Moving beyond a dichotomous classification for gambling disorders

Toce-Gerstein et al. (2003) provide empirical support for a continuum of disordered gambling. While this concept has been debated for some time, data have been lacking regarding behavioral patterns that constitute more and less severe forms of the disorder. Results suggested that chasing losses may be a subthreshold aspect of disordered gambling. This is the criterion endorsed most often, and it is the only one affirmed by the majority of people who otherwise reported no gambling-related problems. In early–mid problematic stages, gamblers cover up the extent of gambling, gamble to escape adverse moods and become preoccupied with gambling. In the stage classified by the Diagnostic and Statistical Manual (DSM) diagnosis, they experience tolerance, withdrawal and loss of control. In the most severe stages, gamblers commit illegal acts to support their gambling. Not only may severity of gambling problems differ among people, but also within the same individual over time. Observations of treatment seekers detail movement in and out of phases of more and less severe gambling (Hodgins et al. 2001). In community samples, comparisons of past-year to life-time rates suggest that 30–50% of pathological gamblers no longer meet diagnostic criteria (Hodgins et al. 1999). Recent data suggest that gambling is a highly variable phenomenon, with individuals experiencing and then not experiencing problems over time (Slutske et al. 2003). Data from the Toce-Gerstein et al. (2003) paper are derived from the National Opinion Research Center Diagnostic and Statistical Manual Screen for Gambling Problems (NODS), and they appear consistent with data derived from the South Oaks Gambling Screen (SOGS) (Lesieur & Blume 1987). In studies utilizing this latter instrument, the most frequently endorsed items are gambling more than intended and going back to win lost money (e.g. Petry & Mallya, in press). Experiencing guilt and arguments about gambling are endorsed to an intermediary degree, while passing bad checks is rarely endorsed. While the SOGS has endured criticism (Ladouceur et al. 2000; Stinchfield 2002), concordance across instruments suggests that problematic gambling indeed may exist along a continuum. The Toce-Gerstein et al. (2003) paper has a number of strengths. It utilizes a representative, community sample drawn from across the United States, and the instrument administered is based on DSM-IV criteria. The study provides statistically sophisticated support for symptom clusters in various stages of gambling problems. Future research should examine the reliability of these findings in other samples, including clinically confirmed pathological gamblers and frequent heavy gamblers. Studies may also examine concordance of symptoms when assessed by other instruments. Although gaining popularity, the NODS has yet to be subjected to extensive psychometric testing. As acknowledged by the authors, the NODS is based strictly on DSM criteria, yet DSM criteria for pathological gambling may not entirely or accurately characterize the disorder (e.g. Lesieur & Rosenthal 1991). DSM criteria may reflect only the most severe form, with other symptoms more prevalent in early stage disordered gamblers, such as borrowing money, having others criticize one's gambling and feeling regret over amounts wagered. Such issues are not assessed by the NODS. Indeed, the national study that used the NODS classified fewer individuals with pathological gambling (0.8% life-time; Gerstein et al. 1999) than other studies (Shaffer et al. 1999; Welte et al. 2001). These differences may be related to strict interpretation of DSM criteria in the NODS or characteristics of the sample to which it was administered (i.e. only those who acknowledged losing $100 or more in a year). Given that a hallmark feature of pathological gambling is hiding one's losses, more individuals may have admitted to some symptoms but did not affirm this screening item. Future research should also evaluate how quantity and frequency measures are related to the continuum of disordered gambling. The Alcohol Use Disorders Identification Test (AUDIT) (Saunders et al. 1993) finds that these variables are associated with alcohol problems, especially among at-risk drinkers. While they are likely to vary vastly across the continuum of gamblers and with respect to preferred gambling activity (Petry 2003), some aspects of frequency and quantity are probably related to development and severity of gambling problems. In sum, the Toce-Gerstein et al. (2003) paper provides a needed advance in the assessment of disordered gambling. To better understand patterns of gambling and clusters of symptoms, especially among less severe or at-risk gamblers, sensitive and specific instruments need to be developed, refined and administered to large samples of gamblers in the community. This paper is one important step in this direction.