Psychometric evaluation of the Multidimensional Fatigue Inventory (MFI) in Portuguese adult samples: a consensus version

Background: Fatigue is a multidimensional symptom common across various conditions, often underdiagnosed despite its significant impact on quality of life. The Multidimensional Fatigue Inventory (MFI) is a self-assessment scale designed to measure multiple fatigue dimensions. However, its factor structure has proven difficult to replicate. Objective: This study evaluates the psychometric properties of the MFI in a Portuguese sample, proposing a refined version and contributing to the debate on its structural validity. Methods: A total of 1,858 participants (aged 18–85, M = 30.72 ± 13.9; 69.5% female) were recruited through non-probabilistic sampling and split into two subsamples. Six factor structures were tested using confirmatory factor analysis (CFA); the three best-fitting models underwent measurement invariance testing across subsamples and sex. The best model was further validated in the full sample, with a second-order factor (Global Fatigue–GlobF) tested. Results: Three models were discarded due to poor fit. The retained models showed strong measurement invariance and good psychometric properties. The final model preserved the original Physical Fatigue and Mental Fatigue dimensions. CFA fit indices were strong (RMSEA = 0.076, SRMR = 0.028, CFI = 0.971, TLI = 0.957). Reliability was high (α = 0.798–0.881; ω = 0.807–0.880), with adequate construct validity (AVE = 0.519–0.648) and discriminant validity (HTMT = 0.589). Conclusions: This study proposes a refined, reliable and valid consensus version of the MFI for the Portuguese population. The revised instrument improves fatigue assessment and may support better fatigue identification and management in clinical and non-clinical settings. Further research should examine its responsiveness and cross-cultural applicability.