The influence of chronological age on cognitive biases and impulsivity levels in male patients with gambling disorder Journal of Behavioral  ROSER GRANERO1,2, FERNANDO FERNANDEZ-ARANDA 1,3,4 , Addictions  3 SUSANA VALERO-SOLIS , AMPARO DEL PINO-  GUTIERREZ 3,5 , GEMMA MESTRE-BACH1,3, ISABEL BAENAS3, 9 (2020) 2, 383-400 S. FABRIZIO CONTALDO3, MONICA   GOMEZ-PE ~ 3, NA DOI: NEUS AYMAMI , LAURA MORAGAS , CRISTINA VINTRO 3 3  3, 10.1556/2006.2020.00028 © 2020 The Author(s) TERESA MENA-MORENO1,3, EDUARDO VALENCIANO- MENDOZA3,6, BERNAT MORA-MALTAS3, JOSE  M. MENCHON  3,4,6  and SUSANA JIMENEZ-MURCIA1,3,4p 1 CIBER Fisiopatologıa Obesidad y Nutrici on (CIBERObn), Instituto de Salud Carlos III, Madrid, FULL-LENGTH REPORT Spain 2 Department of Psychobiology and Methodology, Autonomous University of Barcelona, Barcelona, Spain 3 Department of Psychiatry, University Hospital of Bellvitge-IDIBELL, Barcelona, Spain 4 Department of Clinical Sciences, School of Medicine and Health Sciences, University of Barcelona, Barcelona, Spain 5 Department of Public Health, Mental Health and Perinatal Nursing, School of Nursing, University of Barcelona, Barcelona, Spain 6 CIBER Salud Mental (CIBERSam), Instituto de Salud Carlos III, Madrid, Spain Received: November 6, 2019 • Revised manuscript received: February 16, 2020; April 3, 2020 • Accepted: April 4, 2020 Published online: June 22, 2020 ABSTRACT Background and aims: Due to the contribution of age to the etiology of gambling disorder (GD), there is a need to assess the moderator effect of the aging process with other features that are highly related with the clinical profile. The objective of this study is to examine the role of the chronological age into the relationships between cognitive biases, impulsivity levels and gambling preference with the GD profile during adulthood. Methods: Sample included n 5 209 patients aged 18–77 years-old recruited from a Pathological Gambling Outpatients Unit. Orthogonal contrasts explored polynomial patterns in data, and path analysis implemented through structural equation modeling assessed the underlying mech- anisms between the study variables. Results: Compared to middle-age patients, younger and older age groups reported more impairing irrational beliefs (P 5 0.005 for interpretative control and P 5 0.043 for interpretative bias). A linear trend showed that as people get older sensation seeking (P 5 0.006) and inability to stop gambling (P 5 0.018) increase. Path analysis showed a direct effect between the cognitive bias and measures of gambling severity (standardized effects [SE] between 0.12 and 0.17) and a direct effect between impulsivity levels and cumulated debts due to gambling (SE 5 0.22). Conclusion: Screening tools and intervention plans should consider the aging process. Specific programs should be *Corresponding author. Department of developed for younger and older age groups, since these are highly vulnerable to the consequences of Psychiatry, Bellvitge University gambling activities and impairment levels of impulsivity and cognitive biases. Hospital, c/ Feixa Llarga s/n, Hospitalet de Llobregat, Barcelona, 08907, Spain. Tel.: þ34 93 260 79 KEYWORDS 88; fax: þ34 93 260 76 58. E-mail: sjimenez@bellvitgehospital.cat cognitive biases, gambling disorder, impulsivity, older age, path analysis, younger age Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC 384 Journal of Behavioral Addictions 9 (2020) 2, 383-400 INTRODUCTION and intensity of the gambling severity and the continuation of the gambling activity is even justified by gamblers, arguing that they are learning and developing the required Epidemiological data for gambling disorder skills/abilities to win (Chretien, Giroux, Goulet, Jacques, & Systematic reviews of epidemiological studies have shown Bouchard, 2017; Emond & Marmurek, 2010; Kovacs, Rich- significant increases in the prevalence of gambling disorder man, Janka, Maraz, & Ando, 2017; Leonard & Williams, (GD) worldwide during the last decades, with cross-sectional 2016). At a psychological level, it has been postulated that estimates (over the last 12 months) around of 0.1–6% in the these cognitive distortions related to the gambling severity general population in developed countries (Calado & Grif- could be explained by three main mechanisms. The first fiths, 2016). Prevalence studies also warn of a potentially hypothesis should be the generic poor capacity of humans greater risk of problematic gambling in the near future themselves in processing probability/chance and judging among all sectors of the population as a result of the prev- randomness (Williams & Griffiths, 2013; Yu, Gunn, Osh- alent ease accessibility to gambling platforms and the in- erson, & Zhao, 2018). The second hypothesis refers to the crease in the opportunity to gamble (Suissa, 2015). specific structural characteristics of some games that could Furthermore, there is an awareness of the high-vulnerability promote cognitive distortions (e.g., the stimuli of bright of two age groups for the onset and intensification of the flashing lights and loud noises of slot-machine that disordered gambling: during adolescence (even at ages when accompany each win) (Myles, Carter, & Yucel, 2019). And bets on gambling is illegal) and early adulthood stages thirdly, psychobiological approaches show that cognitive (Giralt et al., 2018), and among the elderly (Subramaniam biases in GD could be the result of the brain reward func- et al., 2015; Tse, Hong, Wang, & Cunningham-Williams, tions: a) neurochemical research has related dysregulation in 2012). These disturbing reports have led to the appearance serotonin, noradrenaline, and glutamate functions with poor of new empirical research in order to provide a compre- decision-making performance (van Timmeren, Daams, van hensive view of the GD phenotypes. New studies should be Holst, & Goudriaan, 2018); and b) functional neuroimaging focused on the most high-risk groups, and on the analysis of studies and neuropsychological measurements of impulsivity the multivariate relationships between several risk factors and risky decision-making have revealed damage in the (including interaction and mediational effects). brain function of GD patients (mainly in the ventromedial prefrontal cortex and striatum), as well as impairments in the executive functions (van Holst, van den Brink, Veltman, Relevance of cognitive performance on the GD profiles & Goudriaan, 2010). Research in the cognitive area also Cognitive biases related to gambling behavior are a classical evidences that aging is strongly related with declines in and challenging area for the study of GD. Patients with GD cognitive abilities, which could render older adults highly systematically report relevant cognitive distortions related susceptible to many cognitive biases (Miquel et al., 2018). with the onset of problematic gambling, its maintenance and Since these deviations from rationality in judgments could the difficulty overcoming this dependence. Studies have moderate the relationship between age and the final deci- shown that irrational thoughts are pervasive in most forms sion-making outcomes (Bangma, Fuermaier, Tucha, Tucha, of problematic gambling, and that a number of erroneous & Koerts, 2017; Bruine de Bruin, Parker, & Fischhoff, 2012), beliefs held by GD patients seem to affect their capacity to it has been postulated that aging-related cognitive decline estimate the real chances of winning, and seriously condition greatly impacts older adults’ daily life, including their GD their fallibility of decision making mechanisms (Mallorqui- profile (Paolini, Leonardi, Visani, & Rodofili, 2018). But Bague et al., 2019; Verdejo-Garcia, Alcazar-Corcoles, & despite the promising results in this area, it remains unclear Albein-Urios, 2019). The cognitive approach to problematic how chronological age influences the direction, strength/s, gambling has identified several types of cognitive biases and precise mechanisms of the relationships between (Clark, 2010; Clark & Limbrick-Oldfield, 2013; Goodie & cognitive styles and problematic gambling-related behaviors. Fortune, 2013; Levesque, Sevigny, Giroux, & Jacques, 2018), which finally give rise to an “illusion of personal control” Relevance of impulsivity as a core mechanism in the over the game. Gamblers usually overestimate their capacity GD area of control and can even confuse chance games with games of skill. This results in a perception of expected value of Another core concept to understanding GD profile is gambling as a positive when expected value is really negative. impulsivity, currently considered as a complex multidi- Studies have systematically observed that GD patients usu- mensional construct explaining behaviors that may be ally overvalue recent results when evaluating the chances of unduly hasty, risky, and/or inappropriate, leading to nega- a certain outcome occurring (recency bias). They only seek tive outcomes. Recent models of impulsivity address both its out information that supports what is called gamblers initial behavioral manifestations and the underlying brain-based gut decision in ignoring evidence to the contrary that might mechanisms, and highlight that many psychopathological be a red flag to a given decision (confirmation bias). Like- conditions (including behavioral addictions) require the wise, they may believe that a win is necessarily due after a description of the impulsivity as a core mechanism (Lee, series of loses (gambler’s fallacy or near-miss effect [un- Hoppenbrouwers, & Franken, 2019; Sharma, Markon, & successful outcome is proximal to a win]). The frequency Clark, 2014; Tiego et al., 2019). In fact, GD has been Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC Journal of Behavioral Addictions 9 (2020) 2, 383-400 385 commonly listed alongside the impulse control disorders, literature in recent years (Kyonka & Schutte, 2018; Steward largely as a consequence of the high level of personality traits et al., 2017). As regards the models of impulsivity used to related to impulsivity reported by GD patients (such as assess the relationships between personality traits and novelty/sensation seeking, lack of perseverance/premedita- problem gambling, one of the most widely used in the tion, or positive/negative urgency), and by the results of the research area nowadays is the UPPS-P scale (Canale, Vieno, neurobiological models measuring the relationship between Bowden-Jones, & Billieux, 2017), originally based on a impulsivity levels and gambling activity (Chamberlain, multi-faceted conception comprising five impulsive per- Stochl, Redden, & Grant, 2018; Ioannidis, Hook, Wickham, sonality traits: lack of premeditation, lack of perseverance, Grant, & Chamberlain, 2019; Maclaren, Fugelsang, Harri- sensation seeking, positive urgency, and negative urgency gan, & Dixon, 2011; Rochat, Billieux, Gagnon, & Van der (Whiteside & Lynam, 2001). This scale was defined based on Linden, 2018). Although few studies have addressed the exploratory factor analysis to identify the personality facets structure of