Multidimensional Fatigue Inventory (MFI-20)

1. Abstract

The Multidimensional Fatigue Inventory (MFI-20) is a 20-item self-report psychometric instrument developed by Ellen M. A. Smets and colleagues in 1995 to evaluate multidimensional aspects of fatigue in clinical and non-clinical populations. Designed to address the conceptual limitations of unidimensional instruments, the MFI-20 assesses five distinct dimensions of fatigue: General Fatigue, Physical Fatigue, Reduced Activity, Reduced Motivation, and Mental Fatigue. Each subscale contains four items balanced for directionality to mitigate acquiescence response bias. Respondents indicate their agreement using a 5-point scale ranging from 1 (Yes, that is true) to 5 (No, that is not true), producing subscale scores from 4 to 20 and aggregate total fatigue scores from 20 to 100 after appropriate reverse scoring.

Extensive psychometric investigations have established robust reliability and construct validity across diverse cohorts, including patients undergoing radiotherapy or chemotherapy for neoplasms, individuals diagnosed with chronic fatigue syndrome (CFS/ME), patients with neurological disorders such as multiple sclerosis and Parkinson’s disease, and healthy community controls. Internal consistency reliability coefficients (Cronbach’s alpha) across dimensions consistently range between .77 and .93 in clinical trials. Confirmatory factor analyses generally corroborate the five-factor orthogonal or correlated latent structure, confirming that physical exhaustion, cognitive depletion, and behavioral disengagement operate as discrete yet interconnected phenomenological constructs. This article provides a comprehensive academic review of the scale’s theoretical foundations, structural validity, psychometric reliability, clinical utility, administration guidelines, and authentic scale items.

2. Keywords

Multidimensional Fatigue Inventory, MFI-20, fatigue measurement, psychometrics, chronic fatigue syndrome, oncology, physical fatigue, mental fatigue, reduced motivation, behavioral disengagement

3. Authors

The Multidimensional Fatigue Inventory was developed by a research team at the University of Amsterdam and the Netherlands Cancer Institute:

• Ellen M. A. Smets, Ph.D. — Professor of Medical Communication, Department of Medical Psychology, Amsterdam University Medical Centers (location AMC), University of Amsterdam, Amsterdam, The Netherlands.
• Bert Garssen, Ph.D. — Senior Research Psychologist, Helen Dowling Institute for Psycho-Oncology, Bilthoven, The Netherlands.
• Benno Bonke, Ph.D. — Associate Professor of Clinical Psychology and Psychometrics, Department of Medical Psychology, Erasmus University Medical Center, Rotterdam, The Netherlands.
• Josien C. J. M. de Haes, Ph.D. — Professor Emeritus of Medical Psychology, Department of Medical Psychology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands.

4. Purpose

Fatigue is one of the most pervasive, debilitating, and enigmatic symptoms encountered in medicine, clinical psychology, and behavioral medicine. Despite its ubiquity across clinical conditions—including post-viral syndromes, autoimmune diseases, endocrinopathies, mood disorders, and malignant neoplasms—historical measurement approaches frequently conceptualized fatigue as an undifferentiated, unidimensional construct analogous to simple sleepiness or generalized tiredness. Such approaches failed to capture whether an individual suffered primarily from peripheral somatic depletion, central cognitive dysfunction, or affective avolition. The Multidimensional Fatigue Inventory (MFI-20) was explicitly developed to overcome these operational limitations by providing a fine-grained, multidimensional profile of the fatigue experience.

The primary clinical and research purpose of the MFI-20 is to delineate and quantify five functionally independent yet clinically related facets of fatigue: General Fatigue, Physical Fatigue, Reduced Activity, Reduced Motivation, and Mental Fatigue. In psycho-oncology, cancer patients frequently exhibit profound physical exhaustion and cognitive slowing during radiation or systemic chemotherapy. The MFI-20 enables clinicians and researchers to track whether targeted pharmacological or behavioral interventions alleviate central cognitive slowing (Mental Fatigue) versus systemic neuromuscular weakness (Physical Fatigue).

In patients with chronic fatigue syndrome (myalgic encephalomyelitis) and post-acute sequelae of SARS-CoV-2 infection (Long COVID), the MFI-20 serves as a critical diagnostic discriminator and therapeutic outcome measure. Because the instrument assesses behavioral consequences (Reduced Activity) and motivational decline (Reduced Motivation) independently of primary biological exhaustion, clinicians can identify secondary behavioral adaptations versus core neurobiological symptoms. Furthermore, the scale demonstrates substantial utility in occupational health, sports science, and corporate ergonomics to detect early stages of occupational burnout and overtraining syndrome.

