ORIGINAL RESEARCH published: 28 May 2020 doi: 10.3389/fpsyt.2020.00482 Gambling Phenotypes in Online Sports Betting Roser Granero 1,2, Susana Jiménez-Murcia 2,3,4,5*, Amparo del Pino-Gutiérrez 3,6, Bernat Mora 3, Eduardo Mendoza-Valenciano 3,7, Isabel Baenas-Soto 3, Mónica Gómez-Peña 3, Laura Moragas 3, Ester Codina 3,6, Hibai López-González 3, Teresa Mena-Moreno 2,3, Gemma Mestre-Bach 2,3, Susana Valero-Solı́s 3, Sandra Rivas 3, Zaida Agüera 2,3, Cristina Vintró-Alcaraz 3, Marı́a Lozano-Madrid 2,3, José M. Menchón 3,7 and Fernando Fernández-Aranda 2,3,4,5 1 Department of Psychobiology and Methodology, Universitat Autònoma de Barcelona—UAB, Barcelona, Spain, 2 Ciber Edited by: Giovanni Martinotti, Fisiopatologı´a Obesidad y Nutrición (CIBERobn), Instituto Salud Carlos III, Madrid, Spain, 3 Department of Psychiatry, Università degli Studi G. d’Annunzio Hospital Universitari de Bellvitge, L’Hospitalet de Llobregat, Barcelona, Spain, 4 Psychiatry and Mental Health Group, Chieti e Pescara, Italy Neuroscience Program, Institut d’Investigació Biomèdica de Bellvitge—IDIBELL, L’Hospitalet de Llobregat, Barcelona, Spain, 5 Department of Clinical Sciences, School of Medicine, Universitat de Barcelona—UB, L’Hospitalet de Llobregat, Barcelona, Reviewed by: Spain, 6 Department of Public Health, Mental Health and Perinatal Nursing, School of Nursing, Universitat de Barcelona—UB, Oussama Kebir, Barcelona, Spain, 7 Ciber Salut Mental (CIBERsam), Instituto Salud Carlos III, Madrid, Spain Institut National de la Santé et de la Recherche Médicale (INSERM), France Background and Objectives: The Internet provides easy access to multiple types of Ornella Corazza, University of Hertfordshire, gambling and has led to changes in betting habits. A severe rise in problematic gambling United Kingdom has been predicted among all sectors of the population, and studies are required to Mauro Pettorruso, assess the emerging phenotypes related to the new structures of gambling activities. This D’Annunzio University of Chieti– Pescara, Italy study aimed to explore the existence of latent classes associated with gambling habits *Correspondence: among treatment-seeking gamblers due to Online Sports Betting (OSB). Susana Jiménez-Murcia sjimenez@bellvitgehospital.cat Method: Initial sample included n = 4,516 patients consecutively admitted for treatment in a hospital unit specialized in behavioral addictions. Two-step clustering analysis was used Specialty section: within the subsample of n = 323 patients who reported problems related with OSB, within This article was submitted to a set of indicators including sociodemographics, psychopathological distress, personality, Addictive Disorders, a section of the journal and severity of the gambling activity. Frontiers in Psychiatry Results: The prevalence of OSB as a main type of gambling problem in the study was Received: 24 February 2020 Accepted: 12 May 2020 7.2% (95% confidence interval: 6.4 to 7.9%). Two latent clusters were identified, with Published: 28 May 2020 differences in sociodemographics and clinical status. Cluster 1 (n = 247, 76.5%) grouped Citation: patients that were more affected due to the OSB behaviors, and it was characterized by Granero R, Jiménez-Murcia S, non-married patients, lower socioeconomic position index, higher comorbidity with other del Pino-Gutiérrez A, Mora B, Mendoza-Valenciano E, substance related addictions, younger age, and early onset of the gambling activity, as Baenas-Soto I, Gómez-Peña M, well as higher debts due to the OSB, higher psychopathological distress, and a more Moragas L, Codina E, López-González H, Mena-Moreno T, dysfunctional personality profile. Cluster 2 (n = 76, 23.5%) grouped patients that were less Mestre-Bach G, Valero-Solı´s S, affected by OSB, mostly married (or living with a stable partner), with higher social position Rivas S, Agüera Z, Vintró-Alcaraz C, levels, older age and older onset of the gambling activity, as well as a more functional Lozano-Madrid M, Menchón JM and Fernández-Aranda F (2020) Gambling psychopathological and personality profile. Phenotypes in Online Sports Betting. Front. Psychiatry 11:482. Conclusion: The increasing understanding of latent classes underlying OSB phenotypes doi: 10.3389/fpsyt.2020.00482 is essential in guiding the development of reliable screening tools to identify individuals Frontiers in Psychiatry | www.frontiersin.org 1 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting highly vulnerable to addictive behaviors among Internet gamblers, as well as in planning prevention and treatment initiatives focused on the precise profiles of these patients. Keywords: clustering, gambling disorder, internet, online sports betting, phenotype INTRODUCTION modalities exist that differ in several aspects, such as the range of stakes involved, the odds of the winning, or the level of Despite the extensive research on the involvement in gambling mental/physical skills required. The advances in technology and related problems, a new phenomenon has recently emerged the universalization of Internet access have facilitated fast and which is causing concern among specialists: Online Sports easy access to almost all traditional manners of gambling Betting (OSB). The expansion of this gambling modality in globally. It has been observed that when gamblers can choose, developed countries has significantly increased in parallel with they tend to select Internet instead of land-based modes, arguing, opportunities to participate in online gambling services. Some as the main reasons, convenience, higher fun-excitement- experts advice that the characteristics of this betting modality entertainment, greater comfort (online is accessible in their could make it potentially more addictive and dangerous than own homes), perception of greater capacity to win money, other gambling activities, or even than betting at physical faster play speed, anonymity, and privacy (9). As a result, locations [online gambling sites are permanently accessible online gambling (also referred to as Internet, interactive, or from anywhere there’s an Internet connection (24 hour-day, 7 remote gambling) is currently a particular area of interest, and day-week), gamblers can play via computer or mobile device at many studies have emerged to investigate the characteristics and different sites (such as work, home), and online provides greater motivations of the growing population of Internet gamblers. convenience, anonymity and comfort than other offsite Online gambling does not represent a new gambling platforms]. But although the potential unhealthy consequences modality, since Internet is only a mode of access to multiple associated with OSB, little research has been centered on this gambling platforms. Evidence suggests that the same activity behavior. This study contributes to developing empirical experienced in online modalities versus venue/land-based forms knowledge regarding the phenotypes of the OSB, classically may have particular features that can lead to different harms (10, included as a subtype of the problematic or disordered gambling. 