First, the Cochrane Database of Systematic Reviews (CDSR) and Epistemonikos were searched in August 2021 for prior published reviews, and PROSPERO was searched for ongoing reviews with similar overarching aims as the present review. At the time of the search, no such reviews were identified.

As shown in Figure 1, the preliminary scoping was run in databases providing health and social science research: ASSIA (ProQuest), CINAHL (EBSCO), EMBASE (Ovid), MEDLINE (EBSCO), PsycINFO (OVID), and SocINDEX (EBSCO). The scope was conducted during September and October 2021 using a combination of relevant keywords: [(gambl* OR betti* OR wage*) AND (disorder* OR addict* OR dependen* OR excessive)) AND (treat* OR therap* OR rehab* OR recover* OR intervention*].

The search results were screened to identify the intended outcomes of prevailing GD interventions, focus the aims of this review and design a main search strategy. The results from the preliminary scoping were reviewed and discussed between all authors (DJ, SO, and EA), leading to an agreement to address the underlying traits that contribute to the maintenance of GDs. These traits include factors related to GD (e.g., cognitive flexibility, decision-making, impulsiveness, or emotional regulation) that prevailing interventions are designed to address and that are central to program theory. The main search (described in the next section) was adjusted according to findings from the preliminary search by adding keywords related to emotional regulation and impulsiveness to the search string.

The main search was conducted in January and February 2022 in the following databases: ASSIA (ProQuest), MEDLINE (Ovid), PsycINFO (OVID), SocINDEX (EBSCO), and PubPsych (see Figure 1). The databases were selected to target the biopsychosocial aspects of GDs (WHO, 1992, 2019; APA, 2013).

The main search string was constructed by combining relevant keywords: (gambl*) AND (problem* OR disorder* OR addict*) AND (therap* OR treatment* OR self-help OR “user driven” OR “health care”) AND (urge* OR emotion* OR affect*). The search strategy was adapted to meet the terms given by each of the respective databases. The complete search strategy is presented in the Supplementary Information SI1.

Relevant information was extracted from the finally included references and recorded in a predesigned data extraction form. In addition to the variables author and year of publication, information that was extracted focused on study design, phenomena of interest (e.g., study aims and objectives, operationalization’s, and definition of GD), sample characteristics (e.g., sample size, gender, age, ethnicity, and socio-demographics), intervention (e.g., type of therapeutic approach, context, and mode of delivery, intended outcomes), and findings and explanation of findings. The data extraction form is presented in the Supplementary Information SI2. Information extraction was carried out in two phases. First, DJ, SO, and EA extracted one-third of the information each from the finally included references. Second, DJ, SO, and EA reviewed the extracted information across the included references by each analyzing a set of variables.

The sample sizes of the studies varied widely, with a range of 1–471 participants and a median of 27.5 (see Supplementary Table 1). Of the 34 intervention studies included in our scoping review, 23 were quantitative studies, eight were qualitative studies (case studies and interview studies), and three used mixed method designs. Of the 23 quantitative studies, 17 were categorized as exploratory studies with various pre–post outcome changes. Five studies had experimental designs, including one quasi-controlled study. According to the five-phase model proposed by Robey and Schultz (1998), most of the included studies of treatment interventions for GD fell into phase I or II in that they focused on therapeutic effect or dimensions of therapeutic effect. Here, three out of the 34 studies had medical interventions (Memantine; Grant et al., 2010a; Disulfiram; Mutschler et al., 2010; Naltrexone; Ward et al., 2018). The most reported psychotherapeutic interventions were cognitive behavioral therapy (CBT, n = 10: Castren et al., 2013; Rossini-Dib et al., 2015; Tarrega et al., 2015; Boughton et al., 2016; Smith et al., 2016, 2018; Bouchard et al., 2017; Mallorqui-Bague et al., 2018; Zhuang et al., 2018; Granero et al., 2020), which was delivered either in groups or individually, either face to face (F2F) or digitally. This was followed by motivational interviewing (MI, n = 4: Grant et al., 2011; Parhami et al., 2012; Pasche et al., 2013; Stewart et al., 2016), mindfulness based interventions (n = 3: de Lisle et al., 2011; Shead et al., 2020; van der Tempel et al., 2020), dialectical behavior therapy (DBT, n = 1: Christensen et al., 2013), acceptance and commitment therapy (ACT, n = 1: Nastally and Dixon, 2012), and transcranial magnetic stimulation (TMS, n = 1: Zack et al., 2016). Desensitization techniques or exposure were a part of multiple studies but were also used exclusively in two studies (Giroux et al., 2013; van Minnen et al., 2020). Several studies did not describe the specific content of the treatment (n = 9). Four of the interventions were delivered digitally, and the rest were delivered via F2F methods.