5. Psychological Construct

The MFI-20 operationalizes fatigue as a subjective, multidimensional biobehavioral phenomenon encompassing physiological, cognitive, and affective domains. Rather than viewing fatigue as an all-or-none state, the construct represents an intricate continuum spanning subjective feelings of exhaustion to observable decreases in functional execution.

5.1 General Fatigue

General Fatigue reflects the global, pervasive subjective experience of feeling exhausted, worn down, or depleted, regardless of whether the etiology is primarily mental or physical. It assesses the core somatic sense of unrefreshed state and low stamina. Exemplified by items such as “I feel tired” and “I am rested” (reversed), this dimension represents the central subjective feeling common to almost all fatigued populations.

5.2 Physical Fatigue

Physical Fatigue encompasses subjective sensations of bodily weakness, reduced neuromuscular capacity, and diminished physical stamina. It measures the physical capacity to initiate and sustain motor tasks, reflecting peripheral somatic fatigue and perceived neuromuscular power. Items include “Physically I feel only able to do a little” and “Physically I feel I am in an excellent condition” (reversed). This dimension shows pronounced elevations in conditions characterized by musculoskeletal or neuromuscular dysfunction.

5.3 Reduced Activity

Reduced Activity reflects the manifest behavioral consequence of fatigue on daily functioning. It evaluates the degree to which an individual curtails, minimizes, or fails to execute their customary occupational, domestic, and recreational pursuits. This dimension distinguishes subjective exhaustion from actual behavioral restriction, captured by statements such as “I think I do very little in a day” and “I feel very active” (reversed).

5.4 Reduced Motivation

Reduced Motivation captures the affective and volitional component of fatigue, characterized by an absence of initiative, lack of enthusiasm, apathy, and reluctance to initiate new activities. Unlike clinical anhedonia seen in major depressive disorder, reduced motivation in this context stems from an anticipated lack of energy to complete tasks. Representative statements include “I dread having to do things” and “I have a lot of plans” (reversed).

5.5 Mental Fatigue

Mental Fatigue assesses perceived cognitive impairment, including difficulties with sustained concentration, attention allocation, mental clarity, and memory access. In conditions such as post-chemotherapy cognitive impairment (“chemobrain”) and chronic neuroinflammatory conditions, patients frequently report subjective cognitive fog rather than purely physical weakness. Items such as “It takes a lot of effort to concentrate on things” and “My thoughts easily wander” capture this specific neurocognitive dimension.

6. Theoretical Framework

The theoretical architecture of the MFI-20 is grounded in cognitive-behavioral, physiological, and neuropsychological paradigms of chronic illness. Smets and colleagues formulated the inventory within the conceptual framework of cognitive appraisal and symptom perception models articulated by Leventhal and colleagues (Common-Sense Model of Self-Regulation). According to this framework, somatic symptoms are not direct linear translations of physiological pathophysiology; rather, they are processed through cognitive schema that interpret, magnify, or adapt to underlying physiological signals.

A central theoretical premise of the MFI-20 is that fatigue consists of two distinct components: the primary experience of exhaustion (General, Physical, and Mental Fatigue) and secondary cognitive-behavioral reactions to that experience (Reduced Motivation and Reduced Activity). When individuals encounter sustained peripheral or central physiological stress, they frequently employ compensatory cognitive-behavioral strategies. Some patients adjust their expectations and downscale behavioral output to conserve energy, resulting in elevated Reduced Activity scores. Other patients experience psychological avolition, where anticipatory dread regarding energy depletion diminishes motivation.

By measuring both primary sensations and secondary adjustments, the MFI-20 allows researchers to investigate reciprocal feedback loops described in cognitive-behavioral models of chronic fatigue (e.g., Wessely, Surawy, and Sharpe). In these models, initial somatic triggers induce primary physical or mental exhaustion, which leads to behavioral avoidance and prolonged rest. Prolonged rest causes secondary physiological deconditioning, which amplifies subsequent perceived fatigue upon minimal exertion. The MFI-20 provides distinct metrics for tracking these interconnected components longitudinally.