impulsivity in problematic and disordered associated with impulsive behaviors from among several gambling (Gullo, Loxton, & Dawe, 2014; Hodgins & Holub, other commonly used measurements of impulsivity, and it 2015; Kr€aplin et al., 2014; MacKillop et al., 2016), the has shown to have robust correlation with different forms of analysis of the impulsivity levels in different personality psychopathology. But despite the extensive literature on GD domains has received much attention. Along this line, large and its relations to impulsivity, few studies have explored the cross-sectional associations have been found between mea- contribution of chronological age to these associations. The surements of impulsivity and gambling severity (including available empirical data show complex underpinnings the level of gambling symptoms, frequency of gambling (Ioannidis et al., 2019; Mitchell & Potenza, 2014). While activity, bets per gambling-episodes, or even debts due to the early impulsivity levels have been proven to increase the risk gambling practices) in both clinical and population-based of impairing gambling behaviors in later life (Dussault et al., samples (Black et al., 2015; Grall-Bronnec et al., 2012; Yan, 2011), it also seems that regardless of age range gambling Zhang, Lan, Li, & Sui, 2016). Several pioneer longitudinal activity induces impulsivity levels that evolve into compul- studies have also suggested that impulsivity levels during sion, chronic forms of addiction and more severe gambling childhood may have a predictive capacity on the problematic behavior (Hodgins & Holub, 2015; Kovacs et al., 2020). This gambling in emerging adulthood (Dussault, Brendgen, scenario accentuates the need for new research with a special Vitaro, Wanner, & Tremblay, 2011). For example, the focus on the mechanisms underlying the multidimensional literature has highlighted the robust link between behavioral components of impulsivity and age in the GD area. addiction and neurodevelopmental disorders characterized by inattention and hyperactivity-impulsivity, such as the Consideration of the preferred forms of gambling in the presence of attention-deficit hyperactivity disorder (ADHD) (Brandt & Fischer, 2019). Various hypotheses have been etiology of the GD suggested to explain the mechanisms linking ADHD Finally, there is currently great interest in the study of (particularly when it persists in adulthood) with GD (Jacob, preferred forms of gambling. Although most studies pub- Haro, & Koyanagi, 2018), high impulsivity levels being one lished to date have a preference-blind research approach, of the core features (Abouzari, Oberg, Gruber, & Tata, empirical research suggest that the gambling subtypes could 2015). It has also been postulated that children with ADHD provide an insight into the etiology and treatment of GD, usually report lower intelligence quotient (IQ) than control and that taking the specific preferred gambling activity into subjects (Biederman, Fried, Petty, Mahoney, & Faraone, account could be constraining individuals’ phenotype (even 2012), and the combination of high impulsivity levels with about the gambling-related harm) and might offer infor- lower IQ could lead to a higher risk of GD (Rai et al., 2014). mation about treatment response and disease course (Ste- Moreover, the presence of ADHD during childhood has vens & Young, 2010; Subramaniam et al., 2016). Two broad been defined as a risk of personality disorders (such as categories have been proposed for grouping gambling ac- borderline, antisocial, avoidant or narcissistic personality) tivity based on the role of chance in the outcome of the game and other psychiatric conditions (the most frequent being (Odlaug, Marsh, Kim, & Grant, 2011): non-strategic games anxiety/mood disorders and substance-use disorders), which (also called chance-based games, since little [or no] decision could play an important role in ADHD, impulsivity and making or skill can be used by gamblers in determining the problem gambling (Fatseas et al., 2016). Finally, typical outcome; e.g., lotteries, slots-machines, bingo) versus stra- ADHD symptoms have been related to the characteristic tegic games (also called skill-based games, since autonomous cognitive impairments of problem gambling (Chamberlain, decision making skills can be by used by gamblers in Derbyshire, Leppink, & Grant, 2015). determining the outcome; e.g., poker, sports/animals betting, Furthermore, high levels of decision-making impulsivity craps, stock market). The study of the correlates of the (largely motor inhibition, attention inhibition or decision- gambling preference have found multiple reasons that lead making tasks) have been related to the onset and progression individuals to a preferred gambling style, including socio- of the GD (Ioannidis et al., 2019). Positive connections be- demographics (gender, age, education level, civil status and tween probability discounting (a cognitive bias defined as social position (Jimenez-Murcia et al., 2019; Kastirke, the subjects’ tendency to overvalue reinforcement with lower Rumpf, John, Bischof, & Meyer, 2015), accessibility/avail- odds) and gambling have also been reported in the scientific ability of the gambling platforms (Moore, Thomas, Kyrios, Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC 386 Journal of Behavioral Addictions 9 (2020) 2, 383-400 Bates, & Meredyth, 2011), certain personality traits (mainly METHODS novelty/sensation seeking and impulsivity levels) (Lorains, Stout, Bradshaw, Dowling, & Enticott, 2014b; Navas et al., Participants 2017), and even the psychological state and level of the disordered gambling (Bonnaire et al., 2017; Chamberlain, The data analyzed in this work correspond to a research Stochl, Redden, Odlaug, & Grant, 2017; Ledgerwood & project developed at the Pathological Gambling Outpatient Petry, 2010; Suomi, Dowling, & Jackson, 2014). Regarding Unit at University Hospital of Bellvitge, with the objective to the effect of chronological age on the gambling preference, examine risk factors for gambling behavior in the adulthood chance-based games are more likelihood selected by older population of individuals with gambling behavior. The individuals, who tend to select low skill (high chance) initial sample considered for the study included n 5 227 gambling activities (Moragas et al., 2015). It has been sug- patients recruited between July 2016 and October 2016, gested that age-related vulnerabilities of the brain typical when they first attended for assessment and before starting among elderly (in particular the poor neuropsychological treatment. Inclusion criteria in the study were age 18þ performance, which means reasoning slowness, difficulty to years-old, met clinical criteria for GD and education level gain explicit insight into the rules of decision tasks, and and cognitive capacity to complete the self-report mea- limited ability in assess the real risks in decision making surements of the study. Only patients who sought treatment activities), could be a reason why elders select games char- for GD as their primary health concern were admitted to acterized by lower decision making processes (Mouneyrac this study. et al., 2018; Schiebener & Brand, 2017). On the other hand, After accepting to be part of the study and completing lower age, higher levels in different impulsivity domains, and the whole assessment, 16 women were excluded, due to the better cognitive performance, have been related to the low frequency of this gender in the study and the difference preference for skill-based games (Jimenez-Murcia, Granero, in the age distribution between sexes (in the range 18–77 for Fernandez-Aranda, & Mench on, 2020). And despite men and 37–65 for women). Another two participants were gambling preferences appearing to be relevant to the study not included in the analysis due to lack of response to the of GD phenotypes and evidence existing of variables that questionnaire measuring cognitive biases. Therefore, the may favor a preferred style of gambling, the underlying final sample for the study was n 5 209 patients. processes considering simultaneously cognitive biases, Three groups of age were defined in the study based on impulsivity levels, and chronological age are unknown. the tertiles in the study, with the aim to divide the sample in three parts each containing approximately a third of the participants (3 groups were considered to guarantee sample Objectives size enough for the statistical comparison and the remaining In summary, empirical evidence support that cognitive biases analyses). The three groups were labeled in this wok as and impulsivity levels play a relevant role for both the devel- “younger age” (18–35 years, n 5 73), “middle age” (36–45 opment and progression of the GD, and it seems that these years, n 5 63), and “older age” (46–77 years, n 5 73). All the constructs could be meaningfully interrelated. Studies also data analyzed in this work correspond to the first assessment suggest that gambling preference may be clinically significant before the patients began the therapy. and could contribute towards the phenotype of GD patients. However, few studies have explored how patients’ age can Measures modulate the underlying mechanism between this set of vari- ables during the adulthood, and to our knowledge no research Diagnostic Questionnaire for Pathological Gambling (ac- has analyzed mediational links through path analysis. cording to DSM criteria) (Stinchfield, 2003). This is a self- The aim of this study was to assess the role of the aging report questionnaire including 19 items coded in a binary process in the relationships between impulsivity profile and scale (yes-no), originally developed for diagnosing GD ac- cognitive bias with gambling preferences and severity. The cording to the DSM-IV-TR (American Psychiatric Associ- specific objectives were: a) to explore polynomial trends ation, 2010). This tool has currently been adapted to assess between patients’ chronological age with impulsivity and the DSM-5 criteria for GD (American Psychiatric Associa- cognitive distortions; and b) to assess the underlying tion, 2013) by deleting the illegal acts symptom and fixing a mechanisms through path analysis (including mediational cut-off of four symptoms to diagnose GD (five was the cut- links) between the study variables: age, impulsivity profile, off for the DSM-IV-TR criteria). This self-report can obtain cognitive biases and gambling severity levels. Analyses were different measurements for the GD based on the DSM-5 performed in a clinical sample of patients with ages between taxonomy: the presence/absence for each DSM criterion, the 18 to 77 years-old treatment seeking due the problematic presence/absence diagnosis for GD, a dimensional mea- gambling. Based on the empirical evidence we hypothesized: surement of the gambling severity (total number of DSM a) positive linear trends between age and impulsivity and criteria, obtained as the sum for the individual criteria), and cognitive bias levels; b) positive correlations between