Cognitive Processes Related to Problematic Pornography Use (PPU): A Systematic Review of Experimental Studies (2021)

2.1. Literature review and study selection

A systematic search was used to compile evidence regarding cognitive processes related to PPU. Studies were eligible if they (1) examined a cognitive process through an experimental task and (2) linked the results from this task with an aspect directly or indirectly related to PPU. We included studies establishing the following relationships between a particular cognitive process and PPU: (a) studies comparing certain cognitive processes in subjects with and without PPU; (b) studies comparing certain cognitive processes in subjects with and without SA/HD/CSBD (provided that the study specified PPU as the primary problematic sexual behavior of a large proportion of the sample and/or when certain aspects of pornography consumption –e. g., frequency of pornography use– let to distinguish between groups); (c) studies conducted in community samples correlating certain cognitive process with a direct indicator of PPU (e.g., scores in scales assessing PPU); (d) studies conducted in community samples correlating certain cognitive process with an indirect indicator of PPU (e.g., time online watching pornography, scores in scales assessing out-of-control sexual behaviors, etc.); and (e) studies conducted in clinical or community samples correlating certain cognitive processes with indicators of PPU after the exposure to pornography (e.g., arousability when exposed to pornography, craving after doing so, etc.).

We identified eligible studies by searching for published studies reported in English from 2000 to October 2020, using four academic search engines: PubMed, PsycINFO, Web of Science, and Google Scholar. To identify relevant articles, we used different combinations of the following search terms: “porn*” or “sexually explicit material” or “erotica” or “Internet sex*” AND “cognitive process*” or “executive functions” or “attention* bias*” or “working memory” or “inhibition” or “inhibitory control” or “decision-making”. An asterisk after the search term means that all the terms beginning with that root were included in the study search. To identify additional articles, we conducted a complementary search using keywords such as: “porn* addiction” or “problematic porn* use” or “sex* addiction” or “hypersexual disorder” or “compulsive sexual behavior disorder”. Studies retrieved through the last three terms (SA, HD, and CSBD) included clinical samples of patients reporting PPU as their primary sexual outlet, but also patients reporting other sexual problems (e.g., excessive use of Internet chats or sexual webcams, persistent and uncontrolled extra-marital affairs, habitual solicitation of commercial sex workers, etc.). Following the inclusion criteria, studies assessing clinical samples whose problems were not focused on PPU were excluded from this review.

A flowchart detailing the study selection process is shown in Fig. 1. In total, 7,675 studies were identified. After removal of duplicates, we obtained 3,755 records. Two of the review authors (JCC and VCC) examined abstracts and titles for relevant content. Only 23 of these studies were identified as potentially relevant. After a full-text review, we removed 12 of these articles (n = 11). To increase the number of studies, we searched the reference list of included articles for relevant literature, identifying 10 additional records that were finally included after a full-text review (n = 21).

1. Download : Download high-res image (365KB)
2. Download : Download full-size image

Fig. 1. Flowchart of study screening and selection process.

2.2. Data extraction

The following two recorded variables (i.e., the cognitive domain assessed in the study and the experimental tasks or paradigms employed in its assessment) constituted central aspects of this review. In order to categorize studies according to the cognitive domain, we followed the taxonomy proposed by Ioannidis et al., 2019, Brand et al., 2020. In particular, we distinguished between the following cognitive domains (and subprocesses): (a) attentional bias; (b) inhibitory control (pre-potent motor inhibitory control, motor inhibitory control, and attentional inhibitory control); (c) working memory; and (d) decision making (delay discounting, approach-avoidance tendencies, and decision making under ambiguity). Then, we described the experimental paradigm used to assess these cognitive domains (type of task, stimuli employed, instructions).

