Abstract
A set of variables from the social environment (e.g., social contacts, confiding relationships, leisure activities, loneliness) was used to differentiate between a consecutive series of 111 patients diagnosed as having a depressive illness and a sample of 98 non-psychiatric controls. The function derived from the discriminant analysis correctly classified 83% of the subjects. Results indicate shortcomings in the social environment of depressives and point to the discriminating power of such variables.