the GD severity grouped in four levels [non-problematic cognitive distortions and impulsivity levels; c) a mediational gambling (for individuals who met 0 criteria), problematic link between age, cognition and impulsivity measures, and gambling (for 1–3 criteria), moderate-GD (for 4–5 criteria), gambling severity measures. mild-GD (for 6–7 criteria), and severe-GD (for 8–9 Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC Journal of Behavioral Addictions 9 (2020) 2, 383-400 387 criteria)]. The Spanish adaptation of the questionnaire used Statistical analysis in this work obtained very good psychometrical properties (Cronbach’s alpha equal to a 5 0.81 for general population Statistical analysis was carried out with Stata16 for windows and a 5 0.77 for clinical sample) (Jimenez-Murcia et al., (Stata-Corp, 2019). Firstly, the association between patients’ age 2009). The internal consistency obtained in the sample of and cognitive biases (GRCS scales) and impulsivity levels this work was adequate (a 5 0.76). (UPPS-P scales) was estimated with Pearson correlation co- South Oaks Gambling Severity Screen (SOGS) (Lesieur & efficients (R, also called the Pearson product-moment correla- Blume, 1987). This is a 20-item self-report questionnaire tion coefficient). The Curve Estimation Fit procedure (CurveFit) developed with the aim of measuring the symptom level of was also used to test goodness-of-fit for functions different to the problem gambling, as well as the related negative con- the linear-model (logarithmic, inverse, quadratic, cubic, power, sequences. A total score is generated as the sum of the items, S, growth, exponential and logistic). The resulting non-adequate typically considered as a measurement of the GD severity. fit for most of the relationships suggested that a better under- Good psychometrical properties were obtained for the tool standing of the associations between the variables should be in different clinical and population-based settings (Lesieur & provided, categorizing age. Since there is no consensus Blume, 1993), as well as for the Spanish validation used in regarding the bounds for age ranges within the gambling area this work (test-retest reliability R 5 0.98, internal consis- (bounds substantially vary between studies according to tency a 5 0.94 and convergent validity R 5 0.92) different definitions and criteria), we decided on a classification (Echebur ua, Baez, Fernandez, & Paez, 1994). The internal in three groups based on the tertiles estimated in the sample consistency in the study was a 5 0.712. itself, in this study labeled younger, middle and older age. Gambling Related Cognitions Scale (GRCS) (Raylu & Oei, Secondly, the comparison of the means registered for the 2004). This is a 23-item self-report questionnaire used to cognitive biases (GRCS scales) and the impulsivity levels assess gambling-related cognitions in both population-based (UPPS-P scales) between the age groups (younger, middle, and clinical disordered gambling samples. Items are struc- and older) was based on analysis of variance (ANOVA), tured into five primary cognitive factors: gambling related which included orthogonal polynomial contrasts to assess expectancies, illusion of control, predictive control, linear and quadratic trends and post-hoc multiple compari- perceived inability to stop gambling, and interpretative bias. sons with the least significant difference estimation method. A total score is also available as the sum of the primary- For this set of statistical comparisons, Finner’s method (a factor scores. The internal consistency in the study was familywise error rate procedure which is more powerful than a 5 0.76 for expectancies, a 5 0.78 for illusion of control, the classical Bonferroni correction) was used to control in- a 5 0.77 for predictive control, a 5 0.81 for illusion of crease in Type-I error due to multiple statistical tests (Finner, control, a 5 0.78 inability to stop gambling, and a 5 0.93 1993). The effect size for the mean differences was also for the total scale. measured using Cohen’s-d (effect size was considered low- Impulsive Behavior Scale (UPPS-P) (Whiteside, Lynam, poor |d| > 0.20, moderate-medium for |d| > 0.5, and large- Miller, & Reynolds, 2005). This is a 59-item self-report high for |d| > 0.8) (Kelley & Preacher, 2012). questionnaire developed to assess different domains of The association between the clinical profiles in the study impulsivity: lack of perseverance, lack of premeditation, (cognitive biases, impulsivity and other gambling related sensation seeking, negative urgency, and positive urgency. variables measured) were estimated with Pearson correla- This tool has obtained satisfactory psychometric properties tions (R), stratified by the age groups. Due to the strong in the original version and in the Spanish adaptation (Ver- association between the statistical significance of R-co- dejo-Garcia, Lozano, Moya, Alcazar, & Perez-Garcia, 2010). efficients and the sample size, the correlation effect sizes The internal consistency in the study was a 5 0.75 for lack were established as follow: poor-low |R| > 0.10, moderate- of premeditation, a 5 0.82 for lack of perseverance, medium |R| > 0.24, and large-high |R| > 0.37 (these cut offs a 5 0.85 for sensation seeking, a 5 0.94 for positive ur- corresponded to a Cohen’s-d of 0.20, 0.50 and 0.80, gency, and a 5 0.87 for negative urgency. respectively) (Rosnow & Rosenthal, 1996). Path analysis Other variables. The other variables analyzed in this assessed the magnitude and significance of the relationships study were assessed face-to-face with a semi-structured between the variables of the study with the gambling severity interview, which included socio-demographics (e.g. level, including direct and indirect effects (mediational gender, education, civil status, and employment status), links). This procedure can be used for both exploratory and and other gambling problem related variables (age of confirmatory modeling, and therefore permits theory testing onset and duration of the gambling behaviors, cumulate and theory development (MacCallum & Austin, 2000). This debts due to the gambling behaviors, and bets per analysis was implemented as a case of structural equation gambling/episode). This specific tool has been described modeling (SEM), using the maximum-likelihood estimation elsewhere (Jimenez-Murcia, Aymamı, G omez-Pe~ na, (MLE) method of parameter estimation (Kline, 2005). A  Alvarez-Moya, & Vallejo, 2006). Socioeconomic status latent variable was defined as a measure of the impulsivity was measured with the questionnaire designed by Hol- levels defined by the UPPS-P scores. Goodness-of-fit was lingshead, which generates a position level index based evaluated using standard statistical measures: chi-square test on the education attainment and occupational prestige (c2), the root mean square error of approximation (Hollingshead, 2011). (RMSEA), Bentler’s Comparative Fit Index (CFI), the Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC 388 Journal of Behavioral Addictions 9 (2020) 2, 383-400 Tucker-Lewis Index (TLI), and the standardized root mean Committee of University Hospital of Bellvitge (reference square residual (SRMR). Adequate model fit was considered number PR095/16) in accordance with the Helsinki Decla- for the following criteria (Barrett, 2007): RMSEA < 0.08, TLI ration of 1975 as revised in 1983. > 0.9, CFI > 0.9, and SRMR < 0.1. The global predictive capacity of the model was measured by the coefficient of determination (CD). The variables used in the study as a measurement of the GD severity were the SOGS-total (as RESULTS indicators of the GD symptom level) and the bets per gambling-episode (other alternative measures were not Characteristics of the participants considered due the lack of fit). The first block of Table 1 contains the descriptive parame- ters for the sociodemographics registered in the sample. Most participants were single (41.1%) or lived with a stable Ethics partner (46.9%), achieved primary (61.7%) or secondary Written informed consent was obtained for all the partici- education levels (30.1%), and had a low social index status pants in the study. The work was approved by the Ethics (64.1%). The only differences that emerged between age Table 1. Descriptive for the sample Total Age 18–35 Age 36–45 Age 46–77 (n 5 209) (n 5 73) (n 5 63) (n 5 73) Sociodemographics n % n % n % n % p Marital status Single 86 41.1% 48 65.8% 26 41.3% 12 16.4% <0.001* Married 98 46.9% 22 30.1% 28 44.4% 48 65.8% Divorced 25 12.0% 3 4.1% 9 14.3% 13 17.8% Education Primary 129 61.7% 43 58.9% 35 55.6% 51 69.9% 0.319 Secondary 63 30.1% 23 31.5% 24 38.1% 16 21.9% University 17 8.1% 7 9.6% 4 6.3% 6 8.2% Social index Mean-high to high 4 1.9% 0 0.0% 2 3.2% 2 2.7% 0.469 Mean 20 9.6% 9 12.3% 6 9.5% 5 6.8% Mean-low 51 24.4% 19 26.0% 18 28.6% 14 19.2% Low 134 64.1% 45 61.6% 37 58.7% 52 71.2% Employment Unemployed 67 32.1% 24 32.9% 14 22.2% 29 39.7% 0.091 Employed 142 67.9% 49 67.1% 49 77.8% 44 60.3% Substances use-abuse n % n % n % n % p Tobacco 112 53.6% 41 56.2% 38 60.3% 33 45.2% 0.182 Alcohol 54 25.8% 15 20.5% 19 30.2% 20 27.4% 0.412 Other illegal drugs 22 10.5% 6 8.2% 9 14.3% 7 9.6% 0.490 Prevalence of comorbid disorders n % n % n % n % p a Any other comorbid disorder 62 29.7% 17 23.3% 23 36.5% 22 30.1% 0.241 Depression 14 6.7% 2 2.7% 4 6.3% 8 11.0% 0.138 Anxiety 20 9.6% 3 4.1% 5 7.9% 12 16.4% 0.035* Specific Phobic 1 0.5% 0 0.0% 1 1.6% 0 0.0% 0.312 Social Phobia 1 0.5% 0 0.0% 0 0.0% 1 1.4% 0.392 Compulsive-essive 7 3.3% 1 1.4% 1 1.6% 5 6.8% 0.119 Other 20 9.6% 7 9.6% 7 11.1% 6 8.2% 0.849 Gambling variables Mean SD Mean SD Mean SD Mean SD p Chronological age (years-old) 41.97 13.38 27.94 5.28 40.91 2.84 56.93 7.43 <0.001* Age of onset (years-old) 25.62 10.31 20.62 4.37 24.93 8.86 31.22 12.76 <0.001* Duration of gambling (years) 16.35 12.57 7.32 5.21 15.98 9.39 25.71 13.50 <0.001* DSM-5 total criteria 7.19 1.64 7.12 1.48 7.25 1.67 7.19 1.78 0.899 SOGS total score 14.56 4.29 14.29 4.64 14.90 4.08 14.53 4.15 0.706 Debts due to gambling 4,125 5,636 4,019 5,417 3,786 5,503 4,525 6,006 0.735 Bets mean/episode 67 85 66 81 83 98 54 73 0.131 Bets maximum/episode 587 636 621 637 638 670 508 605 0.417 Gambling preference n % n % n % n % p Non-strategic 149 71.3% 46 63.0% 42 66.7% 61 83.6% 0.014* Strategic 60 28.7% 27 37.0% 21 33.3% 12 16.4% Note. SD: standard deviation. *Bold: significant comparison (0.05). a This variable has been generated to identify the presence of at least one comorbid disorder. Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC Journal of Behavioral Addictions 9 (2020) 2, 383-400 389 groups was in the civil status (the prevalence of single par- Fig. 2), as well as the results of the ANOVA comparing the ticipants was higher among younger age). groups of age (younger, middle and older). As regards the The second and third block of Table 1 contains the cognitive bias severity, a quadratic trend was found for the prevalence of substances use-abuse and the presence of other predictive control, interpretative bias and the total score comorbid disorders. No differences between groups were (line plots in Fig. 1 seem to be similar polynomial functions found, except for the prevalence of anxiety disorder which to curved parabolas): younger and older groups showed the showed a higher frequency in older participants. highest means, while middle-age group showed the lowest As regards gambling variables, age of onset was lower mean. A positive linear trend was obtained for the perceived with the lower the age of the patients, while duration was inability to stop gambling: as the higher the age the higher higher with the older the age. No differences were found was the mean level in the scale. Considering the impulsivity between the groups for the variables measuring the gambling levels, No differences between the groups of age were found severity (gambling symptoms level, debts due to gambling, in the impulsivity levels, except for the sensation seeking and bets per gambling-episode). Differences between the scale, which showed a negative linear trend: the older the age groups also were found for the gambling preferred subtype, the lower was the mean score in this scale. with the prevalence for strategic games being most frequent in younger patients. Contribution of the age on the associations between cognitive and impulsivity with gambling Association between age with cognitive biases and Table 3 contains the correlation matrix with the variables impulsivity levels measuring cognitive bias, impulsivity levels and gambling Table S1 (supplementary material) contains the Pearson severity, stratified by the groups of age. Italic font is used for correlation coefficients and the results of the CurveFit pro- the correlations obtained between scales of the same ques- cedure measuring the linear and non-linear functions be- tionnaire or construct (which tended to be relevant in the tween chronological age (considered on a continuous scale, three groups of the study, excepting for the association be- in years of age) with the GRCS and UPPS-P scales. No tween the GD severity measures). As a whole, the pattern of relevant correlation emerged (all coefficients were in the relationships was different depending on the participants’ poor-low range) and non-significant results were also ob- age. Among younger age patients: a) cognitive bias severity tained for most of the models tested (except for inability to tended to show positive associations with the impulsivity stop gambling [F 5 3.76, df 5 3/205, P 5 0.012] and levels (mainly with the positive and negative urgency scales); sensation seeking [P < 0.005] for many of the polynomial b) higher level in cognitive bias (except for gambling related contrasts). expectancies and illusion of control) positively correlated Table 2 shows the mean scores in the GRCS and UPPS-P with the GD symptom level and the cumulate debts due to scales (see also mean plots in Fig. 1 and radar-chart in the gambling activity; and c) the impulsivity levels did not Table 2. Differences between groups of age on the cognitive biases related to gambling and impulsivity levels Age 18–35 Age 36–45 Age 46–77 Polynomial Pairwise comparisons Y-Young M-Middle O-Old Trends (post-hoc contrasts) (n 5 73) (n 5 63) (n 5 73) LT QT Y vs. M Y vs. O M vs. O Cognitive bias Mean SD Mean SD Mean SD p p p |d| p |d| p |d| (GRCS) Expectancies 12.38 7.20 11.00 7.31 12.40 7.64 0.991 0.213 0.278 0.19 0.991 0.00 0.273 0.19 Illusion of control 7.89 4.80 6.89 4.77 8.55 6.16 0.455 0.098 0.274 0.21 0.455 0.12 0.070 0.30 Predictive control 17.29 8.68 13.89 7.28 17.64 8.82 0.796 0.005* 0.019* 0.42 0.796 0.04 0.009* 0.46 Inability stop 15.78 7.71 16.52 7.77 19.07 9.25 0.018* 0.472 0.603 0.10 0.018* 0.39 0.076 0.30 gambling Interpretive bias 13.51 7.22 11.78 6.53 14.38 7.38 0.455 0.043* 0.157 0.25 0.455 0.12 0.033* 0.37 Total score 66.85 30.12 60.11 27.18 71.99 31.59 0.299 0.040* 0.190 0.23 0.299 0.17 0.021* 0.40 Impulsivity (UPPS- Mean SD Mean SD Mean SD p p p |d| p |d| p |d| P) Lack of 24.01 7.42 24.59 6.97 24.67 8.37 0.604 0.832 0.663 0.08 0.604 0.08 0.949 0.01 premeditation Lack of perseverance 21.49 5.61 22.62 5.19 22.18 6.95 0.491 0.387 0.277 0.21 0.491 0.11 0.670 0.07 Sensation seeking 29.21 7.53 26.03 8.87 25.42 8.17 0.006* 0.299 0.025* 0.39 0.006* 0.51y 0.666 0.07 Positive urgency 31.16 10.95 31.25 11.01 31.70 11.56 0.773 0.916 0.963 0.01 0.773 0.05 0.817 0.04 Negative urgency 31.21 8.35 32.38 7.71 32.75 8.11 0.248 0.742 0.398 0.15 0.248 0.19 0.789 0.05 Note. SD: standard deviation. LT: linear trend. QT: quadratic trend. *Bold: significant comparison (0.05). yBold: effect size into the mean-moderate (|d| > 0.50) to high-large (|d| > 0.80) range. Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC 390 Journal of Behavioral Addictions 9 (2020) 2, 383-400 Fig. 1. Mean plots (n 5 209) relate to the gambling severity. Among older age patients: a) the GD symptom level with cognitive bias or impulsivity cognitive bias also tended to correlate with impulsivity levels levels. And among middle age patients: a) a few number of (especially with lack of premeditation and lack of persever- associations emerged between cognitive bias severity and ance scales); b) regarding cognitive bias severity, gambling- impulsivity levels (and effect size was also lower than cor- related expectancies correlated with the debts due to the relations obtained in the other two groups); b) GD symptom gambling activity, while inability to stop gambling correlated levels correlated with both cognitive bias severity and with the bets per gambling/episode; c) lack of premeditation impulsivity levels (except for lack of premeditation and lack and lack of perseverance related to the cumulate debts due to of perseverance); c) the bets per gambling-episode correlated gambling, and negative urgency correlated with the bets per with the cognitive bias severity in the predictive of control, gambling-episode; and d) no association was found between interpretative bias and total score. Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC Journal of Behavioral Addictions 9 (2020) 2, 383-400 391 Fig. 2. Radar-chart with the z-standardized means based on the groups of age (n 5 209) Path analysis bets per gambling-episode. Age also contributed in the gambling profile: a) a direct effect related younger ages with Figure 3 contains the path-diagram with the standardized strategic gambling preference; b) a mediational link related coefficients obtained in the SEM. This model included the older age with higher likelihood of a comorbid depression- participants’ age (defined as a continuous variable, in years anxiety disorder, which presence increased the GD symptom of age), impulsivity level (a latent variable defined by the level. Civil status also contributed in the model: being un- UPPS-P scores), global cognitive bias level, gambling related married (single, divorced-separated or widowed) increased variables (preferred subtype and measures of severity), and the bets per gambling episode, as well as in the mediational the variables that had achieved a significant relationship link of increasing the likelihood of comorbidity with depres- with the patients’ age (marital status, comorbidity with sion-anxiety and therefore the GD symptom severity. psychiatric disorders and duration of the gambling behavior). Age of onset was not considered a predictor or a potential confounding variable in the SEM since it is generated as the difference between chronological age and DISCUSSION the duration of the problematic gambling, with the conse- quence of lack of fit due to the collinearity among predictors. This study assessed the role of chronological age in the Only significant parameters were retained in the model, underlying mechanism between cognitive biases, impulsivity which was adjusted for the GD duration. The latent variable levels, and preferred forms of gambling with the GD severity defining the impulsivity levels was positively and sig- measures. The results obtained evidenced a higher impairing nificantly defined by all the UPPS-P scales. Adequate cognitive functioning among younger and older patients, goodness-of-fit was obtained: c2 5 71.39 (P 5 0.171), while similar impulsivity levels were achieved by the RMSEA 5 0.029 (95% CI: 0.001–0.053), CFI 5 0.980, different age groups (except for sensation seeking, with TLI 5 0.972 and SRMR 5 0.046. The global predictive ca- higher scores among younger ages). Moreover, path analysis pacity was CD 5 0.726. The results of the SEM indicate that showed a direct contribution of cognitive bias to strategic higher impulsivity levels directly explained the cumulate debts gambling preference, GD symptom level and bets per due to the gambling activity. Higher cognitive bias severity gambling-episode, a direct effect of impulsivity levels on showed a direct effect increasing the likelihood of strategic debts due to the gambling activity, a negative direct effect of gambling preference, higher GD symptom level, and higher age on gambling preference, and a mediational link between Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC Table 3. Association between cognitive biases with impulsivity and gambling severity: Pearson correlation coefficients 392 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Age: 18–35 (n 5 73) 1. GRCS Expectancies 0.48y 0.66y 0.70y 0.64y 0.84y 0.32y 0.36y 0.15 0.32y 0.34y 0.07 0.16 0.12 0.16 0.15 2. GRCS Illusion of control – 0.65y 0.49y 0.54y 0.72y 0.33y 0.22 0.13 0.35y 0.30y 0.16 0.15 0.08 0.11 0.15 3. GRCS Predictive control – 0.68y 0.75y 0.90y 0.09 0.24y 0.05 0.24y 0.28y 0.16 0.22 0.26y 0.12 0.08 4. GRCS Inability stop – 0.67y 0.86y 0.14 0.34y 0.17 0.39y 0.38y 0.22 0.29y 0.09 0.12 0.13 5. GRCS Interpretive bias – 0.86y 0.24y 0.17 0.06 0.35y 0.41y 0.28y 0.27y 0.25y 0.12 0.19 6. GRCS Total score – 0.25y 0.32y 0.13 0.39y 0.41y 0.21 0.27y 0.20 0.15 0.16 7. UPPS-P Premeditation – 0.52y 0.05 0.39y 0.42y 0.15 0.06 0.02 0.08 0.06 8. UPPS-P Perseverance – 0.01 0.35y 0.34y 0.12 0.10 0.03 0.12 0.18 9. UPPS-P Sens.seeking – 0.45y 0.39y 0.05 0.05 0.11 0.15 0.12 10. UPPS-P Posit.urgency – 0.85y 0.20 0.20 0.02 0.00 0.02 11. UPPS-P Negat.urgency – 0.13 0.13 0.04 0.01 0.08 12. DSM-5 criteria for GD – 0.41y 0.06 0.09 0.08 13. SOGS total score – 0.08 0.10 0.01 Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC 14. Debts due to gambling – 0.37y 0.36y 15. Bets mean/episode – 0.50y 16. Bets max/episode – Age: 36–45 (n 5 63) 1. GRCS Expectancies 0.52y 0.73y 0.48y 0.67y 0.86y 0.26y 0.18 0.19 0.22 0.14 0.42y 0.17 0.18 0.11 0.11 2. GRCS Illusion of control – 0.49y 0.20 0.56y 0.63y 0.03 0.08 0.03 0.16 0.14 0.34y 0.00 0.20 0.06 0.12 3. GRCS Predictive control – 0.54y 0.74y 0.89y 0.16 0.15 0.08 0.12 0.12 0.37y 0.09 0.10 0.09 0.34y 4. GRCS Inability stop – 0.56y 0.73y 0.28y 0.42y 0.19 0.17 0.29y 0.27y 0.02 0.14 0.00 0.11 Journal of Behavioral Addictions 9 (2020) 2, 383-400 5. GRCS Interpretive bias – 0.88y 0.16 0.28y 0.15 0.18 0.27y 0.44y 0.06 0.19 0.05 0.31y 6. GRCS Total score – 0.23 0.29y 0.17 0.21 0.24y 0.43y 0.04 0.19 0.00 0.25y 7. UPPS-P Premeditation – 0.67y 0.22 0.21 