Forty years ago, a perceptive Review of depressive disorders in Science (Akiskal and McKinney, 1973) argued that a psychoanalytic model of MD as object loss (a proximal cause of MD) could be conceptualized as loss of reinforcement, or loss of control over reinforcement, then subject to experimental investigation in animal models, and integrated with anatomical, biochemical, and pharmacological data as a process occurring in the diencephalic centers of reward. In this view, MD is a final common pathway, a decrease in the functional capacity of the reward system. Since then, MD has begun to appear as a relatively thin covering serving to unite a plethora of independently acting mechanisms. Genetic analyses can identify risk variants, both rare and common, and in so doing cast much needed illumination on the biology of the commonest psychiatric disorder. The difficulties of sample size and clinical differentiation are daunting but unavoidable if we are to take advantage of the promise that genetics makes.
