The Prevalence of Legal Performance-Enhancing Substance Use and Potential Cognitive and or Physical Doping in German Recreational Triathletes, Assessed via the Randomised Response Technique

is study investigated the use of performance-enhancing substances in recreational triathletes who were competing in German races at distances ranging from super-sprint to long-distance, as per the International Triathlon Union. The use of legal drugs and over-the-counter supplements over the previous year, painkillers over the previous 3 months, and the potential three-month prevalence of physical doping and or cognitive doping in this group were assessed via an anonymous questionnaire. The Randomised Response Technique (RRT) was implemented for sensitive questions regarding “prescription drugs [: : :] for the purpose of performance enhancement [: : :] only available at a pharmacy or on the black market”. The survey did not directly state the word “doping,” but included examples of substances that could later be classed as physical and or cognitive doping. The subjects were not required to detail what they were taking. Overall, 1953 completed questionnaires were received from 3134 registered starters at six regional events—themselves involving 17 separate races—in 2017. Of the respondents, 31.8% and 11.3% admitted to the use of dietary supplements, and of painkillers during the previous three months, respectively. Potential physical doping and cognitive doping over the preceding year were reported by 7.0% (Con dence Interval CI: 4.2–9.8) and 9.4% (CI: 6.6–12.3) of triathletes. Gender, age, experience in endurance sports, and number of weekly

separate races—in 2017. Of the respondents, 31.8% and 11.3% admitted to the use of dietary supplements, and of painkillers during the previous three months, respectively. Potential physical doping and cognitive doping over the preceding year were reported by 7.0% (Con dence Interval CI: 4.2–9.8) and 9.4% (CI: 6.6–12.3) of triathletes. Gender, age, experience in endurance sports, and number of weekly triathlon training hours were linked to potential physical or cognitive doping. Given the potentially relevant side e ects of painkiller use and physical and or cognitive doping, we recommend that educational and preventative measures for them be implemented within amateur triathlons. Keywords:doping; painkillers; triathlon; recreational athletes; risk factors; RRT 1. Introduction The World Anti-Doping Agency (WADA) declares doping as fundamentally contrary to the spirit of sport [1], and this is the moral basis for them producing an annual list of banned substances. These substances can be divided in two groups according to their mode of action. Physical doping agents (e.g., sympathomimetics, anabolic steroids, or erythropoietin) have a direct e ect on physical aspects Sports2019,7, 241; doi:10.3390 /sports7120241 /journal/sports

Sports2019,7, 241 2 of 11 of the body. In contrast, cognitive doping agents (which include stimulants such as amphetamines, methylphenidate, and antidepressants) target the central nervous system. All such substances often have multiple unexpected side e ects. Even over-the-counter painkillers can provoke life-threatening risks such as hyponatraemia, uncontrolled haemorrhage, and myocardial or renal infarction [2,3]. Often these side e ects are unknown or simply ignored [4]. To protect the athlete's health, it is important to avoid the non-therapeutic use of multiple substances. Investigations into state-sponsored doping in Russia and into positive doping cases at the World Athletics Championships have recently highlighted that doping is a major problem in elite sports [5–7]. Given the high extent of substance abuse amongst elite athletes, it is likely that a substantial number of recreational athletes behave in the same way [8–10]. As in all endurance sports, a considerable amount of doping cases are reported in triathlons [11–13]. However, triathletes are a diverse population, with some athletes competing in a total race distance of less than 15 km, whilst others race over more than 220 km [14]. Although the latter so-called “long-distance” group only represents approximately one-tenth of all recreational triathletes [15], it is the only triathlete group thus far within which the prevalence of doping has been examined [4,11,12,16]. When such athletes were asked whether they had used banned substances over the previous 12 months, the doping prevalence was found to exceed 10% [11,12,16]. Yet the real number of substance abusers remains an estimated value. Even doping controls underestimate the actual number of abuse by a factor of 8 [17]. This, to our knowledge, is the rst study to investigate the doping behaviour of recreational triathletes racing over a wide range of distances (from super-sprint up to long-distance) as per the o cial International Triathlon Union (ITU). 2. Materials and Methods 2.1. Sample, Ethics, Races, Procedure Ethical approval to perform this study was obtained from the local University Ethical Committee (document number: EK 74022017). Triathletes were surveyed at six di erent triathlon events in central Germany in 2017. The race distances that were involved