0.28y 0.07 0.01 0.18 0.09 0.01 8. UPPS-P Perseverance – 0.02 0.25y 0.40y 0.04 0.06 0.21 0.22 0.03 9. UPPS-P Sens.seeking – 0.53y 0.49y 0.29y 0.07 0.20 0.06 0.06 10. UPPS-P Posit.urgency – 0.71y 0.25y 0.00 0.07 0.04 0.07 11. UPPS-P Negat.urgency – 0.25y 0.01 0.10 0.00 0.07 12. DSM-5 criteria for GD – 0.26y 0.04 0.20 0.06 13. SOGS total score – 0.09 0.22 0.36y 14. Debts due to gambling – 0.04 0.04 15. Bets mean/episode – 0.35y 16. Bets max/episode – Age: 46–77 (n 5 73) 1. GRCS Expectancies 0.60y 0.64y 0.41y 0.59y 0.79y 0.43y 0.48y 0.35y 0.33y 0.27y 0.11 0.14 0.24y 0.15 0.23 2. GRCS Illusion of control – 0.69y 0.45y 0.54y 0.79y 0.22 0.24y 0.19 0.12 0.11 0.17 0.04 0.13 0.09 0.08 3. GRCS Predictive control – 0.48y 0.62y 0.85y 0.32y 0.33y 0.22 0.17 0.14 0.04 0.10 0.03 0.04 0.11 4. GRCS Inability stop – 0.62y 0.76y 0.42y 0.29y 0.18 0.18 0.33y 0.07 0.00 0.02 0.14 0.28y 5. GRCS Interpretive bias – 0.84y 0.30y 0.31y 0.18 0.15 0.18 0.15 0.13 0.08 0.02 0.06 6. GRCS Total score – 0.43y 0.41y 0.28y 0.24y 0.27y 0.13 0.10 0.10 0.10 0.20 7. UPPS-P Premeditation – 0.75y 0.25y 0.33y 0.53y 0.07 0.03 0.33y 0.19 0.23 (continued) Journal of Behavioral Addictions 9 (2020) 2, 383-400 393 age and the presence of a comorbid depressive-anxiety dis- 0.13 0.54y 0.15 0.21 0.17 0.23 0.03 0.23 16 – order and GD symptom level. The results obtained in this work showing the re- lationships between high impulsivity and strong cognitive 0.03 0.02 0.29y 0.27y biases with the gambling symptoms level are consistent 0.18 0.04 0.20 15 – with previous evidence reported in scientific literature (Del Prete et al., 2017; Michalczuk, Bowden-Jones, Verdejo- Garcia, & Clark, 2011; Ruiz de Lara, Navas, & Perales, 0.02 y 0.09 0.19 0.21 0.11 0.31 14 – 2019). It has been postulated that impulsivity behavior in decision making tasks could predispose individuals to accept erroneous beliefs without questioning in GD in at- 0.05 0.01 0.36y 0.07 0.06 risk (problem) gambling (Ioannidis et al., 2019). A recent 13 – meta-analysis has also demonstrated the existence of a synergistic relationship of different measures and compo- 0.06 0.09 nents of impulsivity with cognitive distortions related to 0.04 0.07 12 – substance addictions (which seem independent of moder- ator influences), which could be compatible with incentive sensitization theory of addiction processes (Leung et al., y 0.34y 0.66y 0.45 11 – 2017). Our results could suggest an expansion of this the- ory to the behavioral addictions area, particularly to explain the development course of the GD. y 0.51y 0.31 10 But still more relevant: in this study a quadratic trend has – emerged between patient age and gambling related cognitive bias level, while there was no association between impul- 0.22 sivity levels and age, except for sensation seeking (which – 9 adjusted to a negative linear trend). These are novel findings, Table 3. Continued and seemingly contradictory to the evidence previously re- ported in literature. Neurocognitive studies have shown that – 8 age is a key factor for changes in negatively biased infor- mation-processing, and many studies in this area have Note. yBold: effect size into the mean-moderate (|R| > 0.24) to high-large (|R| > 0.37) range. thoroughly identified and described the sharp decline in 7 several cognitive tasks performance with advancing age (Harada, Natelson Love, & Triebel, 2013). Based on these studies, cognitive abilities related to reasoning, memory, and processing performance decline gradually over time, which 6 could make us hypothesized (at least) a positive linear trend between age and cognitive bias (and depending on the processing speed with which cognitive activities performed, 5 also a quadratic trend describing the slowing/rapidity of the processes among the groups of age) (Salthouse, 2010). However, most of the published studies have been inter- 4 preted comparing cognitive bias profiles with standardized age-related trajectories in a typically developing population, which makes it difficult to compare this evidence with the 3 results obtained in our sample (consisting exclusively of GD treatment-seeking patients). And although it is very important to find out what are the cognitive changes of 2 individuals who met clinical criteria for a psychopathological condition compared to the normal process accompanying aging (to evaluate how these unhealthy processes affect daily 12. DSM-5 criteria for GD 11. UPPS-P Negat.urgency 14. Debts due to gambling 10. UPPS-P Posit.urgency functioning, and to identify structural and functional alter- 8. UPPS-P Perseverance 9. UPPS-P Sens.seeking 15. Bets mean/episode ations in brain) (Salthouse, 2012), it is also of increasingly 16. Bets max/episode 13. SOGS total score importance to understand the cognitive changes that accompany aging in samples with mental health conditions. This was a primary novel objective of this work, and our results are also pioneering in this area. As regards the changes in impulsivity levels with age, a similar process can be considered: since it is ‘normal’ for Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC 394 Journal of Behavioral Addictions 9 (2020) 2, 383-400 Fig. 3. Path-diagram with the results of the SEM (n 5 209). Note. Only significant coefficients were retained in the model. Results adjusted by the duration of the GD impulsivity to decline as people grow older, we had hy- the development of reliable/valid assessment tools and pothesized a decrease in the means for the UPPS-P scales effective therapeutic plans, which should consider the when comparing younger, middle and older age. However, particular cognitive style and the impulsivity profiles ex- this pattern was only identified for one of the impulsivity pected within different age groups. components analyzed in the study: sensation seeking scale The results of the SEM in this study indicate that higher reported lower scores as the patient age increased. To un- cognition bias is predictive of higher impairment due to the derstand this apparent discrepant result, it must be realized gambling activity, and it demonstrates the existence of a that our study analyzes a clinical sample of patients treat- mediational link between age, depression-anxiety and the ment-seeking for an addictive disorder (GD) that occurs GD severity. A number of studies have reported that prob- throughout the life cycle, and that is characterized by a lem and disordered gambling showed higher cognitive dis- strong lifespan impulsivity-related pattern. It is therefore not tortions, and perceived experiencing higher levels of negative surprising that impulsivity becomes decreasingly less affective states than recreational gambling and control frequent with aging among patients with problem gambling, samples (Tang & Oei, 2011; Yang, Tang, Gu, Luo, & Luo, and that poor impulsive control should be reported inde- 2015). Studies have also reported that the poorer perfor- pendent of the chronological age. This result outlines the mance in decision-making tasks correlates with the higher relevance of impulsivity in all GD groups, including younger gambling severity (Cosenza, Ciccarelli, & Nigro, 2019). Our and older groups. study contributes in this area, since it shows that, although On the other hand, the pattern of associations between gambling severity might seem independent of the chrono- cognitive distortions and impulsivity with the gambling logical age in clinical samples of GD treatment-seeking pa- severity obtained in this work was moderated by patient age. tients, a mediational/indirect effect is observed through the Most studies evaluating the contribution of impulsivity and presence of a comorbid depressive-anxiety disorder: older cognitive bias on the gambling related behaviors have ob- subjects are particularly vulnerable to increase the gambling tained strong associations, independent of age. But again, impairment if they have a comorbid mental problem. most of these studies only measured these constructs at a Moreover, according to the SEM obtained in this study single point-in-time, and therefore the links between the sample, civil status was also a variable to be considered in manifestation of impulsivity and cognitive performance in this mechanism: being unmarried (single, divorced or wid- GD patients of different age groups was unclear. Our results owed) not only shows a direct effect on the bets per outline that although impulsivity and cognitive bias should gambling-episode, but also increases the risk of a comorbid be (strongly) related with GD and problem/at-risk gambling, depressive-anxiety problems and, therefore, is an indirect age seems to be moderating the frequency and effect size of variable explaining the GD severity. This result is consis- these relationships. This evidence may have implications for tent with previous etiological studies focused on the Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC Journal of Behavioral Addictions 9 (2020) 2, 383-400 395 identification of the variables related with the onset of the Limitations, strengths and implications GD and the severity of the gambling behaviors, which outline that having a stable partner is a preventive factor Three main limitations should be taken into account when (Elton-Marshall et al., 2018). Furthermore, the global results interpreting the results of this work. Firstly, because the of the SEM add the novel evidence that older age and being patients in our sample are all men, extrapolation of the re- unmarried increase the odds of depressed-anxious problems, sults to women is not possible. Secondly, the data analyzed and that this is precisely a profile highly vulnerable to pre- are from a cross-sectional study, which does not allow the sent with the most severe consequences of the GD. It must longitudinal changes in the measures to be determined (a be noted that older adults usually experience financial dif- repeated measures design should allow developmental tra- ficulties or insecurity, loss of physical ability (particularly jectories of individual facets of cognitive bias and impulsivity mobility), as well as a loss of relationships with relatives and to be examined, as well as its relationship with the GD friends. Furthermore, subjects are likely to lose their partner profile). And thirdly, although a number of constructs during later life, and since unpartnered older adults tend to strongly related to the GD severity and the gambling pref- be more socially isolated (with the consequence of increasing erence were analyzed in the study (age, cognitive biases, depression and anxiety levels), the risk of severe problematic impulsivity, sociodemographic features and comorbid psy- gambling due to loneliness and its consequences is pre- chopathological state), other variables that could also be dictable (Botterill, Gill, McLaren, & Gomez, 2016; Parke, involved in the GD mechanism were not addressed. Griffiths, Pattinson, & Keatley, 2018). Some strengths should also be noted. The sample of this Path analysis in this study showed a direct contribution study included patients from a large range