11). A relationship also exists, albeit complex, between the Gambling Disorder (GD) is a behavioral addiction involving a availability of the gambling activities and the level of the repeated and uncontrolled urge to gamble, with the consequence related problems (12). It seems that fast, easy, and constant of clinical impairment or distress [DSM-5; (1)]. The DSM-5 access to gambling, as well as the ability to bet for uninterrupted taxonomy allows specifying the gambling severity level based on periods in private settings, may contribute to the early onset of the number of criteria (mild, moderate, or severe gambling), the gambling activities, and the high progression of the while the ICD-11 (2) adds the subdivision into the subtypes of gambling related problems (10, 11). The particular structural predominantly offline versus predominantly online. characteristics and the interactive and immersive Internet Epidemiological studies regularly update worldwide environment could also adversely affect the gambling related prevalences for problem gambling, which estimates they were harms (13). For example, the payment methods: compared to the between 0.1% and 5.8% across five continents during the year cash procedure typical of many land/based games, the digital before the survey, and between 0.7% and 6.5% during lifetime forms of money used through Internet system (e.g. credit cards, (3). The noteworthy increase in the incidences reported in meta- electronic funds transfers, or e-wallets) appear to lead to greater analytical data during the last decades has led to a large volume expenditures and increased gambling and losses (14). It has also of studies aimed at increasing the understanding of the been stated that online gambling environments contribute to mechanisms explaining the onset and the progression of the problems with self-regulation and self-control on spending GD (4). Much of the pioneer research in the gambling area had decisions (11, 15, 16), and that the rapid-sequential choice often contended that individuals who engage excessively in typical of computerized environments facilitates transactions gambling behaviors experience a common set of symptoms, and significantly contributes to gamblers’ decision to continue which are the result of shared risk factors, and which lead to and intensify gambling behavior (17). similar treatment outcomes (5, 6). This assumption involved the But the results on the differences between online and offsite grouping of different gambling types within a theoretical gambling outcomes are still controversial. Some studies outline homogeneous condition, failing to take into account subtypes that adjusting by the gambling preference and other variables of gamblers based on how they engage in gambling activity and (such as the frequency of participation), the contribution of avoiding exploring how the mode of access could affect the online access does not achieve the predictive capacity of the latent phenotypes. gambling impairment (18–20). It has also been observed that the But in the recent years the gambling subtypes and the new dichotomy of online versus offline access to gambling is far too emerging gambling modes have increased clinical and research limited to appropriately understand differences in the subgroups interest, due to their supposed relevant role in the pathways of of gamblers, and that considering the individuals’ life cycle by the gambling picture (7, 8). At present, multiple gambling combining chronological age, socioeconomic position and Frontiers in Psychiatry | www.frontiersin.org 2 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting marital status, should provide better insights into groups (21). It of OSB treatment-seeking patients, using a broad set of indicator has also been postulated that it should be the mixed-mode of variables, including sociodemographics, problem gambling gambling (using both Internet and land-based modes, compared severity, psychological distress, and personality functioning. to internet-only and land-based only) that predicts the greatest Based on the available empirical evidence obtained in different overall involvement in gambling and the greatest level of modes of gambling, we hypothesized that OSB constitute a gambling problems (22). Among clinical treatment-seeking mixed group in which latent underlying subgroups with samples, online pathological gamblers (compared to landed- different gambling profiles can be recognized. Since no based pathological gamblers) have shown limited differences, previous study using this approach is available for OSB focused on slightly higher educational levels, higher samples, we made no a priori assumptions about the number socioeconomical positions, and larger amounts of bets and of expected groups. debts related to Internet forms (23). And even considering the Identifying the variables related to these empirical classes legal status of Internet gambling (with great differences between would facilitate the development of measurement tools with countries), studies carried out across different European good discriminative ability and the planning of effective and jurisdictions (ranging from prohibition of online gambling to precise prevention and treatment programs. broad legal access) have found no relevant differences in the prevalence rates of GD depending on the mode of access (24). One of the most popular types of Internet gambling is OSB. It is heavily marketed and successfully targeted at the young METHOD adult male, with the consequence of hundreds of websites facilitating access to sportsbooks and fastest developing forms Participants internationally. OSB represents an example of the potential The initial sample comprised n = 4,516 treatment-seeking interaction between the mode of access to the gambling patients consecutively attended to at the Pathological activity (Internet versus land-based) and the presence and Gambling Unit and other Behavioral Addictions situated in the severity of the gambling related problems. One study has Bellvitge University Hospital (Barcelona), between January 2005 obtained a great difference in the prevalence of impairment and August 2019. This hospital oversees the treatment of cases related to the sports betting among Internet gamblers (67%) with behavioral addiction-related problems, and it is certified as a compared to land-based gamblers (23%) (25). Based on a large tertiary care center for the treatment of these psychiatric online survey, it has also been observed that, compared to non- conditions. Data analyzed in the study corresponded to the Internet gamblers (both moderate-risk and problem gamblers), first assessment before starting treatment. A total of 3,982 Internet gamblers that experience gambling-related harms are (88.2%) men were attended to, versus 534 women (11.8%). characterized by younger age individuals, engaged in a greater Most of the participants had achieved a primary or less number of gambling behaviors, and more likely to bet on sports (57.1%) education level, were single (41.9%) or married (26). Finally, it has been postulated that OSB gamblers (44.8%), belonged to low socioeconomic levels (51.5%), and (compared to non-sports Internet GD patients and GD were employed (55%). The mean chronological age was 42.0 patients who did not gamble online) represent a particular years old (SD = 13.9), and the mean duration of the gambling vulnerable subgroup characterized by distinct personality traits was 6.1 years (SD = 6.2). The most prevalent reason for seeking (higher persistence levels), riskier betting behavior, and higher treatment in this behavioral addictions unit was GD (n = 3,987; debt levels (27). 