super-sprint up to long-distance) as per the o cial International Triathlon Union (ITU). 2. Materials and Methods 2.1. Sample, Ethics, Races, Procedure Ethical approval to perform this study was obtained from the local University Ethical Committee (document number: EK 74022017). Triathletes were surveyed at six di erent triathlon events in central Germany in 2017. The race distances that were involved ranged from super-sprint to long-distance triathlons (Table). Table 1. The o cial International Triathlon Union (ITU) distances are presented, with the associated competition distances of swimming, cycling, and running [14] as well as the event locations, where athletes of the corresponding distances were interviewed. Race location: Gera (G), Jena (J), Koberbach (K), Leipzig (L), Moritzburg (M), Nordhausen (N). O cial Distance Swim (km) Bike (km) Run (km) Race Location Super-sprint 0.4 10 2.5 G, K, J Sprint 0.75 20 5 G, M, K, J, L, N Olympic 1.5 40 10 G, M, K, L, N Half-distance 1.9 90 21 M, N Long-distance 3.9 180 42.2 M The paper-based questionnaire was distributed from the race registration o ces on the day before and on the race day, at each event. A written explanation of the study was provided at the same locale. The athletes were informed within the aforesaid explanation that the act of submitting the questionnaire implied their informed consent to participate in the study. The athletes were requested to complete the survey prior to or directly after race registration. All the forms were in the German language. The questionnaires were handed to age-group starters for all the available race distances. For this reason, the dataset obtained is more representative of the average recreational athlete than that collected by previous studies, all of which were exclusive to long-distance athletes [11,12]. For the purposes of anonymity, all the completed forms were collected in a black box and no information about the name, birth date, the distance raced, nor the estimated nish time of the participant was

For the purposes of anonymity, all the completed forms were collected in a black box and no information about the name, birth date, the distance raced, nor the estimated nish time of the participant was

Sports2019,7, 241 3 of 11 requested. The term “doping” was circumvented in the questionnaire in order to reduce any problems with compliance that this might cause [11]. It was replaced by the German equivalent of “prescription drugs [: : :] with the goal of increasing (mental or physical) performance. Substances that can only be obtained from a pharmacy or on the black market”. This indirect method of questioning was chosen because it typically yields higher prevalence rates for sensitive issues and thus a more valid picture of athletes' behaviour [12]. 2.2. Questionnaire At the beginning of the questionnaire, the athletes were informed both of the purpose of the survey and that participation was anonymous and voluntary. The Randomised Response Technique (RRT) was used to estimate the 12-month prevalence of prohibited substances [18]. The complete RRT question to assess the prevalence of physical doping is shown in Table. This assessment period was chosen, instead of life time prevalence, in order to obtain comparable data to those of previous studies. Some information was obtained via closed questions with respect to sex (male/female), A-level (i.e., possession of a German diploma that quali es the holder for university admission, yes/no), training in a group (yes/no), ingestion of painkillers during training/competition (prophylactic/therapeutic/rest in case of pain/no pain), 12-month prevalence for the use of legal and freely available substances for physical (yes/no), and cognitive (yes/no) enhancement. The next set of questions related to biographical data (e.g., age, height, and weight) and training behaviour such as previous years of training in endurance sports, and number of weekly training hours in swimming/cycling/running. Finally, the athletes were asked which triathlon distances (out of super-sprint, sprint, Olympic-distance, half-distance, and long-distance) they had competed over within the past 12 months. The athletes were also asked to give details of painkiller intake and the underlying rationale for said intake. The athletes' motivation for supplementation was consequently divided into prophylactic vs. therapeutic and training vs. competition related use. In order to question the gateway hypothesis, the question about dietary supplements di erentiated between the intake of physical (e.g., bronchodilator) or mental (e.g., concentration-enhancing)

