Étude confirmative de l’échelle de dépendance à l’activité physique Exercise Dependence Scale-Revised pour une population francophone

Abstract

Sur la base des critères diagnostiques de dépendance aux substances du DSM-IV [American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994.], Hausenblas et Downs [Exercise dependence: a systematic review. Psychol Sport Exerc 2002;3(2):89–123.] développent pour une population anglophone, une échelle d’évaluation des symptômes de dépendance à l’activité physique : l’Exercise Dependence Scale-Revised (EDS-R). Cette échelle est composée de sept facteurs. L’objectif de cette étude est de confirmer la présence de ces facteurs pour une version française de l’EDS-R [Factorial validity and psychometric examination of the exercise dependence scale-revised. Meas Phys Educ Exerc Sci 2004;8(4):183–201.]. La population étudiée est composée de 516 participants, engagés à un semi-marathon (moyenne d’âge égale à 39,02). L’analyse en composante principale réalisée sur la version française de l’EDS-R présente un nombre de facteurs (six facteurs) différent de celui de la version originale (sept facteurs). Cependant, les indices d’ajustement obtenus par une analyse factorielle confirmatoire montre que le modèle original demeure le plus pertinent. Les alphas de Cronbach des sept facteurs sont bons et présentent la version française de l’EDS-R comme une mesure valide de la dépendance à l’activité physique pour une population francophone. Since the first writings on excessive exercise, there has been an increased interest in exercise dependence. One of the major consequences of this increased interest has been the development of several definitions and measures of exercise dependence. The work of Veale [Does primary exercise dependence really exist? In: Annet J, Cripps B, Steinberg H, editors. Exercise addiction: Motivation for participation in sport and exercise.Leicester, UK: Br Psychol Soc; 1995. p. 1–5.] provides an advance for the definition and measure of exercise dependence. These studies have adapted the DSM-IV criteria for substance dependence to measure exercise dependence. The Exercise Dependence Scale-Revised is based on these diagnostic criteria, which are: tolerance; withdrawal effects; intention effect; lack of control; time; reductions in other activities; continuance. Confirmatory factor analyses of EDS-R provided support to present a measurement model (21 items loaded in seven factors) of EDS-R (Comparative Fit Index = 0.97; Root mean Square Error of Approximation = 0.05; Tucker-Lewis Index = 0.96). The aim of this study was to examine the psychometric properties of a French version of the EDS-R [Factorial validity and psychometric examination of the exercise dependence scale-revised. Meas Phys Educ Exerc Sci 2004;8(4):183–201.] to test the stability of the seven-factor model of the original version with a French population. A total of 516 half-marathoners ranged in age from 17 to 74 years old (Mean age = 39.02 years, ET = 10.64), with 402 men (77.9%) and 114 women (22.1%) participated in the study. The principal component analysis results in a six-factor structure, which accounts for 68.60% of the total variance. Because principal component analysis presents a six-factor structure differing from the original seven-factor structure, two models were tested, using confirmatory factor analysis. The first model is the seven-factor model of the original version of the EDS-R and the second is the model produced by the principal component analysis. The results of confirmatory factor analysis presented the original model (with a seven-factor structure) as a good model and fit indices were good (X2/ddl = 2.89, Root Mean Square Error of Approximation (RMSEA) = 0.061, Expected Cross Validation Index (ECVI) = 1.20, Goodness-of-Fit Index (GFI) = 0.92, Comparative Fit Index (CFI) = 0.94, Standardized Root Mean Square (SRMS) = 0.048). These results showed that the French version of EDS-R has an identical factor structure to the original. Therefore, the French version of EDS-R was an acceptable scale to measure exercise dependence and can be used on a French population.