Classification of cognitive performance in bipolar disorder.

OBJECTIVE: To understand the etiology of cognitive impairment associated with bipolar disorder, we need to clarify potential heterogeneity in cognitive functioning. To this end, we used multivariate techniques to study if the correlation structure of cognitive abilities differs between persons with bipolar disorder and controls. METHOD: Clinically stable patients with bipolar disorder (type I: n = 64; type II: n = 44) and healthy controls (n = 86) were assessed with a wide range of cognitive tests measuring executive function, speed, memory, and verbal skills. Data were analysed with multivariate techniques. RESULTS: A distinct subgroup (∼30%) could be identified that performed significantly poorer on tests concerning memory function. This cognitive phenotype subgroup did not differ from the majority of bipolar disorder patients with respect to other demographic or clinical characteristics. CONCLUSIONS: Whereas the majority of patients performed similar to controls, a subgroup of patients with bipolar disorder differed substantially from healthy controls in the correlation pattern of low-level cognitive abilities. This suggests that cognitive impairment is not a general trait in bipolar disorder but characteristic of a cognitive subgroup. This has important clinical implications for cognitive rehabilitation and remediation.

Cognitive functioning in clinically stable patients with bipolar disorder I and II.

OBJECTIVES: Bipolar disorder is accompanied by cognitive impairments, which persists during euthymic phases. The purpose of the present study was to identify those neuropsychological tests that most reliably tell euthymic bipolar patients and controls apart, and to clarify the extent to which these cognitive impairments are clinically significant as judged from neuropsychological norms. METHODS: Patients with bipolar disorder (type I: n = 64; type II: n = 44) and controls (n = 86) were examined with a comprehensive neuropsychological test battery yielding 47 measures of executive functioning, speed, memory, and verbal skills. Multivariate analysis was used to build a model of cognitive performance with the ability to expose underlying trends in data and to reveal cognitive differences between patients and controls. RESULTS: Patients with bipolar disorder and controls were partially separated by one predictive component of cognitive performance. Additionally, the relative relevance of each cognitive measure for such separation was decided. Cognitive tests measuring set shifting, inhibition, fluency, and searching (e.g., Trail Making Test, Color-Word) had strongest discriminating ability and most reliably detected cognitive impairments in the patient group. CONCLUSIONS: Both bipolar disorder type I and type II were associated with cognitive impairment that for a sizeable minority is significant in a clinical neuropsychological sense. We demonstrate a combination of neuropsychological tests that reliably detect cognitive impairment in bipolar disorder.