88.3%). The first block of Table 1 contains the description of the But although there is cumulated evidence regarding OSB initial complete sample. phenotype compared to other gambling types, most studies A subsample of n = 323 patients who reported OSB related have been planned under the assumption that OSB constitutes problems as the main reason for treatment-seeking was selected a unique homogeneous phenotype, grouping mostly young, for exploring the existence of latent classes based on a set of male, single, with medium to high education levels, employed, indicators, including sociodemographic and clinical variables. or full-time student (15). Although, lower income, minor The mean chronological age was 32.2 years old (SD = 9.7), and ethnicity groups, immigrant situation, and engaging in the mean duration of the betting behavior was 3.7 years (SD = multiple different gambling forms have also been reported as 3.7). Most patients in this subsample were single (61.0%), distinctive characteristics of the OSB profile (28). To the best of employed (67.2%), and had achieved secondary education our knowledge no study has been conducted to identify latent levels (49.2%). The second block of Table 1 contains the empirical classes within OSB groups in a clinical treatment- description of the OSB subsample. seeking population. Measures Objectives Diagnostic Questionnaire for Pathological Gambling (according To summarize, although there is increasing interest in online to DSM criteria) (29). This is a self-report questionnaire gambling, little has been reported in the scientific literature about developed to identify the presence of GD using 19 items based the heterogeneity of the OSB profiles regarding demographic on the DSM criteria [diagnoses are available for the DSM-IV-TR characteristics and other clinical features. This study aimed to (30) and the DSM-5 versions (1)]. The psychometrical Spanish explore the existence of empirical latent classes in a large sample adaptation of this tool achieved adequate properties (Cronbach’s Frontiers in Psychiatry | www.frontiersin.org 3 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting TABLE 1 | Characteristics of the patients in the study. Total sample OSB subsample (n = 4,516) (n = 323) Sociodemographics n % n % Sex Women 534 11.8 13 4.0 Men 3,982 88.2 310 96.0 Education Primary or less 2,577 57.1 106 32.8 Secondary 1,615 35.8 159 49.2 University 324 7.2 58 18.0 Marital status Single 1,892 41.9 197 61.0 Married-couple 2,023 44.8 99 30.6 Divorced-Separated 601 13.3 27 8.4 Social status High 65 1.4 10 3.1 Mean-high 224 5.0 34 10.5 Mean 484 10.7 49 15.2 Mean-low 1,418 31.4 126 39.0 Low 2,325 51.5 104 32.2 Employment Unemployed 2,032 45.0 106 32.8 Employed 2,484 55.0 217 67.2 Age and evolution Min Max Median Min Max Median Chronological age (yrs-old) 15 88 40 15 80 30 Onset of the addiction (yrs-old) 14 80 27 14 58 23 Duration of the addiction (yrs) 1 46 4 1 23 2 OSB, online sports betting; Min, minimum; Max, maximum. alpha a= 0.81 for a population-based sample and a=0.77 for a index, according to Hollingshead’s scale (based on the clinical sample) (31). The internal consistency achieved in this participants’ level of education and profession) (36). Patients study was good (a=0.814). also completed questions related to OSB problem-related Symptom Checklist-Revised (SCL-90-R) (32). This is a self- variables (age of onset, duration, bets per gambling/episode, report questionnaire developed to assess the psychological state and cumulated debts due to the gambling addiction) and using 90 items factorized into nine primary (first order) substance use (no vs. at least sometimes). dimensions (somatization, obsessive-compulsive, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, Procedure paranoid ideation, and psychoticism), and three global indices The study was carried out in accordance with the Declaration of [global severity index (GSI), total positive symptoms (PST), and Helsinki principles, and approved by the Ethics Committee of p ositive sym p toms d iscom for t ind ex (PSDI)]. T he University Hospital of Bellvitge (Barcelona). All patients psychometrical Spanish adaptation of this tool obtained provided signed informed consent for participating in the adequate properties (the mean Cronbach’s alpha was a = 0.75) research. There was no financial or other compensation for (33). The internal consistency in our sample was also in the being part of the sample of the study. adequate to good range (a = 0.790, for the paranoid ideation The assessment process took place in a single session lasting scale, to a = 0.981 for the global indices). about 90 min. Data for the semi-structured interview were Temperament and Character Inventory-Revised (TCI-R) (34). collected by psychologists and psychiatrists with high This is a self-report questionnaire developed to assess personality experience in the treatment of behavioral addictions. The traits using 240 items based on the Cloninger’s multidimensional clinicians also helped participants to complete the self-report model, and structured into seven factors [four for temperament questionnaires in order to guarantee that no data were missing (novelty seeking, harm avoidance, reward dependence, and (for example, due the lack of understanding). persistence), and three for character (self-directedness, cooperation, and self-transcendence)]. The psychometrical Statistical Analysis Spanish adaptation of the tool obtained adequate properties The statistical analysis was carried out with SPSS24 for (the mean Cronbach’s alpha was a = 0.87) (35). The internal windows (37). The identification of the latent empirical classes consistency in the sample of the study was in the adequate to was based on a two-step cluster procedure, a method used to good range (a = 0.703, for novelty seeking, to a = 0.868 explore the existence of natural groupings within a dataset for persistence). of categorical and continuous variables. This method uses an Other variables. This study also analyzed additional data agglomerative hierarchical clustering system and allows assessed using a semi-structured interview. This tool covered automatic determination of the optimal number of groups. socio-demographic characteristics (sex, marital status, education This study used the log-likelihood distance and the Schwarz level, employment status), as well as the socio-economic position Bayesian Information Criterion (BIC), and Akaike’s Information Frontiers in Psychiatry | www.frontiersin.org 4 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting Criterion (AIC) to determine the best model (the optimal the recruitment of the data in the study (obtained for the number of latent classes was considered for the model with the complete sample, N = 4,516). This plot displays an upward largest ratio of changes for the BIC and AIC, as well as the largest trend over time, with rates of 0.3% during 2005 to 16.1% ratio of distances measured comparing the current number of during 2019. clusters against the previous number). OSB was the reason for treatment-seeking in n = 323 patients The indicator variables in the two-step clustering included (prevalence = 7.2%; 95%CI: 6.4 to 7.9%). A number of n = 230 sociodemographic features (sex, marital status, and social patients reported the presence of OSB concurrent with other position index), global psychopathological distress (SCL-90R secondary comorbid forms of gambling or behavioral addiction GSI), personality profile (TCI-R scales), OSB severity (number [prevalence within the total sample: 5.1% (95%CI: 4.5 to 5.7%); of the DSM-5 criteria for gambling), and substances use. The prevalence within the OSB subsample: 71.2% (95%CI: 66.3 to quality of the clustering was measured using the Silhouette index, 76.1%)]. The number of OSB patients with substance use was n = a cohesion-separation measurement interpreted as how similar 161 [prevalence within the total sample: 3.6% (95%CI: 3.0 to individuals are to their own cluster compared to other clusters) 4.1%); prevalence within the OSB subsample: 49.8% (95%CI: (38). Silhouette values are into the range −1 to +1, and high 44.4 to 55.3%)]. Table S1 (Supplementary Material) contains values are indicative of adequate matching in one’s own cluster the distribution of the secondary comorbid forms of gambling and of poor matching in other clusters (values lower than 0.30 and behavioral addictions within the OSB subsample, as well as are interpreted as poor fits, between 0.30 and 0.50 as fair, and the substances use. higher than 0.50 as good). The comparison between the latent empirical clusters was Clustering Procedure based on Chi-square tests (c2) for categorical variables and on t- Table S2 (Supplementary Material) shows the results of the Test procedures for quantitative measurements. The effect sizes auto-clustering in the OSB subsample. The optimal number of for the proportion and mean differences were based on the clusters chosen by the system was two: it achieved the largest standardized Cohen’s-d coefficient, considering poor-low effect ratio changes for BIC and AIC (1.00 in both estimations), the size for |d| > 0.20, moderate-medium