ABSTRACT
OBJECTIVE: There is disagreement about the conceptualization and measurement of pathological exercise (PE). This study seeks to elucidate the nature of this phenomenon as addictive or compulsive in its primary and secondary forms. METHOD: 1,497 adults (608 men, 885 women, 4 other) completed a set of validated surveys, including the Godin Leisure-Time Exercise Questionnaire, the Exercise Dependence Scale-21, the Exercise Addiction Inventory, the Compulsive Exercise Test (CET), the Obligatory Exercise Questionnaire, the Commitment to Exercise Scale, and an exercise specific adaptation of the Dimensional Obsessive-Compulsive Scale (ESDOCS). Participants completed the Eating Disorders Examination Questionnaire, and DSM-5 eating disorder (ED) diagnoses were determined according to published guidelines. Pearson correlation coefficients, independent samples t-tests, χ(2) tests, 1-way ANOVA's, and ANCOVA's were calculated. RESULTS: All PE measures correlated highly with one another (P < 0.001). Participants with EDs scored higher on all measures than those without (P < 0.001), and those with bulimia nervosa had the highest scores. Participants with PE and an accompanying ED (secondary PE) scored higher than those with PE and no ED symptoms (primary ED) on the CET (P < 0.002) and ESDOCS (P < 0.003). Correlations between EDE-Q and PE scores were stronger among women than men. PE prevalence was 6.4% (1.4% primary PE, 5.0% secondary PE) in our heterogeneous sample. DISCUSSION: Secondary PE appears to be more compulsive while primary PE is more addictive in nature. Men and women are equally at risk for PE, but in men it is more often primary and addictive and in women it is more often secondary and compulsive. © 2016 Wiley Periodicals, Inc. (Int J Eat Disord 2016; 49:778-792).