Ecstasy as a Social Drug: How MDMA Affects Emotional Response
We pooled data from two studies using similar within-subjects, double-blind designs with only minor methodological differences. Occasional MDMA users attended three (Study 1) or four outpatient sessions (Study 2), separated by at least 5 days. In Study 1, they received placebo, 0.75 and 1.5 mg/kg MDMA, and in Study 2, they received placebo, 0.75 and 1.5 mg/kg MDMA and one of two doses of oxytocin (20 or 40 IU; not reported here). Drug doses were administered at one session each, with no drugs co-administered. In both studies, drug doses were counterbalanced relative to session order, and drug sequences were assigned randomly to participants. At each session, we collected measures of subjective effects, cardiovascular effects and responses to emotional pictures. The measures reported here were the only measures shared between the two studies; thus, additional results from these studies are published separately elsewhere (Kirkpatrick et al., in press; M. C. Wardle and H. de Wit, submitted for publication). In both studies, the pictures were presented as part of a block of tests given during expected peak effect, along with additional measures testing responses to social stimuli only (e.g. identification of emotional expressions). The picture task was the only measure to directly compare social to non-social stimuli. Task order was counterbalanced in both studies to minimize any order effects.
Healthy participants (58 male, 43 female), ages 18–35 were recruited through flyers and online advertisements. Participants completed a 2 h in-person psychiatric and medical evaluation, including physical examination, electrocardiogram, modified structured clinical interview for Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV; First et al., 1996) and self-reported drug and health history. Inclusion criteria were 4–40 times self-reported ecstasy use with no adverse responses; high-school education; English fluency; body mass index >19 and <30; no regular medication (except birth control); no medical conditions contraindicating MDMA; no past year DSM-IV Axis I diagnosis, excluding non-treatment-seeking substance abuse; no history of stimulant dependence; no women who were pregnant or planning a pregnancy. Smokers smoking more than 25 cigarettes per week were also excluded, to avoid nicotine withdrawal during study procedures. Participants were primarily Caucasian (n = 85, 84%), in their 20 s (mean = 24.1 years, s.d. = 4.2), with some college education (mean = 14.8 years, s.d. = 1.4) and moderate recreational drug use (Table 1).
Participants were instructed to consume normal amounts of caffeine and nicotine, and to fast for 2 h before the session. Participants were instructed to refrain from alcohol and over-the-counter drugs for 24 h before and 12 h after the session. Participants were also instructed to refrain from marijuana for 7 days before and 24 h after the session, and from all other recreational drugs for 48 h before and 24 h after the session. Compliance was verified using breath (Alcosensor III, Intoximeters Inc., St. Louis, MO, USA) and urine tests (ToxCup, Branan Medical Corporation, Irvine, CA, USA). Although these were the minimum requirements for compliance, typical abstention times for recreational drugs were longer. Seventy three percent of participants reported no illicit drug use in the last month, and among those who did report last month use, mean time since last use was 12 days. Women not using hormonal contraceptives were scheduled during the follicular phase (White et al., 2002). Female participants were tested for pregnancy prior to each session.
During the consent procedure, participants were told that the purpose of the study was to investigate individual differences in drug responses, and that they might receive a stimulant (e.g. amphetamine or ecstasy), a sedative (e.g. valium), a hallucinogen (e.g. LSD), a cannabinoid (e.g. marijuana) or a placebo. In Study 2 only, participants were also told they might receive a hormone (e.g. oxytocin). Participants agreed to receive any drug from the list, and all sessions were conducted double blind, with neither the experimenter nor the participant informed about the contents of the capsule in advance. This blind was maintained until the debriefing, at which point participants were told which drugs and doses they had received. All participants provided written informed consent, and all procedures were carried out in accordance with the Declaration of Helsinki and approved by the University of Chicago Institutional Review Board.
