Pathological gambling is a condition affecting roughly 0.5% of the adult population (1, 2), and up to 3–5% of adults are reported to suffer from past-year problem gambling with or without fulfilment of diagnostic criteria (3). Despite documented knowledge about effective treatment for the disorder (4), several potential barriers may limit treatment-seeking behavior in gambling patients. For example, it has been suggested that in individuals with pathological gambling, treatment-seeking may—paradoxically—be inversely associated with gambling severity (5). In addition to individual factors complicating treatment-seeking, structural and organizational issues also may constitute barriers to treatment for individuals with problem gambling (6), and clearly, widespread availability is crucial to treatment coverage.

Sweden, in contrast to extensive epidemiological studies of gambling in the population (7, 8), is an example of a setting where no national strategy for gambling treatment hitherto has been established. National policies for allocation of gambling treatment have been lacking, and few regions have established actual structured treatment for problem gambling, such that in many settings throughout the country, no actual treatment has been available (9). At the same time, Sweden is seeing a vast increase in media advertisements of primarily online gambling and online sports betting, and an increased market share of overseas online gambling operators. This trend has called for action in gambling prevention and treatment by policy makers (10).

Online gambling could potentially be associated with a higher risk of problem gambling, expressed as symptom criteria of pathological gambling, compared to more traditional, land-based gambling (11–14). However, this is not a consistent finding (13), and other data indicate that online gamblers may be less likely to be categorized as problem gamblers (15). Also, it has been suggested that individuals engaging in problem gambling online may be less likely to seek treatment than problems users of land-based gambling modalities (13), and the proportion of online gamblers among treatment-seeking gamblers has been reported to be low (16). In recent years, there is an increasing attention to psychiatric comorbidities in gambling patients with or without treatment (16), and data are mixed as to whether rates of psychiatric comorbidity in online gamblers are higher or lower (7, 13, 16, 17). Gainsbury argued that poor health and substance use may be more common in online gamblers (13). On the other hand, Hing and coworkers reported lower rates of psychological distress among online gamblers than among land-based gamblers (16). Blaszczynski and coworkers reported that exclusive online gamblers were less likely to be problem gamblers, and less likely to suffer from psychological distress compared to land-based gambles (17). Also, in a Swedish population survey conducted in 2008, psychological distress was less common in Internet gamblers (7), whereas Jiménez-Murcia and coworkers identified no difference in psychopathology between land-based and online gamblers (18).

This study is based on the first 16 months with a regional, out-patient facility established within the public health care system, and in a setting where such treatment for pathological gambling previously has been unavailable in the medical sector (9). The aim of this study was to assess the characteristics of treatment-seeking patients in a treatment modality new to this setting. Particularly, given the vast exposure to online gambling advertising in the media in this setting (10), we aimed to study the extent of online gambling as opposed to more traditional gambling in treatment-seeking patients in a new treatment setting and to investigate whether these different gambling modalities differed with respect to gender and prevalence of psychiatric comorbidity.

During the study period, 106 individuals contacted the facility in order to apply for treatment for problem gambling, and showed up for formal assessment. Seventy-nine percent (n = 84) sought treatment themselves, whereas 14% (n = 15) were referred from other psychiatric facilities, and 7% (n = 7) from primary care.

The median age of included patients was 31.5 years (range 18–72), and 80% of the patients (n = 85) were men. The Mann–Whitney U test demonstrated a nearly significant difference in age between women (median = 43 years, n = 21) and men (median = 30 years, n = 85), p = 0.05. Problematic gambling types were available from the records in 95 cases, and a clearly dominating gambling type was available from the records in 91 cases. The distribution of problem gambling types is displayed in Table 1. Among problem gambling types reported, the dominating gambling type was online casino in 55% of cases (n = 50), online sports betting in 24% (n = 22), land-based EGMs in 4% (n = 4), land-based casino in 3% (n = 3), and other games (such as online poker or bingo) in the remaining 13% (n = 12).

Fifty-eight percent (n = 61) received a psychiatric diagnosis other than pathological gambling (all F diagnoses), and these disorders are displayed in Table 2. Non-substance-related psychiatric diagnoses (F diagnoses other than F10–F19) were present in 52% (n = 55). The chi-square test demonstrated that the difference in psychiatric comorbidity between genders only approached statistical significance (χ² = 3.73, df = 1, p = 0.054); a comorbid diagnosis was present in 76% of females and 53% of males. For non-substance-related psychiatric comorbidity, the chi-square test demonstrated a significant association between gender and comorbidity (χ² = 6.20, df = 1, p = 0.01), as this comorbidity was more common in women (76%) than in men (46%).

