Introduction
Gambling disorder (GD) and eating disorders (EDs) are mental conditions characterised by persistent maladaptive patterns of gambling behaviour and abnormal eating, respectively [1]. Despite representing very different entities with specific clinical symptoms, these disorders share numerous psychopathological, neurobiological, and comorbidity features [Reference Kim, von Ranson, Hodgins, McGrath and Tavares2–Reference Lemón, Fernández-Aranda, Jiménez-Murcia and Håkansson6]. For instance, both disorders are distinguished by elevated behavioural disinhibition [Reference Dawe and Loxton7, Reference Petry8], great levels of negative urgency [Reference Fischer and Smith9, Reference Farstad, Von Ranson, Hodgins, El-Guebaly, Casey and Schopflocher10], as well as difficulties in executive function [Reference Álvarez-Moya, Jiménez-Murcia, Moragas, Gómez-Peña, Aymamí and Ochoa11] and emotion regulation [Reference Vintró-Alcaraz, Munguía, Granero, Gaspar-Pérez, Solé-Morata and Sánchez5]. At a neurobiological level, there is a dysregulation of the dopaminergic and serotonin systems, which are respectively involved in the reward and impulse control systems [Reference Fineberg, Potenza, Chamberlain, Berlin, Menzies and Bechara12–Reference Hodgins, Stea and Grant15]. GD and BSDs also display high levels of co-occurrence [Reference Fernández-Aranda, Jiménez-Murcia, Álvarez-Moya, Granero, Vallejo and Bulik16–Reference Von Ranson, Wallace, Holub, Hodgins, Ranson and Wallace18] and usually overlap with the same pathologies, especially with impulse control disorders and substance abuse [Reference Grant, Potenza, Weinstein and Gorelick19–Reference Krug, Treasure, Anderluh, Bellodi, Cellini and di Bernardo23]. Finally, GD and EDs are distinguished by the presence of inappropriate behaviours related to impulsivity and compulsivity [Reference Farstad, Von Ranson, Hodgins, El-Guebaly, Casey and Schopflocher10, Reference Jiménez-Murcia, Granero, Moragas, Steiger, Israel and Aymamí22, Reference Frascella, Potenza, Brown and Childress24, Reference Hoffman, Gagne, Thornton, Klump, Brandt and Crawford25], and are considered to map into the impulsive–compulsive spectrum, a continuum encompassing different neuropsychiatric conditions characterised by impairments in impulse control mechanisms [Reference Allen, King and Hollander26, Reference Mestre-Bach, Fernández-Aranda, Jiménez-Murcia and Potenza27].
Recent research suggests that impulsivity (i.e., rapid unplanned responses performed without regard to their negative consequences) can be conceptualised as a multidimensional construct composed of three major domains: ‘choice impulsivity’, ‘response impulsivity’, and ‘impulsive personality traits’ [Reference MacKillop, Weafer, Gray, Oshri, Palmer and de Wit28, Reference Testa, Granero, Misiolek, Vintró-Alcaraz, Mallorqui-Bagué and Lozano-Madrid29]. Choice impulsivity is understood as impulsive decision-making without neither planning nor regard for future consequences [Reference MacKillop, Weafer, Gray, Oshri, Palmer and de Wit28], and is usually measured with the Iowa Gambling Task (IGT) [Reference Bechara, Damasio, Damasio and Anderson30]. Numerous studies have shown decision-making impairments in all EDs, but especially in bulimic spectrum disorders (BSDs) [Reference Aloi, Rania, Caroleo, Bruni, Palmieri and Cauteruccio31–Reference Testa, Mora-Maltas, Camacho-Barcia, Granero, Lucas and Agüera33], meaning higher impulsive choice in these ED subtypes. Similarly, decision-making deficits have been observed in GD [Reference Mallorquí-Bagué, Tolosa-Sola, Fernández-Aranda, Granero, Fagundo and Lozano-Madrid34, Reference Ochoa, Alvarez-Moya, Penelo, Aymami, Gómez-Peña and Fernández-Aranda35], mainly when taking risk–reward choices, in which GD patients display high difficulty in choosing the advantageous options [Reference Wiehler and Peters36].
