Abstract
BACKGROUND: Aberrance in switching from default mode network (DMN) to fronto-parietal network (FPN) is proposed to underlie working memory deficits in subjects with substance use disorders, which can be studied using neuro-imaging techniques during cognitive tasks. The current study used EEG to investigate pre-stimulus microstates during the performance of Sternberg's working memory task in subjects with substance use disorders. METHODS: 128-channel EEG was acquired and processed in ten age and gender-matched subjects, each with alcohol use disorder, opioid use disorder, and controls while they performed Sternberg's task. Behavioral parameters, pre-stimulus EEG microstate, and underlying sources were analyzed and compared between subjects with substance use disorders and controls. RESULTS: Both alcohol and opioid use disorder subjects had significantly lower accuracy (P < 0.01), while reaction times were significantly higher only in subjects of alcohol use disorder compared to controls (P < 0.01) and opioid use disorder (P < 0.01), reflecting working memory deficits of varying degrees in subjects with substance use disorders. Pre-stimulus EEG microstate revealed four topographic Maps 1-4: subjects of alcohol and opioid use disorder showing significantly lower mean duration of Map 3 (visual processing) and Map 2 (saliency and DMN switching), respectively, compared to controls (P < 0.05). CONCLUSION: Reduced mean durations in Map 3 and 2 in subjects of alcohol and opioid use disorder can underlie their poorer performance in Sternberg's task. Furthermore, cortical sources revealed higher activity in both groups of substance use disorders in the parahippocampal gyrus- a hub of DMN; superior and middle temporal gyri associated with impulsivity; and insula that maintains balance between executive reflective system and impulsive system. EEG microstates can be used to envisage neural underpinnings implicated for working memory deficits in subjects of alcohol and opioid use disorders, reflected by aberrant switching between neural networks and information processing mechanisms.