Dose-related Behavioral, Subjective, Endocrine and Psychophysiological Effects Of the Kappa Opioid Agonist Salvinorin A in Humans

Salvia divinorum and its principal psychoactive constituent, salvinorin A (SA), have become increasingly used recreationally, particularly by adolescents and young adults. SA is a highly potent, selective kappa opioid receptor (KOR) agonist and, unlike many classical hallucinogens, lacks activity at dopaminergic, serotonergic or NMDA receptor systems. The human literature up to the time of this study consisted mainly of anecdotal reports and a small number of experimental investigations that were limited by uncontrolled designs, sublingual administration with poor bioavailability, small samples, or use of Salvia leaf preparations rather than defined doses of pure SA. Important gaps remained regarding SA’s pharmacokinetics in humans and its objective behavioural, subjective, cognitive, endocrine and psychophysiological effects under controlled conditions. B. and colleagues set out to characterise comprehensively the acute effects and safety of inhaled SA in healthy human volunteers with prior Salvia exposure. The study aimed to measure dose-related subjective and behavioural effects, cognitive performance, resting EEG, plasma SA and its major metabolite salvinorin B (SB), and neuroendocrine responses (cortisol and prolactin) using a double-blind, randomised, placebo-controlled, counterbalanced crossover design. The authors framed the work as addressing prior methodological limitations by using pure SA, objective biological measures and standardised delivery in the laboratory.

The study used a double-blind, randomised, placebo-controlled, counterbalanced crossover design with three test days per subject at a specialised clinical research unit. Ten medically and psychiatrically healthy adults who had previously used Salvia were recruited after a rigorous screening process (details referenced to a Supplement). The extracted text reports inclusion of subjects aged 18–55 years with prior Salvia exposure; because recreational users commonly use other drugs, participants with other drug exposure were allowed to enhance representativeness. Subjects were asked to refrain from alcohol, illicit or prescription drugs for one week before the first test day through study completion. On each test day participants inhaled one of three conditions administered via a commercially available vapouriser: placebo (an empty aluminium container), low-dose SA (8 mg) or high-dose SA (12 mg). SA was prepared by research pharmacists on the morning of each test day; subjects and raters were blinded to dose. Procedures included urine toxicology and pregnancy testing, IV line placement and baseline ratings. Safety assessments were performed the day after the first and last test days and at follow-ups 1 and 3 months after completion. Outcome measures combined subjective, behavioural, cognitive, endocrine, pharmacokinetic and psychophysiological assessments. Subjective states were measured with visual analogue scales (VAS), psychotomimetic symptoms with the PANSS and the Psychotomimetic States Inventory (PSI), and perceptual/dissociative effects with the Clinician Administered Dissociative Symptoms Scale (CADSS) and the Hallucinogen Rating Scale (HRS). Cognitive performance used a brief WAIS-R derived battery (Digits Forward/Backward, Letter Number Sequence). Plasma cortisol and prolactin were assayed at multiple time points. SA and SB plasma and urine levels were quantified by LC-APCI-tandem mass spectrometry with limits of quantitation of 0.5 ng/ml. Resting-state EEG (three minutes, eyes closed) was recorded immediately following inhalation; analyses reported spectral power at midline electrodes (Cz, Pz, Oz). Statistical analyses used descriptive summaries and tests for normality. Approximately normally distributed outcomes (PANSS, PSI, HRS, cognitive measures) were analysed with linear mixed models including within-subject factors of dose (placebo, 8 mg, 12 mg) and time (pre- vs post-inhalation) and random subject effects. Non-normally distributed CADSS and VAS outcomes were analysed using a nonparametric ranked mixed-effects approach with ANOVA-type statistics (ATS). Hormone and SA/SB level trajectories and physiological measures were also compared using similar mixed models. EEG power was analysed with mixed models including dose and electrode as within-subject factors. Analyses were performed in SAS v9.2.

