Fatigue Severity Scale (FSS)

1. Abstract

The Fatigue Severity Scale (FSS) is one of the most widely adopted psychometric instruments designed to evaluate the behavioral, physical, and functional consequences of fatigue across a broad spectrum of clinical and research populations. Developed originally by Lauren B. Krupp and colleagues in 1989 to evaluate individuals diagnosed with multiple sclerosis (MS) and systemic lupus erythematosus (SLE), the FSS has since achieved international prominence across rheumatology, oncology, psychiatry, and post-viral conditions such as Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) and Post-COVID-19 Condition (Long COVID). The instrument comprises nine items scored on a 7-point Likert-type scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). Structurally, the FSS was conceptualized as a unidimensional measurement tool that quantifies how fatigue interferes with motivation, exercise capacity, physical functioning, professional and family duties, and social engagement. Over three decades of psychometric investigation demonstrate high internal consistency (Cronbach’s alpha typically ranging between 0.81 and 0.93), robust test-retest reliability ($r = 0.84$ to $0.89$), and notable sensitivity to clinical change and treatment interventions. Despite ongoing psychometric debate regarding its latent factor structure—specifically whether it operates as a strictly unidimensional scale or exhibits multidimensionality across physical functioning versus global impact—the FSS remains a gold standard in clinical research due to its conciseness, ease of administration, and clear discriminant ability in distinguishing pathological fatigue from normal physiological tiredness.

2. Keywords

Fatigue Severity Scale, FSS, psychometric validation, chronic fatigue, multiple sclerosis, systemic lupus erythematosus, physical impairment, clinical assessment, reliability, factor analysis, internal consistency

3. Authors

The Fatigue Severity Scale was developed and introduced into the scientific literature by a multidisciplinary team of clinical researchers and neurologists:

• Lauren B. Krupp, M.D. — Professor of Neurology, Department of Neurology, Grossman School of Medicine, New York University (NYU Langone Health), New York, NY, USA. Former Director of the Lourie Center for Pediatric Multiple Sclerosis at Stony Brook University.
• Nicholas G. LaRocca, Ph.D. — Clinical Neuropsychologist and former Vice President of Health Care Delivery and Policy Research at the National Multiple Sclerosis Society, New York, NY, USA.
• Judi Muir-Nash, B.S. — Department of Neurology, State University of New York at Stony Brook, Stony Brook, NY, USA.
• Alfred D. Steinberg, M.D. — Cellular Immunology Section, National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS), National Institutes of Health (NIH), Bethesda, MD, USA.

4. Purpose

Fatigue is a ubiquitous, complex, and profoundly debilitating human experience that manifests across nearly every category of medical, neurological, and psychological illness. However, despite its high prevalence and severe disruption of quality of life, fatigue is historically notoriously challenging to quantify, operationalize, and treat. The primary purpose of the Fatigue Severity Scale (FSS) is to provide a brief, psychometrically sound, self-administered measurement tool capable of quantifying the overall severity of fatigue and, crucially, evaluating the extent to which fatigue impairs daily functioning, physical execution, and psychosocial roles.

Prior to the establishment of the FSS in 1989, clinical investigations relied heavily on broad visual analogue scales (VAS) or diffuse vitality subscales embedded within generic health-related quality of life measures, such as the Medical Outcomes Study Short Form (SF-36). While visual analogue scales capture the immediate subjective intensity of fatigue, they fail to delineate the multifaceted consequences fatigue exerts on behavioral functioning, sustained physical activity, or social obligations. Conversely, broader health status measures often conflated fatigue with depressive symptomatology, generalized somatization, or cognitive slowing. The theoretical and clinical rationale behind the creation of the FSS was therefore to engineer a targeted instrument that disentangled fatigue-induced disability from concurrent psychiatric distress, enabling clinicians to discern pathological fatigue specific to underlying organic pathologies from normal, everyday weariness.