for |d| > 0.5, and large-high highest ratio distance (1.60) and the highest cohesion/separation for |d| > 0.80 (39). The increase in the Type-I errors due to the measurement (Silhouette = 0.30, into the fair range). Since this multiple statistical tests for comparing the clusters was controlled two-cluster model also achieved good clinical interpretation, it with the Finner method (included in the stepwise familywise was selected as the best. error rate procedures) (40). The first panel in Figure 2 shows the ordered bar-chart with the relative predictor importance in the clustering, and provides a measure of the discriminative capacity of each variable. RESULTS Relative relevance is reported in a range 0 to 1: the greater the relevance of the indicator, the less likely it is that changes Prevalence of OSB in the Study between clusters for the variable are attributable to chance. The Within the initial complete sample (N = 4,516), the number of predictor with the greatest significance in the study was the patients who reported OSB as primary or secondary reason for marital status, followed by the onset of the gambling, the treatment-seeking was n = 323 [prevalence = 7.2%; 95% novelty seeking score, tobacco use, and novelty seeking (these confidence interval (95%CI): 6.4 to 7.9%]. Figure 1 shows the variables are plotted with the darkest color bars). The line-plot for the prevalence of consultations due to OSB during remaining predictors achieved lower significance, as they FIGURE 1 | Prevalence of the prevalence of consultation due to OSB during the recruitment of data (n=4,516). Frontiers in Psychiatry | www.frontiersin.org 5 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting FIGURE 2 | Results of the clustering procedure within the sports betting online subsample (n = 323). were the variables with the poorest contribution to harm Figures 3 and 4 graphically show the results of the avoidance, reward dependence, and persistence levels. The comparison between the clusters, which contribute to the second panel contains the graphic representation of the understanding of the two latent classes within the OSB Silhouette index in the study, and the third panel the cluster patients. Figure 3 displays the line-chart with the prevalence of distribution (the ratio of sizes was 3.25, since cluster 1 achieved patients outside the normal range in the SCL-90R nearly one third of the OSB subsample). (psychopathological state) and the TCI-R scales (personality traits). As a whole, compared to Cluster 2 a higher percentage Comparison Between the Latent Empirical of patients within Cluster 1 reported mean scores within the Clusters clinical range in the SCL-90R, being the highest impairing level Table 2 shows the comparison between the empirical clusters in the depression scale and in the global indexes GSI and PST; identified in the study. Cluster 1 grouped n = 247 patients, which regarding the personality traits, the highest deviances from represented 76.5% of the OSB patients. This latent cluster was normal ranges were for self-directedness and novelty seeking. characterized by unmarried marital status, poorer socioeconomic Figure 4 displays the radar-plot with the main variables analyzed levels, higher prevalence of substances use (tobacco, alcohol, and in the study (proportions area plotted for categorical variables other illicit drugs), younger age, early onset of the OSB activity, and z-standardized means for quantitative variables, to allow higher severity of the betting activity, higher psychopathological easy interpretation due the difference in the metric scale of the distress and more dysfunctional personality profile (higher variables). As a whole, Cluster 1 represented the profile of OSB scores in novelty seeking and lower levels in self-directedness patients that are more affected, while Cluster 2 represented the and cooperativeness). Cluster 2 grouped the remaining n = 76 profile of those less affected. patients (23.5% of the OSB subsample), and it was characterized by married marital status (or living with a stable partner), higher social position levels, older age and older onset of the OSB DISCUSSION activity, lower severity associated with the OSB, and a more functional psychopathological and personality profile. The The current study estimated the prevalence of the OSB among a prevalence of substance use was also lower. large clinical sample of patients attended in a hospital unit Frontiers in Psychiatry | www.frontiersin.org 6 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting TABLE 2 | Comparison between the latent clusters identified within the OSB subsample. Cluster 1 Cluster 2 p |d| (n = 247; 76.5%) (n = 76; 23.5%) n % n % Sex Women 12 4.9 1 1.3 .169 0.21 Men 235 95.1 75 98.7 Education Primary or less 76 30.8 30 39.5 .150 0.18 Secondary 129 52.2 30 39.5 0.26 University 42 17.0 16 21.1 0.10 Marital status Single 197 79.8 0 0.0 .001* 2.21† Married-couple 26 10.5 73 96.1 2.08† Divorced-Separated 24 9.7 3 3.9 0.23 Social position index High 6 2.4 4 5.3 .012* 0.15 Mean-high 25 10.1 9 11.8 0.06 Mean 29 11.7 20 26.3 0.38 Mean-low 103 41.7 23 30.3 0.24 Low 84 34.0 20 26.3 0.17 Employment Unemployed 90 36.4 16 21.1 .012* 0.34 Employed 157 63.6 60 78.9 Other behavioral addictions 177 71.7 53 69.7 .746 0.04 Tobacco 133 53.8 13 17.1 < .001* 0.80† Alcohol 29 11.7 1 1.3 .006* 0.52† Other drugs 36 14.6 0 0.0 < .001* 0.78† Mean SD Mean SD p |d| Chronological age (yrs-old) 30.41 9.33 37.95 8.76 < .001* 0.83† Age of onset of gambling (yrs-old) 23.66 6.91 29.58 8.22 < .001* 0.78† Duration of the addiction (yrs) 3.89 3.76 3.15 3.27 .125 0.21 Number of DSM-5 criteria 7.32 1.79 6.82 1.85 .032* 0.28 1 Maximum bets (euros-episode) 800 1700 700 1775 .795 0.03 1 Mean bets (euros-episode) 40 140 35 190 .430 0.10 1 Debts due to the OSB 6500 20250 2300 14000 .023* 0.23 SCL-90R GSI 1.07 0.69 0.82 0.55 .005* 0.40 Novelty seeking 113.91 13.43 103.64 13.39 < .001* 0.77† Harm avoidance 99.96 17.21 97.55 14.52 .270 0.15 Reward dependence 96.54 14.21 98.57 13.87 .276 0.14 Persistence 106.30 20.84 105.84 18.26 .863 0.02 Self-directedness 126.27 21.59 139.21 21.59 < .001* 0.60† Cooperativeness 126.62 17.95 134.37 14.14 < .001* 0.48 Self-transcendence 60.39 14.57 57.38 13.55 .111 0.21 OSB, online sports betting; SD, standard deviation. 1 Median and interquartile range. *Bold: significant comparison. † Bold: effect size into the mild-moderate (|d| > 0.50) to large-high range (|d| > 0.80). specialized in the treatment of GD and other behavioral subgroups in a clinical sample of OSB patients. The rationales addictions. The clustering analysis then examined the variance of this study were: a) the assumption that not all problematic within the OSB subsample, with the aim of identifying latent online gamblers form a homogeneous group with common homogeneous subgroups. The phenotypical differences between features and shared vulnerabilities, and therefore to the two empirical clusters of OSB as regards sociodemographics, automatically attribute the global Internet gambling habits and gambling severity, psychopathological state, and personality, traits amongst OSB patients is inaccurate; and b) the requisite to provided reliability and validity to the clustering. distinguish modes of gambling (online versus land-based) and Earlier studies fueled concerns that online gambling gambling forms (e.g., slot-machines, lottery, sports betting, …) to significantly contributes to the onset and progression of the adequately characterize gambling related profiles, as well as more gambling related problems, with the result of prompting research specific and personalized treatment approaches for each type into the characteristics and associated risks of this mode of access of patient. to gambling activities (7, 41). But while the insights provided by The clustering analysis in this work revealed that two these studies into the profiles of problem gamblers that gamble distinctive latent subgroups was the optimal grouping