- Email: [email protected]
Reward sensitivity and outcome expectancies as predictors of ecstasy use in young adults
Addictive Behaviors 36 (2011) 1337–1340
Contents lists available at ScienceDirect
Addictive Behaviors
Short Communication
Reward sensitivity and outcome expectancies as predictors of ecstasy use in young adults Melanie J. Smerdon, Andrew J.P. Francis ⁎ Psychology, Health Sciences, RMIT University, Bundoora, 3083, Australia
a r t i c l e
i n f o
Keywords: Ecstasy Reward sensitivity Outcome expectancies Self-confidence
a b s t r a c t Extending recent studies showing that sensitivity to reward and outcome expectancies are associated with problematic alcohol and cannabis use, we undertook to determine if similar relationships would hold for ecstasy. One hundred and twenty five males and females aged between 18 and 35 years were recruited from RMIT University and through snowball sampling. Participants completed a questionnaire package measuring frequency, amount and first age of ecstasy use, sensitivity to reward and punishment as well as outcome expectancies relating to ecstasy use. Frequency of ecstasy use was significantly related to reward sensitivity (p b .05) and positive outcome expectancies (p b .01). Regression analysis revealed significant prediction of ecstasy use by study variables, with expectations of increased confidence making the largest individual contribution. Multiple intervention points are suggested by the results of this study, within a largely cognitive-based framework. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction The psychological and physiological mechanisms underlying hazardous use of illicit substances are yet to be fully elucidated, although in the past ten years the importance of individual personality traits and cognitions has become apparent with respect to predicting vulnerability (Dawe & Loxton, 2004; Lubman, 2007). In particular, it has been determined that individuals who frequently use alcohol and cannabis report higher sensitivity to reward and more positive outcome expectancies (DePino, 2009; Simons & Arens, 2007); and now interest is turning to other substances that are taken for their rewarding properties (Egan, Kambouropoulos, & Staiger, 2010; Franken & Muris, 2006; Van der Poel, Rodenburg, Dijkstra, Stoele, & Van de Mheen, 2009). This study examines the relevance of sensitivity to reward and outcome expectancies for the prediction of ‘ecstasy’ (3,4-Methylenedioxymethamphetamine) use. Gray's (1987) personality theory of approach-inhibition has been one of the most plausible psychological explanations for why some individuals take drugs and find them highly rewarding whilst others find them less so or not at all (Franken, Muris, & Georgieva, 2006; Perry & Carroll, 2008; Schramm-Sapyta, Walker, Caster, Levin, & Kuhn, 2009). Studies have consistently found that individuals high in behavioural approach are more likely to be vulnerable to the use and abuse of both alcohol (DePino, 2009; Franken & Muris, 2006;
⁎ Corresponding author at: Division of Psychology, School of Health Sciences, RMIT University, Bundoora, 3083, Australia. Tel.: + 61 3 9925 7782; fax: + 61 3 9925 7303. E-mail address: [email protected] (A.J.P. Francis). 0306-4603/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.addbeh.2011.07.027
Kambouropoulos & Staiger, 2007; Lubman, 2007; Pardo, Aguilar, Molinuevo, & Torrubia, 2007) and cannabis (DePino, 2009). Drawing on the theoretical basis of Bandura's (1986) social cognitive theory, another significant psychological predictor of substance use is outcome expectancies. Substance-related outcome expectancies may be positive or negative, and it is theorised that if a person perceives a substance as having more positive than negative potential outcomes then they are more likely to consume that substance (Martins, Storr, Alexandre, & Chilcoat, 2007). In support of these theoretical predictions, DePino (2009) confirmed previous research reporting positive outcome expectancies were significantly related to alcohol use. In addition, hazardous drinkers entertained more positive outcome expectancies than nonhazardous drinkers, and these relationships were also found for cannabis users. These data have led researchers to question whether similar relationships might also hold for another popular recreational drug, ecstasy (Clemens, McGregor, Hunt, & Cornish, 2007; Egan et al., 2010; Hatala, 2008). Ecstasy is the third most commonly used illicit drug by Australians aged between 14 and 29 years, with similar levels of use being observed in other countries (Costa, 2003; Degenhardt, Barker, & Topp, 2003; Kaye, Darke, & Duflou, 2009). To date, elevated impulsivity and positive outcome expectancies have been associated with ecstasy use (Egan et al., 2010; Hanson, Luciana, & Sullwold, 2008; Martins et al., 2007; Morgan, 1998; Schilt, Goudriaan, Koeter, Van der Brink, & Schmand, 2009; Yu & Ko, 2006), although further research is required to confirm these relationships. By extension of the theoretical and empirical rationale established for alcohol and cannabis, the aim of this study was to examine individual differences in reward sensitivity and outcome expectancies
1338
M.J. Smerdon, A.J.P. Francis / Addictive Behaviors 36 (2011) 1337–1340
as predictors of ecstasy use in young adults. In accord with previous research, it was hypothesised that there would be significant positive relationships observed between frequency of ecstasy use and sensitivity to reward. It was further hypothesised that positive outcome expectancies would also be associated with higher frequency of ecstasy use, as is the case for alcohol and cannabis from previous research. Finally, it was expected that sensitivity to reward and positive outcome expectancies would, overall, significantly predict levels of ecstasy use within a multiple regression model. 2. Method 2.1. Participants The sample consisted of 125 participants (Mage = 25.35 years, SDage = 3.62 years), ranging in age from 18 years to 35 years of which 60 were male (Mage = 26.10 years, SD = 4.11 years) and 62 were female (Mage = 24.57 years, SD = 2.85 years). Three participants (2.4%) did not provide information on their gender and six participants (4.8%) did not provide information about their age. Most participants (70.4%) reported they were employed full time, the remaining were either employed part time (7.8%), students (9.3%) or causally employed (9.3%). Persons aged less than 18 years and over 35 years were excluded. 