EEG Gamma Band Alterations and REM-like Traits Underpin the Acute Effect of the Atypical Psychedelic Ibogaine in the Rat

Ibogaine is an atypical psychedelic alkaloid with reported long-lasting antiaddictive effects in humans and robust preclinical evidence in rodents. Subjective reports characterise the ibogaine experience as an intense, dream-like episode while awake, involving memory retrieval and imagination but lacking many of the identity and perceptual distortions typical of classical psychedelics; accordingly, ibogaine is often called oneirogenic. The authors note a hypothesised link between these oneiric experiences and REM sleep, since vivid dreaming predominates during REM and REM-related neural processes (for example, plasticity and memory reconsolidation) could plausibly contribute to antiaddictive effects. González and colleagues set out to test whether the electrocortical activity induced acutely by ibogaine in rats shows features of REM sleep. Building on their prior observation that ibogaine increases wakefulness while suppressing NREM and REM sleep, the investigators applied a computational analysis of intracranial electroencephalogram (iEEG) recordings to characterise spectral power, inter-regional synchronization, temporal complexity, cross-frequency coupling, and state classification during the acute post‑administration period. The aim was to determine whether the ibogaine-induced waking state exhibits REM-like electrophysiological traits, with a focus on gamma‑band activity.

The study used six adult Wistar rats chronically implanted with intracranial electrodes over the olfactory bulb (OB), primary motor (M1), primary somatosensory (S1) and secondary visual cortex (V2), plus a cerebellar reference and neck EMG electrodes. Animals were maintained on a 12‑hour light/dark cycle with recordings conducted during the light period. Surgical and post‑operative care followed institutional and national animal welfare guidelines. Ibogaine was purified from Tabernanthe iboga and administered intraperitoneally at 40 mg/kg, a dose previously shown by the authors to produce pronounced sleep and motor effects. Polysomnographic data were acquired at 1,024 Hz and scored in 10‑second epochs using standard criteria for Wake, NREM and REM based on cortical rhythms and EMG. For analysis of the acute ibogaine effect, the investigators selected artifact‑free wake epochs from the first two hours after injection when continuous wakefulness and abnormal motor/autonomic effects were observed; NREM epochs were taken across six hours due to NREM reduction after ibogaine. REM epochs following ibogaine were not used because REM was absent in several animals, so REM comparisons used control REM epochs. Spectral power was estimated using Welch’s method (1‑s windows, 1 Hz resolution) and spectrograms via multitaper methods. Spectra were whitened (power multiplied by frequency) and normalised to relative power. Conventional frequency bands were defined as delta 1–4 Hz, theta 5–10 Hz, sigma 11–14 Hz, beta 15–29 Hz and gamma 30–100 Hz. Inter-electrode synchronization was quantified with magnitude‑squared coherence and time‑frequency coherograms. To identify significant frequency clusters rather than pre‑defined bands, a cluster‑based permutation test was used: paired t‑tests across frequencies, grouping consecutive significant frequencies into clusters, and generating null distributions by label randomisation (10,000 permutations). Temporal complexity was assessed by permutation entropy after down‑sampling iEEG to 128 Hz; permutation entropy quantifies the diversity of ordinal patterns in a time series and is robust to noise and short recordings. Phase–amplitude coupling within regions was measured with the modulation index, computing phase of slow (1–15 Hz) bands and amplitude of faster (40–180 Hz) bands. Finally, a supervised multi‑layer perceptron (patternnet in Matlab) with 10 hidden layers was trained on control data to classify Wake versus REM using gamma power (OB, M1r, M1l, S1r) and coherence values (nine electrode pairs). The trained network was then used to classify ibogaine wake epochs. Statistical comparisons used paired t‑tests (two‑tailed for power and entropy, one‑tailed for coherence) and the cluster permutation procedure described above.

