The personality of pathological gamblers: A meta-analysis
Research highlights
► Meta-analysis of pathological gamblers' impulsivity and disinhibition traits. ► Large effects of Urgency and Low Premeditation forms of impulsivity. ► Large effects of Negative Affect, Disagreeable and Unconscientious Disinhibition. ► Pattern similar to substance use and Borderline and antisocial personality disorders. ► Pathological gambling is externalizing pathology and not Impulse Control Disorder.
Introduction
Problem gambling is the inability to resist recurrent urges to gamble excessively despite harmful consequences to the gambler or others. The Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2000) classifies pathological gambling (PG) as an Impulse Control Disorder that is defined by the presence of at least five symptoms that cause significant distress or impairment in social, family or occupational areas of life and that are not otherwise explained. Some of the symptoms are similar to phenomena seen in substance use disorders (i.e., mood alteration, tolerance, withdrawal, loss of control and preoccupation with gambling), and some are more specific to gambling behavior and consequent financial difficulties (i.e., chasing losses, lying about losses, harm to relationships or occupation, seeking a financial bailout, and committing illegal acts to obtain money). Symptoms of PG may be assessed as part of a clinical diagnostic interview or by psychometrically validated self-report scales. These include the National Opinion Research Center DSM-IV Screen for Gambling Problems (NODS) (Gerstein et al., 1999), the South Oaks Gambling Screen (SOGS) (Lesieur & Blume, 1987), the Gamblers' Anonymous Scale (GA20) (Ursua & Uribelarrea, 1998), and the Canadian Problem Gambling Severity Index (PGSI) (Ferris & Wynne, 2001).
In recent years, the trend toward increased access to gambling as a legal form of entertainment has contributed to incidence of PG. Estimates of the lifetime prevalence of PG in countries with legalized gambling range from 0.15 to 2.1% according to DSM-IV criteria, from 0.2 to 3.5% when symptoms are reported with the SOGS, and 0.5–1.4% with the PGSI (Stucki & Rihs-Middel, 2007). In one national representative survey, 54.5% of 9282 American adults reported gambling at least ten times in their lives, with 10.1% having gambled more than 1000 times, and 0.6% had at some point met the DSM-IV criteria for PG (Kessler et al., 2008).
General population surveys have found a high comorbidity of PG and other Axis I clinical syndromes. Pathological gamblers have elevated rates of substance use disorders (Bland et al., 1993, Cunningham-Williams et al., 1998, Gerstein et al., 1999, Welte et al., 2001) as well as mood and anxiety disorders (Kessler et al., 2008, Petry et al., 2005). Individual characteristics such as Negative Affect and disinhibition are strongly associated with these syndromes and with externalizing behavior generally (Krueger, Markon, Patrick, Benning, & Kramer, 2007). When present in unusual combinations, these personality traits may be symptomatic of a dimension that ranges from normality to dysfunction. Extremes of personality may be considered a form of psychopathology in their own right, and are currently described by the DSM-IV as Axis II personality disorders (Widiger, Livesley, & Clark, 2009). These individual characteristics have been studied extensively in the context of PG, and are the focus of the present study.
Antisocial and Borderline personality disorders occur at disproportionately high rates in clinical PG samples (Bagby et al., 2008, Blaszczynski and Steel, 1998, Fernandez-Montalvo and Echeburua, 2004, Pietrzak and Petry, 2005, Sacco et al., 2008), and the excessive reward-seeking behavior that is typical of these conditions may be an important contributor to PG. This is illustrated by the 2001–2002 National Epidemiologic Survey on Alcohol and Related Conditions. Among the 43,093 American adults surveyed, a tiny percentage (0.42%) had PG, but 60.8% of those with PG also had features indicating at least one of the Axis II personality disorders, including 23.3% with Antisocial personality disorder (Petry et al., 2005). No assessments were made of Borderline, Narcissistic, or Schizotypal features, so these figures likely underestimate the total prevalence of people with PG who also have symptoms of personality disorder.
Antisocial and Borderline personality disorders are debilitating conditions that are characterized by low Agreeableness and low Conscientiousness in the Five-Factor Model of personality (FFM) (Costa & McCrae, 1992), as well as non-normative scores on four facets of impulsivity (Samuel & Widiger, 2008). A recent meta-analysis has also found this same combination of traits to coincide with alcoholism and other substance use disorders (Kotov, Gamez, Schmidt, & Watson, 2010). If some trait or combination of traits is a mechanism linking PG with these personality and substance use disorders, then the most likely candidates might be impulsive and antagonistic traits. Current classification of PG as an Impulse Control Disorder would suggest some form of impulsiveness as the critical trait, but an alternative view would include PG as part of a wider cluster of externalizing behaviors that co-vary with the traits that typify substance use and cluster B personality disorders.