2.2. Materials Participants completed a questionnaire package which examined reward and punishment sensitivity and ecstasy outcome expectancies. The questionnaire also required participants to answer questions pertaining to frequency of ecstasy use, typical amount consumed per occasion and age first used. Frequency of ecstasy use was indicated by a response in one of the following categories: every day, approximately once per week, approximately once per month, every few months, once or twice a year, or not currently using. Categories were coded as increasing numeric values, with higher scores on the frequency variable therefore indicating less frequent ecstasy use. 2.2.1. The Sensitivity to Punishment and Sensitivity to Reward Questionnaire (SPSRQ) The SPSRQ measures levels of sensitivity to punishment (SP) and sensitivity to reward (SR) (Torrubia, Avila, Molto, & Caseras, 2001). Cronbach's alpha coefficients range between .82 and .83 for SP and .75 and .78 for SR (Torrubia et al., 2001). 2.2.2. The Ecstasy Expectancies Questionnaire (EEQ) Using a similar procedure to that utilised by DePino (2009) for measurement of cannabis expectancies, an ecstasy expectancies questionnaire (EEQ) was developed by adapting the Drinking Expectancies Questionnaire (DEQ; Young & Oei, 1996). The 5 subscales of the EEQ are Negative Consequences of Ecstasy Use (EEQ-NC), Increased Confidence (EEQ-IC), Increased Sexual Interest (EEQ-ISI), Cognitive Enhancement (EEQ-CE) and Tension Reduction (EEQ-TR). 2.3. Procedure Following approval from the RMIT University Human Research Ethics Committee, participants were recruited from RMIT University through paper advertisements and through snowball sampling, which is a commonly used and useful approach with research on drug use (Davidson & Parrott, 1997; Hinchliff, 2001). Consent was implied through the return of the survey.
3. Results 3.1. Data screening 3.1.1. Missing values analysis Discrete missing values were replaced by the mean of the available values, and subscales recalculated. Twenty two participants were found to have not completed at least one of the questionnaires making up the survey. These participants were not included in analyses involving those particular questionnaires, but were not excluded from the sample overall. 3.1.2. Normality and univariate outliers After discrete missing values were replaced, the scale and subscale distributions were visually examined for departures from normality and outliers. Examinations of the histograms and box plots showed there were no variables that deviated substantively from normality. 3.2. Reliabilities of scales and descriptive statistics Internal consistencies for sensitivity to reward (SR) and punishment (SP) were .81 and .84 respectively, similar to the alpha coefficients stated by Torrubia et al. (2001) who suggested a range of .82 and .83 for SP and .75 and .78 for SR. Scale reliabilities for ecstasy outcome expectancies subscales ranged from .55 to .94 which, except for EEQ-TR (.55), are consistent with those found by Young and Oei (1996). Considering EEQ-TR was found to have a lower reliability measure, EEQ-TR was interpreted with caution. Of the 125 participants, 57% indicated they had used ecstasy. Of this user group, 3% used ecstasy monthly, 21% every few months and 21% once or twice per year. A single individual reported using ecstasy weekly and none reported using ecstasy daily. Of the user group, 42% indicated they were ‘not currently using’. 3.3. Correlation analysis Pearson correlations were conducted to examine the associations between frequency of ecstasy use, sensitivity to reward or punishment and outcome expectancies (Table 1). Frequency of ecstasy use showed a small, significant negative correlation with sensitivity to reward, indicating that more frequent users were more sensitive to reward. There were significant medium-magnitude, negative correlations between frequency of ecstasy use and each of EEQ-IC, EEQ-TR and EEQ-TPE, as well as a small, negative correlation with EEQ-TE; indicating that more positive evaluations on these outcome expectancy domains were associated with more frequent ecstasy use. A small, significant and positive correlation was also observed between EEQ-ISI and frequency of ecstasy use. There were no significant correlations between the amount of ecstasy consumed on each occasion of use and either sensitivity to reward or punishment, or any of the outcome expectancy domains. The only significant correlation with reported age of first use was the increased confidence domain (EEQ-IC) of the outcome expectancies questionnaire, r = 0.25, n = 55, p = .03, indicating that expectation of increased confidence was associated with earlier ecstasy use. 3.4. Regression analysis To determine the extent to which sensitivity to reward and ecstasy outcome expectancies (EEQ-IC, EEQ-TR & EEQ-TPE) predicted frequency of ecstasy use, a standard multiple regression analysis was conducted. Outcome expectancies concerned with negative consequences were not included in the regression since the correlational analysis revealed that EEQ-NC had no significant relationship with frequency of ecstasy use. Furthermore, the inclusion of the EEQ subscales ISI and TE would be likely to cause multicollinearity, which is a violation of assumption
M.J. Smerdon, A.J.P. Francis / Addictive Behaviors 36 (2011) 1337–1340
1339
Table 1 Pearson's correlations between personality variables, outcome expectancies and frequency of ecstasy use. 1 Freq. E use SR SP EEQ-NC EEQ-IC EEQ-ISI EEQ-CE EEQ-TR EEQ-TPE EEQ-TE
1 −.18⁎ .13 .15 −.42⁎⁎ .18⁎ .01 −.31⁎ −.38⁎ −.18⁎
2
3
4
5
6
7
8
9
1 .50⁎⁎ .67⁎⁎ .7⁎⁎
1 .61⁎⁎ .57⁎⁎
1
10
1 .11 −.01 .19* .04 −.16 −.02 .12 .08
1 .09 .02 .02 −.08 −.18⁎ −.03 .03
1 .42⁎⁎ .48⁎⁎ .53⁎⁎ .33⁎⁎ .51⁎⁎ .82⁎⁎
1 .13 .49⁎⁎ .47⁎⁎ .97⁎⁎ .84⁎⁎
1
.32⁎⁎ .13 .25⁎⁎ .4⁎⁎
.91⁎⁎
1
Note: Freq. E use = Frequency of ecstasy use; SR = Sensitivity to Reward; SP = Sensitivity to Punishment; EEQ-NC = Ecstasy Outcome Expectancies Questionnaire Negative Consequences subscale; EEQ-IC = Ecstasy Outcome Expectancies Questionnaire Increased Confidence Subscale; EEQ-ISI = Ecstasy Outcome Expectancies Questionnaire Increased Sexual Interest Subscale; EEQ-CE = Ecstasy Outcome Expectancies Questionnaire Cognitive Enhancement Subscale; EEQ-TR = Ecstasy Outcome Expectancies Questionnaire Tension Reduction Subscale; EEQ-TPE = Ecstasy Outcome Expectancies Questionnaire Global Positive Expectancy Subscale; and EEQ-TE = Ecstasy Outcome Expectancies Questionnaire Summative Expectancy Score Subscale. ⁎ p b .05 one tailed. ⁎⁎ p b .01 one tailed.
