Acute Effects of Esketamine on the EEG Power Spectrum and Signal Complexity in Prolonged Disorders of Consciousness: A Prospective Exploratory Cohort Study

Prolonged disorders of consciousness (pDoC) have limited effective treatments, and the mechanisms by which psychedelic or psychedelic-like agents might influence recovery remain uncertain. The paper is framed around the entropic brain hypothesis, which suggests that such agents may increase brain complexity and thereby alter consciousness-related brain states. Earlier work had reported EEG changes with ketamine, psilocybin, or related compounds, but it was not clear how esketamine would affect neurophysiological markers in pDoC, or whether any acute EEG effects would be accompanied by behavioural improvement. Liang and colleagues set out to characterise the acute effects of a subanaesthetic esketamine infusion on EEG spectral features and signal complexity in patients with pDoC. The study aimed to examine whether esketamine would reorganise resting-state EEG power and increase Lempel-Ziv complexity, and whether these changes would differ between vegetative state/unresponsive wakefulness syndrome (VS/UWS) and minimally conscious state (MCS) patients. The work is presented as an exploratory cohort analysis focused on within-patient neurophysiological responses.

This was a prospective exploratory cohort study conducted at Beijing Tiantan Hospital, Capital Medical University, and approved by the local ethics committee. It was registered on ClinicalTrials.gov. Although the broader protocol had initially been designed as a parallel-group study, the report focuses on the cohort of patients with pDoC because the comparison group could not be retained due to adverse events and EEG quality issues. Recruitment occurred from July 2024 to July 2025, and legal surrogates provided written informed consent. The study enrolled 27 adults aged 18 to 65 years with pDoC after acquired brain injury; 22 were included in the final analysis after exclusion of 5 patients with insufficient EEG signal quality. Patients were clinically stable and were scheduled for elective spinal cord stimulator implantation, but all EEG recordings were obtained preoperatively and before general anaesthesia to minimise perioperative confounding. Participants fasted for 8 hours, received oxygen via tracheostomy interface, and underwent continuous physiological monitoring during the experiment. Esketamine was administered as a continuous intravenous infusion at 0.3 mg/kg/h for 1 hour, using a subanaesthetic dose selected on the basis of earlier work showing neurophysiological effects without general anaesthesia. Resting-state EEG was recorded for 5 minutes at three time points: baseline, 1 hour after infusion onset, and 30 minutes after discontinuation. Coma Recovery Scale-Revised (CRS-R) assessments were performed by trained assessors who were blinded to the EEG results. EEG was collected with a 64-channel system using the International 10-20 montage. The authors retained only artifact-free data, using 180 seconds of clean EEG for each condition. The scalp was divided into 7 regions, and analyses focused on power spectral density in delta, theta, alpha, beta, and gamma bands, expressed as relative power, plus signal complexity measured with Lempel-Ziv complexity (LZC). Because this was exploratory, sample size was based on feasibility rather than formal power calculation. The authors used repeated-measures Friedman tests across the three time points, with Wilcoxon signed-rank tests for post hoc pairwise comparisons. Between-group comparisons used Mann-Whitney U tests. Baseline characteristics were analysed with Fisher exact tests, independent-samples t tests, or Mann-Whitney U tests as appropriate. Multiple comparisons were controlled using the false discovery rate (FDR), and corrected P values below 0.05 were considered statistically significant.

Of the 27 enrolled patients, 22 were included in the final EEG analysis. The sample was split into VS/UWS and MCS groups. The groups did not differ significantly in age, sex, body mass index, or aetiology, but pDoC duration was longer in the MCS group. CRS-R total and subscale scores were higher in the MCS group than in the VS/UWS group at baseline. No significant changes in CRS-R scores were observed across baseline, esketamine, and recovery time points. Esketamine produced a marked reorganisation of the EEG power spectrum. At the whole-brain level, delta relative power decreased significantly, while beta and gamma relative power increased significantly after FDR correction. The delta reduction was most evident in frontal, parietal, and left temporal regions, and some effects persisted into the recovery period. The gamma increase was widespread, whereas beta changes were more regionally selective. Theta did not change significantly, and alpha changes did not remain significant after correction in the overall cohort. Subgroup analyses suggested different response patterns by diagnosis. In MCS patients, esketamine mainly affected high-frequency activity, with increases in gamma relative power that did not survive FDR correction. In VS/UWS patients, there was a significant global reduction in delta relative power and a widespread increase in alpha relative power, along with beta-band increases in several cortical regions. Gamma- and theta-band changes in VS/UWS were not significant after correction. Esketamine was also associated with increased Lempel-Ziv complexity. The main increases were observed in the parietal and occipital regions at the uncorrected level. LZC values were higher in the MCS group than in the VS/UWS group across states and regions. Both groups showed increases in LZC after esketamine, but the time course differed: the changes were transient in VS/UWS, whereas in MCS they persisted into the recovery period, most prominently in the occipital region.

The authors interpret the findings as evidence that subanaesthetic esketamine acutely alters brain activity in pDoC, reshaping both spectral organisation and signal complexity without producing measurable behavioural improvement. They argue that the observed pattern of delta suppression with beta and gamma enhancement resembles EEG changes reported previously with ketamine, psilocybin, and other psychedelic states, and may reflect increased cortical excitability and reduced synchronisation of slower oscillations. In this framing, the results suggest a shared oscillatory signature across classic and non-classic psychedelic agents. They highlight the selective alpha increase in VS/UWS patients and note that alpha activity is closely linked to thalamocortical integrity and information gating. The authors suggest this may reflect partial restoration of thalamocortical interactions or more favourable brain-state transitions, although they do not claim this as proof of recovery. They also interpret the LZC increases, especially the more sustained response in MCS patients, as consistent with the entropic brain hypothesis, which proposes that psychedelics expand the repertoire of possible brain states by increasing neural entropy. The more transient response in VS/UWS is presented as possibly reflecting greater structural disconnection and less preserved network integrity. The absence of CRS-R change is emphasised as an important dissociation between neurophysiological change and overt behaviour. The authors suggest that psychedelic agents may induce a form of 'disconnected consciousness', in which internal brain dynamics change without corresponding behavioural expression, and that the CRS-R alone may be insufficient to capture such effects. They argue that future studies should use multimodal outcomes combining neurophysiology, neuroimaging, and behaviour. Several limitations are acknowledged. The study lacked a placebo control group, so causal inference is limited. Subgroup analyses, especially in MCS, were underpowered because of the small sample. Only the acute effects of a single esketamine dose were examined, leaving the risk-benefit balance unclear. The authors also note that although CRS-R assessors were blinded to EEG results, they could not be blinded to treatment timing without a placebo, which may have introduced observer bias. They further caution that multiple comparisons and the small sample make the findings statistically fragile, and they describe the results as hypothesis-generating. Finally, because the analysis was limited to resting-state EEG within a broader protocol, generalisability is restricted to acute neurophysiological effects in pDoC. The authors suggest that future work should include task-based or perturbational paradigms to better probe consciousness-related processing.

The authors conclude that a single subanaesthetic infusion of esketamine is associated with clear acute changes in resting-state EEG in pDoC, including reduced delta power, increased beta and gamma power, and higher Lempel-Ziv complexity in parietal and occipital regions. They also state that the pattern of EEG effects differs between VS/UWS and MCS patients, but that these neurophysiological changes were not accompanied by behavioural improvement.