The relationship between executive functions, decision-making, and changes in symptoms of gambling disorder in online sports bettors.

Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Sci Rep . 2026 Apr 12;16:12076. doi: 10.1038/s41598-026-48449-8 Search in PMC Search in PubMed View in NLM Catalog Add to search The relationship between executive functions, decision-making, and changes in symptoms of gambling disorder in online sports bettors Theresa Wirkus Theresa Wirkus 1 Faculty of Psychology, Technische Universität Dresden, Dresden, Germany Find articles by Theresa Wirkus 1 , Robert Czernecka Robert Czernecka 1 Faculty of Psychology, Technische Universität Dresden, Dresden, Germany Find articles by Robert Czernecka 1 , Gerhard Bühringer Gerhard Bühringer 1 Faculty of Psychology, Technische Universität Dresden, Dresden, Germany 2 IFT Mental Health Solutions, Munich, Germany Find articles by Gerhard Bühringer 1, 2 , Anja Kräplin Anja Kräplin 3 Department of Psychiatry and Psychotherapy, Technische Universität Dresden, Dresden, Germany Find articles by Anja Kräplin 3, ✉ Author information Article notes Copyright and License information 1 Faculty of Psychology, Technische Universität Dresden, Dresden, Germany 2 IFT Mental Health Solutions, Munich, Germany 3 Department of Psychiatry and Psychotherapy, Technische Universität Dresden, Dresden, Germany ✉ Corresponding author. Received 2025 Jul 17; Accepted 2026 Apr 8; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . PMC Copyright notice PMCID: PMC13076668  PMID: 41968207 Abstract Deficits in executive functions (EFs) and impulsive decision-making are well-established cognitive mechanisms underlying gambling disorder (GD), though there are some inconsistent findings. This study investigated the relationship between changes in GD symptoms in online sports bettors and specific EFs as well as decision-making processes over time. A subsample of online sports bettors ( n = 54, mean age = 33, 96% male) from a larger study underwent an in-person assessment. Initial GD symptom severity was measured online according to DSM-5 criteria and was low to mild in most participants. All other assessments took place an average of 443 days later, during the in-person assessment. We analyzed cross-sectional and longitudinal associations between three facets of EFs (response inhibition, cognitive flexibility, and working memory) and four decision-making facets (delay discounting, probability discounting for gains and losses, and loss aversion) and GD symptom severity and changes over time. Cross-sectional results showed that higher GD symptom severity was associated with poorer working memory and reduced loss aversion. Reduced loss aversion was also associated with greater symptom increase over time. However, these associations were not statistically significant after correction for multiple testing, so they should be interpreted with caution. No other significant relationships were found. These findings provide preliminary evidence that reduced loss aversion is one cognitive factor in disordered gambling among online sports bettors, a population with heterogeneous profiles of EFs and decision-making. Further longitudinal studies are needed to clarify the role of specific cognitive processes in the development and progression of GD to better inform treatment strategies. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-026-48449-8. Keywords: Gambling disorder, Decision-making, Impulsivity, Loss aversion, Executive functions Subject terms: Diseases, Health care, Psychology, Psychology, Risk factors Introduction Gambling disorder (GD) as an example of addictive disorders is amongst others characterized through deficits in specific executive functions (EFs) and in distinct impulsive decision-making processes. From a theoretical perspective, impairments in EFs and alterations in decision-making are commonly embedded in dual-process models of addiction and self-control. These models distinguish between top-down cognitive control processes, such as response inhibition, working memory, and cognitive flexibility, and bottom-up valuation and reward-processing systems that guide behavior based on immediate gains and losses. Insufficient top-down control has been proposed as a transdiagnostic mechanism underlying addictive behaviors, allowing reward-driven impulses to dominate behavior despite long-term negative consequences 1 , 2 . Within this framework, impulsive decision-making processes such as delay discounting, probability discounting, and loss aversion reflect alterations in value-based decision-making rather than executive control per se. For instance, reduced loss aversion, as conceptualized in prospect theory 3 , 4 , implies diminished sensitivity to potential losses relative to gains, which may facilitate continued gambling despite accumulating negative outcomes. Importantly, these altered processes may be especially relevant for the persistence and escalation of gambling-related symptoms over time. In line with these theoretical assumptions, individuals with GD show various impairments in EFs, e.g. in cognitive flexibility 5 – 7 , working memory 8 and most prominently in response inhibition 9 – 11 . They also show impulsive decision-making for instance in delay and probability discounting 12 , 13 and reduced loss aversion 14 . These impairments are not only present in individuals with GD, but also in continuous presentations of different severity below the clinical threshold of a GD diagnosis 12 , 13 . To refer to individuals on the entire continuum of disordered gambling behavior both below and above the clinical threshold of a diagnosis, we will subsequently use the term disordered gambling. Impairments in EFs and decision-making are discussed as neurocognitive mechanisms implied in GD etiology and paradigmatic of an insufficient top-down regulation of behavior by long term goals 1 potentially resulting in adverse social, financial, and health consequences. This is especially interesting since GD, as a behavioral addiction, is not associated with substance induced neurological changes, offering the potential to disentangle whether these impaired processes predispose individuals to addictive behaviors or might be a consequence of them 7 . In the following, we will present the literature individually for EFs and impulsive decision-making. Executive functions in gambling disorder Current research suggests individuals with GD show preserved general EFs, but deficits in specific processes like response inhibition, cognitive flexibility, and working memory 15 . There are some inconsistent findings in cognitive flexibility 16 – 19 , working memory 8 , 19 , 20 and also, even if the most consistent of these aspects, in response inhibition 9 , 21 . Nevertheless, deficits in EFs like the ones mentioned above are largely accepted as core mechanisms underlying GD. The inconsistencies could in part be due to the multitude of used measures (concerning both tests and respective outcomes), which partly also measure different aspects of respective EFs (esp. in working memory 8 or tap a number of complex executive processes in one task (e.g. Intra-extradimensional set shift as an alternative measure for cognitive flexibility also requires visual processing 7 . Furthermore, most studies compare individuals with GD with healthy controls 9 , 17 , 18 and a lot of them use treatment-seeking samples 8 , 9 , thereby neglecting potential different nuances of the continuous association between disordered gambling and EFs. Studies with continuous designs suggest a linear association, where more severely affected individuals exhibit more strongly impaired EFs 5 , 6 , 8 , 22 . Recent studies also favor a dimensional perspective on disordered gambling to reflect its complex, multifactorial nature instead of the oversimplified distinct characterization a binary diagnosis implies 23 . Moreover, most studies examining EFs in GD are cross-sectional and consequently do not allow any inferences on the course of GD symptoms. However, several studies found deficits in EFs to predict treatment outcomes like drop-out 24 , low compliance and relapse 25 , 26 , implying the importance of these deficits over the course of GD. There is however a shortage of longitudinal studies exploring the association of deficits in EFs and GD symptoms over time. Evidence from longitudinal studies on substance-related and addictive behaviors / disorders further stresses the importance of deficits in EFs over time as they have been shown to predict the development of addictive behavior 27 . Impulsive decision-making in gambling disorder When it comes to impulsive decision-making, it is important to take into account whether the potential choices are delayed or uncertain 28 and whether they involve a gain or a loss 29 . There is evidence that individuals with disordered gambling or GD show steeper delay discounting 12 , 30 , 31 , more shallow probability discounting, i.e. overvaluing low probability gains and/or undervaluing high probability losses 13 and less loss aversion 14 , 32 . Like with deficits in EFs, there are some mixed findings concerning loss aversion 33 , 34 , but the meta-analyses mentioned above found robust associations of both delay and probability discounting with GD. These effects seem to be stronger when comparing individuals with a GD diagnosis to healthy controls as opposed to continuous designs 12 , 13 . Studies on impulsive decision-making are rather heterogeneous, differing in the magnitude of the rewards in question, the length of the delays, the number of choices offered, hypothetical and realistic scenarios and the tasks used to measure these processes 13 , 31 . Like with EFs, there is a lack of longitudinal studies examining the role of impulsive decision-making in disordered gambling over time. In a recent study of our lab, exploring the same facets of impulsive decision-making mentioned above, lower probability discounting for losses and lower loss aversion predicted symptom increase in addictive behaviors, including gambling 35 . Moreover, more shallow probability discounting was found to predict treatment efficacy in GD 36 and delayed discounting predicted treatment outcomes in other addictive disorders 12 , thus underlining the importance of these processes for the course, and especially the treatment, of GD. Summary and present study Both, deficits in EFs and impulsive decision-making have been proposed as transdiagnostic processes in addiction 1 , 2 , 7 , 11 , 12 , 31 , further emphasizing their importance. This is why, for a better understanding of disordered gamblers and ultimately appropriate treatment options, delineating the temporal role of these distinct processes over the course of disordered gambling is crucial. This need is even bigger, considering the heterogeneous evidence and missing longitudinal studies in the field. There is a further need for studies with continuous