Predictors of early dropout in treatment for gambling disorder: The role of personality disorders and clinical syndromes

Highlights

• Comorbidities and predictors of early dropout in a six months therapy for gambling disorder (GD) were investigated.

• 194 pathological gamblers and 78 healthy controls were recruited.

• High comorbidities were highlighted in the clinical group.

• Negativistic personality disorder, antisocial personality disorder, drug dependence and PTSD were associated with dropout.

• This association was controlled for severity and duration of GD, typology of gambling and psychopharmacological treatment.

Abstract

Several treatment options for gambling disorder (GD) have been tested in recent years; however dropout levels still remain high. This study aims to evaluate whether the presence of psychiatric comorbidities predicts treatment outcome according to Millon's evolutionary theory, following a six-month therapy for GD. The role of severity, duration of the disorder, typology of gambling (mainly online or offline) and pharmacological treatment were also analysed. The recruitment included 194 pathological gamblers (PGs) to be compared with 78 healthy controls (HCs). Psychological assessment included the South Oaks Gambling Screen and the Millon Clinical Multiaxial Inventory-III. The “treatment failure” group (n = 70) comprised PGs who prematurely dropped out of the treatment whereas the “abstinent group” (n = 124) included PGs who completed the treatment regardless of whether the outcome was successful or not. As expected, the presence of psychiatric comorbidities was highlighted as a significant predictor in dropping out of the therapy. Specifically negativistic personality disorder, antisocial personality disorder, drug dependence and PTSD were associated with early dropout. These variables were predictive of treatment outcome independently from the typology of gambling, severity, duration of the disorder and pharmacological treatment. Implications for psychological and psychiatric care are discussed.

Introduction

Fittingly with several studies that highlighted common elements between gambling disorder (GD) and substance use disorder (Leeman and Potenza, 2012, Goudriaan et al., 2005, Petry et al., 2014, Potenza, 2014) the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) considers GD as a behavioral addiction (American Psychiatric Association, 2013. Diagnostic and statistical manual of mental disorders − 5th ed.). GD is conceptualized as a maladaptive pattern of wagering that persists despite negative consequences in multiple areas of functioning such as finances, relationships, psychological and medical health (La Barbera and La Cascia, 2008). It included both problem gamblers, whose prevalence ranges from 1.3% to 3.8% of the general population in Italy, and pathological gamblers (PGs) who are estimated to be between .5–2.2% of the population (Serpelloni, 2013). The current scientific debate conceptualizes behavioral addiction as a repeated behavior leading to significant harm or distress. The behavior is not reduced by the person and persists over a significant period of time. The harm or distress is of a functionally impairing nature (Kardefelt et al., 2017). Treatment options for GD include several options: pharmacological therapies that aim to restore the dysregulation in the neuronal circuits responsible of the endophenotypes typical of this behavioral addictive disorder (Achab and Khazaal, 2011, Lupi et al., 2014) and non-pharmacological approaches that work on different determinants of GD such as cognitive distortions, craving, comorbidities, social vulnerability and inhibitory dyscontrol (Cowlishaw et al., 2012; Rosenberg et al., 2013). The results achieved mainly by the integrated approach of both pharmacological and psychological therapy are highlighted by several studies that compare an active treatment to no treatment or wait-list control conditions (Łabuzek et al., 2014; Rash and Petry, 2014; Yakovenko et al., 2015). Nevertheless, a large proportion of treated patients abandons treatment before completion (Dunn et al., 2012, Ledgerwood and Petry, 2006a, Ledgerwood and Petry, 2010); dropout percentages range from 14% to 50% with a median of 26% (Melville et al., 2007). Notably, there is a lack of consistent definitions (Ladouceur et al., 2001), since dropping out is possible if treatment is stopped without notice or without the consent of the therapist. Dropout can also occur at any time of the therapy. For example, a patient may decide to leave prior to initiating treatment, prior to completion of treatment or prior to completing follow-up assessment. In this paper we will refer to dropout as treatment abandonment within a six month-treatment.

Several variables have been associated to dropout from treatment programs for PGs, including sociodemographic variables (Echeburua et al., 2001, Hodgins et al., 2004, Leblond et al., 2003), gambling-related variables (Milton et al., 2002, Robson et al., 2002) treatment-related variables (Ryan et al., 1995, Leblond et al., 2003) and psychological traits (Echeburua et al., 2001, Pelletier et al., 2008, Petry, 2001). As a matter of fact, it is not a rare event that PGs are diagnosed with psychiatric comorbidities that play a role in the evolution of the disorder. Indeed, the presence of a comorbid psychiatric disorder is showed as a significant predictor of dropout both for GD (Ramos, Grille et al., 2015) and for drug abuse treatment (Ball et al., 2006). From a clinical point of view, a possible explanation regarding the association between psychiatric comorbidities and treatment failure for GD concerns the negative influence on compliance, the effect in increasing impulse dyscontrol finally contributing to produce reward-processing abnormalities. Nevertheless, objective data on the impact of psychiatric comorbidities on therapeutical effectiveness and specifically on dropout in GD therapy are still inconsistent and produce mixed results. Given these premises, this study was based on two hypotheses: a) it was predicted a higher presence of psychiatric disorders in PGs compared to HCs; b) it was predicted a higher rate in dropping out of the therapy for PGs who are diagnosed with a psychiatric comorbidity. Finally, starting from these hypotheses, the role of personality disorders and clinical syndromes, according to Millon's evolutionary theory (Millon, 1981) was explored, in order to predict treatment outcome, taking into account the role of severity of GD, duration and typology of gambling (mainly online or offline). Millon's theoretical model is grounded in evolutionary theory. In essence, it seeks to explicate the structure and styles of personality with reference to deficient, imbalanced, or conflicted modes of survival, ecological adaptation and reproductive strategy. Identifying personality traits or clinical syndromes that can be used as markers for predicting those who will drop out of the therapy can be of great importance in promoting an effective and tailored treatment protocol, with the purpose of maximizing the effectiveness and the treatment efficacy.