As no French validated measurement tool distinguishing healthy orthorexia (HeOr) from orthorexia nervosa (OrNe) currently exists, this study aimed at examining psychometric properties of the French version of the Teruel Orthorexia Scale (TOS). A sample of 799 participants (Mean [SD] age: 28.5 [12.1] years-old) completed the French versions of the TOS, the Düsseldorfer Orthorexia Skala, the Eating Disorder Examination-Questionnaire, and the Obsessive–Compulsive Inventory-Revised. Confirmatory factor analysis and exploratory structural equation modeling (ESEM) were used. Although the bidimensional model, with OrNe and HeOr, of the original 17-item version showed an adequate fit, we suggest excluding items 9 and 15. The bidimensional model for the shortened version provided a satisfactory fit (ESEM model: CFI = .963, TLI = .949, RMSEA = .068). The mean loading was .65 for HeOr and .70 for OrNe. The internal consistency of both dimensions was adequate (αHeOr = .83 and αOrNe = .81). Partial correlations showed that eating disorders and obsessive–compulsive symptomatology measures were positively related to OrNe and unrelated or negatively related to HeOr. The scores from the 15-item French version of the TOS in the current sample appears to present an adequate internal consistency, pattern of associations in line with what was theoretically expected, and promising for differentiating both types of orthorexia in a French population. We discuss why both dimensions of orthorexia should be considered in this area of research. Orthorexia ('right appetite', from the Greek) covers two dimensions: (1) orthorexia nervosa (OrNe), a strong preoccupation with healthy diet with negative emotional, cognitive, and/or social consequences while trying to approach this goal and when the eating behavior deviates from it, and (2) healthy orthorexia (HeOr), which can be defined as a healthy interest in diet, (self-assessed) healthy behavior with regard to diet and eating healthily as part of one’s identity. OrNe is not yet indexed into mental disorder classifications. Some prominent measurement tools in the area of orthorexia present important limitations: it is unclear if they validly assess OrNe and they do not tap HeOr by design. To overcome these limitations, a promising self-reported instrument was developed: the Teruel Orthorexia Scale (TOS), a bidimensional questionnaire whose structure has been replicated in different samples. Our research based on 799 participants aimed to adapt this instrument in a French speaking population. The results revealed that the French version of the TOS is an empirically supported tool allowing to differentiate both forms of orthorexia (healthy and nervosa). It also suggests that OrNe is associated with psychopathological symptoms while opposite patterns were found with HeOr. We discuss the importance of measuring both orthorexia dimensions.

Purpose Since the term orthorexia nervosa (ON) was coined from the Greek (omicron rho theta omicron sigma, right and omicron rho epsilon xi iota sigma, appetite) in 1997 to describe an obsession with "correct" eating, it has been used worldwide without a consistent definition. Although multiple authors have proposed diagnostic criteria, and many theoretical papers have been published, no consensus definition of ON exists, empirical primary evidence is limited, and ON is not a standardized diagnosis. These gaps prevent research to identify risk and protective factors, pathophysiology, functional consequences, and evidence-based therapeutic treatments. The aims of the current study are to categorize the common observations and presentations of ON pathology among experts in the eating disorder field, propose tentative diagnostic criteria, and consider which DSM chapter and category would be most appropriate for ON should it be included. Methods 47 eating disorder researchers and multidisciplinary treatment specialists from 14 different countries across four continents completed a three-phase modified Delphi process, with 75% agreement determined as the threshold for a statement to be included in the final consensus document. In phase I, participants were asked via online survey to agree or disagree with 67 statements about ON in four categories: A-Definition, Clinical Aspects, Duration; B-Consequences; C-Onset; D-Exclusion Criteria, and comment on their rationale. Responses were used to modify the statements which were then provided to the same participants for phase II, a second round of feedback, again in online survey form. Responses to phase II were used to modify and improve the statements for phase III, in which statements that met the predetermined 75% of agreement threshold were provided for review and commentary by all participants. Results 27 statements met or exceeded the consensus threshold and were compiled into proposed diagnostic criteria for ON. Conclusions This is the first time a standardized definition of ON has been developed from a worldwide, multidisciplinary cohort of experts. It represents a summary of observations, clinical expertise, and research findings from a wide base of knowledge. It may be used as a base for diagnosis, treatment protocols, and further research to answer the open questions that remain, particularly the functional consequences of ON and how it might be prevented or identified and intervened upon in its early stages. Although the participants encompass many countries and disciplines, further research will be needed to determine if these diagnostic criteria are applicable to the experience of ON in geographic areas not represented in the current expert panel.

Orthorexia nervosa (ON), defined as an excessive preoccupation with healthy eating, has gained more interest in the literature over these past few years. However, little is known about its risk and protective factors, in particular with regards to personality. A total of 3235 college students (10.32% men, 89.67% women) with a mean age of 21.13 (SD = 2.23) answered self-administered questionnaires assessing ON, psychopathological symptoms, and personality disorders including schizotypal, borderline, paranoid, obsessive–compulsive, and narcissistic personality. A subsample of 106 participants (91.51% women, mean age = 20.91, SD = 2.31) was selected based on the DOS cutoff score, and was then considered as the “orthorexic subsample”. Hierarchical cluster analysis was performed in the orthorexic subsample and led to the identification of four groups: 1—a cluster with a low level of traits (L); 2—a cluster with moderate traits and low narcissistic traits (MD); 3—a cluster with a low level of traits and moderate narcissistic traits (MN); 4—a cluster with high paranoid and narcissistic traits (PN) and a moderate level of schizotypal and borderline traits. Levels of anxiety, obsessional–compulsive, and depressive symptoms were higher in the PN and MD clusters than in the L and MN clusters. Social phobia was higher and self-esteem lower in the MD cluster and hypochondriasis was higher in the PN and MN clusters. This study suggests that ON can be associated with different personality profiles, some of them displaying significant psychopathological levels. It also emphasizes the importance of taking into account personality disorder traits of young adults with orthorexic eating behaviors. Descriptive (cross-sectional) study, Level V.

This study aimed at obtaining a French version of the DOS (F-DOS) and evaluating its psychometric properties in a sample of university students, then assessing the prevalence of orthorexic eating behavior among the participants. The F-DOS was obtained using back translation, and then administered to 3235 university students (10.32% men, 89.67% women) with a mean age of 21.13 (SD = 2.23). The Eating Habits Questionnaire and Eating Attitudes Test were used to assess convergent and divergent validities, respectively. Confirmatory factor analysis was performed to explore the factor structure. Ordinal ⍵ of F-DOS was 0.87, indicating very good internal consistency. F-DOS and EHQ total scores were strongly correlated (rs = 0.74, p < 0.001), indicating very good construct validity. Factor analysis revealed a well-fitted one-factor model. Regarding Orthorexia Nervosa (ON) prevalence, according to DOS cut-off score, 3.28% of participants could be considered having ON, while an additional 11.31% could be at risk of developing ON. The F-DOS appears to be a valid and reliable instrument to assess orthorexic eating behavior. ON prevalence in college students appeared similar to Germany and lower than in the US. Descriptive (cross-sectional) study, Level V.