Journal of Clinical and Experimental Neuropsychology, 2014 Vol. 36, No. 8, 794–805, http://dx.doi.org/10.1080/13803395.2014.943694
Neurocognitive moderation of associations between cannabis use and psychoneuroticism Stephan C. J. Huijbregts1,2, Merel F. H. Griffith-Lendering1, Wilma A. M. Vollebergh3, and Hanna Swaab1,2 1
Department of Clinical Child and Adolescent Studies–Neurodevelopmental Disorders, Faculty of Social Sciences, Leiden University, Leiden, The Netherlands 2 Leiden Institute for Brain and Cognition, Leiden University Medical Center, Leiden, The Netherlands 3 Department of Interdisciplinary Social Science, University of Utrecht, Utrecht, The Netherlands (Received 31 January 2013; accepted 7 July 2014) Background: Cannabis use has been associated with neurocognitive impairments and psychopathology. The strength of such associations does however appear to depend on many different factors, such as duration and intensity of use, but also differential susceptibility due to genetic make-up and environmental influences. The present study investigated whether specific cognitive weaknesses moderated associations between cannabis use and psychoneuroticism, which may be considered one of the “softer” expressions of an extended psychosis phenotype. Method: One hundred and fifty (150) young adults (mean age: 24.7 years, SD: 3.7), mostly college students, performed three computerized neuropsychological tasks: a relatively easy social perception task (Face Recognition), a more complex social perception task combining emotion recognition and executive functioning (Matching Facial Emotions), and a more complex task requiring sustained attention and executive functioning (Sustained Attention–Dots). Participants self-reported on the extent to which they experienced psychological problems using the Symptom Checklist-90 (SCL-90). The SCL-90 total score (psychoneuroticism) was used as dependent variable in analyses of variance. Results: Frequent and current cannabis users performed more poorly than nonusers on the three tasks. They also reported more psychoneuroticism than nonusers whether they were classified according to their lifetime use, their use during the past 12 months, or use during the past 4 weeks. Moderate and former users did not differ from nonusers. Relatively poor performance on the Matching Facial Emotions task, as opposed to performance on the Face Recognition and Sustained Attention tasks, augmented levels of experienced psychoneuroticism among frequent and current cannabis users. Conclusions: Relatively poor cognitive abilities appear to represent increased vulnerability to effects of frequent and current cannabis use on affective mental health, even among highly educated individuals. There seems to be some specificity as to which (combinations of) neurocognitive abilities may be most indicative, as moderating effects were only observed when participants had relatively poor complex social perception ability. Keywords: Cannabis; Emotion recognition; Social perception; Psychoneuroticism; Cognitive ability.
Cannabis is the most widely used substance after tobacco and alcohol in Western countries, with a particularly high prevalence among adolescents and young adults (European Monitoring Centre for Drugs and Drug Addiction, 2009; Substance Abuse and Mental Health Services Administration, 2008). There is increasing evidence for beneficial
effects of medicinal cannabis use, particularly for the management of neuropathic pain and multiple sclerosis spasticity (Grant, Hampton Atkinson, Gouaux, & Wilsey, 2012; Ware, Adams, & Guy, 2005). And even though the debate on this topic is still ongoing (e.g., Kleber & DuPont, 2012), legalization of cannabis use is increasingly advocated and
Address correspondence to: Stephan C. J. Huijbregts, Department of Clinical Child and Adolescent Studies–Neurodevelopmental Disorders, Faculty of Social Sciences, Leiden University, P.O. Box 9555, 2300 RB Leiden, The Netherlands (E-mail: [email protected]).
