Drug and Alcohol Dependence 111 (2010) 250–256 Contents lists available at ScienceDirect Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep Use patterns and self-reported effects of Salvia divinorum: An internet-based survey Matthew J. Baggott a,b,∗, Earth Erowid c, Fire Erowid c, Gantt P. Galloway b, John Mendelson b a Helen Wills Neuroscience Institute, University of California - Berkeley, Berkeley, CA, USA b Addiction and Pharmacology Research Laboratory, California Pacific Medicial Center Research Institute, San Francisco, CA, USA c Erowid Center, Grass Valley, CA, USA article info abstract Article history: Background: There is growing use of Salvia divinorum (SD), a psychoactive plant that produces Received 30 October 2009 hallucinogen-like effects through a kappa opioid receptor (KOR) mechanism. Little is known about KOR Received in revised form 4 May 2010 agonist effects in humans and about users of SD. Accepted 6 May 2010 Objectives: To characterize the reasons, methods, and reported consequences of SD use. Methods: Individuals reading SD-related pages of a drug-information website were invited to anony- mously complete an online questionnaire if they had used SD. Keywords: Results: Participants (N = 500) were 92.6% male and 23.4 ± 8.7 (mean ± s.d.) years old. They had used a Salvia divinorum Hallucinogen median of six times (range 1–250). 80.6% probably or definitely would use SD again. Most participants Kappa opioid (92.6%) typically smoked or vaporized SD product. When smoked, the drug’s main effects were estimated Patterns of use to last 14.1 ± 12.8 (range 0.5–120) minutes. When asked to compare SD effects to other methods of alter- Subjective effects ing consciousness, the most common answer was that SD was unique (38.4%). 25.8% reported persisting (≥24 h) positive effects (often described as increased sense of well-being) on at least one occasion. 4.4% reported persisting negative effects (most often anxiety). Conclusions: SD is typically smoked, acute effects are brief, and persistent adverse effects are uncommon. In addition to acute hallucinogenic effects, SD may produce subacute increases in subjective well-being. Such a subacute effect would be unusual for a drug that is used non-medically, as withdrawal from other drugs typically either does not affect mood or causes dysphoria. Findings from this convenience sample should be confirmed and extended using surveys of random samples and controlled clinical studies. © 2010 Elsevier Ireland Ltd. All rights reserved. 1. Introduction et al., 2005; Kovacs et al., 2005; Mitchell et al., 2005; Zapata and Shippenberg, 2006). Variations in the genes encoding the KOR and Pharmacological manipulations of the kappa opioid recep- two ligands have been associated with a risk for alcohol dependence tor (KOR) system have been proposed as novel treatments for (Xuei et al., 2006). In addition, a KOR agonist reduces convul- depression and drug and alcohol dependence (Carroll et al., 2005; sion severity in animals susceptible to ethanol withdrawal seizures Gaveriaux-Ruff and Kieffer, 2002; Kieffer and Gaveriaux-Ruff, (Beadles-Bohling and Wiren, 2006). 2002). KOR antagonists show antidepressant-like effects and atten- Despite this acknowledged therapeutic potential, few clinical uate behavioral responses to stress (Mague et al., 2003), while KOR studies have employed KOR agonists and little is known about their stimulation modulates the reward system by reducing dopamine effects in humans. One apparent reason is that, at high doses, KOR release in the nucleus accumbens, and by generating aversive states agonists such as cyclazocine, enadoline, and spiradoline produce (Aldrich and Vigil-Cruz, 2003; Spanagel et al., 1990). KOR manip- sensory effects such as visual and auditory distortions and feel- ulations modulate ethanol consumption in rodents, and ethanol- ings of depersonalization (Chappell et al., 1993; Martin et al., 1965; and cocaine-induced alteration in brain dopamine levels (Chefer Pfeiffer et al., 1986; Walsh et al., 2001). At lower doses, KOR ago- nists also produce decreased body temperature, diuresis, sedation, stupor, and difficulty concentrating (Ashton et al., 1989; Chappell et al., 1993; Kramer et al., 2000; Martin et al., 1965; Pfeiffer et al., ∗ Corresponding author at: Addiction and Pharmacology Research Laboratory, 1986; Reece et al., 1994; Rimoy et al., 1994; Walsh et al., 2001). California Pacific Medical Center Research Institute, 7th Floor, St Luke’s Hospital, Cardiovascular effects from kappa agonists are reportedly minimal 3555 Cesar Chavez Street, San Francisco, CA 94110, USA. Tel.: +1 415 641 3370; fax: +1 415 641 3380. (Rimoy et al., 1994; Walsh et al., 2001). Selective KOR antagonists E-mail address: [email protected] (M.J. Baggott). are not available for use in clinical trials at present. 0376-8716/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.drugalcdep.2010.05.003 M.J. Baggott et al. / Drug and Alcohol Dependence 111 (2010) 250–256 251 Salvia divinorum (SD) is a hallucinogenic plant that has been Table 1 traditionally used for divination and healing in the Mazatec zone of Estimated duration of Salvia divinorium effects when smoked or vaporized. Mexico (Johnson, 1939), and was first collected by Westerners in Estimated duration of effects when smoked or vaporized (min) 1955 (Epling and Jativa, 1962). In the United States, a 2008 gov- SD extract (N = 296) SD leaf (N = 180) Salvinorin A (N =6) ernment report estimated that 1.8 million people had tried SD, with most initiating use in the last two years (Substance Abuse Mean 14.1 15.7 27.5 St. Dev. 12.8 14.8 16.1 and Mental Health Services Administration, 2008). SD use has been Median 10 10 21 reported in college students in the United States (Khey et al., 2008; Mode 10 5 N/A Lange et al., 2008) and is also occurring in Europe (Pavarin, 2006) Min 0 0 13 and Japan (Tsujikawa et al., 2008). At least in a sample of Italian Max 120 90 50 youth attending concerts, SD users appear to have fewer medical, legal, or social impairments than do individuals who consume other found any of the survey questions difficult to understand, and whether they had recreational drugs (Pavarin, 2006). previously completed this survey. The putative primary psychoactive agent in SD is a structurally When participants finished the survey, they clicked a button on the bottom novel KOR agonist named salvinorin A (Ortega et al., 1982; Valdés et of the page affirming that their responses were complete. Their responses were al., 1984). Consistent with KOR agonist activity, users describe SD in then anonymously stored in an electronic database. If participants did not click this button, none of their responses were stored. The automatically collected web-server lay literature as hallucinogenic: it produces perceptual distortions, activity log counted the number of times the announcement was read, the number pseudo-hallucinations, and a profoundly altered sense of self and of times the information sheet was read, and the number of times the survey was environment, including out-of-body experiences (Aardvark, 1998; completed. Erowid, 2008; Siebert, 1994b; Turner, 1996). SD therefore appears Data from a participant were excluded from analysis if: (1) data were incom- to have the potential to elucidate the role of the KOR receptor sys- plete; (2) the participant reported having previously completed the survey; (3) the participant reported he or she was not fluent in English or had difficulty understand- tem in health and disease (Butelman et al., 2004; Chavkin et al., ing the questions; or (4) another response was submitted from the same computer 2004; Roth et al., 2002). (IP address) within ten minutes. The survey ended when 500 participants completed Aside from books and web sites providing information to poten- the survey and met inclusion and exclusion criteria. tial users (Aardvark, 1998; Erowid, 2008; Ott, 1993; Pendell, 1995; Data were analyzed using R (R Development Core Team, 2007). Siebert, 2009; Sociedad, 1998; Turner, 1996), there is very lit- tle published information on users’ experiences with the drug 3. Results (Gonzalez et al., 2006; Lange et al., 2009). For that reason, study- ing SD users appears to be a novel method for understanding the We collected a total of 520 responses (40.1% of those viewing the effects of high and repeated exposure to a KOR agonist in humans. survey), of which 500 met inclusion/exclusion criteria. Participants We therefore conducted an anonymous web-based survey of SD were primarily (92.6%) male. The mean age was 23.4 ± 8.7 years use. Participants were asked to report basic demographic and drug- (range 13–68), with 20.2% under 18 years of age. 77.4% lived in the effect information as well has when, how, and why they had used United States. SD. 3.1. Usage patterns 2. Methods Participants reported a mean (±standard deviation) lifetime use We recruited anonymous participants from a popular drug-information web site ± that does not take a position against drug use (www.erowid.org). Individuals visiting of SD products to be 13.3 22.9 (median 6; range 1–250) days the web site’s pages relating to SD saw an announcement about the questionnaire, (Fig. 1). The number of days SD was used in the last 30 and 365 which appeared as a hyperlink reading, “Survey for people who have used Salvia days was 1.5 ± 2.6 (median 1; range 0–24) and 7.5 ± 11.6 (median divinorum”. Prospective participants who clicked on this link then read an informa- 3; range 0–110), respectively. Most individuals (92.6%) typically tion sheet describing the study and its potential risks. Prospective participants from Australia were asked not to participate due to legal restrictions on SD in that country. smoked or vaporized the SD products, with 61.4% of this subset Participants went on to fill out a 39-item survey, requiring approximately 20 min.