Psychological effects of (S)-ketamine and N,N-dimethyltryptamine (DMT): a double-blind, cross-over study in healthy volunteers

Gouzoulis-Mayfrank and colleagues frame the study within a long tradition of using hallucinogenic drugs as pharmacological models of psychosis. Earlier research established that classic serotonergic hallucinogens (for example LSD, psilocybin, DMT) act mainly at serotonin 5-HT2A receptors and tend to produce experiences that differ phenomenologically from schizophrenia (more visual phenomena, preserved affect and social engagement), whereas PCP-like drugs (PCP, ketamine) are NMDA receptor antagonists and have been argued to induce a more schizophrenia-like state with thought disorder, negative symptoms and motor abnormalities. Despite this conceptual distinction, no prior experimental within-subject comparison had directly contrasted the psychological profiles produced by a serotonergic agonist and an NMDA antagonist in the same healthy volunteers. The present study therefore set out to compare intravenous N,N-dimethyltryptamine (DMT, a 5-HT2A agonist) with (S)-ketamine (an NMDA antagonist) in a randomised, double-blind, cross-over design. The investigators hypothesised that neither agent would be an overall superior model of psychosis; rather, (S)-ketamine would provoke more negative and cognitive-like phenomena, while DMT would produce more positive symptom-like phenomena (for example thought disorder and perceptual hallucinations). The comparison used dose-titrated infusions to obtain comparable overall intensities of subjective effect across the two drug states.

Fifteen medically and psychiatrically healthy volunteers (9 men, 6 women; mean age 38.0 years) were recruited and screened with medical examination and structured psychiatric interview. Exclusions included current or past Axis I disorder, significant family history of severe psychiatric disorder in first-degree relatives, substance use disorders or regular medication. Most participants had some prior, mostly sporadic, experience with hallucinogens or cannabis; all were clinical or allied psychiatric staff and were unpaid. The study received ethical approval and written informed consent was obtained. Each subject attended two experimental days (one DMT, one (S)-ketamine) in a double-blind, cross-over and pseudo-randomised order separated by two to four weeks. Within each drug day subjects received a low and a high dose (two-hour break between doses) in a single-blind titration procedure intended to produce comparable overall intensity of effects across individuals. Doses were given as an initial bolus over 5 minutes followed by a continuous infusion (specific bolus and infusion-rate ranges for low and high regimens were pre-defined for both drugs). The ketamine infusion was reduced to 75% after 60 minutes to avoid plasma accumulation; no such reduction was required for DMT due to faster elimination. An unblinded physician administered the infusions and did not otherwise interact with subjects or raters; an experienced psychiatrist and a medical student, both blind to substance, accompanied subjects throughout. Psychopathology was assessed during the first 30 minutes of each infusion and during breaks using standardised self-report and clinician-rated instruments. Self-assessments included the Hallucinogen Rating Scale (HRS) and the APZ-OAV altered-states questionnaire (global score plus subscales OSE, AIA and VUS). Clinician-administered ratings comprised the Scale for the Assessment of Positive Symptoms (SAPS), the Scale for the Assessment of Negative Symptoms (SANS) and the Schizophrenia Prediction Instrument—Adult version (SPI-A). Five single items from SPI-A and APZ-OAV were also used to capture catatonia-like motor phenomena. Blood samples for plasma concentrations were drawn at approximately +10, +15, +60, +90 and +150 minutes relative to infusion start. One rater performed all interviews and psychometric ratings. Subjects remained under medical supervision for at least two hours after the second dose; follow-up interviews were conducted the next day, at seven days and at 10–12 months. Statistical analysis used repeated-measures ANOVAs with factors substance (DMT, (S)-ketamine) and dose (low, high). Correlations between plasma exposure metrics (AUC, Cmax) and psychometric scores were reported. Statistical significance was set at p ≤ .05, and the investigators applied a conservative Bonferroni correction for multiple substance comparisons where noted. The extracted text indicates that psychometric data reported here relate to the nine subjects who completed both experimental days with both doses.

