Prevalence of doping use in elite sports: a review of numbers and methods.

The prevalence of doping in elite sports is relevant for all those involved in sports, particularly for evaluating anti-doping policy measures. Remarkably, few scientific articles have addressed this subject so far, and the last review dates back to 1997. As a consequence, the true prevalence of doping in elite sports is unknown. Even though it is virtually impossible to uncover the exact prevalence of a prohibited activity such as doping, various methods are available to uncover parts of this particular problem, which enables the circumvention (to a certain degree) of the issues of truthfulness, definition problems and the limits of pharmacological evidence. This review outlines the various methods that exist and presents the scarce data available in this area. It is concluded that a combination of questionnaires using the Randomised Response Technique and models of biological parameters is able to provide the statistical possibilities to reveal accurate estimates of this often undisclosed practice. Data gathered in this way yield an estimation of 14-39% of current adult elite athletes who intentionally used doping. These period prevalences have been found in specific sub-groups of elite athletes, and the available data suggest that the prevalence of doping is considerably different between sub-groups with varying types of sport, levels and nationalities. The above-mentioned figure of 14-39% is likely to be a more accurate reflection of the prevalence of intentional doping in elite sports than that provided by doping control test results (estimate of doping: 1-2% annually) or questionnaire-based research (estimations between 1 and 70% depending on sport, level and exact definitions of intent and doping). In the future, analytical science may play a more important role in this topic if it may become feasible to detect very low concentrations of prohibited substances in sewage systems downstream of major sporting events. However, it is clear that current doping control test results show a distinct underestimation of true doping prevalence. It does not seem feasible to distil better estimates of the prevalence of doping based on performance indicators or ego documents because of the various existing effects that influence athletic performance. Such information can only be used as extra information to augment the accuracy of prevalence rates that have been found by using other techniques. True doping prevalence studies have been scarce in elite sports so far. With the correct application of the available scientific methods, preferably using harmonised definitions of the terms 'doping' and 'elite sports', more information on this topic may be gathered in a relatively short time. This would assist anti-doping professionals in the future in order to evaluate the effects of possible anti-doping measures, and better anti-doping policies would serve athletes who compete without doping. The existing anti-doping measures seriously impact the lives of elite athletes and their immediate entourage, which imposes a moral burden to evaluate these measures in the best possible way.