Classification of PG as an Impulse Control Disorder may be called into question if its temperamental risk profile is more similar to that of externalizing behaviors. Antagonistic personality traits have long been recognized to play a role in substance use and in Borderline and Antisocial personality disorders (Ball, 2005), but not in Impulse Control Disorders. The present review tested adequacy of these two conceptions in a series of meta-analyses of studies that measured the personality traits of pathological gamblers. These meta-analyses were organized around two integrative accounts of adult personality: the Whiteside and Lynam (2001) model of impulsivity, and the Markon, Krueger, and Watson (2005) Hierarchical Structural Model (HSM) of personality. We hypothesized that PG would be associated with some aspects of impulsivity because these traits are known to be common to both Impulse Control Disorders as well as substance use and cluster B personality disorders. We further hypothesized that disagreeable disinhibitory traits would be associated with PG, and that such a finding would support the conception of PG as a condition more akin to other externalizing syndromes rather than to the behaviors that are classed as Impulse Control Disorders.
There are several well-established self-report measures of impulsivity that have been used in the study of PG. The UPPS model of impulsivity (Whiteside & Lynam, 2001) integrates many of these measures into an empirically derived set of four dimensions. These dimensions were identified in a seminal factor analytic study of 10 prevalent impulsivity scales and their subscales. The scales loaded onto four different factors, each one reflecting a somewhat different aspect of impulsivity. In addition, Whiteside and Lynam identified individual questionnaire items that had the strongest loadings on each of the four factors and used these to develop the UPPS scales. Both the UPPS scales, as well as the original scales from which they were derived, indirectly tap latent constructs that correspond to four aspects of individuals' predisposition toward impulsive behavior. These constructs are assumed to reflect objectively real individual differences, but their measurement is atheoretical in the sense that no specific psychological or biological mechanisms are explicitly assumed to underlie them.
The four factors of the UPPS model are Negative Urgency, Low Premeditation, Low Perseverance, and Excitement Seeking. A fifth factor, Positive Urgency, has more recently been proposed (Cyders & Smith, 2008), but it is not considered in the current review because no studies have been published that compared the Positive Urgency of people with PG against that of nonpathological gamblers. UPPS Negative Urgency is the tendency toward rash and emotionally motivated action. According to the Whiteside and Lynam factor analysis, it is statistically associated with the Urgency facet of Neuroticism in the FFM and with the Attentional Impulsivity subscale of the Barratt Impulsiveness Scale (BIS) (Patton, Stanford, & Barratt, 1995). UPPS Low Premeditation is the tendency to act without adequate consideration of consequences. It is associated with the Nonplanning Impulsiveness subscale of the BIS, with high scores on the Eysenck Impulsivity scale (Eysenck, Pearson, Easting, & Allsopp, 1985), the Impulsiveness scale of the Temperament and Character Inventory (TCI) (Cloninger, Przybeck, Svrakic, & Wetzel, 1994), and with low scores on the Deliberation facet of FFM Conscientiousness. UPPS Low Perseverance is the tendency toward the quick extinction of nonrewarded behavior. It is associated with high scores on the Boredom Susceptibility subscale of the Sensation Seeking Scale (SSS-V) (Zuckerman, 1994), and with low scores on the Self-Discipline facet of FFM Conscientiousness. UPPS Sensation Seeking is the tendency toward behavior that results in novel and varied sensory stimulation and psychomotor arousal, and it is associated with high scores on the SSS-V Disinhibition subscale, Eysenck's Venturesomeness scale, and the Excitement Seeking facet of FFM Extraversion.
Several attempts have been made to identify a limited set of traits that may parsimoniously encompass the full variety of human emotion, cognition, and behavior. The most influential of these models assumes the number of factors necessary to explain individual variation to be two (Gray & McNaughton, 2000), three (Eysenck & Eysenck, 1976) or five (Costa & McCrae, 1992). These and other trait taxonomies (e.g., Cloninger et al., 1994, Digman, 1997, Krueger, 1999, Tellegen, 1982, Zuckerman et al., 1993), may all be objectively correct to some extent, even though they emphasize different psychobiological mechanisms. These disparate views were integrated into a unified model by Markon et al. (2005). This Hierarchical Structural Model (HSM) was tested both meta-analytically and by factor analysis, and included a variety of widely used personality trait inventories, as well as two measures of personality disorder. These measures, the Schedule for Nonadaptive and Adaptive Personality (SNAP) (Clark, 1993) and the Dimensional Assessment of Personality Pathology (DAPP) (Livesley & Jackson, 2002) are quite different from the other instruments. They were originally derived from clinical observations of psychiatric patients rather than factor analytic studies of questionnaires given to samples drawn from normative populations, and their inclusion broadens the scope of the HSM. The HSM is an attempt to integrate several important models of personality structure, and to expand their scope to cover the abnormal range of behavior as well as normality.