© 2014 Taylor & Francis
CANNABIS, SOCIAL PERCEPTION, AND PSYCHONEUROTICISM
effectuated (e.g., Clark, Capuzzi, & Fick, 2011). Despite this changing view of cannabis, its use has been associated with psychosis (Arseneault et al., 2002; Degenhardt, Hall, & Lynskey, 2003; Griffith-Lendering et al., 2013; Moore et al., 2007), antisocial behavior (Fergusson, Horwood, & Ridder, 2007; Griffith-Lendering, Huijbregts, Mooijaart, Vollebergh, & Swaab, 2011; Monshouwer et al., 2006; Rey, Sawyer, Raphael, Patton, & Lynskey, 2002), and depression (Hayatbakhsh et al., 2007; Rey et al., 2002). Cannabis use has frequently been related to cognitive difficulties as well, for example with respect to executive functioning, implicit cognition, episodic memory, information processing speed, and emotional processing (Becker, Collins, & Luciana, 2014; Kelleher, Stough, Sergejew, & Rolfe, 2004; Pope, Gruber, Hudson, Huestis, & YurgelunTodd, 2001; Solowij & Battisti, 2008; Stacy & Wiers, 2010; Verdejo-Garcia, Lawrence, & Clark, 2008). It should be noted, however, that the aforementioned associations have not always been straightforward. With respect to psychopathology and psychological problems, associations often appear to be limited to specific aspects of cannabis use, such as an earlier starting age or a longer duration of use. Quantity or frequency of use also appears to play a key role, as well as preexisting vulnerability to psychological problems or psychopathology (Lynskey, Vink, & Boomsma, 2006; Schubart et al., 2011; Swift et al., 2012). With respect to cognitive outcomes, similar factors have been investigated (Nordstrom & Hart, 2006; Solowij & Battisti, 2008), but most attention has been given to acute versus nonacute effects of cannabis (e.g., Bossong, Jager, Bhattacharyya, & Allen, 2014; Crean, Crane, & Mason, 2011) and to defining the cognitive profile of cannabis users and whether a profile could be established that is specific for cannabis use (as opposed to a more generic profile associated with use of other substances as well; e.g., Griffith-Lendering, Huijbregts, Vollebergh, & Swaab, 2012; Morgan et al., 2012; Stacy & Wiers, 2010; Verdejo-Garcia et al., 2008). The goal of the present study was to investigate the role of specific aspects of cognition in associations between cannabis use and psychological problems. With respect to type of psychological problems, we focused on psychoneuroticism, which shows considerable genetic and phenotypic overlap with psychosis/schizophrenia (Macare, Bates, Heath, Martin, & Ettinger, 2012; Van Nierop et al., 2012), the disorders that have most consistently been associated with cannabis exposure (e.g., Moore et al., 2007). Psychoneuroticism
795
has been related to cannabis use before (Fridberg, Vollmer, O’Donnell, & Skosnik, 2011) and appears to represent one of the “softer” expressions of an extended psychosis phenotype (van Nierop et al., 2012), which is important as our sample consisted of individuals without diagnosed psychopathology. Moreover, even without diagnosed disorders, acute effects of cannabis appear to mimic symptoms associated with schizophrenia/psychosis (D’Souza et al., 2004; Wadsworth, Moss, Simpson, & Smith, 2006). With respect to cognitive functioning, we focused on social perception. The rationale behind this choice is that cannabis particularly appears to exert its influence on the functioning of the limbic system (including the amygdala), which plays an important role in social perception (Adolphs, 2002; Ochsner, 2008). Social perception problems, in turn, also constitute a central deficit in disorders that have been associated with cannabis use, including psychosis/schizophrenia (Demenescu, Kortekaas, Den Boer, & Aleman, 2010; Germine & Hooker, 2011; Marsh & Blair, 2008; Rössler et al., 2011). Limbic brain regions, including the amygdala, have high cannabinoid receptor (CB1) densities (Herkenham et al., 1990; Glass, Dragunow, & Faull, 1997). Animal studies indicate that cannabinoid signaling (as induced by cannabis exposure) induces long-term depression of gammaaminobutyric acid (GABA)-ergic interneurons within the amygdala’s basolateral nucleus, thereby upregulating the amygdala’s activity and sensitivity (Tan, Lauzon, Bishop, Bechard, & Laviolette, 2010). Phan and colleagues (2008) reported reduced amygdala reactivity during social information processing after administration of delta-9tetra-hydrocannabinol (Δ9-THC), the principle constituent of cannabis. Also, Gruber, Rogowska, and Yurgelun-Todd (2009) showed reduced amygdala activity during emotion perception in chronic cannabis users. Among long-term, chronic cannabis users abnormalities in amygdala volumes have also been observed (McQueeny et al., 2011; Yucel et al., 2008). Whereas studies yielded relatively consistent results with respect to abnormalities in amygdala structure and function, thereby often using stimuli that required the ability to recognize emotions from facial expressions, they did not focus on the quality of social perception. Only one recent study, by Platt, Kamboj, Morgan, and Curran (2010), did focus on performance during an emotion recognition task. Cannabis users were significantly slower than controls at identifying emotional expressions in a paradigm where facial expressions gradually changed from neutral to more intense expressions
796
HUIJBREGTS ET AL.
