Psychometric properties of the Expanded Exercise Addiction Inventory 3 (EAI-3) in a Danish sample

The risk of exercise addiction, characterized by an uncontrollable urge to engage in physical activity, poses significant health risks yet lacks clinical diagnostic criteria. The need for its assessment is increasing in research and applied settings. The present study evaluated the psychometric properties and reliability of the Expanded Exercise Addiction Inventory (EAI-3) within a Danish population. The present study involved 392 Danish adults who were all regular exercisers. Participants completed the EAI-3, the Exercise Dependence Scale-Revised (EDS-R), the SCOFF Questionnaire for eating disorders, the Obsessive–Compulsive Inventory-Revised (OCI-R), and the Ten-Item Personality Inventory (TIPI). Confirmatory factor analysis (CFA) and measurement invariance testing were performed to assess the factor structure and reliability of the EAI-3 across biological sex. The results indicated strong reliability and validity for the EAI-3, with good fit indices across models (CFI = .981, RMSEA = .054). The scale scores demonstrated configural, metric, and scalar invariance, indicating consistent performance across male and female exercisers. Reliability analyses yielded high internal consistency (α = .85, ω = .88), and ROC analysis established a cut-off score of 33.5 for potential exercise addiction risk, with high specificity (.856) and sensitivity (.889). Similar good results emerged from the bifactor model, but the original structure was still preferable. The present study supports the EAI-3 as a valid and reliable tool for screening the risk of exercise addiction among Danish adults, facilitating early identification and potential intervention. Further research should focus on longitudinal studies and clinical validations to enhance the understanding and management of exercise addiction.