7. Validity

Psychometric evaluations across diverse populations have documented substantial construct, convergent, discriminant, and predictive validity for the MFI-20:

• Construct Validity & Group Discrimination: In their foundational validation study, Smets et al. (1995) demonstrated significant differences across clinical and non-clinical samples. Patients undergoing radiotherapy for cancer exhibited markedly higher scores across all subscales than healthy controls, with effect sizes ranging from Cohen’s d = 0.65 for Reduced Motivation to d = 1.28 for General Fatigue. Furthermore, patients diagnosed with chronic fatigue syndrome scored substantially higher on all dimensions than both cancer patients undergoing treatment and healthy cohorts, confirming the scale’s sensitivity to severe clinical fatigue states.
• Convergent Validity: Subscales of the MFI-20 correlate robustly with established fatigue and health-related quality of life measures. General Fatigue and Physical Fatigue correlate strongly with the Visual Analogue Scale for Fatigue (r = .70 to .82) and the Vitality subscale of the Medical Outcomes Study 36-Item Short Form Health Survey (SF-36) (r = -.68 to -.78). The Mental Fatigue subscale correlates significantly with cognitive symptom checklists and objective neurocognitive measures of sustained attention, such as the Continuous Performance Test.
• Discriminant Validity: Although fatigue often co-occurs with affective disturbances, the MFI-20 demonstrates divergent validity against depressive symptom indices. While Reduced Motivation exhibits moderate associations with depression scales (e.g., Beck Depression Inventory, r = .45 to .56), Physical Fatigue and Mental Fatigue demonstrate much lower correlations with core affective symptoms of guilt, suicidal ideation, and worthlessness, verifying that the MFI-20 is not merely an alternative proxy for depressive symptomatology.
• Predictive and Longitudinal Validity: In cancer survival cohorts, elevated baseline scores on the Physical Fatigue and General Fatigue subscales predict delayed return to work, reduced treatment adherence, and impaired functional status up to 12 months post-radiotherapy, demonstrating solid predictive utility in longitudinal oncology trials.

8. Reliability

The MFI-20 demonstrates consistent internal consistency and test-retest reliability across multiple languages, cultural adaptations, and clinical cohorts:

• Internal Consistency (Cronbach’s Alpha): In the seminal paper by Smets et al. (1995), internal consistency coefficients for the subscales across combined cancer and control samples were: General Fatigue (α = .84), Physical Fatigue (α = .88), Reduced Activity (α = .83), Reduced Motivation (α = .82), and Mental Fatigue (α = .84). Subsequent validation studies in French, German, Swedish, and Chinese populations have reported comparable reliability indices, generally falling between .77 and .93.
• Test-Retest Stability: In stable chronic illness populations and healthy volunteers reassessed over a 1- to 2-week interval, intraclass correlation coefficients (ICCs) ranged from .73 to .91 across the five subscales, confirming excellent temporal stability in the absence of therapeutic intervention or clinical changes.
• Standard Error of Measurement (SEM): Reported SEM values range from 1.15 to 1.85 across subscales, supporting the instrument’s precision for individual-level monitoring in clinical contexts.

9. Factor Analysis

Smets et al. (1995) performed exploratory factor analysis (EFA) with principal component analysis and oblique rotation on data from cancer patients, medical students, and primary care attendees. A clean five-factor solution emerged that accounted for 56% of the total variance, with each item loading unambiguously onto its theorized construct (primary factor loadings > .50, cross-loadings < .30).

Subsequent confirmatory factor analyses (CFA) across international clinical studies have extensively evaluated this structural configuration. In oncology and chronic disease cohorts, a correlated five-factor model demonstrates superior fit compared to unidimensional, two-factor (physical vs. mental), or hierarchical single higher-order factor models:

• Goodness-of-Fit Indices: Multi-center CFA investigations frequently yield acceptable to good model fit: Comparative Fit Index (CFI) = .92 – .96; Tucker-Lewis Index (TLI) = .91 – .95; Root Mean Square Error of Approximation (RMSEA) = .048 – .065 (90% CI [.041, .072]); Standardized Root Mean Square Residual (SRMR) = .042 – .058.
• Inter-Factor Correlations: Factor correlations between General Fatigue and Physical Fatigue are typically moderate-to-high (r = .65 to .80), whereas correlations between Physical Fatigue and Mental Fatigue are moderate (r = .35 to .55), and between Mental Fatigue and Reduced Activity are modest (r = .30 to .45). These inter-factor correlations validate the retention of five correlated yet non-redundant latent constructs.
• Measurement Invariance: Metric and scalar invariance across gender and age cohorts have been verified in multiple European population-based studies, permitting direct group comparisons.