Secondly, response impulsivity would reflect the inability to inhibit a strong motor response, meaning a low inhibitory control [Reference MacKillop, Weafer, Gray, Oshri, Palmer and de Wit28]. The Stroop Colour and Word Test (SCWT) [Reference Golden37, Reference Golden38] has shown poor inhibitory control and high response impulsivity in patients with EDs, mainly in BSD patients [Reference Lavagnino, Arnone, Cao, Soares and Selvaraj39, Reference Wu, Hartmann, Skunde, Herzog and Friederich40]. Inhibitory control deficits are particularly outstanding when examining stimuli related to the disorder (e.g., food consumption) [Reference Qian, Hu, Wan, Li, Wu and Ren41], suggesting they may constitute a maintenance factor in BSDs [Reference Eichen, Matheson, Appleton-Knapp and Boutelle42]. GD has also been linked with decreased inhibitory control, especially in certain cognitively demanding situations (e.g., gambling activities) [Reference Goudriaan, Oosterlaan, De Beurs and Van Den Brink43, Reference Kertzman, Vainder, Aizer, Kotler and Dannon44].
Impulsive personality traits are the last form of impulsivity, and refer to dispositional tendencies towards impulsive behaviour [Reference MacKillop, Weafer, Gray, Oshri, Palmer and de Wit28]. Novelty seeking, assessed using the Temperament and Character Inventory-Revised (TCI-R) [Reference Cloninger, Przybeck, Svrakic and Wetzel45], is the trait most closely related to impulsivity, since it refers to the search for new stimuli and rewards. Among EDs, the highest levels of novelty seeking are displayed by patients with BSDs [Reference Cassin and Von Ranson46–Reference Munguía, Lucas, Jiménez-Murcia, Mora-Maltas, Granero and Miranda-Olivos48], reflecting a more impulsive personality. Similarly, high novelty seeking has been identified as an outstanding feature in GD [Reference Mallorquí-Bagué, Fernández-Aranda, Lozano-Madrid, Granero, Mestre-Bach and Baño49, Reference Del Pino-Gutiérrez, Jiménez-Murcia, Fernández-Aranda, Agüera, Granero and Hakansson50]. Lastly, another impulsive trait highlighted in GD and particularly in BSDs is low persistence [Reference Fernández-Aranda, Granero, Mestre-Bach, Steward, Müller and Brand51], indicating lack of perseverance in the presence of adversities [Reference Cloninger, Przybeck, Svrakic and Wetzel45].
Similarly to impulsivity, compulsivity (i.e., repetitive maladaptive behaviours performed to avoid negative consequences) can be described as a multifactorial construct made up of different dimensions [Reference Hollander52, Reference Berlin and Hollander53]. Cognitive flexibility is one of the most relevant components of compulsivity. It refers to the ability to adapt cognitive strategies in response to feedback [Reference Potenza54], and is commonly measured by the Wisconsin Card Sorting Task (WCST) [Reference Heaton and Staff55]. Numerous studies have displayed poor cognitive flexibility in patients with BSDs [Reference Aloi, Rania, Caroleo, Bruni, Palmieri and Cauteruccio31, Reference Roberts, Tchanturia and Treasure56, Reference Tchanturia, Liao, Forcano, Fernández-Aranda, Uher and Treasure57] and GD [Reference Álvarez-Moya, Jiménez-Murcia, Moragas, Gómez-Peña, Aymamí and Ochoa11, Reference Forbush, Shaw, Graeber, Hovick, Meyer and Moser58], reflecting high levels of cognitive rigidity and compulsivity in both disorders. On the other hand, due to the lack of psychometric tools to measure compulsivity, some authors have suggested the use of harm avoidance (TCI-R) to assess compulsive traits [Reference Potenza54, Reference Tavares and Gentil59]. In this regard, several studies have shown that BSD and GD patients display higher levels of harm avoidance than healthy controls (HCs) [Reference Cassin and Von Ranson46, Reference Fassino, Daga, Pierò, Leombruni and Rovera47, Reference Claes, Jimenez-Murcia, Agüera, Sánchez, Santamaría and Granero60].