athletes were also asked to give details of painkiller intake and the underlying rationale for said intake. The athletes' motivation for supplementation was consequently divided into prophylactic vs. therapeutic and training vs. competition related use. In order to question the gateway hypothesis, the question about dietary supplements di erentiated between the intake of physical (e.g., bronchodilator) or mental (e.g., concentration-enhancing) substances. The said hypothesis states that the usage of freely available substances may lead to the later abuse of prohibited substances for the purpose of performance enhancement [19]. Table 2. The Randomised Response Technique (RRT) procedure to assess for potential physical doping and cognitive doping. Physical doping Please consider a certain birthday (yours, your mother's, etc.). Is this birthday in the rst third of a month ( rst to tenth day)? If yes, please proceed to Question A; if no, please proceed to Question B. Question A Is this birthday in the rst half of the year (prior to the rst of July)? Question B Have you taken substances to increase your physical performance within the past 12 months that are only available at a pharmacy, at the doctor's o ce, or on the black market (e.g., anabolic steroids, erythropoietin, stimulants, growth hormones)? Note that only you know which of the questions you will answer Yes No Cognitive doping Please consider a certain birthday (yours, your mother's, etc.). Is this birthday in the rst third of a month ( rst to tenth day)? If yes, please proceed to Question A; if no, please proceed to Question B. Question A Is this birthday in the rst half of the year (prior to the rst of July)? Question B Have you taken substances to increase your mental performance in the past 12 months that are only available at a pharmacy, at the doctor's o ce, or on the black market (e.g., stimulants, cocaine, methylphenidate, antidepressants, beta-blockers, moda nil)? Note that only you know which of the questions you will answer Yes No

months that are only available at a pharmacy, at the doctor's o ce, or on the black market (e.g., stimulants, cocaine, methylphenidate, antidepressants, beta-blockers, moda nil)? Note that only you know which of the questions you will answer Yes No

Sports2019,7, 241 4 of 11 2.3. Randomised Response Technique (RRT) RRTs are speci cally developed to obtain more valid estimates when sensitive topics are studied, through their guarantee of a maximum amount of anonymity to the respondent [11,18,20]. In the present survey, a paper-and-pencil version of the unrelated question model (UQM) was used to estimate the potential prevalence of physical and cognitive doping (ˆ s) [21]. The UQM as it was used for the present study has been explicitly described by previous articles [11,12,22]. An explicit example of its calculation is provided by Franke et al. [23]. Similarly to their study, we used a probability for receiving the sensitive question (p) of 245.25/365.25. Afterwards the participants who were randomised to the sensitive group were asked a sensitive question (Questions B, Table), whilst the others were asked a neutral question. The probability for answering the neutral question with “yes” ( n) was 181.25/365.25. With this model, even the interviewer is unable to know whether the interviewee has answered the sensitive question or not. Furthermore, the RRT can be used to assess separate doping-prevalence values for sub-categories (e.g., females vs. males, users vs. non-users of painkillers), if the number of participants for several groups is high enough. For the purposes of this study the term “potential doping” is used for those athletes who (after completion of the RRT) gave a positive answer to the sensitive question. This de nition di ers from the WADA de nition of doping i.e., “doping is the occurrence of one or more of the anti-doping rule violations set forth in Article 2.1 through Article 2.10 of the Code” [24]. In order to minimise the time that was taken to complete the survey, and thereby maximise compliance, we did not ask the athletes to identify what substance(s) they were taking. Strictly speaking, therefore, our data relate to the potential prevalence, rather than the actual prevalence, of doping in German recreational triathletes. Our methodology, however, allows our results to be directly compared to those of the only previously published triathlete speci c studies in this area [11,12]. 2.4. Statistics