The chi-square test demonstrated a significant association between problem online casino gambling and gender (χ² = 13.95, df = 1, p < 0.001). Problem online casino gambling was more common in females (95%) than in male patients (50%). Also, with the chi-square test, online sports betting was significantly associated with gender (χ² = 18.78, df = 1, p < 0.001), as this was reported by 53% of males and by none of the females. No significant association was seen between psychiatric comorbidity and problem online casino gambling (χ² = 2.23, df = 1, p = 0.13). A psychiatric disorder was present in 63% of online casino clients and in 47% in others.

This study described the treatment-seeking patient population from the first 16 months of the first health care-based unit for pathological gambling treatment in a Swedish region, in order to describe patient characteristics in a situation where the health care system opens its first treatment option for this disorder in the region. The main finding is the overwhelming majority of treatment-seeking patients who report online casino or online sports betting to be one of their problem games, or their predominating problem game. In addition, the study lends support to previous findings of a high degree of psychiatric comorbidity in pathological gambling patients and demonstrated potentially higher comorbidity in treatment-seeking women than in men.

The high rates of online gambling in this population stand out compared to several reports from other previous treatment settings. For example, in a large material of treatment-seeking gambling patients in Barcelona, Internet-based gambling, as well as sports betting specifically, was uncommon (21), and Internet-based games constituted only a few percent of problem gambling in helpline callers in a US study (22), 5% of helpline callers in South Africa (23), and 16–17% of helpline users in an Australian study (24). Thus, the high proportion of online gambling—especially the high rates of online casino as the sole game or one of the problematic games reported by patients—stands out in this cohort of treatment-seeking patients, as these rates reported above from other settings have been considerably lower but also as some data on treatment-seeking behavior have indicated that online gambling may in fact be associated with a lower likelihood of seeking treatment (5), i.e., in contrast to the figures seen here.

The high rates of online gambling in treatment-seeking gamblers may be seen as worrying; problem gambling online may be associated with the use of unregulated sites (7), potentially making problem prevention more difficult. Also, although not a consistent finding, several reports have documented a potentially higher degree of problem gambling in online gamblers (11–14). The rates of non-online games were very low in this study, indicating that the predominance of online gambling may be specific to this setting. A recent government evaluation and bill proposal investigated the gambling market of Sweden and reported a clearly increasing market share for gambling companies operating online and overseas, and outside Swedish oligopoly based regulations for gambling. This share of the gambling market was reported to have increased dramatically over the past few years, including vast exposure in media commercials (10). Thus, while the high rates of online gambling in this treatment-seeking cohort may be surprising in comparison to previous publications from other settings, they clearly display an interesting picture of real-world gambling patients showing up when gambling treatment is started, in a setting exposed to large amounts of online gambling commercials.

While online gambling was the overwhelmingly most common type of game in this study, its association with gender differed depending on the type of game. Previously, online gambling has been associated with male gender (12, 25). Here, the same association was seen for online sports betting, whereas Internet casino was clearly associated with female gender; in fact, male and female patients in this cohort represented clearly different patterns of online gambling, especially as almost all females reported online casino, whereas none of them reported online sports betting.

The study of psychiatric comorbidity in different modalities of gambling is a relatively recent research topic. Blaszczynski and coworkers reported a lower degree of psychological distress in gamblers engaging in online gambling (17). In contrast, in the previous study, rates of comorbidity were high in online gamblers; however, without showing a significant difference compared to other gamblers.

However, female gender was linked to higher rates of psychiatric comobidity. Treatment-seeking female gamblers previously have been reported to have a higher degree of psychiatric problems than their counterparts (20, 22, 26), and women may be more likely to gamble in response to psychological distress (27). Thus, this study further supports the need for psychiatric assessment in female problem gamblers. In this study, a majority of treatment-seeking patients were men, consistently with previous publications (28). An important finding from previous literature has been that although gambling is generally associated with male gender, progression from gambling onset to problem gambling may be more rapid in women (29), and that time from problem gambling to treatment was shorter in women (30). Also, gambling for the purpose of mood regulation has been reported to be more common in female online gamblers than among their male counterparts (31). These findings, along with the findings of higher comorbidity in women in this study and others, demonstrate the need for further research and focus on treatment of gambling in women.

This study displays the characteristics of treatment-seeking individuals at a gambling treatment facility opening in a setting without prior health care-based treatment for problem gambling. The high rates of online gambling and psychiatric comorbidity shed light on the treatment needs of problem gamblers in a setting where treatment has been largely underdimensioned, and where gamblers are extensively exposed to online gambling commercials in media. Treatment of pathological gambling may need to address psychiatric comorbidity with a particular emphasis on female clients.

The study was approved by the regional ethics committee, Lund, Sweden (file number 2017/341). Consistent with this permission, based on the study being a retrospective chart review, no consent procedure was carried out.

AH carried out assessment of the hospital records included, statistical analyses, and wrote the manuscript. EM ad MZ carried out assessments of treatment-seeking patients. All authors reviewed and approved the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.