Despite the similarities between GD and BSDs, only a few studies have explored impulsivity and compulsivity comparing both disorders. Studies focused on personality traits observed that GD patients show higher novelty seeking than BSD patients [Reference Fernández-Aranda, Jiménez-Murcia, Álvarez-Moya, Granero, Vallejo and Bulik16, Reference Jiménez-Murcia, Granero, Moragas, Steiger, Israel and Aymamí22, Reference Del Pino-Gutiérrez, Jiménez-Murcia, Fernández-Aranda, Agüera, Granero and Hakansson50, Reference Fernández-Aranda, Granero, Mestre-Bach, Steward, Müller and Brand51, Reference Álvarez-Moya, Jiménez-Murcia, Granero, Vallejo, Krug and Bulik61], meaning a higher tendency to search for new stimuli and rewards. On the other hand, BSD patients display lower persistence [Reference Fernández-Aranda, Granero, Mestre-Bach, Steward, Müller and Brand51] and higher harm avoidance [Reference Del Pino-Gutiérrez, Jiménez-Murcia, Fernández-Aranda, Agüera, Granero and Hakansson50, Reference Fernández-Aranda, Granero, Mestre-Bach, Steward, Müller and Brand51], indicating higher sensitivity and vulnerability to adversities. Only one research compared the neuropsychological performance of GD and BN patients [Reference Álvarez-Moya, Jiménez-Murcia, Moragas, Gómez-Peña, Aymamí and Ochoa11], revealing that GD patients exhibit poorer cognitive flexibility, which is a higher difficulty in adapting cognitive strategies in response to feedback. Finally, although sex differences are especially highlighted in these pathologies at a clinical and neuropsychological level [Reference Fernández‐Aranda, Aitken, Badía, Giménez, Solano and Collier62–Reference Mallorquí-Bagué, Lozano-Madrid, Granero, Mestre-Bach, Vintró-Alcaraz and Sánchez66], most of the cited studies did not consider the effect of sex on their results, possibly leading to a bias.
Given that numerous impulsive and compulsive features overlap across GD and BSDs, there is a need for comparative studies which examine the differences and similarities between both. Thus, this study aimed to compare impulsivity and compulsivity dimensions among HCs and patients with BSDs and GD. Firstly, we hypothesised that both clinical groups would display abnormalities in these domains relative to HCs. Based on previous studies, we also expected to find some differences between GD and BSDs, such as lower cognitive flexibility and higher novelty seeking in GD, but lower persistence and higher harm avoidance in BSDs. Secondly, we aimed to examine sex effects on the impulsivity and compulsivity of GD and BSD patients.
Methodology
Participants
The final sample consisted of 368 participants (52.2% females) aged between 18 and 72 years old, of which 59 were diagnosed with a BSD (62.7% females) and 159 with a GD (11.6% females) according to the DSM-5 diagnostic criteria [1]. The BSD subsample comprised of bulimia nervosa (47.5%) and binge ED patients (52.5%). The control group was made up of 150 individuals without a lifetime mental disorder.
Participants included in the clinical groups were referred to assessment and treatment at the Eating Disorder Unit and the Pathological Gambling Unit of the Department of Clinical Psychology of the Bellvitge University Hospital in Barcelona. They were diagnosed by psychologists/psychiatrists with over 10 years’ experience by means of a semi-structured interview. Healthy controls were recruited via word-of-mouth and advertisements located in the same geographic area. As they arrived at the Eating Disorder and Pathological Gambling Units, they were interviewed by psychologists/psychiatrists to verify their eligibility for the study. They could not have a lifetime history of an ED, current obesity or any behavioural addiction. Healthy controls were also matched for age and education level with participants of the clinical groups.
All participants were informed about the research procedures and gave their informed consent in writing. Exclusion criteria were the following: (1) history of chronic medical illness or neurological condition that might affect cognitive function; (2) head trauma with loss of consciousness for more than 2 min; (3) learning disability or intellectual disability; (4) use of psycho-active medications or drugs; (5) age under 18.
Procedure and assessment
All individuals who arrived at the hospital and were diagnosed with an ED or a GD were screened for the inclusion criteria. Those included in the study underwent a clinical and neuropsychological assessment within the first week of their outpatient treatment. The instruments included were specifically selected to cover a wide range of impulsivity and compulsivity domains (see Supplementary Table S1). Additional sociodemographic information was also taken.
Instruments
The Spanish adaptation of the Symptom Checklist-90 Revised (SCL-90-R) [Reference Derogatis67, Reference González de Rivera, De las Cuevas, Rodríguez and Rodríguez68] assesses psychopathological symptoms grouped as follows: somatisation, obsessive-compulsive, interpersonal sensitive, depression, anxiety, hostility, phobic anxiety, paranoia, and psychotic. Additionally, this scale contains a global severity index (GSI) that measures overall psychological distress. Cronbach’s alpha in this study was good to excellent (between a = 0.77 for phobic anxiety and a = 0.93 for depressive symptoms).