In order to provide a more nuanced overview of the reviewed studies, we also recorded the use of additional assessment measures (interviews, self-report scales, neurological or psychophysiological measures, etc.). The last variable coded in Table 1 comprised the main findings derived from each study. Data extraction and categorizing took place in the following ways. Initially, all results derived from each study were identified from the results and conclusions sections and tabulated in text format. Subsequently, an in-depth analysis was performed to identify findings relevant to the study aims. These findings were included in table 1, whereas information beyond the scope of this review was excluded.

3.1. Study characteristics

Table 1 summarizes studies included in the review. As for date of publication, more than half of reviewed studies (66.66%; n = 14) were published in the last five years. Studies were carried out in six countries and three continents: Europe (57.14%; n = 12), North America (23.80%; n = 5), and Asia (19.04%; n = 4).

In terms of sample size and representativeness, the studies included in this review assessed a total of 1,706 participants. Participant’s distribution for sex and age was far from being equivalent: only 26.20% of participants were females (n = 447), and 15 studies (71.42%) only assessed male participants. Most studies assessed participants below 30 years old (Mage = 25.15). In terms of sexual orientation, 12 studies (57.14%) only assessed heterosexual participants. As for the sample characteristics, 52.38% of the studies (n = 11) reported the assessment of clinical samples, including a total of 226 patients diagnosed with PPU.

For the cognitive domains that the studies focused on, 42.85% (n = 9) explored decision-making, 23.80% (n = 5) attentional bias, 19.04% (n = 4) inhibitory control, and 14.28% (n = 3) working memory. Regarding the use of complementary assessment measures, 76.19% of the studies (n = 16) administered self-report scales to screen for the presence of PPU or symptoms of SA, HD, or CSBD, 38.09% (n = 8) included measures of other sexual dispositions (e.g., sexual excitation/inhibition), 28.57% (n = 6) measured impulsivity, and 19.04% (n = 4) used self-reports to explore psychiatric symptoms.

3.2. Attentional bias

Attentional bias is defined as “the tendency for some stimuli to be preferentially processed, therefore capturing attention” (Kagerer et al., 2014). This preconscious process explains priority when processing competing stimuli: given that our attentional resources are limited, stimuli with a greater salience are preferentially processed. This is the case of stimuli that are relevant for the survival of species (e.g., stimuli indicating a potential threat). As proposed by evolutionary models of human attention (Yorzinski, Penkunas, Platt, & Coss, 2014), this attentional bias is biologically predisposed: thus, everybody shares this predisposition. However, individual differences in the salience of certain stimuli have also been observed, influencing the allocation of attention among competing stimuli. This is a phenomenon extensively studied in SUDs (Field, Marhe, & Franken, 2014). A tendency to preferentially process drug-related cues have been documented for multiple substances (Cox, Fadardi, & Pothos, 2006). These studies show that people with SUDs notice and attend to substance-related stimuli more readily than non-substance users, and that addiction-related cues prevails over other stimuli. More recently, attentional bias toward addiction-related stimuli have been shown in different BAs, such as gambling (Hønsi et al., 2013), gaming, or problematic social networks use (Wegmann & Brand, 2020). The incentive sensitization theory have been employed to explain underlying attentional bias toward addiction-related cues (Robinson & Berridge, 2001). According to this theory, classical conditioning processes explain that addiction-cues end up eliciting attentional biases: in particular, repeated pairings of certain addiction cues with the effects derived from the drug consumption lead to an increase in the salience of these stimuli, thus ‘grabbing’ attention and becoming especially attractive and ‘wanted’.