of ages (between of cognitive distortions level on the likelihood of strategic 18 to 77 years-old), and therefore results may be extrapo- gambling preference, which is consistent with other pre- lated to a substantial proportion of GD treatment-seeking vious empirical studies (Levesque, Sevigny, Giroux, & Jac- patients. The multivariate analysis of two constructs strongly ques, 2017; Navas et al., 2017). However, our SEM did not related to the GD (cognitive bias and age) is another relevant retain a direct association between the impulsivity levels strength, since it allows obtaining the specific contribution and the gambling subtype, as was suggested by previous of each domain in the gambling profile, as well as the po- studies that concluded that higher levels of impulsivity tential moderator role of the patients’ age. behavior may increase likelihood of strategic gambling, as This study has clinically relevant implications in the area well as a higher risk of accumulating losses (Lorains, of the development of assessment tools with high reliability Dowling, et al., 2014a; Lorains, Stout, et al., 2014b; to identify the core components of the GD, as well as for the Worhunsky, Potenza, & Rogers, 2017). It must be noted, advance in effective and precise therapeutic programs however, that our study includes simultaneously the focused on the specific characteristics of problematic and cognitive bias severity and the impulsivity levels in the path disordered gambling patients. analysis, which allows obtaining the specific contribution of each construct on the gambling type. Indeed, the path in Authors’ contribution: Conceptualization, Susana Jimenez- our work show a correlation between impulsivity and Murcia; Data curation, Roser Granero and Susana Valero disturbed thoughts (higher impulsivity levels are related to Solis; Funding acquisition, Susana Jimenez-Murcia worse cognitive reasoning performance), and therefore it and Fernando Fernandez-Aranda; Investigation, Amparo would also be possible to consider both a direct effect of del Pino-Gutierrez, Gemma Mestre-Bach, Isabel Baenas, cognitive bias on the gambling preference, and also that S.Fabrizio Contaldo, Monica Gomez-Pe~ na, Neus Aymamı, cognitive style could act as a mediational variable in the Laura Moragas, Cristina Vintro, Teresa Mena-Moreno, relationship between the impulsivity levels and the Eduardo Valenciano-Mendoza and Bernat Mora-Maltas; preferred gambling style. On the other hand, considering Project administration, Jose M. Menchon and Susana that strategic gambling in the sample of this work was Jimenez-Murcia; Supervision, Susana Jimenez-Murcia; directly related to younger age, the results of the SEM seem Writing – original draft, Roser Granero and Susana Jimenez- consistent with a new phenotype described in the GD Murcia; Writing – review & editing, Fernando Fernandez- profile characterized by strong gambling-related cognitive Aranda and Roser Granero. distortions, young age, and the preference for skill-based games (Mallorquı-Bague, Mestre-Bach et al., 2018a; Conflict of interests: The authors declare no conflict of in- Mallorquı-Bague, Tolosa-Sola et al., 2018b; Myrseth, terest. The funders had no role in the design of the study; in Brunborg, & Eidem, 2010; Perales, Navas, Ruiz de Lara, the collection, analyses, or interpretation of data; in the writing Maldonado, & Catena, 2017). Evidence published to date of the manuscript; or in the decision to publish the results. suggest that this phenotype may be explained by height- ened sensitivity to the rewarding features of gambling ac- Acknowledgments and funding sources: We thank CERCA tivities, greater disinhibition and sensation seeking, and Programme/Generalitat de Catalunya for institutional sup- therefore, with a higher risk for progression and escalating port. This manuscript and research was supported by grants the disordered gambling, and could be even related to the from the Ministerio de Economıa y Competitividad new forms of gambling activities (such as online gambling) (PSI2015-68701-R), Research funded by the Delegaci on del (Jimenez-Murcia et al., 2019). Gobierno para el Plan Nacional sobre Drogas (2017I067 and Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC 396 Journal of Behavioral Addictions 9 (2020) 2, 383-400 2019I47), Instituto de Salud Carlos III (ISCIII) (FIS PI14/ The mediating role of loneliness. Journal of Gambling Studies, 00290 and PI17/01167) and co-funded by FEDER funds/ 32(3), 1027–1038. https://doi.org/10.1007/s10899-015-9575-5. European Regional Development Fund (ERDF), a way to Brandt, L., & Fischer, G. (2019). Adult ADHD is associated with build Europe. CIBERobn and CIBERSAM are both initia- gambling severity and psychiatric comorbidity among tives of ISCIII. The funders had no role in the study design, treatment-seeking problem gamblers. Journal of Attention data collection and analysis, decision to publish, or prepa- Disorders, 23(12), 1383–1395. https://doi.org/10.1177/ ration of the manuscript. TMM and CVA are supported by a 1087054717690232. predoctoral Grant of the Ministerio de Educaci on, Cultura y Bruine de Bruin, W., Parker, A. M., & Fischhoff, B. (2012). Deporte (FPU16/02087; FPU16/01453). With the support of Explaining adult age differences in decision-making compe- the Secretariat for Universities and Research of the Ministry tence. Journal of Behavioral Decision Making, 25(4), 352–360. of Business and Knowledge of the Government of Catalonia. https://doi.org/10.1002/bdm.712. Calado, F., & Griffiths, M. D. (2016). Problem gambling worldwide: An update and systematic review of empirical research (2000– APPENDIX A. SUPPLEMENTARY DATA 2015). Journal of Behavioral Addictions, 5(4), 592–613. https:// doi.org/10.1556/2006.5.2016.073. Supplementary data to this article can be found online at Canale, N., Vieno, A., Bowden-Jones, H., & Billieux, J. (2017, https://doi.org/10.1556/2006.2020.00028. February). The benefits of using the UPPS model of impulsivity rather than the Big Five when assessing the relationship be- tween personality and problem gambling. Addiction (Abingdon, England), Vol. 112, pp. 372–373. https://doi.org/10.1111/add. REFERENCES 13641. Chamberlain, S. R., Derbyshire, K., Leppink, E., & Grant, J. E. Abouzari, M., Oberg, S., Gruber, A., & Tata, M. (2015). Interactions (2015). Impact of ADHD symptoms on clinical and cognitive among attention-deficit hyperactivity disorder (ADHD) and aspects of problem gambling. Comprehensive Psychiatry, 57, problem gambling in a probabilistic reward-learning task. 51–57. https://doi.org/10.1016/j.comppsych.2014.10.013. Behavioural Brain Research, 291, 237–243. https://doi.org/10. Chamberlain, S. R., Stochl, J., Redden, S. A., & Grant, J. E. (2018). 1016/j.bbr.2015.05.041. Latent traits of impulsivity and compulsivity: Toward dimen- American Psychiatric Association. (2010). Diagnostic and statistical sional psychiatry. Psychological Medicine, 48(5), 810–821. manual of mental disorders (4th ed.). Washington, DC: https://doi.org/10.1017/S0033291717002185. American Psychiatric Association. Chamberlain, S. R., Stochl, J., Redden, S. A., Odlaug, B. L., & Grant, American Psychiatric Association. (2013). Diagnostic and statistical J. E. (2017). Latent class analysis of gambling subtypes and manual of mental disorders (5th ed.). Washington DC: Author. impulsive/compulsive associations: Time to rethink diagnostic Bangma, D. F., Fuermaier, A. B. M., Tucha, L., Tucha, O., & Koerts, boundaries for gambling disorder? Addictive Behaviors, 72, J. (2017). The effects of normal aging on multiple aspects of 79–85. https://doi.org/10.1016/j.addbeh.2017.03.020. financial decision-making. PLoS One, 12(8), e0182620. https:// Chretien, M., Giroux, I., Goulet, A., Jacques, C., & Bouchard, S. doi.org/10.1371/journal.pone.0182620. (2017). Cognitive restructuring of gambling-related thoughts: A Barrett, P. (2007). Structural equation modelling: Adjudging model systematic review. Addictive Behaviors, 75, 108–121. https://doi. fit. Personality and Individual Differences, 42(5), 815–824. org/10.1016/j.addbeh.2017.07.001. https://doi.org/10.1016/j.paid.2006.09.018. Clark, L. (2010). Decision-making during gambling: An integration Biederman, J., Fried, R., Petty, C., Mahoney, L., & Faraone, S. V. of cognitive and psychobiological approaches. Philosophical (2012). An examination of the impact of attention-deficit hy- Transactions of the Royal Society of London. Series B, Biological peractivity disorder on IQ: A large controlled family-based Sciences, 365(1538), 319–330. https://doi.org/10.1098/rstb.2009. analysis. Canadian Journal of Psychiatry. Revue Canadienne 0147. de Psychiatrie, 57(10), 608–616. https://doi.org/10.1177/ Clark, L., & Limbrick-Oldfield, E. H. (2013). Disordered gambling: 070674371205701005. A behavioral addiction. Current Opinion in Neurobiology, 23(4), Black, D. W., Coryell, W. H., Crowe, R. R., Shaw, M., McCormick, 655–659. https://doi.org/10.1016/j.conb.2013.01.004. B., & Allen, J. (2015). Personality disorders, impulsiveness, and Cosenza, M., Ciccarelli, M., & Nigro, G. (2019). Decision-making novelty seeking in persons with DSM-IV pathological gambling styles, negative affectivity, and cognitive distortions in adoles- and their first-degree relatives. Journal of Gambling Studies, cent gambling. Journal of Gambling Studies, 35(2), 517–531. 31(4), 1201–1214. https://doi.org/10.1007/s10899-014-9505-y. https://doi.org/10.1007/s10899-018-9790-y. Bonnaire, C., Kovess-Masfety, V., Guignard, R., Richard, J. B., du Del Prete, F., Steward, T., Navas, J. F., Fernandez-Aranda, F., Rosco€at, E., & Beck, F. (2017). Gambling type, substance abuse, Jimenez-Murcia, S., Oei, T. P. S., et al. (2017). The role of affect- health and psychosocial correlates of male and female problem driven impulsivity in gambling cognitions: A convenience- gamblers in a nationally representative French sample. Journal sample study with a Spanish version of the gambling-related of Gambling Studies, 33(2), 343–369. https://doi.org/10.1007/ cognitions scale. Journal of Behavioral Addictions, 6(1), 51–63. s10899-016-9628-4. https://doi.org/10.1556/2006.6.2017.001. Botterill, E., Gill, P. R., McLaren, S., & Gomez, R. (2016). Marital Dussault, F., Brendgen, M., Vitaro, F., Wanner, B., & Tremblay, R. E. status and problem gambling among Australian older adults: (2011). Longitudinal links between impulsivity, gambling Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC Journal of Behavioral Addictions 9 (2020) 2, 383-400 397 problems and depressive symptoms: A transactional model gambling: A meta-analysis. Neuropsychopharmacology: Official from adolescence to early adulthood. The Journal of Child Publication of the American College of Neuropsycho- Psychology and Psychiatry and Allied Disciplines, 52(2), 130– pharmacology, 44(8), 1354–1361. https://doi.org/10.1038/ 138. https://doi.org/10.1111/j.1469-7610.2010.02313.x. s41386-019-0393-9. Echebur ua, E., Baez, C., Fernandez, J., & Paez, D. (1994). Cues- Jacob, L., Haro, J. M., & Koyanagi, A. (2018). Relationship be- tionario