The Spanish validation of the TCI-R [Reference Cloninger, Przybeck, Svrakic and Wetzel45, Reference Gutiérrez-Zotes, Bayón, Montserrat, Valero, Labad and Cloninger69] is a reliable questionnaire composed of 240 items. It includes four dimensions of temperament (novelty seeking, harm avoidance, reward dependence, and persistence) and three of character (self-directedness, cooperativeness, and self-transcendence). In this study sample, Cronbach’s alpha was good to excellent (between a = 0.73 for novelty seeking and a = 0.89 for self-directedness).
The SCWT [Reference Golden37, Reference Golden70] is a neuropsychological tool that evaluates inhibitory control. It comprises three different lists: a word list containing names of colours, a colour list that comprises letter Xs printed in colour, and a colour-word list composed of names of colours in a colour ink that does not match the written name. Three scores are calculated using the number of items correctly read from each list in 45 s. An additional ‘interference score’ is obtained using all three lists. Better capacity of inhibitory control is related to higher scores in the colour-word list and in the interference score.
The Wisconsin Card Sorting Test (WCST) [Reference Heaton and Staff55] is a computerised set-shifting task which allows the assessment of cognitive flexibility. Participants have to sort each of the 128 cards provided in one of the four available decks. Each card presents a figure with different properties (colour, shape, and number). The aim is to select the correct deck considering one of the properties. When participants select a deck, feedback (‘correct’ or ‘incorrect’) is displayed, hence they can deduce the selection criteria. The rule changes after completing a category (10 consecutive correct sorts) or if this is not discovered after six trials. The task finishes when all 128 cards are sorted or after six categories are completed. The main variable related to compulsivity is perseverative errors (i.e., failures to change sorting strategy after negative feedback).
The IGT [Reference Bechara, Damasio, Damasio and Anderson30] is a computerised task designed to assess decision-making processes and impulsive choices [Reference Eisinger, Anna, Máté, Edina, Zsolt and Gyöngyi71]. The task involves 100 trials in which participants have to select one of the four presented decks (A, B, C, or D); afterwards, a specified amount of play money is awarded or subtracted. Two of the decks result in money wins (C, D), while the others result in losses (A, B). Participants are instructed to gather as much money as possible. The test score is computed by subtracting the number of times that participants selected the disadvantageous decks from the number of advantageous deck choices. Lower scores translate to impulsive decision-making.
Statistical analysis
Statistical analysis was carried out with Stata16 for Windows [72]. Firstly, the frequency distribution of the sociodemographic and clinical variables was based on chi-square tests (χ 2) for categorical variables and analysis of variance (ANOVA) for quantitative measures. The effect size of the mean differences was tested through Cohen’s d coefficient, considering low-poor effect for |d| > 0.20, moderate-medium for |d| > 0.5, and large-high for |d| > 0.8 [Reference Kelley and Preacher73]. The effect size of the proportion differences was estimated through Cohen’s h coefficient, obtained as the difference of the arcsine transformation for the proportions observed in each group; it was interpreted with the same threshold ranges as Cohen’s d measure [Reference Cohen74]. The comparison of the impulsivity and compulsivity measures was based on 3 × 2 ANOVA procedures, controlling for the covariates chronological age and education. Two between-subjects factors were defined: group (with three levels: HCs, BSDs, and GD) and sex (with two levels: women and men). In addition, Finner correction was used to control the increase in the Type I error due to the multiple statistical comparisons [Reference Finner75], which is a family-wise procedure that has demonstrated higher power than classical Bonferroni correction.
Ethics
The study procedures were carried out in accordance with the Declaration of Helsinki. The Clinical Research Ethics Committee of Bellvitge University Hospital approved the study.
Results
Characteristics of the sample
Table 1 contains the comparison of the sociodemographic and clinical profile of the three groups. Statistical differences emerged for all the measures among all three groups, with most effect sizes within the moderate to high range (|d| > 0.50 or |h| > 0.50).
BSD, bulimic spectrum disorders; GD, gambling disorder; HC, healthy controls; SD, standard deviation.
a Significant comparison (.05 level).
b Effect size into the range mild–moderate (|h| or |d| higher than 0.50) to high-large (|h| or |d| higher than 0.80).
