The Dimensionality of the Multidimensional Fatigue Inventory (MFI-20) Derived From Healthy Adults and Patient Subpopulations: A Challenge for Clinicians

Introduction: Fatigue is associated with various diseases of different origins extending from immune disorders to cancer. The purpose of this study was to explore the psychometric properties/dimensionality of the Multidimensional Fatigue Inventory (MFI-20) questionnaire in samples of healthy adults and patient subgroups in Greece. Methods: This was a multicenter cross-sectional study that included 80 women with breast cancer, 108 patients with multiple sclerosis (MS), 100 people with thalassemia diagnosis, and 185 healthy adults, aged 18-45 years. All patients were adults. Patients were recruited from a breast surgery clinic, a neurological clinic, and a thalassemia unit, while healthy adults were recruited from the University of West Attica students and personnel. The MFI-20, the modified fatigue impact scale (MFIS), the fatigue severity scale (FSS), and the Hamilton anxiety-depression scale (HANDS) were used. Internal consistency, repeatability, test-retest reliability, construct, and convergent validity were investigated. Results: MFI-20 exhibited excellent reliability properties (internal consistency: Cronbach’s alpha MFI-20 subscales ranged from 0.83 to 0.94; repeatability: Pearson’s r = 0.335 [p < 0.001]). Significant correlations were found between MFI-20 and MFIS: Pearson’s r = 0.870, FSS: Pearson’s r = 0.582 - 0.335, and HANDS: Pearson’s r = 0.734 - 0.442 (all p-values < 0.0001) on all subsamples. However, its dimensionality is questionable depending on the subpopulation tested, and the one-dimension perspective is possible. MS patients exhibited the highest total score (55.26 ± 16.53), while thalassemia patients exhibited the lowest score (45.09 ± 13.54). In all subscales, thalassemia patients differed statistically significantly from the MS patients (p < 0.01), while in the reduced activity subscale, thalassemia patients differed significantly from all other groups (p < 0.01). Conclusions: As strict fatigue subscale classification is questionable, the use of MFI-20 total score is suggested for the assessment of fatigue in clinical populations. As MFI-20 is a very useful research tool for studying fatigue, the use of the total and/or partial scores depends on the clinical population. Total score instead of (or additionally) partial scores is suggested in clinical practice.