Cognitive variability in bipolar I disorder: A cluster-analytic approach informed by resting-state data. (15th September 2019)

• Record Type: Journal Article
• Authors: Kollmann, Bianca; Yuen, Kenneth; Scholz, Vanessa; Wessa, Michèle
• Is Part Of: Neuropharmacology. Volume 156(2019)
• Publication Date: 2019-09-15
• Subjects: Bipolar I disorder -- Cognitive profile -- Cluster analysis -- Executive functions -- Resting-state; Neuropsychopharmacology -- Periodicals; Autonomic Agents -- Periodicals; Neuropsychopharmacologie -- Périodiques; Neuropsychopharmacology; Periodicals; Electronic journals; 615.78
• DOI: 10.1016/j.neuropharm.2019.03.028
• Languages: English
• ISSNs: 0028-3908

Abstract

Abstract: Background: While the presence of cognitive performance deficits in bipolar disorder I (BD-I) is well established, there is no consensus about which cognitive abilities are affected. Heterogeneous phenotypes displayed in BD-I further suggest the existence of subgroups among the disorder. The present study sought to identify different cognitive profiles among BD-I patients as well as potentially underlying neuronal network changes. Methods: 54 euthymic BD-I patients underwent cognitive testing and resting state neuroimaging. Hierarchical cluster-analysis was performed on executive function scores of bipolar patients. The derived clusters were compared against 54 age-, gender- and IQ-matched healthy controls (HC) to facilitate the interpretation of results. Further, resting state network properties were compared to identify differences probably underlying cognitive profiles. Results: A three-cluster solution emerged. Cluster 1 (n = 22) was characterized by deficits in cognitive flexibility and motor inhibition, cluster 2 (n = 12) displayed impulsive decision-making, while cluster 3 (n = 20) showed good visuospatial planning. Weaker connections in cluster 1 compared to cluster 2 were found between regions activated during tasks cluster 1 showed deficits on. Cluster 3 had a higher modularity than cluster 2, which correlated positively with problem solving performance and risk-taking in this cluster. Conclusion: Obtained clusters showed distinct cognitive profiles, characterized by deficits and strengths, most of which remained precluded in a general comparison. Weaker interregional connections and separated subnetworks might underly behavioral deficits and strengths, respectively. The findings help explain the phenotypic heterogeneity observed in BD-I. This article is part of the Special Issue entitled 'Current status of the neurobiology of aggression and impulsivity'. Highlights: 3 distinct clusters among BD-I patients based on cognitive performance. Clusters show differing cognitive profiles, irrespective of performance deficits. Clusters also differ in network connectivity in regions of interest related to cognitive tasks investigated. Clusters show distinct connectivity hubs within the brain. Results ameliorate the understanding of the different phenotypes observed for BD-I.