testing for regression analysis (Coakes & Steed, 2007), and were therefore also excluded. Results of the multiple regression are displayed in Table 2. The model significantly predicted frequency of ecstasy use, F(4,96) = 10.31, p b .01, accounting for 27.1% of the variance in ecstasy use. Further examination revealed that EEQ-TR, EEQ-IC and EEQ-TPE were the only significant predictors of ecstasy use, but EEQ-IC had the largest unique contribution to the prediction ecstasy use behaviour. 4. Discussion Results supported the hypothesis that there would be a relationship between frequency of ecstasy use and sensitivity to reward. Results also supported the hypothesis that domains of positive outcome expectancy would be associated with higher frequency of ecstasy use, as is the case for alcohol and cannabis. Finally, the results supported the hypothesis that study variables would, overall, significantly predict levels of ecstasy use within a regression model analysis. The hypothesis that sensitivity to reward would be associated with ecstasy use was supported, although the relationship was relatively weak. The association found between reward sensitivity and ecstasy use is consistent with recent findings by Egan et al. (2010) which showed that more frequent ecstasy use was associated with higher self-reported sensitivity to reward. Support is also found for Gray's (1987) neurological model of substance use, which suggests being sensitive to reward is a risk factor for potential drug use (Acton, 2003; Franken, 2002). Despite the significant association between the variables, regression analysis suggested that sensitivity to reward did not significantly and uniquely predict ecstasy use behaviour in the current sample. There are several potential explanations for this finding. First, the correlation analysis revealed only a small magnitude negative correlation between
Table 2 Standard and unstandardised beta coefficients, standard error, squared semi-partial correlations and significance associated with prediction of frequency of ecstasy use.
ecstasy use and reward sensitivity, and so it might be expected that the regression analysis (as a more conservative statistic) could reveal a nonsignificant result. Second, the current study illustrates the important extent to which individual differences in cognitions contribute to substance use. Results revealed that positive outcome expectancies were significantly associated with frequency of ecstasy use and were a significant predictor of ecstasy use behaviour. Regression analysis revealed having outcome expectancies of increased confidence, tension reduction and overall positive expectancies predicted how frequently a person would use ecstasy, consistent with findings by Martins et al. (2007) and Egan et al. (2010). The results are crucial as they provide information as to why people continue to take ecstasy despite all risks: expected positive outcomes superseding any expected negative outcomes. Thus within the context of the regression model, reward sensitivity was not significant as a unique predictor relative to the more substantial variation predicted by the cognitive variables. 5. Conclusions Findings of this study indicate reward sensitivity and outcome expectancies are associated with ecstasy use, consistent with the theoretical framework of previous research on alcohol and cannabis. Of particular interest is the finding of an expectation that ecstasy use would increase self-confidence. This expectation was associated with greater frequency of ecstasy use (making the largest unique contribution to the prediction ecstasy use behaviour in regression modelling) and with earlier use of the drug. Self-confidence, as an individual factor and also as a drug-associated expectation, is therefore indicated as an important and appropriate focus for both further research and potential intervention. Modification of outcome expectancies towards drugs, including ecstasy, within a broader cognitive-behavioural framework could hold the key to changing an individual's drug usage pattern. Role of funding sources Funding for this study was provided by the Division of Psychology, RMIT University.
2
Predictor
B
SE B
β
Sr
SR EEQ-IC EEQ-TR EEQ-TPE
−.03 −.24 −.36 .18
.03 .06 .09 .05
−.08 − 1.7⁎⁎ −.57⁎⁎ 1.6⁎
.007 .12 .11 .09
Note: N = 101; B = unstandardised beta coefficients; SE B = standard error of unstandardised beta coefficients; β = standardised beta coefficients; and Sr2 = squared semi partial correlations. ⁎ p b .01. ⁎⁎ p b .001.
Contributors Andrew Francis designed the study. Melanie Smerdon performed literature searches, collected data and formulated the first draft of the manuscript. All authors contributed to and have approved the final manuscript. Andrew Francis and Melanie Smerdon analysed the data.
Conflict of interest There are no conflicts of interest by any author.