The most popular paradigm to assess these preconscious attentional biases is the dot-probe task (van Rooijen, Ploeger, & Kret, 2017). In this task, two stimuli (e.g., words, pictures, faces) are simultaneously presented for a brief period (typically, <500 ms) in different locations of a computer screen. One of these stimuli is emotionally neutral (e.g., kitchen items), whereas the other comprises the stimulus supposed to elicit the attentional bias (e.g., a wine bottle in an alcohol-related dot-probe task). Immediately after these stimuli disappear, a neutral object (a ‘dot’) is presented in the space previously occupied by one of these stimuli, and participants should press a response button as soon as they perceive this object. Attentional bias is measured through reaction times: participants are thought to respond quickly when the ‘dot’ appears next to the stimulus they were viewing (i.e. the stimuli grabbing attention at a preconscious level). In our review, four studies employed the dot-probe task to assess attentional bias in PPU. Two of these studies used a very similar experimental design (neutral vs. sexual stimuli and 500 ms of stimuli presentation) (Doornwaard et al., 2014, Kagerer et al., 2014), whereas the other two employed a more complex design (inclusion of three types of stimuli [explicit, erotic, and neutral] and 150 ms of stimuli presentation) (Banca et al., 2016, Mechelmans et al., 2014). One study assessed attentional bias trough a different experimental paradigm (i.e., visual probe task; Pekal, Laier, Snagowski, Stark, & Brand, 2018), and two studies included complementary tasks to assess other aspects of attentional bias: a word search task measuring selective attention (Doornwaard et al., 2014) and a line orientation task measuring stimuli categorization (Kagerer et al., 2014).

Findings derived from all the reviewed studies suggest that individuals with PPU, with a greater pornography consumption, or with traits related to PPU are more likely to present attentional bias toward sexual stimuli. In a sample of 46 men and 41 heterosexual women, Kagerer et al. (2014) found that sexual sensation seekers tended to answer faster to the dot-probe task when the dot appeared next to a sex picture, and to categorize faster the pictures depicting sex in the line-orientation task. Doornwaard et al. (2014) found that participants who consumed pornography on a more regular basis (moderate and high pornography users vs. low pornography users) were faster answering to the dot probe task, independently of whether the dot appeared next to a neutral or a sexual picture. In a study comparing 22 patients with CSBD (PPU as their primary sexual problem) and 44 healthy controls, the former displayed greater attentional bias to explicit sexual stimuli (Mechelmans et al., 2014). Notably, this attentional bias was observed only when participants were presented with sexually explicit stimuli; when presented with an erotic stimulus (i.e., lower level of explicitness) or a neutral stimulus, participants with CSBD and healthy volunteers responded similarly. Reanalyzing data from this study, Banca et al. (2016) found that subjects having a greater preference for conditioned sexual stimuli (mainly, those with CSBD and PPU) also showed enhanced attentional bias for sexual stimuli. In contrast, the preference for novel vs. familiar stimuli was not associated with attentional bias for sexual stimuli. Therefore, they concluded that attentional bias toward sexual stimuli was associated with a greater preference for cues conditioned to sexual images, but not with novelty preference. This conclusion resonates with the incentive sensitization theory (Robinson & Berridge, 2001), proposing that attentional bias toward drug stimuli are the result of classical conditioning processes; however, it goes against findings from the study by Kagerer et al. (2014), which found a relationship between attentional bias and sexual sensation seeking (aka novelty preference). Finally, Pekal et al. (2018) found that attentional bias toward sexual stimuli was correlated with severity of pornography addiction, craving (i.e., desire to masturbate when presented with pornography), and subjective sexual arousal. These findings were consistent in both males and females, and partially mediated by craving and subjective sexual arousal (i.e., the effect of attentional bias upon pornography addiction was boosted by cue-reactivity and craving).

3.3. Inhibitory control

Inhibitory control plays a central role when it comes to regulating human behavior as it is considered responsible for suppressing thoughts, actions, and emotions in response to environmental demands: when a certain behavior is no longer relevant or is harmful (especially in the latter case), inhibitory control allows to stop and replace it with an alternative –more adapted– behavior (Verbruggen & Logan, 2008). Deficient inhibitory control is often found in multiple psychiatric conditions, including SUDs (Bechara, 2005) and BAs (Brand et al., 2016, 2019). Experimental studies have identified three levels of inhibitory control (Chamberlain and Sahakian, 2007, Howard et al., 2014): (a) motor inhibitory control (i.e., the ability to withhold not-already-triggered responses); (b) pre-potent motor inhibitory control (i.e., the ability to suppress already-triggered responses); and (c) attentional inhibitory control (i.e., the ability to suppress irrelevant cognitive processing and shift attention away from salient yet irrelevant features of the situation).

