Orthorexia nervosa and disordered eating attitudes among Lebanese adults: Assessing psychometric proprieties of the ORTO-R in a population-based sample.

backgroundPrevious research demonstrated a relationship between ON and disordered eating symptoms (eating concern, restraint, cognitive preoccupations about body shape and weight) and disordered eating attitudes (DEA). Since screening for orthorexia nervosa is now part of clinical practice, the measurement instruments to be used must be clinically significant, reliable, valid and sensitive to our target population. The main objective of the present study was to confirm the factor structure of the Arabic version of the ORTO-R using a first sample of Lebanese adults and confirm those results on another sample. The secondary objective was to assess sex differences in terms of ON and DEA, as well as to examine whether symptoms of ON were related to DEA in Lebanese adults.

methodsA total of 783 Lebanese adults was selected to participate in this cross-sectional study (January-May 2018) using a proportionate random sample from all Lebanese governorates.

resultsThe mean age of the total sample was 27.78 ± 11.60 years (Min. 18 -Max. 84) (33.5% females) and their mean BMI was 24.36 ± 5.31 kg/m2. All items of the ORTO-R were extracted during the factor analysis and yielded a two-factor solution with Eigenvalues > 1 (variance explained  =  50.07%; KMO = 0.570; Bartlett's sphericity test p<0.001; αCronbach = 0.755). This factor structure was confirmed by a confirmatory factor analysis; the Maximum Likelihood Chi-Square  =  26.894 and Degrees of Freedom  =  8, which gave a χ2/df  =  3.36. The Tucker Lewis Index (TLI) value was 0.914, whereas the standardized root mean square residual (SRMR) value was 0.032. The root mean square error of approximation (RMSEA) value was 0.077 [95% CI 0.046-0.111] (pclose = 0.07) and comparative fit index (CFI) value was 0.967 respectively, indicating a good fit of the model. There was also no measurement invariance between genders. Female gender was significantly associated with lower ORTO-R scores (more orthorexia nervosa) compared to males (B = -0.65; p = 0.026, 95% CI -1.22- -0.08; ɳ2 = 0.006). However, no significant difference was found between genders in terms of EAT-26 scores (B = 0.23; p = 0.813, 95% CI -1.66-2.12; ɳ2 = 0.0001). Higher ORTO-R scores (lower/ less pathological ON tendencies and behaviors) were significantly related to higher EAT-26 total scores (higher levels of DEA) as well as higher dieting, bulimia and oral control scores in both females (from a weak to a moderate positive correlation) and males (a weak positive correlation).

conclusionThis cross-sectional population-based study confirmed the factor structure of the Arabic version of the ORTO-R, demonstrated an association between ON and DEA and revealed more ON among females compared to males.