Electrophysiological indices of interference resolution covary with individual fluid intelligence: investigating reactive control processes in a 3-back working memory task ☆ Bernardo Perfetti a, ⁎, Sara Varanese b , Elisa Mancino a , Pasqua Mercuri a , Marcello Tesse c , Raffaella Franciotti d , Laura Bonanni a , Astrid Thomas a , Marco Onofrj a a Ce.S.I – University of Chieti-Pescara, Department of Neuroscience and Imaging, Via Luigi Polacchi, 11 (ex Via Colle dell'Ara) – 66100 Chieti, Italy b New York University, Department of Neurology, Division of Movement Disorders, School of Medicine, 550 First Avenue RR311, New York, NY 111016, USA c Institute of Psychiatry, King's College London, Department of Psychological Medicine, Weston Education Centre, Cutcombe Road, SE5 9RJ London, UK d ITAB – University of Chieti-Pescara, Department of Neuroscience and Imaging, Via Luigi Polacchi, 12 (ex Via Colle dell'Ara) – 66100 Chieti, Italy abstract article info Article history: Accepted 16 February 2014 Available online 25 February 2014 Keywords: Attention Executive function Problem solving Cognition Cortical synchronization Evoked potentials It has been proposed that the well-established relationship between working memory (WM) and fluid intelli- gence (gf) is mediated by executive mechanisms underlying interference control. The latter relies upon the integ- rity of a frontoparietal brain network, whose activity is modulated by general cognition. In regards to the chronology of this activation, only few EEG studies investigated the topic, although none of them examined the regional interaction or the effects of individual differences in gf. The current investigation sought at extending previous research by characterizing the EEG markers (temporal activation and regional coupling) of interference control and the effects of the individual variation in gf. To this end, we recorded the EEG activity of 33 participants while performing verbal and spatial versions of a 3-back WM task. In a separate session, participants were admin- istered with a test of fluid intelligence. Interference-inducing trials were associated with an increased negativity in the frontal scalp region occurring in two separate time windows and probably reflecting two different stages of the underlying cognitive process. In addition, we found that scalp distribution of such activity differed among in- dividuals, being the strongest activation of the left and right frontolateral sites related to high gf level. Finally, high- and low-gf participants showed different patterns in the modulation of regional connectivity (electrodes coherence in the range of 4.5–7.5 Hz) according to changes in attention load among types of trials. Our findings suggest that high-gf participants may rely upon effective engagement and modulation of attention resources to face interference. © 2014 Elsevier Inc. All rights reserved. Introduction Working Memory (WM) and Fluid Intelligence (gf) are fundamental to higher cognitive functioning. The former has been defined as the abil- ity to hold and manipulate a limited amount of information in memory (Baddeley, 1986), while the latter has been described as a general psy- chometric factor (Carroll, 2003) related to novel problem solving skills and to the optimization of individual performance (Duncan, 2005, 1995). In spite of the different definitions, WM and gf have consistently shown a strong association, likely reflecting the recruitment of common underlying cognitive mechanisms. Recent evidence suggests that these mechanisms are the attentional processes engaged during resolution of interference (Burgess et al., 2011). For instance, behavioral indices of interference resolution covary with both WM span and gf score (Kane et al., 2007; May et al., 1999; Perfetti et al., 2011) and interference-related brain activity has been revealed in prefrontal and parietal cortices as for WM and gf (Gray et al., 2003; Jonides and Nee, 2006; Perfetti et al., 2009). Most importantly, a recent study (Burgess et al., 2011) reported that interference-related performance and brain activity accounted for a significant proportion of the shared variance be- tween gF and WM span, highlighting the importance of investigating in- terference control to ultimately provide a mechanistic model of higher cognitive functioning. Much of the imaging research on the topic has employed functional magnetic resonance (fMRI), revealing the recruitment of a cortical NeuroImage 93 (2014) 146–153 ☆ The work was performed at the University of Chieti-Pescara “G.d'Annunzio”, Centro Scienze dell'Invecchiamento, (Ce.S.I.) Via Luigi Polacchi, 11 (ex Via Colle dell'Ara) – 66100 Chieti (Italy). ⁎ Corresponding author at: stanza 237, Centro Scienze dell'Invecchaimento, Via Luigi Polacchi, 11 (ex Via Colle dell'Ara) – 66100 Chieti, Italy. Fax: +39 0871 562019. E-mail addresses: perfetti@inwind.it (B. Perfetti), sara.varanese@nyumc.org (S. Varanese), elisa.mancino@libero.it (E. Mancino), mercuripsycho@libero.it (P. Mercuri), marcello.tesse@email.it (M. Tesse), raffaella.franciotti@itab.unich.it (R. Franciotti), l.bonanni@unich.it (L. Bonanni), athomas@unich.it (A. Thomas), onofrj@unich.it (M. Onofrj). http://dx.doi.org/10.1016/j.neuroimage.2014.02.020 1053-8119/© 2014 Elsevier Inc. All rights reserved. Contents lists available at ScienceDirect NeuroImage journal homepage: www.elsevier.com/locate/ynimg