Ketamine-induced changes in accumbal glutamate and their association with altered states of consciousness

Ketamine is a non-competitive NMDA receptor antagonist with recognised neuroplastic and antidepressant effects, and it has also shown promise in substance use disorders (SUDs). A central uncertainty is how ketamine produces these effects at the neurochemical level. The paper notes that ketamine can modulate glutamatergic transmission and that disturbed glutamate homeostasis has been implicated in psychiatric disorders, particularly in corticostriatal circuits involving the nucleus accumbens (NAc), a region important for reward and craving. However, evidence for ketamine’s effects on glutamate in humans remains mixed, and little is known about whether changes in NAc glutamate relate to ketamine’s altered subjective states. The study aimed to test whether a single intravenous dose of ketamine increases glutamate in the left NAc of healthy volunteers, measured with proton magnetic resonance spectroscopy (1H-MRS), and whether any glutamate changes are associated with altered states of consciousness measured using the Five-Dimensional Altered States of Consciousness (5D-ASC) questionnaire. The authors describe this as a pilot study designed to help clarify a possible mechanism relevant to future work in SUDs and other psychiatric disorders.

This was a pilot study conducted in 10 healthy adults aged 18 to 55 years; 5 were female and the mean age was 29.4 ± 6.8 years. Participants were screened for extensive psychiatric, neurological, medical, medication-related, pregnancy-related, and MRI-related exclusion criteria. The study took place at the Psychiatric University Clinic Zurich and had ethics and regulatory approval. Written informed consent was obtained from all participants. The study involved two visits: a screening visit and an infusion/MRS visit. On the study day, participants underwent two identical 22-minute 1H-MRS scans of the left NAc, one at baseline and one during ketamine administration. During both scans they listened to a standardised instrumental music playlist developed for psychedelic-assisted therapy. Before the infusion, participants also completed a 10-minute audio-guided mindfulness-based exercise. Blood pressure was measured before and after infusion, and participants were monitored for 90 minutes afterwards. During the second MRS scan, participants received racemic ketamine (Ketalar) intravenously over 40 minutes at a total dose of 0.71 mg/kg bodyweight. Five participants received a steady infusion, and five received a bolus after 10 minutes of steady infusion, followed by return to the steady rate for the remaining 20 minutes. The bolus was included to assess tolerability and feasibility for a later clinical trial. Subjective effects were assessed one hour after infusion using the 5D-ASC, which covers five main dimensions, 11 subscales, and a global score. Neuroimaging was performed on a 3.0 T scanner using a tailored sLASER 1H-MRS sequence for the left NAc voxel. Glutamate was quantified relative to the unsuppressed water peak. The researchers analysed whole-scan averages and also three overlapping sub-blocks to capture temporal changes during infusion. Spectral quality was assessed using signal-to-noise ratio, linewidth, and Cramér-Rao lower bounds. Statistical analyses included a one-tailed paired t-test for whole-block baseline versus ketamine glutamate, a one-tailed two-way ANOVA for sub-block effects, and Spearman rank correlations between glutamate change and 5D-ASC scores, with Benjamini-Hochberg false discovery rate correction. Baseline glutamate was also explored as a predictor of subjective effects.

Spectral quality was reported as acceptable in all participants: signal-to-noise ratio exceeded 18, linewidth was below 10 Hz, and Cramér-Rao lower bounds for glutamate were below 7%. No scans were excluded for quality reasons, and the average overlap between baseline and ketamine voxels was 77%. At the group level, ketamine did not significantly increase NAc glutamate. In the whole-block analysis, the paired t-test showed no difference between baseline and ketamine (t(9) = -0.24, p = 0.59, d = -0.08, 95% CI [-1.14, 0.58]). Mean glutamate was 8.2 mmol/kg at baseline and 8.3 mmol/kg during ketamine. The sub-block analysis was also null: there was no significant main effect of condition (F(1, 9) = 0.01, p = 0.92), no significant sub-block effect (F(2,18) = 1.95, p = 0.17), and no significant interaction (F(2,18) = 0.51, p = 0.61). Mean glutamate across conditions and sub-blocks ranged from 8.3 to 8.8 mmol/kg. Subjective effects varied substantially across individuals, especially for oceanic boundlessness, visionary restructuralisation, and reductions in vigilance. In the correlation analyses, greater glutamate change from baseline to ketamine was significantly associated with anxious ego dissolution (r = 0.77, p = 0.01) and reductions in vigilance (r = 0.70, p = 0.03). After multiple-comparison correction, only the anxious ego dissolution association remained significant (p corr = 0.04), whereas vigilance did not (p = 0.08). Similar positive associations were found when using absolute glutamate values during ketamine: anxious ego dissolution and vigilance again correlated positively, with corrected p-values reported as significant. Baseline glutamate was not significantly associated with later subjective effects, suggesting the observed relationships were not simply pre-existing traits. The paper notes that the exploratory sub-block results hinted at a subtle glutamate increase, but this was not statistically significant.

The authors interpret the study as the first human investigation of ketamine’s effects on glutamate signalling in the NAc. Their main finding was that ketamine did not produce a significant group-level increase in NAc glutamate in 10 healthy volunteers, contrary to their initial hypothesis. They note that this differs from much of the preclinical literature and some human MRS work in other brain regions, but they argue that the discrepancy may reflect regional specificity, differences between MRS and microdialysis, and the difficulty of measuring small structures such as the NAc. They emphasise that the NAc voxel was small, which reduced signal and required long acquisition times. This may have made it easier for brief or rapidly changing glutamate effects to be averaged out, and the prolonged scan also increased the chance of motion-related distortion. The authors suggest that the exploratory sub-block analysis may hint at a subtle effect that was obscured in the averaged analysis, but they treat this cautiously. Although the group-level neurochemical effect was null, the authors highlight the association between individual glutamate changes and subjective ketamine effects. In particular, greater glutamate increases were linked to more anxious ego dissolution and reduced vigilance. They suggest that elevated NAc glutamate may alter excitatory signalling and disturb mesolimbic-thalamocortical circuits, producing negative depersonalisation-like experiences and clouded consciousness rather than positive or mystical experiences. They also note that music could potentially influence subjective and neural effects, but argue that this was held constant across conditions. The authors place these findings alongside broader ketamine research suggesting that the quality of subjective experience may matter for treatment outcomes, but they caution that dissociation alone may not fully explain response. They also mention work indicating that anxiety and other affective reactions during ketamine administration may be relevant, and that broader psychometric tools than the CADSS may be needed. Key limitations they acknowledge include the very small sample size, the exploratory nature of the study, the methodological constraints of imaging the NAc, and the uncertainty about whether the findings generalise to people with SUDs. They state that the MRS protocol is now being used in a Phase II trial in cocaine use disorder and argue that future studies should clarify the mechanisms behind the glutamate-subjective experience association and whether these effects relate to therapeutic outcomes.