1340
M.J. Smerdon, A.J.P. Francis / Addictive Behaviors 36 (2011) 1337–1340
References Acton, G. S. (2003). Measurement of impulsivity in a hierarchical model of personality traits: Implications of substance use. Substance Use & Misuse, 38, 67–83. Bandura, A. (1986). Social foundations of thought and action; a social cognitive theory. Englewood Cliffs: Prentice Hall. Clemens, K. J., McGregor, I. S., Hunt, G. E., & Cornish, J. L. (2007). MDMA, methamphetamine and their combination: Possible lessons for party drug users from recent preclinical research. Drug and Alcohol Review, 26, 9–15. Coakes, S., & Steed, L. (2007). Multiple regression. In SPSS analysis without anguish: Version 14.0 for windows (pp. 135–146). Milton: John Wiley & Sons Australia, Ltd. Costa, A. M. (2003). Ecstasy and amphetamines global survey 2003. Presented to the United Nations office of crime and drugs in ViennaUnited Nations Publication Sales No. E.03.XI.1592-1-148164-3. Davidson, D., & Parrott, A. C. (1997). Ecstasy (MDMA) in recreational users: Self reported psychological and physiological effects. Human Psychopharmacology, 12, 221–226. Dawe, S., & Loxton, N. J. (2004). The role of impulsivity in the development of substance use and eating disorders. Neuroscience Biobehavioral Review, 28, 343–351. Degenhardt, L., Barker, B., & Topp, L. (2003). Patterns of ecstasy use in Australia: Findings from a national household survey. University of NSW, Australia: National Drug and Alcohol Research Centre. doi:10.1046/j.1360-0443.2003.00622.x. DePino, E. (2009). Reward sensitivity and outcome expectancies predict alcohol and cannabis use in young adults. Unpublished doctoral thesis, RMIT University, Victoria. Egan, S. T., Kambouropoulos, N., & Staiger, P. K. (2010). Rash-impulsivity, reward-drive and motivations to use ecstasy. Personality and Individual Differences, 48, 670–675. Franken, I. H. A. (2002). Behavioral approach system (BAS) sensitivity predicts alcohol craving. Personality and Individual Differences, 32, 349–355. Franken, I. H. A., & Muris, P. (2006). BIS/BAS personality characteristics and college students' substance use. Personality and Individual Differences, 40, 1497–1503. Franken, I. H. A., Muris, P., & Georgieva, I. (2006). Gray's model of personality and addiction. Addictive Behaviors, 31, 399–403. Gray, J. A. (1987). Perspectives on anxiety and impulsivity: A commentary. Journal of Research in Personality, 21, 493–509. Hanson, K. L., Luciana, M., & Sullwold, K. (2008). Reward-related decision making deficits and elevated impulsivity among MDMA and other drug users. Drug and Alcohol Dependence, 96, 99–110. Hatala, E. M. (2008). Characteristics and predictors of ecstasy (MDMA) use during college. Unpublished Doctoral Thesis, Temple University, Missouri. Hinchliff, S. (2001). The meaning of ecstasy use and clubbing to women in the last 1990's. The International Journal on Drug Policy, 12, 455–468.
Kambouropoulos, N., & Staiger, P. K. (2007). Personality, behavioural and affective characteristics of hazardous drinkers. Personality and Individual Differences, 42, 213–224. Kaye, S., Darke, S., & Duflou, J. (2009). Methylenedioxymethamphetamine (MDMA)related fatalities in Australia: Demographics, circumstances, toxicology and major organ pathology. Drug and Alcohol Dependence, 104, 254–261. Lubman, D. I. (2007). Addiction neuroscience and its relevance to clinical practice. Drug and Alcohol Review, 26, 1–2. Martins, S. S., Storr, C. L., Alexandre, P. K., & Chilcoat, H. D. (2007). Do adolescent ecstasy users have different attitudes towards drugs compared to marijuana users? Drug and Alcohol Dependence, 94, 63–72. Morgan, M. J. (1998). Recreational use of ecstasy (MDMA) is associated with elevated impulsivity. Neuropsychopharmacology, 19, 252–264. Pardo, Y., Aguilar, R., Molinuevo, B., & Torrubia, R. (2007). Alcohol use as a behavioural sign of disinhibition: Evidence from J.A. Gray's model of personality. Addictive Behaviors, 32, 2398–2403. Perry, J. L., & Carroll, M. E. (2008). The role of impulsive behavior in drug abuse. Psychopharmacology, 200, 1–26. Schilt, T., Goudriaan, A. E., Koeter, M. W., Van der Brink, W., & Schmand, B. (2009). Decision making as a predictor of first ecstasy use: A prospective study. Psychopharmacology, 203, 519–527. Schramm-Sapyta, N. L., Walker, Q. D., Caster, J. M., Levin, E. D., & Kuhn, C. M. (2009). Are adolescents more vulnerable to drug addiction than adults? Evidence from animal models. Psychopharmacology, 206, 1–21. Simons, J. S., & Arens, A. M. (2007). Moderating effects of sensitivity to punishment and sensitivity to reward on associations between marijuana effect expectancies and use. Psychology of Addictive Behaviours, 21, 409–414. Torrubia, R., Avila, C., Molto, J., & Caseras, X. (2001). The sensitivity to punishment and sensitivity to reward (SPSRQ) as a measure of Gray's anxiety and impulsivity dimensions. Personality and Individual Differences, 31, 837–862. Van der Poel, A., Rodenburg, G., Dijkstra, M., Stoele, M., & Van de Mheen, D. (2009). Trends, motivations and setting of recreational cocaine use by adolescents and young adults in The Netherlands. The International Journal on Drug Policy, 20, 143–151. Young, R. M., & Oei, T. P. -S. (1996). Drinking Expectancy Profile (DEP) test manual. Queensland, Australia: Behaviour and Research Therapy Centre, The University of Queensland. Yu, R. L., & Ko, H. C. (2006). Cognitive determinants of MDMA use among college student in southern Taiwan. Addictive Behaviors, 31, 2199–2211.