approaches, taking the different nuances within the spectrum of disordered gambling into account. Further, there is a need to examine samples of online sports bettors, which - as a popular example of online gambling - gain more and more relevance 37 – 39 . Guided by dual-process models of addiction and value-based decision-making frameworks, the present study aimed to examine whether each pre-specified executive control process (response inhibition, cognitive flexibility, working memory) and each alteration in valuation-based decision-making (delay discounting, probability discounting for gains, probability discounting for losses, loss aversion) is differentially associated with gambling disorder symptoms and their change over time in online sports bettors. Sports betting represents a form of gambling that is characterized by repeated decisions under uncertainty, delayed and probabilistic outcomes, and a strong illusion of skill-based control. These features place particular demands on both EFs as well as on value-based decision-making processes, including probability weighting and sensitivity to potential losses. Unlike purely chance-based gambling forms, sports betting requires continuous evaluation of changing information (e.g., odds, session dynamics) and repeated adjustment of decisions, which makes it a theoretically relevant context for studying the interplay between executive function processes and decision-making mechanisms in relation to gambling-related symptoms. As a first step, we aimed to validate cross-sectional findings concerning deficits in executive function performance and impulsive decision-making in online sports bettors. Concerning our main aim, we explored these processes in association to the change in GD symptoms over time. The present analyses are based on GD symptoms assessed during an initial online survey and a subsequent in-person assessment approximately one year later. We expected the number of GD criteria fulfilled at the in-person assessment, as well as the change in GD criteria from the initial online survey to the in-person assessment, to be negatively associated with performance in executive function tasks. This means that people with an increase (i.e. an aggravation) of GD symptoms would perform worse on executive function tasks. We also expected the number of GD criteria fulfilled at the in-person assessment, as well as the change in GD criteria from the initial online survey to the in-person assessment, to be associated with more impulsive decision-making. That is to say we expected a positive association with delay discounting and negative associations with probability discounting and loss aversion. Cross-sectional hypotheses correspond to H46-52; hypotheses concerning change in GD symptoms to H81-87 of the preregistration of the in-person study ( https://osf.io/g3nfv ). Importantly, each cognitive process was examined as a theoretically distinct construct, and no overarching inference about “cognitive functioning” as a unified domain was intended. Methods This study is part of the RIGAB project, which includes an initial online survey, a one-year follow-up online survey, and a nested in-person study with a subsample (see https://osf.io/k6c23/ and the study protocol for more details 40 . The analyses presented here are based on data from the initial online survey and the in-person subsample. Hypotheses and analysis plans for each study component were preregistered separately on the Open Science Framework (initial online survey https://osf.io/jbfhe , in-person study https://osf.io/g3nfv) 41 , 42 . Prior to data collection, we also received aggregated anonymous player tracking data as part of the RIGAB project, though these data were not used in the current analyses. Participants and recruitment Sample size rationale Sample size considerations were based on an a priori power analysis conducted for the preregistered in-person study 42 . Using G*Power 3.1 43 , we calculated the required sample size for multiple linear regression assuming a medium effect size (R² = 0.15), a power of 1 − β = 0.80, and a significance level of α = 0.05, with four covariates. This calculation was informed by the effect sizes reported in previous studies on impulsive decision-making that applied the same paradigms to individuals with addictive behaviours 35 . This yielded a target sample size of n = 80 (40 participants per group). Due to substantial difficulties in recruiting participants for in-person testing, this target sample size could not be achieved. Therefore, a secondary power calculation using the same parameters but assuming a single main predictor was conducted to estimate the minimum required sample size. This analysis indicated a minimum sample size of n = 50. The final in-person sample ( n = 54) slightly exceeded this minimum threshold. Recruitment Participant recruitment was conducted in collaboration with Tipico, an international gambling provider, which provided us with anonymized player tracking data for their entire German customer base. Inclusion criteria for our participants were as follows: aged 18 to 55 years, residing near one of the study locations in Germany (Dresden, Leipzig, Chemnitz, Munich, Berlin, Hamburg, Frankfurt am Main, Düsseldorf), having logged into their account within the two months preceding the study, and having held their account for at least six months. The last two locations were added after initial recruitment began, due to low response rates in the other study locations 44 , 45 . To facilitate the inclusion of individuals at risk for GD, we used a classification system based on a machine learning algorithm developed by the provider as part of its responsible gambling strategy. This artificial intelligence (AI) algorithm, which continuously evaluates behavior and communication data, categorized customers into groups with or without indicators of risky gambling. Tipico reports that the algorithm performs well in identifying gambling-related issues 46 . The classification was used solely for sampling purposes. After classification, participants were randomly selected for invitation to the study. Due to data protection constraints, the provider was responsible for contacting selected participants. Participants Originally, the in-person subsample was designed to include two distinct groups: participants with and without GD (please see the preregistration of the initial online survey for the original power analysis https://osf.io/jbfhe ). However, due to low participation among those with GD, the recruitment strategy was adapted. Participants were instead selected across the full spectrum of GD criteria, starting with individuals who met 3–9 criteria and gradually including those with fewer criteria. Recruitment within each symptom level continued until all eligible individuals had been contacted. Because of anticipated low turnout and logistical barriers (e.g., failed contact attempts), participants were not randomized. Only individuals from Dresden, Leipzig, Chemnitz, and Berlin were invited to the in-person study. A flowchart of the participants of the RIGAB study can be found in Fig. 1 . Fig. 1. Open in a new tab Flowchart of the participants of the RIGAB study (without the online follow-up survey, which was not included in this paper). GD = gambling disorder. Compliance with ethical standards All study activities adhered to the principles outlined in the Declaration of Helsinki. The study protocol was approved by the Institutional Review Board (IRB00001473) at TUD Dresden University of Technology (reference number SR-EK-190032021). Before beginning the initial online survey, participants were fully informed about the study and provided written informed consent. Measures Gambling disorder GD criteria were assessed using an internally translated German version of the Diagnostic and Statistical Manual of Mental Disorders, fifths edition (DSM-5)[ 47 ] developed by Stichfield, referring to the last 12 months 48 . The initial online assessment measured symptoms using the German version of the Stinchfield DSM-5 GD criteria. During the in-person assessment, the DSM-5 GD criteria were evaluated using a self-developed GD section of the German WHO Composite International Diagnostic Interview (DIA-X, M-CIDI) 49 , which was based on the same German Stinchfield DSM-5 items. Therefore, both instruments assess the same DSM-5 criteria using identical wording. However, the item structure differs: the Stinchfield online questionnaire uses one item per criterion, whereas the M-CIDI uses one item for three criteria and two items for six criteria, with criterion-specific endorsement rules. The DSM-5 criteria have demonstrated satisfactory reliability, validity, and classification accuracy 50 . While various DSM-5-based screenings have been employed in several epidemiological studies in Germany 51 , there are no German validation studies. Therefore, we computed Cronbach’s alpha based on our dataset and found the internal consistency to be 0.84 during the initial assessment and 0.87 during the subsequent online follow-up. Participants responded to questions covering all nine DSM-5 diagnostic criteria for GD using a binary response format (yes/no). To calculate the change in GD criteria, we used the difference between the number of fulfilled criteria from the initial online survey to the in-person study. Positive values indicate an increase in the number of DSM-5 criteria and thus an aggravation of symptoms, negative values indicate a decrease in the number of DSM-5 criteria and therefore an improvement in symptoms. Zero indicates no change between the two times of measurement. Executive functions The following three tasks were used to assess individual differences in EFs: GoNoGo (response inhibition), Number-Letter (cognitive flexibility) and Two-back (working memory). For a detailed description of the tasks and the respective outcomes see Table 1 . For analyses we used inverse efficiency scores (IES)[ 52 ], where error rates (ERs) and reaction times (RTs) for each task are combined. IES were calculated using the following procedures already established in our lab 53 , 54 : to account for individual differences in the balance of the speed-accuracy trade-off 55 the mean RT of correct responses was divided by the proportion of correct responses (RT/[1 –ER]). Table 1. Executive functions tasks and respective outcomes. (adapted from Wolff et al. 2021). Task (Executive functions) Description Outcome GoNoGo (response inhibition) During each trial, a fixation cross was presented for 750 ms, followed by the display of two dots arranged either vertically or horizontally for 500 ms. Participants were directed to press the response key for “Go” trials (vertically arranged dots), while inhibiting the response on “No-Go” trials (horizontally arranged dots). There were 280 “Go” trials and 40 “No-Go” trials. IES using “go” RT and “no-go” ER Number-Letter (cognitive