and fewer years of education. ... [Show full abstract] Structural neuroimaging may not have an added advantage in the detection of MCI in middle-aged community-dwelling subjects. It may be that this age group is too young for such brain changes to be identi-fi ed.