Contents lists available at ScienceDirect
Addictive Behaviors
Short Communication
Reward sensitivity and outcome expectancies as predictors of ecstasy use in young adults Melanie J. Smerdon, Andrew J.P. Francis ⁎ Psychology, Health Sciences, RMIT University, Bundoora, 3083, Australia
a r t i c l e
i n f o
Keywords: Ecstasy Reward sensitivity Outcome expectancies Self-confidence
a b s t r a c t Extending recent studies showing that sensitivity to reward and outcome expectancies are associated with problematic alcohol and cannabis use, we undertook to determine if similar relationships would hold for ecstasy. One hundred and twenty five males and females aged between 18 and 35 years were recruited from RMIT University and through snowball sampling. Participants completed a questionnaire package measuring frequency, amount and first age of ecstasy use, sensitivity to reward and punishment as well as outcome expectancies relating to ecstasy use. Frequency of ecstasy use was significantly related to reward sensitivity (p b .05) and positive outcome expectancies (p b .01). Regression analysis revealed significant prediction of ecstasy use by study variables, with expectations of increased confidence making the largest individual contribution. Multiple intervention points are suggested by the results of this study, within a largely cognitive-based framework. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction The psychological and physiological mechanisms underlying hazardous use of illicit substances are yet to be fully elucidated, although in the past ten years the importance of individual personality traits and cognitions has become apparent with respect to predicting vulnerability (Dawe & Loxton, 2004; Lubman, 2007). In particular, it has been determined that individuals who frequently use alcohol and cannabis report higher sensitivity to reward and more positive outcome expectancies (DePino, 2009; Simons & Arens, 2007); and now interest is turning to other substances that are taken for their rewarding properties (Egan, Kambouropoulos, & Staiger, 2010; Franken & Muris, 2006; Van der Poel, Rodenburg, Dijkstra, Stoele, & Van de Mheen, 2009). This study examines the relevance of sensitivity to reward and outcome expectancies for the prediction of ‘ecstasy’ (3,4-Methylenedioxymethamphetamine) use. Gray's (1987) personality theory of approach-inhibition has been one of the most plausible psychological explanations for why some individuals take drugs and find them highly rewarding whilst others find them less so or not at all (Franken, Muris, & Georgieva, 2006; Perry & Carroll, 2008; Schramm-Sapyta, Walker, Caster, Levin, & Kuhn, 2009). Studies have consistently found that individuals high in behavioural approach are more likely to be vulnerable to the use and abuse of both alcohol (DePino, 2009; Franken & Muris, 2006;
⁎ Corresponding author at: Division of Psychology, School of Health Sciences, RMIT University, Bundoora, 3083, Australia. Tel.: + 61 3 9925 7782; fax: + 61 3 9925 7303. E-mail address: [email protected] (A.J.P. Francis). 0306-4603/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.addbeh.2011.07.027
Kambouropoulos & Staiger, 2007; Lubman, 2007; Pardo, Aguilar, Molinuevo, & Torrubia, 2007) and cannabis (DePino, 2009). Drawing on the theoretical basis of Bandura's (1986) social cognitive theory, another significant psychological predictor of substance use is outcome expectancies. Substance-related outcome expectancies may be positive or negative, and it is theorised that if a person perceives a substance as having more positive than negative potential outcomes then they are more likely to consume that substance (Martins, Storr, Alexandre, & Chilcoat, 2007). In support of these theoretical predictions, DePino (2009) confirmed previous research reporting positive outcome expectancies were significantly related to alcohol use. In addition, hazardous drinkers entertained more positive outcome expectancies than nonhazardous drinkers, and these relationships were also found for cannabis users. These data have led researchers to question whether similar relationships might also hold for another popular recreational drug, ecstasy (Clemens, McGregor, Hunt, & Cornish, 2007; Egan et al., 2010; Hatala, 2008). Ecstasy is the third most commonly used illicit drug by Australians aged between 14 and 29 years, with similar levels of use being observed in other countries (Costa, 2003; Degenhardt, Barker, & Topp, 2003; Kaye, Darke, & Duflou, 2009). To date, elevated impulsivity and positive outcome expectancies have been associated with ecstasy use (Egan et al., 2010; Hanson, Luciana, & Sullwold, 2008; Martins et al., 2007; Morgan, 1998; Schilt, Goudriaan, Koeter, Van der Brink, & Schmand, 2009; Yu & Ko, 2006), although further research is required to confirm these relationships. By extension of the theoretical and empirical rationale established for alcohol and cannabis, the aim of this study was to examine individual differences in reward sensitivity and outcome expectancies
1338
M.J. Smerdon, A.J.P. Francis / Addictive Behaviors 36 (2011) 1337–1340
as predictors of ecstasy use in young adults. In accord with previous research, it was hypothesised that there would be significant positive relationships observed between frequency of ecstasy use and sensitivity to reward. It was further hypothesised that positive outcome expectancies would also be associated with higher frequency of ecstasy use, as is the case for alcohol and cannabis from previous research. Finally, it was expected that sensitivity to reward and positive outcome expectancies would, overall, significantly predict levels of ecstasy use within a multiple regression model. 2. Method 2.1. Participants The sample consisted of 125 participants (Mage = 25.35 years, SDage = 3.62 years), ranging in age from 18 years to 35 years of which 60 were male (Mage = 26.10 years, SD = 4.11 years) and 62 were female (Mage = 24.57 years, SD = 2.85 years). Three participants (2.4%) did not provide information on their gender and six participants (4.8%) did not provide information about their age. Most participants (70.4%) reported they were employed full time, the remaining were either employed part time (7.8%), students (9.3%) or causally employed (9.3%). Persons aged less than 18 years and over 35 years were excluded. 