Ketamine-induced analgesia and dissociation show distinct behavioral and neural correlates
Noam Goldway, Itamar Jalon, Yara Agbaria, Yotam Pasternak, Roi Sar-El, Dan Mirelman, Noam Sarna, Nili Green, Talma Hendler, Haggai Sharon
Neuropsychopharmacology July 20, 2026 DOI: 10.1038/s41386-026-02496-x (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Within-subject, placebo-controlled fMRI study Peer reviewed |
|---|---|
| Sample size | 37 |
| Population | Healthy volunteers (21 female) |
| Intervention | Ketamine |
| Dose | 0.4 mg/kg bolus over 10 min followed by 0.4 mg/kg/h continuous infusion |
| Duration | Single session per condition (ketamine or placebo), with pain stimuli and dissociation measures during scanning |
| Topics | Esketamine Ketamine |
| Key points | Ketamine-induced analgesia and dissociation show separable behavioral and neural signatures, with pain-related brain activity not covarying with dissociation-related brain connectivity. |
Abstract
Abstract Ketamine produces both analgesic and dissociative effects, but whether analgesia depends on dissociation remains debated. This question is part of a broader discussion on whether the subjective experiences elicited by psychoactive drugs are necessary for their therapeutic benefits. Here, we tested whether ketamine-induced analgesia and dissociation show separable behavioral and neural signatures. In a within-subject, placebo-controlled fMRI study, 37 healthy volunteers (21 female) underwent two sessions: intravenous ketamine (0.4 mg/kg bolus over 10 min followed by 0.4 mg/kg/h continuous infusion) or saline placebo. Individually calibrated thermal pain stimuli were applied to the right leg during scanning. Dissociative states were measured repeatedly using the Clinician-Administered Dissociative States Scale. Whole-brain univariate and multivariate analyses, as well as network-based functional connectivity analyses, were performed. Ketamine induced both analgesia (session × pain intensity interaction: F (1,54) = 11.22, p = 0.001) and dissociation (main effect of session: F (1,32) = 57.44, p < 0.001), and the two corresponded to distinct neural indices. Greater pain was associated with increased univariate activity in regions such as the anterior insula, as well as with stronger expression of a pain-predictive multivoxel pattern ( ρ = 0.6, p < 0.001), whereas higher dissociation intensity was selectively associated with reduced default mode network connectivity ( ρ = .49, p < 0.01). Neurobehavioral markers of pain and dissociation did not covary ( ρ = −0.24 to 0.32, all p > 0.09), consistent with distinct neural correlates of ketamine’s analgesic and dissociative effects.