Comparison of the impulsivity and compulsivity measures
Table 2 includes the results of the 3 × 2 ANOVA comparing the impulsivity and compulsivity measures among the three groups, controlling for age and education. Multivariate tests showed a significant effect of the factor group in most variables. When single effects were estimated, differences between HCs and the clinical groups were found for most measures, except for the following: IGT block 1, SCWT interference, TCI-R reward dependence, and WCST conceptual. Comparisons between BSDs and GD reported differences only for TCI-R novelty seeking (higher in GD) and harm avoidance (higher in BSDs).
BSD, bulimic spectrum disorders; GD, gambling disorder; HC, healthy controls; SD, standard deviation.
a Significant parameter.
b p-values.
Regarding the sex factor, differences between men and women were obtained for all the WCST measures, except for WCST trials, with men showing a better performance relative to women. Compared to women, men achieved lower scores in SCWT colours, TCI-R harm avoidance, and reward dependence, but higher levels in TCI-R persistence.
Multivariate tests also displayed a significant interaction sex-by-group for some variables: IGT block 2 and 4, IGT total, SCWT words, and SCWT colours. Single effects were estimated (see Supplementary Table S2).
Figure 1 contains the learning curve of each group in the IGT, HCs reporting the best performance. When the two clinical groups were compared, GD displayed a trend towards better performance.
As a summary, Figure 2 illustrates the radar chart for the variables measuring impulsivity and compulsivity features (z-standardised means are plotted to allow an easier interpretation).
Discussion
In this study we aimed to compare, through self-reported and neurocognitive measures, impulsivity and compulsivity among HCs, GD, and BSDs, emphasising the differences and similarities between the two clinical groups. We also aimed to analyse the potential influence of sex in these impulsive and compulsive domains.
Firstly, when exploring general psychopathology, we observed notable differences among the three groups in all the SCL-90-R subscales. In agreement with previous studies [Reference Fernández-Aranda, Jiménez-Murcia, Álvarez-Moya, Granero, Vallejo and Bulik16, Reference Jiménez-Murcia, Granero, Moragas, Steiger, Israel and Aymamí22, Reference Fernández-Aranda, Granero, Mestre-Bach, Steward, Müller and Brand51], HCs exhibited the lowest levels of global psychopathology and specific symptomatology, while BSD patients displayed the highest levels. This could be interpreted as an increased perception of psychopathological distress in BSDs.
Regarding choice impulsivity, our results illustrated that GD and BSD patients had worse IGT performance than HCs, indicating a reduced capacity to learn the reward/punishment contingencies of their choices [Reference Bechara, Damasio, Damasio and Lee76]. In addition, BSD patients performed slightly worse than patients with GD, demonstrating a trend to more impulsive decision-making. Attending to earlier findings, it was expected that both clinical groups presented difficulties in decision-making [Reference Aloi, Rania, Caroleo, Bruni, Palmieri and Cauteruccio31, Reference Guillaume, Gorwood, Jollant, Van den Eynde, Courtet and Richard-Devantoy32, Reference Munguía, Lucas, Jiménez-Murcia, Mora-Maltas, Granero and Miranda-Olivos48, Reference Kovács, Richman, Janka, Maraz and Andó77]. Evidence also indicates that GD and BSD patients share a preference for immediate rewards (i.e., binging or gambling) regardless of future consequences [Reference Grecucci, Giorgetta, Rattin, Guerreschi, Sanfey and Bonini78, Reference Manwaring, Green, Myerson, Strube and Wilfley79]. This might reflect their increased choice impulsivity, which appears to be slightly enhanced in BSDs. When comparing women and men, significantly increased choice impulsivity was displayed by women only in GD, although BSD women also showed a trend to higher choice impulsivity than men. Previous studies failed to find sex differences in the decision-making of AN patients [Reference Tchanturia, Liao, Forcano, Fernández-Aranda, Uher and Treasure57], indicating that sex influences might be only noticeable in ED patients with binge-purging symptoms. Regarding GD, several metanalyses failed to find any sex effect on this domain [Reference Kovács, Richman, Janka, Maraz and Andó77, Reference Ioannidis, Hook, Wickham, Grant and Chamberlain80], which is likely to be explained by methodological issues (e.g., metanalyses compared female samples to male samples). We also observed that men exhibited similar choice impulsivity regardless of the diagnostic group; however, among women, both clinical groups displayed higher choice impulsivity than HCs. Altogether, our results indicate that enhanced choice impulsivity might be highlighted only in female patients.