of sadness, anger, or happiness. Although the authors discussed the possible implications of their findings for vulnerability to psychological problems in cannabis users, they did not investigate this further. We sought to extend the research by Platt and colleagues by examining social perception in relation to the previously described psychoneuroticism among cannabis users. Based on the available evidence, it was to be expected that cannabis users would have more social perception difficulties and would report more affective mental health problems than nonusers. However, as most associations with impaired cognition or psychological problems have been established after acute use, in individuals with chronic/heavy use, or among those with indicators of vulnerability to mental health problems (Caspi et al., 2005; Henquet, Di Forti, Morrison, Kuepper, & Murray, 2008), effects were particularly expected among those who classified themselves as recreational users but used cannabis relatively frequently. It was further hypothesized that quality of social perception would influence the extent to which psychological problems were experienced, again particularly among relatively heavy users. In order to investigate specificity of social perception as a mediating or moderating cognitive factor in these associations, three tasks were used: a relatively easy face recognition task, a task requiring emotion recognition and working memory, and a task requiring sustained attention and inhibitory control.
METHOD Participants Participants were 150 young adults (74 men, 76 women; mean age: 24.7 years, SD = 3.7) recruited among University of Leiden undergraduate students and through advertisements on internet forums concerning cannabis topics. Advertisements were aimed at recruitment of recreational cannabis users and nonusing controls. Polydrug users and users of drugs other than cannabis were not excluded. Exclusion criteria (established by means of selfreport) were history of neurological illness or serious head injury, past or present major medical illness, and diagnosed psychopathology (without further specification of whether this was substance induced). Five potential participants (3 cannabis users, 2 nonusers) reported a diagnosed psychiatric disorder (3 depression, 1 bipolar disorder, 1 schizophrenia). These were excluded from statistical analyses. No participants reported a neurological or other major medical illness. Written informed
consent was obtained from all participants before the start of the study. Ethical approval for this study was granted by Leiden University’s Education and Child Studies Ethics Committee. Measures Cannabis use and use of other substances Cannabis use was assessed by asking participants about their lifetime use, their use during the past 12 months, and use during the past 4 weeks. For each time period, there were 14 answer categories in order to establish the frequency of use. The answer categories were as follows: 0 = I have never/not used cannabis, 1 = one time, 2 = two times, 3 = three times, … 10 = 10 times, 11 = 11– 19 times, 12 = 20 times or more, and 13 = 40 times or more. Furthermore, participants were asked about the age they first used cannabis and about the time of last use (answer categories: 1 = more than 12 months ago; 2 = 4–12 months ago; 3 = 1–3 months ago; 4 = 1–4 weeks ago; 5 = 3–7 days ago; 6 = 2 days ago; 7 = yesterday). The same questions with the same answer categories were asked for Ecstasy (3,4-methylenedioxymethamphetamine, MDMA), cocaine, and a maximum of three other illegal drugs (to be filled out by the respondents). For alcohol, answer categories were the same as those for cannabis and other drugs when use during the last 4 weeks was concerned, but for lifetime use and use during the last 12 months, the answer categories referred to the average frequency of use per month. For tobacco use, the answer categories referred to the average frequency of use per day. Aggregate scores are often used for the upper ranges of substance use reports, particularly those involving extended or nonrecent periods of time (e.g., lifetime use, use during the last 12 months), as inaccuracy of estimates strongly increases with increasing amounts of use and the passing of time. The measures used in the present study were similar to those used in many other studies (e.g., Creemers et al., 2010; Huizink, Ferdinand, Ormel, & Verhulst, 2006; Jaddoe et al., 2006; Monshouwer et al., 2006; O’Malley, Bachman, & Johnston, 1983; Otten, Barker, Maughan, Arseneault, & Engels, 2010; Prince van Leeuwen et al., 2011; Tait, MacKinnon, & Christensen, 2011). Psychoneuroticism The Symptom Checklist-90 (SCL-90; Derogatis, Lipman, & Covi, 1973; Elliott et al., 2006), a 90item self-report symptom inventory developed to measure psychological symptoms and distress, was
