Gambling in primary care patients: why should we care and what can we do about it?

Disordered gambling has been associated with a host of serious health and social consequences. A limited body of research suggested that problem and pathologic gamblers experience higher rates of several types of physical symptoms, including fatigue, insomnia, minor respiratory ailments, intestinal distress, migraine headaches, high blood pressure and cardiovascular disease [1–3]. The social costs of problem and pathologic gambling include work and educational disruptions, criminal arrests, severe financial difficulties and familial problems (including domestic violence and child maltreatment) [4–6]. Problematic gambling is also associated with higher rates of comorbid psychiatric conditions, including concomitant substance use, eating disorders, depression and anxiety disorders [4,7–9]. Individuals with problematic gambling also have higher rates of attempted and completed suicides as compared with the general population [10,11].

In contrast to the harm associated with problem and pathologic gambling, recreational gambling has been argued to have significant societal benefits, including the creation of jobs and generation of revenues for state and local governments that are used for public benefit [12]. Some research has also suggested that nonpathologic recreational gambling may be related to improvements in social and health functioning, at least for a subset of the population [13].

The study by Morasco, vom Eigen and Petry featured in this issue (pages 94–100) provides further evidence that problem and pathologic gambling can have a significant negative impact on both the physical and mental domains of functioning and calls into question prior research suggesting that recreational gambling may confer health benefits. Morasco et al. extend the research on the relationship between gambling and health and mental health functioning in three important ways. First, the study uses a well-established measure of gambling pathology, the South Oaks Gambling Screen (SOGS), to evaluate rates of problem and pathologic gambling in a general adult primary care setting [14]. The SOGS is the most widely used measure of adult gambling behavior, rendering results easily comparable with prevalence rates in the general population outside primary care. Use of the SOGS also allows for classification of individuals as non-gamblers, recreational gamblers (Level 1;SOGS scores <3), problem gamblers (Level 2; SOGS scores 3–4) or probable pathologic gamblers (Level 3; SOGS scores≥5), enabling researchers to evaluate health outcomes across the full range of gambling involvement [15]. In addition, Morasco et al. use a well-validated measure of physical and mental health functioning, the Short-Form Health Survey Version 2, to describe health outcomes in their sample [16]. This is a significant improvement over prior research and allows Morasco et al. to more fully describe the range of consequences associated with different levels of gambling involvement. Finally, Morasco et al. assess a wide range of demographic characteristics and conduct the research in an urban primary care setting with an ethnically and socioeconomically diverse sample, enabling them to evaluate demography as a potential explanation for the widely disparate rates of gambling pathology reported in prior studies on primary care patients [17,18].

Morasco et al. found five times the rate of probable pathologic gambling and double the rate of problem gambling in their urban primary care sample in comparison with the general adult population. Together, more than 15% of patients in this study reported some level of disordered gambling. Rates were highest among males, ethnic minority patients and individuals receiving disability payments. Contrary to a previous finding of beneficial health effects for recreational gambling, Morasco et al. found an inverse relationship in which overall physical and mental health functioning decreased as gambling involvement increased [12]. Recreational gamblers did not fare better than non-gamblers on any dimension of functioning.

Although the study results are an important addition to the literature, there are limitations that must be considered and questions remaining. First, the contrasting findings between the work of Morasco et al. and that of Desai et al., demonstrating improved functioning among recreational gamblers, should be considered in the context of differing sampling procedures (voluntary vs. random digit dialing) and population characteristics (Desai et al. focused on older adults; Morasco et al., on primarily low-income participants) [12]. Further research is needed to evaluate the extent to which demography moderates the relationship between level of gambling involvement and negative (or positive) health and mental health outcomes. A second limitation of Morasco et al.’s study is that gambling behavior is treated as a homogenous phenomenon, in which, for example, purchasing lottery tickets is combined with casino gambling. In future research, we might be better served by examining different types of gambling behavior in relation to health and mental health outcomes rather than assuming a one-size-fits-all relationship. The potential for sampling bias is another possible limitation of Morasco et al.’s study as they did not assess characteristics of those who refused the survey, whom they estimate to be 20% of those approached. To the extent that sampling bias exists, prevalence rates of problem and pathologic gambling reported in this sample could be an overestimate or underestimate of the actual rates, although the relation between gambling and health outcomes would not be expected to be markedly affected by any such bias. Finally, questions remain regarding why and how gambling is related to poorer health and mental health functioning. It is possible that decreased health functioning may not be related to gambling per se but rather to the environment in which gambling occurs—for example, smoking, free or reduced-price alcohol and all-you-can-eat buffets at many casinos may contribute to negative health outcomes. Further research is needed to begin to answer these questions.

Despite limitations, the research reported by Morasco et al. documenting prevalence of problem and pathologic gambling and related health concerns among primary care patients suggests important issues that need to be addressed. First, although evidence suggests that problem and pathologic gambling are associated with negative psychological and physical health consequences, the magnitude of the problem in terms of health care use and associated costs has yet to be determined. Documentation of greater use and higher costs of health care among problem and pathologic gamblers would provide additional and stronger justification for aggressively pursuing treatment among these patients as part of regular primary care visits. In addition, although screening for other behavioral disorders, including substance use disorders, is becoming more common in primary care, screening for problem and pathologic gambling is considerably less common. The addition of a short screening instrument such as the two-item Lie–Bet Questionnaire on existing intake assessments seems feasible and could identify a relatively large number of patients at risk for a disorder that might not otherwise be uncovered [19]. Of course, assuming we can do a better job of identifying problem and pathologic gamblers begs the question of what we do with them once identified. More research is needed to determine whether and what kinds of services these patients might be willing to pursue in addressing their gambling problems. At the very least, these patients might be offered referral information from their primary care provider. Unfortunately, at present, few referral options are available for problem and pathologic gambling. Gamblers Anonymous (GA) meetings are available in many areas, but typically only at a fraction of the frequency of other 12-step groups such as Alcoholics Anonymous (i.e., biweekly or monthly vs. daily). Although the success rates of GA have not been well documented, the availability of GA, free services and lack of available alternatives may make GA a good referral in many locations. The development and evaluation of brief advice interventions, similar to those available for smoking and substance use disorders (e.g., 5 min), for problem and pathologic gambling, which could be delivered in the course of a primary care visit, would be invaluable. Longer, but still brief (e.g., 1 h), interventions have begun to show promise in the treatment of gambling disorders [20,21] and might also be made available but would require additional resources and planning. Finally, some evidence suggests that traditional cognitive–behavioral therapy for treating problem and pathologic gambling is also effective [22]. Nonetheless, further evaluation of gambling treatments is sorely needed, as is additional training in interventions for gambling disorders among medical and mental health practitioners. The findings reported by Morasco et al. should serve as an impetus to devise service delivery systems and training programs to help reduce the public health threat posed by disordered gambling.