Italian validation and exploratory measurement invariance of the food addiction symptom inventory (FASI)

Abstract Background The Food Addiction Symptom Inventory (FASI) is a clinician-administered semi-structured interview developed to assess addiction-like eating behaviors related to ultra-processed food consumption, but a validated Italian version is unavailable. This study aimed to validate the Italian version of the FASI and examine its psychometric properties and clinical correlates in a sample of patients with eating disorders (EDs). Methods One hundred and eighteen patients with anorexia nervosa-restrictive subtype and binge-eating spectrum disorder were recruited. Participants completed the FASI interview and questionnaires on eating and general disorder psychopathology (YFAS 2.0, EDE-Q, EBA-O, BES, BDI-II). Confirmatory factor analysis (CFA) tested the dimensional structure of the FASI and the measurement of invariance between diagnostic groups. Reliability and convergent validity were examined. Results CFA supported the one-factor structure of the FASI and indicated a good model fit (CFI=0.958, TLI=0.947, RMSEA=0.064 and SRMR=0.076) and internal consistency (categorical ω =0.84). Exploratory measurement invariance analyses provided preliminary support for configural and partial metric invariance across restrictive- and binge-spectrum ED groups. Correlations between total FASI symptom criteria and YFAS 2.0 symptom scores supported convergent validity. Moderate-to-strong associations emerged in the overall sample (r=0.581), with somewhat lower convergence observed in the AN group (r=0.408) than the binge-spectrum group (r=0.505). Conclusions The Italian version of the FASI demonstrates satisfactory psychometric properties, with exploratory analyses providing preliminary support for configural and partial metric invariance across ED diagnostic groups. Findings in the restrictive-spectrum group highlighted potential challenges in distinguishing addiction-like intake from restraint-driven subjective overeating, warranting particular clinical attention. Future studies with larger and more diverse samples should further evaluate measurement invariance, inter-rater and test–retest reliability, and predictive validity for clinical outcomes. Level of Evidence III: Evidence obtained from well-designed cohort or case–control analytic studies.