Coaches strongly influence athletes' attitudes toward doping and can shape athlete's beliefs, behaviors, and decisions to be for or against doping (Barkoukis et al., 2019; García-Grimau et al., 2021). The impact of the athlete's entourage on attitudes toward doping has been reported in different studies (Backhouse and McKenna, 2012; Mazanov et al., 2014; Engelberg and Moston, 2016) and the coach is considered part of the Athlete Support Personnel (ASP) as well as parents/guardians, physiotherapists and other professionals supporting and working directly with athletes, as defined by the World Anti-Doping Code (WADC) (World Anti-Doping Agency, 2021a). Some specific roles and responsibilities of the ASP are the following: cooperating with anti-doping organizations, complying with anti-doping regulations that are applicable to them or their athletes, and using their influence on athletes' values and behavior to foster anti-doping rules (World Anti-Doping Agency, 2021a). Coaches could engage on doping behavior, act as social facilitators in doping in sport (Vakhitova and Bell, 2018) and incite their athletes to commit anti-doping rule violations (ADRVs) (i.e., those displayed in articles ranging from 2.5 to 2.11 of the Code). In these cases, coaches would have conflicted with their responsibility of encouraging their athletes to avoid the use of doping. According to the current anti-doping legislation (World Anti-Doping Agency, 2021a), seven ADRVs are applicable to ASP: tampering or attempted tampering, possession of a prohibited substance or method, trafficking or attempted trafficking, administration, or attempted administration to any athlete of any prohibited substance or method, complicity, or attempted complicity, prohibited association, and acts to discourage or retaliate against reporting to authorities. These non-analytical ADRVs are typically reported in this population. For example, 16 sanctions were imposed worldwide on ASP in 2018 (World Anti-Doping Agency, 2020) and 155 members of ASP are currently serving a period of ineligibility (World Anti-Doping Agency, 2021b).

Beyond the legislative framework, different studies reported a lack of anti-doping knowledge in coaches (Mazanov et al., 2014; Morente-Sánchez and Zabala, 2015; Engelberg and Moston, 2016). Other studies analyzed their beliefs about threats to health (Scarpino et al., 1990; Laure et al., 2001) and their attitudes toward doping and doping prevention. Fung and Yuan (2006) examined perceived knowledge, actual knowledge, attitudes, subjective norms, and behavior in relation to doping and anti-doping in sport in 114 coaches through the conceptual framework of the theory of Planned Behavior (Ajzen, 1991). Results revealed a negative correlation between the perceived and actual knowledge, which implies that coaches may under-estimate or over-estimate their knowledge regarding performance enhancing drugs/substances (PES) and doping control. This finding reflects the complexity of the phenomenon of doping in sport (Fung and Yuan, 2006). The authors suggest the need to design mandatory educational programs specifically designed for coaches. Furthermore, coaches are typically aware that they have an important role in doping prevention and generally display negative attitudes toward doping (Backhouse et al., 2015). Nevertheless, they are not fully committed to the prevention of doping (Patterson et al., 2014; Engelberg et al., 2019), and they need more resources and support to be able to proactively prevent doping among their athletes (Barkoukis et al., 2019).

Attitudes toward doping represents a strong predictor of doping susceptibility and behavior (Gucciardi et al., 2010; Barkoukis et al., 2013; Blank et al., 2016; García-Grimau et al., 2020). However, studies analyzing factors affecting coaches' doping attitudes and behavior are scarce. Nonetheless, a few psychological factors among coaches which may influence the use of doping in their athletes have been highlighted. Sullivan et al. (2015) explored the doping confrontation efficacy in coaches through the Doping Confrontation Efficacy Scale. They reported that coaches who are more prone toward task-involved climates tend to show higher efficacy on confronting athletes, thus preventing them from doping use (Sullivan et al., 2015). Findings from a cluster randomized controlled trial reported that coaches adopting a motivational climate reduced athlete willingness to dope (Ntoumanis et al., 2021). In addition, moral disengagement was strongly correlated with doping attitudes and intentions toward doping in athletes (Kavussanu et al., 2019; García-Grimau et al., 2021) and plays an important role in coaching style and climate (Hodge and Lonsdale, 2011; Chen et al., 2017). Moreover, recent qualitative evidence (Patterson and Backhouse, 2018; Barkoukis et al., 2019) reported that displacement and diffusion of responsibility, which are considered two of the mechanisms involved in the moral disengagement construct (Bandura et al., 1996), influence coaches' opinions and behavior toward doping. Coaches delegate their role in anti-doping prevention on authorities, which reflects an unwillingness to proactively engage in anti-doping education; furthermore, Barkoukis et al. (2019) reported that the stigmatization of doping behavior prevents athletes and coaches from being effectively educated about doping. Horcajo and De La Vega (2016) carried out an experimental study in which the conviction about attitudes related to doping and the effect of deliberative thinking (i.e., high- vs. low-elaboration likelihood) on that conviction were analyzed in soccer coaches. They reported that coaches with high elaboration likelihood showed more conviction about their attitudes than those with low elaboration likelihood. Morente-Sánchez and Zabala (2015) examined knowledge, attitudes, and beliefs toward doping in 237 soccer technical staff (including 101 coaches) and highlighted their lack of doping-related knowledge.