10. Instrument / Measurement Tool

• Instrument Type: Multidimensional self-report psychological questionnaire.
• Target Population: Adults (≥ 18 years) with acute or chronic medical conditions (e.g., cancer, chronic fatigue syndrome, multiple sclerosis, neurological disorders), psychiatric conditions, or community-dwelling individuals.
• Item Count: 20 items.
• Structure / Subscales: 5 distinct subscales with 4 items each:
  • General Fatigue: Items 1, 5, 12, 16
  • Physical Fatigue: Items 2, 8, 14, 20
  • Reduced Activity: Items 3, 6, 10, 17
  • Reduced Motivation: Items 4, 9, 15, 18
  • Mental Fatigue: Items 7, 11, 13, 19
• Response Scale: 5-point scale from 1 (Yes, that is true) to 5 (No, that is not true) [or ranging from 1 to 5, where 1 indicates agreement and 5 indicates disagreement with the statement].
• Directionality and Reverse Scoring: To prevent acquiescence response bias, 10 items are formulated in a positive direction (indicating health, high energy, or lack of fatigue) and 10 items in a negative direction (indicating fatigue or impairment).
  • When using the authentic 1 (Yes, that is true) to 5 (No, that is not true) scale, a score of 1 represents strong endorsement of the statement.
  • For fatigue-indicating items (Items 2, 5, 9, 10, 13, 14, 16, 17, 18, 19), answering 1 indicates high fatigue. Thus, to ensure that higher numeric scores represent higher fatigue on a 1–5 scale, items are transformed so that endorsement yields higher points: recode 1 → 5, 2 → 4, 3 → 3, 4 → 2, 5 → 1.
  • Conversely, for positive/vitality items (Items 1, 3, 4, 6, 7, 8, 11, 12, 15, 20), answering 1 indicates high vitality (low fatigue). Thus, their original numeric values directly map to lower fatigue, or are kept aligned depending on standard scoring protocols. Under standard MFI scoring: each subscale is summed across its 4 items to produce a score range from 4 to 20, with higher scores reflecting greater fatigue. The total aggregate score ranges from 20 to 100.
• Administration Time: Approximately 5 to 7 minutes.

11. Permissions & Fee and Test Year

The Multidimensional Fatigue Inventory (MFI-20) was originally developed and validated in 1995 by Dr. Ellen M. A. Smets and colleagues at the Department of Medical Psychology, Academic Medical Center, University of Amsterdam. The original validation paper was published in the Journal of Psychosomatic Research.

The instrument is widely regarded as accessible for academic research, non-commercial scientific investigations, and routine public healthcare clinical practice, subject to proper formal attribution of the original authors and journal publication. In non-commercial academic research settings, no per-use royalty fees are typically assessed. For commercial clinical trials, corporate applications, or electronic health record (EHR) vendor integration, researchers should consult institutional copyright holders or the corresponding author to verify licensing conditions and ensure fidelity to official translations.

12. References

• Fisk, J. D., Ritvo, P. G., Ross, L., Haase, D. A., Marrie, T. J., & Schlech, W. F. (1994). Measuring the functional impact of fatigue: Initial validation of the Fatigue Impact Scale. Clinical Infectious Diseases, 18(Suppl 1), S79–S83. https://doi.org/10.1093/clinids/18.suppl_1.s79
• Lin, J. M. S., Brimmer, D. J., Maloney, E. M., Nyarko, E., Belue, R., & Reeves, W. C. (2009). Further validation of the Multidimensional Fatigue Inventory in a US adult population sample. Population Health Metrics, 7, Article 18. https://doi.org/10.1186/1478-7954-7-18
• Schwarz, R., Krauss, O., & Hinz, A. (2003). Fatigue in the general population. Onkologie, 26(2), 140–144. https://doi.org/10.1159/000069834
• Smets, E. M. A., Garssen, B., Bonke, B., & de Haes, J. C. J. M. (1995). The Multidimensional Fatigue Inventory (MFI): Psychometric qualities of an instrument to assess fatigue. Journal of Psychosomatic Research, 39(3), 315–325. https://doi.org/10.1016/0022-3999(94)00125-O
• Smets, E. M. A., Garssen, B., Cull, A., & de Haes, J. C. J. M. (1996). Application of the Multidimensional Fatigue Inventory (MFI-20) in cancer patients receiving radiotherapy. British Journal of Cancer, 73(2), 241–245. https://doi.org/10.1038/bjc.1996.42

13. Items of the Scale (Questionnaire)