Demography of DSM-III borderline personality disorder (PD): a comparison with Axis II PDs, affective illness and schizophrenia convergent and discriminant validation.

Demographic characteristics of borderline personality disorder (PD) defined according to DSM-III criteria were compared with those found for schizophrenia, affective illness, and other Axis II PDs. Borderline PD, unlike affective illness and most other Axis II PDs, usually occurs before the age of 30. By contrast to antisocial PD and schizophrenia, borderline PD usually occurs after the age of 25. For borderline PD (N = 280) average age was significantly more homogeneous compared with affective illness (N = 157) and Axis II PDs (N = 71) across 9 studies. By contrast, variability for 63 predominantly male schizophrenics in 3 studies was significantly less, reflecting the younger age at admission compared with borderline PD (N = 84). According to predictions based upon an age-of-risk hypothesis (Dahl, 1985) for 106 borderline PD patients, a significantly lower percentage were > or = 40 years of age than diagnostic controls (N = 181) predominantly with DSM-III affective illness. Borderline PD is predominantly diagnosed in females either single or who have been divorced compared with Axis II PDs and affective illness, to a lesser extent. Unlike antisocial PD, as well as schizophrenia, the preponderance of male and single/divorced patients usually occur significantly less than for borderline PD. Borderline PD usually occur significantly less than Axis II PDs, affective illness and schizophrenia and ethnic minorities, particularly Afro-Americans. These differences in ethnic/racial distribution are explained in terms of two hypotheses. From the perspective of demographic variables, borderline PD closely converges with neither (a) schizophrenia, (b) antisocial PC, (c) other Axis II PDs, nor (d) affective illness. Evidence for discriminant and convergent validation of these data is provided by (a) cluster analyses and intersample pairwise contrasts, as well as comparisons with (b) clinical samples selected on the basis of DSM-III-R and criteria of the (c) Diagnostic Interview for Borderline Patients, (d) a longitudinal case registry study conducted in Denmark, and (e) prospective surveys conducted among (i) North Carolina community residents and (ii) first degree (nonpatients) relatives of psychiatric patients in Iowa.