Sessions were conducted from 9:00 am to 2:00 pm in a comfortable 'living room' style laboratory. At arrival, participants provided breath and urine samples for drug and pregnancy testing, and at 9:15 am, they completed baseline measures of subjective and cardiovascular effects. At 9:30 am, participants ingested a capsule containing MDMA powder (0.75 and 1.5 mg/kg, maximum dose of 125 mg, with lactose filler) or placebo (lactose only), encapsulated in 00 opaque capsules by the University of Chicago Hospitals Investigational Pharmacy. When no measures were scheduled, participants relaxed, watched a movie from a selection available or read. At 10:00 am and every 30–60 min thereafter, subjective and cardiovascular effects were assessed. From 10:40 am to 11:30 am, participants completed computerized tasks including picture ratings. At 2:00 pm, participants completed an end of session questionnaire, which asked the participant to identify the drug that they had received that day. Participants were then discharged provided their subjective and cardiovascular measures had returned to baseline.
To measure subjective mood, we used a Visual Analog Scale (VAS) comprised of 13 adjectives rated on a 1–100 (not at all–extremely) line. This included two 'entactogenic' effects, 'playful' and 'loving' which Bedi et al. (2010) found to be sensitive to the unique effects of MDMA on social emotions, and two typical stimulant-like effects 'elated' and 'stimulated'.
Blood pressure and heart rate were measured using portable monitors (Life Source, A&D Company, Tokyo, Japan). Heart rate results were similar in dose dependence and time course to blood pressure, so we use mean arterial pressure (MAP; [Systolic BP + 2 × Diastolic BP]/3) as our measure of cardiovascular effects of the drug.
We used pictures from the International Affective Picture System (IAPS; Lang et al., 1999) as emotional stimuli. IAPS pictures are normatively rated on valence (positivity vs negativity) and arousal. Although IAPS pictures are not normatively rated for social relevance, based on previous research (Cacioppo et al., 2009; Gros et al., 2009) we defined 'social' pictures as those depicting at least two people or parts of people (e.g. two people talking, a hand pointing a gun at another person), and 'non-social' pictures as those depicting no people or parts of people (e.g. a slice of pizza, a car accident with no bodies visible). Thus, there were six subtypes: social/negative, non-social/negative, social/neutral, non-social/neutral, social/positive and non-social/positive. To avoid adaptation, at each session the participant saw a different set of pictures. We constructed 3 sets of 54 pictures for Study 1, with 9 pictures per subtype per set, and 4 sets of 36 pictures for Study 2, with 6 pictures per subtype per set. We attempted to match valence and arousal across sets and social vs non-social pictures, using the normative ratings provided with the IAPS pictures (Lang et al., 1999). We counterbalanced picture set with drug dose, such that each picture set was paired approximately the same number of times with each drug dose. Pictures were presented in fixed random order, with no more than two of the same valence in a row. Picture trials consisted of a 3 s pre-picture fixation, a 6 s picture period, then subjective ratings. Participants rated pictures using the evaluative space grid (Larsen et al., 2009), which allows independent 0 (not at all) to 4 (extreme) ratings of positivity and negativity, and a 0 (not at all) to 9 (extreme) rating of arousal.
At the end of each session, we asked participants to identify the class of drug that they thought they had received that day as '1. a stimulant (e.g. amphetamine or ecstasy), 2. A hallucinogen (e.g. LSD), 3. A sedative (e.g. Valium), 4. A cannabinoid (e.g. marijuana), or 5. A placebo'.
We used linear mixed effect models (LMEMs) in the lme4 package (v 0.999999-0; Bates et al., 2011) of the R statistical computing environment (v. 2.15.2; R Development Core Team, 2011) as our primary statistical approach.
For subjective and cardiovascular measures, which were taken repeatedly across sessions, we first summarized each session by calculating area under the curve (AUC) relative to the participant's baseline score for that session. We then used the AUC scores in LMEMs (one each for VAS playful, VAS loving, VAS elated, VAS stimulated and MAP) with dose as an independent (fixed) factor, and participant as a random effect.
For response to emotional stimuli we constructed mean ratings of arousal, positivity and negativity for each picture subtype within each session. We then used these means in LMEMs (one each for arousal, positivity and negativity) using dose, valence of picture and social content of picture as independent (fixed) factors and participant and dose within participant as random effects.