Ten subjects were enrolled (mean age 23.8 ± 3.2 years; 90% male; mean education 15.3 ± 1.2 years; mean IQ 117.2 ± 7.1). Nine completed all three test days and one dropped out after two sessions; all 10 were included in analyses. None met criteria for substance dependence; all had prior SA and other illicit drug exposure. No serious adverse events occurred; no test days were terminated early. Follow-up at 1, 3 and 6 months detected no new psychiatric symptoms or increased Salvia consumption. Subjective and behavioural effects: Participants reported somaesthetic changes, dissociative experiences and perceptual alterations, illustrated by brief subject quotations (e.g. cold prickling, detachment, heightened pattern perception). On VAS measures, SA produced a main effect on ‘‘drowsy’’ such that both 8 mg and 12 mg doses produced less drowsiness than placebo (ATS statistics reported: overall ATS=4.55, p=0.01; low dose ATS=4.9, p=0.03; high dose ATS=3.99, p<0.05). SA did not increase ratings of feeling ‘‘high’’, ‘‘calm’’, ‘‘sad’’, ‘‘irritable’’ or ‘‘anxious’’. On psychotomimetic scales the PANSS positive subscale increased by 3.5 points and the PSI by approximately 10 points, magnitudes the authors compared to effects seen with delta-9-THC and ketamine. Pharmacokinetics: Plasma SA levels rose rapidly after inhalation and peaked at about +15 minutes post-administration. Mixed-model comparisons showed a significant increase in SA levels after active dosing compared with pre-administration [F(3,38)=29.4, p<0.0001], without significant differences between the two active doses. SB levels also increased after SA administration [F(3,38)=8.66, p=0.0002]. Neuroendocrine effects: Low-dose SA significantly elevated plasma cortisol [F(2,120)=3.11, p<0.05], with cortisol returning to baseline by 60 minutes post-inhalation [F(4,120)=18.69, p<0.0001]. Both SA doses significantly elevated plasma prolactin [F(8,120)=4.07, p=0.0003], with prolactin also returning to baseline by 60 minutes. Physiology and cognition: Neither SA dose produced significant changes in heart rate or systolic/diastolic blood pressure. The extracted text reports no euphoria or persistent cognitive deficits; cognitive battery outcomes are not detailed numerically in the extracted text. EEG: Resting-state EEG spectral power decreased across frequencies examined, with statistically significant reductions in beta power (13–29 Hz) at both active doses versus placebo [F(68,2)=5.47, p<0.006]. Theta band power showed a trend toward reduction [F(68,2)=2.44, p=0.09]. Effects on delta, alpha and gamma bands were not statistically significant. Safety and tolerability: SA was well tolerated acutely and in follow-up. One subject dropped out after reporting no effects; no rescue medications were required and no serious adverse events were recorded. The subjective and objective effects were transient, peaking within about 10 minutes and generally returning to baseline within 30–60 minutes.

B. and colleagues interpret their findings as demonstrating that inhaled SA produces very rapid, short-lasting psychoactive effects in healthy humans characterised by somaesthetic changes, dissociation and perceptual alterations. The onset was within seconds to minutes, with peak intensity typically within 10 minutes and return toward baseline by 30 minutes; no lingering effects were reported at discharge (90 minutes) or during follow-ups. The magnitude of psychotomimetic effects on PANSS and PSI was described as comparable to those produced by delta-9-THC and ketamine on the same measures, although peak intensities reported by subjects in the laboratory were estimated at only 20–30% of their recalled recreational peak effects. The authors highlight objective biological corroboration of central KOR activity: rapid rises in plasma SA and SB, elevations in prolactin and cortisol consistent with KOR agonism, and reductions in resting-state EEG beta power. They note that the prolactin and cortisol responses align with preclinical and human data for other KOR agonists and suggest hypothalamic–pituitary–adrenal (HPA) activation. The decrease in beta-band EEG power differs from patterns observed with some other hallucinogens (which produce increases or no change in beta), supporting the view that SA has a distinct psychophysiological profile linked to its unique KOR mechanism. Regarding abuse liability, the authors argue that SA’s lack of euphoria and its putative KOR-mediated reductions in mesolimbic dopamine are consistent with lower addictive potential relative to drugs that increase accumbal dopamine. They caution that recreational use is variable and can range up to presentations prompting emergency care; they also note that SA exposure could be especially harmful in people vulnerable to psychotic illness. The discussion extends to potential relevance for understanding psychosis: SA produces psychosis-like effects despite decreasing dopamine, raising questions about KOR involvement in psychotic phenomena and suggesting that KOR antagonists warrant further investigation as possible antipsychotic agents. Strengths cited include the randomised double-blind crossover design, use of multiple doses and pure SA, measurement of blood levels and inclusion of both subjective and objective measures (hormones, EEG, pharmacokinetics). Limitations acknowledged by the authors include limited generalisability to recreational use because of standardised setting and method of delivery, and the inability to characterise a dose–response relationship since plasma levels and responses did not differ between the two active doses. The authors recommend future studies to examine a wider dose range, to test whether SA’s effects are blocked by KOR antagonists, and to use receptor imaging to clarify mechanisms.