Gambling is a common leisure activity involving social interaction and positive feelings (Hilbrecht & Mock, 2019). However, in some cases, gambling becomes compulsive, and it has been reported as a behavioral addiction and considered a pressing public health issue (APA, 2013). However, despite the growing recognition of the negative health impact of compulsive gambling among adults, according to the frequency and framework in which it occurs, problem gambling as at-risk behavior for addiction is underrated and neglected among children and adolescents (Armitage, 2021). As emerged by the review of Delfabbro et al. (2016), despite international studies indicating a wide engagement of young people in gambling activities, inconsistencies in methods and samples did not clearly indicate the risk related to problem gambling among the population under 18 years of age. The most recent epidemiological studies around the world indicated that approximately two out of five adolescents between the ages of 11 and 16 had some gambling experience (Commission Gambling, 2022; Melendez-Torres et al., 2020), confirming a large diffusion of the phenomenon among younger people (King et al., 2020). In 2020, 22% of 16-year-olds in Europe reported engaging in gambling activities in the preceding 12 months (European School Survey Project on Alcohol and Other Drugs) (Molinaro et al., 2020). Italy is climbing the ranks settling at the top in prevalence, reporting that from one out of three (Ferrara et al., 2018; Mastrobattista et al., 2021) to 50% (De Luigi et al., 2018) of the minors aged under 18 had gambled at least once within the year. The high prevalence and the increasing trend may be ascribed to the improvement of online activities that broadened gambling accessibility and, therefore, the number of adolescents who engage in this behavior, indicating a risk of lasting and concerning changes (Mastrobattista et al., 2021).

Numerous studies have emphasized the diffusion of problem gambling, with data indicating a worldwide prevalence between 0.12 and 5.8% and a European prevalence between 0.12 and 3.4% (Calado & Griffiths, 2016), with a significant weight of gambling among adolescents. Already in the early 2000s, Hardoon et al. (2004) reported that 4.9% of adolescents met the criteria for pathological gambling. Data from most recent studies indicated that the number of teenagers at-risk of gambling addiction quadrupled from 2016 to 2019 (Armitage, 2021), and the prevalence of problem gambling in children increased from 1.6 to 2.2% during the same period.

Italian data reported heterogeneous but still concerning prevalences. In Cerrai and colleagues’ study, 7.1% of gamblers aged 15–19 had a “problematic” gambling profile, while including “at-risk” behavior increased the prevalence to 13.5% (Cerrai et al., 2018). The epidemiological Italian survey of Mastrobattista et al. (2021) highlighted that among 17-years-olds, 3% met the criteria for severe gambling (i.e., gambling addiction), and 3.5% were at risk of developing an addiction (Mastrobattista et al., 2021). A compulsive approach to gambling in younger, indicated by a “problematic” and “at-risk” behavior, may be ascribed to a lack of knowledge about the reasons and contexts that lead to gambling and the changes occurring in recent years that determine early access to gambling activities.

Despite the alarming data on compulsive gambling, features and dimensions related to gambling experience remain understudied in the context of adolescent health and well-being, especially considering the implications related to the transition to compulsive behavior (Armitage, 2021). Early problem gambling can be profoundly detrimental to adolescents’ mental, emotional, and social health, similar to what happens for other problematic addictions (Forte et al., 2023; Pisarska & Ostaszewski, 2020; Purwaningsih & Nurmala, 2021), and it can negatively impact emotional and cognitive development, thereby impacting school performance and increasing the risk of acquiring other addictions. Neuropsychologically, gambling involves the common neural pathway of reward, and over time, it may “hijack” the natural reward system of the brain (Hammond et al., 2014). This can cause an imbalance of cognitive control and dysfunction in motivation and gratification systems, justifying typical behavioral addiction symptoms. Adolescents are particularly susceptible to reward up-regulation and addiction because of enhanced brain plasticity (Hammond et al., 2014). Therefore, the consequences of recurring or compulsive gambling might extend well beyond childhood and adolescence and cause adverse mental, economic, relational, and social outcomes in adulthood (Potenza et al., 2019).