In the HSM, there are four levels of analysis and each level specifies individual differences using a different number of traits. The first level includes two factors, which Markon et al. (2005) referred to as Alpha and Beta after Digman (1997). At this level, Alpha includes traits from various instruments that all share some similarity with Eysenck and Eysenck's (1976) notion of Psychoticism and Neuroticism, while Beta includes traits similar to Extraversion. At the second level of the HSM, Alpha is referred to as Positive Emotionality, while Beta is split into two sub-traits: Negative Emotionality and Disinhibition. This second level is similar to Tellegen's (1982) model of personality (i.e., Positive Affect, Negative Affect, and Constraint). At the third level of the HSM, Disinhibition is further subdivided into Unconscientious Disinhibition and Disagreeable Disinhibition. At this level, the three Alpha traits are approximately the same as Neuroticism, low Agreeableness, and low Conscientiousness in the FFM, as Harm Avoidance, low Cooperativeness, and Novelty Seeking in Cloninger's psychobiological model (Cloninger et al., 1994), and as Neuroticism-anxiety, Aggression-hostility, and Impulsive Sensation Seeking in the Alternative Five model of Zuckerman et al. (1993). Finally, at the lowest level of the HSM, Beta divides into FFM Extraversion and Openness to Experience.
Two features of the HSM are important for the present review. First, because the UPPS impulsivity traits are very similar to four of the FFM facets, and the HSM includes the FFM domains, all four of the UPPS impulsivity traits are ultimately subsumed within the HSM. Urgency (Negative Urgency in the UPPS) is a facet of FFM Neuroticism (Negative Emotionality in the HSM). Excitement Seeking (UPPS Sensation Seeking) is a facet of FFM Extraversion (HSM Positive Emotionality). Self-Discipline (UPPS Low Perseverance) and Deliberation (UPPS Low Premeditation) are both facets of FFM Conscientiousness (HSM Unconscientious Disinhibition). The HSM factors include the impulsivity facets but they are not equivalent to them, since each of the FFM domains includes a total of six facets. For instance, an individual might have elevated Urgency, but a Neuroticism score that is not atypical. Likewise, PG might be characterized by elevations on some of the UPPS impulsivity traits, but does not necessarily show elevations on the respective FFM domains or their respective HSM factors.
The second important feature of the HSM is that it allows personality to be modeled at one of four different levels of specificity. Because of the high comorbidity of PG and personality disorders, the appropriate level of analysis for the present review is the level at which the HSM most parsimoniously allows the Axis II personality disorders to be characterized. According to a meta-analysis of the FFM and personality disorders (Samuel & Widiger, 2008), these conditions can be specified as combinations of non-normative scores on four of the FFM domains: Neuroticism, Extraversion, Agreeableness and Conscientiousness. This corresponds closely to the third level of the HSM, which includes Negative Emotionality, Positive Emotionality, Disagreeable Disinhibition, and Unconscientious Disinhibition. None of the DSM-IV personality disorders was reliably associated with high or low scores on Openness or any of its facets. For the present review, we adopted the four-factor level of the HSM.
Pathological gambling is an Impulse Control Disorder that has high comorbidity with personality disorders, especially Antisocial and Borderline. Both Antisocial and Borderline personality disorders are characterized by very high impulsivity and they both were found to have elevations on all four of the UPPS-related facets in the Samuel and Widiger meta-analysis. It is possible that the association between PG and the personality disorders may be an artifact of the impulsivity that is symptomatic of these conditions. If that is true, then the present meta-analyses might show elevations in the UPPS traits in PG groups relative to NPG groups, but no significant differences on other traits that are known to be associated with personality disorder (i.e., low FFM Agreeableness and other Disagreeable Disinhibition traits). Since Neuroticism (HSM Negative Emotionality) and low Conscientiousness (HSM Unconscientious Disinhibition) are both typical of Antisocial and Borderline personality disorders, and since both have impulsivity traits as facets, we predicted these to show group differences across studies comparing PG to NPG. This pattern of results would support an impulsivity–PG hypothesis and would be consistent with the current diagnostic classification of PG as an Impulse Control Disorder.