In all analyses we examined any dose effects using orthogonal polynomial contrasts, which constituted our primary analyses of interest. We tested for both linear effects of drug (which would suggest a dose-dependent relationship between dose and outcome), and quadratic effects (which would suggest a U shaped relationship between dose and outcome). If a significant effect of drug was identified, we then used paired t-tests comparing each dose to placebo to further describe the effect and identify the effective doses. We also included participant sex and study (Study 1 vs Study 2) as potential moderators. We additionally examined number of self-reported previous occasions of ecstasy use as a potential continuous moderator, but it did not affect any of the outcomes in this study, and is omitted from the final models for simplicity. Finally, we included a fixed session effect, to account for any order effects. Effect sizes are reported as unstandardized coefficients (B) with standard errors (s.e.). We calculated P-values using the t distribution with n − 1 degrees of freedom (see Wardle and de Wit, 2012 for rationale).
Picture sets for Study 1 were the same as in Wardle and de Wit (2012), and can be found in the footnote on p. 143 of that article. IAPS numbers for each subset of Study 2, followed by IAPS normative mean valence (V, 1 = extremely unpleasant − 9 = extremely pleasant) and mean arousal (A, 1 = extremely unarousing − 9 = extremely arousing):
Set 1 non-social/negative = 7380, 9911, 9180, 9373, 1280, 7360, V = 3.06, A = 5.29;
Set 1 social/negative = 9425, 9903, 6561, 9584, 2694, 9926, V = 3.15, A = 5.24;
Set 1 non-social/neutral = 7830, 7190, 7285, 7207, 1935, 7055, V = 5.23, A = 3.77;
Set 1 social/neutral = 7620, 2580, 9700, 2595, 2397, 2597, V = 5.28, A = 3.58;
Set 1 non-social/positive = 1640, 7352, 5450, 7480, 5700, 5260, V = 6.90, A = 5.26;
Set 1 social/positive = 4606, 8467, 3291, 8116, 8420, 2216, V = 7.03, A = 5.37;
Set 2 non-social/negative = 9560, 9301, 1274, 1111, 1220, 9008, V = 2.96, A = 5.23;
Set 2 social/negative = 9420, 2053, 3216, 6562, 6836, 2718, V = 3.06, A = 5.22;
Set 2 non-social/neutral = 7500, 1616, 5661, 9472, 7546, 7054, V = 5.02, A = 3.89;
Set 2 social/neutral = 8010, 4605, 2485, 2393, 2606, 2593, V = 5.38, A = 3.81;
Set 2 non-social/positive = 1660, 7289, 8500, 7508, 5480, 8502, V = 6.97, A = 5.28;
Set 2 social/positive = 8600, 4598, 4601, 4599, 7502, 8496, V = 7.00, A = 5.38;
Set 3 non-social/negative = 9300, 9620, 9290, 9471, 9110, 9480, V = 3.05, A = 5.09;
Set 3 social/negative = 6212, 6022, 2700, 2455, 9045, 9594, V = 3.00, A = 5.07;
Set 3 non-social/neutral = 7100, 5120, 7590, 7183, 7037, 7242, V = 5.01, A = 3.51;
Set 3 social/neutral = 2850, 9582, 2695, 9913, 2579, 2396, V = 4.71, A = 4.06;
Set 3 non-social/positive = 7284, 7481, 8162, 5849, 8170, 5660, V = 6.87, A = 5.01;
Set 3 social/positive = 2358, 2605, 2344, 4624, 2352, 8499, V = 6.88, A = 4.95;
Set 4 non-social/negative = 9140, 9320, 9470, 9621, 9010, 9390, V = 3.12, A = 4.90;
Set 4 social/negative = 9421, 6838, 2691, 4621, 4635, 2312, V = 3.07, A = 4.97;
Set 4 non-social/neutral = 5510, 7233, 7283, 7182, 5395, 1947, V = 5.35, A = 3.67;
Set 4 social/neutral = 4000, 7496, 2272, 2435, 2704, 7506, V = 5.21, A = 4.34;
Set 4 non-social/positive = 7450, 5991, 7410, 8531, 5600, 7270, V = 7.00, A = 5.00;
Set 4 social/positive = 4625, 2594, 4650, 2373, 2345, 4626, V = 6.90, A = 5.04
Materials and Methods
Study Design
We pooled data from two studies using similar within-subjects, double-blind designs with only minor methodological differences. Occasional MDMA users attended three (Study 1) or four outpatient sessions (Study 2), separated by at least 5 days. In Study 1, they received placebo, 0.75 and 1.5 mg/kg MDMA, and in Study 2, they received placebo, 0.75 and 1.5 mg/kg MDMA and one of two doses of oxytocin (20 or 40 IU; not reported here). Drug doses were administered at one session each, with no drugs co-administered. In both studies, drug doses were counterbalanced relative to session order, and drug sequences were assigned randomly to participants. At each session, we collected measures of subjective effects, cardiovascular effects and responses to emotional pictures. The measures reported here were the only measures shared between the two studies; thus, additional results from these studies are published separately elsewhere (Kirkpatrick et al., in press; M. C. Wardle and H. de Wit, submitted for publication). In both studies, the pictures were presented as part of a block of tests given during expected peak effect, along with additional measures testing responses to social stimuli only (e.g. identification of emotional expressions). The picture task was the only measure to directly compare social to non-social stimuli. Task order was counterbalanced in both studies to minimize any order effects.
