Prevalence and predictors of problem gambling among Swedish male elite football players: a cross-sectional study

Abstract

Objectives

This study aimed to assess the prevalence of problem gambling and gambling behaviours among Swedish elite male football players and explore their awareness of gambling-related policies within their clubs. A secondary objective was to identify factors associated with problem gambling in this group.

Methods

A cross-sectional survey was conducted in 2021 among players from the 32 clubs in Sweden’s top two football leagues (Allsvenskan and Superettan). Data were collected through an online survey utilising the Problem Gambling Severity Index (PGSI) alongside additional questions on gambling types, mental health, alcohol use, match-fixing and policy awareness. Multivariable logistic regression was employed to analyse predictors of problem gambling.

Results

A total of 416 footballers participated, yielding a response rate of 58%. The prevalence of gambling in the past year and problem gambling (PGSI score≥3) was 60.6% and 6.6%, respectively. While 72% of past-year gamblers reported engaging in sports betting, only slots (OR=7.46, p=0.01) and bingo (OR=6.08, p=0.02) were significant predictors of problem gambling. Only 32% of respondents were aware of a gambling policy within their clubs. Additionally, 17.8% reported depressive symptoms, while 13.9% experienced anxiety.

Conclusions

This study highlights the high prevalence of problem gambling among elite Swedish footballers, with slots and bingo posing higher risks. Sports betting, while prevalent, may contribute to the gamblification of football. The role of gambling sponsorships in normalising gambling among footballers requires further investigation. There is a need for longitudinal studies on gambling behaviours and policy awareness within clubs, and comprehensive preventive strategies to address gambling-related harm among footballers.

WHAT IS ALREADY KNOWN ON THIS TOPIC.

WHAT THIS STUDY ADDS

• The prevalence of problem gambling among Swedish elite footballers was 6.6%, lower than in previous studies but still higher than in the general population.

• Slots and bingo, two types of high-speed games, were identified as predictors of problem gambling in this group.

• Player awareness of gambling-related policies within their clubs was low.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• There is a need for preventive intervention strategies within football clubs, including the development and promotion of gambling policies and restrictions for players.

• This study could inform policy changes in professional football leagues, advocating for the removal of gambling sponsorships and stricter regulations on athlete betting to protect the sport’s integrity.

• Future research could explore the cocreation and implementation of preventive interventions in football clubs and assess their effectiveness.

Introduction

Gambling is a potentially harmful activity that affects not only individuals but also society as a whole. Gambling-related harm ranges from financial distress and broken relationships to more severe outcomes such as addiction and even suicide.¹ In recent decades, gambling has become increasingly digitised, a trend that was accelerated by the COVID-19 pandemic.² What was once confined to land-based venues, such as sports betting and casino games, is now widely available online.³ In Europe, the Gross Gambling Revenue for online gambling rose from €25.7 billion in 2019 to €38.2 billion in 2022, an increase of 49%.⁴ This growth has been accompanied by a rise in the harms associated with online gambling, as reflected in data from Nordic helplines, which indicate an increase in gambling-related issues, particularly with online casino games and betting formats.^{3 5} Research indicates that certain types of gambling, such as electronic gambling machines (both land-based and online-based) and sports betting—especially in-game and micro betting—are most commonly associated with individuals who have gambling problems or are at risk.5,8 These forms of gambling share features such as high speed, frequent events, high payout rates, variable bet sizes and gamification elements.^{9 10}

Although gambling is widespread, the gambling industry and its advertising are not well regarded by the public.¹¹ To improve their reputation, gambling companies often form partnerships with trusted sectors, such as non-profit organisations and sports. Associating with sports, particularly football, helps reduce the stigma around gambling and contributes to its normalisation.^{12 13} This is achieved through sponsorships of football associations, teams and athletes.¹³ On a global scale, football is the largest sport and sports betting has long been intertwined with it. Sharman¹⁴ reported that 10 Premier League clubs in the UK had betting logos on their jerseys, and all but three had betting partners. Such sponsorships create the impression that betting is an integral part of sports, fostering positive attitudes towards gambling among fans.¹⁵ Research exhibits that children as young as 5 can associate betting companies with football teams, suggesting that vulnerable groups, such as minors, are particularly susceptible to the normalisation.¹³ Furthermore, increased exposure to gambling advertising is associated with higher gambling participation and greater severity of problem gambling.^{16 17} Gambling has also become normalised among athletes themselves,¹⁸ yet only 2%–4% of athletes report that their clubs have rules, policies or educational programmes related to gambling.¹⁹ Athletes are considered an at-risk group for gambling due to specific personality traits, such as strong drive for challenges and competition, sensation-seeking, reward dependence, impulsivity and risk-taking behaviour.²⁰ Gambling among athletes, particularly footballers, is problematic for many reasons, with several high-profile cases of players betting on their own games or engaging in match-fixing, which undermines the integrity of the sport.^{14 21}

