The relationship between executive functions and fluid intelligence in euthymic Bipolar Disorder patients

Abstract

Distinct cognitive deficits have been described in Bipolar disorder (BD), including executive impairments, commonly attributed to frontal dysfunction. However, recent attention has been paid to the heterogeneity of cognitive functioning in this population, suggesting that the executive deficits observed in BD might be due to a loss in fluid intelligence (g). Following our previous line of investigation in multiple neurological and psychiatric conditions we aimed at determining the role of g in frontal deficits in BD. Euthymic BD patients (n = 51) and healthy controls (n = 37) were assessed with Wisconsin Card Sorting Test (WCST), Verbal Fluency, Trail Making Test B (TMTB), a multitasking test, and a theory of mind test. A general cognitive battery was used to derive a measure of g. As in other neuropsychiatric conditions, significant patient-control differences in WCST, Verbal Fluency and TMTB were removed when g was introduced as a covariate. Deficits remained significant in the multitasking test. We suggest that neuropsychological assessment in BD should include tests of general intelligence, together with one or more specific tasks that allow for the assessment of residual frontal deficits, putatively associated with anterior frontal functioning.

Keywords: Frontal deficits; Multitasking; Spearman's g; Theory of mind.

Fig. 1

Scatter plots relating performance in (a) the Wisconsin Card Sorting Test (WCST), (b) Verbal Fluency and (c) Trail Making Test part B (TMTB) to gGTB for patients with bipolar disorder (squares) and controls (triangles). Regression lines (broken for bipolars and solid for controls) reflect the average within-group association of the variables, as determined by ANCOVA, constrained to have the same slope across groups.

Fig. 2

Scatter plots relating performance in (a) the Hotel task, and (b) Faux Pas to gGTB for patients with bipolar disorder (squares) and controls (triangles). Regression lines (broken for bipolars and solid for controls) reflect the average within-group association of the variables, as determined by ANCOVA, constrained to have the same slope across groups.