An alternative view to the impulsivity–PG hypothesis is that traits other than impulsivity may contribute to both PG and to comorbid personality disorder. In the Samuel and Widiger meta-analysis, strong associations were found between low Agreeableness and both Antisocial and Borderline personality disorder. Agreeableness does not contain an impulsivity-related facet in the FFM, nor do any of the Disagreeable Disinhibition traits from other taxonomies. We predicted that PG would be associated with trait scores indicating elevated Disagreeable Disinhibition, in addition to the impulsivity and Negative Affect that are predicted by the impulsivity–PG hypothesis. Evidence for this notion would support a characterization of PG not as an Impulse Control Disorder, but as part of a broader spectrum of externalizing psychopathologies that have common etiology with features of Antisocial and Borderline personality disorders. We will refer to this set of predictions as the externalizing–PG hypothesis.
It is widely expected that the next edition of the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2000) will list PG as a behavioral addiction rather than as an Impulse Control Disorder (Petry, 2010). Justification for such a change would be bolstered by evidence that individual characteristics that contribute to the etiology of PG are similar to those that predict emergence of other addictive behaviors (i.e., the externalizing–PG hypothesis). Maintaining PG as an Impulse Control Disorder would be supported by evidence that pathological gamblers have elevated impulsiveness, but without additional traits known to predict alcoholism and other addictive behaviors (i.e., the impulsivity–PG hypothesis).
Section snippets
Literature search
An exhaustive search of the PsychInfo database was conducted in July, 2010. Search criteria were the presence of the words Gambling, Gambler, or Gamble anywhere in the articles, plus any of the following: Impulsivity, Impulsiveness, Eysenck Personality Questionnaire, Venturesomeness, Psychoticism, Barratt Impulsivity Scale, Sensation Seeking, SSS-V, ZKPQ, NEO-FFI, NEO PI-R, Neuroticism, Extraversion, Agreeableness, Conscientiousness, Tridimensional Personality Questionnaire, TPQ, Temperament
Negative Urgency
Each of the studies that were included in this analysis reported PG and NPG group means and standard deviations for the UPPS Urgency scale, the BIS Attentional Impulsivity subscale or the NEO PI-R Impulsiveness facet or its public domain counterpart, NEO-IPIP Immoderation. These measures were included because they loaded most heavily onto the factor identified as Negative Urgency by Whiteside and Lynam (2001). These 11 studies contained 363 PGs and 581 NPGs (see Table 2). The mean age was
Discussion
Meta-analysis was used to assimilate the results from studies that compared the personality traits of people with PG versus NPG control groups. Studies in these analyses compared PG and NPG groups that were similar in age and gender, that were drawn from a variety of clinical, community, convenience and student samples, and that used a variety of self-report and clinical interview methods of determining PG status. Total numbers of participants in the studies forming each meta-analysis were
Acknowledgement
This work was supported by a grant from the Ontario Problem Gambling Research Center.
References⁎ (108)
- et al.
Comparison of personality risk factors in bulimia nervosa and pathological gambling
Comprehensive Psychiatry
(2007) - et al.
Pathological gambling and the five factor model of personality
Personality and Individual Differences
(2007) Personality traits, problems and disorders: Clinical applications to substance use disorders
Journal of Research in Personality
(2005)- et al.
Plasma testosterone and pathological gambling
Psychiatry Research
(2001) - et al.
Decision-making impairments in patients with pathological gambling
Psychiatry Research
(2005) - et al.
Intermittent explosive disorder — revised: Development, reliability, and validity of research criteria
Comprehensive Psychiatry
(1998) - et al.
Prevalence and predictors of pathological gambling: Results from the St. Louis Personality, Health and Lifestyle (SLPHL) study
Journal of Psychiatric Research
(2005) - et al.
Age norms for impulsiveness, venturesomeness and empathy in adults
Personality, & Individual Differences
(1985) Faites vos jeux! Another look at sensation seeking and pathological gambling
Personality and Individual Differences
(2004)- et al.