2.2. Materials Participants completed a questionnaire package which examined reward and punishment sensitivity and ecstasy outcome expectancies. The questionnaire also required participants to answer questions pertaining to frequency of ecstasy use, typical amount consumed per occasion and age first used. Frequency of ecstasy use was indicated by a response in one of the following categories: every day, approximately once per week, approximately once per month, every few months, once or twice a year, or not currently using. Categories were coded as increasing numeric values, with higher scores on the frequency variable therefore indicating less frequent ecstasy use. 2.2.1. The Sensitivity to Punishment and Sensitivity to Reward Questionnaire (SPSRQ) The SPSRQ measures levels of sensitivity to punishment (SP) and sensitivity to reward (SR) (Torrubia, Avila, Molto, & Caseras, 2001). Cronbach's alpha coefficients range between .82 and .83 for SP and .75 and .78 for SR (Torrubia et al., 2001). 2.2.2. The Ecstasy Expectancies Questionnaire (EEQ) Using a similar procedure to that utilised by DePino (2009) for measurement of cannabis expectancies, an ecstasy expectancies questionnaire (EEQ) was developed by adapting the Drinking Expectancies Questionnaire (DEQ; Young & Oei, 1996). The 5 subscales of the EEQ are Negative Consequences of Ecstasy Use (EEQ-NC), Increased Confidence (EEQ-IC), Increased Sexual Interest (EEQ-ISI), Cognitive Enhancement (EEQ-CE) and Tension Reduction (EEQ-TR). 2.3. Procedure Following approval from the RMIT University Human Research Ethics Committee, participants were recruited from RMIT University through paper advertisements and through snowball sampling, which is a commonly used and useful approach with research on drug use (Davidson & Parrott, 1997; Hinchliff, 2001). Consent was implied through the return of the survey.
3. Results 3.1. Data screening 3.1.1. Missing values analysis Discrete missing values were replaced by the mean of the available values, and subscales recalculated. Twenty two participants were found to have not completed at least one of the questionnaires making up the survey. These participants were not included in analyses involving those particular questionnaires, but were not excluded from the sample overall. 3.1.2. Normality and univariate outliers After discrete missing values were replaced, the scale and subscale distributions were visually examined for departures from normality and outliers. Examinations of the histograms and box plots showed there were no variables that deviated substantively from normality. 3.2. Reliabilities of scales and descriptive statistics Internal consistencies for sensitivity to reward (SR) and punishment (SP) were .81 and .84 respectively, similar to the alpha coefficients stated by Torrubia et al. (2001) who suggested a range of .82 and .83 for SP and .75 and .78 for SR. Scale reliabilities for ecstasy outcome expectancies subscales ranged from .55 to .94 which, except for EEQ-TR (.55), are consistent with those found by Young and Oei (1996). Considering EEQ-TR was found to have a lower reliability measure, EEQ-TR was interpreted with caution. Of the 125 participants, 57% indicated they had used ecstasy. Of this user group, 3% used ecstasy monthly, 21% every few months and 21% once or twice per year. A single individual reported using ecstasy weekly and none reported using ecstasy daily. Of the user group, 42% indicated they were ‘not currently using’. 3.3. Correlation analysis Pearson correlations were conducted to examine the associations between frequency of ecstasy use, sensitivity to reward or punishment and outcome expectancies (Table 1). Frequency of ecstasy use showed a small, significant negative correlation with sensitivity to reward, indicating that more frequent users were more sensitive to reward. There were significant medium-magnitude, negative correlations between frequency of ecstasy use and each of EEQ-IC, EEQ-TR and EEQ-TPE, as well as a small, negative correlation with EEQ-TE; indicating that more positive evaluations on these outcome expectancy domains were associated with more frequent ecstasy use. A small, significant and positive correlation was also observed between EEQ-ISI and frequency of ecstasy use. There were no significant correlations between the amount of ecstasy consumed on each occasion of use and either sensitivity to reward or punishment, or any of the outcome expectancy domains. The only significant correlation with reported age of first use was the increased confidence domain (EEQ-IC) of the outcome expectancies questionnaire, r = 0.25, n = 55, p = .03, indicating that expectation of increased confidence was associated with earlier ecstasy use. 3.4. Regression analysis To determine the extent to which sensitivity to reward and ecstasy outcome expectancies (EEQ-IC, EEQ-TR & EEQ-TPE) predicted frequency of ecstasy use, a standard multiple regression analysis was conducted. Outcome expectancies concerned with negative consequences were not included in the regression since the correlational analysis revealed that EEQ-NC had no significant relationship with frequency of ecstasy use. Furthermore, the inclusion of the EEQ subscales ISI and TE would be likely to cause multicollinearity, which is a violation of assumption
M.J. Smerdon, A.J.P. Francis / Addictive Behaviors 36 (2011) 1337–1340
1339
Table 1 Pearson's correlations between personality variables, outcome expectancies and frequency of ecstasy use. 1 Freq. E use SR SP EEQ-NC EEQ-IC EEQ-ISI EEQ-CE EEQ-TR EEQ-TPE EEQ-TE
1 −.18⁎ .13 .15 −.42⁎⁎ .18⁎ .01 −.31⁎ −.38⁎ −.18⁎
2
3
4
5
6
7
8
9
1 .50⁎⁎ .67⁎⁎ .7⁎⁎
1 .61⁎⁎ .57⁎⁎
1
10
1 .11 −.01 .19* .04 −.16 −.02 .12 .08
1 .09 .02 .02 −.08 −.18⁎ −.03 .03
1 .42⁎⁎ .48⁎⁎ .53⁎⁎ .33⁎⁎ .51⁎⁎ .82⁎⁎
1 .13 .49⁎⁎ .47⁎⁎ .97⁎⁎ .84⁎⁎
1
.32⁎⁎ .13 .25⁎⁎ .4⁎⁎
.91⁎⁎
1
Note: Freq. E use = Frequency of ecstasy use; SR = Sensitivity to Reward; SP = Sensitivity to Punishment; EEQ-NC = Ecstasy Outcome Expectancies Questionnaire Negative Consequences subscale; EEQ-IC = Ecstasy Outcome Expectancies Questionnaire Increased Confidence Subscale; EEQ-ISI = Ecstasy Outcome Expectancies Questionnaire Increased Sexual Interest Subscale; EEQ-CE = Ecstasy Outcome Expectancies Questionnaire Cognitive Enhancement Subscale; EEQ-TR = Ecstasy Outcome Expectancies Questionnaire Tension Reduction Subscale; EEQ-TPE = Ecstasy Outcome Expectancies Questionnaire Global Positive Expectancy Subscale; and EEQ-TE = Ecstasy Outcome Expectancies Questionnaire Summative Expectancy Score Subscale. ⁎ p b .05 one tailed. ⁎⁎ p b .01 one tailed.
