Writeup of gambling seminar

From Dr Andrew Byrne:

Concord Seminar summary on gambling issues. Tuesday 3rd August 2010.

The 4th Concord Seminar of 2010, “Is pathological gambling an addiction? You bet it may or may not be!” was presented by Alex Blaszczynski, Professor of Clinical Psychology and Director of the Gambling Treatment Clinic at the School of Psychology, University of Sydney. He has written a self-help manual, “Overcoming Compulsive Gambling”. He is editor of International Gambling Studies and Assistant Regional Editor for the journal ‘Addiction’.

The presentation covered the definitions and epidemiology of problem and pathological gambling, their impacts on self and others; the multifactorial etiology of gambling problems; cognitive distortions and implications for treatment; and the “pathways” model for understanding etiology and matching treatment interventions.

While pathological gambling (PG) (recurrent gambling despite severe negative consequences and/or repeated unsuccessful attempts to cease) remains classified in DSM -IV-TR (A.P.A., 2000) among the impulse control disorders (along with kleptomania and pyromania), its diagnostic criteria since DSM III have come more closely to resemble those substance use disorders. In DS-V, it is proposed to reclassify the condition as a non-substance behavioural addiction. By contrast, ‘problem’ gambling is defined by harms to the individual player, their family and/or the wider community. This resembles the definition of harmful substance use in ICD-10.

There has been a worldwide increase of legalized forms of gambling, starting in the USA in 1968 with the New Hampshire lotteries; in Australia with the 1973 West Point Casino; in Britain with the1978 Royal Commission into Gambling and large increases in Europe in the 1990s and in Asia in 2000s. Electronic gaming machines have become increasingly common while the current spectacular growth area is in online gambling.

The prevalence of moderate to severe problem gambling is similar in Australia (2.7%) and in the USA (3.0%). Estimates vary widely for prevalence of pathological gambling (Australia 0.6-1.2%; USA 0.1-1.9%), reflecting the assessment tools used (e.g., South Oaks Gambling Screen versus DSM criteria).

The DSM diagnosis of pathological gambling (PG) requires five or more of the following:
1. Preoccupation (psychological dependence & salience)
2. Increased amount gambled (tolerance)
3. Irritability/restlessness on cessation (withdrawal)
4. Escape from stress (negative reinforcement & motivation)
5. Chasing losses (erroneous & distorted cognitions)
6. Lying
7. Repeated failure to cease (impaired control)
8. Illegal acts
9. Risked significant relationships
10. ‘Bailout’ (relatives or friends paying gambling debts)

The criteria of salience, tolerance, withdrawal, impaired control and continuing despite knowledge of harm have obvious parallels with substance dependence, and suggest the likely involvement of meso-limbic/orbito-frontal reward systems in positive and negative reinforcement, underpinning classical and operant conditioning in the development of craving, and impulsive decision-making in pathological gambling.

However, certain other features of pathological gambling bear less close comparison with substance dependence. One example is the mediating function of erroneous and distorted cognitions such as the “gambler’s fallacy”, the mistaken belief that the chances of winning over time increase (in fact the chances of winning remain the same at each point in time, and losses are cumulative over time). A recent published paper (Slutske et al 2010), reported that recovery from PG is commonly achieved in the absence of abstinence, ie with a return to “controlled gambling”, a further difference from most instances of substance dependence, where a return to controlled use is exceptional (see Stanton Peele for the contrary view for alcohol and drug use).

Indeed the significance of tolerance or withdrawal, two defining elements of “gambling as an addiction”, remains unclear. A recent study (Blaszczynski et al 2008) found that increased bet size was not related to the need to maintain excitement or arousal levels, as in an addictive model, but rather were consistent with a cognitive model in which accumulating debts coupled with erroneous perceptions lead the gambler to increase bet size, with larger bets required to win enough to meet financial obligations. While withdrawal features in gambling are comparable in severity and character (depression, general discomfort, irritability/agitation, restlessness, anxiety and headache) to alcohol withdrawal, it remains unclear whether these symptoms “result from the inability to gamble or from the loss of an avoidant stress coping strategy”.

As for substance dependence, gambling has a multifactorial etiology. There is a strong association of PG with parental gambling and genetic transmission is estimated to account for 40-54% of variance of risk for developing PG (Shah et al., 2005). Other factors include environmental factors such as access to venues, ease of accessing money, advertising, community and cultural attitudes, ethnicity and lower socioeconomic status.

In terms of comorbidity, 40% of PG have current substance use disorders, 75% suffer major depression, 40% report serious suicidal ideation. It is estimated that approximately 1.7% of Australian suicides are gambling-related. PGs score high for impulsivity, risk-taking, substance use disorders, and borderline, anti-social, narcissistic personalities. Some 60% commit illegal acts to support their habit, usually non-violent property crimes.