In clinical practice, the FSS serves several distinct functions:

• Screening and Diagnostic Differentiation: It provides empirical cutoff thresholds that differentiate normal, baseline physiological exhaustion from severe, clinically meaningful fatigue requiring therapeutic attention.
• Longitudinal Monitoring: The scale enables clinicians to track disease trajectory, identifying gradual worsening or acute relapses in neurodegenerative and autoimmune diseases such as multiple sclerosis and systemic lupus erythematosus.
• Outcome Assessment in Interventional Trials: The FSS serves as a primary or secondary endpoint in clinical trials evaluating pharmacological agents (such as modafinil, amantadine, or biologic disease-modifying therapies) and behavioral interventions (such as cognitive-behavioral therapy, graded exercise therapy, or energy conservation education).
• Multidisciplinary Care Planning: By delineating specific areas of interference—such as motivation versus physical functioning versus social participation—the FSS assists physical therapists, occupational therapists, and neuropsychologists in tailoring specific adaptive strategies to patient needs.

5. Psychological Construct

The construct measured by the Fatigue Severity Scale is multidimensional in its impact yet focused primarily on behavioral, physical, and functional impairment secondary to fatigue. In psychometric literature, fatigue is defined as an overwhelming, persistent sense of exhaustion, tiredness, and depleted physical and mental energy that is disproportionate to exerted effort and not alleviated by typical periods of rest or sleep. The FSS specifically conceptualizes fatigue not merely as a passive sensory perception (such as sleepiness or lethargy), but as an active, intrusive disruptor of human performance.

Although the original authors intended the FSS to yield a single composite score reflecting overall fatigue severity, an analysis of the constituent items reveals distinct facets of the broader fatigue-related disability construct:

Physical Functioning and Sustained Capacity

Items within this dimension (notably Items 4, 6, and 2) assess the physiological boundary conditions imposed by fatigue on somatic output. Item 4 (“Fatigue interferes with my physical functioning”) captures acute physical restriction, reflecting how the sensation of exhaustion directly blocks the execution of standard somatic activities such as ambulation, climbing stairs, or carrying groceries. Item 6 (“My fatigue prevents sustained physical functioning”) captures the temporal breakdown of somatic endurance; patients often report that while they can initiate motor tasks, their muscular and central motor output diminishes rapidly over continuous exertion. Furthermore, Item 2 (“Exercise brings on my fatigue”) assesses exertional vulnerability, identifying the rapid induction of pathologic exhaustion in response to physical demands that would normally induce minimal strain in healthy controls.

Behavioral and Role Functioning

The construct encompasses how fatigue degrades institutional, occupational, and interpersonal responsibilities. Item 7 (“Fatigue interferes with carrying out certain duties and responsibilities”) captures domestic, occupational, and administrative tasks, probing whether the individual fails to meet personal or vocational deadlines due to systemic depletion. Item 9 (“Fatigue interferes with my work, family, or social life”) reflects ecological disruption across macro-level ecological domains, quantifying the strain placed on familial relationships, sustained employment, and recreational social bonds. When these roles collapse, patients frequently experience secondary psychological morbidity, including social isolation, role identity loss, and demoralization.

Motivational and Perceptual Impact

Fatigue fundamentally alters the reward-effort computation in the central nervous system. Item 1 (“My motivation is lower when I am fatigued”) taps into the affective-motivational construct of fatigue, where the perceived effort required to initiate or complete goal-directed activities exceeds the anticipated subjective reward. Furthermore, Item 3 (“I am easily fatigued”), Item 5 (“Fatigue causes frequent problems for me”), and Item 8 (“Fatigue is among my three most disabling symptoms”) capture the subjective threshold, frequency, and cognitive salience of fatigue within the individual’s global symptom hierarchy. By evaluating whether fatigue is viewed as one of the top three most disabling symptoms, the instrument contextualizes the profound subjective burden fatigue imposes relative to pain, motor weakness, sensory disturbances, or cognitive deficits.