CANNABIS, SOCIAL PERCEPTION, AND PSYCHONEUROTICISM
used to measure psychoneuroticism. It was designed to be appropriate not only for use in clinical populations but also for use within community samples. The SCL-90, for which items are rated on 5-point scales reflecting the extent to which problems were experienced in the past 7 days, generates the following scales: somatic complaints (12 items), insufficiency of thoughts and actions (9 items), distrust (18 items), depression (16 items), anxiety (10 items), hostility (6 items), agoraphobia (6 items), and sleeping problems (3 items). In addition, a global score is obtained, called psychoneuroticism, using the overall score of the 90 items. Internal reliability of the different scales ranges from .77–.97 (Arrindell & Ettema, 1986; Olsen, Mortensen, & Bech, 2004). Neuropsychological tasks Three tasks from the Amsterdam Neuropsychological Tasks (ANT; De Sonneville, 1999), a battery of computerized tests, were used to assess neurocognitive abilities. Test–retest reliability, construct-, criterion-, and discriminant validity of the ANT tasks are satisfactory and have extensively been described elsewhere (e.g., De Sonneville et al., 2002; Huijbregts, Jahja, De Sonneville, De Breij, & Swaab-Barneveld, 2010). Before each part of a task the participants were given a standard verbal instruction and were given the opportunity to ask questions and to practice. Face Recognition (FR). This task (duration: 5 min) examined the ability to recognize neutral faces. A target face was presented on the monitor for 2.5 s. Following the presentation of the target face, a set of four photographs of individuals was
797
presented, and participants had to indicate whether or not the target individual appeared in the set of four (Figure 1). The gender and age category of the target (i.e., boys, girls, men, or women) match those of the subsequently shown set of four faces. A yes-response was given by pressing the mouse button below the index finger of the preferred hand; a no-response required a press of the mouse button below the index finger of the nonpreferred hand. There were 40 trials, in half of which the display set contained the target face. Matching of Facial Emotions (MFE). This task (duration: 10 min) measured the ability to match emotions using facial expressions. The expressed emotions are happiness, sadness, anger, and fear. In each of the 160 trials, two (digitized photographs of) faces expressing a particular emotion were presented simultaneously on the computer screen. The participants had to press the yes-button when the two faces expressed the same emotion and the nobutton when the facial emotions did not match (Figure 2). MFE may be considered a more demanding task than FR, as it requires executive functioning (working memory) in addition to social perception. The tasks can also be distinguished based on the fact that MFE specifically involves emotion recognition, whereas FR does not. Sustained Attention–Dots (SA-Dots). In this task (duration: approximately 12 min), 600 patterns of 3, 4, or 5 dots were presented in random configurations on the computer screen. The 3, 4, or 5 dots were each presented 200 times, but in random order. Participants were required to press the mouse button under the index finger of the nondominant hand when 3 or 5 dots appeared and to
Figure 1. Stimulus example (target face + display set) and timing of the trials for the Face Recognition task. © Sonares B.V. Reproduced by permission of Sonares B.V.
798
HUIJBREGTS ET AL.
Figure 2. Stimulus examples for the Matching Facial Emotions task. © Sonares B.V. Reproduced by permission of Sonares B.V.