flexibility) Throughout the task, the screen was partitioned into four quadrants by a horizontal and a vertical line. Each trial presented a digit-letter pair (e.g., 2a or g9) for 3,000 ms in one quadrant. Digits were categorized as even or odd, while letters were classified as vowels or consonants. Participants were instructed to utilize two response keys to indicate whether the digit was even or odd (when the stimulus was displayed above the horizontal line), or to determine whether the letter was a vowel or a consonant (when the stimulus was below the horizontal line). “Switch” trials alternated with “no-switch” trials as the stimulus location was rotated in a clock-wise direction from trial to trial. In total, there were 128 trials. IES difference between “switch” and “no-switch” trials Two-back (working memory) During the task, 8 circles positioned around the centre of the screen were displayed. Initially, for the first 1,500 ms of each trial, all circles remained empty. Subsequently, one of the circles was filled for 500 ms, creating a flashing impression. Participants were directed to press the right key when the flashing circle matched the one that had flashed two trials earlier (“yes” trials), and to press the left key otherwise (“no” trials). There were 40 “yes” trials and 120 “no” trials. IES across all trials Open in a new tab MS = milliseconds, IES = inverse efficiency scores, RT = response time, ER = error rate. Depending on the number of response keys in a task, response keys were “space” (one key) or “Y” and “M” (two keys). Note that “Y” and “Z” keys are transposed on German compared to English-layout keyboards. ERs included only wrong-key errors. RTs for error trials and trials immediately following wrong-key errors were excluded. RTs below 100 ms and RTs deviating from the median by more than 3.32 median absolute deviations were also excluded 56 . Please note that individuals with lower EFs are assumed to display higher IES. Impulsive decision-making To assess impulsive decision-making, we employed four tasks developed by Pooseh et al. (MATLAB scripts accessible from https://github.com/spooseh/VBDM) . 57 These tasks encompassed: (1) a delay discounting task with delays of 3, 7, 14, 31, 61, 180, and 365 days between choice options; probability discounting tasks for (2) gains and (3) losses, involving five probability values (2/3, 1/2, 1/3, 1/4, and 1/5). Each task comprised 30 trials, with monetary gains/losses ranging from 0.30 to 10 €. In the (4) mixed gambles task, participants began with 10 € and played 40 trials with gains ranging from 1 to 40 € and losses from 5 to 20 €. Scenarios were hypothetical, there was no extra compensation for these tasks. In all four tasks, participants were required to choose between two given options, which were simultaneously presented on a computer screen, using the Psychophysics Toolbox 58 in MATLAB R2010a 59 . A Bayesian adaptive algorithm was implemented, updating parameter estimation after each trial, and utilizing it for calculating options in the subsequent trial. This approach focused on providing the most informative offers near the individual’s indifference point (the point of indifference between two choice alternatives), enabling efficient inference of decision-making parameters without post-hoc estimations. For the delay and probability discounting tasks, a hyperbolic value function 60 was utilized, reflecting that subjective values of delayed (or probabilistic) rewards decrease hyperbolically according to the discounting rate ( k ). In the mixed gambles task, a simple linear function was used, in which loss aversion ( λ ) represents the relative weighting of losses to gains in the participant’s decision 61 . Individuals with higher impulsive decision-making are assumed to display higher k values in the delay discounting task, lower k values in probability discounting tasks, and lower λ values in the mixed gambles task. Procedure The RIGAB (“Characteristics and prediction of RIsky GAmbling Behavior”) study is a longitudinal research project composed of three phases 44 , 62 . The initial online survey was conducted at baseline. Approximately 12 months later, participants who had consented to being followed up were invited to complete a second online survey. In a nested subsample, an in-person assessment took place on average 443 days (SD = 116) after the initial online survey, with variability in the interval due to logistical factors. Although the broader RIGAB study includes a 12-month online follow-up, the present analyses focus exclusively on data from the initial online survey and the in-person subsample (Fig. 1 ). Fig. 2. Open in a new tab Study design of the RIGAB study 63 . Grey boxes: These parts of the study were not analysed in this paper. GD = gambling disorder assessed with DSM-5 Stinchfield screening questionnaire [ 48 ]. Initial online survey Data collection for both online surveys was conducted using REDCap 64 , 65 , a secure, web-based platform hosted by TU Dresden. After providing informed consent, participants completed the assigned questionnaires. A full list of the instruments used is available at https://osf.io/ac8gj 66 . The questionnaires covered sociodemographic, personality, and gambling-related characteristics. During the assessment, participants provided information on the following covariates: gender (female, male or other), age, and level of education. Data on race and ethnicity were not collected. The survey sequence was standardized for all participants. The initial online survey took approximately 30 min to complete. Participants received a €30 Amazon voucher as compensation. After finishing the initial survey, individuals could choose to share their contact information for possible inclusion in future study phases. In-person study The in-person study took place at TUD Dresden University of Technology and the collaborating Psychologische Hochschule Berlin. During this session, participants undertook a battery of executive function and impulsive decision-making tasks (described above), a structured clinical interview using the Composite International Diagnostic Interview (DIA-X / M-CIDI) 67 , and an interview regarding specific responsible gambling strategies via REDCap. Interview data are not included in the present analysis. The full procedure lasted around two hours and was administered in a fixed order by trained psychologists. Participants were compensated with a €50 Amazon voucher. Those traveling from Leipzig and Chemnitz also received reimbursement for transportation expenses. Statistical analysis All statistical analyses were performed using STATA version 14.2 software 68 . Initially, we sought to examine the cross-sectional relationships between the number of GD criteria fulfilled, according to the DSM-5, and each pre-specified parameter of EFs and impulsive decision-making. To do so, we ran separate multiple linear regression models, using the IES from each executive function task and the k/λ parameters from each impulsive decision-making task as dependent variables and the number of GD criteria met as the independent variable. Next, we investigated whether changes in GD criteria between the initial online survey and the subsequent in-person assessment (independent variable) were linked to the respective executive function and impulsive decision-making parameters (dependent variables). Change in gambling-related symptoms was operationalized as the difference between symptom counts at baseline and follow-up. Given variability in the follow-up interval across participants, the number of days between assessments was included as a covariate in all longitudinal regression models. As mentioned above we are using IES, where higher IES values indicate impaired EFs. This means that due to the nature of the outcome we expect a positive association of (change in) GD criteria and IES. For example, someone with an increase in GD criteria from the initial online survey to the in-person study is expected to perform worse on executive function tasks, which means to have a higher IES. Our preregistered hypotheses concerning this were as follows: We expect the number of GD criteria at the in-person study / the change in GD criteria from the initial online survey to the in-person study to be …. … positively associated with IES at the GoNoGo task (H46, H81), the Number-Letter task (H47, H82) and the 2-back task (H48, H83). … positively associated with delay discounting (H49, H84) and negatively associated with probability discounting for gains (H50, H85), losses (H51, H86) and with loss aversion (H52, H87). Each regression model corresponded to a pre-registered, theoretically distinct hypothesis referring to a specific cognitive parameter. No composite cognitive score or overarching domain-level inference was tested and no joint hypothesis across cognitive domains were evaluated. Covariates included in all analyses were age, gender, educational level, and the age of the participant’s betting account, used as a proxy measure for exposure to online gambling, based on prior literature 69 – 71 . Additional covariates for the analyses examining criteria changes comprised the number of days elapsed between assessments and the baseline number of GD criteria, to account for regression to the mean effects. Predictor variables were entered in their original metric and were not standardized. To ensure robustness of the regression analyses, models were estimated using robust standard errors. The robust regression models yielded no meaningful differences, so only the results from the multiple linear regressions are reported. Although hypotheses were directional and preregistered, all p-values reported are based on two-tailed tests, reflecting a conservative testing approach. Given that each model tested a pre-registered and theoretically distinct hypothesis pertaining to a specific cognitive parameter, no correction for multiple testing was applied in the primary analyses. However, as a sensitivity analysis, we additionally controlled the false discovery rate across the seven cognitive outcomes using the Benjamini–Hochberg procedure within each analysis set (cross-sectional and longitudinal models). To examine whether comorbid substance use disorders influenced the observed associations, a further sensitivity analyses were conducted excluding participants with comorbid alcohol or tobacco use disorders ( n = 13). All regression models were re-estimated in this reduced sample using the same model specifications as in the primary analyses. These sensitivity analyses were not preregistered and are reported for robustness purposes. Results Descriptive characteristics of the sample from the in-person study can be found in Table 2 . Participants were predominantly male (96%) with an average age of 33 years. The majority had a high level of education ( n = 38 (70%)) and on average the age of their betting account was 