6. Theoretical Framework

The theoretical framework underpinning the Fatigue Severity Scale is rooted in the Biopsychosocial Model of Illness and contemporary models of Central Fatigue and Neuro-Immune Signaling. Lauren Krupp and her colleagues designed the FSS to capture the subjective manifestation of a neurobiological phenomenon that directly intersects with behavioral performance and psychological appraisal.

The Central Fatigue Hypothesis and Effort-Reward Dysregulation

From a neurobiological standpoint, pathological fatigue in conditions such as multiple sclerosis and systemic lupus erythematosus involves widespread central nervous system dysfunction. Chronic systemic inflammation—characterized by elevated circulating pro-inflammatory cytokines such as interleukin-1 beta (IL-1β), interleukin-6 (IL-6), and tumor necrosis factor-alpha (TNF-α)—triggers neuroinflammatory cascades within the basal ganglia, anterior cingulate cortex, and prefrontal circuits. These structural networks govern the neural calculation of subjective effort, dopamine transmission, and motor readiness. When these neural pathways are disrupted, the brain’s internal cost-benefit algorithm is altered: tasks that require modest energy expenditure are perceived as overwhelmingly demanding. The FSS’s emphasis on motivation (Item 1) and physical task performance (Items 4 and 6) reflects this exact neuro-computational disruption, where physiological fatigue translates directly into reduced behavioral engagement.

The Behavioral Impact and Adaptation Paradigm

Psychologically, the FSS operates within a cognitive-behavioral framework of chronic illness adaptation. In models such as those articulated by Wessely, Hotopf, and Sharpe (1998), fatigue is maintained and exacerbated through a cyclical interaction of physiological triggers, catastrophic cognitive appraisals, behavioral avoidance, and secondary physical deconditioning. When an individual experiences sudden, overwhelming fatigue following mild exertion (Item 2), fear of symptom exacerbation often drives behavioral withdrawal. Over time, the contraction of social, occupational, and physical activities (Items 7 and 9) leads to functional decline and psychological demoralization. The FSS operationalizes this operational breakdown, measuring the exact interface where biological symptom generation leads to functional restriction.

Distinction Between Sleepiness, Depression, and Fatigue

A central theoretical premise of the FSS is that fatigue must be clearly distinguished from daytime sleepiness (the propensity to fall asleep, as measured by the Epworth Sleepiness Scale) and primary clinical depression (characterized primarily by depressed mood and pervasive anhedonia). While depression and sleepiness may co-occur with fatigue, fatigue represents a distinct neuro-behavioral entity defined by energetic exhaustion and functional interference. The FSS was deliberately formulated to exclude vegetative symptoms of depression (such as appetite changes, insomnia, or guilt), ensuring that elevations on the scale reflect genuine fatigue-induced functional impairment rather than general psychological distress.

7. Validity

The psychometric validity of the Fatigue Severity Scale has been rigorously evaluated across hundreds of empirical studies spanning diverse clinical and non-clinical cohorts.

Construct and Known-Groups Validity

Construct validity was initially established in the seminal study by Krupp et al. (1989). The authors demonstrated that the FSS effectively discriminated between healthy control subjects and clinical populations known to suffer from profound pathological fatigue. In the foundational validation cohort:

• Healthy Controls: Exhibited a mean FSS score of $2.3 \pm 0.7$.
• Multiple Sclerosis (MS) Patients: Exhibited a mean FSS score of $4.7 \pm 1.5$ ($p < 0.001$).
• Systemic Lupus Erythematosus (SLE) Patients: Exhibited a mean FSS score of $4.6 \pm 1.2$ ($p < 0.001$).

This stark divergence confirmed the scale’s sensitivity to pathological fatigue states. Furthermore, known-groups validity has been replicated across Parkinson’s disease, post-stroke populations, chronic hepatitis C, fibromyalgia, traumatic brain injury, and post-polio syndrome, consistently showing statistically significant differences ($d > 1.0$) between clinical patients and age-matched healthy controls.

