BACKGROUND: There is little information of the prevalence of personality disorder (PD) in those with depressive disorder in community samples; neither is there any data on the impact of PD on service utilisation or outcome in this setting. METHODS: A two stage screening study to identify cases of depressive disorder using SCAN in five European countries. Personality assessed 6 months after the diagnostic interview. Follow-up for 1 year using symptom and social function measures. RESULTS: Personality disorder is present in 22% of a community sample with depressive disorders but the range varied from 13.7% to 33.3% across countries. Cluster C formed 43% of the total. Long-term psychotropic drug use was more common in the PD group even after depression was controlled. Those with PD had higher symptom scores at the outset and, although the PD group was more likely to be cases at follow-up, this disappeared when the depression score was co-varied. Only initial social function predicted outcome at 6 and 12 months. LIMITATIONS: The use of a non-treatment seeking population may limit the application of the findings to clinical populations. CONCLUSIONS: PD is common even in a non-treatment seeking population with depressive disorder. It impacts upon outcome at 6 and 12 months but this is related to the initial severity of depressed mood. Social function is the only independent predictor of outcome and should be assessed separately.
BACKGROUND: There is little information of the prevalence of personality disorder (PD) in those with depressive disorder in community samples; neither is there any data on the impact of PD on service utilisation or outcome in this setting. METHODS: A two stage screening study to identify cases of depressive disorder using SCAN in five European countries. Personality assessed 6 months after the diagnostic interview. Follow-up for 1 year using symptom and social function measures. RESULTS:Personality disorder is present in 22% of a community sample with depressive disorders but the range varied from 13.7% to 33.3% across countries. Cluster C formed 43% of the total. Long-term psychotropic drug use was more common in the PD group even after depression was controlled. Those with PD had higher symptom scores at the outset and, although the PD group was more likely to be cases at follow-up, this disappeared when the depression score was co-varied. Only initial social function predicted outcome at 6 and 12 months. LIMITATIONS: The use of a non-treatment seeking population may limit the application of the findings to clinical populations. CONCLUSIONS:PD is common even in a non-treatment seeking population with depressive disorder. It impacts upon outcome at 6 and 12 months but this is related to the initial severity of depressed mood. Social function is the only independent predictor of outcome and should be assessed separately.
Authors: Christopher Dowrick; C Shiels; H Page; J L Ayuso-Mateos; P Casey; O S Dalgard; G Dunn; V Lehtinen; P Salmon; M Whitehead Journal: Soc Psychiatry Psychiatr Epidemiol Date: 2009-12-25 Impact factor: 4.328