Motor inhibitory control is typically measured through the go/no-go paradigm. In this task, subjects are presented with a series of stimuli and instructed to respond as quickly as possible when a ‘go stimulus’ is presented, and to withhold their response when a ‘no-go stimulus’ is presented (e.g., “press the response button when a horizontal line appears on the screen” and “do not press the response button when a vertical line appears on the screen”). In this task, impaired response inhibition is measured through the number of omissions (participants fail to respond in a ‘go trial’) and commissions (participants fail to inhibit response in a ‘no-go trial’). In our review, only one study employed this task to explore the relationship between PPU and motor inhibitory control (Seok & Sohn, 2020). In this study, participants (30 men meeting criteria for the diagnosis of HD and a notable weekly pornography use vs. 30 healthy men reporting a moderate pornography use) completed an adapted version of this task in which neutral stimuli (letters) were presented in a neutral or sexual background. Authors found that patients with HD and an increased weekly pornography consumption performed worse in the go/no-go task than healthy controls, especially in ‘no-go trials’ (those requiring inhibition) and when the task was presented together with sexual images in the background. Therefore, they concluded that patients with HD seem to be more prone to experience problems with motor response inhibition, especially when inhibition should occur during the exposition to sexual cues.

The most popular paradigm to measure pre-potent motor inhibitory control is the stop-signal task. In a stop-signal task, subjects typically perform a choice reaction task (e.g., “press ‘R’ after the presentation of a red circle and ‘B’ after the presentation of a blue circle”). During certain trials (i.e., ‘stop signal trials’), subjects are presented with a stop signal after the presentation of the stimuli (e.g., an auditory signal) indicating that they should inhibit the already initiated response to the stimuli. In this task, pre-potent motor response inhibition is measured through the number of commission errors and the stop-signal reaction time (i.e., an estimate of the time taken to suppress a response that would normally be made) (Verbruggen & Logan, 2008). In our review, only one study assessed pre-potent motor inhibitory control in PPU (Antons & Brand, 2020). This research found that severity of Internet pornography use (measured through the S-IATporn –a scale assessing addiction symptoms–) and craving (i.e., strong desire to use pornography) correlated with reaction times during ‘stop-signal trials’ in both the neutral and pornographic conditions. Surprisingly, increased severity of Internet pornography use and craving was associated with faster reaction times (i.e., better pre-potent motor inhibitory control). Authors explained these contradictory findings by suggesting that subjects with a higher severity of Internet pornography use and craving may have developed certain tolerance towards pornography, meaning that the exposition to these contents was less interfering.

Attention inhibitory control is typically measured through the classical Stroop paradigm. In this task, participants are instructed to name the font color of different colored words. Participants are encouraged to respond as quickly as possible, while response time and errors are measured as outcome measures. The font color of the colored word may be congruent (e.g., the word ‘BLUE’ in blue font) or incongruent (i.e., the word ‘BLUE’ in red font), and subjects typically present delayed reaction times and increased errors in the latter condition. Attention inhibitory control is computed as the difference between subjects’ performance in congruent and incongruent conditions. In this review, only one study employed this paradigm to assess attention inhibitory control in a sample of patients with PPU meeting criteria for the diagnosis of HD (Seok & Sohn, 2018). This study found that individuals with HD and healthy controls showed similar reaction times when answering to a stroop task, but the former were less accurate when answering to incongruent stroop trials. These findings should be considered as preliminary, but they point out that patients with HD may experience certain problems to shift attention away from irrelevant stimuli. Future studies should address whether these problems are increased when using sexual stimuli as distractors.