Participants
Healthy participants (58 male, 43 female), ages 18–35 were recruited through flyers and online advertisements. Participants completed a 2 h in-person psychiatric and medical evaluation, including physical examination, electrocardiogram, modified structured clinical interview for Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV; First et al., 1996) and self-reported drug and health history. Inclusion criteria were 4–40 times self-reported ecstasy use with no adverse responses; high-school education; English fluency; body mass index >19 and <30; no regular medication (except birth control); no medical conditions contraindicating MDMA; no past year DSM-IV Axis I diagnosis, excluding non-treatment-seeking substance abuse; no history of stimulant dependence; no women who were pregnant or planning a pregnancy. Smokers smoking more than 25 cigarettes per week were also excluded, to avoid nicotine withdrawal during study procedures. Participants were primarily Caucasian (n = 85, 84%), in their 20 s (mean = 24.1 years, s.d. = 4.2), with some college education (mean = 14.8 years, s.d. = 1.4) and moderate recreational drug use (Table 1).
Participants were instructed to consume normal amounts of caffeine and nicotine, and to fast for 2 h before the session. Participants were instructed to refrain from alcohol and over-the-counter drugs for 24 h before and 12 h after the session. Participants were also instructed to refrain from marijuana for 7 days before and 24 h after the session, and from all other recreational drugs for 48 h before and 24 h after the session. Compliance was verified using breath (Alcosensor III, Intoximeters Inc., St. Louis, MO, USA) and urine tests (ToxCup, Branan Medical Corporation, Irvine, CA, USA). Although these were the minimum requirements for compliance, typical abstention times for recreational drugs were longer. Seventy three percent of participants reported no illicit drug use in the last month, and among those who did report last month use, mean time since last use was 12 days. Women not using hormonal contraceptives were scheduled during the follicular phase (White et al., 2002). Female participants were tested for pregnancy prior to each session.
During the consent procedure, participants were told that the purpose of the study was to investigate individual differences in drug responses, and that they might receive a stimulant (e.g. amphetamine or ecstasy), a sedative (e.g. valium), a hallucinogen (e.g. LSD), a cannabinoid (e.g. marijuana) or a placebo. In Study 2 only, participants were also told they might receive a hormone (e.g. oxytocin). Participants agreed to receive any drug from the list, and all sessions were conducted double blind, with neither the experimenter nor the participant informed about the contents of the capsule in advance. This blind was maintained until the debriefing, at which point participants were told which drugs and doses they had received. All participants provided written informed consent, and all procedures were carried out in accordance with the Declaration of Helsinki and approved by the University of Chicago Institutional Review Board.
