Depictions of people with schizophrenia as violent or dangerous are incorrect and are major drivers of stigma (Torrey, 2011; Walsh et al., 2002; Whiting et al., 2022). The vast majority of patients with schizophrenia are not violent or aggressive and most people with schizophrenia who do commit violent acts do so rarely (Silverstein et al., 2015). However, rates of aggression and violence are higher in patients with schizophrenia than in the general population (Silverstein et al., 2015; Whiting et al., 2022). Accurately assessing risk of violence is particularly important for inpatient psychiatric treatment and multiple meta-analyses have demonstrated that a diagnosis of schizophrenia is an independent risk factor for assaults on nursing staff (Iozzino et al., 2015; Jang et al., 2022). Reducing the likelihood of aggressive acts is one goal of treatment and clinician assessments of risk of aggressive behavior are often used to guide treatment decisions. However, such judgments are clinician dependent and subject to individual biases (Spector, 2001). An objective biological correlate of aggression would allow for interventions to reduce violence to be applied more effectively, would decrease the potential for overmedication in minoritized communities and would shed light on the underlying pathology that drives aggressive behavior (Bonner et al., 2002).

The relationship between schizophrenia and risk for aggression is complicated and mediated by societal and biological factors. Aggression is a multifaceted construct with distinct types of aggression related to distinct neural circuits (Hoptman, 2015). Multiple studies have demonstrated that cognitive deficits are a major driver of aggression in people with schizophrenia (Ahmed et al., 2018; Mcdermott and Holoyda, 2014; Stahl, 2013). Increased impulsivity, especially in emotionally charged settings, is highly correlated with violence in patients with schizophrenia (Juríčková et al., 2022; Quanbeck et al., 2007). The precise neural correlates of impulsive aggression in schizophrenia are poorly understood although neuroimaging studies suggest dysfunction within a corticothalamic loop in which bottom up impulses are generated by the amygdala and striatum and are regulated by top down control from the medial and frontal cortices (Fjellvang et al., 2018; Leclerc et al., 2018). Additionally, overall burden of psychotic symptoms contributes to the risk for aggressive behavior (Juríčková et al., 2022; Stahl, 2013).

Electroencephalography (EEG) has been used to study the neurobiology of aggression, in part because it is relatively easy to apply. Studies of people with schizophrenia who have committed violent acts have produced inconsistent results with various reports of associations with delta, beta, and alpha power (Convit et al., 1991; Palix et al., 2016; Schug et al., 2011). The alpha findings are particularly compelling given that alpha activity is the dominant feature of resting state EEG. Alpha waves are 8–12 Hz oscillations that are largest over the occipital cortex (Nunez et al., 2001). The peak of activity within the alpha band (the individual alpha peak frequency, IAPF) is correlated with multiple aspects of cognition. IAPF is correlated with performance on working memory tasks (Clark et al., 2007), reading comprehension (Rathee et al., 2020), and non-verbal IQ (Dickinson et al., 2018). Additionally, IAPF is decreased in patients with schizophrenia (Harris et al., 2006; Murphy and Öngür, 2019; Ramsay et al., 2021). Given the connections between aggression, cognition, alpha activity, and schizophrenia, we hypothesized that, in patients with schizophrenia, slower IAPF would be associated with higher levels of impulsivity and therefore aggression.

In addition to analyzing EEG based on frequency content, EEG can also be analyzed spatially. The spatial patterns of scalp voltages can be parsimoniously explained by a small set of recurring EEG topographies called microstates (Michel and Koenig, 2018a). Four canonical microstate topographies (A, B, C, D) have been repeatedly found by different groups (Koenig et al., 2023). Each microstate topography overlaps with one or more resting state functional magnetic resonance imaging networks (Custo et al., 2017). Schizophrenia is associated with increased microstate C and decreased microstate D (da Cruz et al., 2020). These microstates have been linked to salience and attention respectively. We hypothesized that more severe schizophrenia-related neuropathology, that is increased microstate C and decreased microstate D, would be linked to increased aggression. We previously showed that psychosis is associated with abnormal transitions between microstates (Murphy et al., 2020). These transition abnormalities may partially explain abnormal thoughts and behaviors in schizophrenia. We therefore hypothesized that transitions between microstates would be more disordered in aggressive patients with schizophrenia consistent with improper top-down processing of sensory input and subsequent incongruent emotion responsiveness including misplaced aggression.