Recordings from six rats showed robust changes in iEEG following 40 mg/kg i.p. ibogaine. During the first two hours after administration, wakefulness predominated and gamma oscillations (30–80 Hz) increased markedly in OB, M1, S1 and V2, with the gamma elevation lasting at least two hours. Mean theta power also increased, and the theta peak frequency shifted downward from about 9 Hz to 8 Hz, particularly apparent in S1 and V2. A decrease in very high‑frequency power (>100 Hz up to 512 Hz) was observed in M1, S1 and V2, but the authors note that this likely reflects changes in muscular activity rather than a true neural spectral peak. Despite local increases in gamma power, inter‑regional synchronization in the gamma band decreased. Cluster‑based permutation analyses showed significant coherence reductions across sigma‑beta, gamma and high‑frequency ranges in multiple cortical areas including OB, M1 and S1. Specifically, inter‑regional gamma coherence decreased in 9 of 21 electrode pairs, for example between right OB and right S1, and in inter‑hemispheric M1–M1 and M1–S1 combinations; some intra‑hemispheric pairs did not show the same reduction. Permutation entropy, estimating temporal complexity after down‑sampling to 128 Hz, was significantly reduced in OB, M1 and S1 during ibogaine wakefulness compared with control wakefulness, indicating lower dynamical motif diversity; no significant change occurred in V2. Because the down‑sampling retains primarily gamma‑band content for this measure, the complexity reduction pertains chiefly to gamma dynamics. When compared to physiological REM sleep (from control recordings), ibogaine wakefulness showed similar gamma power between the two states, while theta, sigma and beta powers were lower under ibogaine than in REM. High‑frequency (>100 Hz) power was higher during ibogaine wakefulness, probably due to muscle activity. Gamma coherence and permutation‑entropy values were broadly similar between ibogaine wakefulness and REM sleep, with the exception that permutation entropy in M1 was slightly larger during ibogaine wakefulness. Using the neural network classifier trained to separate Wake and REM from control data, most ibogaine wake epochs were labelled as REM: in 5 of 6 animals the majority of ibogaine epochs were classified as REM, and in 3 animals all ibogaine epochs were classified as REM. These results indicate that features of gamma power, coherence and complexity in the ibogaine‑induced waking state closely resemble REM sleep.

The investigators interpret their findings as evidence that acute systemic administration of ibogaine in rats produces a waking brain state with electrophysiological traits characteristic of REM sleep. In particular, the state is marked by increased local gamma power in OB, M1 and S1 alongside reduced inter‑regional gamma coherence and diminished temporal complexity; these are features shared with REM sleep and were also partially observable in NREM after ibogaine (for example, OB gamma power increases and reduced gamma coherence in some derivations). Comparing to other psychedelics, the authors note parallels with human studies of serotonergic 5‑HT2A agonists (LSD, psilocybin), which reduce alpha/beta power and connectivity; ibogaine likewise reduced sigma and beta coherence, although the principal cortical locus differs between species (olfactory areas in the rat versus visual cortex in humans). Cross‑frequency coupling of gamma to slower rhythms was not altered by ibogaine, suggesting that slow sensory inputs (for example respiratory‑linked OB oscillations) may still reach sensory cortex but be integrated differently under the drug. Pharmacologically, the authors emphasise that ibogaine is rapidly metabolised to noribogaine and that both compounds are present in the brain during the recorded period. They estimate free brain concentrations for ibogaine/noribogaine of roughly 3.1–3.5 μM and discuss receptor interactions consistent with the electrophysiology: noncompetitive NMDA receptor antagonism by ibogaine (Ki ~1.0–5.2 μM) and noribogaine (higher Ki) could contribute to increased gamma power and decreased coherence, an effect resembling ketamine. Enhanced serotonergic transmission via SERT inhibition by noribogaine (IC50 ~50–300 nM) and modest 5‑HT2A interaction could also explain some similarities to classical psychedelics. The authors acknowledge that multiple receptor systems may contribute and that disentangling ibogaine versus noribogaine effects will require further experiments. Limitations discussed include the lack of prior quantitative rodent iEEG data for classical psychedelics (which forced some comparisons to human studies), the practical absence of REM sleep after ibogaine (precluding within‑drug REM observations), and the need for future work to parse pharmacological contributions. Conceptually, the authors propose that REM sleep, psychosis and psychedelic states may share overlapping electrocortical characteristics that can be reached by distinct neurochemical routes, and they present their results as supporting the long‑standing oneirogenic description of the ibogaine experience.