Keywords

A Long and Winding Road

The path out of problem gambling is a complex, iterative journey with many twists and turns. So much so that when a sample of recovered problem gamblers were asked if they had any advice to help an active problem gambler, half of them indicated that, “there was nothing that could be done to aid in this process” (Toneatto et al. 2008, page 116).

The gambling treatment literature is replete with high rates of treatment dropouts and relapse (Aragay et al. 2015; Petry 2005). The success rates of treated gamblers can be compared to the success rates of problem gamblers who recover without treatment. Slutske (2006) looked at prevalence rates for lifetime and past year pathological gambling in two US nationwide surveys. She found that between 36% and 39% of those who had at one time in their life met the criteria for a diagnosis of pathological gambling no longer met those criteria in the past year. One third of recoveries from problem gambling could be classified as natural recovery (i.e. without treatment), half of which were stable recoveries of 5 years or more. Stable recovery, however, is by no means a typical pathway towards recovery (Nixon and Solowoniuk 2006; Reith and Dobbie 2013; Blaszczynski et al. 1991).

Complex trajectories speak to relapse rates and the difficulty of change, but they also reveal individuals successfully learning from experience (Oakes et al. 2012a, b; Vasiliadis and Thomas 2016). This is true for both those who recover unaided and with the assistance of treatment. People learn from treatment episodes, even if they do not recover on their first attempt. Hodgins and El-Guebaly (2004) reported that recovered gamblers who had sought treatment in the past used a greater number of change strategies in their attempts to quit. Jackson et al. (2008) suggested that representing for a second or later episode of treatment has traditionally been viewed as a sign of failure. Jackson et al. found, however, that representers were likely to achieve positive outcomes from treatment. They propose that treatment episodes can be understood as interactive, incremental and cumulative.

Tucker and King (1999) suggest that given the ubiquity of unsuccessful quit attempts, many addicts encounter circumstances that motivate change, but fewer encounter circumstances that maintain it. Specifically, Tucker and colleagues noted that negative events motivate attempts to control or cease substance abuse, while positive events help maintain the new lifestyle. Hodgins and colleagues (1999, 2000) reported a similar pattern in recovered gamblers.

The ideas that recovery is a complex journey, that lapses are highly probable and that one can learn to respond differently to lapses and that changes in behaviour contribute to changes in life events which in turn help in recovery are all very much part of the widely influential relapse prevention model proposed by Marlatt and Gordon (1985).

Marlatt and Gordon’s Relapse Prevention Model

When Marlatt and Gordon published Relapse Prevention: Maintenance Strategies in the Treatment of Addictive Behaviours in 1985, viewing relapse as a process rather than an outcome was a new approach. Indeed, research into instances of natural recovery or trajectories of recovery as described above was novel. Key to their approach was identifying triggers and early warning signs, using coping skills in high-risk situations, differentiating between a lapse and a full-blown relapse and avoiding throwing in the towel and giving up altogether following a lapse – the abstinence violation effect (AVE). Clinical trials of relapse prevention (RP) for a range of substance use disorders were successful (Carroll 1996).

RP for Problem Gambling

Writing in 2005, Shaffer and LaPlante noted that there had been a paucity of research addressing the effectiveness of RP in the gambling field. Nonetheless, in a review of the problem gambling treatment literature, Korn and Shaffer noted that the strength of the evidence for RP was moderate (Korn and Shaffer 2004). RP modules have justifiably been routinely included in treatment protocols (Milton 2001; Ladouceur et al. 2002; Ladouceur and Lachance 2007; Petry 2005; Raylu and Oei 2010).

A randomised trial of RP for problem gamblers was conducted by Echeburua, Fernandez-Montalvo and Baez in 2000. This was the first study of RP as an isolated component of therapy rather than as one component embedded in a treatment package. After successful treatment, gamblers were randomly assigned to individual or group RP aftercare or a control group (assessment only). The individual and group RP conditions were equivalent and achieved significantly higher success rates (82.6% and 78.3% recovered, respectively) than the control group (55% recovered.) The RP package was based on Marlatt and Gordon’s RP model.

Relapse in PG

Ledgerwood and Petry in 2006 lamented that there were few empirical studies of relapse in pathological gambling. This remains the case (Aragay et al. 2015). It is yet not clear, for example, how closely the high-risk situations for alcohol and substance use described in Marlatt and Gordon’s original model apply to the high-risk situations commonly identified in problem gambling. While many relapse triggers overlap (e.g. escaping negative emotions), there is one particular cognitive difference between relapse to problem gambling and relapse to substance use. One last drink can never make good all the past hurts. But in gambling, the fantasy of quitting after one last big win, big enough to reverse the damage done to loved ones, before finally quitting persists. This relapse-inducing fallacy even has its own name – the gambler’s conceit.