solution online, they did not account for the relevant issue of the for the study that, respectively, represented latent phenotypes of heterogeneity and within-subjects variance among the samples. OSB. The characterization of these subgroups seemed to suggest This paper is the first, to our knowledge, to explore latent a dimensional factor varying in the psychological and functional Frontiers in Psychiatry | www.frontiersin.org 7 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting FIGURE 3 | Line-chart within the sports betting online subsample (n = 323). FIGURE 4 | Radar-chart within the sports betting online subsample (n = 323). affectation. This result is in line with previous studies, which have distress (43), but the higher the immersion in online gambling published a higher likelihood for experiencing psychological activity the higher are the increases in the gambling harms and distress among problematic online gamblers compared to non- their correlates (including the general psychopathological state). problematic online gamblers (42). To explain these results, it has Associated with the worse mental state among the patients been argued that online access to gambling may become within the cluster with the higher affectation, this latent subgroup particularly motivating for escaping and alleviating negative also reported higher comorbidity with substance use (tobacco, emotional states, since Internet provides privacy, is less socially alcohol, and other drugs). In fact, the tobacco consumption was demanding than many land-based gambling activities, and one predictor with high importance in the clustering, after the allows gambling sessions without d istractions and marital status and the onset of the OSB activity. This finding is interruptions. This results in a vicious circle: individuals with also consistent with previous studies, which have related a higher high risk of negative mood and anxiety states should find in likelihood of substance use while gambling (mainly for alcohol online gambling an easy way to escape and control emotional and illicit drugs) among people with online gambling habits (44). Frontiers in Psychiatry | www.frontiersin.org 8 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting Increased prevalences of substance-related disorders among gamblers controls, and it has also been identified as a strong Internet problematic gamblers compared with other gambling predictor of the gambling severity (57). Self-directness is the ability forms and with non-gamblers have also been reported (45). to adapt-regulate owns’ behavior to the demands of a situation in Epidemiological and etiological research has also shown the order to achieve personally chosen goals and values, while relevance of the co-occurrence between online gambling and cooperativeness is described as the capacity of the individuals substances consumption (mainly with tobacco), as well as the for being empathic, helpful, socially tolerant and compassionate. relationship between substances status with more severe Adequate functioning in self-directedness and cooperativeness gambling habits (46). Our results, together with this set of seem play a relevant role in fast and adaptive emotional findings, should again suggest that the solitary and private responses and in the choice of cognitive regulation strategies settings allowed by Internet gambling may undermine rational (58), and its relevance in the stress response has also been decision-making and increase the ease of substance use. Online consistently reported (59). Along this line, some researchers access at home may also facilitate gambling alone at any time of have suggested that a combination of both low self-directedness the day for long continuous sessions, and these contextual and low cooperativeness could form a general factor representing features should make substances more likely to be consumed. low psychological maturity, a temperamental vulnerability Future research should assess how gambling alone, timing, and predictive of many psychiatric disorders (60, 61). This intrinsic duration of online play contribute to the gambling severity aspect of the global mental health has also been interpreted as an among OSB. epiphenomenon, a “marker” of the neuropsychiatric dysfunctions As regards the comorbid concurrence of OSB with other in individuals who show a lack of sense of responsibility, self- addictive behaviors, previous studies have shown a strong control, and social skills, which ultimately are part of the relationship between online gambling and engaging in a wide definitions of the addictive disorders (substance and non- range of other behavioral addictions (land-based and online substance) (62). Finally, our results regarding the personality activities) (47). In our study, both latent empirical clusters profile related to the more affected cluster (higher novelty achieved high prevalences of comorbid forms of gambling and/or seeking score and lower self-directedness and cooperativeness) other behavioral addictions (71.7% and 69.7%, with a low difference are consistent with previous studies which reported the of 2.0% between the classes). The strong links (co-occurrences) contribution of these domains on the gambling area (63–65). A within addictions obtained in the scientific studies have suggested recent path-analysis study has also observed that at a young age, the notion that some people are more prone to these problems, the combination of a profile defined by immature character (low regardless of whether these involve substances or other behavioral self-directedness and cooperativeness) with extreme temperament activities (48). This higher vulnerability has been explained on the (high novelty seeking) may be a predictor of substance addiction basis of a spectrum that grouped a number of disorders drawn from across sex (with direct and indirect effects on the mental health several diagnostic categories that share core impulsive-compulsive status) (66). Future research should analyze the longitudinal features. This construct has been supported by many studies (49– predictive capacity of this particular endophenotype on the 51), who have placed GD, substance use disorders and other onset and progression of the OSB related problems. behavioral addictions (sex, buying or gaming) toward the upper Previous research into the OSB area have also emphasized band of the impulsive trait in the spectrum (the opposite upper distinguishing demographic features for this gambling subtype, band of the compulsive trait in the spectrum included disorders such as the individuals’ male sex, young age, not-married marital such as obsessive-compulsive, body dysmorphic or restrictive-type status (mostly single), medium or higher education levels and anorexia nervosa). This theoretical assumption could explain the employed as full-time student (11, 47). The results of this study results of our study, which showed a joint association between are in line with these findings. The description of the OSB the presence of OSB and the high likelihood of other subsample revealed that most patients were male (96%), with multiple addictions. secondary or university education levels (67.2%), unmarried Regarding personality traits, the higher affectation cluster was (69.4%), and employed (67.2%). The mean of the chronological characterized by a more dysfunctional profile defined by higher age among OSB subsample was also younger compared to the scores in novelty seeking and lower scores in self-directedness and mean obtained among the total sample (32.2 versus 42.0 years- cooperativeness. Novelty seeking is a personality trait strongly old), as well as the age of onset of the addictive gambling (25.1 related with the exploratory level of the individuals in response to versus 29.6 years-old), and the progression of the disorder (3.7 novel situation and the impulsive decision making (52), and it has versus 6.1 years). The results obtained in the OSB clustering