Procedure
Sessions were conducted from 9:00 am to 2:00 pm in a comfortable 'living room' style laboratory. At arrival, participants provided breath and urine samples for drug and pregnancy testing, and at 9:15 am, they completed baseline measures of subjective and cardiovascular effects. At 9:30 am, participants ingested a capsule containing MDMA powder (0.75 and 1.5 mg/kg, maximum dose of 125 mg, with lactose filler) or placebo (lactose only), encapsulated in 00 opaque capsules by the University of Chicago Hospitals Investigational Pharmacy. When no measures were scheduled, participants relaxed, watched a movie from a selection available or read. At 10:00 am and every 30–60 min thereafter, subjective and cardiovascular effects were assessed. From 10:40 am to 11:30 am, participants completed computerized tasks including picture ratings. At 2:00 pm, participants completed an end of session questionnaire, which asked the participant to identify the drug that they had received that day. Participants were then discharged provided their subjective and cardiovascular measures had returned to baseline.
Subjective Mood
To measure subjective mood, we used a Visual Analog Scale (VAS) comprised of 13 adjectives rated on a 1–100 (not at all–extremely) line. This included two 'entactogenic' effects, 'playful' and 'loving' which Bedi et al. (2010) found to be sensitive to the unique effects of MDMA on social emotions, and two typical stimulant-like effects 'elated' and 'stimulated'.
Cardiovascular Measures
Blood pressure and heart rate were measured using portable monitors (Life Source, A&D Company, Tokyo, Japan). Heart rate results were similar in dose dependence and time course to blood pressure, so we use mean arterial pressure (MAP; [Systolic BP + 2 × Diastolic BP]/3) as our measure of cardiovascular effects of the drug.
Responses to Emotional Stimuli
We used pictures from the International Affective Picture System (IAPS; Lang et al., 1999) as emotional stimuli. IAPS pictures are normatively rated on valence (positivity vs negativity) and arousal. Although IAPS pictures are not normatively rated for social relevance, based on previous research (Cacioppo et al., 2009; Gros et al., 2009) we defined 'social' pictures as those depicting at least two people or parts of people (e.g. two people talking, a hand pointing a gun at another person), and 'non-social' pictures as those depicting no people or parts of people (e.g. a slice of pizza, a car accident with no bodies visible). Thus, there were six subtypes: social/negative, non-social/negative, social/neutral, non-social/neutral, social/positive and non-social/positive. To avoid adaptation, at each session the participant saw a different set of pictures. We constructed 3 sets of 54 pictures for Study 1, with 9 pictures per subtype per set, and 4 sets of 36 pictures for Study 2, with 6 pictures per subtype per set. We attempted to match valence and arousal across sets and social vs non-social pictures, using the normative ratings provided with the IAPS pictures (Lang et al., 1999). We counterbalanced picture set with drug dose, such that each picture set was paired approximately the same number of times with each drug dose. Pictures were presented in fixed random order, with no more than two of the same valence in a row. Picture trials consisted of a 3 s pre-picture fixation, a 6 s picture period, then subjective ratings. Participants rated pictures using the evaluative space grid (Larsen et al., 2009), which allows independent 0 (not at all) to 4 (extreme) ratings of positivity and negativity, and a 0 (not at all) to 9 (extreme) rating of arousal.
Drug Identifications
At the end of each session, we asked participants to identify the class of drug that they thought they had received that day as '1. a stimulant (e.g. amphetamine or ecstasy), 2. A hallucinogen (e.g. LSD), 3. A sedative (e.g. Valium), 4. A cannabinoid (e.g. marijuana), or 5. A placebo'.
Statistical Analyses
We used linear mixed effect models (LMEMs) in the lme4 package (v 0.999999-0; Bates et al., 2011) of the R statistical computing environment (v. 2.15.2; R Development Core Team, 2011) as our primary statistical approach.
For subjective and cardiovascular measures, which were taken repeatedly across sessions, we first summarized each session by calculating area under the curve (AUC) relative to the participant's baseline score for that session. We then used the AUC scores in LMEMs (one each for VAS playful, VAS loving, VAS elated, VAS stimulated and MAP) with dose as an independent (fixed) factor, and participant as a random effect.
For response to emotional stimuli we constructed mean ratings of arousal, positivity and negativity for each picture subtype within each session. We then used these means in LMEMs (one each for arousal, positivity and negativity) using dose, valence of picture and social content of picture as independent (fixed) factors and participant and dose within participant as random effects.
