Cognitive network reconstruction in individuals who use opioids compared to those who do not: Topological analysis of cognitive function through graph model and centrality measures.

Introduction: Cognitive dysfunction related to opioid use disorder (OUD) requires investigation of the interconnected network of cognitive domains through behavioral experiments and graph data modeling. Methods: We conducted n-back, selective and divided attention, and Wisconsin card sorting tests and reconstructed the interactive cognitive network of subscales or domains for individuals who use opioids and controls to identify the most central cognitive functions and their connections using graph model analysis. Each two subscales with significant correlations were connected by an edge that incorporated in formation of interactive networks. Each network was analyzed topologically based on the betweenness and closeness centrality measures. Results: Results from the network reconstructed for individuals who use opioids show that in the divided attention module, reaction time and number of commission errors were the most central subscales of cognitive function. Whereas in controls, the number of correct responses and commission errors were the most central cognitive measure. We found that the subscale measures of divided attention module are significantly correlated with those of other tests. These findings corroborate that persons who use opioids show impaired divided attention as higher reaction time and errors in performing tasks. Divided attention is the most central cognitive function in both OUD subjects and controls, although differences were observed between the two groups in various subscales. Discussion: Although equal proportions of males and females may be used in future studies, divided attention and its subscales may be the most promising target for cognitive therapies, treatments and rehabilitation as their improvement can enhance overall cognitive domain performance.