Abstract
Gambling is a major contemporary public health issue with approximately 1.5% of the Australian population meeting criteria for pathological gambling problems, and a further 2.3% exhibiting significant gambling problems [1]. While there have been a number of attempts to explain the aetiology of pathological gambling in the context of a non-addiction framework, the addiction model of gambling remains the predominant paradigm [2–6].
In this context, high rates of comorbidity between pathological gambling and substance use have been consistently reported in community survey studies [7–9] and in samples of pathological gamblers attending for treatment at drug and alcohol facilities [10, 11]. Lifetime and current rates of 47% and 39%, respectively for substance use have been reported in samples of pathological gamblers attending veteran's administration or specialist gambling treatment programmes [12, 13] with rates of 8% to 20% noted among Gamblers Anonymous attendees [14–16].
Conversely, the rate of pathological gambling among populations of substance abusers has also been reported to be four to 10 times greater than that found in the general population [17, 18]. These rates have ranged from 6% to 33% subject to the population source and diagnostic criteria used [19–23]. More recently, Hall et al. [24] found rates of 8.0% for lifetime diagnosis and 3.8% for current diagnosis of pathological gambling in a sample of 313 cocaine-dependent outpatients (200 were also opiate-dependent) recruited for a treatment outcome study.
Although these data suggest that substance abusers may be at greater risk for developing gambling problems, and conversely, problem gamblers may be at greater risk for developing substance-abuse problems than the general population, a notable feature of the data is the large variability found in constructs that have been measured. Measured constructs have ranged from use, abuse, and misuse to dependence. Such differences in severity prevent direct comparisons across studies and limit generalization to other populations. With respect to gambling, measures have inconsistently assessed for the presence or absence of either lifetime and/or current diagnosis of ‘pathological’ or ‘problem’ gambling [25], using either standardized psychiatric interviews or SOGS [26]. There is even less consistency in the assessment of substance use and/or substance problems. A number of studies have used standardized interviews to establish psychiatric diagnoses of ‘dependence’ and ‘abuse’ [27]. Others have assessed broader constructs such as ‘problematic use’, or ‘hazardous use’ [7, 25]. In addition the types of substances used have been classified into broad categories such as ‘alcohol’ and ‘non-prescription illicit drugs’ without specifying the type[s] of drug used, or reported in such a way that it is not possible to separate the rates of alcohol from other drug use. This variability has the potential to lead to confusion and inflated comorbidity rates.
The purpose of this study was to investigate systematically rates of substance-use problems in a sample of diagnosed pathological gamblers seeking treatment at a specialist gambling outpatient clinic in Sydney, Australia. Structured standardized interviews were used to establish 12-month diagnoses of substance abuse and dependence. In addition, as there is a growing body of research indicating that non-problematic use of substances such as alcohol can encourage increased gambling [23, 28, 29], participants were also asked about their levels of general substance use.
Method
Participants
Seventy-five poker-machine treatment-seeking gamblers (48 males and 27 females) took part in this study. All met DSM-IV [30] criteria for pathological gambling, and the mean score on the SOGS [26] was M = 11.90 (SD = 2.51) indicating that a high level of gamblingrelated problems were experienced by participants. There was no difference between males and females on the SOGS. The mean age of participants was M = 37.69 (SD = 10.61), with males M = 35.90 (SD = 10.40), significantly younger than females M = 42.56 (SD = 10.61), f (1,74) = 10.30, p < 0.003).
Participants had recognized their gambling to be a problem for a mean of 3.70 years (SD = 3.90 years) prior to presentation for treatment. There was no difference between males and females on this variable.
Procedure
Consecutive poker-machine gamblers attending the impulse control research clinic, a specialist treatment service for problem gamblers, completed the SOGS and a semistructured clinical interview assessing gambling history and self-reported substance use. This interview contained probe questions for each of the DSM-IV diagnostic criteria for pathological gambling. Participants who met diagnostic criteria were invited to participate in a large-scale research project being conducted by the authors. As part of this study they attended a second assessment interview during which substance abuse and dependence were systematically assessed using the clinician-administered computerized composite international diagnostic interview (CIDI-auto v2.1, 12 month version) [31]. The CIDI-auto v2.1, 12 month version is a standardized interview that assesses for the presence of mental disorders during the previous 12 months and provides psychiatric diagnoses according to the definitions and criteria of DSM-IV [30].
The South-western Sydney Area Health Service Ethics Committee granted approval for the study.
Results
The results of this study are presented in two sections. The first section presents subjects' self-reported substance use. The second section presents data on rates of substance-use psychiatric diagnosis during the previous 12 months.
Self-reported substance use
Self-reported estimates for substance use during the previous 12 months were obtained from responses to items in the semistructured demographic interview. Data are presented in Table 1. As may be expected the legal substances, alcohol (73.3%) and nicotine (65.30%), were the most commonly used substances. However, males (52.08%) were more likely than females (19.23%) to combine alcohol use and gambling (χ2 = 7.56, p < 0.007). The majority of participants who smoked tobacco (85.71%) indicated that they smoked many more cigarettes during a gambling session. Self-reported usage rates for alcohol were similar to recent national estimates for at least semiregular use in the Australian population (81%), while rates of nicotine use were higher than national estimates (26%) [32]. Just over 13% of the sample reported marijuana use. There was no difference in the percentage of males and females reporting marijuana use. Reported usage of other illicit substances was minimal.
Self-report estimates for substance use and alcohol problems
Participants were asked to indicate whether they felt their use of substances was problematic. Twelve per cent of clients (12.50% of males and 11.11% of females) felt their current use of alcohol was problematic, and 8.0% (8.33% of males and 7.41% of females) reported an alcohol problem in the past. Only a small percentage of clients (5.33%) reported previous drug and alcohol treatment. All of these participants were male.
