Decision-Making in Gambling Disorder, Problematic Pornography Use, and Binge-Eating Disorder: Similarities and Differences (2021)

2020 Sep;7(3):97-108.

doi: 10.1007/s40473-020-00212-7.

Abstract

Purpose of Review

The present review attempts to provide a comprehensive and critical overview of the neurocognitive mechanisms of gambling disorder (GD), problematic pornography use (PPU), and binge-eating disorder (BED), focusing specifically on decision-making processes.

Recent Findings

GD, PPU, and BED have been associated with decision-making impairments both under risk and ambiguity. Features such as intelligence, emotions, social variables, cognitive distortions, comorbidities, or arousal may condition decision-making processes in these individuals.

Summary

Impairments in decision-making seem to be a shared transdiagnostic feature of these disorders. However, there is varying support for the degree to which different features may affect decision-making. Therefore, the study of decision-making processes can provide crucial evidence for understanding addictions and other disorders with addiction-like symptomatology.

Introduction

Behavioral addictions and eating disorders (EDs) are significant public health concerns worldwide [1]. Increases in gambling opportunities (with legalization of online gambling in many jurisdictions), the increased availability and affordability of pornographic materials, and the instantiation of eating habits strongly associated with more sedentary lifestyles and accessibility of high-calorie palatable foods have impacted addictive behaviors and disorders (especially gambling disorder (GD) and problematic pornography use (PPU)) and EDs (especially binge-eating disorder (BED)) [2,3,4].

Common mechanisms underlying substance-use disorders (SUDs such as alcohol, cocaine, and opioids) and addictive or maladaptative disorders or behaviors (such as GD and PPU) have been suggested [5,6,7,8, 9••]. Shared underpinnings between addictions and EDs have also been described, mainly including top-down cognitive-control [10,11,12] and bottom-up reward-processing [13, 14] alterations. Individuals with these disorders often show impaired cognitive control and disadvantageous decision-making [12, 15,16,17]. Deficits in decision-making processes and goal-directed learning have been found across multiple disorders; thus, they could be considered clinically relevant transdiagnostic features [18,19,20]. More specifically, it has been suggested that these processes are found in individuals with behavioral addictions (e.g., in dual-process and other models of addictions) [21,22,23,24].

Regarding the addiction model, GD has been studied in greater depth and has even been classified in the category “substance-related and addictive disorders” of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) [1]. However, in the case of BED and especially PPU, the existing literature is limited, particularly in neurocognition and neuroscience. Understanding of neurocognitive mechanisms underlying these psychiatric disorders has been slower, and fewer neurobiological models have been proposed, and those that have been cite decision-making as relevant [23, 25, 26].

Recent studies have suggested a biopsychosocial explanatory model of BED, where different factors (such as a genetic susceptibility to food reward, chronic stress, and specific features of highly processed foods with high level of fats and sugars) would promote a behavioral pattern of dysfunctional intake and alterations in dopamine levels, facilitating the learning of erroneous eating behaviors [27]. Therefore, some authors claim that the intake of certain high-calorie food and addictive drugs produce similar neural responses, linked to reward pathways modulated by dopamine [28, 29], and could contribute to develop an addiction [30]. Similar neurobiological features have been identified between BED and GD [31, 32], such as diminished ventral striatal activity during anticipatory phases of reward processing, which may be considered a biomarker associated with addictive processes [33]. BED has also shown similarities with food addiction, such as diminished control over consumption, excessive and continued consumption patterns despite negative consequences, and difficulties reducing the frequency or quantity of consumption [34,35,36].

There is considerable debate whether PPU and compulsive sexual behaviors (CSBs) more generally should be considered as a behavioral addiction (37••, 38). CSB disorder (CSBD) has recently been incorporated in the eleventh revision of the International Classification of Diseases (ICD-11) as an impulse-control disorder [39]. Similarities between CSBD and addictions have been described, and impaired control, persistent use despite adverse consequences, and tendencies to engage in risky decisions may be shared features (37••, 40). While some authors have contended that based on similarities in behavioral neuroscientific and other features—such as the possible involvement of the reward system and the prefrontal–striatal circuits in cognitive control over motivational brain circuitry—that CSBD and PPU should be classified as addictive disorders [41], the addictive nature of sexually explicit materials remains debated.