3.4. Working memory

Working memory is necessary to keep things in mind while performing complex tasks, such as reasoning, comprehension, or learning (Baddeley, 2010). It is defined as “a system for temporary storage and a mechanism for the ‘on-line’ manipulation of stored information that occurs during a wide variety of cognitive activities” (Owen et al., 1998, p. 567) and involves two central components: a memory component (limited to events occurring in a short period of time –and sometimes equated to the concept of ‘short-term memory store’–) and a working component (necessary for understanding, problem solving, and decision making) (Cowan, 2014). At a practical level, individuals with better working memory are more efficient when it comes to integrating the analysis of current environmental information/demands with past experiences; on the contrary, individuals with working memory deficits often neglect past experiences when making present decisions, giving in to the urge of engaging in appetitive behaviors without considering potential negative consequences. As a result, working memory impairments increase the risk to engage in multiple problematic behaviors, including SUDs (Khurana, Romer, Betancourt, & Hurt, 2017) and BAs (Ioannidis et al., 2019).

The n-back task is one of the most popular paradigms to assess working memory (Owen, McMillan, Laird, & Bullmore, 2005). In this task, participants are instructed to monitor a series of stimuli (e.g., words or pictures) and to respond whenever a new stimulus is presented that is the same as the one presented n trials before. Cognitive demand required to conduct this task increases as a function of the n trials required to be remembered: tasks in which participants are required to respond to stimuli presented two (2-back) or three trials earlier (3-back) are considered complex. Subjects should indicate whether each stimulus was previously presented or not, and working memory is assessed by reaction times and response accuracy (Meule, 2017). In this review, we found three studies using a n-back task to measure working memory in PPU. Experimental tasks used to assess this cognitive domain greatly varied between studies: Sinke, Engel, Veit, Hartmann, Hillemacher, Kneer, and Kruger (2020) compared performance on a 1-back and a 2-back task while participants were presented with a neutral or a pornographic background; Au and Tang (2019) used a 3-back task after the induction of positive, negative, sexual, or neutral emotional states; and Laier, Schulte, and Brand (2013) conducted a 4-back task including pornographic pictures as stimuli. Despite these notable differences, results were highly consistent: participants with a greater pornography use and/or patients with PPU (two independent but related categories) tend to perform worse in tasks assessing working memory, especially when this cognitive domain is assessed during the presentation of concurrent sexual stimuli. Laier et al. (2013) found that subjective sexual arousal after seeing pornography and craving for porn (two basic features of PPU) correlated with different indicators of poor working memory performance. Furthermore, the interaction between these two variables predicted 27% of the variance in the performance of the 4-back task. Au and Tang (2019) confirmed that pornography users with greater problems of sexual compulsivity performed worse in working memory (less precision and increased time to answer), independently of the emotional context and the type of stimuli employed in the n-back test. Finally, Sinke et al. (2020) found that patients with CSBD performed worse than healthy controls when the n-back test was conducted with a sexual picture in the background, but not when the task was conducted with a neutral picture in the background. Notably, this study found that sexually compulsive patients performed better than healthy controls in a task measuring long-term recognition of sexual stimuli, suggesting that patients with PPU may have a better memorization/recall of sexual cues despite short-term problems with working memory.

3.5. Decision making

Decision making constitutes one of the most central cognitive processes as it influences on multiple aspects of goal-oriented behavior. In brief, decision making is defined as the ability to select optimal choices considering all available information (Ioannidis et al., 2019). Individuals with decision making impairments tend to show preference for short-term small gains rather than long-term large gains, experience approach tendencies toward appetitive stimuli (e.g., drugs) despite medium or long-term negative consequences, are more likely to select risky options, tend to be inaccurate when judging the probability and magnitude of potential outcomes, and tend to perseverate in their responses despite the negative results. Multiple studies demonstrate that these features are typical of individuals with SUDs (Bechara, 2005, Ernst and Paulus, 2005) and BAs (e.g., Internet gaming disorder; Schiebener & Brand, 2017), constituting the ‘core’ cognitive underpinnings of some of their self-regulation problems.