Overall, coaching style and climate, and psychological and moral factors have been quantitatively and independently analyzed in the doping literature to finally understand their relationships with doping attitudes in athletes. However, studies examining a large number of factors influencing coaches' doping attitudes and behavior under a specific theoretical framework are scarce. Literature reviews carried out by Backhouse et al. (2007, 2015) confirmed the lack of quantitative research guided by a theoretical framework on doping attitudes and behavior in coaches. The Sport Drug Control Model (SDCM) was developed by Donovan et al. (2002) and incorporates different frameworks from the behavioral sciences such as the Theory of Reasoned Action (Fishbein, 1979) and the Theory of Planned Behavior (Ajzen, 1991). The SDCM analyzes several factors influencing athletes' attitudes toward and susceptibility to doping: morality, legitimacy, benefits and threat appraisals, motivational profiles, beliefs about reference groups' endorsement of doping methods/substances, use of legal supplements, beliefs about the availability of PES and relevant authorities' control over trafficking of doping methods/substances, beliefs about the affordability of doping methods/substances, attitudes toward doping, susceptibility to doping, and self-reported use of banned PES or methods (PESM). The World Anti-Doping Agency integrated this model as a guideline for anti-doping organizations (World Anti-Doping Agency, 2015) and it has been tested in athletes (Gucciardi et al., 2010; Jalleh et al., 2013; García-Grimau et al., 2021) showing validity and reliability. Jalleh et al. (2013) and García-Grimau et al. reported that the most influential factors were morality, reference group opinion, and legitimacy. Moreover, Donovan et al. (2002) observed that the SDCM could be adapted for application to ASP but has not yet been applied to coaches. In this study, the SDCM is applied for the first time in coaches with the aim of assessing the reproducibility of the model in ASP. Secondary study purpose was to determine the factors in the SDCM that most influence coaches' attitudes and susceptibility toward doping, and their doping prevalence.

Descriptive statistics for the different variables analyzed (see Table 1) indicate that coaches reported on average negative attitudes toward doping and low levels of susceptibility to doping and moral disengagement. With respect to psychological factors, coaches stated on average a high self-efficacy to refrain from doping and moderate ego-oriented goals. Regarding social norms, they reported a high subjective norm. They believed that, on average, reference groups would disapprove doping behaviors. With respect to descriptive norms, coaches perceived an average doping prevalence of 19.5% (1.95 ± 1.74 [mean ± SD]) (see Table 2 for further details). Measures showed good internal consistency and reliability, with omega (ω) values > 0.6, AVE values > 0.4 and CR values > 0.7 (see Table 1). Self-reported doping prevalence among coaches was 4.5%, and 3% acknowledged having had an athlete who has tested positive for a prohibited substance. The SEM analysis of the SDCM in coaches (Figure 1) revealed a good fit of the data: χ²/df = 1.76, p < 0.001, CFI = 0.93, TLI = 0.96, RMSEA = 0.062 (90% CI = 0.054, 0.070), and SRMR = 0.09. Covariance between moral disengagement and subjective norms, and between subjective norms and descriptive norms did not change the model fitness and improved the standardized parameter estimates and significance. Among the standardized parameter estimates (Figure 1), all the relationships were significant excepting threat appraisal, benefit appraisal, and subjective norms. Moral disengagement (β = 0.58, p < 0.001) and descriptive norms (β = 0.42, p < 0.001) were the strongest predictors of attitudes toward doping in coaches. Regarding motivational and psychological profiles, ego-oriented goals (β = 0.34, p ≤ 0.01) and self-efficacy to refrain from doping (β = 0.26, p < 0.05) were also good predictors of coaches' doping attitudes, while ego-oriented goals displayed a greater level of significance and higher standardized parameter estimate than self-efficacy. Moreover, attitudes toward doping predicted doping susceptibility significantly (β = 0.39, p < 0.001).