The addiction model requires more data about possible transdiagnostic clinical features. A lack of consensus regarding this theoretical framework has hindered BED and especially PPU becoming a more substantial part of clinical debate. Therefore, the present review attempts to provide a comprehensive and critical overview of neurocognitive mechanisms, focusing specifically on decision-making processes [42].

Decision-Making in GD, PPU, and BED

The DSM-5 establishes six neurocognitive domains that have been studied in the field of addictions and EDs: complex attention, social cognition, learning and memory, language, perceptual-motor function, and executive function [1, 43]. Among them, special interest has been given to executive functioning, delving into planning, cognitive flexibility, inhibition, responding to feedback, and decision-making [44••, 45, 46].

The specific conceptualization of the decision-making construct is controversial and has led to heterogeneous definitions, limiting the generalization of results. Decisions, even those linked with a potentially addictive behavior, result from a competition between different possible actions for behavioral expression [47]. Instrumental behaviors may be less sensitive to contingency manipulations over time, if they turn into addictive behaviors [47]. Therefore, decision-making may be understood as a complex set of processes that promotes the choice of the most optimal behavior, contemplating the possible alternatives [48]. Decision-making may involve both habitual or “automatic” and deliberate processes [49]. The former are typically quicker and more effortless, while top-down executive-control processes are typically goal-dependent, slower, and effortful [50]. Executive-control processes may allow individuals to avoid distracting information from the environment and to suppress actions or habits [50, 51]. However, the impairment of these executive-control processes may lead to activation of habitual processes in guiding behavior [50].

Distinctions have been made regarding decision-making under objective and ambiguous risk conditions [52, 53]. In decision-making under objective risk, measured with tasks such as the Columbia Card Task [54] and the Probability-Associated Gambling Task [52], individuals have information on probabilities and explicit rules associated with each option. Therefore, decision-making processes may involve considerable reasoning. However, decisions under ambiguity are missing information about probabilities or possible associated consequences. Therefore, emotional experiences may contribute considerably in analyses of possible punishments or rewards linked with each option. They are often more uncertain, may be perceived as more aversive [55], and are associated with intuitive processes. Decisions under ambiguity are commonly assessed using the Iowa Gambling Task (IGT), where decisions can result in immediate and high rewards that are associated to greater losses in the long term. The IGT involves learning as well. Poor performance on the IGT typically involves greater sensitivity to immediate rewards, without learning from or contemplating probable losses [44••]. Therefore, the findings on decision-making under ambiguity included in the present review used the IGT as the main assessment tool.

Impulsivity and decision-making are related, and some studies intermingle delay-discounting and decision-making processes. Delay discounting is related to choice impulsivity [56] and refers to the tendency to select smaller-immediate rewards over larger-later rewards [56, 57]. While delay-discounting tasks involve decision-making, they involve sequential selection of one of two rewards of differing magnitudes separated in time. Individuals with high levels of choice impulsivity show greater tendencies not to consider the longer-term consequences of their decisions and to focus on shorter-term rewards [58].

The present review focuses on decision-making in 3 conditions: GD, PPU, and BED. Precise boundaries between the constructs of decision-making and choice impulsivity are not entirely distinct. In this review, we will review decision-making under ambiguity as measured by the IGT and decision-making under more defined contingencies as measured by delay-discounting tasks. We have tabulated main findings (Table 1).

Table 1 Summary of the main studies

Decision-Making and GD

Decision-making processes that underpin gambling share similarities with those underlying day-to-day choices [59]. They may be conceptualized as cost/benefit decisions, based on choosing between risking losing things of value and obtaining greater rewards [59]. In general, individuals usually prefer to gamble in risky than in ambiguous ways, since in decision-making processes, ambiguity is often perceived as more aversive than risk [55]. However, individual differences in personalities or tendencies (e.g., punishment insensitivity and sensation-seeking) and cognitive factors (e.g., reversal learning inflexibility) may influence decision-making in individuals with GD [60]. Moreover, although specific influences of variables such as age, sex, or educational level have not often been directly linked to decision-making deficits in GD [58], features including intelligence, emotions, social variables, cognitive distortions, cognitive processing, comorbidities, length of abstinence, or arousal may also condition decision-making [50, 55, 58, 61, 62].