4.70 (SD = 2.86) years. The sample was largely homogeneous with respect to cultural background. All participants reported German citizenship ( n = 54, 100%), and the majority were born in Germany ( n = 53, 98%) and indicated German as their native language ( n = 50, 93%). Two participants (4%) reported an additional second citizenship (Iran and Finland). Parental citizenship showed a similar pattern. In most cases, both parents held German citizenship ( n = 46, 85%). In a small number of cases, one parent held German citizenship while the other parent had a different citizenship, including Turkey ( n = 1), Poland ( n = 1), Finland ( n = 1), and Russia ( n = 1). Additionally, there were isolated cases in which both parents held the same non-German citizenship (Iran, n = 1; Vietnam, n = 1). Overall, these data indicate that the sample was predominantly composed of participants with German citizenship and largely German parental background, with only limited representation of more diverse migration-related characteristics. Table 2. Descriptive characterization of the in-person sample. M (SD) / Absolute frequencies (%) Range Age 33 (10.04) 20–55 Gender (male) n = 52 (96%) Education Middle n = 16 (30%) High n = 38 (70%) Sum of DSM-5 Stinchfield criteria at initial online survey 2.22 (2.18) 0–9 Sum of DSM-5 Stinchfield criteria at in-person study 2.78 (2.30) 0–8 Number of days between initial online survey and in-person study 443.41 (116.31) 234–699 Age of betting account 4.70 years (2.86) 0.93–12.81 German as a native language n = 50 (93%) German citizenship n = 54 (100%) Most common family status n = 42 single (78%) Most common employment status n = 32 fully employed (59%) Open in a new tab N total sample = 54. M = Mean, SD = Standard deviation. At the initial online survey, the average number of fulfilled DSM-5 criteria for GD was 2.2 (2.18), at the in-person study it was 2.78 (2.3). The number of fulfilled GD criteria changed between − 4 and + 6 between the initial online assessment and the in-person assessment (see Table S3 in the supplemental material). Eighteen participants (33.3%) showed no change in the number of criteria fulfilled. Eleven participants (20.4%) showed a decrease, while 25 (46.3%) showed an increase. Overall, 63.0% of participants exhibited changes of between − 1 and + 1 criteria. For descriptive data on the initial and the follow-up online survey respectively, please see our other publications 23 , 45 . Descriptive statistics of the respective executive function and impulsive decision-making tasks can be found in Table 3 . The correlations between the cognitive outcome parameters were generally small (see Table S4 in the supplementary material), which indicates that the measures were largely independent of each other. Table 3. Descriptive statistics and reliabilities of the executive function and impulsive decision-making tasks. Outcome M SD Executive functions (IES) GoNoGo 351.71 a 59.74 Number-Letter 382.11 b 201.92 Two-back 555.98 c 148.47 Impulsive decision-making Delay discounting (log k ) -5.21 2.89 Probability discounting for losses (log k ) − 0.32 2.58 Probability discounting for wins (log k ) 0.01 0.84 Loss aversion (log λ ) 0.24 0.54 Open in a new tab If not otherwise indicated N = 54. M = Mean, SD = Standard Deviation, IES = Inverse Efficiency Scores. a n = 52, 2 persons had to be excluded due to exclusion procedures explained above. b n = 47, 2 persons could not complete this task due to technical difficulties, 5 persons had to be excluded due to exclusion procedures explained above. c n = 51, 3 persons had to be excluded due to exclusion procedures explained above. Cross-sectional associations of executive functions, decision-making and gambling disorder criteria We found no evidence of a significant association between the number of fulfilled GD criteria at the in-person study and performance in executive function tasks, except for the two-back task (working memory). Participants with more GD criteria showed higher IES on the two-back task ( b = 23.98, p = .027 [2.82, 45.15]). This association did not remain statistically significant after false discovery rate correction (q = 0.096). Concerning impulsive decision-making we found a significant association between the number of GD criteria at the in-person study and loss aversion; participants with more GD criteria were less loss averse and vice versa ( b = − 0.09, p = .024, [-0.16, − 0.01]). This association did not remain statistically significant after false discovery rate correction (q = 0.096). All other associations were not significant. Results of the multiple regressions for the association of GD criteria at the in-person study and EFs and impulsive decision-making can be found in Table 4 . Table 4. Results of the multiple linear regression analyses for cross-sectional association of executive functions, impulsive decision-making and gambling disorder criteria at the in-person study. Outcome Regression coefficient p q (FDR) 95% CI Executive functions (IES) GoNoGo − 0.06 a 0.988 0.989 [– 8.26, 8.14] Number letter 12.40 b 0.431 0.754 [– 19.11, 43.91] Two-back 23.98 c 0.027 0.096 [2.82, 45.15] Impulsive decision-making Delay discounting (log k ) -0.003 0.989 0.989 [– 0.39, 0.39] Probability discounting for losses (log k ) 0.19 0.254 0.592 [– 0.14, 0.52] Probability discounting for wins (log k ) − 0.03 0.649 0.909 [– 0.14, 0.09] Loss aversion (log λ ) − 0.09 0.024 0.096 [– 0.16, – 0 0.01] Open in a new tab If not otherwise indicated N = 54. CI = 95% Confidence Interval, IES = Inverse Efficiency Scores. All analyses controlled for the influence of age, gender, education and age of account. All p values are two-tailed. The q values represent false discovery rate (FDR)–adjusted p values (Benjamini–Hochberg procedure) across the seven cognitive outcomes. Results are generally reported using two decimal places, with three decimal places used for p and q values or when rounding to two decimals would result in a value of zero. a n = 52, 2 persons had to be excluded due to exclusion procedures explained above. b n = 47, 2 persons could not complete this task due to technical difficulties, 5 persons had to be excluded due to exclusion procedures explained above. c n = 51, 3 persons had to be excluded due to exclusion procedures explained above. Sensitivity analyses excluding participants with comorbid alcohol or tobacco use disorders ( n = 13) revealed that the association between loss aversion and the number of GD criteria remained nominally significant, whereas the association with working memory did not (see Table S4 in the supplement). Associations of executive functions, decision-making and change in GD criteria over time In contrast to our hypotheses, only the association between loss aversion and the change in GD criteria from initial online survey to in-person study was significant. Participants with a positive change in GD criteria and thus an aggravation of symptoms were less loss averse and vice versa ( b = − 0.09, p = .045, [-0.17, -0.002]). This association did not remain statistically significant after false discovery rate correction (q = 0.314). All other associations between the change in GD criteria and EFs and impulsive decision-making respectively were not significant. Results of the multiple regressions for the association of change in GD criteria from the initial online survey to the in-person study and EFs and impulsive decision-making can be found in Table 5 . Sensitivity analyses that excluded participants with comorbid alcohol or tobacco use disorders ( n = 13) corroborated the main finding. The longitudinal association between loss aversion and changes in gambling-related symptoms remained nominally significant (see Table S5 in the supplement). Table 5. Results of the multiple linear regression analyses for the association of executive functions, impulsive decision-making and the change in gambling disorder criteria from the initial online survey to the in-person study. Outcome Regression coefficient p q (FDR) 95% CI Executive functions (IES) GoNoGo − .65 a 0.895 0.948 [-10.52, 9.21] Number Letter -1.22 b 0.948 0.948 [-38.82, 36.37] Two-back 16.56 c 0.173 0.453 [-7.58, 40.69] Impulsive decision-making Delay discounting (log k) 0.16 0.473 0.828 [-0.29, 0.61] Probability discounting for losses (log k) 0.25 0.194 0.453 [-0.13, 0.64] Probability discounting for wins (log k) 0.03 0.633 0.886 [-0.10, 0.16] Loss aversion (log λ) − 0.09 0.045 0.314 [-0.17, − 0.002] Open in a new tab If not otherwise indicated N = 54. CI = 95% Confidence Interval. All analyses controlled for the influence of age, gender, education, age of account, days between initial online survey and in-person study and GD criteria at initial online survey. All p values are two-tailed. The q values represent false discovery rate (FDR)–adjusted p values (Benjamini–Hochberg procedure) across the seven cognitive outcomes. Results are generally reported using two decimal places, with three decimal places used for p and q values or when rounding to two decimals would result in a value of zero. a n = 53, 1 person had to be excluded due to exclusion procedures explained above. b n = 47, 2 persons could not complete this task due to technical difficulties, 5 persons had to be excluded due to exclusion procedures explained above. c n = 51, 3 persons had to be excluded due to exclusion procedures explained above. Discussion Our main aim for the present study was to explore the association between EFs, decision-making and the change in GD symptoms over the past year in a sample of online sports bettors; thereby adding to the literature on the temporal role of specific deficits in these processes in relation to changes in GD symptoms. Importantly, the present study focuses on online sports bettors with the majority of participants not meeting the diagnostic threshold for GD. Consistent with dimensional models of GD symptoms 73 , disordered gambling is therefore treated as a continuous construct rather than a categorical diagnosis. In the first step, a cross-sectional validation, we found nominal associations indicating that more disordered gambling over the past year was associated with more deficits in working memory and reduced loss aversion, but no evidence for deficits in other executive function or decision-making processes. Furthermore, we found nominal associations between loss aversion and more disordered gambling over the past year, but no evidence for an association with other impulsive decision-making processes. However, these associations were no longer statistically significant after controlling for multiple testing, so they should be interpreted with caution. For the first time our study explored these associations in relation to the change of GD symptoms over time at an average of 443 days later. Our results showed a nominal association between reduced loss aversion and an increase in GD symptoms over time. This association was