testing for regression analysis (Coakes & Steed, 2007), and were therefore also excluded. Results of the multiple regression are displayed in Table 2. The model significantly predicted frequency of ecstasy use, F(4,96) = 10.31, p b .01, accounting for 27.1% of the variance in ecstasy use. Further examination revealed that EEQ-TR, EEQ-IC and EEQ-TPE were the only significant predictors of ecstasy use, but EEQ-IC had the largest unique contribution to the prediction ecstasy use behaviour. 4. Discussion Results supported the hypothesis that there would be a relationship between frequency of ecstasy use and sensitivity to reward. Results also supported the hypothesis that domains of positive outcome expectancy would be associated with higher frequency of ecstasy use, as is the case for alcohol and cannabis. Finally, the results supported the hypothesis that study variables would, overall, significantly predict levels of ecstasy use within a regression model analysis. The hypothesis that sensitivity to reward would be associated with ecstasy use was supported, although the relationship was relatively weak. The association found between reward sensitivity and ecstasy use is consistent with recent findings by Egan et al. (2010) which showed that more frequent ecstasy use was associated with higher self-reported sensitivity to reward. Support is also found for Gray's (1987) neurological model of substance use, which suggests being sensitive to reward is a risk factor for potential drug use (Acton, 2003; Franken, 2002). Despite the significant association between the variables, regression analysis suggested that sensitivity to reward did not significantly and uniquely predict ecstasy use behaviour in the current sample. There are several potential explanations for this finding. First, the correlation analysis revealed only a small magnitude negative correlation between
Table 2 Standard and unstandardised beta coefficients, standard error, squared semi-partial correlations and significance associated with prediction of frequency of ecstasy use.
ecstasy use and reward sensitivity, and so it might be expected that the regression analysis (as a more conservative statistic) could reveal a nonsignificant result. Second, the current study illustrates the important extent to which individual differences in cognitions contribute to substance use. Results revealed that positive outcome expectancies were significantly associated with frequency of ecstasy use and were a significant predictor of ecstasy use behaviour. Regression analysis revealed having outcome expectancies of increased confidence, tension reduction and overall positive expectancies predicted how frequently a person would use ecstasy, consistent with findings by Martins et al. (2007) and Egan et al. (2010). The results are crucial as they provide information as to why people continue to take ecstasy despite all risks: expected positive outcomes superseding any expected negative outcomes. Thus within the context of the regression model, reward sensitivity was not significant as a unique predictor relative to the more substantial variation predicted by the cognitive variables. 5. Conclusions Findings of this study indicate reward sensitivity and outcome expectancies are associated with ecstasy use, consistent with the theoretical framework of previous research on alcohol and cannabis. Of particular interest is the finding of an expectation that ecstasy use would increase self-confidence. This expectation was associated with greater frequency of ecstasy use (making the largest unique contribution to the prediction ecstasy use behaviour in regression modelling) and with earlier use of the drug. Self-confidence, as an individual factor and also as a drug-associated expectation, is therefore indicated as an important and appropriate focus for both further research and potential intervention. Modification of outcome expectancies towards drugs, including ecstasy, within a broader cognitive-behavioural framework could hold the key to changing an individual's drug usage pattern. Role of funding sources Funding for this study was provided by the Division of Psychology, RMIT University.
2
Predictor
B
SE B
β
Sr
SR EEQ-IC EEQ-TR EEQ-TPE
−.03 −.24 −.36 .18
.03 .06 .09 .05
−.08 − 1.7⁎⁎ −.57⁎⁎ 1.6⁎
.007 .12 .11 .09
Note: N = 101; B = unstandardised beta coefficients; SE B = standard error of unstandardised beta coefficients; β = standardised beta coefficients; and Sr2 = squared semi partial correlations. ⁎ p b .01. ⁎⁎ p b .001.
Contributors Andrew Francis designed the study. Melanie Smerdon performed literature searches, collected data and formulated the first draft of the manuscript. All authors contributed to and have approved the final manuscript. Andrew Francis and Melanie Smerdon analysed the data.
Conflict of interest There are no conflicts of interest by any author.