In all analyses we examined any dose effects using orthogonal polynomial contrasts, which constituted our primary analyses of interest. We tested for both linear effects of drug (which would suggest a dose-dependent relationship between dose and outcome), and quadratic effects (which would suggest a U shaped relationship between dose and outcome). If a significant effect of drug was identified, we then used paired t-tests comparing each dose to placebo to further describe the effect and identify the effective doses. We also included participant sex and study (Study 1 vs Study 2) as potential moderators. We additionally examined number of self-reported previous occasions of ecstasy use as a potential continuous moderator, but it did not affect any of the outcomes in this study, and is omitted from the final models for simplicity. Finally, we included a fixed session effect, to account for any order effects. Effect sizes are reported as unstandardized coefficients (B) with standard errors (s.e.). We calculated P-values using the t distribution with n − 1 degrees of freedom (see Wardle and de Wit, 2012 for rationale).
Picture sets for Study 1 were the same as in Wardle and de Wit (2012), and can be found in the footnote on p. 143 of that article. IAPS numbers for each subset of Study 2, followed by IAPS normative mean valence (V, 1 = extremely unpleasant − 9 = extremely pleasant) and mean arousal (A, 1 = extremely unarousing − 9 = extremely arousing):
Set 1 non-social/negative = 7380, 9911, 9180, 9373, 1280, 7360, V = 3.06, A = 5.29;
Set 1 social/negative = 9425, 9903, 6561, 9584, 2694, 9926, V = 3.15, A = 5.24;
Set 1 non-social/neutral = 7830, 7190, 7285, 7207, 1935, 7055, V = 5.23, A = 3.77;
Set 1 social/neutral = 7620, 2580, 9700, 2595, 2397, 2597, V = 5.28, A = 3.58;
Set 1 non-social/positive = 1640, 7352, 5450, 7480, 5700, 5260, V = 6.90, A = 5.26;
Set 1 social/positive = 4606, 8467, 3291, 8116, 8420, 2216, V = 7.03, A = 5.37;
Set 2 non-social/negative = 9560, 9301, 1274, 1111, 1220, 9008, V = 2.96, A = 5.23;
Set 2 social/negative = 9420, 2053, 3216, 6562, 6836, 2718, V = 3.06, A = 5.22;
Set 2 non-social/neutral = 7500, 1616, 5661, 9472, 7546, 7054, V = 5.02, A = 3.89;
Set 2 social/neutral = 8010, 4605, 2485, 2393, 2606, 2593, V = 5.38, A = 3.81;
Set 2 non-social/positive = 1660, 7289, 8500, 7508, 5480, 8502, V = 6.97, A = 5.28;
Set 2 social/positive = 8600, 4598, 4601, 4599, 7502, 8496, V = 7.00, A = 5.38;
Set 3 non-social/negative = 9300, 9620, 9290, 9471, 9110, 9480, V = 3.05, A = 5.09;
Set 3 social/negative = 6212, 6022, 2700, 2455, 9045, 9594, V = 3.00, A = 5.07;
Set 3 non-social/neutral = 7100, 5120, 7590, 7183, 7037, 7242, V = 5.01, A = 3.51;
Set 3 social/neutral = 2850, 9582, 2695, 9913, 2579, 2396, V = 4.71, A = 4.06;
Set 3 non-social/positive = 7284, 7481, 8162, 5849, 8170, 5660, V = 6.87, A = 5.01;
Set 3 social/positive = 2358, 2605, 2344, 4624, 2352, 8499, V = 6.88, A = 4.95;
Set 4 non-social/negative = 9140, 9320, 9470, 9621, 9010, 9390, V = 3.12, A = 4.90;
Set 4 social/negative = 9421, 6838, 2691, 4621, 4635, 2312, V = 3.07, A = 4.97;
Set 4 non-social/neutral = 5510, 7233, 7283, 7182, 5395, 1947, V = 5.35, A = 3.67;
Set 4 social/neutral = 4000, 7496, 2272, 2435, 2704, 7506, V = 5.21, A = 4.34;
Set 4 non-social/positive = 7450, 5991, 7410, 8531, 5600, 7270, V = 7.00, A = 5.00;
Set 4 social/positive = 4625, 2594, 4650, 2373, 2345, 4626, V = 6.90, A = 5.04
SHARE