Therapeutic infusions of ketamine: do the psychoactive effects matter?

A single sub‑anaesthetic intravenous infusion of ketamine (commonly 0.5 mg/kg over ~40 min) has been shown in prior work to produce rapid antidepressant effects that peak about 24 h after infusion and attenuate by 72 h. Such effects have been attributed to enhanced neuroplasticity, modulation of prefrontal glutamate balance, and changes in default mode network function. Ketamine may therefore also ameliorate dependence‑related deficits by similar biological pathways; the investigators previously reported that ketamine improved low motivation to quit and cue‑induced craving in non‑depressed cocaine‑dependent volunteers. This analysis asks whether the transient psychoactive effects produced by therapeutic ketamine infusions—specifically mystical‑type experiences similar to those elicited by serotonergic hallucinogens—contribute to clinical benefit. Dakwar and colleagues set out to (1) test whether two sub‑anaesthetic intravenous ketamine doses produce dose‑dependent mystical‑type effects measured shortly after infusion, and (2) evaluate whether the intensity of those mystical‑type experiences, as distinct from dissociative symptoms, mediates ketamine’s 24‑hour effects on motivation to quit cocaine and on cue‑induced craving.

Design and participants: The study was a double‑blind, randomised, inpatient experiment involving eight adult cocaine‑dependent individuals actively using freebase cocaine and not seeking treatment. Participants underwent psychiatric screening (including DSM‑IV SCID) and a medical evaluation; those enrolled were in good health, had no history of ketamine use, and had no psychiatric disorders other than substance use disorders. Subjects were hospitalised for nine days and were abstinent on days 1–3 before randomisation to infusion order. Institutional review board approval was obtained. Interventions and procedures: Each participant received three 52‑minute infusions separated by 48 h in a counterbalanced, double‑blind manner: low‑dose ketamine (K1, 0.41 mg/kg: 0.11 mg/kg bolus over 2 min then 0.3 mg/kg over 50 min), higher‑dose ketamine (K2, 0.71 mg/kg: 0.11 mg/kg bolus then 0.6 mg/kg slow drip), and lorazepam 2 mg (LZP: saline bolus then 2 mg over 50 min) as an active control. There were three possible infusion sequences constrained so that K1 always preceded K2 for safety. Participants were monitored for 2 h post‑infusion and interviewed at 20 min and 1 h with attention to persistent sedation, psychosis or dissociation. Assessments: Primary behavioural outcomes were motivation to quit cocaine, assessed by the University of Rhode Island Change Assessment (URICA) at baseline and 24 h post‑infusion, and cue‑induced craving, assessed as the summed visual analogue scale (VAS) ratings collected every 3 min during 15 min of cocaine cue exposure (baseline and 24 h). Acute psychoactive effects were measured within 15 min of infusion termination. Dissociation was measured with the Clinician Administered Dissociative Symptoms Scale (CADSS). Mystical‑type phenomena were measured using a modified, shortened version of Hood’s Mysticism Scale (HMS) adapted to refer specifically to the infusion experience; nine items represented the scale’s dimensions and scores were summed (maximum 36). Other brief psychiatric items and a VAS of drug liking were also collected. Statistical analysis: Given the small sample (n = 8), nonparametric tests were used. HMS scores were compared across medication conditions and infusion order with Friedman’s two‑way ANOVA and paired Wilcoxon signed‑rank tests with Bonferroni correction, two‑tailed α = 0.05. Mediation analyses focused on K1 versus lorazepam because previous work indicated K1 produced a robust ~60% change in the primary outcomes and K2 could not be cleanly evaluated due to order/carryover. Four mediation models were tested (CADSS and HMS each tested as mediators for URICA and for summed VAS craving), using the standard approach whereby mediation is supported if the mediator predicts outcome when controlling for infusion condition and infusion condition ceases to predict the outcome when the mediator is included. To limit Type I error across multiple mediation tests, α was adjusted to 0.0125.

