A COSMIN-Based Systematic Review of the Measurement Properties of Cross-Cultural Versions of the Orthorexia Nervosa Inventory (ONI)

Abstract

Background Orthorexia nervosa (ON) has received increasing attention, and the Orthorexia Nervosa Inventory (ONI) is among the most widely used instruments to assess ON. As ON research expands globally, multiple countries have translated and culturally adapted the ONI. However, the methodological quality of these adaptations and the psychometric evidence of the translated versions have not been systematically synthesized. Guided by the COSMIN framework, this review evaluated the measurement quality of cross-culturally adapted ONI versions to inform standardized use in research and practice. Methods This systematic review followed COSMIN and PRISMA guidance and was registered in PROSPERO (CRD420251218905). PubMed, Scopus, Web of Science, Embase, and CNKI were searched to November 2025 using “Orthorexia Nervosa Inventory” and related terms. We included peer-reviewed English or Chinese original studies reporting ONI translation/cultural adaptation and psychometric properties. Two reviewers independently screened studies and extracted data. Translation/adaptation procedures were appraised using Beaton et al.’s guidelines. Measurement properties were evaluated using Terwee et al.’s criteria and the COSMIN Risk of Bias checklist. Findings were narratively synthesized. Results From 280 records, five studies met inclusion criteria, covering four countries and five language versions. Translation and adaptation procedures varied substantially; only one Turkish version fully reported all five Beaton steps. Most studies reported forward–backward translation and synthesis, whereas expert committee review and pretesting were each reported in ~40%. COSMIN ratings indicated ~60% of studies were “very good” for structural validity; EFA and/or CFA generally supported the intended factor structure. All studies reported acceptable internal consistency. Only the Chinese version reported test–retest reliability and measurement error/minimal detectable change, although measurement error was mainly estimated via SEM derived from α without interpretation against minimal important change. Construct validity evidence was limited and often rated “adequate” or “doubtful.” No study assessed cross-cultural validity/measurement invariance, responsiveness, or ROC-based cut-offs. Conclusion Cross-cultural ONI versions show generally acceptable internal consistency and preliminary structural validity, but major gaps remain in translation rigor, reproducibility, measurement error, construct validity, and especially cross-cultural validity and responsiveness. Future studies should adhere to established adaptation standards, test measurement invariance, and comprehensively evaluate reliability, measurement error, responsiveness, and clinically meaningful cut-offs.