A key finding addresses cultural biases with generative potential for the access to knowledge about people with GD. Considering the aims of this review, the cultural biases constitute the influencing contextual factor, while access to culturally adapted knowledge and interventions comprise the outcome of GD interventions.

We identified two cultural biases in the selected studies related to Region and ethnicity and Gender perspectives. The first bias, Region and ethnicity, related to the lack of studies examining non-Western societies and cultural subgroups (e.g., ethnicity, religion, and Indigenous peoples). Three regions were overrepresented in the 34 included studies (see Supplementary Table 1): North America (n = 14), Europe (n = 10), and Australia (n = 7). Most studies were conducted in Canada (n = 10), Australia (n = 7), Spain (n = 5), and the United States (n = 4). In total, three studies were from countries in South America (Brazil), Africa (South Africa), and Asia (China/Hong Kong).

Many studies (n = 19) provided no information about the ethnic composition of the sample. In two studies conducted in Canada, the sample consisted solely of Caucasian males (Stewart et al., 2016) and Caucasian women (Piquette and Norman, 2013). None of these studies included cultural factors in the analyses. Several studies included different ethnic groups (Grant et al., 2011; Christensen et al., 2013; Gomes and Pascual-Leone, 2015; Rossini-Dib et al., 2015; Tarrega et al., 2015; Boughton et al., 2016; Bouchard et al., 2017; van Minnen et al., 2020). However, none of these studies addressed cultural background in the analysis of the results.

We found that cultural factors were given attention in two of the three studies conducted in non-Western countries (Pasche et al., 2013; Zhuang et al., 2018) and in a study that specifically addressed an ethnic minority group (Parhami et al., 2012). A common feature of these studies is that they examined culturally adapted programs. For example, Parhami et al. (2012) investigated the effects of an intervention specifically designed for Chinese immigrants in the United States. However, there were no cross-cultural studies or studies about indigenous peoples in the sample. ²

The second bias, Gender perspectives, points to a lack of research examining the impact of gender in the intervention of gambling problems. Most of the studies had samples of both males and females (n = 21). The majority of the gender-mixed studies had more male than female participants (n = 15); five studies had more female than male participants; and one study (Smith et al., 2016) had an equal gender composition. The sample size in these studies ranged from 3 to 471 participants, with an average of 43.5. Finally, one of the articles in the sample (Smith et al., 2018) presented results from three sub studies, two of which had approximately the same gender composition and with the last consisting of only male participants. The majority of the gender-mixed studies did not examine gender differences in the analysis, with Castren et al. (2013) and Rodda et al. (2017) as the only exceptions.

Seven studies used an all-male sample (Tarrega et al., 2015; Stewart et al., 2016; Zack et al., 2016; Mallorqui-Bague et al., 2018; Zhuang et al., 2018; Granero et al., 2020; Melero Ventola et al., 2020), and three studies used an all-female sample (Piquette and Norman, 2013; Boughton et al., 2016; van der Tempel et al., 2020). In the gender-homogeneous studies, the number ranged from 1 to 192 participants, with an average of 42.6. Further, there were two single case studies, one of which involved a 48-year-old man (Mutschler et al., 2010) and one involving a 61-year-old woman (de Lisle et al., 2011).

The gender perspective was not present in any of the studies that exclusively used male samples. In comparison, two of the three studies with only female participants (Piquette and Norman, 2013; van der Tempel et al., 2020) discussed gambling behavior in relation to gender roles in society.

An additional key finding addresses the commonalities across successful avenues for change. This finding presents contextual factors with generative potential for the provision of GD interventions. Considering the aims of this review, the structure and organization of health care services constitute the influencing contextual factor, while the three elements represent provision of GD interventions.

Across the included studies, certain features were echoed in the treatment courses identified as successful. These can be summarized in three overarching elements with relevance to recovery from GD: Competence, Perception, and Utilization. The element Competence encompasses insight, awareness or knowledge related to recovery from GD (e.g., understanding emotions, cognition, behavior, coping strategies, literacy, GD-specific factors, and the extent of available help or services). Examples from the included studies contain sharing experience with peers to gain knowledge about GD and learn efficient GD-specific coping strategies (e.g., Piquette and Norman, 2013; Boughton et al., 2016; Syvertsen et al., 2020). Additionally, cognitive flexibility (e.g., Grant et al., 2011; Mallorqui-Bague et al., 2018; Melero Ventola et al., 2020) and enhanced awareness of GD were associated with successful outcomes in terms of a reduced urge to gamble and GD treatment compliance (e.g., Castren et al., 2013; Jara-Rizzo et al., 2019).

The element Perception include opportunities to experience, test and adjust newly acquired understandings. For example, exposure to high-risk situations or distinctive emotional vulnerabilities during treatment sessions may offer opportunities to reevaluate and adjust one’s understanding of trigger traits (Pasche et al., 2013; e.g., Bouchard et al., 2017; van der Tempel et al., 2020). Additionally, testing and practicing coping strategies can provide opportunities to gain personal experiences with one’s coping strategies, and tailor the strategies accordingly (e.g., Grant et al., 2011; Boughton et al., 2016; Melero Ventola et al., 2020).