de juego patologico de South Oaks (SOGS): Validaci on tween attention-deficit hyperactivity disorder symptoms and espa~nola (South Oaks Gambling Screen (SOGS): Spanish vali- problem gambling: A mediation analysis of influential factors dation). Analisis y Modificacion de Conducta, 20, 769–791. among 7,403 individuals from the UK. Journal of Behavioral Elton-Marshall, T., Wijesingha, R., Sendzik, T., Mock, S. E., van der Addictions, 7(3), 781–791. https://doi.org/10.1556/2006.7. Maas, M., McCready, J., et al. (2018). Marital status and 2018.72. problem gambling among older adults: An examination of so- Jimenez-Murcia, S., Aymamı, M. N., G omez-Pe~  na, M., Alvarez- cial context and social motivations. Canadian Journal on Aging, Moya, E. M., & Vallejo, J. (2006). Protocols de tractament 37(3), 318–332. https://doi.org/10.1017/S071498081800017X. cognitiuconductual pel joc patologic i d’altres addicions no Emond, M. S., & Marmurek, H. H. C. (2010). Gambling related toxiques [Guidelines of cognitive-behavioral treatment of path- cognitions mediate the association between thinking style and ological gambling and other non-toxic addicitions]. Barcelona, problem gambling severity. Journal of Gambling Studies, 26(2), Spain: Hospital Universitari de Bellvitge, Departament de Salut, 257–267. https://doi.org/10.1007/s10899-009-9164-6. Generalitat de Catalunya. Fatseas, M., Alexandre, J.M., Venisse, J.L., Romo, L., Valleur, M., Jimenez-Murcia, S., Granero, R., Fernandez-Aranda, F., & Magalon, D., et al. (2016). Gambling behaviors and psycho- Mench on, J. M. (2020). Comparison of gambling profiles based pathology related to Attention-Deficit/Hyperactivity Disorder on strategic versus non-strategic preferences. Current Opinion (ADHD) in problem and non-problem adult gamblers. Psy- in Behavioral Sciences, 31, 13–20. https://doi.org/10.1016/j. chiatry Research, 239, 232–238. https://doi.org/10.1016/j. cobeha.2019.09.001. psychres.2016.03.028. Jimenez-Murcia, S., Granero, R., Fernandez-Aranda, F., Finner, H. (1993). On a monotonicity problem in step-down Stinchfield, R., Tremblay, J., Steward, T., et al. (2019). Phe- multiple test procedures. Journal of the American Statistical notypes in gambling disorder using sociodemographic and Association, 88, 920–923. https://doi.org/10.1080/01621459. clinical clustering analysis: An unidentified new subtype? 1993.10476358. Frontiers in Psychiatry, 10, 173. https://doi.org/10.3389/fpsyt. Giralt, S., Muller, K. W., Beutel, M. E., Dreier, M., Duven, E., & 2019.00173. Wolfling, K. (2018). Prevalence, risk factors, and psychosocial Jimenez-Murcia, S., Stinchfield, R., Alvarez-Moya, E., Jaurrieta, N., adjustment of problematic gambling in adolescents: Results Bueno, B., Granero, R., et al. (2009). Reliability, validity, and from two representative German samples. Journal of Behavioral classification accuracy of a Spanish translation of a measure of Addictions, 7(2), 339–347. https://doi.org/10.1556/2006.7.2018. DSM-IV diagnostic criteria for pathological gambling. Journal 37. of Gambling Studies, 25(1), 93–104. https://doi.org/10.1007/ Goodie, A. S., & Fortune, E. E. (2013). Measuring cognitive dis- s10899-008-9104-x. tortions in pathological gambling: Review and meta-analyses. Kastirke, N., Rumpf, H.J., John, U., Bischof, A., & Meyer, C. (2015). Psychology of Addictive Behaviors: Journal of the Society of Demographic risk factors and gambling preference may not Psychologists in Addictive Behaviors, 27(3), 730–743. https://doi. explain the high prevalence of gambling problems among the org/10.1037/a0031892. population with migration background: Results from a German Grall-Bronnec, M., Wainstein, L., Feuillet, F., Bouju, G., Rocher, B., nationwide survey. Journal of Gambling Studies, 31(3), 741–757. Venisse, J.L., et al. (2012). Clinical profiles as a function of level https://doi.org/10.1007/s10899-014-9459-0. and type of impulsivity in a sample group of at-risk and Kelley, K., & Preacher, K. J. (2012). On effect size. Psychological pathological gamblers seeking treatment. Journal of Gambling Methods, 17(2), 137–152. https://doi.org/10.1037/a0028086. Studies, 28(2), 239–252. https://doi.org/10.1007/s10899-011- Kline, R. B. (2005). Principles and practice of structural equation 9258-9. modeling. In Methodology in the social sciences. (2nd ed.). New Gullo, M. J., Loxton, N. J., & Dawe, S. (2014). Impulsivity: Four ways York: The Guilford Press. five factors are not basic to addiction. Addictive Behaviors, Kovacs, I., Demeter, I., Janka, Z., Demetrovics, Z., Maraz, A., & 39(11), 1547–1556. https://doi.org/10.1016/j.addbeh.2014.01.002. Ando, B. (2020). Different aspects of impulsivity in chronic Harada, C. N., Natelson Love, M. C., & Triebel, K. L. (2013). alcohol use disorder with and without comorbid problem Normal cognitive aging. Clinics in Geriatric Medicine, 29(4), gambling. PloS One, 15(1), e0227645. https://doi.org/10.1371/ 737–752. https://doi.org/10.1016/j.cger.2013.07.002. journal.pone.0227645. Hodgins, D. C., & Holub, A. (2015). Components of impulsivity in Kovacs, I., Richman, M. J., Janka, Z., Maraz, A., & Ando, B. (2017). gambling disorder. International Journal of Mental Health and Decision making measured by the Iowa gambling task in Addiction, 13(6), 699–711. https://doi.org/10.1007/s11469-015- alcohol use disorder and gambling disorder: A systematic re- 9572-z. view and meta-analysis. Drug and Alcohol Dependence, 181, Hollingshead, A. B. (2011). Four factor index of social status. Yale 152–161. https://doi.org/10.1016/j.drugalcdep.2017.09.023. Journal of Sociology, 8, 21–51. Kr€aplin, A., Buhringer, G., Oosterlaan, J., van den Brink, W., Ioannidis, K., Hook, R., Wickham, K., Grant, J. E., & Chamberlain, Goschke, T., & Goudriaan, A. E. (2014). Dimensions and dis- S. R. (2019). Impulsivity in gambling disorder and problem order specificity of impulsivity in pathological gambling. Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC 398 Journal of Behavioral Addictions 9 (2020) 2, 383-400 Addictive Behaviors, 39(11), 1646–1651. https://doi.org/10. Psychopharmacology, 233(18), 3361–3370. https://doi.org/10. 1016/j.addbeh.2014.05.021. 1007/s00213-016-4372-0. Kyonka, E. G. E., & Schutte, N. S. (2018). Probability discounting Maclaren, V. V, Fugelsang, J. A., Harrigan, K. A., & Dixon, M. J. and gambling: A meta-analysis. Addiction (Abingdon, (2011). The personality of pathological gamblers: A meta- England), 113(12), 2173–2181. https://doi.org/10.1111/add. analysis. Clinical Psychology Review, 31(6), 1057–1067. https:// 14397. doi.org/10.1016/j.cpr.2011.02.002. Ledgerwood, D. M., & Petry, N. M. (2010). Subtyping pathological Mallorquı-Bague, N., Mestre-Bach, G., Lozano-Madrid, M., gamblers based on impulsivity, depression, and anxiety. Psy- Fernandez-Aranda, F., Granero, R., Vintr o-Alcazaz, C., et al. chology of Addictive Behaviors: Journal of the Society of Psy- (2018a). Trait impulsivity and cognitive domains involving chologists in Addictive Behaviors, 24(4), 680–688. https://doi. impulsivity and compulsivity as predictors of gambling disorder org/10.1037/a0019906. treatment response. Addictive Behaviors, 87, 169–176. https:// Lee, R. S. C., Hoppenbrouwers, S., & Franken, I. (2019). A sys- doi.org/10.1016/j.addbeh.2018.07.006. tematic meta-review of impulsivity and compulsivity in addic- Mallorquı-Bague, N., Tolosa-Sola, I., Fernandez-Aranda, F., Granero, tive behaviors. Neuropsychology Review, 29(1), 14–26. https:// R., Fagundo, A. B., Lozano-Madrid, M., et al. (2018b). Cognitive doi.org/10.1007/s11065-019-09402-x. deficits in executive functions and decision-making impairments Leonard, C. A., & Williams, R. J. (2016). The relationship between cluster gambling disorder sub-types. Journal of Gambling Studies, gambling fallacies and problem gambling. Psychology of 34(1). https://doi.org/10.1007/s10899-017-9724-0. Addictive Behaviors: Journal of the Society of Psychologists in Mallorqui-Bague, N., Vintro-Alcaraz, C., Verdejo-Garcia, A., Addictive Behaviors, 30(6), 694–704. https://doi.org/10.1037/ Granero, R., Fernandez-Aranda, F., Magana, P., et al. (2019). adb0000189. Impulsivity and cognitive distortions in different clinical phe- Lesieur, H. R, & Blume, S. B. (1987). The south Oaks gambling notypes of gambling disorder: Profiles and longitudinal pre- screen (SOGS): A new instrument for the identification of diction of treatment outcomes. European Psychiatry: The pathological gamblers. American Journal of Psychiatry, 144(9), Journal of the Association of European Psychiatrists, 61, 9–16. 1184–1188. https://doi.org/10.1176/ajp.144.9.1184. https://doi.org/10.1016/j.eurpsy.2019.06.006. Lesieur, H. R., & Blume, S. B. (1993). Revising the south Oaks Michalczuk, R., Bowden-Jones, H., Verdejo-Garcia, A., & Clark, L. gambling screen in different settings. Journal of Gambling (2011). Impulsivity and cognitive distortions in pathological Studies, 9, 213–223. https://doi.org/10.1007/BF01015919. gamblers attending the UK national problem gambling clinic: A Leung, D., Staiger, P. K., Hayden, M., Lum, J. A. G., Hall, K., preliminary report. Psychological Medicine, 41(12), 2625–2635. Manning, V., et al. (2017). Meta-analysis of the relationship https://doi.org/10.1017/S003329171100095X. between impulsivity and substance-related cognitive biases. Miquel, S., Champ, C., Day, J., Aarts, E., Bahr, B. A., Bakker, M., et al. Drug and Alcohol Dependence, 172, 21–33. https://doi.org/10. (2018). Poor cognitive ageing: Vulnerabilities, mechanisms and 1016/j.drugalcdep.2016.11.034. the impact of nutritional interventions. Ageing Research Reviews, Levesque, D., Sevigny, S., Giroux, I., & Jacques, C. (2017). 42, 40–55. https://doi.org/10.1016/j.arr.2017.12.004. Gambling-related cognition scale (GRCS): Are skills-based Mitchell, M. R., & Potenza, M. N. (2014). Addictions and per- games at a disadvantage? Psychology of Addictive Behaviors: sonality traits: Impulsivity and related constructs. Current Journal of the Society of Psychologists in Addictive Behaviors, Behavioral Neuroscience Reports, 1(1), 1–12. https://doi.org/10. 31(6), 647–654. https://doi.org/10.1037/adb0000297. 1007/s40473-013-0001-y. Levesque, D., Sevigny, S., Giroux, I., & Jacques, C. (2018). Psy- Moore, S. M., Thomas, A. C., Kyrios, M., Bates, G., & Meredyth, D. chological vulnerability and problem gambling: The media- (2011). Gambling accessibility: A scale to measure gambler tional role of cognitive distortions. Journal of Gambling Studies, preferences. Journal of Gambling Studies, 27(1), 129–143. 34(3), 807–822. https://doi.org/10.1007/s10899-017-9740-0. https://doi.org/10.1007/s10899-010-9203-3. Lorains, F. K., Dowling, N. A., Enticott, P. G., Bradshaw, J. L., Moragas, L., Granero, R., Stinchfield, R., Fernandez-Aranda, F., Trueblood, J. S., & Stout, J. C. (2014a). Strategic and non- Fr€oberg, F., Aymamı, N., et al. (2015). Comparative analysis of strategic problem gamblers differ on decision-making under distinct phenotypes in gambling disorder based on gambling risk and ambiguity. Addiction, 109(7), 1128–1137. https://doi. preferences. BMC Psychiatry, 15(1), 86. https://doi.org/10.1186/ org/10.1111/add.12494. s12888-015-0459-0. Lorains, F. K., Stout, J. C., Bradshaw, J. L., Dowling, N. A., & Mouneyrac, A., Lemercier, C., Le Floch, V., Challet-Bouju, G., Mor- Enticott, P. G. (2014b). Self-reported impulsivity and inhibitory eau, A., Jacques, C., et al. (2018). Cognitive characteristics of control in problem gamblers. Journal of Clinical and Experi- strategic and non-strategic gamblers. Journal of Gambling Studies, mental Neuropsychology, 36(2), 144–157. https://doi.org/10. 