As delineated by recent theoretical models, decision making occurs on different steps comprising functionally distinct cognitive subprocesses (Ernst & Paulus, 2005). The first step of decision making (i.e., assessment and formation of preferences among the possible options) is influenced by the preference for small immediate rewards rather than large delayed rewards (i.e., discounting). Discounting is assessed by discounting tasks. These tasks measure “the extent to which an individual devalues a reinforcer as a function of the delay to or probability of receiving it” (Lawyer, 2008, p. 36). In a classical ‘delay discounting task’, participants are presented with a situation in which they must make a choice (e.g., “do you want 1€ now or 10€ tomorrow?”). In the first trials, participants typically choose delayed larger gains. Over the course of the experiment, the smaller immediate amount increases systematically (1€, 2€, 3€…) and, at some point (e.g., 8€ now or 10€ tomorrow), individuals tend to switch to the immediate outcome over the delayed outcome. In a ‘probability discounting task’, the likelihood of receiving certain outcomes changes over the course of the experiment (e.g., “do you prefer 1€ for sure or 10€ with a 25% chance?”). In this review, two studies used these tasks to assess discounting in PPU. One study measured delay and probability discounting for both money and erotica (Lawyer, 2008), whereas the other only measured delay discounting for money (Negash, Van, Sheppard, Lambert, & Fincham, 2016). Lawyer (2008) found that in both the monetary and erotica delay discounting tasks, erotica users preferred smaller reinforcers available immediately than larger reinforcers provided after some delay. Similarly, erotica users preferred small but certain outcomes rather than larger but uncertain outcomes. Further, the degree in which sexual behavior was problematic correlated with discounting. All in all, erotica users (especially, those showing more symptoms of PPU) tended to show more impulsive choice patterns than non-erotica users. Similarly, Negash et al. (2016) found that frequency of pornography consumption measured in time 1 predicted delay discounting four weeks later: again, participants reporting viewing more pornography demonstrated higher discounting of future rewards. Furthermore, they found that after refraining from pornography consumption for 21 days, participants reported reduced levels of delay discounting (i.e., showed an increase in their preferences for delayed longer gains). This suggests that decision making impairments related to PPU may constitute temporal deficits derived from persistent pornography use, and exercising self-control over pornography use may have a medium-term positive effect on this cognitive ability.

The first step of decision making is also influenced by another cognitive process: approach bias toward appetitive stimuli. Approach bias is defined as “an automatically activated action tendency to approach reward-related cues” (Kahveci, van Bocstaele, Blechert, & Wiers, 2020, p. 2). The most popular paradigm to assess this aspect is the approach-avoidance task (AAT). In the AAT, participants use a joystick to pull certain stimuli presented on a computer screen toward themselves (approach bias) or to push away (avoidance bias). The use of a joystick (i.e., physical movement) and the inclusion of a zooming feature (i.e., visual movement) enhance the effect of approaching/avoiding the stimuli. In the case of PPU, studies have focused on approach bias toward sexual stimuli: in particular, four studies used an AAT to explore the link between approach bias toward sexual stimuli and PPU. Studies varied in terms of the stimuli employed and the type of instructions provided to participants. As for the stimuli, three studies included both neutral and sexual stimuli (in particular, pictures), whereas the fourth study only included sexual stimuli. As for the task instructions, two studies used ‘task-irrelevant instructions’ (pull or push the stimuli according to image orientation –horizontal vs. vertical–) (Sklenarik et al., 2019, 2020) and two used ‘task-relevant instructions’ (pull or push the stimuli according to their content –sexual vs. neutral or clothed vs. nude–) (Kahveci et al., 2020, Snagowski and Brand, 2015). These differences may explain some of the inconsistent results found in these studies. In a study including 123 male pornography users, Snagowski and Brand (2015) found a curvilinear relationship between approach-avoidance tendencies and severity of pornography consumption: in particular, individuals with PPU showed either extreme approach or extreme avoidance tendencies toward pornographic stimuli. On the contrary, the series of studies conducted by Sklenarik et al. suggested that, both in males (2019) and females (2020), the severity of pornography consumption showed a linear (not a curvilinear) relationship with approach bias toward sexual stimuli. Furthermore, in males but not in females, individuals with PPU demonstrated stronger approach bias toward sexual stimuli than non-problematic pornography users: in particular, problematic pornography users showed more than a 200% stronger approach bias than individuals without PPU. Finally, Kahveci et al. (2020) found that individuals reporting using pornography on a more regular basis showed stronger approach biases towards sexual stimuli; however, severity of pornography consumption (measured through the Problematic Pornography Use Scale –PPUS–) did not significantly correlate with approach bias, and problematic and non-problematic pornography users did not differ in terms of approach biases toward sexual stimuli. These findings suggest that frequency –but not severity– of pornography consumption may constitute the key factor when predicting approach biases toward sexual stimuli.