To adequately address this growing public health problem, this study moves beyond population prevalence as a measure of gambling among Italian adolescents. Instead, it also tries to capture changes in the intensity of gambling behavior and aspects related to this pattern, such as gambling knowledge and cognitive beliefs about gambling.

Methods

Participants and Procedure

The survey was conducted between January 2022 and September 2022. Participants were recruited from various high schools, with previous consent from the school board and the adolescents’ parents or legal guardians to participate in the investigation. No strict inclusion or exclusion criteria were defined due to the exploratory and inclusive nature of the study.

Survey Information

In General information, information about age, gender, and lifestyle was collected. Specifically, information about hobbies and leisure activities was collected.

Gambling Behavior

To classify respondents according to gambling experience, in line with previous studies (e.g., King et al., 2014), a single item was administered. The instruction provides the definition of gambling (i.e., “The gaming experience involving a cash win or loss”) and requests to indicate the frequency of gambling in the last year on a four-Likert scale. With the aim to clarify the specific quantitative range for each level of the scale, the instructions have been defined as follows: (i) No play = 0 experience of gambling; (ii) Rarely/Unusually = less than 5 times in the last year; (iii) Sometimes/Casually = from 5 times in the last year to one monthly in the last year; (iv) Frequently/Regularly = from two times monthly to weekly. Responses to this screening question were adopted to classify participants into (i) No-Gambler (i.e., no experience of gambling in the preceding year); (ii) Unusual gambler (i.e., occasional gambling experience in the preceding year); (iii) Casual gambler (i.e., different gambling experience in the preceding year), and (iv) Regular Gambler (i.e., regular experience of gambling in the preceding year) (King et al., 2014; Molinaro et al., 2020).

A series of questions was developed to assess the characteristics of the young participants’ gambling experience. Type of gambling activity (poker, sports betting, scratch card) and items focused on (i) Attempt to hide gambling Experience in the last year (Yes; No), (ii) Compulsive urge to gamble in the last year (Yes; No), and (iii) To attend illegal places to gamble in the last year (Yes; No) were administered.

Beliefs Related to Gambling

Information about personal beliefs about gambling (e.g., aid support) were included. Specifically, there were requested if participants perceived (i) high social diffusion of gambling habits (i.e., among close people such as relatives, friends, and classmates) (Yes; No); (ii) environmental diffusion of gambling (Limited; Everywhere); (iii) infiltration of criminal enterprise (Occurring; None); Aid support (Yes; No).

Knowledge (Know-How) About Gambling

To assess the current knowledge of the sample about gambling, a set of eleven items was included in the survey with three alternative responses (True, False, Unknown). Each item had only one correct response. Some examples of the items are as follows: “In gambling, the chance is the only thing that matters”; “Games involving cash stakes are prohibited for anyone under 18 years of age,” “Don’t leave the slot until it “pays” is the only winning strategy for that type of game,” “For those who want to cure their gambling addiction there are specially dedicated public services.” The overall accuracy percentage in each participant’s question set provided a Gambling Know-How score (sum of correct responses divided by the total number of items × 100). The know-how index shows good reliability within the sample (α = 0.87).

Cognitive Dimension of Problem Gambling

The Gambling Related Cognitions Scale (GRCS) (Raylu & Oei, 2004; Italian version: Iliceto et al., 2015) was adopted. It is a 23-item self-report questionnaire that uses as response format a seven-point Likert scale from 1 (strongly disagree) to 7 (strongly agree) to assess five subscales: (1) Gambling-related Expectancies; (2) Illusion of Control; (3) Predictive Control; (4) Perceived Inability to stop Gambling; (5) Interpretative Bias. Moreover, a global score can be calculated through the sum of the subscales. Higher scores on all the scales indicated a high risk of problem gambling and worse cognition on gambling. Italian validation provided two different cut-off scores of at-risk for problem gambling, respectively of 50 for males and 43 for females, considering the global score of the Scale. The internal consistency (Cronbach’s α) of each factor was acceptably high (ranging from 0.71 to 0.85).