Despite the widespread nature of the issue, there is limited research on gambling habits and problem gambling among elite male football players. Additionally, previous studies have used relatively small samples of senior players. Jensen et al²² surveyed 323 Danish and Swedish elite players, including 205 senior players, and found a 16.1% prevalence of at-risk gambling, while Vinberg et al²³ found a prevalence of 14% among 148 Swedish footballers. The latter study used the Problem Gambling Severity Index (PGSI),²⁴ with a score of 3 or more indicating at-risk gambling. Additionally, Kilic et al²⁵ surveyed 149 Australian footballers using three items from the NORC Diagnostic Screen for Gambling Problems²⁶ and found that 23.6% reported problem gambling. These rates are higher than those in the general population, where a pooled global estimate suggests that 2.4% experience at-risk or problem gambling.²⁷ A review synthesising results from longitudinal studies examining risk and protective factors for problem gambling found that factors such as use of alcohol and other substances, depression, impulsivity, sensation seeking and the number of gambling activities predicted problem gambling.²⁸ Within the context of footballers, the Vinberg et al study also found that alcohol consumption was associated with gambling problems.²³ Jensen et al²² also found that depression and sports anxiety were associated with at-risk gambling among football players, while Kilic et al²⁵ identified recent injuries as a risk factor for gambling problems. To the best of our knowledge, there are no studies that have examined the association between types of gambling and their relationship to problem gambling among footballers.

Given the limited understanding of elite footballers’ gambling habits, this study aims to expand knowledge on gambling behaviours, problem gambling and awareness of gambling policies among elite male football players. The primary objective is to assess the prevalence of problem gambling and engagement in various types of gambling over the past year among players in Sweden’s top two football leagues. The secondary objective is to explore factors associated with problem gambling in this population.

Methods

Study design

A cross-sectional study was conducted with players in 32 clubs in Sweden’s top two football leagues (Superettan and Allsvenskan). Data were collected from players through an online survey conducted between February and November 2021.

Setting

The data collection occurred during the COVID-19 pandemic. Although games followed the regular schedule, there were restrictions on the number of spectators allowed. At the start of the season, games were played without spectators, but restrictions were eased in the summer, and by the latter part of the season, they were completely lifted.

Participants

With the support of the Swedish Professional Football Leagues, which only oversees the top two male football leagues (thereby restricting the study to males), email contact lists for all players registered as of February 2021 were obtained. All players, N=717, were invited to participate in the survey through email, with up to five reminders sent to those who did not respond to the initial invitation. Prior to participation, all players were provided with information about the study and gave informed consent. Neither the public nor patient representatives were involved in the development of the survey.

Data sources/measurement

In addition to background questions, including age, educational level, training hours, club changes in the past year and awareness of or personal experiences with match-fixing, several scales were employed to assess mental health and alcohol consumption. Problem gambling was measured using the PGSI, which assesses gambling over the past year. The PGSI was designed for use in the general population²⁹ as a screening tool.³⁰ Participants answered nine items on a four-point Likert scale: 0 (never), 1 (sometimes), 2 (often) and 3 (always), with total scores ranging from 0 to 27. A score of 0 indicates non-problem gambling, 1–2 low-risk gambling, 3–7 moderate-risk gambling and ≥8 problem gambling.²⁴ Given that problem gambling is relatively rare in the population, a cut-off score of three points was used to define problem gambling, consistent with prior studies on footballers’ gambling habits.^{22 23} Participants also reported their engagement with eight types of gambling over the past 12 months: sports betting, lotteries, horse betting, casino games, poker, slots, bingo and number games.

Depressive symptoms over the previous 2 weeks were assessed using the Patient Health Questionnaire (PHQ-9), which has demonstrated good reliability and validity.³¹ Scores range from 0 to 27, with 0–4 indicating no or minimal depression, 5–19 mild depression, 10–14 moderate depression, 15–19 moderate to severe depression and 20–27 severe depression.³¹ A cut-off score of 5 points was used to identify depressive symptoms in this study.