did not ask the athletes to identify what substance(s) they were taking. Strictly speaking, therefore, our data relate to the potential prevalence, rather than the actual prevalence, of doping in German recreational triathletes. Our methodology, however, allows our results to be directly compared to those of the only previously published triathlete speci c studies in this area [11,12]. 2.4. Statistics Descriptive data are presented as mean SD values for continuous scaled variables and as numbers and percentages for non-continuous scaled variables. They were obtained using SPSS software, version 22. Prevalence estimates (ˆ s) for physical and cognitive doping are presented as percentages with 95% con dence intervals (CI) and standard error (SE), as obtained via MATLAB version R2015a. The continuous variables “age” and “years doing endurance sports” were dichotomized by median. The splitting enabled us to calculate separate prevalence estimates, for example for younger/older athletes. Post-hoc power analyses [25] were performed for all RRT calculations, in order to test whether the sample sizes were adequate. 3. Results A total of 3134 recreational athletes, on the start lists for six di erent triathlon events, were surveyed. Overall, 1989 (63.5%) questionnaires were received from them (Tables). Of these, 1953 forms (98.2%) were su ciently completed and evaluated. More than half of all the questionnaires (n=1046; 53.6%) were collected at the Schlosstriathlon Moritzburg (sprint, Olympic, half-distance, and long-distance) event, and about one- fth (n=419; 21.5%) were obtained at the Leipzig (sprint and Olympic-distance) triathlon. The other races that were surveyed yielded smaller amounts of data: 170 forms at Powertriathlon Gera (8.7%, involving super-sprint, sprint, and Olympic-distance races); 127 forms at Koberbachtalsperre (6.5%, from super-sprint, sprint, and Olympic-distance triathlons); 106 forms at Paradiestriathlon Jena (5.4%, from super-sprint and sprint races); and 84 forms at ICAN Nordhausen (4.3%, for Olympic-distance and half-distance triathlons). Of the study participants, 76.4% (n=1491) were male. The mean athlete age was 39.6 years. Subject and event characteristics are presented in Table.

ICAN Nordhausen (4.3%, for Olympic-distance and half-distance triathlons). Of the study participants, 76.4% (n=1491) were male. The mean athlete age was 39.6 years. Subject and event characteristics are presented in Table.

Sports2019,7, 241 5 of 11 Table 3. The distribution of athletes from the di erent locations, with biographical data, and training behaviour. Race location: Gera (G), Jena (J), Koberbach (K), Leipzig (L), Moritzburg (M), Nordhausen (N). Race Athletes n=3134 Participants (Total) n=1989 Response Rate 63.5% Location Moritzburg 53.6% (n=1046) Leipzig 21.5% (n=419) Gera 8.7% (n=170) Koberbach 6.5% (n=127) Jena 5.4% (n=106) Nordhausen 4.3% (n=84) Gender 76.4% male (n=1477) 23.6% female (n=456) Age in years, (mean; SD) 18–80 (39.6 10.7) Height cm, (mean; SD) 150–202 (177.9 8.4) Weight in kg, (mean; SD) 46–130 (78.7 11.6) BMI kg/m 2 , (mean; SD) Male 14.8–41.3 (24.0 2.4) Female 14.7–34.6 (21.9 2.4) A-Level (German diploma, quali es the holder for university admission) 69.2% yes (n=1351) 30.0% no (n=586) Years of triathlon-speci c training, years (mean; SD) 0–50 (11.9 9.7) Hours swimming/week (mean; SD) 0–12 (1.56 1.23) Hours bike/week, (mean; SD) 0–20 (4.20 3.00) Hours running/week (mean; SD) 0–20 (2.79 1.87) Hours of training in total (mean; SD) 0–39 (8.56 2.14) distances No distance raced 14.8% (n=289) Super-sprint (race location G, J, K) 4.8% (n=94) Sprint (race location G, J, K, L, M, N) 51.2% (n=999) Olympic (race location G, K, L, M, N) 43.7% (n=853) Half-Distance (race location M, N) 24.1% (n=470) Long-Distance (race location M) 8.9% (n=173) 3.1. Dietary Supplements and Painkillers Of the study respondents, 31.8% declared that they had taken dietary supplements, 6.9% reported the use of cognitive enhancers and 9.7% stated that they had used physical enhancers. Intake of substances from both of the latter groups was reported by 14.2% of the athletes who took part in the study. Painkiller use within the previous three months was reported by 11.3% of participants. We found slight di erences between training and competition in the underlying rationale that was given for such intake. The prevalence of within competition painkiller use for therapeutic reasons (3.5%) was similar to that for prophylactic (3.6%) reasons. However, during training sessions more athletes used painkillers to treat their pain (4.7%), than to avoid it (2.0%). Furthermore, we identi ed that the intake of painkillers was associated with

The study assesses the prevalence of legal performance-enhancing substance use among German recreational triathletes.