34(1), 199–208. https://doi.org/10.1007/s10899-017-9710-6. 1080/13803395.2013.873773. Myles, D., Carter, A., & Yucel, M. (2019). Cognitive neuroscience MacCallum, R. C., & Austin, J. T. (2000). Applications of structural can support public health approaches to minimise the harm of equation modeling in psychological research. Annual Review of “losses disguised as wins” in multiline slot machines. European Psychology, 51(1), 201–226. https://doi.org/10.1146/annurev. Journal of Neuroscience, 50(3), 2384–2391. https://doi.org/10. psych.51.1.201. 1111/ejn.14191. MacKillop, J., Weafer, J., Gray, J., Oshri, A., Palmer, A., & de Wit, Myrseth, H., Brunborg, G. S., & Eidem, M. (2010). Differences H. (2016). The latent structure of impulsivity: Impulsive choice, in cognitive distortions between pathological and non-patho- impulsive action, and impulsive personality traits. logical gamblers with preferences for chance or skill games. Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC Journal of Behavioral Addictions 9 (2020) 2, 383-400 399 Journal of Gambling Studies, 26(4), 561–569. https://doi.org/10. Sharma, L., Markon, K. E., & Clark, L. A. (2014). Toward a theory 1007/s10899-010-9180-6. of distinct types of “impulsive” behaviors: A meta-analysis of Navas, J. F., Billieux, J., Perandres-Gomez, A., Lopez-Torrecillas, F., self-report and behavioral measures. Psychological Bulletin, Candido, A., & Perales, J. C. (2017). Impulsivity traits and 140(2), 374–408. https://doi.org/10.1037/a0034418. gambling cognitions associated with gambling preferences and Stata-Corp. (2019). Stata statistical software: Release 16. College clinical status. International Gambling Studies, 17, 102–124. Station, TX: StataCorp LLC. https://doi.org/10.1080/14459795.2016.1275739. Stevens, M., & Young, M. (2010). Who plays what? Participation Odlaug, B. L., Marsh, P. J., Kim, S. W., & Grant, J. E. (2011). profiles in chance versus skill-based gambling. Journal of Strategic vs nonstrategic gambling: Characteristics of patho- Gambling Studies, 26(1), 89–103. https://doi.org/10.1007/ logical gamblers based on gambling preference. Annals of s10899-009-9143-y. Clinical Psychiatry, 23(2), 105–112. Retrieved from http://www. Steward, T., Mestre-Bach, G., Fernandez-Aranda, F., Granero, R., ncbi.nlm.nih.gov/pubmed/21547270. Perales, J. C., Navas, J. F., et al. (2017). Delay discounting and Paolini, D., Leonardi, C., Visani, E., & Rodofili, G. (2018). The impulsivity traits in young and older gambling disorder pa- gambling disorder: Family styles and cognitive dimensions. Eu- tients. Addictive Behaviors, 71, 96–103. https://doi.org/10.1016/ ropean Review for Medical and Pharmacological Sciences, 22(4), j.addbeh.2017.03.001. 1066–1070. https://doi.org/10.26355/eurrev_201802_14390. Stinchfield, R. (2003). Reliability, validity, and classification accu- Parke, A., Griffiths, M., Pattinson, J., & Keatley, D. (2018). Age- racy of a measure of DSM-IV diagnostic criteria for patholog- related physical and psychological vulnerability as pathways to ical gambling. American Journal of Psychiatry, 160, 180–182. problem gambling in older adults. Journal of Behavioral Ad- https://doi.org/10.1176/appi.ajp.160.1.180. dictions, 7(1), 137–145. https://doi.org/10.1556/2006.7.2018.18. Subramaniam, M., Abdin, E., Shahwan, S., Vaingankar, J. A., Perales, J. C., Navas, J. F., Ruiz de Lara, C. M., Maldonado, A., & Picco, L., Browning, C. J., et al. (2015). Culture and age in- Catena, A. (2017). Causal learning in gambling disorder: fluences upon gambling and problem gambling. Addictive Beyond the illusion of control. Journal of Gambling Studies, Behaviors Reports, 1, 57–63. https://doi.org/10.1016/j.abrep. 33(2), 705–717. https://doi.org/10.1007/s10899-016-9634-6. 2015.04.004. Rai, D., Hall, W., Bebbington, P., Skapinakis, P., Hassiotis, A., Subramaniam, M., Abdin, E., Vaingankar, J. A., Shahwan, S., Picco, L., Weich, S., et al. (2014). Estimated verbal IQ and the odds & Chong, S. A. (2016). Strategic versus nonstrategic of problem gambling: A population-based study. Psychological gambling: Results from a community survey. Journal of Medicine, 44(8), 1739–1749. https://doi.org/10.1017/ Addiction Medicine, 10(3), 174–181. https://doi.org/10.1097/ S0033291713002195. ADM.0000000000000211. Raylu, N., & Oei, T. P. S. (2004). The gambling related cognitions Suissa, A. J. (2015). Cyber addictions: Toward a psychosocial scale (GRCS): Development, confirmatory factor validation and perspective. Addictive Behaviors, 43, 28–32. https://doi.org/10. psychometric properties. Addiction, 99(6), 757–769. https://doi. 1016/j.addbeh.2014.09.020. org/10.1111/j.1360-0443.2004.00753.x. Suomi, A., Dowling, N. A., & Jackson, A. C. (2014). Problem Rochat, L., Billieux, J., Gagnon, J., & Van der Linden, M. (2018). A gambling subtypes based on psychological distress, alcohol multifactorial and integrative approach to impulsivity in abuse and impulsivity. Addictive Behaviors, 39(12), 1741–1745. neuropsychology: Insights from the UPPS model of impulsivity. https://doi.org/10.1016/j.addbeh.2014.07.023. Journal of Clinical and Experimental Neuropsychology, 40(1), Tang, C. S.K., & Oei, T. P. (2011). Gambling cognition and sub- 45–61. https://doi.org/10.1080/13803395.2017.1313393. jective well-being as mediators between perceived stress and Rosnow, R. L., & Rosenthal, R. (1996). Computing contrasts, effect problem gambling: A cross-cultural study on white and Chinese sizes, and counternulls on other people’s published data: Gen- problem gamblers. Psychology of Addictive Behaviors: Journal of eral procedures for research consumers. Psychological Methods, the Society of Psychologists in Addictive Behaviors, 25(3), 511– 1(4), 331–340. https://doi.org/10.1037/1082-989X.1.4.331. 520. https://doi.org/10.1037/a0024013. Ruiz de Lara, C. M., Navas, J. F., & Perales, J. C. (2019). The Tiego, J., Oostermeijer, S., Prochazkova, L., Parkes, L., Dawson, A., paradoxical relationship between emotion regulation and Youssef, G., et al. (2019). Overlapping dimensional phenotypes gambling-related cognitive biases. PloS One, 14(8), e0220668. of impulsivity and compulsivity explain co-occurrence of https://doi.org/10.1371/journal.pone.0220668. addictive and related behaviors. CNS Spectrums, 24(4), 426– Salthouse, T. A. (2010). Influence of age on practice effects in 440. https://doi.org/10.1017/S1092852918001244. longitudinal neurocognitive change. Neuropsychology, 24(5), Tse, S., Hong, S.I., Wang, C.W., & Cunningham-Williams, R. M. 563–572. https://doi.org/10.1037/a0019026. (2012). Gambling behavior and problems among older adults: Salthouse, T. A. (2012). Consequences of age-related cognitive A systematic review of empirical studies. Journals of Geron- declines. Annual Review of Psychology, 63, 201–226. https://doi. tology Series B: Psychological Sciences and Social Sciences, 67(5), org/10.1146/annurev-psych-120710-100328. 639–652. https://doi.org/10.1093/geronb/gbs068. Schiebener, J., & Brand, M. (2017). Age-related variance in de- van Holst, R. J., van den Brink, W., Veltman, D. J., & Goudriaan, A. E. cisions under ambiguity is explained by changes in reasoning, (2010). Why gamblers fail to win: A review of cognitive and executive functions, and decision-making under risk. Cognition neuroimaging findings in pathological gambling. Neuroscience & & Emotion, 31(4), 816–824. https://doi.org/10.1080/02699931. Biobehavioral Reviews, 34(1), 87–107. https://doi.org/10.1016/j. 2016.1159944. neubiorev.2009.07.007. Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC 400 Journal of Behavioral Addictions 9 (2020) 2, 383-400 van Timmeren, T., Daams, J. G., van Holst, R. J., & Goudriaan, A. E. four-factor model of impulsivity. European Journal of Person- (2018). Compulsivity-related neurocognitive performance deficits ality, 19(7), 559–574. https://doi.org/10.1002/per.556. in gambling disorder: A systematic review and meta-analysis. Williams, J. J., & Griffiths, T. L. (2013). Why are people bad at Neuroscience & Biobehavioral Reviews, 84, 204–217. https://doi. detecting randomness? A statistical argument. Journal of org/10.1016/j.neubiorev.2017.11.022. Experimental Psychology. Learning, Memory, and Cognition, Verdejo-Garcia, A., Alcazar-Corcoles, M. A., & Albein-Urios, 39(5), 1473–1490. https://doi.org/10.1037/a0032397. N. (2019). Neuropsychological interventions for decision- Worhunsky, P. D., Potenza, M. N., & Rogers, R. D. (2017). Alterations in making in addiction: A systematic review. Neuropsychology functional brain networks associated with loss-chasing in gambling Review, 29(1), 79–92. https://doi.org/10.1007/s11065-018- disorder and cocaine-use disorder. Drug and Alcohol Dependence, 9384-6. 178, 363–371. https://doi.org/10.1016/j.drugalcdep.2017.05.025. Verdejo-Garcia, A., Lozano, O., Moya, M., Alcazar, M. A., & Perez-Garcia, Yang, Q., Tang, P., Gu, R., Luo, W., & Luo, Y. (2015). Implicit M. (2010). Psychometric properties of a Spanish version of the UPPS- emotion regulation affects outcome evaluation. Social Cognitive P impulsive behavior scale: Reliability, validity and association with and Affective Neuroscience, 10(6), 824–831. https://doi.org/10. trait and cognitive impulsivity. Journal of Personality Assessment, 1093/scan/nsu124. 92(1), 70–77. https://doi.org/10.1080/00223890903382369. Yan, W.-S., Zhang, R.-R., Lan, Y., Li, Y.H., & Sui, N. (2016). Com- Whiteside, S. P., & Lynam, D. R. (2001). The five factor model and parison of impulsivity in non-problem, at-risk and problem gam- impulsivity: Using a structural model of personality to under- blers. Scientific Reports, 6, 39233. https://doi.org/10.1038/srep39233. stand impulsivity. Personality and Individual Differences, 30(4), Yu, R. Q., Gunn, J., Osherson, D., & Zhao, J. (2018). The consis- 669–689. https://doi.org/10.1016/S0191-8869(00)00064-7. tency of the subjective concept of randomness. Quarterly Whiteside, S. P., Lynam, D. R., Miller, J. D., & Reynolds, S. K. Journal of Experimental Psychology, 71(4), 906–916. https://doi. (2005). Validation of the UPPS impulsive behaviour scale: A org/10.1080/17470218.2017.1307428. Open Access statement. This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (https://creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted use, distribution, and reproduction in any medium for non-commercial purposes, provided the original author and source are credited, a link to the CC License is provided, and changes – if any – are indicated. Brought to you by Universitat Autonoma de Barcelona | Unauthenticated | Downloaded 05/18/22 05:39 PM UTC