Exercise addiction prevalence and correlates in the absence of eating disorder symptomology. A systematic review and meta-analysis

Background: Exercise addiction (EA) can be debilitating and can be a symptom of an eating disorder. To date, the prevalence rates of EA without indicated eating disorders in the general population and associated correlates remain unreported. Methods: Two authors searched major databases from inception to 31/12/2018 to identify studies investigating the prevalence of EA in any population without indicated eating disorders. We conducted a random effects meta-analysis to report (i) prevalence rates of EA using the exercise addiction inventory and exercise dependence scale and compare sub-populations, (ii) compare methods of EA measurement and explore heterogeneity, and (iii) report on correlates. Results: A total of 13 studies including 3635 people were included. The prevalence of EA among general exercisers was 8.1% (95% CI 1.5%–34.2%), amateur competitive athletes was 5.0% (95% CI 1.3%–17.3%), and university students was 5.5% (95% CI 1.4–19.1%%). Overall prevalence rates varied depending on the EA measurement tool. EA subjects were more likely to have lower levels of overall wellbeing (only in amateur competitive athletes), higher anxiety levels, and have greater frontal brain activity. Conclusions: EA is prevalent in the absence of indicated eating disorders across populations but varies depending on measurement tool. Further research is needed to explore EA without indicated eating disorders in different populations using homogenous measurement tools, further determine psychological correlates, and examine which measures of EA without indicated eating disorders predict poor health outcomes.