Data Analyses

To describe gambling behavior in adolescents, we conducted different statistical analyses. Descriptive statistics were computed for demographic variables and explorative questions on gambling experiences. Estimates of the prevalence of gambling and problem gambling were carried out considering the proportion of the population fitting with the classification defined in the methods section.

Between-group comparison (No Gambling; Unusual Gambling; Casual Gambling; Regular Gambling) was conducted by using χ2 test or analysis of variance (ANOVA) according to the categorial or continuous nature of the dependent variables referring to the items on experience and beliefs on gambling (χ2) and know-how on gambling and cognition (ANOVA). According to GRCS, respondents were classified as “non at-risk” and “at-risk”gamblers, and statistical analysis was conducted considering the previous dependent variables. For post hoc analyses, the Student’s t-test for independent samples was used.

Data were presented as means (SD), frequencies, and percentages, as appropriate. Statistical significance was set at p < 0.05, and Jamovi software was used for the statistical analyses.

Results

Participants’ Data

From a sample of 4974 respondents, 149 adolescents did not complete the survey due to different reasons (absence of informed consent from the legal tutor, refusal to participate, absence from school on the day of data collection (final response rate, 97%). Data from 4825 Italian 15-year-olds were included in the analysis (age, 15.2 (Std.Dev = 1.07); 44.2% females) (see Table 1).

Table 1 Sample distribution and characteristics according to gambling experiences
Full size table

Gambling Behavior

Diffusion of Gambling Among Adolescents

The screening question regarding the direct experience of gambling in the preceding 12 months allowed the respondents categorization into four groups: (i) No Gambling (47.7% of the total sample); (ii) Unusual gambling (33.0 % of the total sample); (iii) Casual gambling (12.9% of the total sample) and (iv) Regular Gambling (6.4% of the total sample).

Data showed that the most frequent type of gambling was sports betting (n = 549; 21.7% of the respondents with Unusual to Regular gambling), followed by poker (n = 392; 15.5%) and scratch cards (n = 379; 15.0%).

Beliefs Behind Gambling Among Adolescents

Table 2 illustrates that respondents’ personal beliefs about gambling varied according to their gambling experience. The data highlighted that a larger proportion of respondents who engaged in Regular Gambling perceived the phenomenon of problem gambling in closer people compared with the No Gambling group (χ2 = 105.77; p < 0.0001) and Unusual Gambling (χ2 = 32.49; p < 0.0001), but not compared with Casual Gambling (χ2 < 1; p = 0.65). Additionally, the proportion of participants who engaged in Regular Gambling, who perceived a high diffusion of problem gambling in the context in which they lived, was significantly higher than those in the No Gambling (χ2 = 71.29; p < 0.0001), Unusual Gambling (χ2 = 44.58; p < 0.0001), and Casual Gambling (χ2 = 21.89; p < 0.0001) groups. Regarding the perception of illegality, a large proportion of Regular Gambling respondents perceived a lower diffusion of illegal gambling compared with the No Gambling (χ2 = 23.04; p < 0.0001), Unusual Gambling (χ2 = 24.01; p < 0.0001), and Casual Gambling (χ2 = 11.92; p = 0.005) groups. Finally, more respondents in the Regular Gambling group exhibited knowledge of supportive opportunities (e.g., aid and control centers) than in the No Gambling group (χ2 = 30.99, p < 0.0001), but they did not differ from respondents in the Unusual Gambling (χ2 = 1.57, p = 0.21) and Casual Gambling (χ2 = 1.50, p = 0.22) groups.

Table 2 Beliefs about gambling considering gambling experience and problem gambling
Full size table

Know-How on Gambling Among Adolescents

Accuracy of gambling knowledge among the overall sample was 59.1% (SD = 18.3). Gambling knowledge differed significantly based on the classification of gambling experience (F3,482 = 12.65; p < 0.001). Specifically, the group with No Gambling experience had lower know-how of gambling than the Unusual Gambling (mean difference = −2.89, t = −4.86; p < 0.001), Casual Gambling (mean difference = −3.97; t = −4.84; p < 0.001), and Regular Gambling (mean difference = −2.92; t = −2.65; p = 0.04) groups. No other significant differences were observed (t < 1.26, p > 0.58).

Gambling Related Cognitions Scale (GRCS) to Define the Risk for Problem Gambling

ANOVAs on the GRCS indices revealed significant differences in all subscales among groups classified based on gambling experience (refer to Table 3 for ANOVA results).

Table 3 Means and std.dev. of the global score and the subscales scores of the GRCS. ANOVAs results between groups are also reported
Full size table

Post hoc comparisons highlighted higher scores in Regular than Casual Gambling ((a) Gambling-related Expectancies, t = 10.38, p < 0.001; (b) Illusion of Control, t = 8.09, p < 0.001; (c) Predictive Control, t = 7.83, p < 0.001; (d) Perceived inability to stop gambling, t = 7.56; p < 0.001; (e) Interpretative control/bias, t = 7.56; p < 0.001) and Unusual Gambling ((a) Gambling-related Expectancies, t = 14.3; p < 0.001; (b) Illusion of Control, t = 7.02; p < 0.001; (c) Predictive Control, t = 10.62; p < 0.001; (d) Perceived inability to stop gambling, t = 8.40, p < 0.001; (e) Interpretative control/bias, t = 10.36; p < 0.001). When comparing Casual and Unusual Gambling groups, a higher score in the former group emerged for Gambling Expectancies (t = 3.58; p = 0.001) and Predictive Control (t = 2.76; p = 0.04), while no other significant differences were observed (all t < 2.27; all p > 0.07).

Problem Gambling

Diffusion of Problem Gambling Among Adolescents

According to the Gambling Related Cognitions Scale (GRCS) score, 791 out of 2524 (31.0%) respondents who indicated some gambling experiences were found to be at-risk for problem gambling (Table 4).

Table 4 Gambling risk prevalence across groups using 1.64 standard deviations from the GRCS mean as cut-off
Full size table

The proportion of respondents who were found to be at risk for problem gambling differed significantly based on their gambling experience. Respondents who had regular gambling experience were more likely to be at risk than those who had unusual (χ2 = 87.07; p < 0.0001) or casual gambling experiences (χ2 = 48.87; p < 0.0001). No significant difference was observed between the Casual and Unusual groups (χ2 = 2.48; p = 0.11).

Beliefs Behind Gambling Among Problem Gamblers

At-risk respondents were found to be more likely than non-at-risk respondents to perceive gambling diffusion among closer individuals (χ2 = 43.08; p < 0.0001) and in the social context, while no differences emerged considering the perception of illegality (χ2 < 1, p = 0.36) and supportive opportunities (χ2 = 2.82; p = 0.09) (Table 2).

Know-How on Gambling Among Problem Gamblers

ANOVA revealed that participants classified as at-risk had a significantly higher accuracy than the non-at-risk group (F1, 2522 = 61.84; p < 0.001; ηp2 = 0.03).

The Characteristics of Gambling Among Problem Gamblers

Table 5 shows that individuals at-risk for problem gambling reported a tendency to conceal their behavior, compulsivity to gamble, and frequency of attending illegal places to gamble.

Table 5 Gambling-associated behaviors
Full size table

Discussion

Given the social nature of gambling diffusion in adolescents and the impact of problem gambling on mental health and healthcare policy, analysis of their spread and characteristics across the young population has been widely recommended (Molinaro et al., 2020). The current study aimed to consider the spread of gambling and problem gambling in adolescents but also go into the characteristics behind this popular activity. In this sense, two aspects emerged from the results of the study: a high prevalence of gambling and problem gambling and an interesting pattern of adolescents who gamble.

Regarding the prevalence of gambling, our findings confirm worldwide data and previous Italian surveys (Colasante et al., 2014; Molinaro et al., 2014) indicating that about 50% of adolescents experienced at least one gambling experience in the previous year (e.g., sports betting to scratch cards). This is the first but relevant evidence indicating that even in Italy, where participation in public games with cash winnings by minors under 18 years is prohibited by law (Law No. 111 of 15/07/2011), gambling is a common leisure activity among adolescents. In this sense, as previously suggested by Raylu and Oei (2004)), analyzing not only personal but also cultural and social dimensions that may explain them is useful. Following this last consideration, the analysis of the framework of gambling experience, as well as the beliefs and cognitions that characterize different degrees of gambling, can serve the purpose. As our results suggest, increased gambling frequency seems to be inevitably associated with problem gambling. In fact, the highest prevalence of problem gambling and worse cognitive beliefs are reported by those who regularly gamble. Moreover, according to previous studies (Buja et al., 2022; Hardoon et al., 2004; Spritzer et al., 2011), gambling in adolescence is more common among males than females, with boys more at risk of developing problem gambling than girls.

It is interesting to note that based on the GRCS score, the percentage of problem gambling in the current survey is almost twice as high as previous findings in Italian samples (Molinaro et al., 2014; De Luigi et al., 2018). This aspect should be considered cautiously because of the differences in the instruments adopted to assess the phenomenon, but it again points to the importance of detecting the reason for the increase in the prevalence of problem gambling. Moreover, this evidence results in a dramatic report of the phenomenon with high implications for the healthy growth of young people.

Interesting results emerged also on beliefs about gambling. Adolescents who experienced gambling in the last year perceive it as a common activity, even adopted by close family members. At the same time, they perceive it to be illegal. Moreover, they highlight greater gambling-related expectancies, illusion of control, predictive control, perceived inability to stop gambling, and interpretative control/bias. This result seems to be the most crucial evidence of this study. Adolescents who regularly gamble showed a greater understanding of aspects related to gambling outcomes, which seems to provide a better picture of their behavior than that displayed by their unexposed peers. However, despite this apparent clear perception of the condition, a misperception was reported when gambling-related cognitions and beliefs were detected. This could be explained by an illusion of control characterizing the adolescents that leads them to believe, despite the associated risks, that they can control the negative outcomes of the gaming experience according to the vision that “If I know something, I can control it.”

On the other side, the differences emerged by the comparison between at-risk and non-at-risk adolescents of compulsive gambling suggest that at-risk adolescents live in a social condition of overexposure and misperception of the problem (e.g., illegal aspects were not perceived). These data highlight the extent of this problem. In general, when social problems, particularly those related to potentially addictive behaviors, affect the youth, there is certainly greater social mobilization. Accordingly, researchers have sought explanations and methods to counteract this phenomenon, and the results of this study can be read and interpreted in light of these preliminary considerations. Gambling increased with the growing availability of online technology and Internet access. This has made gambling more accessible, offering an on-demand and immersive experience. The digital environment engulfs young people and renders them susceptible to problematic gambling (Derevensky et al., 2019). Moreover, in recent years, an increment may also be attributed to the pandemic experience, which reduced the possibility of interacting with the real world and has increasingly directed adolescents toward the digital world (Masaeli & Farhadi, 2021). Another factor that may contribute to adolescent gambling is motivation. Contrary to monetary gains that drive adults, adolescent gambling is often viewed as an opportunity to socialize with others rather than a chance to win money. Our data support these findings by demonstrating that sports betting, the most practiced gambling activity, involves a strong social and peer exchange component (Russell et al., 2019). However, the motivational features of gambling may also serve as coping mechanisms, that is, not only as tools to interact with others, but also as a way to dissociate from stressful events. From this perspective, a vicious cycle is created in the development of problematic gambling, which could alter beliefs about gambling and cognitions, as highlighted in our study (Langher et al., 2019). From a neuropsychological perspective, problem gambling as an at-risk addiction behavior in adolescence is associated with a certain degree of brain vulnerability, as well as with some psychiatric and neurodevelopmental disorders (e.g., Yen et al., 2007), which can result in risk-taking behaviors (see for example ADHD; for reviews see Altable et al., 2022; Dekkers et al., 2022).

Adolescence is a critical period characterized by changes in affective, cognitive, and social domains affecting behavior. Changes in cognitive schemas may lead to underestimating risks and consequences, generating addictive behaviors (Andrie et al., 2019; Dekkers et al., 2022). This aspect can be used to predict risky and effective problematic behaviors and, consequently, vulnerability to addictive behaviors. The concept of vulnerability is rooted in the discourse of prevention, identifiable in the dictum “better safe than sorry.” Adolescents are vulnerable to maladaptive behavior over time (Emond & Griffiths, 2020).

One critical aspect related to brain development, motivation, and expertise is cognition, which directly or indirectly drives behavior and is associated with the risk of addiction. Research has consistently provided evidence linking problematic gambling with erroneous gambling cognitions (Hardoon & Derevensky, 2001, 2002). Individuals with problem gambling often have distorted cognitions about gambling and its effects on themselves. Our results would confirm that cognition about gambling is compromised in adolescents, particularly when they engage in regular gambling experiences, suggesting a possible negative trend that may lead to problem gambling.

Limitation

The findings from this survey are intriguing. However, the cross-sectional design of the study may lead to confusion between group-level and individual-level behaviors. For this reason, longitudinal studies are necessary to investigate approach how gambling evolves from a leisure activity to problematic behavior. These studies should track the various factors identified in this research that contribute to protective or risky behaviors leading to pathological addiction. Moreover, exploring comparable patterns in other high-risk behaviors from adolescence to early adulthood could be an intriguing area of investigation (Forte et al., 2021; Forte et al., 2023; Mastropietro et al., 2022).

Another limit should be ascribed to some of the measures adopted by the survey. Gambling behavior is examined by studies considering different exploratory questions, but no standardized tools were available to define some of the aspects analyzed in this work (beliefs on gambling, experience, and know-how on gambling). The lack of standardized measures requires caution in generalizing the results, although they have provided interesting insights and are in line with evidence from previous surveys. Still considering the methodological limitations, no data were obtained concerning on the actual frequency of gambling, which limits quantitative explorations and temporalities. Another limitation of this study is the classification of adolescents as non-gamblers, occasional gamblers, and gamblers based on a single question. Employing a standardized questionnaire would have been more appropriate for this study. However, if a relatively large amount of tool were validated in adult population to assess gambling, less questionnaires are available for children and adolescents, and most importantly, these instruments are focused on gambling as a problem and its risk of behavioral addiction. The aim of the study was to focus on a first stage on gambling experience, and the item in line with the previous study was useful to this account.

Also, sociodemographic traits, family relations, and other aspects related to risk behavior and problem gambling (such as psychiatric conditions, ADHD, etc.) should be further analyzed. Despite these limitations, this study is intended to be a snapshot of the current state in Italy in order to enable further research and increase interest in this area.

Conclusions

In conclusion, this study has provided many useful insights about the nature of gambling in adolescents, including its diffusion in a large sample of adolescents and correlates that characterize its behavioral expression. Moreover, unlike previous studies that have focused on individual aspects and experience, this study suggests an overlap of multiple aspects that characterize gambling, such as beliefs and cognitions that influence the risk of addiction.

The major achievement of this study is the picture of gambling behaviors and cognitions in early Italian adolescents that will hopefully encourage further research into gambling to elucidate the determinants of these phenomena in greater detail. These findings can be evaluated and read from a preventive perspective. Previous studies suggested the importance of intervention programs to address problem gambling and the necessity for strengthened prevention policies (Blinn-Pike et al., 2010; Bodor et al., 2021; St-Pierre & Derevensky, 2016). The importance of suggesting the prevalence, but more importantly, the cognitions, beliefs, and patterns that have emerged in the continuum of the frequency of gambling and its problematic expression may provide new insights for these programs.