A systematic review and meta-analysis of 34 studies using indirect estimation methods found that roughly 1 in 5 multi-sport competitive athletes admitted to ever doping, while past-year prevalence varies between 8.7% and 15.5% depending on sport type and competition level.
Researchers conducted a comprehensive review of 46 records examining doping prevalence in sport using indirect estimation models (IEM), statistical techniques designed to elicit honest responses about sensitive behaviors like performance-enhancing drug use. The final analysis included 34 independent studies published through March 2025, making this the first systematic review and meta-analysis of its kind to synthesize prevalence estimates across different athlete populations and study methodologies.
The findings reveal a nuanced landscape of doping prevalence that varies significantly based on athlete characteristics. Lifetime prevalence (whether an athlete has ever doped) reached 22.6% among multi-sport competitive athletes but dropped to 12.7% among single-sport competitive athletes. This suggests that athletes competing across multiple sports face different risk factors or pressures compared to those focused on a single discipline. When looking at past-year prevalence, the pattern reverses: single-sport recreational athletes reported the highest rates at 15.5%, while multi-sport recreational athletes reported the lowest at 8.7%. The overall takeaway is striking: approximately one in five multi-sport competitive athletes admitted to ever doping, while about one in six single-sport recreational athletes reported doping in the past year.
The meta-analysis also characterized the methodological landscape of existing research. Most studies (k=18) applied the World Anti-Doping Agency's definition of doping use, and prevalence assessment focused primarily on past-year or past-season behavior (k=20). The Unrelated Question technique and Forced Response with Cheater Detection models were the most common statistical approaches. Notably, 22 of the 30 meta-analyzed studies came from Europe, suggesting potential geographic limitations in the current evidence base. Data collection occurred outside sport events in 28 studies, which may influence prevalence estimates by reducing social desirability bias and other confounding factors present during competitive environments.
A methodological challenge that emerged during the review involved secondary re-analyses: four studies re-examined previously published datasets rather than collecting new data. The authors flagged this as a novel consideration for future meta-analyses, as it complicates dependency assumptions and may inflate or deflate pooled estimates if not properly accounted for statistically.
This research provides evidence-based prevalence estimates rather than moral or regulatory guidance, but several practical implications emerge:
For athletes: The variation in prevalence across sport types and competition levels suggests that risk factors for doping differ meaningfully by context. Multi-sport competitive athletes face the highest lifetime prevalence, which could reflect greater competitive pressure, more opportunities to use performance-enhancing substances, or different motivational structures compared to single-sport athletes.
For sport administrators and researchers: The geographic concentration of studies in Europe reveals a research gap. Prevalence estimates from other regions would clarify whether findings generalize globally or reflect regional policy, anti-doping enforcement, or cultural differences in performance enhancement attitudes.
For interpreting future research: Indirect estimation models provide more honest prevalence estimates than direct surveys, since they reduce social desirability bias through randomization techniques. However, the reliance on past-year or past-season measurements in most studies means lifetime doping behavior may be underestimated if participants do not accurately recall distant drug use.
The study does not evaluate the effectiveness of anti-doping interventions or explain why prevalence rates differ across athlete types. Future research should investigate the mechanisms driving these differences and test prevention strategies in populations showing elevated prevalence.
| Aspect | Detail |
|---|---|
| Study Type | Systematic review and meta-analysis |
| Records Reviewed | 46 |
| Independent Studies Analyzed | 34 |
| Geographic Focus | Primarily Europe (k=22) |
| Primary Outcome | Lifetime and past-year doping prevalence |
| Key Prevalence Findings | Lifetime: 12.7%-22.6%; Past-year: 8.7%-15.5% |
| Main Statistical Methods | Unrelated Question (k=8); Forced Response with Cheater Detection (k=6) |
| Definition Used | World Anti-Doping Agency definition (k=18 studies) |
| Data Collection Timing | Outside sport events (k=28 studies) |
| PROSPERO Registration | CRD42022373691 |
| Publication Date | 2025 |
Hensel, J., et al. (2025). Doping Prevalence in Sport from Indirect Estimation Models: A Systematic Review and Meta-analysis. Sports Medicine Open. PubMed ID: 42251609
ProtocolEngine provides general health information based on published research. This is not medical advice. Consult a healthcare professional before starting any supplement or health protocol.