not statistically significant after correction for multiple testing, so it should be considered preliminary. No evidence was available that other processes were associated with the change of GD symptoms over time. Thus, while the pattern of results may suggest a potential role for reduced loss aversion in symptom dynamics, the findings do not provide robust statistical support for a specific cognitive mechanism underlying changes in GD symptoms. These new findings tentatively underline the relevance of reduced loss aversion in relation to changes in gambling disorder symptoms over time and further stress the need for longitudinal studies on EF and decision-making processes in individuals with disordered gambling. From a theoretical standpoint, the present findings may point to alterations in value-based decision-making, rather than broad deficits in executive control, as being particularly relevant for changes in gambling-related symptoms over time. However, as none of the observed associations remained significant after correction for multiple testing, any interpretation at the domain level should be treated with caution. Loss aversion is a core construct of prospect theory and reflects the extent to which potential losses are weighted more strongly than equivalent gains during decision-making 3 , 4 . Reduced loss aversion implies a diminished sensitivity to potential losses, which may facilitate the continuation of gambling behavior despite repeated negative outcomes. Importantly, this valuation-related mechanism is conceptually distinct from EFs and was examined as a separate, pre-specified hypothesis rather than as part of a global cognitive construct. We will discuss our findings individually for each construct below. Executive functions Contrary to our hypotheses, we did not find evidence for associations between deficits in EFs and GD symptom counts assessed over the past year, neither cross-sectionally nor with change in symptoms over time, with the exception of working memory (cross-sectionally). The cross-sectional association between working memory and GD symptoms was nominal and did not survive correction for multiple testing. This is in line with other studies finding only isolated deficits in EFs 17 , 19 . Overall, the current findings do not suggest that there is a general pattern of impaired EFs among online sports bettors. The cross-sectional association between working memory performance and disordered gambling did also not persist after excluding participants with comorbidities. This pattern suggests that working memory deficits observed in the full sample may, at least in part, reflect comorbidity-related cognitive effects rather than gambling-related processes per se. Whether these differences in working memory performance precede gambling-related problems, needs to be assessed in future longitudinal studies with adolescents. The otherwise absent evidence for deficits in EFs might be due to the fact that strategic gambling activities like sports betting demand cognitive processes and individuals preferring these might thus not present any or less deficits in EFs as opposed to non-strategic gamblers 74 . For instance, working memory and mental flexibility are necessary to calculate and compare odds and probable payouts; while declarative memory is needed to apply past knowledge and experiences (e.g. performance track records, absence of key players) to assess the probability of wins 18 . Whether these individuals’ underlying neurobiology supports efficient EFs, fostering a preference for strategic gambling, or if regular use of these EFs during betting has contributed in preserving cognitive processes is unclear 18 , 75 . Our findings add to existing inconsistencies concerning the facets of EFs explored here and are in line with a number of studies reporting no significant differences between individuals with disordered gambling and healthy controls 5 , 8 , 16 – 21 , 75 , 76 . Within dual-process frameworks, this pattern may indicate relatively preserved top-down control in this sample. However, the absent evidence for lower inhibitory control is contradictory to a meta-analysis reporting robust associations 9 . Our results might also be partially explained by sample and measurement limitations. We examined a relatively small sample, where the majority of participants displayed a low to modest number of GD symptoms below the cut-off of a diagnosis (76% ≤ 3 symptoms at initial online survey, 65% at in-person study, see Tables S1, S2 in the supplement). The association of deficits in EFs and disordered gambling severity 5 , 26 and studies finding differences between individuals with GD and healthy controls but not subclinical levels of disordered gambling 6 suggest a linear association of GD symptoms and deficits in EFs. Consequently, impairments in EFs in our sample might not have been that severe, which in turn might have made it hard to detect an association. The same might have been the case for the change in GD symptoms over time, where most participants exhibited only a minor change in symptoms (no difference for one third, 63% between − 1 and + 1). This is however in line with studies on the course of GD, suggesting that most individuals ”at-risk” of and with GD move toward a lower (or less intensive) level of gambling behavior over time, while those who gamble recreationally are unlikely to move to a more severe level 77 , 78 . In combination with the modest sample size and the correction for multiple testing applied in the present study, these characteristics may have further reduced statistical power to detect small effects. Concerning the measurement limitations, performance variability was relatively high for some executive function tasks. However, this pattern is consistent with previous studies using the same paradigm 27 and likely reflects task-inherent characteristics. Nevertheless, high variability and task-specific exclusions may have reduced sensitivity to detect associations. Furthermore, IESs were used to combine reaction time and accuracy into a single performance index. This approach assumes a relatively consistent speed–accuracy trade-off across individuals, which may not hold for all participants, particularly in gambling-related populations characterized by heterogeneity in response strategies. As a result, the reliability and sensitivity of the executive function tasks to detect associations with disordered gambling may have been reduced, and null findings for this measure should be interpreted with caution. In addition, the already mentioned sizeable heterogeneity in almost all aspects of methodology in studies concerned with EFs in gambling 15 , further hampers comparisons across studies. For instance, conclusions based on treatment-seeking populations might skew the picture as they are inferred from individuals with a potentially longer duration and higher severity of GD and assumed associated deficits in EFs 27 . However, the heterogeneity of findings also strengthens the argument that GD is a disorder characterized by substantial heterogeneity, for instance also depicted in efforts to subtype its affected individuals (e.g. pathways model of problem and pathological gambling) 79 , 80 . Potentially this includes a subtype where only highly affected individuals show deficits in EFs, but mildly affected individuals, who may experience difficulties in other areas due to gambling, do not (cf. considerations concerning quasi- or semi-continuous relationships for delay discounting as proposed) 12 . Conceivably, this could interact with other factors like the preferred form of gambling. Future studies should explore the association of EFs and the course of disordered gambling with a sufficiently distributed variation of severity in longitudinal designs, ideally with larger sample sizes including strategic and non-strategic gamblers. To facilitate comparisons across studies a comprehensive and unitary network of EFs in disordered gambling is needed. Nevertheless, assuming online sports bettors do have preserved and robust EFs also means individuals with disordered gambling are well capable of participating in and benefiting from cognitive demanding treatment like Cognitive Behavioral Therapy 18 . However, this implication should be interpreted with caution, given the limitations of the present study mentioned above. Impulsive decision-making Reduced loss aversion showed nominal associations with disordered gambling, cross-sectionally and with the change of disordered gambling over time. However, these associations did not withstand correction for multiple testing and therefore cannot be considered statistically robust. These associations were also observed in additional sensitivity analyses excluding participants with comorbid alcohol or tobacco use disorders, suggesting that this effect is not driven by comorbid substance use. This finding is in line with other cross-sectional studies showing that individuals with GD exhibit less loss aversion than healthy controls 33 , 34 , 81 . In a study with adolescents with internet gaming disorder the majority of participants showed reduced loss aversion, but no evidence for impaired response inhibition 82 . As only severely affected individuals also showed impairments in response inhibition the authors concluded that mildly affected individuals mainly differ in their reward-seeking system but not in EFs. They posit reduced loss aversion as an important factor for initiation and development of this behavioral addiction similar to GD, which is linked to underestimating long-term negative outcomes due to the disorder. This might further speak for the heterogeneity of cognitive function in disordered gamblers with specific but not general deficits and consequently also domains entirely unaffected. Our findings suggest that reduced loss aversion may be one of several cognitive processes associated with worsening GD symptoms over time in sports bettors. Reduced loss aversion may represent a cognitive characteristic linked to the persistence of gambling behavior despite negative consequences, a feature commonly observed in more severe forms of disordered gambling. This association may help to explain individual differences in the extent to which negative outcomes are weighed against potential gains. This could in turn be linked to motivational processes in treatment. Accordingly, interventions that focus on strengthening personally relevant values and goals outside of gambling, for example through motivational interviewing 83 , may be particularly useful in supporting motivation for change. However, as the longitudinal association was not statistically significant after correction for multiple testing, these implications should be interpreted with caution. Despite the longitudinal nature of the study, only two time points were available, which precludes conclusions about the direction of causality. The present