The Generalised Anxiety Disorder 7-item scale (GAD-7) was used to measure anxiety over the last 2 weeks. This scale has demonstrated good reliability and validity. Scores range from 0 to 21, with 0–4 indicating no or minimal anxiety, 5–9 mild anxiety, 10–14 moderate anxiety and 15–21 severe anxiety.³² A cut-off score of 5 points was used in this study to assess anxiety symptoms.

Sleep disturbance over the prior 2 weeks was measured using the Insomnia Severity Index (ISI), a reliable and valid measure in clinical contexts.³³ The ISI comprises seven items, with total scores ranging from 0 to 28. Scores of 0–7 indicate no insomnia, 8–14 subthreshold insomnia, 15–21 moderate insomnia and 22–28 severe insomnia.³³ In this study, a cut-off of eight points was used to indicate insomnia.

Alcohol consumption over the past year was measured using the Alcohol Use Disorders Identification Test Consumption (AUDIT-C), which evaluates the frequency and quantity of alcohol consumption and binge drinking.^{34 35} Scores range from 0 to 12, with scores of 5 and above indicating at-risk drinking. This AUDIT-C has demonstrated good psychometric properties in the Swedish context.^{35 36}

Bias

We recognise potential biases such as selection bias due to voluntary participation and response bias from those who chose to respond. We sent up to five reminders to non-respondents to reduce non-response bias. While we did not address all biases, the large sample size and support from the Swedish Professional Football Leagues help provide a representative snapshot.

Statistical methods

Descriptive statistics were used to summarise the data. For variables following a normal distribution, means and SDs were reported. For variables with a skewed distribution, medians and IQRs were presented. Categorical variables with binary outcomes (yes/no) were expressed as counts and proportions of ‘yes’ responses.

For the inference statistics, the PGSI, dichotomised at three points (≥3 points coded as 1), served as the outcome variable. Based on previous research^{22 23 28} and the identified knowledge gap, we focused on the following mental health, substance use and gambling-related variables: PHQ-9 (dichotomised at 5 points), GAD-7 (dichotomised at 5 points), AUDIT-C (dichotomised at 5 points), ISI (dichotomised at 8 points), current use of snuff (yes=1), participation in each of eight gambling types (past year participation=1) and no gambling (yes=1). We also added club changes in the past 12 months (yes=1) as a potentially important predictor variable.

Multivariable logistic regression was employed to assess the association between the outcome variable and the various covariates. The variables were added in two blocks: in the first block (model 1), we tested the mental health variables and club changes in the past 12 months separately; in the second block (model 2), we added the gambling and substance-related variables. To assess the performance of the regression models, we calculated McFadden’s R-squared. This pseudo R-squared measure provides an indication of the proportion of the log-likelihood explained by the model compared with a null model (a model with no predictors).

Since 36 (8.7%) rows contained missing variables, multiple imputation (20 imputations) with predictive mean matching was performed using the R-package Mice.³⁷ The dataset with imputed values was used to perform sensitivity analyses, comparing the coefficients between the original and imputed datasets to ensure the findings were consistent.

The variance inflation factor (VIF) was calculated for each covariate in both blocks to check for multicollinearity. A VIF of four or higher suggests potential multicollinearity, while a score of 10 or higher indicates that multicollinearity must be addressed. None of the covariates exceeded a VIF of four, so all covariates were retained. Regression model estimates are presented as ORs with 95% CIs and p values. All statistical analyses were performed using R V.4.2.3.³⁸

Results

Participants

All 717 players registered in one of the 32 teams as of February 2021 were invited to participate in the study. A total of 416 players completed the online survey, resulting in a response rate of 58.0%.

Descriptive data

The mean age of participants was 25.5 years (SD=5.3). The majority had completed secondary school (69.6%), and nearly half (49.4%) lived with a partner. Full background information is presented in table 1. Regarding mental health, 17.8% of participants reported depressive symptoms, 13.9% reported anxiety symptoms, 13.3% reported sleep disturbances and 9.6% at-risk alcohol consumption (table 2). One-third of the participants reported being aware that their club had a written gambling policy. When asked about match-fixing, 0.5% indicated they had been contacted within the past year and 3.6% reported being contacted several years ago. Additionally, 11.7% stated they knew someone who had been approached regarding match-fixing.

