Second, our results provide—for the first time to our knowledge—some support, from a genetic perspective, for the decision in DSM-III to distinguish between axis I and axis II disorders. The genetic substrate for axis II disorders is, in our analyses, at least partially separable from those factors that predispose to axis I disorders. However, the axis I and II disorders that loaded on our genetic factors are not isomorphic with those articulated by DSM-III. Two axis I disorders—dysthymia and social phobia—were included in the internalizing axis II cluster. The concept of dysthymia evolved in part from the concept of “depressive personality” (16, 43). Our results suggest that from a genetic perspective, it may be better placed with the personality disorders than in the mood disorders section. A debate has long simmered about the relationship between social phobia and avoidant personality disorder (see, for example, references 44–47). Our results suggest that from a genetic perspective, social phobia belongs with avoidant personality disorder on axis II.