Social and emotional factors are usually integrated into decision-making processes. In a recent study evaluating decision-making processes in poker players, it was observed that when participants experienced anger, they made mathematically poorer decisions [61]. Moreover, the social nature of some forms of gambling, and more specifically the social identity of some people who gamble (e.g., on poker), may have a significant moderating influence on the expression of emotions and decision-making processes [61].

In assessing the specific role of arousal in risk and ambiguity decision-making, notable differences have been observed. In the case of decisions under risk, arousal is usually closely associated with the choice of safer options, when the risk is high and the probability of winning is low, thus decreasing gambling behavior [55]. However, in the case of decisions under ambiguity, arousal may present a qualitatively different nature and is often associated with increased gambling [55]. Therefore, arousal may condition the perception of value in decisions involving greater or lesser degrees of uncertainty [55].

Individuals with gambling problems often wager large amounts and exhibit difficulties ceasing betting, and control and appetitive centers may contribute to decisions to gamble. Cognitive training that includes response inhibition may alter the amounts wagered, as well as stopping behaviors that may generalize beyond gambling [50].

Decision-making processes in the context of GD may also involve erroneous beliefs and cognitive distortions that may promote overconfidence in the ability to predict and control wins and losses, the denial of luck and chance, and generate high expectancies of winning [63,64,65,66]. Gender differences in cognitive distortions have been reported [67], with females showing more magical thinking and procrastination and procrastination mediating the association between magical thinking and GD. The gender-related difference may explain tendencies for women to rely more on luck than on skill during gambling [67].

Overactivation of motivational and valuation networks has been reported in GD, with individuals presenting greater risk-seeking and a focus on immediate rewards [68, 69]. Both tendencies may influence decision-making and delay discounting [68,69,70]. Specifically, links between risk-seeking and delay discounting were driven by GD status, and factors specific to the disorder, such as illusion of control, may contribute [68]. Other studies have also highlighted the relevance of factors such as age in the association between delay discounting and GD, with younger individuals showing relationships between forms of impulsivity [71].

Laboratory-based decision-making studies have shown that individuals with GD exhibit decision-making impairments both under risk and ambiguity. They typically perform more poorly than comparison subjects on the IGT (although not always [72]), preferring short-term rewards, even if they are not profitable in the long term, evidencing insensitivity to the future consequences of their gambling behavior [73,74,75,76]. In spite of making more disadvantageous choices, individuals with GD often learn from feedback more slowly than do comparison subjects [77, 78]. Disadvantageous decision-making on the IGT may relate to loss-chasing behaviors [74]. Some authors have found that the relationship between IGT performance and GD severity is mediated by loss chasing, the tendency to continue to bet in attempts to recover previous losses [74]. Others have reported that disadvantageous decision-making may involve diminished striatal signaling during reward and loss prospect and may operate across individuals with and without GD [72]. In adolescents, a correlation between disadvantageous decision-making and problem gambling was observed [64]. Disadvantageous decision-making on the IGT was linked to interpretative biases, a cognitive distortion characterized by tendencies to associate losses with bad luck and gains with personal skill. Both factors, along with alcohol consumption, were powerful predictors of problem-gambling severity in adolescents.

Although most studies of decision-making in GD have focused on the outcomes derived from decisional processes, individual differences in habitual response patterns may also contribute [79•]. Decision-making styles are related to cognitive styles, and rational, intuitive, dependent, avoidant, and spontaneous styles have been described [80, 81]. Problem-gambling severity has been positively related to spontaneous decision-making styles and negatively to rational decision-making styles in adolescents [79•]. Therefore, problematic gambling may be associated with non-rational and non-adaptive decision-making tendencies.