Hodgins and el-Guebaly (2004) provided the first description of relapse precipitants for gambling problems obtained from a naturalistic sample followed prospectively. They found that the categorization of relapse precipitants in a sample of problem gamblers were very different to Marlatt’s taxonomy. In Marlatt’s categorization, intrapersonal and interpersonal negative emotional states predominate, whereas for problem gamblers, cognitive and financial aspects (such as optimism about winning or a need to make money) were more likely to precede lapses. Similarly, Echeburua and Fernandez-Montalvo (2005) found that the main trigger of relapse was inadequate money management.

Toneatto (2005) noted that “unlike the chemical addictions, where cessation can reasonably be expected to lead to improvement in most areas of functioning, the cessation of gambling may be only the beginning of the process of coping with serious or intractable financial problems which may be shared with significant others and endure for many years”. He notes, “the financial repercussions may be beyond what psychological treatments can reasonably impact upon, yet such long-term debt may contribute to relapse and contribute to ambivalence” (page 79).

Pathological gamblers are more likely to rehearse positive rather than negative outcomes of gambling scenarios as indicated by responses on a modified Stroop task (Atkins and Sharpe, cited in Sharpe 2002). Furthermore, research has shown that problem gamblers show heightened autonomic arousal in the presence of gambling cues (Ledgerwood and Petry 2006). Urges are the combination of such physiological states and the gambling-related cognitions that are elicited. Sharpe (2002) describes the interplay between triggers, arousal and expectancies that make up a gambling urge: “Gambling-related problems can make the weighted importance of winning seem so enormous that losing further seems inconsequential, contributing to the cognitive biases that perpetuate gambling” (Sharpe 2002, page 20).

The identification of high-risk situations and the acknowledgement that one’s thinking can be altered in these situations are a critical component of Ladouceur and colleagues’ cognitive behavioural therapy for problem gambling (Ladouceur et al. 2002). Describing the behavioural chain of excessive gambling, Ladouceur and colleagues explain to clients “a situation or event marks the first step in the chain. The urge to gamble always surfaces within a particular context… this context, which is usually a risky situation, generates risky thoughts that activate the urge to gamble” Ladouceur et al. 2002, page 31.

Predicting Relapse

Research attempts to predict the likelihood of relapse to problem gambling using personality measures have generally been inconsistent (Blasszczynski et al. 1991; Dowling et al. 2009) with the possible exception of impulsivity (Sharpe 2002; Ledgerwood and Petry 2006). Echeburua and Ferndandez-Montalvo (2005) concluded that situational elements rather than personality dimensions were more important in predicting relapse and that this should generate therapeutic optimism. Focusing on situations, the RP model did provide a helpful heuristic and treatment framework. Nonetheless, criticisms of Marlatt’s earlier classifications of relapse triggers included a lack of emphasis on urges as well as not allowing adequately for idiosyncrasies in an interactive, fluctuating process (Marlatt and Witkiewitz 2005).

Taking on board such criticisms, the previously linear RP model was reconceptualised into a multidimensional and dynamic model which includes feedback loops between tonic processes, phasic responses, distal risks and contextual factors (Witkiewitz and Marlatt 2004). This model is compatible with mathematical models that describe the behaviour of complex systems such as catastrophe and chaos theories. Witkiewitz et al. (2007) successfully applied one such “dynamical systems” mathematical model to data sets from Project MATCH (a multisite alcohol treatment study). West (2006) also suggested that chaos theory is useful, at least as a metaphor for the transition from an addicted state to a not-addicted state.

McCown and Chamberlain (2000) applied chaos theory specifically to relapse in gambling addiction. They noted that “one of many implications of chaos theory is the apparent paradox of phenomena that are simultaneously completely causally determined but essentially still mostly entirely unpredictable” (page 184). “Chaotic phenomena although unpredictable are not random” (page 188). Examples might be the pattern of smoke rising or the flow of eddies in a stream of water as well as lapses and recovery from problem gambling. They note that, “at specific and identifiable critical periods in the process of recovery, problem gamblers may be extraordinarily influenced by very small events, which may have extreme and unpredictable implications for future functioning… sensitivity to initial conditions explains both relapse and the desire to recover” (page 193).

