EEG Correlates of the Confusional State after Traumatic Brain Injury

Abstract

Post-traumatic amnesia (PTA), recently conceptualized as part of the broader syndrome known as post-traumatic confusional state (PTCS), marks a critical phase of recovery following traumatic brain injury (TBI). Indeed, this state is characterized not only by anterograde memory impairment but also by disorientation, agitation, and attention deficits. Given the phenotypic overlap between PTA/PTCS and delirium—both marked by fluctuating cognitive and attentional disturbances—electroencephalography (EEG) represents a promising tool for elucidating shared pathophysiological mechanisms. While delirium is typically associated with diffuse EEG slowing and the presence of slow-wave activity (SWA), thought to reflect underlying global cortical disruption, it remains unclear whether PTCS exhibits similar EEG underpinnings. In this prospective longitudinal study, we assessed dynamic EEG correlates of PTCS using the grand total EEG (GTE) score, a composite measure that incorporates background slowing and superimposed SWA. We enrolled 42 consecutive TBI patients (mean age = 40.3 ±15 years) classifying them at baseline (T0) into two groups based on the Confusion Assessment Protocol (CAP): those in PTA/PTCS (N = 22; median time from injury 24 days; median CAP total score 5) and those already emerged from PTA/PTCS (i.e., TBI controls, N = 20; median time from injury 24 days; median CAP total score 0). At T0, patients with PTA/PTCS exhibited significantly higher baseline GTE scores compared with TBI controls, 16.6 ± 4.5 versus 5.1 ± 2.8; t(35.75) = 10.04, p < 0.0001; d = 3.04, reflecting severe EEG abnormalities characterized by diffuse slowing and disrupted rhythmic activity, as captured by the GTE subdomains. Longitudinal follow-up (T1) at emergence from PTCS revealed a significant EEG improvement paralleling clinical recovery, with GTE scores dropping from 16.5 (interquartile range [IQR]: 6.5) to 8, IQR: 3.75; t(21) = 8.03, p < 0.0001; d = 1.71, confirming EEG’s sensitivity to dynamic clinical changes. Furthermore, the severity of EEG abnormalities at follow-up (T1) significantly correlated with the total duration of PTA/PTCS (ρ = 0.56, p < 0.0001), underscoring EEG’s potential as an objective biomarker for disease burden and for monitoring recovery trajectories. Notably, these findings were independent of pharmacological confounders, as medication regimens were not significantly different across groups and time points. Our results support a reconceptualization of PTA/PTCS as a functional (i.e., non-structural) encephalopathy that shares key clinical and neurophysiological features with delirium, with EEG slowing reflecting widespread, often reversible cortical dysfunction. By capturing these transient yet clinically critical changes, clinical EEG—quantified via the granular, multifaceted GTE—offers a novel tool for diagnosing PTA/PTCS, stratifying its severity, and objectively monitoring its evolution in intensive care unit and subacute rehabilitation settings.