been considered in research and clinical settings as a measure of reinforced these findings as regards the contribution of the the individuals’ (in)capacity to bring responses into standards and sociodemographic features in the variability of the phenotype. to support the pursuit of long-term goals and as a consequence a The higher affectation latent group was related to more deprived powerful risk factor for psychopathology (53, 54). High levels on social positions, unemployed situation, unmarried marital status, the novelty seeking dimension has been linked to all stages of younger age, and early age of onset of the gambling activity. addictions, from the acquisition phase of a single addiction to the Other studies have also showed a relationship between lower escalation to multiple concurrent addictive behaviors (51, 55, 56). income and the severity of the problematic online gambling (67), This characteristic has also been systematically obtained by and being divorced/separated or living common-law has been comparing problematic and disordered gamblers with non- described as the marital status most predictive of the worse Frontiers in Psychiatry | www.frontiersin.org 9 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting harms related to the Internet gambling (68). Lastly, it should be adequately cover the variance of naturally occurring individual emphasized that research studies focused on specific populations differences (likelihood of small extreme groups are minimized). highly vulnerable to gambling problems have found that The third strength is the assessment of other behavioral and unemployed/low income/poverty, unmarried status, young age, substance related addictions different to OSB. The high and early age are powerful risk factors for experiencing the comorbidity rates found in our study warn of the high greatest severe consequences of the gambling activity, vulnerability of the patients for the concurrent presence of independent of the gambling forms/types and the online versus multiple addictive problems and the need of early screening offline access (69). tools and prevention plans. Finally, our results are also consistent with previous research focused on the identification of separate profiles among sports Conclusion and Implications betting. The secondary data analysis published by LaBrie and In conclusion, Internet gambling has become a relative newcomer Shaffer on a longitudinal study among a large sample of to the world of gambling opportunities. The amount (in number subscribers to an Internet sports gambling site observed, using and variety) of online applications has progressed with hasty speed discriminant function analysis, a particular subgroup of during the last decades, offering changes and increases in sports individuals that made more and larger bets, bet more betting opportunities. This study is the first to systematically frequently, and were more likely to exhibit intense betting soon analyze individual variance of OSB in a large clinical sample, after enrollment (70). The results of our study, obtained in a and the clusters obtained provides empirical evidence about the clinical sample of treatment-seeking patients, adds empirical existence of different latent phenotypes associated with the evidence about the heterogeneity of the gambling habit profiles Internet sports betting habits. The identification of a latent among OSB, with a worse psychopathological state (higher subgroup of patients with higher affectation could suggest that distress and comorbid patterns with substance and non- OSB may be largely attractive for some highly vulnerable substance behaviors), a higher affectation latent subgroup, also individuals, who can isolate and immerse in these activities in characterized also by a higher duration of the addiction patterns, their home environment as an “escape problem” way, with the higher debts related to the gambling activity, and personality consequence of increased harms. The results of this work sought to traits defined by higher scores in novelty seeking, and lower self- provide a more accurate assessment of these patients in whom the directedness and cooperativeness. gambling problems stemmed from OSB, as well as to identify highly vulnerable individuals from the general population. The Limitations and Strengths findings of this work should also prove useful for planning Some limitations should be considered when interpreting the effective prevention programs for developing effectiveness results of this study. Firstly, the analysis of cross-sectional data intervention therapies focused on the needs of the patients. restricts the temporal analysis of causative associations, and future longitudinal studies should explore the predictive capacity of the identified phenotypes (for example in treatment DATA AVAILABILITY STATEMENT outcomes or developmental trajectories of the gambling problem). Secondly, the low prevalence of women within the The datasets generated for this study will not be made publicly OSB subsample restricts generalization of the findings, and available because the data used in this study is part of the hospital makes results potentially non-representative of the female database, and it is restricted to protect patients’ confidentiality. population with OSB. It should be considered that the frequency of women in the study is consistent with the point prevalence estimates in clinical treatment-seeking samples in the ETHICS STATEMENT gambling area (GD is highly more frequent among men). We decided retaining the women in the study to increase the This study was carried out in accordance with the latest version ecological validity of the study, and to be able to provide of the Declaration of Helsinki principles. The Ethics Committee pioneer results in the area for OSB women. of the Bellvitge University Hospital approved the study and A strength of the current study is the use of clustering written informed consent was obtained and signed from all procedure to identify the latent empirical groups among OSB final participants. patients, based on a relatively large set of predictors, including sociodemographics and clinical features. Compared with usual analytical procedures, cluster analysis does not require a priori AUTHOR CONTRIBUTIONS assumptions regarding the underlying profiles in the sample, and therefore it allows empirically identifying the systematic Conceptualization: SJ-M, FF-A. Data curation: AP-G. Statistical covariation of multiple features contributing to the inter- analysis: RG. Investigation: BM, EM-V, IB-S, MG-P, LM, EC, individual variance in the gambling habits. A second strength HL-G, TM-M, GM-B, SV-S, SR, CV-A, ML-M. Project is the relatively large sample size for the two latent subgroups administration: SJ-M, FF-A, JM. Resources: AP-G, ZA. Writing identified (247 and 76 patients), which suggest that the clusters of the first draft: SJ-M, RG. Review and critique: FF-A, JM. Frontiers in Psychiatry | www.frontiersin.org 10 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting ACKNOWLEDGMENTS Research of the Ministry of Business and Knowledge of the Government of Catalonia. TM-M, CV-A, and ML-M are This study has been funded by Ministerio de Economı́a, supported by a predoctoral grant awarded by the Ministerio de Industria y Competitividad (MINECO), which is part of Educació n, Cultura y Deporte (FPU16/02087; FPU16/01453; Agencia Estatal de Investigació n (AEI), through the project FPU15/02911). We thank CERCA Programme/Generalitat de RTI2018-101837-B-I00 (Co-funded by European Regional Catalunya for institutional support. Development Fund. ERDF, a way to build Europe). The research was also funded by the Delegació n del Gobierno para el Plan Nacional sobre Drogas (projects 2017I067 and 2019I47), Instituto de Salud Carlos III (ISCIII) (FIS PI14/00290 and PI17/ SUPPLEMENTARY MATERIAL 01167). CIBERobn and CIBERSAM are both initiatives of ISCIII. The funders had no role in the study design, data collection and The Supplementary Material for this article can be found online analysis, decision to publish, or preparation of the manuscript. at: https://www.frontiersin.org/articles/10.3389/fpsyt.2020. Support was received from the Secretariat for Universities and 00482/full#supplementary-material REFERENCES 16. Siemens JC, Kopp SW. The influence of online gambling environments on Self-Control. J Public Policy Mark (2011) 30(2):279–93. doi: 10.1509/ 1. American Psychiatric Association. Diagnostic and Statistical Manual of jppm.30.2.279 Mental Disorders (Fifth Edition). 