Another dimension of impulsivity examined in this work was response impulsivity. Our results revealed that both clinical groups showed lower inhibitory control than HCs, as observed in the SCWT word-colour scores. Recent research also noted significant inhibitory control deficits in GD and BSDs [Reference Lavagnino, Arnone, Cao, Soares and Selvaraj39, Reference Wu, Hartmann, Skunde, Herzog and Friederich40, Reference Kertzman, Vainder, Aizer, Kotler and Dannon44], and the only study which compared both disorders also found analogous outcomes [Reference Álvarez-Moya, Jiménez-Murcia, Moragas, Gómez-Peña, Aymamí and Ochoa11]. These similarities in response impulsivity found in GD and BSDs may be related to common underlying processes involving impairments in behavioural inhibition systems [Reference Goodman3]. Finally, women and men did not differ in inhibitory control, neither in the whole sample nor within each clinical group. Although a recent study has shown lower inhibitory control in women with BSDs [Reference Mallorquí-Bagué, Lozano-Madrid, Granero, Mestre-Bach, Vintró-Alcaraz and Sánchez66], there is very little evidence to examine these effects in GD, presenting similar evidence to this work [Reference Mallorquí-Bagué, Mestre-Bach, Lozano-Madrid, Granero, Vintró-Alcaraz and Fernández-Aranda64, Reference Ioannidis, Hook, Wickham, Grant and Chamberlain80].
The results of this work also illustrated that, in terms of impulsive personality traits, patients with GD showed the highest levels of novelty seeking in the TCI-R, differing from BSD patients and HCs. In addition, BSD patients showed lower levels of persistence compared to HCs. It is noted that, although BSDs are considered to be the EDs with the highest impulsive personality [Reference Cassin and Von Ranson46, Reference Fassino, Daga, Pierò, Leombruni and Rovera47], BSD patients displayed similar novelty seeking to HCs. Our finding is consistent with former investigations [Reference Fassino, Daga, Pierò, Leombruni and Rovera47, Reference Fernández-Aranda, Granero, Mestre-Bach, Steward, Müller and Brand51, Reference Claes, Jimenez-Murcia, Agüera, Sánchez, Santamaría and Granero60, Reference Álvarez-Moya, Jiménez-Murcia, Granero, Vallejo, Krug and Bulik61] and suggests that novelty seeking might not be a core factor of impulsive personality in BSDs; other traits, such as low persistence, seem to be more relevant. As a whole, these findings give support to previous research [Reference Fernández-Aranda, Jiménez-Murcia, Álvarez-Moya, Granero, Vallejo and Bulik16, Reference Jiménez-Murcia, Granero, Moragas, Steiger, Israel and Aymamí22, Reference Del Pino-Gutiérrez, Jiménez-Murcia, Fernández-Aranda, Agüera, Granero and Hakansson50, Reference Fernández-Aranda, Granero, Mestre-Bach, Steward, Müller and Brand51, Reference Álvarez-Moya, Jiménez-Murcia, Granero, Vallejo, Krug and Bulik61] and uphold the notion that impulsive personality in GD and BSDs appears to be from a different nature, with high novelty seeking as the distinctive trait in GD, and low persistence in BSDs. Concerning sex, no differences in novelty seeking were found within the clinical groups, although lower persistence was displayed by women in the GD group. Prior studies in GD and EDs partially support our findings, since no sex differences were yielded in any of these impulsive traits [Reference Fernández‐Aranda, Aitken, Badía, Giménez, Solano and Collier62, Reference Granero, Penelo, Martínez-Giménez, Álvarez-Moya, Gómez-Peña and Aymamí63]. In addition, although earlier studies reported novelty-seeking differences between GD and BSDs in both sexes [Reference Fernández-Aranda, Jiménez-Murcia, Álvarez-Moya, Granero, Vallejo and Bulik16, Reference Jiménez-Murcia, Granero, Moragas, Steiger, Israel and Aymamí22, Reference Claes, Jimenez-Murcia, Agüera, Sánchez, Santamaría and Granero60, Reference Álvarez-Moya, Jiménez-Murcia, Granero, Vallejo, Krug and Bulik61], our study only displayed differences among men. This might indicate that, regarding impulsive traits, GD and BSDs may be more closely associated in females than in males. However, future research exploring sex differences in these disorders would be required to make solid assumptions.