findings therefore suggest a possible association between reduced loss aversion and changes in gambling disorder symptoms, but cannot determine whether reduced loss aversion precedes symptom worsening or emerges as a consequence of increasing gambling-related problems. Furthermore, while the observed associations were statistically significant, they did not withstand correction for multiple testing, and their magnitude was small. They should therefore be interpreted with appropriate caution. Reduced loss aversion is likely to be one of several contributing factors, rather than a strong standalone indicator, of symptom change. Future studies should employ cross-lagged panel designs with larger samples and with three or more measurement occasions to disentangle the temporal and potentially causal relationships between EFs, decision-making processes such as loss aversion, and gambling disorder symptoms. In contrast to our hypotheses and a vast body of literature 12 , 13 , we did not find any evidence for associations between delay or probability discounting and disordered gambling, both cross-sectionally and concerning the change in symptoms over time. This might be partly attributable to the already mentioned heterogeneity in studies concerning amongst others measures, samples and sample sizes. Akin to deficits in EFs, the low levels of disordered gambling in our sample might have made it difficult to detect impulsive decision-making, as they might have entailed that putative impairments in these processes are not as severe. Although counter-intuitive as continuous designs supposedly have higher power to detect associations, other studies also found larger effect sizes for studies with categorical designs 12 , 13 . Furthermore, as is the case with EFs, some authors propose there might be differences between strategic and non-strategic gamblers in decision-making 74 , 81 , which could explain the difference to other studies. Limitations Several limitations should be considered when interpreting the present findings concerning the design, the measures, and the sample. Concerning the design, change over time was assessed using a simple difference score across two time points with varying follow-up intervals. Although the interval was statistically controlled, individual trajectories were not modeled, so observed effects reflect differences between non-uniform time points. Another limitation is the absence of important contextual control variables, such as economic status or family background, to heighten ecological validity. While the focus of our longitudinal analyses on within-person change scores reduces the influence of stable between-person differences, we cannot rule out the impact of time-varying, unmeasured factors on symptom changes (e.g., acute life events, major wins or losses, temporary substance use, or short-term treatment exposure). Consequently, observed associations between cognitive processes and symptom change should be interpreted as indicative of potential mechanisms. Concerning the measurement limitations, GD symptoms were assessed using the number of DSM-5 criteria met, which reflects the past 12 months rather than current severity. In contrast, EFs and decision-making were assessed cross-sectionally at the time of the in-person assessment so that the observed associations are correlational and not concurrent or causal. Moreover, DSM-5 symptom counts provide a diagnostic framework rather than a fine-grained measure of severity. More sensitive dimensional instruments such as the Gambling Symptom Assessment Scale (G-SAS) 84 may be better suited to capture variability in gambling symptom severity, particularly in non-clinical samples. A further limitation concerns the assessment of GD criteria across assessments. Although all measures were based on identical DSM-5 criteria and wording, symptom counts were derived from different instruments and assessment modes (online self-report vs. in-person interview). Differences in item structure between the Stinchfield screening and the M-CIDI may influence symptom endorsement patterns and thus affect the interpretation of symptom change. Although our executive function and decision-making paradigms are well established and in some cases even said to be advantageous over other measures (e.g. Two-back task as a comparatively holistic assessment of working memory) 8 , the kind of behavioral measures used within this study have been criticized for their lack of ecological validity 18 . This concerns for instance the use of generic instead of complex cues or processes in milliseconds instead of more realistic larger time frames 9 . In the case of impulsive decision-making, this might imply that these measures are not suited to detect or activate modest deficiencies in a population well-accustomed to winning/losing real money 18 . Our sample mainly consists of individuals with a low number of gambling-related symptoms, which limit the estimation of the association between cognitive deficits and moderate-to-severe gambling disorder. Therefore, the present results should be validated in clinical samples of individuals meeting diagnostic criteria for GD. In addition, the sampling strategy oversampled individuals with elevated gambling-related symptoms, which may have contributed to regression-to-the-mean effects, even though baseline symptom levels were controlled for in the analyses. In addition, participants were recruited via a proprietary risk-classification algorithm provided by a gambling provider. Although this algorithm was used exclusively for sampling and did not inform any study measures or analyses, it might have resulted in a systematic preselection of individuals whose gambling behavior aligns with the operator’s internal risk thresholds. Consequently, the findings may not generalize to all sports bettors or to individuals with gambling-related problems identified through alternative screening or clinical pathways. Furthermore, our sample showed a substantial gender imbalance, which limits the generalizability of the findings. Although gender was included as a covariate in all regression models, the sample size and distribution did not allow for the examination of potential gender-specific associations. Given evidence that risk factors for GD may differ by gender 85 , future studies with more balanced samples are needed to investigate whether the observed associations vary across genders. Our sample also lacked direct measures of race or ethnicity, reflecting common research practice in Germany due to historical, ethical, and methodological considerations. While we used proxy indicators such as citizenship, country of birth, native language, and parental citizenship to describe the sample, these do not capture processes of racialization or allow conclusions regarding differences related to race or ethnicity. Future research should consider context-sensitive approaches to better assess diversity and inequality in German samples. Finally, the modest sample size limits statistical power, particularly for detecting small effects, and may have contributed to null findings for some executive function and decision-making measures. Conclusion and future research To the best of our knowledge our study was the first to explore associations between EFs, decision-making, and changes in gambling disorder symptoms over time in online sports bettors. We observed associations between reduced loss aversion and an exacerbation of disordered gambling symptoms. As this association was not statistically robust after correction for multiple testing, it should not be interpreted as evidence of a reliable predictor. Rather, it should be considered a potential variable of interest for future studies requiring replication. The present findings are based on a sample of individuals who predominantly engaged in sports betting, which represents only one specific form of gambling. Other gambling forms differ in their structural characteristics, cognitive demands, and reinforcement schedules. Future studies should therefore test and compare the present hypotheses across different gambling activities to determine whether the observed associations between executive functions, value-based decision-making, and changes in disordered gambling generalize beyond sports betting or are specific to certain gambling forms. To test these hypotheses, future research should examine large samples with varied severity and ideally incorporate not exclusively treatment-seeking, strategic and non-strategic gamblers and assess long follow-up periods. Our findings further support that individuals with disordered gambling are characterized by substantial heterogeneity with highly individual presentations, where clustering efforts for instance by preferred gambling type might facilitate a more comprehensive understanding of subgroups. While the present results do not provide statistically robust evidence for specific cognitive mechanisms, they are consistent with the broader notion of heterogeneity in cognitive profiles among individuals with gambling-related problems. As (online) gambling becomes increasingly widespread and easily accessible, comprehending the cognitive characteristics of specific subgroups of gamblers with increasing relevance like online sports bettors remains crucial to the implementation of effective treatment strategies. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 (25.2KB, docx) Acknowledgements Open Access publication was funded by the project DEAL.The web-based software platform REDCap, used within this study for data collection, was funded through grants for unrelated projects at TUD Dresden University of Technology (DFG SFB/TRR 205).We thank Dr Juliane Hilde Fröhner for support during implementation and analyses of the executive functions battery. We thank Dr Max Wolff, Dr Shakoor Pooseh and Prof Michael Smolka for the provision of the task batteries and support during implementation. Chat GPT version 3.5 was selectively used for paraphrasing purposes (prompt: “rewrite”, input: specific direct quotes from references identified by the author) in the methods section and for paraphrasing purposes of selective quotes in the discussion on executive functions in strategic gamblers only (prompt and input as above). Outputs were adapted for language / style / expressions. DeepL Write was used to support language editing and improve clarity of the manuscript. Author contributions TW wrote the manuscript, preregistered all study parts and conducted the RIGAB study under supervision of GB, RC and AK. TW analyzed and interpreted the data under supervision of AK. RC and AK conceptualized the study including data and measures to be assessed. AK revised the manuscript. All authors critically contributed to earlier drafts of the manuscript and approved the final version of the paper. All authors had full access to all data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Funding Open Access funding enabled and organized by Projekt DEAL. This research was supported by an unrestricted research donation to the TUD Dresden University of Technology (donation that does not assign purpose of use) provided by Tipico. Tipico provided the initial anonymized player tracking data and invited participants for the initial online survey, who were randomly preselected by us. Per contract, Tipico had and has no influence on the design, data analyses, data interpretation and publication of the RIGAB study. Tipico did not review or approve the manuscript. Partial funding for the manuscript writing was provided by the DFG as part of the TRR 265 (project number 402170461). Data availability The datasets presented in this article are not publicly available due to their inclusion of sensitive information, such as player tracking data. Since the raw data could potentially allow for the identification of individuals, further anonymization, such as data aggregation, is required before any public release. We are currently preparing the data in accordance with FAIR principles and applying the necessary anonymization procedures. This process is expected to be completed by autumn 2026. Afterward, the anonymized dataset will be made publicly available. Sensitive data will be accessible only in aggregated form or through controlled access within an open repository for qualified research purposes. Until then, data may be made available upon reasonable request to the corresponding author. Declarations Competing interests TW was funded by an unrestricted research donation to TUD Dresden University of Technology provided by Tipico. In the last 5 years RC was partly funded by an unrestricted research grant of members of the ‘Düsseldorfer Kreis’ (a group of key stakeholders from public and private gambling providers, research, and the support system) to Technische Universität Dresden until January 2022. From January 2022 until March 2023, he was partly funded by an unrestricted research grant from the Federal Ministry for Economic Affairs and Climate Action in the context of the evaluation of gambling hall regulations and from March 2023 to March 2024 by residual funds of TU Dresden. Additionally, since 2016 he is partly and since March 2024 fully funded by the AntiSkid program and is currently employed by Mental Care AntiSkid GmbH, Frankfurt, Germany to coordinate the AntiSkid treatment programme for airline pilots with mental disorders. The program is funded by German airlines including Lufthansa AG and her affiliated companies. He is a Ph.D. student at TU Dresden. GB received unrestricted grants for gambling research activities from various public and commercial gambling providers and public regulatory agencies. He is a member of the ‘Düsseldorfer Kreis’ (a group of key stakeholders from public and private gambling providers, research, and the support system) and was also partly funded by an unrestricted research grant from the Federal Ministry of Economic Affairs and Climate Action as part of the evaluation of gambling hall regulations.In the last 5 years AK was mainly funded by grants from the German Research Foundation (Deutsche Forschungsgemeinschaft – DFG) for the Collaborative Research Centers SFB 940 (project number 178833530) and the TRR 265 (project number 402170461). She was additionally funded by a grant from the International Center for Responsible Gaming (ICRG) and by an unrestricted research grant from the Federal Ministry of Economic Affairs and Climate Action as part of the evaluation of gambling hall regulations. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References 1. Goschke, T. Dysfunctions of decision-making and cognitive control as transdiagnostic mechanisms of mental disorders: Advances, gaps, and needs in current research. Int. J. Methods Psychiatr. Res. 23 , 41–57 (2014). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Bickel, W. K. et al. Excessive discounting of delayed reinforcers as a trans-disease process: Update on the state of the science. Curr. Opin. Psychol. 30 , 59–64 (2019). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. Kahneman, D. & Tversky, A. Prospect Theory: An Analysis of Decision under Risk. Econometrica 47 , 263–291 (1979). [ Google Scholar ] 4. Tversky, A. & Kahneman, D. Advances in prospect theory: Cumulative representation of uncertainty. J. Risk Uncertain. 5 , 297–323 (1992). [ Google Scholar ] 5. Leppink, E. W., Redden, S. A., Chamberlain, S. R. & Grant, J. E. Cognitive flexibility correlates with gambling severity in young adults. J. Psychiatr. Res. 81 , 9–15 (2016). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Odlaug, B. L., Chamberlain, S. R., Kim, S. W., Schreiber, L. R. N. & Grant, J. E. A neurocognitive comparison of cognitive flexibility and response inhibition in gamblers with varying degrees of clinical severity. Psychol. Med. 41 , 2111–2119 (2011). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. van Timmeren, T., Daams, J. G., van Holst, R. J. & Goudriaan, A. E. Compulsivity-related neurocognitive performance deficits in gambling disorder: A systematic review and meta-analysis. Neurosci. Biobehavioral Reviews . 84 , 204–217 (2018). [ DOI ] [ PubMed ] [ Google Scholar ] 8. Ngetich, R. et al. Working memory performance in disordered gambling and gaming: A systematic review. Compr. Psychiatr. 126 , 152408 (2023). [ DOI ] [ PubMed ] [ Google Scholar ] 9. Chowdhury, N. S., Livesey, E. J., Blaszczynski, A. & Harris, J. A. Pathological gambling and motor impulsivity: A systematic review with meta-analysis. J. Gambl. Stud. 33 , 1213–1239 (2017). [ DOI ] [ PubMed ] [ Google Scholar ] 10. Ioannidis, K., Hook, R., Wickham, K., Grant, J. E. & Chamberlain, S. R. Impulsivity in Gambling Disorder and problem gambling: a meta-analysis. Neuropsychopharmacol 44 , 1354–1361 (2019). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Smith, J. L., Mattick, R. P., Jamadar, S. D. & Iredale, J. M. Deficits in behavioural inhibition in substance abuse and addiction: A meta-analysis. Drug. Alcohol. Depend. 145 , 1–33 (2014). [ DOI ] [ PubMed ] [ Google Scholar ] 12. Amlung, M., Vedelago, L., Acker, J., Balodis, I. & MacKillop, J. Steep delay discounting and addictive behavior: a meta-analysis of continuous associations. Addiction 112 , 51–62 (2017). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 13. Kyonka, E. G. E. & Schutte, N. S. Probability discounting and gambling: a meta-analysis. Addiction 113 , 2173–2181 (2018). [ DOI ] [ PubMed ] [ Google Scholar ] 14. Genauck, A. et al. Reduced loss aversion in pathological gambling and alcohol dependence is associated with differential alterations in amygdala and prefrontal functioning. Sci. Rep. 7 , 16306 (2017). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Verdejo-Garcia, A. & Manning, V. Executive Functioning in Gambling Disorder: Cognitive Profiles and Associations with Clinical Outcomes. Curr. Addict. Rep. 2 , 214–219 (2015). [ Google Scholar ] 16. Boog, M. et al. Cognitive Inflexibility in Gamblers is Primarily Present in Reward-Related Decision Making. Front. Hum. Neurosci. 8 , 569 (2014). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Ledgerwood, D. M. et al. Executive Function in Pathological Gamblers and Healthy Controls. J. Gambl. Stud. 28 , 89–103 (2012). [ DOI ] [ PubMed ] [ Google Scholar ] 18. Manning, V., Teoh, H. C., Guo, S., Wong, K. E. & Li, T. K. Executive functioning in Asian pathological gamblers. Int. Gambl. Stud. 13 , 403–416 (2013). [ Google Scholar ] 19. van Holst, R. J., van den Brink, W., Veltman, D. J. & Goudriaan, A. E. Why gamblers fail to win: A review of cognitive and neuroimaging findings in pathological gambling. Neurosci. Biobehavioral. Rev. 34 , 87–107 (2010). [ DOI ] [ PubMed ] [ Google Scholar ] 20. Albein-Urios, N., Martinez-González, J. M., Lozano, Ó. & Clark, L. Verdejo-García, A. Comparison of impulsivity and working memory in cocaine addiction and pathological gambling: Implications for cocaine-induced neurotoxicity. Drug. Alcohol. Depend. 126 , 1–6 (2012). [ DOI ] [ PubMed ] [ Google Scholar ] 21. Sharif-Razi, M., Hodgins, D. C. & Goghari, V. M. Reactive and proactive control mechanisms of response inhibition in gambling disorder. Psychiatry Res. 272 , 114–121 (2019). [ DOI ] [ PubMed ] [ Google Scholar ] 22. Mallorquí-Bagué, N. et al. Cognitive Deficits in Executive Functions and Decision-Making Impairments Cluster Gambling Disorder Sub-types. J. Gambl. Stud. 34 , 209–223 (2018). [ DOI ] [ PubMed ] [ Google Scholar ] 23. Wirkus, T., Czernecka, R., Bühringer, G. & Kräplin, A. The relationship between player tracking data and individual risk factors in online sports bettors: an explorative study. 10.31234/osf.io/52n7x (2024). 24. Álvarez-Moya, E. M. et al. Effect of executive functioning, decision-making and self-reported impulsivity on the treatment outcome of pathologic gambling. JPN 36 , 165–175 (2011). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Goudriaan, A. E., Oosterlaan, J., De Beurs, E. & Van Den Brink, W. The role of self-reported impulsivity and reward sensitivity versus neurocognitive measures of disinhibition and decision-making in the prediction of relapse in pathological gamblers. Psychol. Med. 38 , 41–50 (2008). [ DOI ] [ PubMed ] [ Google Scholar ] 26. Mallorquí-Bagué, N. et al. Trait impulsivity and cognitive domains involving impulsivity and compulsivity as predictors of gambling disorder treatment response. Addict. Behav. 87 , 169–176 (2018). [ DOI ] [ PubMed ] [ Google Scholar ] 27. Kräplin, A. et al. The relationship between executive functioning and addictive behavior: new insights from a longitudinal community study. Psychopharmacology 239 , 3507–3524 (2022). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Shead, N. W. & Hodgins, D. C. Probability Discounting of Gains and Losses: Implications for Risk Attitudes and Impulsivity. J. Exp. Anal. Behav. 92 , 1–16 (2009). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Ohmura, Y., Takahashi, T. & Kitamura, N. Discounting delayed and probabilistic monetary gains and losses by smokers of cigarettes. Psychopharmacology 182 , 508–515 (2005). [ DOI ] [ PubMed ] [ Google Scholar ] 30. MacKillop, J. et al. Delayed reward discounting and addictive behavior: a meta-analysis. Psychopharmacol. (Berl) . 216 , 305–321 (2011). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Weinsztok, S., Brassard, S., Balodis, I., Martin, L. E. & Amlung, M. Delay Discounting in Established and Proposed Behavioral Addictions: A Systematic Review and Meta-Analysis. Front. Behav. Neurosci. 15 , 786358 (2021). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Brevers, D. et al. Impulsive Action but Not Impulsive Choice Determines Problem Gambling Severity. PLOS ONE . 7 , e50647 (2012). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Giorgetta, C. et al. To play or not to play: A personal dilemma in pathological gambling. Psychiatry Res. 219 , 562–569 (2014). [ DOI ] [ PubMed ] [ Google Scholar ] 34. Takeuchi, H. et al. Heterogeneity of Loss Aversion in Pathological Gambling. J. Gambl. Stud. 32 , 1143–1154 (2016). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Kräplin, A. et al. Impulsive decision-making predicts the course of substance-related and addictive disorders. Psychopharmacology 237 , 2709–2724 (2020). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Petry, N. M. Discounting of probabilistic rewards is associated with gambling abstinence in treatment-seeking pathological gamblers. J. Abnorm. Psychol. 121 , 151–159 (2012). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Castrén, S., Heiskanen, M. & Salonen, A. H. Trends in gambling participation and gambling severity among Finnish men and women: cross-sectional population surveys in 2007, 2010 and 2015. BMJ Open. 8 , e022129 (2018). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Chóliz, M., Marcos, M. & Lázaro-Mateo, J. The Risk of Online Gambling: a Study of Gambling Disorder Prevalence Rates in Spain. Int. J. Ment Health Addict. 19 , 404–417 (2021). [ Google Scholar ] 39. Gemeinsame Geschäftsstelle Glücksspielen. Jahresreport Der Glücksspielaufsichtbehörden Der Länder. Der Deutsche Glücksspielmarkt - Eine Ökonomische Darstellung https://innen.hessen.de/sites/innen.hessen.de/files/2022-12/jahresreport_2021_0.pdf . (2022). 40. Czernecka, R., Wirkus, T., Bühringer, G. & Kräplin, A. Characteristics and prediction of risky gambling behaviour study: A study protocol. Int. J. Methods Psychiatr. Res. 33 , e1995 (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Wirkus, T., Czernecka, R., Kräplin, A. & Bühringer, G. Characteristics and prediction of risky gambling behavior (RIGAB study) - Initial online survey. 10.17605/OSF.IO/JBFHE (2021). 42. Wirkus, T., Czernecka, R., Kräplin, A. & Bühringer, G. Characteristics and prediction of risky gambling behavior (RIGAB study) - In-person study. 10.17605/OSF.IO/G3NFV10.17605/OSF.IO/G3NFV (2022). [ DOI ] 43. Faul, F., Erdfelder, E., Lang, A. G. & Bucher, A. G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior. Res. Methods 39 , (2007). [ DOI ] [ PubMed ] 44. Czernecka, R., Wirkus, T., Bühringer, G. & Kräplin, A. Characteristics and prediction of risky gambling behaviour (RIGAB study): a study protocol. Inter. J. Methods Psych. Res. . [ DOI ] [ PMC free article ] [ PubMed ] 45. Wirkus, T., Czernecka, R. & Kräplin, A. Individual risk factors and prediction of gambling disorder in online sports bettors - the longitudinal RIGAB study. Front Psychiatry 15 , (2024). [ DOI ] [ PMC free article ] [ PubMed ] 46. Tipico Spielerschutzkonzept. (2023). 47. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (5th ed.). Washington, D.C., (2013). 48. Buchner, U., Irles-Garcia, V., Koytek, A. & Kroher, M. Praxishandbuch Glücksspiel (Landesstelle Glücksspielsucht in Bayern, 2009). 49. Wittchen, H. U. & Pfister, H. Compos. Int. Diagn. Interview (1997). 50. Stinchfield, R. et al. Reliability, Validity, and Classification Accuracy of the DSM-5 Diagnostic Criteria for Gambling Disorder and Comparison to DSM-IV. J. Gambl. Stud. 32 , 905–922 (2016). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 51. Meyer, G., Kalke, J. & Buth, S. Problem gambling in Germany: results of a mixed-mode population survey in 2021. Int. Gambl. Stud. 0 , 1–18 (2023). [ Google Scholar ] 52. Bruyer, R. & Brysbaert, M. Combining Speed and Accuracy in Cognitive Psychology: Is the Inverse Efficiency Score (IES) a Better Dependent Variable than the Mean Reaction Time (RT) and the Percentage Of Errors (PE)? Psychol. Belg. 51 , 5 (2011). [ Google Scholar ] 53. Wolff, M. et al. Action versus state orientation moderates the impact of executive functioning on real-life self-control. J. Exp. Psychol. Gen. 145 , 1635–1653 (2016). [ DOI ] [ PubMed ] [ Google Scholar ] 54. Wolff, M. et al. Chronic stress, executive functioning, and real-life self-control: An experience sampling study. J. Pers. 89 , 402–421 (2021). [ DOI ] [ PubMed ] [ Google Scholar ] 55. Bogacz, R. Speed-accuracy trade-off. in Encyclopedia of computational neuroscience (eds (eds Jaeger, D. & Jung, R.) 2798–2801 (Springer, New York, (2015). [ Google Scholar ] 56. Wilcox, R. R. & Keselman, H. J. Modern robust data analysis methods: measures of central tendency. Psychol. Methods . 8 , 254–274 (2003). [ DOI ] [ PubMed ] [ Google Scholar ] 57. Pooseh, S., Bernhardt, N., Guevara, A., Huys, Q. J. M. & Smolka, M. N. Value-based decision-making battery: A Bayesian adaptive approach to assess impulsive and risky behavior. Behav. Res. 50 , 236–249 (2018). [ DOI ] [ PubMed ] [ Google Scholar ] 58. Brainard, D. H. The Psychophysics Toolbox. Spat. Vis. 10 , 433–436 (1997). [ PubMed ] [ Google Scholar ] 59. The MathWorks Inc. MATLAB R2010a. (2010). 60. Mazur, J. E. An Adjusting Procedure for Studying Delayed Reinforcement. in The Effect of Delay and of Intervening Events on Reinforcement Value (Psychology, (1987). 61. Tom, S. M., Fox, C. R., Trepel, C. & Poldrack, R. A. The Neural Basis of Loss Aversion in Decision-Making Under Risk. Science 315 , 515–518 (2007). [ DOI ] [ PubMed ] [ Google Scholar ] 62. Wirkus, T., Czernecka, R., Kräplin, A. & Bühringer, G. Characteristics and prediction of risky gambling behavior (RIGAB study) https://osf.io/k6c23/ . (2021). [ DOI ] [ PMC free article ] [ PubMed ] 63. Czernecka, R., Wirkus, T., Bühringer, G. & Kräplin, A. The RIGAB study: study design. https://osf.io/3sy8w (2023). [ DOI ] [ PMC free article ] [ PubMed ] 64. Harris, P. A. et al. The REDCap consortium: Building an international community of software platform partners. J. Biomed. Inform. 95 , 103208 (2019). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 65. Harris, P. A. et al. Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. J. Biomed. Inform. 42 , 377–381 (2009). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 66. Czernecka, R., Wirkus, T., Bühringer, G. & Kräplin, A. The RIGAB study: all measures. https://osf.io/ac8gj (2023). 67. Wittchen, H. U. & Pfister, H. DIA-X-Interviews: Manual Für Screening-Verfahren Und Interview (Swets & Zeitlinger, 1997). 68. StataCorp. Stata Statistical Software. StataCorp LP. (2015). 69. Dowling, N. A. et al. Early risk and protective factors for problem gambling: A systematic review and meta-analysis of longitudinal studies. Clin. Psychol. Rev. 51 , 109–124 (2017). [ DOI ] [ PubMed ] [ Google Scholar ] 70. Gray, H. M., LaPlante, D. A. & Shaffer, H. J. Behavioral characteristics of Internet gamblers who trigger corporate responsible gambling interventions. Psychol. Addict. Behav. 26 , 527–535 (2012). [ DOI ] [ PubMed ] [ Google Scholar ] 71. Hing, N., Russell, A. M. T., Vitartas, P. & Lamont, M. Demographic, behavioural and normative risk factors for gambling problems amongst sports bettors. J. Gambl. Stud. 32 , 625–641 (2016). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 72. Rubin, M. When to adjust alpha during multiple testing: a consideration of disjunction, conjunction, and individual testing. Synthese 199 , 10969–11000 (2021). [ Google Scholar ] 73. American Psychiatric Association (APA). Diagnostic and Statistical Manual of Mental Disorders (5th Ed.). American Psychiatric Association, Washington, DC, (2013). 74. Sharman, S. et al. Heterogeneity in Disordered Gambling: Decision-Making and Impulsivity in Gamblers Grouped by Preferred Form. Front. Psychiatry 10 , (2019). [ DOI ] [ PMC free article ] [ PubMed ] 75. Kapsomenakis, A., Simos, P. G., Konstantakopoulos, G. & Kasselimis, D. S. In Search of Executive Impairment in Pathological Gambling: A Neuropsychological Study on Non-treatment Seeking Gamblers. J. Gambl. Stud. 34 , 1327–1340 (2018). [ DOI ] [ PubMed ] [ Google Scholar ] 76. Lorains, F. K., Stout, J. C., Bradshaw, J. L., Dowling, N. A. & Enticott, P. G. Self-reported impulsivity and inhibitory control in problem gamblers. J. Clin. Exp. Neuropsychol. 36 , 144–157 (2014). [ DOI ] [ PubMed ] [ Google Scholar ] 77. Black, D. W. & Shaw, M. The Epidemiology of Gambling Disorder. in Gambling Disorder (eds (eds Heinz, A., Romanczuk-Seiferth, N. & Potenza, M. N.) 29–48 (Springer International Publishing, Cham, 10.1007/978-3-030-03060-5_3. (2019). [ Google Scholar ] 78. LaPlante, D. A., Nelson, S. E., LaBrie, R. A. & Shaffer, H. J. Stability and Progression of Disordered Gambling: Lessons from Longitudinal Studies. Can. J. Psychiatry . 53 , 52–60 (2008). [ DOI ] [ PubMed ] [ Google Scholar ] 79. Blaszczynski, A. & Nower, L. A pathways model of problem and pathological gambling. Addiction 97 , 487–499 (2002). [ DOI ] [ PubMed ] [ Google Scholar ] 80. Nower, L., Blaszczynski, A. & Anthony, W. L. Clarifying gambling subtypes: the revised pathways model of problem gambling. Addiction 117 , 2000–2008 (2022). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 81. Lorains, F. K. et al. Strategic and non-strategic problem gamblers differ on decision-making under risk and ambiguity. Addiction 109 , 1128–1137 (2014). [ DOI ] [ PubMed ] [ Google Scholar ] 82. Wang, L., Tian, M., Zheng, Y., Li, Q. & Liu, X. Reduced loss aversion and inhibitory control in adolescents with internet gaming disorder. Psychol. Addict. Behav. 34 , 484–496 (2020). [ DOI ] [ PubMed ] [ Google Scholar ] 83. Yakovenko, I., Quigley, L., Hemmelgarn, B. R., Hodgins, D. C. & Ronksley, P. The efficacy of motivational interviewing for disordered gambling: Systematic review and meta-analysis. Addict. Behav. 43 , 72–82 (2015). [ DOI ] [ PubMed ] [ Google Scholar ] 84. Kim, S. W., Grant, J. E., Potenza, M. N., Blanco, C. & Hollander, E. The Gambling Symptom Assessment Scale (G-SAS): A reliability and validity study. Psychiatry Res. 166 , 76–84 (2009). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 85. Hing, N., Russell, A., Tolchard, B. & Nower, L. Risk Factors for Gambling Problems: An Analysis by Gender. J. Gambl. Stud. 32 , 511–534 (2016). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1 (25.2KB, docx) Data Availability Statement The datasets presented in this article are not publicly available due to their inclusion of sensitive information, such as player tracking data. Since the raw data could potentially allow for the identification of individuals, further anonymization, such as data aggregation, is required before any public release. We are currently preparing the data in accordance with FAIR principles and applying the necessary anonymization procedures. This process is expected to be completed by autumn 2026. Afterward, the anonymized dataset will be made publicly available. Sensitive data will be accessible only in aggregated form or through controlled access within an open repository for qualified research purposes. Until then, data may be made available upon reasonable request to the corresponding author. Articles from Scientific Reports are provided here courtesy of Nature Publishing Group ACTIONS View on publisher site PDF (1.8 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top