There are gender differences, in that men are more likely to engage in wagering and online and sports gambling; women have a bimodal distribution of young and 45yo gambling. Early onset (before age 20) is almost universal in PG, fostered by family examples of gambling, and gifts such as scratch lotteries. The average age at treatment seeking is in the mid to late 30s.

The problems associated with problem and pathological gambling are wide ranging, as the person slips into borrowing and financial strife, sometimes into theft and lying, with impacts on work, legal problems, family problems including neglect, domestic violence and family breakdown, increasing stress, worry and depression, even personality change (irritability, becoming withdrawn).

The impacts on spouses can be enormous, including loss of trust and sense of security, loss of savings, superannuation, even the marital home, or the partner forced to resume or increase work hours. Domestic violence, emotional and physical and verbal abuses are common (often against the gambler). Children of gamblers may suffer confusion, insecurity and poor self esteem, emotional neglect, exposure to domestic arguments/violence, as well as adverse role modeling and vicarious learning.

By way of example, Professor Blaszczynski drew our attention to the structural characteristics of electronic gaming machines (EGMs). They operate within a social, alcohol-licensed environment and provide continuous, rapid cycle, multi-line multi-credits, many near wins, requiring minimal skill and fostering erroneous beliefs. The random ratio schedule of reinforcement (wins) is the most resistant to extinction of all reinforcement schedules, perhaps because of the intensity of the mounting excitement and arousal created by the unpredictability of a reward. This forms an interesting contrast to substance use disorders in that the effect of most psychoactive substances is, comparatively, predictable and constant (as long as the drug supply is secure).

Professor Blaszczynski pointed to the multiple factors interacting in an etiological model for PG: neurobiological/genetic factors as with substance dependence, interacting with personality and with environmental factors including family and peer group influences, and the wider socio-cultural setting of gambling.

This model resembles the bio-psycho-social framework generally used for conceptualising substance use disorders. One distinct difference however is the central role of belief schemas that have a mediating function in the development of problem and pathological gambling. These include the “gambler’s fallacy” mentioned above, but also superstitious beliefs (rituals, talismanic objects, cognitive ‘prayers’, promises, bargaining), biased evaluation, illusions of control and belief in the role of personal skill.

Erroneous cognitions are common in pathological gamblers (PG) and non-pathological gamblers alike, although superstitious beliefs are more common in PG, and PG are more likely to show make increasing estimates of the chances of winning during a session of play. Knowledge of the statistical reality of gambling itself does not prevent irrational beliefs during play.

The approaches to reducing harms from gambling, like those for substance use disorders, range of from public health measures to psychological and pharmacological therapies. As the risk of PG increases with consumption, measures to reduce overall consumption would be expected to have benefit: as with substance use disorders, consumption is skewed, with mean higher than median, and a small number of people accounting for a large amount of consumption. Taxation revenue incentives severely impede a regulatory public-health approach to gambling problems.

Self-help groups such as Gamblers Anonymous are effective for a significant minority of people. However, drop-out rates are very high.

Cognitive therapy is beneficial in 75-80% of cases resulting in the reduction of cognitive distortions and levels of gambling behaviour, motivation and urges to gamble. This form of therapy aims to inform gamblers that gaming machines are recreational devices on which you spend money: while it is possible to win in the short-term, in the long term, in all but the most unusual cases and extraordinary circumstances, this outcome is virtually impossible.

Behavioural interventions are designed to diminish the arousal associated with gambling, and include aversive therapy, imaginal desensitization, and stimulus control and cue exposure techniques. Positive outcomes are achieved in 20%-70% of PG with reduced arousal associated with gambling stimuli and consequently diminished urges to gamble.

The posited underlying neurobiological mechanisms of gambling suggest potential benefit of psychopharmacological interventions, however studies of lithium, SSRIs, naltrexone and olanzepine have given mixed and overall disappointing results. The studies to date have been limited by small size, high drop-out rates, short follow-up and varied outcome measures.

A further problem in evaluating treatments is that PGs do not form a homogeneous group. Accordingly, Blaszczynski and Nower (2002) have proposed a “pathways model” which distinguishes among three more or less distinct groups of PG, with implications for treatment matching.

A first pathway, encompassing mainly behaviourally conditioned gamblers, is characterized by a social context of gambling, with wins generating excitement, reinforcement and cognitive distortions leading to poor decisions. These people have less dissociation and more absorption in their gambling, briefer histories and either less severe gambling or rapid escalation in response to defined stress. They have a background of childhood and family stability, with less severe psychopathology. Substance abuse onset tends to follow rather than precede gambling problems. Cognitive-behavioural interventions are most likely to be effective with this group.

For the second and third pathways and for the second half of the seminar, and references, see the web site: