When Friends Roll the Dice: Social Networks Drive Teen Gambling, Italian Study Finds

More than half of Italian teenagers have gambled at least once in their lives, and the strongest predictor of whether they gamble regularly is not their personality, their family income, or even their age—it is whether their friends gamble every day. That is the central finding of a large cross-sectional study of 4,176 adolescents aged 14 to 17 in Pavia, northern Italy, published in the International Journal of Mental Health and Addiction. The research, conducted by a team at the University of Pavia and the IRCCS Mondino Foundation using data from the Selfie Project surveillance program, offers one of the most detailed portraits yet of how gambling behavior spreads through the interpersonal worlds of young people.

The numbers are striking. Among the students surveyed in ten high schools between January and June 2022, 53.7 percent reported having gambled at some point in their lives, despite the fact that gambling is illegal for anyone under 18 in Italy. More alarming still, one in five of those with lifetime experience—20.1 percent—reported gambling regularly, defined as at least once a month. The study covered five gambling formats, from slot machines and video lottery terminals to lotteries, sports betting, scratch cards, and card games of chance such as poker and blackjack, and asked participants about their use of each format regardless of whether it happened online or in person.

The researchers went beyond simple prevalence figures by asking a deceptively simple question: do these teenagers know people who gamble every day or almost every day? They mapped exposure across four mutually exclusive social circles—close relatives such as parents and siblings, other relatives such as grandparents and cousins, friends, and acquaintances. Nearly half of the adolescents reported knowing at least one habitual gambler somewhere in their network. Thirty percent knew an acquaintance who gambled daily, 21 percent knew a friend, 15.8 percent knew a more distant relative, and 8.8 percent knew a close relative with a daily gambling habit.

To quantify these relationships rigorously, the team used modified Poisson regression with generalized estimating equations, a statistical approach that yields adjusted prevalence ratios while accounting for the fact that students within the same school are not independent observations. Because only ten school clusters were available, the researchers employed a bias-reduced cluster-robust covariance estimator with Satterthwaite degrees of freedom—a technical safeguard against overstating statistical confidence in small-cluster settings. The models adjusted for legal sex, year of birth, school type, and nationality, and the social exposure variables were tested both individually and simultaneously to determine which social circles exerted independent effects.

The results were unambiguous. After mutual adjustment across all four social circles, knowing a habitual gambler among friends nearly doubled the prevalence of regular gambling, with an adjusted prevalence ratio of 1.90. Acquaintances came next at 1.64, followed by close relatives at 1.42. Only the association with more distant relatives—grandparents, aunts, uncles, cousins—disappeared once the other circles were accounted for, suggesting that the influence of extended family operates through the same channels as friends and acquaintances rather than independently. Perhaps most compelling was the dose-response pattern: regular gambling prevalence climbed steadily with the number of social circles containing a habitual gambler, from 1.81 times higher with one circle to 3.86 times higher with all four, yielding an overall trend of 1.47 per additional circle.

The sociodemographic patterns revealed a telling split between gambling initiation and gambling escalation. Lifetime gambling was more common among older students and less common among those born abroad, who showed a prevalence ratio of 0.64 compared with Italian-born peers—a difference the authors attribute to Italy’s exceptionally liberalized gambling culture relative to the countries of origin of many migrant families. Yet among those who had ever gambled, nationality no longer mattered for regular play. Sex and school type told the opposite story: while boys and girls were equally likely to have tried gambling, boys were 1.81 times more likely to gamble regularly, and students in technical and vocational institutes—schools that in Italy disproportionately enroll students from lower socioeconomic backgrounds—were 1.39 and 1.53 times more likely to do so than lyceum students.

The authors interpret these findings through the lens of the social ecological model, which frames health behaviors as products of interactions between individual characteristics and influences operating at multiple levels of the social environment. Knowing people who gamble habitually may shape two distinct kinds of social norms: descriptive norms, which concern perceptions of how common gambling is, and injunctive norms, which concern perceptions of how socially approved it is. When a teenager sees daily gambling among friends, relatives, and acquaintances simultaneously, both types of norms are reinforced, normalizing the behavior across every layer of daily life. The graded association across multiple circles echoes what behavioral scientist Damon Centola has called complex contagion—the idea that behaviors spread more reliably when reinforced by multiple social contacts, much as an epidemic accelerates when a susceptible person encounters multiple sources of infection.

This epidemiological framing leads directly to a provocative policy prescription: treat adolescent gambling prevention the way public health treats infectious disease. Just as vaccination raises individual resistance while sanitation reduces exposure, the authors argue, gambling prevention should combine interventions that strengthen resistance—gambling literacy programs, critical thinking training, and emotional regulation skills targeted at vulnerable groups such as boys in technical and vocational schools—with environmental measures that cut the pathways of transmission. Those measures could include restrictions on gambling advertising and online promotion, limits on physical access to gambling venues, peer-led campaigns delivered through digital and non-digital channels, and family-based programs designed to reshape social norms and reduce the visibility of gambling within households and friendship groups.

The study’s limitations deserve honest weight. Because the data are cross-sectional, causality cannot be established: adolescents who gamble may simply seek out gambling friends, a process of social selection rather than social influence, and both dynamics may operate simultaneously. Both the participants’ own gambling and their perceptions of gambling in their networks came from the same respondent, opening the door to false consensus bias, in which gamblers overestimate how much the people around them gamble. The sample came from a single city—famously dubbed one of Italy’s gambling capitals in the early 2000s—which may limit generalizability, and socioeconomic status was not directly measured, leaving residual confounding possible. Data were also collected in 2022, before the most recent shifts in digital gambling availability.

Nevertheless, the scale of the sample, the high response rate of 95 percent among enrolled students, and the consistency of findings across multiple sensitivity analyses—including a full-sample analysis designed to rule out collider bias—lend the conclusions considerable credibility. The message for parents, educators, and policymakers is that underage gambling is not merely an individual failing or a matter of personal vulnerability. It is a socially patterned behavior that clusters within networks, escalates with cumulative exposure, and follows the same contagion logic that governs smoking, vaping, and other adolescent risk behaviors. If the goal is to stop gambling from taking root in the next generation, the evidence suggests that changing what teenagers see around them every day may matter as much as anything we teach them in a classroom.

Subject of Research: Social exposure to habitual gambling and its association with gambling involvement among Italian adolescents

Article Title: Social Exposure to Gambling and Gambling Involvement Among 4176 Italian Adolescents: A Cross-Sectional Study

Article References: Mosconi, G., Silvestri, G., Barello, S., Taverna, F., Feder, S., & Odone, A. (2026). Social Exposure to Gambling and Gambling Involvement Among 4176 Italian Adolescents: A Cross-Sectional Study. International Journal of Mental Health and Addiction. https://doi.org/10.1007/s11469-026-01734-z

Image Credits: AI Generated

DOI: 10.1007/s11469-026-01734-z

Keywords: adolescent gambling, social networks, Italy, public health, prevention, social ecological model, peers, risk factors, cross-sectional study, gambling addiction, social norms, Social