Table 1. Participant background characteristics.

Variable	Value (N=416)
Age, mean (SD)	25.5 (5.3)
Education, n (%)
Not completed primary school	4 (1.0)
Elementary school	50 (12.1)
Secondary school	289 (69.6)
Folk high school	7 (1.7)
University, higher education	61 (14.7)
Other tertiary education	4 (1.0)
Housing, n (%)
Living alone	113 (27.2)
Living with parents/guardians	80 (19.3)
Living with partner	205 (49.4)
Living with friends	9 (2.2)
Other	8 (1.9)
Changed place of residence due to football, n (%)	273 (65.8)
Accommodation through the club, n (%)	142 (34.3)
Training hours per week, median (IQR)	12 (4)
Clubs in the last 5 years, median (IQR)	2 (1)
Changed clubs in the past 12 months, n (%)	111 (26.9)
Current use of cigarettes, n (%)	4 (1.0)
Current use of snuff, n (%)	122 (29.5)
Aware of a written gambling policy in the club, n (%)	133 (32.0)
Contacted about match fixing the prior year, n (%)	2 (.5)
Contacted about match fixing several years ago, n (%)	14 (3.6)
Knows someone who has ever been contacted about match fixing, n (%)	47 (11.7)

Table 2. Mental health, sleep problems and alcohol consumption among elite footballers.

Variable	Value
Depression
PHQ-9, mean (SD)	2.5 (2.9)
PHQ-9≥ 5, n (%)	74 (17.8)
Anxiety
GAD-7, mean (SD)	2.0 (2.5)
GAD-7 ≥ 5, n (%)	57 (13.9)
Insomnia
ISI, mean (SD)	5.5 (3.6)
ISI≥8, n (%)	54 (13.3)
Alcohol
AUDIT-C≥5, n (%)	39 (9.6)

AUDIT-C, Alcohol Use Disorders Identification Test Consumption; GAD-7, Generalised Anxiety Disorder 7-item scale; ISI, Insomnia Severity Index; PHQ-9, Patient Health Questionnaire-9 .

Main results

In the past year, gambling participation was 60.6%. Among those who gambled, the most common type of gambling was sports betting (72.0%), followed by lotteries (49.6%) and horse betting (31.3%). Problem gambling, indicated by a PGSI score of 8 or more points, was reported by 1.7% of participants, while 4.9% reported at-risk gambling (ie, 3–7 points). Overall, 6.6% of participants reported three or more PGSI points (table 3).

Table 3. Levels of gambling problems and engagement in types of gambling among elite footballers.

Variable	Value
Problem gambling
PGSI, median (IQR)	0 (0)
PGSI=0, n (%)	329 (85.3)
PGSI=1–2, n (%)	33 (8.1)
PGSI=3–7, n (%)	20 (4.9)
PGSI $\ge$ 8, n (%)	7 (1.7)
PGSI≥3, n (%)	27 (6.6)
Past year gambling
No past year gambling, n(%)	160 (39.4)
Past year gambling, n (%)	246 (60.6)
Types of gambling
Sports betting	177 (72.0)
Lotteries	122 (49.6)
Horse betting	77 (31.3)
Casino games	47 (19.1)
Poker	38 (15.4)
Slots	32 (13.0)
Bingo	29 (11.8)
Number games	25 (10.2)
Number of game types, mean (SD)	2.22 (1.44)

PGSI, Problem Gambling Severity Index.

The multivariable logistic regression model indicated that past-year participation in slots (OR=7.46, p=0.01) and bingo gambling (OR=6.08, p=0.02) were the two covariates significantly associated with problem gambling (PGSI≥3). None of the mental health-related variables in model 1 were statistically significant. The model performance, assessed using McFadden’s R-squared, improved from 0.015 in model 1 to 0.147 in model 2, indicating better explanatory power with the addition of the gambling and substance-related predictors. For detailed regression model results, see table 4.

Table 4. Results of the multivariable logistic regression analysis with problem gambling (PGSI≥3 points) as the outcome.