Convergent and Discriminant Validity

Convergent validity is well-supported by robust correlations between the FSS and other validated fatigue assessment instruments:

• Correlations with the Fatigue Impact Scale (FIS) typically range between $r = 0.65$ and $0.78$.
• Correlations with the Modified Fatigue Impact Scale (MFIS) range between $r = 0.68$ and $0.82$.
• Correlations with the vitality subscale of the SF-36 health survey are strongly negative, typically between $r = -0.60$ and $-0.74$, demonstrating that lower vitality corresponds with higher fatigue severity.
• Correlations with a 100-mm Visual Analogue Scale for Fatigue (VAS-F) consistently fall between $r = 0.55$ and $0.70$.

Discriminant validity has been established through comparisons with measures of depression and somatic disability. While FSS scores correlate moderately with depression scales such as the Beck Depression Inventory (BDI) or the Hospital Anxiety and Depression Scale (HADS) (coefficients generally ranging between $r = 0.30$ and $0.45$), multiple regression analyses demonstrate that fatigue remains independent of depressive affect. Furthermore, in clinical studies of MS, FSS scores show low-to-moderate correlations with the Expanded Disability Status Scale (EDSS) ($r = 0.18$ to $0.35$), confirming that fatigue severity is not simply a direct reflection of physical neurological impairment.

Predictive and Evaluative Validity

The FSS demonstrates high responsiveness to therapeutic interventions. In clinical trials evaluating amantadine, modafinil, and aerobic exercise programs in neuro-immunological conditions, changes in FSS scores correlate significantly with objective physiological markers of recovery and patient-reported global impression of change (PGIC). The minimal clinically important difference (MCID) for the mean FSS score has been estimated in various neurological cohorts to range between 0.45 and 0.70 points on the 1-to-7 scale.

8. Reliability

The reliability of the Fatigue Severity Scale has been repeatedly confirmed across diverse linguistic adaptations, medical populations, and research designs.

Internal Consistency

The internal consistency of the FSS is uniformly high. In the original validation paper by Krupp et al. (1989), Cronbach’s alpha was reported as 0.88 for patients with multiple sclerosis and systemic lupus erythematosus. Subsequent independent validation studies across international populations have consistently yielded comparable or superior values:

• Neurological cohorts (Parkinson’s, Stroke, MS): Cronbach’s $\alpha$ ranges between 0.88 and 0.94.
• Rheumatological cohorts (Lupus, Rheumatoid Arthritis, Sjögren’s): Cronbach’s $\alpha$ ranges between 0.86 and 0.92.
• Oncological populations: Cronbach’s $\alpha$ ranges between 0.89 and 0.93.
• General and healthy populations: Cronbach’s $\alpha$ generally falls between 0.81 and 0.88.

Corrected item-total correlations across the nine items are almost universally above 0.50, with Items 4, 5, 6, and 7 frequently demonstrating the highest item-total correlations ($r > 0.70$). Item 1 (“My motivation is lower when I am fatigued”) and Item 2 (“Exercise brings on my fatigue”) occasionally demonstrate somewhat lower item-total correlations (ranging from 0.42 to 0.58), reflecting their specific behavioral and exercise-related focus.

Test-Retest Reliability

Test-retest stability has been evaluated over intervals ranging from 24 hours to several months. In clinically stable patients with chronic neurological and autoimmune conditions:

• Over a 1- to 2-week interval, the intraclass correlation coefficient (ICC) typically spans 0.82 to 0.89.
• In the original cohort studied by Krupp et al., re-testing within several weeks showed high rank-order stability ($r = 0.84$).
• Longer-term test-retest assessments over 3 to 6 months in stable cohorts continue to demonstrate acceptable temporal stability (ICC > 0.70), reflecting that the FSS captures trait-like aspects of chronic pathological fatigue while retaining responsiveness to acute clinical relapses.