Sample and safety: Eight participants completed the protocol. Extracted participant characteristics indicate high baseline cocaine use (mean 22 use days in the prior 4 weeks; estimated cost per use day reported as US$158.83). All subjects tolerated the procedures without adverse events, including no unexpected psychiatric disturbances or initiation of ketamine/benzodiazepine misuse. Acute psychoactive effects: All ketamine‑related psychoactive effects, including dissociative and mystical‑type phenomena, resolved within 20 min after infusion termination. Both ketamine doses produced significantly higher HMS scores than lorazepam: median HMS scores were 4 for lorazepam, 22 for K1, and 30 for K2 (overall comparison p = 0.012). The higher ketamine dose (K2) produced greater mystical‑type effects than the lower dose (K1) (median 30 vs 22, p = 0.027). There was no significant effect of infusion order on HMS. A linear regression testing the relationship between HMS and CADSS scores for K1 was non‑significant (R2 = 0.261, p = 0.12), indicating no clear correlation between mystical‑type and dissociative scores for that dose. Mediation analyses: Four mediation models were run (HMS or CADSS mediating changes in URICA or summed VAS craving). The only model meeting the authors’ mediation criteria was HMS mediating the effect of infusion condition (K1 vs lorazepam) on URICA change. The total mediation model for URICA had an adjusted R2 = 0.775 (p < 0.005). In simple regression infusion condition predicted URICA change (β = 0.776, p < 0.001), but when HMS score was included the direct effect of infusion condition fell to β = 0.274 (p = 0.19), consistent with mediation. No mediation was observed for cue‑induced craving, and CADSS did not fulfil mediation criteria for either outcome.

Dakwar and colleagues highlight two principal findings: sub‑anaesthetic intravenous ketamine produced dose‑dependent mystical‑type experiences, and the intensity of those transient experiences statistically mediated ketamine’s 24‑hour enhancement of motivation to quit cocaine but did not mediate reductions in cue‑induced craving. The authors interpret these results as evidence that certain psychoactive effects of ketamine may contribute to its therapeutic impact via psychological mechanisms rather than solely via neurobiological processes. They situate the findings within broader literature on psychedelic and mystical experiences, noting that earlier work with psilocybin linked such experiences to lasting personal significance and therapeutic benefit. A prior intramuscular ketamine study reportedly did not find mystical‑type effects versus placebo; Dakwar and colleagues propose that differences in route and dose might explain that discrepancy because IV administration and the doses used in the present protocol produced clear dose dependence. The authors further draw on historical and psychological theories (for example, writings of William James) to suggest mystical experiences may alter existential dimensions such as self‑identity and purpose, which plausibly explain stronger effects on motivation to quit than on cue‑elicited craving, the latter being only partly influenced by changes in perspective. The discussion also considers clinical translation: approaches such as Ketamine Psychedelic Therapy (KPT), which pair ketamine‑facilitated existential reappraisal with psychotherapy, are mentioned as examples of frameworks that seek to harness and consolidate pharmacologically induced transformative experiences. The investigators suggest that careful screening, preparation, a controlled setting, and post‑infusion support might enhance benefits while minimising risks such as misuse or behavioural toxicity. Several limitations are acknowledged. The sample was small and homogeneous, which limits generalisability and statistical power. The shortened HMS may have reduced sensitivity to nuanced differences between doses. The design could not disentangle mediation from marker effects definitively—although the lack of correlation between HMS and CADSS for K1 and the specificity of mediation for URICA (not craving) provide some support for a mediating role, the authors caution that larger, properly powered studies are required. Other potentially relevant variables (for example, family history of alcoholism or serum ketamine levels) were not examined. Finally, the study was not designed to assess longer‑term drug use outcomes because there was no non‑ketamine control group across follow‑up. The authors conclude that their preliminary results justify further research into whether enhancing and integrating ketamine‑induced mystical‑type experiences can improve clinical outcomes in substance use and related disorders.