Advances in molecular genetics and neuroimaging now support novel integrative studies of complex behavioral traits and brain disease.1 These phenomena connect multiple etiologies with common clinical end points of which Alzheimer disease (AD) is probably the best known. Twin studies reliably estimate the effect of environment on AD at approximately 20%,2 but how to disentangle the genetic and environmental factors in AD remains one of the greatest challenges in clinical neuroscience. The task requires understanding a complex matrix of data that comprises the myriad physical manifestations of neurodegenerative diseases arising in a multidimensional research space and occurring at multiple possible levels of enquiry that includes neurogenetics and neuroimaging and covers neurodevelopment and aging, childhood adversities, occupation, physical and mental health, and more. The potential reward is great: identification of neural pathways underpinning the harmful effects and novel types of intervention to prevent or delay AD. In this issue of Neurology, Shen et al.3 used UK Biobank data to make a unique contribution to this problem and show how, following a complex integrative analysis of epidemiologic, neuroimaging, and molecular genetic data, social isolation is linked to incident dementia. They added strong evidence of plausible pathways by showing an association between social isolation and reduced gray matter (GM) volumes and underexpressed genes already linked to AD.