1340
M.J. Smerdon, A.J.P. Francis / Addictive Behaviors 36 (2011) 1337–1340
References Acton, G. S. (2003). Measurement of impulsivity in a hierarchical model of personality traits: Implications of substance use. Substance Use & Misuse, 38, 67–83. Bandura, A. (1986). Social foundations of thought and action; a social cognitive theory. Englewood Cliffs: Prentice Hall. Clemens, K. J., McGregor, I. S., Hunt, G. E., & Cornish, J. L. (2007). MDMA, methamphetamine and their combination: Possible lessons for party drug users from recent preclinical research. Drug and Alcohol Review, 26, 9–15. Coakes, S., & Steed, L. (2007). Multiple regression. In SPSS analysis without anguish: Version 14.0 for windows (pp. 135–146). Milton: John Wiley & Sons Australia, Ltd. Costa, A. M. (2003). Ecstasy and amphetamines global survey 2003. Presented to the United Nations office of crime and drugs in ViennaUnited Nations Publication Sales No. E.03.XI.1592-1-148164-3. Davidson, D., & Parrott, A. C. (1997). Ecstasy (MDMA) in recreational users: Self reported psychological and physiological effects. Human Psychopharmacology, 12, 221–226. Dawe, S., & Loxton, N. J. (2004). The role of impulsivity in the development of substance use and eating disorders. Neuroscience Biobehavioral Review, 28, 343–351. Degenhardt, L., Barker, B., & Topp, L. (2003). Patterns of ecstasy use in Australia: Findings from a national household survey. University of NSW, Australia: National Drug and Alcohol Research Centre. doi:10.1046/j.1360-0443.2003.00622.x. DePino, E. (2009). Reward sensitivity and outcome expectancies predict alcohol and cannabis use in young adults. Unpublished doctoral thesis, RMIT University, Victoria. Egan, S. T., Kambouropoulos, N., & Staiger, P. K. (2010). Rash-impulsivity, reward-drive and motivations to use ecstasy. Personality and Individual Differences, 48, 670–675. Franken, I. H. A. (2002). Behavioral approach system (BAS) sensitivity predicts alcohol craving. Personality and Individual Differences, 32, 349–355. Franken, I. H. A., & Muris, P. (2006). BIS/BAS personality characteristics and college students' substance use. Personality and Individual Differences, 40, 1497–1503. Franken, I. H. A., Muris, P., & Georgieva, I. (2006). Gray's model of personality and addiction. Addictive Behaviors, 31, 399–403. Gray, J. A. (1987). Perspectives on anxiety and impulsivity: A commentary. Journal of Research in Personality, 21, 493–509. Hanson, K. L., Luciana, M., & Sullwold, K. (2008). Reward-related decision making deficits and elevated impulsivity among MDMA and other drug users. Drug and Alcohol Dependence, 96, 99–110. Hatala, E. M. (2008). Characteristics and predictors of ecstasy (MDMA) use during college. Unpublished Doctoral Thesis, Temple University, Missouri. Hinchliff, S. (2001). The meaning of ecstasy use and clubbing to women in the last 1990's. The International Journal on Drug Policy, 12, 455–468.
Kambouropoulos, N., & Staiger, P. K. (2007). Personality, behavioural and affective characteristics of hazardous drinkers. Personality and Individual Differences, 42, 213–224. Kaye, S., Darke, S., & Duflou, J. (2009). Methylenedioxymethamphetamine (MDMA)related fatalities in Australia: Demographics, circumstances, toxicology and major organ pathology. Drug and Alcohol Dependence, 104, 254–261. Lubman, D. I. (2007). Addiction neuroscience and its relevance to clinical practice. Drug and Alcohol Review, 26, 1–2. Martins, S. S., Storr, C. L., Alexandre, P. K., & Chilcoat, H. D. (2007). Do adolescent ecstasy users have different attitudes towards drugs compared to marijuana users? Drug and Alcohol Dependence, 94, 63–72. Morgan, M. J. (1998). Recreational use of ecstasy (MDMA) is associated with elevated impulsivity. Neuropsychopharmacology, 19, 252–264. Pardo, Y., Aguilar, R., Molinuevo, B., & Torrubia, R. (2007). Alcohol use as a behavioural sign of disinhibition: Evidence from J.A. Gray's model of personality. Addictive Behaviors, 32, 2398–2403. Perry, J. L., & Carroll, M. E. (2008). The role of impulsive behavior in drug abuse. Psychopharmacology, 200, 1–26. Schilt, T., Goudriaan, A. E., Koeter, M. W., Van der Brink, W., & Schmand, B. (2009). Decision making as a predictor of first ecstasy use: A prospective study. Psychopharmacology, 203, 519–527. Schramm-Sapyta, N. L., Walker, Q. D., Caster, J. M., Levin, E. D., & Kuhn, C. M. (2009). Are adolescents more vulnerable to drug addiction than adults? Evidence from animal models. Psychopharmacology, 206, 1–21. Simons, J. S., & Arens, A. M. (2007). Moderating effects of sensitivity to punishment and sensitivity to reward on associations between marijuana effect expectancies and use. Psychology of Addictive Behaviours, 21, 409–414. Torrubia, R., Avila, C., Molto, J., & Caseras, X. (2001). The sensitivity to punishment and sensitivity to reward (SPSRQ) as a measure of Gray's anxiety and impulsivity dimensions. Personality and Individual Differences, 31, 837–862. Van der Poel, A., Rodenburg, G., Dijkstra, M., Stoele, M., & Van de Mheen, D. (2009). Trends, motivations and setting of recreational cocaine use by adolescents and young adults in The Netherlands. The International Journal on Drug Policy, 20, 143–151. Young, R. M., & Oei, T. P. -S. (1996). Drinking Expectancy Profile (DEP) test manual. Queensland, Australia: Behaviour and Research Therapy Centre, The University of Queensland. Yu, R. L., & Ko, H. C. (2006). Cognitive determinants of MDMA use among college student in southern Taiwan. Addictive Behaviors, 31, 2199–2211.