Of the 15 enrolled volunteers, 12 completed both DMT doses and 10 completed both (S)-ketamine doses; nine subjects completed both experiments with both doses and constitute the analysed sample reported here. Dropouts occurred because of intense unpleasant psychological effects (three cases), nausea (one DMT), hypotonia (one DMT) and one participant declining a second day because of delayed mild adverse effects; all acute undesirable effects resolved rapidly after stopping the infusion and no lasting sequelae were observed. Next-day interviews reported transient tiredness and headache more commonly after DMT (10/13) than after (S)-ketamine (4/11); otherwise no treatment-related psychopathology or substance-use problems emerged at seven-day or 10–12 month follow-up. Measured plasma concentrations stabilised between +60 and +90 minutes. Mean (S)-ketamine plasma levels (mean of +60 and +90 minutes) were 169.8 ± 29.9 ng/ml for low dose and 197.4 ± 31.6 ng/ml for high dose. Mean DMT levels were 42.8 ± 26.1 ng/ml (low) and 60.0 ± 28.0 ng/ml (high). Ten minutes before the second-dose infusion, DMT levels had declined to about 5.1 ± 3.1 ng/ml and (S)-ketamine to 49.1 ± 12.2 ng/ml. Self-report scales: On the HRS there were no significant substance effects for somaesthesia, volition or intensity. Affect and perception tended to be higher for DMT but did not reach significance; cognition showed a significant main effect of dose (higher at high dose) but no substance effect. On the APZ-OAV global score there was no substance main effect and a marginal dose effect; the AIA (dread/ego-disintegration) subscale showed a dose effect, while the VUS (visionary restructuralization) subscale showed a significant main effect of substance (higher for DMT). Clinician ratings—positive symptoms: Repeated-measures ANOVAs showed significant main effects of substance and dose on the SAPS global score, indicating higher positive-symptom ratings after DMT and at the higher dose. Positive formal thought disorder showed strong substance and dose effects. Inappropriate affect was higher after DMT. Hallucinations showed a marginal substance effect (p = 0.058) and a significant dose effect; delusions increased with dose but did not differ between substances. Clinician ratings—negative and cognitive symptoms: The SANS global score and multiple SANS subscales (affective blunting, alogia, avolition-apathy, attention) all showed significant main effects of substance and of dose, with higher scores after (S)-ketamine. Significant substance × dose interactions for affective blunting and alogia indicated a stronger dose-related increase with (S)-ketamine than with DMT. SPI-A subscales all showed significant dose effects; substance main effects favoured (S)-ketamine for emotional deficits, cognitive impediments, cognitive disturbances and body perception disturbances. A substance × dose interaction was seen for SPI-A body perception disturbances. Catatonia-like items: Four of five single items (motor interference, motor blockages, feeling paralysed, staying frozen in an unnatural position) showed significant main effects of substance with stronger effects after (S)-ketamine. Dose effects were significant for motor interference, motor blockages and the frozen-position item. Motor blockages showed a significant substance × dose interaction (increase only for higher (S)-ketamine dose). Plasma–effect correlations: Higher DMT exposure correlated with greater positive symptoms and overall altered consciousness (DMT AUC vs SAPS global r = 0.629, p < .05; DMT Cmax vs APZ-OAV global r = 0.590, p < .05). Higher (S)-ketamine exposure correlated with more pronounced negative symptoms and cognitive/body-perception disturbances ((S)-ketamine AUC vs SANS global r = 0.691, p < .01; AUC vs SPI-A subscores r = 0.643–0.720, p < .05 to < .01). Correction for multiple comparisons: After a conservative Bonferroni correction for the 33 psychometric comparisons, several substance main effects remained significant at the corrected threshold (p < 0.0016): SAPS positive formal thought disorder; SANS global score and SANS emotional blunting; SPI-A cognitive impediments, SPI-A body perception disturbances and SPI-A motor blockages. Qualitative impressions: Subject descriptions and investigator observations converged in showing that DMT produced vivid visual and auditory phenomena, synaesthetic reports in some cases, active interpersonal interest and frequent positive formal thought disorder; high-dose DMT could elicit transient paranoid ideation. By contrast, (S)-ketamine produced dose-dependent hypomimia, marked psychomotor poverty, reduced speech and social withdrawal, prominent bodily and vestibular distortions, and catatonia-like motor phenomena; most subjects judged the (S)-ketamine experience unpleasant. Overall global intensity of the two drug states was similar.

Gouzoulis-Mayfrank and colleagues interpret the findings as evidence that the two pharmacological models produce overlapping but distinct psychological profiles rather than one being categorically superior as a model of schizophrenia. Both agents induced dose-dependent, powerful alterations in perception, affect and cognition, and global intensity measures were comparable. DMT preferentially provoked phenomena resembling positive symptoms—most notably positive formal thought disorder, inappropriate affect and visual perceptual alterations—whereas (S)-ketamine preferentially induced negative- and cognitive-like phenomena including affective blunting, alogia, avolition-apathy, attention deficits and body-perception disturbances, together with catatonia-like motor abnormalities. The investigators note that some (S)-ketamine effects could reflect nonspecific sedation, and that psychometric ratings alone cannot fully disentangle sedation from genuine negative or catatonic symptoms. Nevertheless, they argue that subjective reports (feeling emotionally blunted, isolated and lacking drive) and observed motor phenomena support interpretation beyond mere sedation. Correlations between plasma exposure and symptom domains provide additional support for differential pharmacological relationships. Several methodological limitations acknowledged by the authors bear on interpretation. Dose equivalence across drugs cannot be established definitively despite titration to comparable subjective intensity. Flexible dosing improves individual matching but prevents firm dose–response conclusions. The study omitted a placebo arm (practical recruitment constraints and limited blinding feasibility were cited), the sample was small and drawn from psychiatric staff (risking reporting bias and limited generalisability), and no drug screens were performed on experimental days. Finally, catatonia-like phenomena were not assessed with a validated catatonia scale, so those findings should be treated cautiously. Despite these caveats, the authors conclude that the serotonergic (LSD-type) and NMDA antagonist (PCP-type) models appear to capture different aspects or subtypes of schizophrenia—DMT aligning more with paranoid-hallucinatory presentations and (S)-ketamine with negative/cognitive and possibly catatonic or undifferentiated presentations—and that multiple pharmacological models may be useful for laboratory study of heterogeneous psychotic syndromes and symptom-specific therapeutic strategies.

The authors conclude that the 5-HT2A agonist (LSD-type) and NMDA antagonist (PCP-type) pharmacological models portray different aspects of schizophrenia: DMT induced phenomena resembling positive/paranoid symptoms, while (S)-ketamine induced phenomena resembling negative and cognitive symptoms and catatonia-like features. They suggest that no single agent can model the full heterogeneity of schizophrenia and that having multiple models may aid neurobiological research and the development of symptom-specific treatments.