Variable	Model 1				Model 2
	OR	95% CI (lower)	95% CI (upper)	P value	OR	95% CI (lower)	95% CI (upper)	P value
Intercept	0.06	0.03	0.10	<0.001	0.04	0.01	0.14	<0.001
PHQ-9 ≥ 5	0.69	0.19	2.14	0.54	0.74	0.17	2.65	0.65
GAD-7 ≥ 5	2.17	0.67	6.28	0.17	2.48	0.64	8.86	0.17
ISI≥8	1.63	0.48	4.71	0.40	1.36	0.35	4.68	0.64
Changed clubs in the past 12 months	0.90	0.32	2.23	0.84	0.78	0.25	2.16	0.65
AUDIT-C≥5	–	–	–	–	2.38	0.57	8.67	0.21
Current use of snuff	–	–	–	–	1.01	0.35	2.71	0.99
Bingo	–	–	–	–	6.08	1.26	28.34	0.02
Sports betting	–	–	–	–	0.71	0.18	3.10	0.62
Horse betting	–	–	–	–	0.91	0.22	3.31	0.89
Lotteries	–	–	–	–	0.47	0.13	1.63	0.24
Number games	–	–	–	–	1.26	0.20	5.86	0.79
Slots	–	–	–	–	7.46	1.68	35.14	0.01
Casino games	–	–	–	–	1.91	0.41	8.17	0.39
Poker	–	–	–	–	0.69	0.14	2.95	0.64
No gambling	–	–	–	–	1.20	0.28	6.06	0.81
Model performance	Model 1				Model 2
McFadden’s R-squared	0.015				0.147

Model 1 includes initial predictors (PHQ-9, GAD-7, ISI scores, change in club membership). Model 2 adds predictors related to gambling types and substance use.

AUDIT-C, Alcohol Use Disorders Identification Test Consumption; GAD-7, Generalised Anxiety Disorder 7-item scale; ISI, Insomnia Severity Index; PHQ-9, Patient Health Questionnaire-9.

Sensitivity analysis

The logistic regression analysis based on the original dataset and the imputed dataset yielded consistent results, with significant associations found for slots and bingo gambling in both analyses. The imputed dataset showed slightly lower McFadden’s R-squared values, indicating a marginally reduced explanatory power compared with the original dataset, see online supplemental table 1. Overall, the consistency of key findings across both datasets suggests robustness in the results.

Discussion

Gambling is a significant concern among elite athletes. This study examined gambling behaviours and problem gambling among elite footballers, as well as their awareness of relevant policies within their clubs. The results indicate that 6.6% of Swedish elite footballers report problem gambling, which is considerably higher than the general population,²⁷ but lower than two other studies conducted in Sweden and Denmark, which reported a point prevalence of 14% and 16% among elite footballers.^{22 23} The discrepancy between the current study and the two previous studies can potentially be explained by our data being collected during the COVID-19 pandemic, which impacted gambling opportunities.³⁹ Furthermore, our study was conducted 5–6 years after the two previous ones, and given the ongoing debate in the media and among clubs about gambling among footballers, this may have influenced the level of gambling. Other explanations for the observed discrepancy could be the relatively low response rate in the Vinberg et al study and the relatively smaller Swedish sample in the Jensen et al study, which also included a different scale for problem gambling.^{22 23} Interestingly, while sports betting was the most frequently reported type of gambling, it was not a significant predictor of problematic gambling. Instead, participation in slots and bingo—both characterised by high speed and event frequency—was associated with problem gambling, corroborating prior findings^{5 7 8} and underscoring the addictive nature of these gambling types.

Although sports betting was not associated with problem gambling in this study, its prevalence of 72% among past-year gamblers highlights that the ‘gamblification’¹⁵ of sports among athletes, a trend observed elsewhere^{18 19} that is also evident in Sweden. This normalisation of gambling within football can be attributed to a strong betting culture among footballers,⁴⁰ reinforced by gambling company sponsorships that contribute to the normalisation of betting behaviours and, consequently, higher levels of problem gambling among athletes.^{12 13 18} Given this context, professional football leagues should reconsider the appropriateness of sponsorships from gambling companies, and clubs should proactively implement prevention measures within their contexts. For example, football leagues and clubs might prohibit players from participating in certain gambling activities. This not only protects the integrity of the sport but also aims to reduce gambling overall, as many gambling companies offer sports betting, online bingo and various casino games (such as slots) on the same platform. Such restrictions could indirectly reduce slot gambling among football players. While gambling company sponsorship of football is inappropriate, it remains essential for football clubs, their confederations, gambling companies and responsible authorities to collaborate in detecting and counteracting match-fixing to secure the sport’s integrity. Another potential measure within the club context is policy development that raises awareness and fosters discussions about gambling and related problems. However, in the current study, only 32% of footballers were aware of a written policy document on gambling within their clubs, suggesting either a lack of dissemination or the absence of such policies.