press the mouse button under the index finger of the dominant hand when 4 dots appeared. A response had to be generated between 250 and 6000 ms after a signal. The fixed postresponse interval was 250 ms. Auditory feedback was provided after an error. This task requires sustained attention and executive functioning (inhibitory control; Huijbregts et al., 2002). Statistical analysis First, Spearman correlations were calculated to investigate associations among cannabis use measures (lifetime frequency of cannabis use, frequency of use during the last 12 months, frequency of use during the last 4 weeks, starting age, and most recent use), between cannabis use measures and use of other substances, and between cannabis and other substance use measures and neurocognitive and SCL-90 outcomes. Group differences regarding the aggregate SCL90 score (psychoneuroticism) and performance on the neurocognitive tasks were investigated using general linear model (GLM) analyses of variance. Separate analyses were performed for lifetime use, use during the last 12 months, and use during the last 4 weeks. Individuals were classified as nonusers, moderate users (
Neurocognitive moderation of associations between cannabis use and psychoneuroticism Stephan C. J. Huijbregts1,2, Merel F. H. Griffith-Lendering1, Wilma A. M. Vollebergh3, and Hanna Swaab1,2 1
Department of Clinical Child and Adolescent Studies–Neurodevelopmental Disorders, Faculty of Social Sciences, Leiden University, Leiden, The Netherlands 2 Leiden Institute for Brain and Cognition, Leiden University Medical Center, Leiden, The Netherlands 3 Department of Interdisciplinary Social Science, University of Utrecht, Utrecht, The Netherlands (Received 31 January 2013; accepted 7 July 2014) Background: Cannabis use has been associated with neurocognitive impairments and psychopathology. The strength of such associations does however appear to depend on many different factors, such as duration and intensity of use, but also differential susceptibility due to genetic make-up and environmental influences. The present study investigated whether specific cognitive weaknesses moderated associations between cannabis use and psychoneuroticism, which may be considered one of the “softer” expressions of an extended psychosis phenotype. Method: One hundred and fifty (150) young adults (mean age: 24.7 years, SD: 3.7), mostly college students, performed three computerized neuropsychological tasks: a relatively easy social perception task (Face Recognition), a more complex social perception task combining emotion recognition and executive functioning (Matching Facial Emotions), and a more complex task requiring sustained attention and executive functioning (Sustained Attention–Dots). Participants self-reported on the extent to which they experienced psychological problems using the Symptom Checklist-90 (SCL-90). The SCL-90 total score (psychoneuroticism) was used as dependent variable in analyses of variance. Results: Frequent and current cannabis users performed more poorly than nonusers on the three tasks. They also reported more psychoneuroticism than nonusers whether they were classified according to their lifetime use, their use during the past 12 months, or use during the past 4 weeks. Moderate and former users did not differ from nonusers. Relatively poor performance on the Matching Facial Emotions task, as opposed to performance on the Face Recognition and Sustained Attention tasks, augmented levels of experienced psychoneuroticism among frequent and current cannabis users. Conclusions: Relatively poor cognitive abilities appear to represent increased vulnerability to effects of frequent and current cannabis use on affective mental health, even among highly educated individuals. There seems to be some specificity as to which (combinations of) neurocognitive abilities may be most indicative, as moderating effects were only observed when participants had relatively poor complex social perception ability. Keywords: Cannabis; Emotion recognition; Social perception; Psychoneuroticism; Cognitive ability.
Cannabis is the most widely used substance after tobacco and alcohol in Western countries, with a particularly high prevalence among adolescents and young adults (European Monitoring Centre for Drugs and Drug Addiction, 2009; Substance Abuse and Mental Health Services Administration, 2008). There is increasing evidence for beneficial
effects of medicinal cannabis use, particularly for the management of neuropathic pain and multiple sclerosis spasticity (Grant, Hampton Atkinson, Gouaux, & Wilsey, 2012; Ware, Adams, & Guy, 2005). And even though the debate on this topic is still ongoing (e.g., Kleber & DuPont, 2012), legalization of cannabis use is increasingly advocated and
Address correspondence to: Stephan C. J. Huijbregts, Department of Clinical Child and Adolescent Studies–Neurodevelopmental Disorders, Faculty of Social Sciences, Leiden University, P.O. Box 9555, 2300 RB Leiden, The Netherlands (E-mail: [email protected]).