The third element, Utilization, involve possibilities to integrate new understandings and strategies to cope with triggering emotions, thoughts or experiences into real-life situations. Successful utilization was described as processes in which new coping strategies became embodied and integrated as an individual’s default responses to triggering emotions, thoughts, or situations. The participants from the study of van der Tempel et al. (2020) suggested that “more practice transferring skills from group sessions to home would have been valuable.” Further examples from the included studies are building traits such as abstinence self-efficacy (Gomes and Pascual-Leone, 2015) or confidence (Jackson et al., 2013), and daily practicing (Melero Ventola et al., 2020) or adherence to positive change (de Lisle et al., 2011).

First, the findings show that studies on GD interventions can be characterized by biases related to region, ethnicity, and gender perspectives (Cultural biases). Second, three elements have been identified in avenues to recovery (Competence, Perception, and Utilization). The key findings and relation between them are outlined in Figure 3.

As pointed out by De Munck and Bennardo (2019), culture is a term used by many people and in varying ways. Anthropologist Clifford Geertz famously claimed that “man is an animal suspended in webs of significance he himself has spun” (Geertz, 1973). In line with Geertz, we refer to culture as a shared set of thoughts, meanings, norms, attitudes, values, and myths (see also Abbott et al., 2018).

A key finding from the present review is a lack of studies examining non-Western societies and specific cultural subgroups (e.g., ethnicity, religion, and Indigenous peoples). As pointed out by the researchers in the South African study (Pasche et al., 2013)—and still valid today—there is a lack of research on the effectiveness of treatment for GDs in low- and middle-income countries. This bias is unfortunate because research suggested that attitudes toward gambling and treatment can vary considerably by ethnicity and cultural traditions (Williams et al., 2012; Breen and Gainsbury, 2013; Clark, 2013; Kim et al., 2016; Abbott et al., 2018).

Another finding is a lack of research examining the impact of gender in the intervention of gambling problems. Although 25 out of 34 studies included both males and females in the sample, most did not explicitly address gender differences in the analysis. Overall, the gender perspective was the most present in the studies that exclusively used female samples. For example, Piquette and Norman (2013) claimed that most of the research in Western cultures has focused on the situation of American male gamblers and that there is a need for more research on intervention mechanisms that could support female problem gamblers. Similarly, van der Tempel et al. (2020) argued that gender often influences the trajectory of a GD, as several previous studies have also found (e.g., Baxter et al., 2016; Hing et al., 2016; Carneiro et al., 2020; Håkansson and Widinghoff, 2020). In line with these two studies, we argue that future research should integrate a gender perspective to a greater extent. This would allow for more insights into gender-related experiences of gambling, treatment, and GD, as well as how gender intersects with other factors (such as age, social class, and gambling type).

Health care policies and clinical practice are grounded on knowledge about the challenges they aim to solve and characteristics of the population they intend to serve (Brown and Ashford, 2019). The observed cultural bias in gambling research is most likely influenced by a complex interplay of factors. The dominance of English in academic publishing (Politzer-Ahles et al., 2020) and global research funding disparities (Hilbrecht et al., 2020) may play a pivotal role. Variations in cultural norms and gambling regulations across regions may also contribute. For instance, in areas with strict gambling laws, like many Middle Eastern countries, conducting gambling research is likely less feasible (Aldhehayan and Tamvada, 2023; Zeng et al., 2023). Additionally, in less affluent countries, pressing issues like poverty and inadequate healthcare infrastructure are likely to take precedence. A limited or biased knowledge base may have generative potential for provision and, therefore, outcomes of GD interventions, especially for populations for which there is a lack of knowledge about them (Grant and Chamberlain, 2023). To address this bias, a multifaceted approach involving inclusive funding, diverse publication channels, and international collaborations is essential (Hilbrecht et al., 2020). Such efforts could foster a more balanced understanding of GDs. The findings of the present review clearly suggest that there is a need for more attention to the role of cultural factors and gender in future gambling research.