Together, these findings suggest that decision-making is an important consideration in GD. However, it is necessary not to operationalize risky decision-making patterns as a feature of GD solely, since it could represent an intermediate phenotype present across pathologies [59].

Decision-Making and PPU

A specific role of arousal on decision-making under risk and ambiguity has rarely been studied in PPU [82, 83]. Sexual arousal may influence motivational drives towards sexual gratification; thus, responses to sexual context cues, such as pornography or other sexual arousing stimuli, are important to consider in decision-making [84].

Experimental studies of sexual decision making have been conducted [85], including when inducing sexual arousal by presenting images with sexual content [86]. A modified version of the IGT included neutral and sexual pictures. When sexual images were associated with disadvantageous alternatives, decision-making performance was worse than when they were associated with advantageous alternatives, especially for individuals who were more sexually aroused. A preference in decision-making for images with sexual content may be associated with drives to receive and maintain gratification. Therefore, sexual stimuli may act as distractors, leading individuals, especially those who are more sexually aroused, to neglect feedback provided by the task during decision-making processes.

Sexual risk-taking when experiencing strong arousal may operate across genders. Sexual arousal may directly impact assessment of risky sexual situations and perceived advantages and disadvantages of chosen behaviors. Effects of “sexual myopia” may be similar to “alcohol myopia” and increase risk-taking [84]. In one study [87], when sexual arousal was heightened, the effects of alcohol on risk behavior (in this case, intentions to have unprotected sex) were stronger.

When comparing individuals with recreational/occasion use of pornography and those with PPU, differences in impulsive choice were observed [88]. These findings resonate with associations between impulsivity and severity of PPU described earlier [89]. Longitudinal studies suggest that individuals are immediately rewarded by use of pornography, which may predict steeper delayed discounting rate over time. Furthermore, effects of pornography use on decision-making may last longer than the duration of sexual arousal [17]. These findings resonate with those proposing long-term effects of pornography on reward system [90]. In addition, self-control training through non-use of pornography reduced delay discounting more so than other approaches, such as food abstinence [17].

In the case of problematic sexual behaviors, similarly to GD, it has been suggested that cognitive biases may contribute to decision-making in PPU, consistent with attentional impacts of erotic stimuli [91]. Individuals who reported greater cybersex-addiction symptomatology showed approach/avoidance biases to erotic stimuli [92]. A curvilinear relationship between PPU and approach-avoidance patterns was described [92]. Impaired cognitive control has also been observed when individuals with cybersex addiction are faced with multi-tasking including pornographic and neutral stimuli [93]. These findings were recently extended in male college students who used pornography; PPU was linked more to speed of approach than avoidance of erotic stimuli, with erotic stimuli being perceived as more positive and arousing [94•]. Similar findings have recently been reported in female college students [95]. In a separate study, being sexually aroused and the desire to masturbate reduced self-confidence about the ability to avoid pornographic stimuli even in individuals whose pornography use is once or less per week [96]. Some authors hypothesize that reward-related brain activations involved in PPU lead over time to a greater desire for increasingly new and extreme external sexual stimulation [97]. However, others propose that it could be seen as a precondition rather than a consequence of PPU [97]. Consequently, more research is needed to examine how decision-making relate to the onset or the maintenance of PPU.

Finally, when evaluating associations between sexual arousal and gambling in the general population, it has been observed that the incorporation of sexual stimuli reduced differences in arousal between gains and losses associated with gambling, when more arousal is usually observed towards losses. The presence of sexual stimuli could make losses associated with gambling be perceived as less salient [82].

Decision-Making and BED

Making advantageous decisions when eating and evaluating possible long-term consequences is important due to the increasing availability of palatable food and rates of obesity worldwide [98, 99]. Employing advantageous decision-making processes is especially important in the case of BED, particularly with respect to bingeing [98].

Individuals with BED often report feeling unable to control their food intake [26]. Individuals with BED may use more rigid decision-making strategies [16]. Specifically, people with BED may demonstrate enhanced switching between choices leading to impaired behavioral adaptation, reflecting a bias towards exploratory decisions in the context of dynamic environments [16]. Therefore, further investigation of decision-making in BED is important [16, 100].