5th ed. Arlington, VA: Author (2013). 17. LaPlante DA, Nelson SE, Gray HM. Breadth and depth involvement: 2. WHO. (2020). WHO. International classification of Diseases (11th Revision) Understanding Internet gambling involvement and its relationship to [Internet]. Retrieved from https://icd.who.int/browse11/l-m/en. gambling problems. Psychol Addict Behav (2014) 28(2):396–403. doi: 3. Calado F, Griffiths MD. Problem gambling worldwide: An update and 10.1037/a0033810 systematic review of empirical research (2000–2015). J Behav Addict (2016) 18. Baggio S, Dupuis M, Berchtold A, Spilka S, Simon O, Studer J. Is gambling 5(4):592–613. doi: 10.1556/2006.5.2016.073 involvement a confounding variable for the relationship between Internet 4. Yau YHC, Potenza MN. Gambling disorder and other behavioral addictions: gambling and gambling problem severity? Comput Hum Behav (2017) recognition and treatment. Harv Rev Psychiatry (2015) 23(2):134–46. doi: 71:148–52. doi: 10.1016/j.chb.2017.02.004 10.1097/HRP.0000000000000051 19. McCormack A, Shorter GW, Griffiths MD. An examination of participation in 5. Kardefelt-Winther D, Heeren A, Schimmenti A, van Rooij A, Maurage P, online gambling activities and the relationship with problem gambling. Carras M, et al. How can we conceptualize behavioural addiction without J Behav Addict (2013) 2(1):31–41. doi: 10.1556/JBA.2.2013.1.5 pathologizing common behaviours? Addiction (2017) 112(10):1709–15. doi: 20. Philander KS, MacKay TL. Online gambling participation and problem 10.1111/add.13763 gambling severity: Is there a causal relationship? Int Gambl Stud (2014) 14 6. Zou Z, Wang H, d’Oleire Uquillas F, Wang X, Ding J, Chen H. Definition of (2):214–27. doi: 10.1080/14459795.2014.893585 Substance and Non-substance Addiction. Adv Exp Med Biol (2017) 1010:21– 21. Gainsbury SM, Liu Y, Russell AMT, Teichert T. Is all Internet gambling 41. doi: 10.1007/978-981-10-5562-1_2 equally problematic? Considering the relationship between mode of access 7. Braverman J, LaPlante DA, Nelson SE, Shaffer HJ. Using cross-game and gambling problems. Comput Hum Behav (2016) 55:717–28. doi: 10.1016/ behavioral markers for early identification of high-risk internet gamblers. j.chb.2015.10.006 Psychol Addict Behav (2013) Sep27(3):868–77. doi: 10.1037/a0032818 22. Gainsbury SM, Russell A, Wood R, Hing N, Blaszczynski A. How risky is 8. Jimé n ez-Murcia S, Granero R, Ferná n dez-Aranda F, Menchó n JM. Internet gambling? A comparison of subgroups of Internet gamblers based on Comparison of gambling profiles based on strategic versus non-strategic problem gambling status. New Media Soc (2015) 17(6):861–79. doi: 10.1177/ preferences. Curr Opin Behav Sci (2020) 31:13–20. doi: 10.1016/ 1461444813518185 j.cobeha.2019.09.001 23. Jimé nez-Murcia S, Stinchfield R, Ferná ndez-Aranda F, Santamarı́a JJ, Penelo 9. McCormack A, Griffiths MD. Motivating and Inhibiting Factors in Online E, Granero R, et al. Are online pathological gamblers different from non- Gambling Behaviour: A Grounded Theory Study. Int J Ment Health Addict online pathological gamblers on demographics, gambling problem severity, (2012) 10(1):39–53. doi: 10.1007/s11469-010-9300-7 psychopathology and personality characteristics? Int Gambl Stud (2011) 11 10. Gainsbury SM. Online Gambling Addiction: the Relationship Between (3):325–37. doi: 10.1080/14459795.2011.628333 Internet Gambling and Disordered Gambling. Curr Addict Rep (2015) 2 24. Planzer S, Gray HM, Shaffer HJ. Associations between national gambling (2):185–93. doi: 10.1007/s40429-015-0057-8 policies and disordered gambling prevalence rates within Europe. Int J Law 11. Hing N, Russell AMT, Gainsbury SM, Blaszczynski A. Characteristics and Psychiatry (2014) 37(2):217–29. doi: 10.1016/j.ijlp.2013.11.002 help-seeking behaviors of Internet gamblers based on most problematic mode 25. Gainsbury SM, Russell A, Hing N, Wood R, Lubman DI, Blaszczynski A. The of gambling. J Med Internet Res (2015) 17(1):e13. doi: 10.2196/jmir.3781 prevalence and determinants of problem gambling in Australia: assessing the 12. Gray HM, LaPlante DA, Shaffer HJ. Behavioral characteristics of Internet impact of interactive gambling and new technologies. Psychol Addict Behav gamblers who trigger corporate responsible gambling interventions. Psychol (2014) 28(3):769–79. doi: 10.1037/a0036207 Addict Behav (2012) 26(3):527–35. doi: 10.1037/a0028545 26. Gainsbury SM, Russell A, Hing N, Wood R, Blaszczynski A. The impact of 13. McCormack A, Griffiths MD. Scoping study of the structural and situational internet gambling on gambling problems: a comparison of moderate-risk and characteristics of internet gambling. Int J Cyber Behav Psychol Learn (2013) 3 problem Internet and non-Internet gamblers. Psychol Addict Behav (2013) 27 (1):29–49. doi: 10.4018/ijcbpl.2013010104 (4):1092–101. doi: 10.1037/a0031475 14. Gainsbury SM, Russell A, Hing N, Wood R, Lubman D, Blaszczynski A. How 27. Esté vez A, Rodrı́guez R, Dı́az N, Granero R, Mestre-Bach G, Steward T, et al. the Internet is Changing Gambling: Findings from an Australian Prevalence How do online sports gambling disorder patients compare with land-based Survey. J Gambl Stud (2015) 31(1):1–15. doi: 10.1007/s10899-013-9404-7 patients? J Behav Addict (2017) 6(4):639–47. doi: 10.1556/2006.6.2017.067 15. Hing N, Cherney L, Gainsbury SM, Lubman DI, Wood RT, Blaszczynski A. 28. Hing N, Russell AMT. Psychological factors, sociodemographic Maintaining and losing control during internet gambling: A qualitative study characteristics, and coping mechanisms associated with the self-stigma of of gamblers’ experiences. New Media Soc (2015) 17(7):1075–95. doi: 10.1177/ problem gambling. J Behav Addict (2017) 6(3):416–24. doi: 10.1556/ 1461444814521140 2006.6.2017.056 Frontiers in Psychiatry | www.frontiersin.org 11 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting 29. Stinchfield R. Reliability, Validity, and Classification Accuracy of a Measure of compulsivity. Psychopharmacol (Berl) (2012) 219(2):469–90. doi: 10.1007/ DSM-IV Diagnostic Criteria for Pathological Gambling. Am J Psychiatry s00213-011-2550-7 (2003) 160(1):180–2. doi: 10.1176/appi.ajp.160.1.180 52. Costa VD, Tran VL, Turchi J, Averbeck BB. Dopamine modulates novelty 30. American Psychiatric Association. Diagnostic and Statistical Manual of seeking behavior during decision making. Behav Neurosci (2014) 128(5):556– Mental Disorders (Fourth Edition-Rev). 4th Rev. Washington, DC: American 66. doi: 10.1037/a0037128 Psychiatric Association (2010). 