The main study objectives were to analyze attitudes and susceptibility toward doping in coaches while testing the SDCM for the first time in ASP, and to determine which factors were the strongest predictors of their doping attitudes. So far, the SDCM had been applied to athletes, its application in a new population is an added value to this research. The most important findings in regard to the relationship between the variables were that attitudes toward doping predicted high susceptibility to doping, and were highly influenced by moral disengagement and descriptive norms in athletics coaches. The observed strength of the relationship between attitudes toward doping and doping susceptibility agrees with results from other studies carried out in athletes (Barkoukis et al., 2013; Blank et al., 2016; García-Grimau et al., 2021). However, to the best of the authors' knowledge by the time of writing, this relationship was not previously tested in coaches and the SDCM was not ever adapted and examined in coaches, hence this study brings innovation to anti-doping research.

In addition, the fact that morality and more particularly moral disengagement was found to be the factor with the strongest influence on attitudes toward doping in coaches reveals that the more they are morally disconnected, the more their favorable attitudes toward doping. These results are in line with other studies in athletes (Jalleh et al., 2013; García-Grimau et al., 2021) and coaches (Patterson and Backhouse, 2018; Barkoukis et al., 2019) in which moral disengagement displayed a strong influence on doping likelihood acting as a direct predictor or mediator (Ring and Kavussanu, 2018b). The relevance of moral variables on doping prevention has been proven in empirical research. For example, a recent anti-doping interventional study which focused on developing morality in British and Greek athletes reported a reduction in the likelihood of doping use (Kavussanu et al., 2021). In this sense, morality variables should be considered when implementing educational programs targeting ASP and coaches in particular.

Social norms were measured in the present study through subjective and descriptive norms. Our results show that descriptive norms were a significant and positive predictor of attitudes toward the use of doping in coaches, which reflects that the perception of high prevalence of doping in others could enhance their attitudes toward doping, as they tend to normalize this behavior. Their perception of incidence of performance enhancing drug use across all sports and for athletics were 23.4 and 20.0%, respectively (see Table 2). These results are consistent with those by Moston et al. (2015b), who reported that doping prevalence perception in 92 Australian coaches across all sports was 20.9%. Additionally, coaches believe that 25.3% of athletes will be engaged in doping in the next 2 years.

Subjective norms (i.e., social approval of significant others) did not display a significant influence on doping attitudes, which contrasts with findings from other studies performed in athletes (Lazuras et al., 2010; García-Grimau et al., 2021). For example, a recent SDCM analysis in Spanish national standard and elite track and field athletes reported that moral disengagement and subjective norms were the variables with the greatest influence on attitudes toward doping (García-Grimau et al., 2021). Therefore, the absence of influence of this construct on coaches' doping attitudes, which contrasts with that in athletes, reflects their need to receive anti-doping education from sporting bodies and anti-doping organizations given their important role as anti-doping educators. Overall, this finding highlights that normative beliefs can be considered determinants of attitudes and behavior toward doping in the sport society.

Regarding motivational profiles, self-efficacy, and ego-oriented goals were significant predictors of coaches' doping attitudes in the present study. Coaches with ego-oriented goals and less ability to resist temptations may be more prone to doping. This finding is consistent with results from Matosic et al. (2016, 2020), who reported that coach narcissism is positively associated with controlling coaching behavior, and athletes' perceptions of coach behavior predict athletes' attitudes toward doping. Moreover, self-regulatory efficacy in coaches helps to create a motivational climate (Sullivan et al., 2015), reduce athlete willingness to dope (Ntoumanis et al., 2021) and, according to our findings, prevent them from displaying positive attitudes toward doping. Hence, those motivational factors are strong predictors of attitudes and susceptibility to doping in both coaches and athletes.

Benefit and threat appraisal factors did not significantly predict coaches' doping attitudes. That means that the potential benefits or positive outcomes that a coach could achieve by cheating, and the level of threat perceived by the coach due to deterrent effect or risk to health, are not significant factors influencing their attitudes toward doping. Moston et al. (2015a) analyzed the perceived incidence of deterrents on drug use in Australian athletes and coaches and reported that coaches saw deterrents as less credible than athletes, with a low perceived likelihood of detection. Similarly, 75% of the coaches who participated in our study perceived the likelihood of being tested out of competition as low (not at all likely or not likely) and generally considered that doping has positive effects on athletes' performance. However, the latter would not result in substantial economic benefits or positive outcomes for coaches given that 81.3% of the coaches surveyed reported an average annual income derived from their coach role of <10,000 euros.