Testing the validity of the need inventory (NISS) within a clinical context: The role of pathological gambling
Journal of Research in Personality
(2010)
Decision-making and neuroendocrine responses in pathological gamblers
Psychiatry Research
Behavioral assessment of impulsivity in pathological gamblers with and without substance use disorder histories versus healthy controls
Drug and Alcohol Dependence
Reward drive, rash impulsivity and punishment sensitivity in problem gamblers
Personality and Individual Differences
Problem gambling and the Five Factor Model in university students
Personality and Individual Differences
The dimensional assessment of personality in pathologic and social gamblers: The role of novelty seeking and self-transcendence
Comprehensive Psychiatry
Disinhibition and negative affectivity in substance abusers with and without a gambling problem
Addictive Behaviors
The subtyping of pathological gambling: A comprehensive review
Clinical Psychology Review
Personality factors as predictors of pathological gambling
Personality and Individual Differences
The association between gambling pathology and personality disorders
Journal of Psychiatric Research
A meta-analytic review of the relationships between the five factor model and DSM-IV-TR personality disorders: A facet level analysis
Clinical Psychology Review
Relations between personality and drinking motives in young people
Personality and Individual Differences
Why gamblers fail to win: A review of cognitive and neuroimaging findings in pathological gambling
Neuroscience and Biobehavioral Reviews
The five factor model and impulsivity: Using a structural model of personality to understand impulsivity
Personality and Individual Differences
Pathological gamblers are neither impulsive nor sensation-seekers
Australian and New Zealand Journal of Psychiatry
Diagnostic and statistical manual of mental disorders
Personality disorders and pathological gambling: A review and re-examination of prevalence rates
Journal of Personality Disorders
Personality disorder and dimension differences between type A and type B substance abusers
Journal of Personality Disorders
Epidemiology of pathological gambling in Edmonton
Canadian Journal of Psychiatry
Pathological gamblers, heroin addicts and controls compared on the E.P.Q. “Addiction Scale”
British Journal of Addiction
Boredom proneness in pathological gambling
Psychological Reports
A pathways model of problem and pathological gambling
Addiction
Personality disorders among pathological gamblers
Journal of Gambling Studies
Sensation seeking in a French population of pathological gamblers: Comparison with regular and nongamblers
Psychological Reports
Introduction to meta-analysis
Factors determining the severity of pathological gambling in males
Journal of Gambling Studies
Low platelet monoamine oxidase activity in pathological gambling
Acta Psychiatrica Scandinavica
Cognitive, dispositional and psychophysiological correlates of dependent slot machine gambling in young people
Journal of Applied Social Psychology
Manual for the schedule for nonadaptive and adaptive personality
Impulsiveness, locus of control, motivation and problem gambling
Journal of Gambling Studies
The Temperament and Character Inventory (TCI): A guide to its development and use
Statistical power analysis for the behavioral sciences
NEO PI-R professional manual
Taking chances: Problem gamblers and mental health disorders — Results from the St. Louis Epidemiologic catchment area study
American Journal of Public Health
Emotion-based dispositions to rash action: Positive and negative urgency
Psychological Bulletin
Higher-order factors of the big five
Journal of Personality and Social Psychology
A milti-site study of alcohol subtypes: Classification and overlap of unidimensional and multi-dimensional typologies
Addiction
Psychoticism as a dimension of personality
Pathological gambling and personality disorders: An exploratory study with the IPDE
Journal of Personality Disorders
The Canadian problem gambling index: Final report
Neuropsychological characteristics and personality traits in pathological gambling
CNS Spectrums
Cited by (170)
Do attention-deficit/hyperactivity symptoms influence treatment outcome in gambling disorder?
2024, Comprehensive PsychiatryGambling disorder duration and cognitive behavioural therapy outcome considering gambling preference and sex
2023, Journal of Psychiatric ResearchPsychometric validation of the desire thinking questionnaire in a Turkish adolescent sample: Associations with internet gaming disorder
2022, Addictive BehaviorsCitation Excerpt :In these recent studies, the samples consisted of both adolescent and adult individuals, and psychosocial correlates of IGD such as personality traits and impulsivity were not controlled. However, particular personality traits (high levels of neuroticism and low levels of conscientiousness) have been found to be related to addictive behaviors including alcohol use (Malouff et al., 2007), nicotine use (Munafò et al., 2007), gambling (MacLaren et al., 2011), and problematic Internet use (Andreassen et al., 2013). Additionally, one recent study examined the role of both positive and negative affect on IGD and it was found that both affect types act as mediators of the pathway between impulsivity and IGD among young adults (Shin et al., 2019).
Individual risk factors and prediction of gambling disorder in online sports bettors - the longitudinal RIGAB study
2024, Frontiers in Psychiatry
- ⁎
Article included in meta-analysis.