Borderline: an adjective in search of a noun.

Outpatients diagnosed as borderline (N = 100) were prospectively followed for 6-36 months and examined from phenomenologic developmental, and family history perspectives. At index evaluation, 66 met criteria for recurrent depressive, dysthymic, cyclothymic, or bipolar II disorders, and 16 for those of schizotypal personality. Other subgroups included sociopathic, somatization, panic-agoraphobic, attention deficit, epileptic, and identify disorders. Compared with nonborderline personality controls, borderlines had a significantly elevated risk for major affective but not for schizophrenic breakdowns during follow-up. Prominent substance abuse history, tempestuous biographies, and unstable early home environment were common to all diagnostic subgroups. In family history, borderlines were most like bipolar controls, and differed significantly from schizophrenic, unipolar, and personality controls. It is concluded that, despite considerable overlap with subaffective disorders, the current adjectival use of this rubric does not identify a specific psychopathologic syndrome.