Coaches have a key role in doping prevention (Kirby et al., 2011), but they can also represent a risk factor regarding doping use in their athletes as they may encourage them to dope or supply them with banned PESM (Allen et al., 2017; Vakhitova and Bell, 2018). In our study, the percentage of self-reported doping administration was 4.5% among coaches. Despite the lack of studies analyzing doping prevalence in coaches in the current literature, Morente-Sánchez and Zabala (2015) reported that 8.1% of the Spanish soccer coaches analyzed in their study used banned substances while a literature review (Backhouse and McKenna, 2012) indicated that 4–6% of the coaches and rest of ASP analyzed reported personal use of banned substances. Moreover, doping in sport is a prohibited and socially rejected practice, methods for assessing doping prevalence remain unclear and therefore data prevalence is potentially underestimated due to the sensitivity of the question. The prevalence of doping in elite sports is likely to be between 14 and 39% (De Hon et al., 2015). A recent meta-analysis study shows a disparate range between 0 and 73% (Gleaves et al., 2021) where authors suggest best practice recommendations and guidelines to improve the evidence quality in this field. The use of multiple measures to triangulate doping prevalence data, and indirect measures like randomized response technique (RRT) may provide more reliable measurements (Gleaves et al., 2021).

Overall, adaptation and application of the SDCM in ASP proves its reproducibility in other population. Our results show that moral disengagement, social norms, and motivational profiles are the strongest predictors of attitudes toward doping among coaches and 4.5% of them supply their athletes with prohibited substances or methods. Therefore, it is necessary to address these psychological-, attitude-, and behavior-related issues through educational programs targeting coaches. Furthermore, sanctions applied to coaches and the rest of professionals conforming the ASP seem scarce (World Anti-Doping Agency, 2020), probably due to the non-analytical nature of these ADRV, since it is difficult to prove their doping behaviors and therefore to sanction them. In addition, most of the social science research literature on doping in sport has focused on athletes, while coach-centered studies remain limited (Backhouse et al., 2015; García-Grimau et al., 2020). The results of the present study alongside those from others reveal that coaches tend to morally disengage through a lack of commitment and a diffusion of their responsibilities as educators in doping prevention (Barkoukis et al., 2019), and consider that they do not have adequate tools to prevent their athletes from doping use, while being aware of their role as antidoping educators though (Engelberg et al., 2019; Patterson et al., 2019). All this scientific evidence paints a worrying picture, as coaches could rather represent a doping risk. On the one hand, the threshold prevalence of doping among coaches is considerable (i.e., between 4.5 and 8.1%). On the second hand, they manifested an absence and lack of interest in doping prevention. In the complex context of elite sport in which the influences of sport environment and reference group on athletes are crucial, the absence of doping prevention may involve the presence of risk of its use. Perhaps it is time to focus more efforts on coaches, without putting aside the athletes, and therefore turn coaches into reliable doping prevention factors. To this end, it is necessary to enhance scientific research and then develop, implement, and promote more educational programs targeting coaches, on a mandatory basis while covering the specific needs of coaches so that they can perform their role as anti-doping educators in an effective, committed, and proactive manner.

Attitudes toward the use of doping in Spanish track and field national coaches were analyzed for the first time to the best of authors' knowledge through the SDCM adaptation for coaches, which displayed reproducibility. Moral disengagement, social norms, and motivational profiles were the strongest predictors of positive attitudes toward doping among these coaches. Accordingly, further design of anti-doping research and education should target on developing and improving these abilities in coaches. It is necessary to enhance coach-centered research, provide more assistance to sport coaches, and establish effective and mandatory anti-doping education in them.

The datasets presented in this article are not readily available for ethical reasons according to an explicit condition set by the Ethics Committee from the World Anti-Doping Agency (2019-A2). Requests to access the datasets should be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Isabel I de Castilla International University (UI1-PI016) World Anti-Doping Agency (2019-A2). The patients/participants provided their written informed consent to participate in this study.

EG-G: designed, conceptualized, performed analysis, and wrote the paper. EG-G and AC: collected the data. AC and RD: contributed to the writing. All authors contributed to the article and approved the submitted version.

This work was funded by the World Anti-Doping Agency, within their Social Science Research Grants program and Isabel I de Castilla International University.

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

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.