It is crucial to consider the role of mental health in gambling problems. In our study, 17.8% of participants reported depressive symptoms, and 13.9% experienced anxiety symptoms. These findings align with prior research; for instance, a study by Gouttebarge et al⁴¹ exploring mental health among elite footballers across five countries found a 25% prevalence of anxiety and/or depression in the Swedish subsample. Our results were lower than those reported in a recent study by Kilic et al, which found a 27% prevalence of depressive and anxiety symptoms among Australian footballers.²⁵ This difference could potentially be explained by contextual factors and the smaller sample size in the Australian study. Another recent study compared levels of depression and anxiety among professional European professional footballers before and during the COVID-19 pandemic.⁴² During the pandemic, 41% of players experienced symptoms of depression, and 47% reported anxiety, which was considerably higher than our findings, despite both studies being conducted during the pandemic.^{25 42} One possible explanation for this discrepancy could be the differences between football leagues and how the pandemic was handled. In Sweden, the leagues’ start was postponed from April to June, after which matches were played without spectators. In contrast, most other European leagues follow a seasonal format from August to May, meaning they either had to pause during the pandemic or were even cancelled. As a result, players in European leagues may have experienced greater uncertainty, particularly regarding when or if football activities and competitions would resume and the potential impact on their future contracts.⁴² Unlike other studies,^{22 25} we did not observe an association between problem gambling and mental health issues among athletes in this study. In our analyses, adding gambling types and substance use variables increased the explanatory value of the regression model considerably.

Policy implications

Given the elevated levels of problem gambling compared with the general population^{18 27} and the low awareness of gambling policies within football clubs, there is an urgent need for preventive measures at multiple levels.⁴³ At a universal level, clubs could develop comprehensive gambling policies accompanied by action plans to ensure their effective dissemination among players and staff. This approach would help normalise discussions about gambling risks and potentially mitigate the development of gambling problems. At the selective level, clubs could implement early detection routines like those proposed for identifying mental health symptoms among elite athletes.⁴² These routines could include screenings, education and discussions about gambling within teams and clubs to identify individuals at risk. At an indicated level, clubs could establish partnerships with external organisations that provide treatment for problem gambling. It is essential that these treatment services are offered confidentially to reduce the stigma associated with gambling problems.⁴⁴ From a legislative perspective, it could be considered to reduce the addictive potential of slot and bingo games. This could be done by several means, for instance by making the games slower, lowering the bet size, limiting bonus offers, reducing the opening hours (ie, not allowing gambling at night) and by having (reasonable) global mandatory loss and time limits.

Limitations

This study has several limitations. First, data were collected during the COVID-19 pandemic, which reduced the availability of betting opportunities. Further, compared with most other European leagues, the Swedish leagues were less affected since the start of the competitive season was postponed from April to June, whereas other European leagues were paused mid-season or even cancelled. This makes our results difficult to compare to previously reported findings. This may also partly explain the lack of an association between sports betting and problem gambling. Moreover, as a cross-sectional study, it limits the ability to infer causality between gambling problems and associated factors. The study also relied on self-reported data, known to have issues related to underestimation on gambling self-reports due to factors such as social desirability.⁴⁵ Finally, although the study aimed to include all players in the top two male Swedish football leagues, the response rate was 58%. While this is a reasonable participation rate, there may be issues with selection bias, and the findings may not be generalisable to all elite football players in Sweden or elsewhere.

Conclusions

This study highlights the high prevalence of problem gambling among elite Swedish football players. We found that participation in slot and bingo gambling predicted problem gambling, while mental health issues did not show a significant association. Although sports betting is prevalent among players, it has not been linked to gambling problems. Nonetheless, sports betting may contribute to the gamblification of football. The role of gambling sponsorships in normalising gambling among athletes requires further investigation. Furthermore, there is a lack of longitudinal studies examining the long-term impact of gambling behaviours and interventions among elite athletes, and the low awareness of gambling policies within clubs indicates a need for research on their dissemination and effectiveness. Overall, this study underscores the importance of comprehensive preventive strategies to address gambling and its associated adverse effects in elite football. Future research should investigate the cocreation and implementation of preventive interventions in football clubs and assess their effectiveness.

The funders had no role in the design, data collection, data analysis, interpretation or reporting of this study.

Data availability statement

Data are available on reasonable request.