An abductive inference approach to assess the performance-enhancing effects of drugs included on the World Anti-Doping Agency Prohibited List

Research output: Contribution to journalReviewResearchpeer-review

Some have questioned the evidence for performance-enhancing effects of several substances included on the World Anti-Doping Agency's Prohibited List due to the divergent or inconclusive findings in randomized controlled trials (RCTs). However, inductive statistical inference based on RCTs-only may result in biased conclusions because of the scarcity of studies, inter-study heterogeneity, too few outcome events, or insufficient power. An abductive inference approach, where the body of evidence is evaluated beyond considerations of statistical significance, may serve as a tool to assess the plausibility of performance-enhancing effects of substances by also considering observations and facts not solely obtained from RCTs. Herein, we explored the applicability of an abductive inference approach as a tool to assess the performance-enhancing effects of substances included on the Prohibited List. We applied an abductive inference approach to make inferences on debated issues pertaining to the ergogenic effects of recombinant human erythropoietin (rHuEPO), beta2-agonists and anabolic androgenic steroids (AAS), and extended the approach to more controversial drug classes where RCTs are limited. We report that an abductive inference approach is a useful tool to assess the ergogenic effect of substances included on the Prohibited List - particularly for substances where inductive inference is inconclusive. Specifically, a systematic abductive inference approach can aid researchers in assessing the effects of doping substances, either by leading to suggestions of causal relationships or identifying the need for additional research.

Original languageEnglish
JournalSports Medicine
Volume51
Issue number7
Pages (from-to)1353-1376
Number of pages24
ISSN0112-1642
DOIs
Publication statusPublished - 2021

Number of downloads are based on statistics from Google Scholar and www.ku.dk


No data available

ID: 259571025