Pathological gambling and the DSM-V

The American Psychiatric Association (APA) is in the process of updating the fifth of edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM). The DSM is the primary classification system for diagnosing mental disorders in the United States, and it is widely used in other countries as well.

Currently, pathological gambling is classified in the ‘Impulse-Control Disorders Not Elsewhere Classified’ section of the DSM-IV (American Psychiatric Association, 1994). The Substance Use Disorders Workgroup of the APA's DSM committee has made a recommendation to move pathological gambling to its chapter. This suggestion was based on a vast body of literature demonstrating high rates of comorbidity between substance use disorders and pathological gambling (Petry, Stinson & Grant, 2005; Kessler et al., 2008), similarities in some symptom presentations (Petry, 2006), some potential parallel biological dysfunction (Potenza et al., 2003), evidence of shared genetic liability (Slutske et al., 2000), and overlap in treatment approaches (Hodgins, Curry & el-Guebaly, 2001; Petry et al., 2006; Petry, Weinstock, Ledgerwood & Morasco, 2008).

In addition to recommending a change in location in DSM-V, the Workgroup has suggested that the name of the disorder be altered. The term ‘pathological’ conjures pejorative thoughts among providers, patients and the public. A working suggestion is ‘disordered gambling’. Terms such as ‘problem gambling’ and ‘compulsive gambling’ were considered but decided against. The former has often been used to refer to a sub-diagnostic threshold condition and hence could lead to future confusion related to severity, and the latter shares overlap with disorders in the anxiety disorders section of the DSM. The final name for pathological gambling will be determined when the overall chapter name is decided.

The most recent version of the DSM (revision IV) lists 10 criteria for pathological gambling (American Psychiatric Association, 1994). Statistical analyses generally find these 10 criteria tap a single uniform underlying construct (Strong & Kahler, 2007). However, one of the ten criteria is rarely endorsed and appears to add little to diagnostic classification accuracy. The criterion related to committing illegal acts (e.g. forgery, fraud, theft or embezzlement) to finance gambling is the least often endorsed criterion in a number of population surveys (Blanco, Hasin, Petry, Stinson & Grant, 2006; Gerstein et al., 1999; Strong & Kahler, 2007), and item response theory reveals it adds little to classification accuracy (Grant, 2010). Hence, the Workgroup has suggested that this criterion be dropped in the next version of the DSM.

The other recommendation of the Workgroup is to reduce the number of criteria necessary for a diagnosis of pathological gambling. Several population based studies reveal that meeting four criteria, rather than the five currently necessary for a diagnosis, improves classification accuracy (Jiménez-Murcia et al., 2009; Stinchfield, 2003; Stinchfield, Govoni & Frisch, 2005). Additional analyses were undertaken on the NESARC data when the legal criterion was eliminated, and similar results were found as noted in the published studies with the threshold of four criteria (Grant, unpublished data, 2010). Hence, the Workgroup recommends that ‘pathological gambling’ be diagnosed when one meets four of nine criteria in the next version of the DSM.

The Substance Use Disorders Workgroup received over 400 comments related to its proposed changes during a period of open public commentary. Although not all responses were favorable about all aspects of proposed changes to substance use disorders, the overwhelming vast majority of comments related to pathological gambling were in support of moving gambling to an ‘Addiction and Related Disorders’ section. Some argued to reduce the number of criteria necessary for a diagnosis even lower than four (e.g. to two or three criteria), but analyses from large population databases reveal that such changes would substantially impact the prevalence rate of the disorder, by more than doubling the rate of this disorder (Blanco et al., 2006; Welte, Barnes, Wieczorek, Tidwell & Parker, 2001). Moreover, lowering the threshold for the first non-substance related ‘addictive’ behaviour may set a precedence for setting the bar too low for future putative ‘behavioural addictions’ that may be introduced in subsequent versions of the DSM. Ultimately, low thresholds for diagnoses could undermine the system of psychiatric classification more broadly and prove a disservice to individuals with substantial psychiatric distress (Martin & Petry, 2005).

Moving pathological gambling to the substance use disorders section and reducing the criteria and diagnosis threshold are likely to improve screening and treatment efforts in high-risk populations. A new name for the disorder is also likely to be a welcome change. It is hoped that these suggested alterations will improve diagnosis, treatment and research efforts to better understand and ultimately prevent this disorder.