Regarding decision-making under risk, individuals with BED who were overweight or obese made more risky decisions relative than those without BED who were overweight or obese as evidenced by performance on the game of dice task (GDT), which presents explicit probabilities and provides feedback to participants [98]. Individuals with BED also showed greater risk-seeking under monetary reward anticipation [101]. Thus, BED may involve impaired discrimination of reward values and tendencies to attribute more importance to subjective relative to objective probabilities (that is, when they perceive the likelihood of a probabilistic reward to be higher than the actual likelihood) [101, 102].

When evaluating decision-making under ambiguity with the IGT, patients with BED obtain lower scores, showing a greater tendency to make disadvantageous decisions, compared to individuals without BED, and difficulties in processing feedback received after making decisions [103, 104]. When studying individuals with obesity with and without BED, both show similar task performance [102]. In addition, BED severity correlates positively with the degree of impairment of decision-making processes [105].

With respect to delay discounting, individuals with BED versus those without tend to discount rewards more steeply [26, 106]. Furthermore, this tendency transcends domains, such as food, money, massages, or sedentary activity [107]. Higher levels of delay discounting have been observed in individuals with obesity, with and without BED. In the case of morbid obesity, higher delay discounting is observed if they also have BED, in comparison with individuals with non-BED obesity [102]. Therefore, an association between BED, severity of obesity, and impaired decision-making has been suggested [102]. Some authors have stressed that in the case of BED, the subjective perception of impulsivity and difficulties in controlling behavior (self-reported impulsivity) may be more relevant than conscious decision-making processes (impulsive task performance) [108]. Individuals’ preferences for short-term rewards, discounting possible long-term consequences, may explain the occurrence of binge-eating episodes, associated with a sense of loss of control, even when individuals start to experience negative consequences, such as weight gain or feelings of guilt [109].

Despite these findings, studies assessing BED and decision-making are relatively scarce and heterogeneous [109], so they should be interpreted with caution. In addition, findings of impaired decision-making processes may be less applicable to adolescent populations with BED, as a recent meta-analysis of EDs suggests [110, 111]. The possibility exists that decision-making processes remain relatively intact in early stages of BED [111], although this too warrants more examination. Over time and during development, individuals with BED may develop maladaptive patterns of decision-making in response to rewarding food cues [111].

Binge-eating behaviors may be driven by multiple neurocognitive alterations associated with decision-making and impulsivity and compulsivity, as well as other neurocognitive domains [26]. Some authors report, however, that in EDs, this impairment in decision-making processes may lessen when patients recover, with decision-making processes similar to non-affected individuals. Therefore, decision-making may be malleable and targeted in interventions for BED [112].

Limitations and Future Research

A current limitation in the field of neurocognition, and specifically in decision-making, is the existence of multiple tasks and models, which may hinder comparability of results across studies. More empirical studies are needed to understand the precise role for this neurocognitive domain in GD, PPU, and BED. Differences in conceptualizations of decision-making may also limit the assessment of this construct. The division between decisions under risk and ambiguity is not addressed in all the studies, and multiple neuropsychological instruments have been used to assess both processes, which may overlap to some extent. Moreover, the direct comparison between these three clinical entities is challenging since the literature is focused on different factors which may affect decision-making. Therefore, future studies should also address these conceptualization and assessment limitations. Finally, it should be noted that laboratory findings may not translate to real-world contexts, and these should be assessed.

Conclusions

Understanding decision-making has important implications for the assessment and treatment of individuals with GD, PPU, and BED. Similar alterations in decision-making under risk and ambiguity, as well as greater delay discounting, have been reported in GD, BED, and PPU. These findings support a transdiagnostic feature that may be amenable to interventions for the disorders. However, there are relevant gaps in the decision-making literature across these three clinical conditions, and a direct comparison of these groups on decision-making may benefit from directly assessing specific constructs in parallel across the conditions.