on the maximum symptom count is meaningful and more informative. In doing so, the density of one’s FH is more closely captured and the sensitivity of density measures as risk indices is increased even for individuals without any AUD diagnosed relatives. Indeed, the density measure FHDrPSND–AUDMaxSxCnt–based on log-transformed DSM-5 maximum symptom counts of primary and secondary non-descendants, emerged as the most informative measure with the highest diagnostic accuracy at classifying AUD diagnosis and with most robust associations with clinical, behavioral, and neural phenotypes, for all subjects. Second, the density measures used a weighting scheme that corresponded to the affected family members’ degree of relatedness such that the weights decreased exponentially with the increase in degree of relatedness. This formulation was designed to also approximate the amount of biopsychosocial information shared between an individual and his/her affected relative. That is, the biological (genes) and psychosocial (environment) information shared between individuals and their siblings may be more similar than between an individual and a distant cousin. Third, because density measures are ratio/proportion scores, they control for inflation of scores due to variation in subjects’ total number of alcohol-affected relatives. The benefit of using ratio scores, compared to the dichotomous measures which are