53. Baumeister RF, Vohs KD, Tice DM. The strength model of self-control. Curr 31. Jiménez-Murcia S, Stinchfield R, Alvarez-Moya E, Jaurrieta N, Bueno B, Granero Dir Psychol Sci (2007) 16(6):351–5. doi: 10.1111/j.1467-8721.2007.00534.x R, et al. Reliability, validity, and classification accuracy of a Spanish translation of a 54. Norbury A, Husain M. Sensation-seeking: Dopaminergic modulation and risk measure of DSM-IV diagnostic criteria for pathological gambling. J Gambl Stud for psychopathology. Behav Brain Res (2015) 288:79–93. doi: 10.1016/ (2009) 25(1):93–104. doi: 10.1007/s10899-008-9104-x j.bbr.2015.04.015 32. Derogatis LR. SCL-90-R. In: . Cuestionario de 90 sı́ntomas. Madrid: TEA 55. Black DW, Coryell WH, Crowe RR, Shaw M, McCormick B, Allen J. Ediciones (1997). Personality Disorders, Impulsiveness, and Novelty Seeking in Persons with 33. Gonzalez De Rivera JL, Derogatis LR, De las Cuevas C, Gracia Marco R, DSM-IV Pathological Gambling and Their First-Degree Relatives. J Gambl Rodrı́guez-Pulido F, Henry-Benitez M, et al. The Spanish version of the SCL- Stud (2015) 31(4):1201–14. doi: 10.1007/s10899-014-9505-y 90-R. Normative data in the general population. Towson: Clinical 56. Wingo T, Nesil T, Choi J-S, Li MD. Novelty Seeking and Drug Addiction in Psychometric Research (1989). Humans and Animals: From Behavior to Molecules. J Neuroimmune 34. Cloninger CR, Przybeck TR, Syrakic DM, Wetzel RD. The Temperament and Pharmacol (2016) 11(3):456–70. doi: 10.1007/s11481-015-9636-7 Character Inventory (TCI). A guide to its development and use. Washington 57. Nower L, Martins SS, Lin K-H, Blanco C. Subtypes of disordered gamblers: University, St. Louis, MO: Center for Psychobiology of Personality (1994). results from the National Epidemiologic Survey on Alcohol and Related 35. Gutié rrez-Zotes JA, Bayó n C, Montserrat C, Valero J, Labad A, Cloninger CR, Conditions. Addiction (2013) 108(4):789–98. doi: 10.1111/add.12012 et al. [Temperament and Character Inventory Revised (TCI-R). 58. Chae H, Park SH, Garcia D, Lee SJ. Cloninger’s TCI associations with adaptive Standardization and normative data in a general population sample]. Actas and maladaptive emotion regulation strategies. PeerJ (2019) 7:e7958. doi: Españolas Psiquiatr (2004) 32(1):8–15. 10.7717/peerj.7958 36. Hollingshead AB. Four Factor Index of Social Status. Yale J Sociol (2011) 8:21–51. 59. Eley DS, Leung J, Hong BA, Cloninger KM, Cloninger CR. Identifying the 37. IBM-Corp. IBM SPSS Statistics for Windows, version 24.0. Arkmonk, NY: IBM Dominant Personality Profiles in Medical Students: Implications for Their Corp (2016). Well-Being and Resilience. PloS One (2016) 11(8):e0160028. doi: 10.1371/ 38. Rousseeuw PJ. Silhouettes: a graphical aid to the interpretation and validation journal.pone.0160028 of cluster analysis. J Comput Appl Math (1987) 20:53–65. doi: 10.1016/0377- 60. Garcia D, Anckarsater H, Lundstrom S. Self-directedness and cooperativeness, 0427(87)90125-7 psychosocial dysfunction and suffering in ESSENCE. ScientificWorldJournal 39. Kelley K, Preacher KJ. On effect size. Psychol Methods (2012) 17(2):137–52. (2013) 2013:416981. doi: 10.1155/2013/416981 doi: 10.1037/a0028086 61. Laidlaw TM, Dwivedi P, Naito A, Gruzelier JH. Low self-directedness (TCI), 40. Finner H. On a monotonicity problem in step-down multiple test procedures. mood, schizotypy and hypnotic susceptibility. Pers Individ Dif (2005) 39 J Am Stat Assoc (1993) 88:920–3. doi: 10.1080/01621459.1993.10476358 (2):469–80. doi: 10.1016/j.paid.2005.01.025 41. Wardle H, Moody A, Griffiths M, Orford J, Volberg R. Defining the online 62. Ando B, Must A, Kurgyis E, Szkaliczki A, Drotos G, Rozsa S, et al. Personality gambler and patterns of behaviour integration: Evidence from the British traits and coping compensate for disadvantageous decision-making in long- Gambling Prevalence Survey 2010. Int Gambl Stud (2011) 11(3):339–56. doi: term alcohol abstinence. Alcohol Alcohol (2012) 47(1):18–24. doi: 10.1093/ 10.1080/14459795.2011.628684 alcalc/agr144 42. Nower L, Blaszczynski A. Gambling motivations, money-limiting strategies, 63. del Pino-Gutié rrez A, Jimé nez-Murcia S, Ferná ndez-Aranda F, Agüera Z, and precommitment preferences of problem versus non-problem gamblers. Granero R, Hakansson A, et al. The relevance of personality traits in J Gambl Stud (2010) 26(3):361–72. doi: 10.1007/s10899-009-9170-8 impulsivity-related disorders: From substance use disorders and gambling 43. Sauvaget A, Jimé nez-Murcia S, Ferná ndez-Aranda F, Fagundo AB, Moragas L, disorder to bulimia nervosa. J Behav Addict (2017) 6(3):396–405. doi: 10.1556/ Wolz I, et al. Unexpected online gambling disorder in late-life: A case report. 2006.6.2017.051 Front Psychol (2015) 6(MAY):655. doi: 10.3389/fpsyg.2015.00655 64. Mestre-Bach G, Steward T, Granero R, Ferná ndez-Aranda F, Taló n-Navarro 44. Castré n S, Basnet S, Salonen AH, Pankakoski M, Ronkainen J-E, Alho H, et al. MT, Cuquerella Á , et al. Gambling and impulsivity traits: A recipe for criminal Factors associated with disordered gambling in Finland. Subst Abuse Treat behavior? Front Psychiatry (2018) 9:6. doi: 10.3389/fpsyt.2018.00006 Prev Policy (2013) 8:24. doi: 10.1186/1747-597X-8-24 65. Moragas L, Granero R, Stinchfield R, Ferná ndez-Aranda F, Fröberg F, 45. Scholes-Balog KE, Hemphill SA. Relationships between online gambling, Aymamı́ N, et al. Comparative analysis of distinct phenotypes in gambling mental health, and substance use: a review. Cyberpsychol Behav Soc Netw disorder based on gambling preferences. BMC Psychiatry (2015) 15(1):86. (2012) 15(12):688–92. doi: 10.1089/cyber.2012.0232 doi: 10.1186/s12888-015-0459-0 46. Petry NM, Oncken C. Cigarette smoking is associated with increased severity 66. Boson K, Wennberg P, Fahlke C, Berglund K. Personality traits as predictors of gambling problems in treatment-seeking gamblers. Addiction (2002) 97 of early alcohol inebriation among young adolescents: Mediating effects by (6):745–53. doi: 10.1046/j.1360-0443.2002.00163.x mental health and gender-specific patterns. Addict Behav (2019) 95:152–9. 47. Hing N, Russell AM, Browne M. Risk Factors for Gambling Problems on doi: 10.1016/j.addbeh.2019.03.011 Online Electronic Gaming Machines, Race Betting and Sports Betting. Front 67. Currie SR, Hodgins DC, Wang J, el-Guebaly N, Wynne H, Miller NV. Psychol (2017) 8:779. doi: 10.3389/fpsyg.2017.00779 Replication of low-risk gambling limits using canadian provincial gambling 48. Starcevic V, Khazaal Y. Relationships between Behavioural Addictions and prevalence data. J Gambl Stud (2008) 24(3):321–35. doi: 10.1007/s10899-008- Psychiatric Disorders: What Is Known and What Is Yet to Be Learned? Front 9091-y Pychiat (2017) 8:53. doi: 10.3389/fpsyt.2017.00053 68. Wood RT, Williams RJ. A comparative profile of the internet gambler: 49. Bottesi G, Ghisi M, Ouimet AJ, Tira MD, Sanavio E. Compulsivity and Demographic characteristics, game-play patterns, and problem gambling Impulsivity in Pathological Gambling: Does a Dimensional-Transdiagnostic status. New Media Soc (2011) 13:1123–41. doi: 10.1177/146144481 Approach Add Clinical Utility to DSM-5 Classification? J Gambl Stud (2015) 0397650 31(3):825–47. doi: 10.1007/s10899-014-9470-5 69. Petry NM, Zajac K, Ginley MK. Behavioral Addictions as Mental Disorders: 50. Lee RSC, Hoppenbrouwers S, Franken I. A Systematic Meta-Review of To Be or Not To Be? Annu Rev Clin Psychol (2018) 14:399–423. doi: 10.1146/ Impulsivity and Compulsivity in Addictive Behaviors. Neuropsychol Rev annurev-clinpsy-032816-045120 (2019) 29(1):14–26. doi: 10.1007/s11065-019-09402-x 70. Labrie R, Shaffer HJ. Identifying behavioral markers of disordered internet 51. Leeman RF, Potenza MN. Similarities and differences between pathological sports gambling. Addict Res Theory (2011) 19(1):56–65. doi: 10.3109/ gambling and substance use disorders: a focus on impulsivity and 16066359.2010.512106 Frontiers in Psychiatry | www.frontiersin.org 12 May 2020 | Volume 11 | Article 482 Granero et al. Online Sports Betting Conflict of Interest: The authors declare that the research was conducted in the Vintró-Alcaraz, Lozano-Madrid, Menchón and Fernández-Aranda. This is absence of any commercial or financial relationships that could be construed as a an open-access article distributed under the terms of the Creative Commons potential conflict of interest. Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) Copyright © 2020 Granero, Jime ń ez-Murcia, del Pino-Gutie ŕ rez, Mora, are credited and that the original publication in this journal is cited, in Mendoza-Valenciano, Baenas-Soto, Góm ez-Peña, Moragas, Codina, accordance with accepted academic practice. No use, distribution or Lóp ez-González, Mena-Moreno, Mestre-Bach, Valero-Solı ́s, Rivas, Agüera, reproduction is permitted which does not comply with these terms. Frontiers in Psychiatry | www.frontiersin.org 13 May 2020 | Volume 11 | Article 482