Overall, the implication is that it is difficult to predict relapse. Clinically, however, individuals can usefully learn from their lapses or relapses. To paraphrase the Danish philosopher Kierkegaard – life can only be understood backwards, but it must be lived forwards. Problem gambling clients can identify their own distal factors or early warning signs, their own proximal factors or high-risk situations and their own internal and external triggers. A recent extension of relapse prevention programme does this by incorporating mindfulness (Bowen et al. 2011)

Mindfulness-Based Relapse Prevention

Mindfulness meditation has been incorporated into treatment protocols and successfully applied to pain management (mindfulness-based stress reduction (MBSR)) and to depression (mindfulness-based cognitive therapy (MBCT)). Based on the success of these therapies, Witkiewitz, Marlatt and colleagues proposed a protocol which formally integrates mindfulness and relapse prevention (Witkiewitz et al. 2005).

How does mindfulness-based relapse prevention (MBRP) differ from standard RP? Bowen et al. (2014) note that despite the supporting evidence from clinical trials, RP has some potential shortcomings. Although urge surfing has always been a component of RP, there has been an emphasis on avoidance of high-risk situations or the use of alternative coping behaviours to attempt to control the causes of negative affect. MBRP, however, extended the mindfulness approach to the full range of life experiences beyond just a strategy for dealing with urges.

MBRP in the format of an 8-week aftercare programme was evaluated in a pilot study where it was compared to treatment as usual (TAU) (8 weeks of standard aftercare delivered in a 12-step group format) (Bowen et al. 2009). In this study, MBRP was shown to be as effective at 4-month follow-up as TAU. It was also found to be acceptable to clients and feasible in that most clients did some mindfulness meditation practice at home. More recently, Bowen et al. (2014) compared MBRP with standard RP and with TAU (12-step oriented programme) at 3-, 6- and 12-month follow-up. At 3 months, there was no difference between the groups. At 6 months, both the RP groups showed significantly less relapse than the TAU group. By 12 months, the MBRP group had better outcomes than the standard RP group and the TAU group. The authors suggest that the more enduring effect of MBRP is explained by the participants’ improved ability to recognise and tolerate discomfort associated with craving or negative affect due to continued practice of the MBRP skills over time. Although they do not indicate what skills, they reported that over 88% of the participants in the MBRP group said they were using skills from the programme at least once a week.

A clinician’s guide to the MBRP programme is available (Bowen et al. 2011). This describes the eight-session programme in detail. Each session consists of psychoeducation with handouts, a mindfulness meditation practice, group discussion and home practice. Throughout the programme, mindfulness practices are applied to the range of experiences associated with relapse – feelings, thoughts, reactions and triggers.

MBRP has not yet been evaluated for problem gamblers. However, both standard RP and mindfulness have been found to be effective in the treatment of problem gamblers (Echeburura et al. 2000; Echeburua and Fernandez-Montalvo 2005; McIntosh et al. 2016. See the Metacognitive and Mindfulness chapter in this book).

Life Is More than Not Gambling

As noted above, people recovering from gambling may experience more difficulties than those recovering from other substances in restoring relationships or taking up new passions for several reasons. McCown and Chamberlain (2000) observe that whereas people in early recovery from substance abuse often feel somewhat better quite quickly because they begin to reclaim some of their health, it may take months for problem gamblers to experience a sense of hopefulness in the face of overwhelming financial and relationship problems (page 130). Trust may take longer to re-establish for gamblers than for other substances. The families of those recovering from substance use can more easily see signs of relapse, but because relapse to gambling is harder to detect immediately, it may take years before the gambler earns back the trust of family members or business partners. Furthermore, the financial situation left by the gambling may take many years to repair, and this may severely limit new options and impact on the family (Toneatto 2005).

RP programmes emphasise the importance of lifestyle balance in long-term maintenance. In a proof of concept study, Jackson et al. (2013) noted that problem gamblers often engage in few social activities other than gambling. After treatment, they can be left with considerable unstructured time and inadequate social skills and feelings of emptiness and boredom. The (Re)Making Meaning Project was a 9-month programme focusing on normalising non-gambling leisure in which former problem gamblers sampled various activities such as barbeques, computer lessons and French Jive lessons with community volunteers. Despite dropouts, the results were promising enough to warrant further research.

People quit gambling for a reason. Quitting gambling may initially involve an avoidance goal (they might quit to avoid financial ruin and emotional pain). Continuing to refrain from gambling may involve an approach goal (reconnecting relationships, doing long-needed repairs around the home, taking a family holiday, travelling, studying or resuming an abandoned passion such as music or sport). It is not enough just to stop gambling. One also stops gambling when one dies. Recovering from problem gambling is done in the service of life.