Another finding to emerge from the present study are the differences in cognitive flexibility. Compared to HCs, patients with GD and BSDs displayed a worse performance in most measures of the WCST, including perseverative errors, meaning less cognitive flexibility and more compulsivity. This outcome is in line with previous research [Reference Álvarez-Moya, Jiménez-Murcia, Moragas, Gómez-Peña, Aymamí and Ochoa11, Reference Aloi, Rania, Caroleo, Bruni, Palmieri and Cauteruccio31, Reference Roberts, Tchanturia and Treasure56, Reference Forbush, Shaw, Graeber, Hovick, Meyer and Moser58, Reference Duchesne, Mattos, Appolinário, De Freitas, Coutinho and Santos81–Reference Tchanturia, Davies, Roberts, Harrison, Nakazato and Schmidt83] and might explain the inability of GD and BSD patients to learn from mistakes and handle negative affect, increasing loss of control over gambling or eating [Reference Marazziti, Catena Dell’Osso, Conversano, Consoli, Vivarelli and Mungai84–Reference Wu, Brockmeyer, Hartmann, Skunde, Herzog and Friederich87]. Contrary to our hypothesis, we did not succeed in identifying greater cognitive flexibility impairments in GD compared to BSDs. A former work had revealed such results [Reference Álvarez-Moya, Jiménez-Murcia, Moragas, Gómez-Peña, Aymamí and Ochoa11], although they were considered preliminary, due to the fact that the sample was considerably small and only included females. Since comparative studies of compulsivity in BSDs and GD are extremely scarce, solid conclusions still cannot be made. Focusing on sex effects, worse cognitive flexibility in females than males was observed in the whole sample. Nevertheless, when examining each diagnostic group, this pattern only remained present in the GD group. Only a few studies had explored sex differences in compulsivity between women and men with BSDs and GD, showing women lower cognitive flexibility in both pathologies [Reference Mallorquí-Bagué, Mestre-Bach, Lozano-Madrid, Granero, Vintró-Alcaraz and Fernández-Aranda64, Reference Mallorquí-Bagué, Lozano-Madrid, Granero, Mestre-Bach, Vintró-Alcaraz and Sánchez66]. Moreover, when comparing GD and BSDs within each sex, the clinical groups showed lower cognitive flexibility than HCs in both sexes. As a whole, these results reveal that females with GD appear to have more compulsivity features in common with BSD females than with GD males.
Regarding compulsive personality traits, differences in harm avoidance were observed among all three groups, with BSD patients showing the highest scores, followed by GD patients. Our findings dovetail with previous reports and uphold that compulsive traits are a distinguishing feature in these pathologies [Reference Claes, Jimenez-Murcia, Agüera, Sánchez, Santamaría and Granero60], especially in BSDs [Reference Munguía, Lucas, Jiménez-Murcia, Mora-Maltas, Granero and Miranda-Olivos48, Reference Del Pino-Gutiérrez, Jiménez-Murcia, Fernández-Aranda, Agüera, Granero and Hakansson50, Reference Fernández-Aranda, Granero, Mestre-Bach, Steward, Müller and Brand51]. Therefore, the elevated compulsivity noted in these pathologies is likely to underlie the use of abnormal behaviours (e.g., binge eating, compensatory purging, gambling) with the aim of avoiding negative affect and relieving emotional distress [Reference Frascella, Potenza, Brown and Childress24, Reference Hoffman, Gagne, Thornton, Klump, Brandt and Crawford25]. Finally, when we explored sex effects on harm avoidance, higher scores were displayed by women in the overall sample and within the GD group. These results are in line with previous research [Reference Granero, Penelo, Martínez-Giménez, Álvarez-Moya, Gómez-Peña and Aymamí63] and with our findings in cognitive flexibility, indicating that GD women may differ from GD males and present a specific profile characterised by higher compulsivity. These findings could be indirectly related to the increased emotion regulation difficulties found in women with GD when compared to male patients [Reference Venne, Mazar and Volberg88]. Finally, differences among all three groups emerged when exploring women and men separately. As observed in the overall sample, the highest harm avoidance was displayed by BSDs, followed by GD. These results indicate that high levels of compulsive traits are noticeable in both GD and BSD patients regardless of their sex, being significantly highlighted in female and male patients with BSDs.
These findings should be interpreted considering some limitations. Firstly, our study sample comprised mostly young adults, not allowing for generalisation to older populations. In addition, since the age of participants could be a relevant factor affecting impulsivity and compulsivity in these disorders, it would be important to collect bigger samples and explore this point in future studies. Secondly, although results have been controlled for age and education, other confounding variables might be affecting. For instance, it would be useful to measure severity level following the DSM 5 criteria: mild, moderate, and severe. Having a bigger sample would allow us to find out if severity could be interacting with the variables studied (type of disorder and gender) and affecting impulsivity/compulsivity levels. Moreover, the subgroup sizes are relatively large but not evenly distributed due to the sex bias usually observed in these disorders. A bigger sample would allow us to reduce this bias and create subgroups more equally distributed. Lastly, the cross-sectional nature of this study does not allow for cause–effect inferences. Longitudinal studies are needed to explore if impulsivity and compulsivity deficits are a cause or a consequence of the disorder.
Conclusions
In conclusion, GD and BSDs seem to be distinguished by elevated levels of impulsivity and compulsivity in all the dimensions, showing differences in some domains mainly related to personality. In this regard, these pathologies show distinct impulsive personality traits, such as high novelty seeking in GD and low persistence in BSDs. Moreover, BSDs are characterised by higher compulsive personality traits as well as a trend towards greater impulsive choice. Given the lack of neuropsychological and clinical studies comparing GD and BSDs, future research would be needed to consolidate our findings. Nevertheless, clinical interventions should consider these deficits to enhance their effectiveness, for instance, including adjunctive treatment to target these difficulties, such as inhibitory control training [Reference Turton, Nazar, Burgess, Lawrence, Cardi and Treasure89], emotion regulation training [Reference Claudat, Brown, Anderson, Bongiorno, Berner and Reilly90, Reference Mena-Moreno, Munguía, Granero, Lucas, Fernández-Aranda and Gómez-Peña91], or cognitive remediation therapy [Reference Hagan, Christensen and Forbush92]. Finally, it seems that GD women present higher difficulties than GD men in almost all the impulsivity and compulsivity components, displaying in some cases more similarities with BSD women than with GD men. This finding is in line with former research providing support to the relevance of sex in GD [Reference Fattore, Melis, Fadda and Fratta93], which should also be considered in clinical interventions. For instance, GD women would benefit from a more personalised approach which targets impulse control deficits, combining pharmacological treatment, cognitive behavioural therapy, family involvement, and support groups [Reference Mestre, Strafella, Thomsen, Voon and Miyasaki94].
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1192/j.eurpsy.2023.2458.
Acknowledgements
We thank CERCA Programme/Generalitat de Catalunya for institutional support. We also thank Instituto de Salud Carlos III (ISCIII), CIBERobn (an initiative of ISCIII), FEDER funds/European Regional Development Fund (ERDF), a way to build Europe, and European Social Fund (ESF, investing in your future).
Author contribution
Conceptualisation: M.L.-M., S.J.-M., F.F.-A.; Data curation: M.L.-M.; Formal analysis: R.G.; Funding acquisition: F.F.-A., M.L.-M., S.J.-M; Investigation: I.S., J.S.-G., M.G.-P., L.M., N.M.-B., J.T.; Methodology: M.L.-M., S.J.-M., F.F.-A.; Writing—original draft preparation: M.L.-M., R.G., I.L.; Writing—review and editing: M.L.-M., S.J.-M., F.F.-A. All authors have read and agreed to the published version of the manuscript.
Financial support
This manuscript and research were supported by grants from the Delegación del Gobierno para el Plan Nacional sobre Drogas (2019I47 and 2021I031), Ministerio de Ciencia e Innovación (PDI2021-124887OB-I00), Instituto de Salud Carlos III (ISCIII) (FIS PI20/00132), European Union’s Horizon 2020 research and innovation programme under grant agreement no. 847879 (PRIME/H2020, Prevention and Remediation of Insulin Multimorbidity in Europe). Additional funding was received by AGAUR-Generalitat de Catalunya (2021-SGR-00824). M.L.-M. was supported by a predoctoral grant from the Ministerio de Educación, Cultura y Deporte (FPU15/02911). R.G. was supported by The Catalan Institution for Research and Advanced Studies (ICREA-2021 Academia Program). I.L. was supported by the Ministerio de Ciencia e Innovación (Juan de la Cierva-Formación Program, FJC2021–046494-I). The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
Competing interest
F.F.-A. and S.J.-M. received consultancy honoraria from Novo Nordisk. The rest of the authors declare no conflict of interest.
Comments
No Comments have been published for this article.