© 2014 Taylor & Francis
CANNABIS, SOCIAL PERCEPTION, AND PSYCHONEUROTICISM
effectuated (e.g., Clark, Capuzzi, & Fick, 2011). Despite this changing view of cannabis, its use has been associated with psychosis (Arseneault et al., 2002; Degenhardt, Hall, & Lynskey, 2003; Griffith-Lendering et al., 2013; Moore et al., 2007), antisocial behavior (Fergusson, Horwood, & Ridder, 2007; Griffith-Lendering, Huijbregts, Mooijaart, Vollebergh, & Swaab, 2011; Monshouwer et al., 2006; Rey, Sawyer, Raphael, Patton, & Lynskey, 2002), and depression (Hayatbakhsh et al., 2007; Rey et al., 2002). Cannabis use has frequently been related to cognitive difficulties as well, for example with respect to executive functioning, implicit cognition, episodic memory, information processing speed, and emotional processing (Becker, Collins, & Luciana, 2014; Kelleher, Stough, Sergejew, & Rolfe, 2004; Pope, Gruber, Hudson, Huestis, & YurgelunTodd, 2001; Solowij & Battisti, 2008; Stacy & Wiers, 2010; Verdejo-Garcia, Lawrence, & Clark, 2008). It should be noted, however, that the aforementioned associations have not always been straightforward. With respect to psychopathology and psychological problems, associations often appear to be limited to specific aspects of cannabis use, such as an earlier starting age or a longer duration of use. Quantity or frequency of use also appears to play a key role, as well as preexisting vulnerability to psychological problems or psychopathology (Lynskey, Vink, & Boomsma, 2006; Schubart et al., 2011; Swift et al., 2012). With respect to cognitive outcomes, similar factors have been investigated (Nordstrom & Hart, 2006; Solowij & Battisti, 2008), but most attention has been given to acute versus nonacute effects of cannabis (e.g., Bossong, Jager, Bhattacharyya, & Allen, 2014; Crean, Crane, & Mason, 2011) and to defining the cognitive profile of cannabis users and whether a profile could be established that is specific for cannabis use (as opposed to a more generic profile associated with use of other substances as well; e.g., Griffith-Lendering, Huijbregts, Vollebergh, & Swaab, 2012; Morgan et al., 2012; Stacy & Wiers, 2010; Verdejo-Garcia et al., 2008). The goal of the present study was to investigate the role of specific aspects of cognition in associations between cannabis use and psychological problems. With respect to type of psychological problems, we focused on psychoneuroticism, which shows considerable genetic and phenotypic overlap with psychosis/schizophrenia (Macare, Bates, Heath, Martin, & Ettinger, 2012; Van Nierop et al., 2012), the disorders that have most consistently been associated with cannabis exposure (e.g., Moore et al., 2007). Psychoneuroticism
795
has been related to cannabis use before (Fridberg, Vollmer, O’Donnell, & Skosnik, 2011) and appears to represent one of the “softer” expressions of an extended psychosis phenotype (van Nierop et al., 2012), which is important as our sample consisted of individuals without diagnosed psychopathology. Moreover, even without diagnosed disorders, acute effects of cannabis appear to mimic symptoms associated with schizophrenia/psychosis (D’Souza et al., 2004; Wadsworth, Moss, Simpson, & Smith, 2006). With respect to cognitive functioning, we focused on social perception. The rationale behind this choice is that cannabis particularly appears to exert its influence on the functioning of the limbic system (including the amygdala), which plays an important role in social perception (Adolphs, 2002; Ochsner, 2008). Social perception problems, in turn, also constitute a central deficit in disorders that have been associated with cannabis use, including psychosis/schizophrenia (Demenescu, Kortekaas, Den Boer, & Aleman, 2010; Germine & Hooker, 2011; Marsh & Blair, 2008; Rössler et al., 2011). Limbic brain regions, including the amygdala, have high cannabinoid receptor (CB1) densities (Herkenham et al., 1990; Glass, Dragunow, & Faull, 1997). Animal studies indicate that cannabinoid signaling (as induced by cannabis exposure) induces long-term depression of gammaaminobutyric acid (GABA)-ergic interneurons within the amygdala’s basolateral nucleus, thereby upregulating the amygdala’s activity and sensitivity (Tan, Lauzon, Bishop, Bechard, & Laviolette, 2010). Phan and colleagues (2008) reported reduced amygdala reactivity during social information processing after administration of delta-9tetra-hydrocannabinol (Δ9-THC), the principle constituent of cannabis. Also, Gruber, Rogowska, and Yurgelun-Todd (2009) showed reduced amygdala activity during emotion perception in chronic cannabis users. Among long-term, chronic cannabis users abnormalities in amygdala volumes have also been observed (McQueeny et al., 2011; Yucel et al., 2008). Whereas studies yielded relatively consistent results with respect to abnormalities in amygdala structure and function, thereby often using stimuli that required the ability to recognize emotions from facial expressions, they did not focus on the quality of social perception. Only one recent study, by Platt, Kamboj, Morgan, and Curran (2010), did focus on performance during an emotion recognition task. Cannabis users were significantly slower than controls at identifying emotional expressions in a paradigm where facial expressions gradually changed from neutral to more intense expressions
796
HUIJBREGTS ET AL.