Change-promoting interventions mostly consist of several phases or steps that may produce the intended outcomes (Pawson et al., 2004). Our findings imply that understanding the traits and behaviors that contribute to GD and having knowledge and experience with resilient coping strategies are important steps for navigating the complex processes of change. Furthermore, an equally important phase of the transformation process is gaining awareness and understanding of the personal triggers related to emotions, thoughts, or situations. These findings are in line with previous research on recovery from GD (see, e.g., Cornil et al., 2018; Oakes et al., 2020; Pickering et al., 2020). Individuals’ reactions to emotions are situational, especially in pressured circumstances, such as unexpected trigger situations (Gross, 2013; Tenenbaum et al., 2013). Trigger situations may impair cognitive function, hence influencing an individual’s capacity to regulate emotions and utilize newly gained knowledge and experiences with resilient coping strategies (Oakes et al., 2020). Emotional regulation includes strategies to modify or replace an affective state and efforts taken to identify the underlying cause of the affective state (Gross, 2013). The understanding of emotions arising in trigger situations affects the reactions to the emotion and, consequently, the actions taken to cope in such circumstances (Tenenbaum et al., 2013). Emotional reactions are influenced by both psychological (e.g., utilization of resilient coping strategies) and contextual (e.g., social support) conditions and vary depending on how triggering the specific situation is (Tenenbaum et al., 2013). In line with this, experiences with successful coping in situations that previously have triggered an unpleasant affective state may change the meaning the individual gives to similar situations (e.g., unmanageable). The new understanding of the situation (e.g., manageable) may facilitate or enable the individual’s ability to cope with similar challenges in the future (Gross, 2013), such as gambling-related trigger situations.

Different interventions are intended to initiate the features and processes described in two of the three elements. Competence contains interventions aiming to increase an individual’s insight, knowledge, and awareness (e.g., psychoeducation, cognitive therapy, conversational group therapy, etc.). Perception comprises interventions intending to enable opportunities to practice new strategies or enhance experience with emotions, thoughts, or situations (e.g., exposure therapy, virtual reality, body-based therapy, etc.). However, it seems that understanding these traits and behaviors and having knowledge of coping strategies alone may not be enough to successfully implement these strategies in the face of unexpected triggers in real-life situations. As the findings reveal, for GD patients to effectively use their knowledge and awareness, they need opportunities that enable the process where newly learned coping strategies are transformed and integrated as the standard response. Such opportunities should be formalized and integrated as services in the health care system.

The features and processes summarized earlier in Utilization are characterized by unplanned or unexpected episodes and situations where the newly gained insight or strategies are put to the test. In considering the prevailing treatment in conventional health care systems, interventions targeting traits embedded in Competence and Perception are planned according to a specific structure and progress in controlled environments (i.e., therapeutic setting, treatment facility, etc.). In contrast, features and processes covered by Utilization are not planned and occur in outside-controlled environments (i.e., in-between therapeutic sessions or after discharge from treatment). This points to a shortcoming of the formal health care system, representing a tendency that may be especially challenging for people recovering from disorders partly triggered by traits or activities that are conventional in most societies, such as gambling. Indeed, gambling is a legal activity in most countries worldwide, and the majority of players gamble with recreational motives (Calado and Griffiths, 2016; Health Survey for England 2018, 2019; Pallesen et al., 2020). Therefore, people who have experienced problems with gambling and have recovered from a GD are dependent on resilient strategies when facing unexpected trigger situations in their everyday lives.

Interventions that promote the features described in the elements Competence and Perception are well established within the formal health care system (see, e.g., Cowlishaw et al., 2012; Abbott, 2019; Bodor et al., 2021). However, treatment success extends beyond a successful outcome at the time of discharge from formal treatment. More specifically, treatment success depends not only on positive results within the treatment context, but also on successful outcomes outside of it. Our findings suggest that this requires insights and experiences gained in the treatment facility or therapeutic setting (i.e., in controlled and structured environments) become internalized and applied in real-life situations outside of the treatment context (i.e., outside-controlled environments). In line with this, we argue that these finding have structural and practical implications for stakeholders that are engaged in with GD services. Interventions promoting the processes described in Utilization may further facilitate avenues for recovery from GD, hence representing an important area for further research and a valuable contribution to the already established interventions in the formal health care system.

Some limitations must be taken into account when considering the findings from the present study. First, this scoping review has applied an explorative approach in designing the search strategy and the analysis of the search results. However, the search strategy was informed by the findings from the preliminary search for treatment studies of prevailing GD interventions. Additionally, the search strategy and summary of the search results were driven by the intent to provide insights with relevance to today’s status in the practice field. Second, an alternative selection of databases, a different selection and combination of text words or studies written in other languages than English would generate other search results. Also, the first screening phase (titles) was primarily conducted by one author that evaluated all titles in the search result, while two authors evaluated 10% each of the titles. This entails a risk that eligible studies were not identified during the systematic search or excluded during the early screening process. However, the objective of scoping reviews is not to provide a complete overview of the existing evidence (Arksey and O'Malley, 2005; Tricco et al., 2018; Lockwood et al., 2019), and missing some studies for inclusion is the rule rather than the exception. Finally, the quality of the research design and data basis of studies were not assessed and, therefore, not considered in the analyses of the final included references. However, as shown in the discussion, the findings from this review are in line with established theories relevant to GD-related factors, as well as previous research findings with relevance to the aims of the current review.