Redish et al. put forward a framework of addiction that unites competing theories by focusing on key vulnerabilities in the development and maintenance of addictive behaviors and emphasizing their links to the habit and planning systems underlying decision-making. Consideration of pathological gambling within this framework seems particularly valuable as historically this disorder has been conceptualized and categorized separately from substance use disorders (American Psychiatric Association 2000; Hollander & Wong Reference Hollander and Wong1995). Based largely on data describing irrational cognitions in pathological gamblers and therapies based on altering these cognitions, Redish et al. argue that vulnerability to pathological gambling falls largely within the domain of misclassifying situations (Vulnerability 6). However, many recreational gamblers experience irrational cognitions when gambling, which raises questions regarding the centrality of this feature to pathological gambling (Sharpe Reference Sharpe2002). Furthermore, the most thoroughly tested cognitive behavioral therapy for pathological gambling targets more than irrational cognitions, addressing coping with cravings and managing finances (Petry Reference Petry2005; Petry et al. Reference Petry, Alessi, Carroll, Hanson, MacKinnon, Rounsaville and Sierra2006). Clinical experience suggests that changing erroneous perceptions is not always sufficient in ceasing pathological gambling, as some patients report knowing that they will lose but continue gambling problematically nonetheless.

Consistent with the integrative goals of the target article, we consider Vulnerability 9 (over-fast discounting) as a failure point with relevance to pathological gambling. While other vulnerabilities (e.g., those related to craving, obsessions, or withdrawal) could be considered, over-fast discounting applies to gambling in important ways and thus was selected for elaboration.

Redish et al. describe the relevance of over-fast discounting of rewards, particularly temporally (i.e., delay discounting; Ainslie Reference Ainslie1975; Rachlin & Green Reference Rachlin and Green1972), to substance use disorders, citing pre-clinical and clinical data (Kirby et al. Reference Kirby, Petry and Bickel1999; Richards et al. Reference Richards, Sabol and de Wit1999). Delay discounting reflects one aspect of impulsivity (Bornovalova et al. Reference Bornovalova, Lejuez, Daughters, Rosenthal and Lynch2005; Moeller et al. Reference Moeller, Barratt, Dougherty, Schmitz and Swann2001). A principal component analysis of self-reported and behavioral measures of impulsivity identified two components, termed “impulsive disinhibition” and “impulsive decision-making,” with delay discounting contained in the latter category (Reynolds et al. Reference Reynolds, Ortengren, Richards and de Wit2006). Self-reported and behavioral measures of impulsivity, including delay discounting, have not correlated strongly (Krishnan-Sarin et al. Reference Krishnan-Sarin, Reynolds, Duhig, Smith, Liss, McFertridge, Cavallo, Carroll and Potenza2007). This phenomenon might reflect hypothetical versus real-life differences in risk/reward decision-making, as are often observed, for example, in trying to maintain New Year's dieting resolutions in the setting of chocolate cake being served. Unsurprisingly, preliminary data associate behavioral measures of impulsivity and substance abuse treatment outcome, whereas outcomes were not associated with self-reported measures in the same study (Krishnan-Sarin et al. Reference Krishnan-Sarin, Reynolds, Duhig, Smith, Liss, McFertridge, Cavallo, Carroll and Potenza2007).

Although several studies of drug abusers have utilized hypothetical drug reinforcers (a small amount of heroin immediately versus a larger amount tomorrow), most work has considered financial choices ($10 immediately versus $12 tomorrow) with the assumption that the drug and money are functionally related. As such, one might hypothesize that the typical delay discounting procedure utilizing financial rewards might be even more relevant to gambling than to substance use behaviors. Relatively few studies have directly examined problem or pathological gambling and delay discounting, and existing studies have generated inconsistencies (Reynolds Reference Reynolds2006). One study indicated a link between problem gambling and increased discounting of delayed monetary rewards, even when controlling for self-reported impulsivity (Alessi & Petry Reference Alessi and Petry2003). A separate study found that pathological gamblers discounted delayed rewards more steeply than non-gamblers (Dixon et al. Reference Dixon, Marley and Jacobs2003). However, a third study of young adults found the converse (Holt et al. Reference Holt, Green and Myerson2003).

A role for co-occurring substance use disorders is also indicated. One study found that among a sample of substance abusers, probable pathological gamblers discounted rewards more rapidly than did those without gambling problems (Petry & Casarella Reference Petry and Casarella1999), although the difference was limited to hypothetical rewards of larger magnitudes. A related study similarly indicated that pathological gamblers discounted delayed rewards at higher rates than did control participants, and gamblers with substance use disorders discounted delayed rewards at higher rates than did non-substance-abusing gamblers (Petry Reference Petry2001). The only experimental study found that, among pathological gamblers, a classic discounting profile was evident only when conducted in a real-life gambling context (Dixon et al. Reference Dixon, Jacobs and Sanders2006). It has also been suggested that individuals who engage in different forms of gambling (e.g., slot machine vs. sports) might be discounting rewards differently (Cooper Reference Cooper2007). Thus, although multiple studies indicate that problem and pathological gamblers, like substance abusers, discount rewards more rapidly than do control subjects, the results are not entirely consistent and suggest that specific environmental, developmental, or individual factors influence these processes. This interpretation fits well with the assertion of Redish et al. for addictions in general, that specific vulnerability factors (including over-fast discounting) may be more salient for specific sub-groups, even within diagnostic categories.

The scope of Vulnerability 9 also warrants further consideration. Both positive and negative reinforcement processes have been implicated in addiction (Koob & Le Moal Reference Koob and Le Moal2001), and considering both with respect to rapid discounting seems important. Although delay discounting has typically been applied to positive reinforcers in comparing smaller immediate and larger delayed rewards, it also may be considered in relation to aversive stimuli in the context of negative reinforcement. In this case, impulsivity involves the selection of a larger, delayed aversive stimulus over a smaller, yet immediate aversive stimulus. Said differently, this describes the tendency to delay experiencing a mildly aversive event, even though this delay likely will result in a more aversive event in the future. Currently, few conceptualizations of delay discounting consider this reciprocal focus on negative reinforcement across substance disorders or pathological gambling. This conceptualization appears particularly relevant to important aspects of pathological gambling such as “chasing” losses, wherein one continues to gamble further, typically leading to greater future gambling losses, instead of accepting the immediate consequences associated with a recent gambling loss.

Redish et al. have advanced the field with this ambitious integrative effort. Further developing this model for pathological gambling as it relates to other vulnerabilities could strengthen the impact of the model and its utility in advancing prevention and treatment strategies for pathological gambling.