Exercise addiction: A narrative overview of research issues

Aviv Weinstein, Attila Szabo

Research output: Contribution to journalReview articlepeer-review

13 Scopus citations


This narrative overview summarises the work on exercise addiction (EA) over the past 12 years and exposes critical conceptual and methodological issues. More than 1000 articles exist on EA, conceptualised as uncontrolled training harming the individual. Still, EA has no clinical diagnosis criteria at this time. Research is increasing continuously, but it is stale in advancing knowledge. Scalar measurement and lack of differentiation between addictive and instrumental exercise could be reasons for insufficient progress. Exercise addiction fits in the framework of behavioural addictions, but excessive exercise patterns also co-occur with other morbidities, including eating or body-image disorders. In these cases, exercise is instrumental; it functions to achieve a non-exercise-related goal. Therefore, it is essential to separate primary from secondary EA. Based on the interactional model, significant stress and capacity-exceeding ambitions fuel primary EA, while chief motives behind secondary EA embed body image dissatisfaction and eating disorders. Few reports exist on EA’s brain mechanisms, which could delay its classification as a distinct psychiatric dysfunction. Treatment of EA involves cognitive-behavioural approaches, but we know little about their effectiveness. Conceptually focussed psychophysiological research and in-depth interviews, complementing scalar data, could answer several open questions in this widely studied but relatively stagnant scholastic field.

Original languageEnglish
Pages (from-to)1-13
Number of pages13
JournalDialogues in Clinical Neuroscience
Issue number1
StatePublished - 2023


  • Behavioural addiction
  • compulsive exercise
  • depression
  • eating disorders
  • exercise dependence


Dive into the research topics of 'Exercise addiction: A narrative overview of research issues'. Together they form a unique fingerprint.

Cite this