of sadness, anger, or happiness. Although the authors discussed the possible implications of their findings for vulnerability to psychological problems in cannabis users, they did not investigate this further. We sought to extend the research by Platt and colleagues by examining social perception in relation to the previously described psychoneuroticism among cannabis users. Based on the available evidence, it was to be expected that cannabis users would have more social perception difficulties and would report more affective mental health problems than nonusers. However, as most associations with impaired cognition or psychological problems have been established after acute use, in individuals with chronic/heavy use, or among those with indicators of vulnerability to mental health problems (Caspi et al., 2005; Henquet, Di Forti, Morrison, Kuepper, & Murray, 2008), effects were particularly expected among those who classified themselves as recreational users but used cannabis relatively frequently. It was further hypothesized that quality of social perception would influence the extent to which psychological problems were experienced, again particularly among relatively heavy users. In order to investigate specificity of social perception as a mediating or moderating cognitive factor in these associations, three tasks were used: a relatively easy face recognition task, a task requiring emotion recognition and working memory, and a task requiring sustained attention and inhibitory control.
METHOD Participants Participants were 150 young adults (74 men, 76 women; mean age: 24.7 years, SD = 3.7) recruited among University of Leiden undergraduate students and through advertisements on internet forums concerning cannabis topics. Advertisements were aimed at recruitment of recreational cannabis users and nonusing controls. Polydrug users and users of drugs other than cannabis were not excluded. Exclusion criteria (established by means of selfreport) were history of neurological illness or serious head injury, past or present major medical illness, and diagnosed psychopathology (without further specification of whether this was substance induced). Five potential participants (3 cannabis users, 2 nonusers) reported a diagnosed psychiatric disorder (3 depression, 1 bipolar disorder, 1 schizophrenia). These were excluded from statistical analyses. No participants reported a neurological or other major medical illness. Written informed
consent was obtained from all participants before the start of the study. Ethical approval for this study was granted by Leiden University’s Education and Child Studies Ethics Committee. Measures Cannabis use and use of other substances Cannabis use was assessed by asking participants about their lifetime use, their use during the past 12 months, and use during the past 4 weeks. For each time period, there were 14 answer categories in order to establish the frequency of use. The answer categories were as follows: 0 = I have never/not used cannabis, 1 = one time, 2 = two times, 3 = three times, … 10 = 10 times, 11 = 11– 19 times, 12 = 20 times or more, and 13 = 40 times or more. Furthermore, participants were asked about the age they first used cannabis and about the time of last use (answer categories: 1 = more than 12 months ago; 2 = 4–12 months ago; 3 = 1–3 months ago; 4 = 1–4 weeks ago; 5 = 3–7 days ago; 6 = 2 days ago; 7 = yesterday). The same questions with the same answer categories were asked for Ecstasy (3,4-methylenedioxymethamphetamine, MDMA), cocaine, and a maximum of three other illegal drugs (to be filled out by the respondents). For alcohol, answer categories were the same as those for cannabis and other drugs when use during the last 4 weeks was concerned, but for lifetime use and use during the last 12 months, the answer categories referred to the average frequency of use per month. For tobacco use, the answer categories referred to the average frequency of use per day. Aggregate scores are often used for the upper ranges of substance use reports, particularly those involving extended or nonrecent periods of time (e.g., lifetime use, use during the last 12 months), as inaccuracy of estimates strongly increases with increasing amounts of use and the passing of time. The measures used in the present study were similar to those used in many other studies (e.g., Creemers et al., 2010; Huizink, Ferdinand, Ormel, & Verhulst, 2006; Jaddoe et al., 2006; Monshouwer et al., 2006; O’Malley, Bachman, & Johnston, 1983; Otten, Barker, Maughan, Arseneault, & Engels, 2010; Prince van Leeuwen et al., 2011; Tait, MacKinnon, & Christensen, 2011). Psychoneuroticism The Symptom Checklist-90 (SCL-90; Derogatis, Lipman, & Covi, 1973; Elliott et al., 2006), a 90item self-report symptom inventory developed to measure psychological symptoms and distress, was