Substance-use disorders
Table 2 shows the percentages of male and female participants who met criteria for a substance use disorder on the CIDI-auto. As this table indicates, nicotine dependence was the most common substancerelated disorder. Thirty-seven per cent of participants (41.67% males and 29.63% females) met criteria for this disorder. Alcohol abuse was the next most common disorder, with 16.00% of the sample meeting criteria for this disorder during the 12 months prior to assessment. Males (22.92%) were significantly more likely to meet criteria for alcohol abuse than females (3.70%; Fisher's exact p < 0.03). Rates of alcohol dependence were somewhat lower (10.42% males, 3.70% females) and there was no sex differences on this variable.
Rate of DSM-IV substance use disorders
As Table 2 indicates, comparatively few subjects met diagnostic criteria for abuse and dependence disorders for substances other than alcohol and tobacco (1.3%–5.3%), and no subject met criteria for a 12-month opiate disorder. With the exception of opiate disorders, the rates of substance disorders were higher for all substances in this sample than in the general Australian population [33].
Discussion
This study investigated the rates of substance use and substance-use disorders in a sample of pathological gamblers seeking treatment at a specialist gambling treatment. Although direct comparison of comorbidity rates with previous literature is hampered by lack of consistency in measurement of substance-use problems, the rates of alcohol dependence in this sample appear somewhat lower than reported elsewhere in the literature [11]. The rate of alcohol abuse in this sample is comparable to the rate of lifetime ‘alcohol problems’ reported by Toneatto and Skinner [25]. However, in contrast to these authors, who found no differences between males and females in lifetime rates of alcohol problems for treatment seeking gamblers, the present findings indicated that recent or current alcohol-abuse problems are more prevalent among male treatment seeking pathological gamblers than females.
There is little data on rates of comorbidity between gambling and substances other than alcohol. It is therefore difficult to make comment on the rates of nonalcohol substance disorders observed in this study with rates found in other settings. Compared to available data it appears that rates of non-alcohol problems were also lower in the current sample than previous literature. For example, while high rates of comorbidity between gambling and opiate use have been reported elsewhere [24], no participant in this sample reported opiate use during the 12 months prior to assessment. Rates of substance use and substance-use disorders were also lower than reported by Toneatto and Skinner [25].
These differences between rates of comorbidity found in this sample and in previous research may reflect actual differences in comorbidity of substance use and gambling across the different settings. Pathological gamblers in treatment may not be representative of the total population pool of gamblers with problems with less than 10% of ‘cases’ identified in community-based samples reporting that they received treatment [1]. Pathological gamblers in treatment represent the more severe end of the spectrum of gambling and exhibit higher rates of psychopathology, notably antisocial personality disorder, mood and anxiety disorders [34, 35]. In the current sample, participants were attending a specialized gambling treatment clinic. Many of the previous studies have been undertaken in veteran administration drug and alcohol treatment centres. However, as there has been little consistency in the constructs that have been measured in the literature, higher rates found in previous studies may reflect less stringent assessment criteria or less ‘serious’ categories of substance use, for example, abuse versus dependence.
Although rates of substance-use problems were lower in this sample than found in previous studies, they were higher than found in the general Australian population [33]. This finding supports previous research suggesting that pathological gamblers may be at higher risk for developing substance-use problems than the general population. However, research into the relationship between gambling and substance misuse is in its infancy and the causal relationship between gambling and alcohol and other drug use remains at best a complex and little understood issue. The literature to date has focused mainly on identifying comorbidity rather than offering conceptual models explaining its aetiology.
Pathological gambling is conceptualized as an addictive disorder on the grounds that it exhibits many phenomenological similarities to substance-abuse disorders. Without entering into any discourse on the validity of the addiction model of pathological gambling, a number of explanations for the association between gambling and substance use can be offered without necessary reference to the presence of an inherent underlying addiction process. It could be argued on the basis of joint probabilities that poker-machine gambling and alcohol coexist in some individuals simply because both are offered in the same licensed venues. This increases the chance that gamblers will be exposed to opportunities for alcohol consumption, and alcoholics to gambling behaviour. Alternatively, alcohol may be used by the gambler as a way of coping with the guilt and depression following a big loss, or an alcohol user may use gambling as a means of trying to win money to buy alcohol [18]. Further research is needed to investigate these issues.
Identifying comorbidity is not only of conceptual and theoretical importance. Recognizing and assessing the association between gambling and substance use is also important from a clinical standpoint as empirical evidence is emerging to suggest that alcohol use and gambling is interrelated and that comorbidity places an individual at greater risk for relapse [18]. Within gambling sessions, alcohol is known to impair rational judgement and control and increase risk-taking among gamblers [23]. Baron and Dickerson [28] found that the ingestion of alcohol prior to gambling reduced resistance to begin and end a session of gambling, while Kyngdon and Dickerson [29] demonstrated that even in regular uses a small amount of alcohol during a session prolonged the duration and intensity of gambling. Alcohol use may thus precipitate a gambling lapse in a pathological gambler by impairing judgement, increasing risk-taking and selfconfidence in their ability to control their gambling. Alternatively persistence at gambling after cessation or controlled alcohol consumption may precipitate family, work, financial and/or emotional problems, that may in turn trigger a return to drinking, particularly where gambling is undertaken in a licensed venue.
Despite the potential impacts, current evidence indicates that many problem gamblers are not receiving assistance for concurrent substance problems [1] and many substance abusers are not being routinely assessed for the presence of a comorbid pathological gambling disorder [18]. Screening for gambling problems should be included in standard approaches in the assessment of substance users and appropriate interventions offered to identified cases of problem gambling.
Footnotes
Acknowledgements
Financial assistance for this project was provided by the NSW government from the Casino Community Benefit Fund.