The second step of decision making refers to the selection and execution of an action (Ernst & Paulus, 2005). In this step, the appraisal of risk, reward magnitudes, and the probability of different outcomes constitutes a central feature of decision making. These aspects may be assessed under two conditions: objective risk and ambiguous risk (Schiebener & Brand, 2017). Given that no studies have assessed decision making ‘under objective risk’ in PPU, we will focus on decision making ‘under ambiguous risk’. In these tasks, individuals are not provided with explicit information about the probabilities for positive/negative consequences derived from their choices before starting the task; thus, they should base their first decisions on ‘feelings’ and, over the course of the task, they can learn the implicit rules behind each decision through periodical feedback (i.e., contingency-reversal learning) (Bechara, Damasio, Tranel, & Damasio, 2005). The most popular task to assess this aspect is the Iowa Gambling Test (IGT). In the IGT, participants are given 2000€ with the indication that they should maximize their benefits over the course of the task. Participants choose cards from four decks lying face down: decks A and B are disadvantageous (high gains but even larger losses), whereas decks C and D are advantageous (moderate gains and small losses) (Buelow & Suhr, 2009). Choosing cards from decks A/B leads to overall losses, whereas cards from decks C/D leads to overall gains. Therefore, people with appropriate decision-making abilities tend to preferentially select cards from decks C/D (Steingroever, Wetzels, Horstmann, Neumann, & Wagenmakers, 2013). In this review, we found two studies measuring decision making under ambiguity through the IGT. Mulhauser et al. (2014) used a classical version of the IGT to compare decision making in a sample of 18 patients with HD (PPU as the primary sexual problem) and 44 healthy controls. These researchers found that hypersexual patients were more likely to select decks with frequent loss penalties, a pattern of response that leads to a poor performance on the IGT. Laier, Pawlikowski, and Brand (2014) employed a modified version of the IGT in which two types of stimuli (neutral vs. pornographic pictures) were alternatively assigned to the advantageous or disadvantageous desk. They assessed a sample of non-problematic pornography users, finding that performance on the IGT was better when sexual stimuli were associated with advantageous decisions and worse when associated with disadvantageous decisions (i.e., sexual cues conditioned decision making). This effect was moderated by individuals’ reactivity to pornographic contents: in individuals reporting high sexual arousal after sexual picture presentation, the influence of sexual stimuli on decision making was greater. In summary, these two studies suggest that individuals exhibiting higher reactivity in front of sexual stimuli or with PPU exhibit poor decision-making, especially when this process is guided by sexual cues. This may explain why these individuals experience problems to control their sexual behavior despite the wide range of negative consequences related to their pornography consumption.