The main motivation to screen brain images was to find some heritable measure of disease burden that might be closer to the underlying genetic effect than clinical diagnosis based on cognitive and clinical tests. The endophenotype hypothesis, long advocated by psychiatric geneticists such as Irving Gottesman (Gottesman and Gould 2003; Blangero 2004; Goldman 2012; White and Gottesman 2012; Kendler and Neale 2010) suggested that one might fruitfully apply genetic screening to any reliable and heritable biomarkers of a disease—measures from the blood or cerebrospinal fluid (CSF), or even from brain scans, which by now had become quite plentiful. The original definition of “endophenotype” for an illness or disorder (see Gottesman and Gould 2003) suggested that an endophenotype should (i) be associated with the illness/disorder of interest, (ii) be heritable, (iii) be state-independent, i.e., seen in people even when they do not show symptoms of the illness/disorder, (iv) co-segregate with illness/disorder within families, and (v) be observed in relatives of affected family members at a higher rate than in the general population.