CANNABIS, SOCIAL PERCEPTION, AND PSYCHONEUROTICISM
used to measure psychoneuroticism. It was designed to be appropriate not only for use in clinical populations but also for use within community samples. The SCL-90, for which items are rated on 5-point scales reflecting the extent to which problems were experienced in the past 7 days, generates the following scales: somatic complaints (12 items), insufficiency of thoughts and actions (9 items), distrust (18 items), depression (16 items), anxiety (10 items), hostility (6 items), agoraphobia (6 items), and sleeping problems (3 items). In addition, a global score is obtained, called psychoneuroticism, using the overall score of the 90 items. Internal reliability of the different scales ranges from .77–.97 (Arrindell & Ettema, 1986; Olsen, Mortensen, & Bech, 2004). Neuropsychological tasks Three tasks from the Amsterdam Neuropsychological Tasks (ANT; De Sonneville, 1999), a battery of computerized tests, were used to assess neurocognitive abilities. Test–retest reliability, construct-, criterion-, and discriminant validity of the ANT tasks are satisfactory and have extensively been described elsewhere (e.g., De Sonneville et al., 2002; Huijbregts, Jahja, De Sonneville, De Breij, & Swaab-Barneveld, 2010). Before each part of a task the participants were given a standard verbal instruction and were given the opportunity to ask questions and to practice. Face Recognition (FR). This task (duration: 5 min) examined the ability to recognize neutral faces. A target face was presented on the monitor for 2.5 s. Following the presentation of the target face, a set of four photographs of individuals was
797
presented, and participants had to indicate whether or not the target individual appeared in the set of four (Figure 1). The gender and age category of the target (i.e., boys, girls, men, or women) match those of the subsequently shown set of four faces. A yes-response was given by pressing the mouse button below the index finger of the preferred hand; a no-response required a press of the mouse button below the index finger of the nonpreferred hand. There were 40 trials, in half of which the display set contained the target face. Matching of Facial Emotions (MFE). This task (duration: 10 min) measured the ability to match emotions using facial expressions. The expressed emotions are happiness, sadness, anger, and fear. In each of the 160 trials, two (digitized photographs of) faces expressing a particular emotion were presented simultaneously on the computer screen. The participants had to press the yes-button when the two faces expressed the same emotion and the nobutton when the facial emotions did not match (Figure 2). MFE may be considered a more demanding task than FR, as it requires executive functioning (working memory) in addition to social perception. The tasks can also be distinguished based on the fact that MFE specifically involves emotion recognition, whereas FR does not. Sustained Attention–Dots (SA-Dots). In this task (duration: approximately 12 min), 600 patterns of 3, 4, or 5 dots were presented in random configurations on the computer screen. The 3, 4, or 5 dots were each presented 200 times, but in random order. Participants were required to press the mouse button under the index finger of the nondominant hand when 3 or 5 dots appeared and to
Figure 1. Stimulus example (target face + display set) and timing of the trials for the Face Recognition task. © Sonares B.V. Reproduced by permission of Sonares B.V.
798
HUIJBREGTS ET AL.
Figure 2. Stimulus examples for the Matching Facial Emotions task. © Sonares B.V. Reproduced by permission of Sonares B.V.
press the mouse button under the index finger of the dominant hand when 4 dots appeared. A response had to be generated between 250 and 6000 ms after a signal. The fixed postresponse interval was 250 ms. Auditory feedback was provided after an error. This task requires sustained attention and executive functioning (inhibitory control; Huijbregts et al., 2002). Statistical analysis First, Spearman correlations were calculated to investigate associations among cannabis use measures (lifetime frequency of cannabis use, frequency of use during the last 12 months, frequency of use during the last 4 weeks, starting age, and most recent use), between cannabis use measures and use of other substances, and between cannabis and other substance use measures and neurocognitive and SCL-90 outcomes. Group differences regarding the aggregate SCL90 score (psychoneuroticism) and performance on the neurocognitive tasks were investigated using general linear model (GLM) analyses of variance. Separate analyses were performed for lifetime use, use during the last 12 months, and use during the last 4 weeks. Individuals were classified as nonusers, moderate users (