9. Factor Analysis

The internal structural validity of the FSS has been subjected to extensive exploratory factor analysis (EFA), confirmatory factor analysis (CFA), and Item Response Theory (IRT) / Rasch analyses over the past three decades.

Original Unidimensional Model

Krupp et al. (1989) initially designed the scale under the premise of a single, uniform construct: overall fatigue severity and its functional interference. Principal Component Analysis (PCA) conducted in early studies consistently produced a single dominant factor accounting for 55% to 68% of the total variance, with an eigenvalue far exceeding that of subsequent factors (e.g., initial eigenvalue > 4.5, with subsequent eigenvalues < 1.0). In this classical model, all nine items load positively and substantially on the central factor, typically with factor loadings exceeding 0.60:

• Item 1: Factor loading ~ 0.58 – 0.68
• Item 2: Factor loading ~ 0.52 – 0.65
• Item 3: Factor loading ~ 0.68 – 0.78
• Item 4: Factor loading ~ 0.78 – 0.86
• Item 5: Factor loading ~ 0.80 – 0.88
• Item 6: Factor loading ~ 0.79 – 0.87
• Item 7: Factor loading ~ 0.76 – 0.85
• Item 8: Factor loading ~ 0.65 – 0.75
• Item 9: Factor loading ~ 0.74 – 0.83

Multifactorial Perspectives and CFA Fit Indices

Despite the functional utility of a single summary score, subsequent modern confirmatory factor analyses have questioned the strict unidimensionality of the 9-item tool. Several investigators (e.g., Valko et al., 2008; Lerdal et al., 2005) have reported that a single-factor CFA model sometimes demonstrates marginal fit in heterogeneous cohorts:

• Comparative Fit Index (CFI): 0.88 to 0.93
• Tucker-Lewis Index (TLI): 0.85 to 0.91
• Root Mean Square Error of Approximation (RMSEA): 0.08 to 0.12
• Standardized Root Mean Square Residual (SRMR): 0.05 to 0.07

Consequently, alternative models have been advanced. The most frequent alternative is a two-factor correlated model dividing the scale into:

1. Physical/Exertional Fatigue Impact: Comprising Items 2, 4, 6, and 7.
2. Global/Motivational Impact: Comprising Items 1, 3, 5, 8, and 9.

This two-factor structure often produces superior fit indices ($ ext{CFI} > 0.95$,$ ext{RMSEA} < 0.06$). Additionally, Rasch and Item Response Theory (IRT) analyses conducted in multiple languages (such as the FSS-7 variant proposed by Mills et al., 2009) have revealed that Item 1 (motivation) and Item 2 (exercise induction) frequently exhibit differential item functioning (DIF) or poor fit to the Rasch model. Nevertheless, because the 9-item composite score exhibits such pronounced clinical utility and historical continuity, the standard unidimensional scoring remains the prevailing convention in clinical trials worldwide.

10. Instrument / Measurement Tool

• Instrument Name: Fatigue Severity Scale (FSS)
• Alternative Abbreviation: FSS-9
• Authors: Lauren B. Krupp, Nicholas G. LaRocca, Judi Muir-Nash, and Alfred D. Steinberg
• Initial Publication Year: 1989
• Construct Measured: Severity of fatigue and its disabling functional impact on physical performance, motivation, daily responsibilities, and social engagement
• Target Population: Adult and adolescent clinical populations (multiple sclerosis, systemic lupus erythematosus, ME/CFS, post-stroke, Parkinson’s disease, rheumatoid arthritis, cancer-related fatigue, Long COVID) and healthy reference groups
• Administration Format: Self-report questionnaire (paper-and-pencil, computerized, or clinical interview administration)
• Completion Time: Approximately 2 to 3 minutes
• Item Count: 9 statements
• Response Scale: 7-point Likert scale (1 = Strongly disagree, 2 = Disagree, 3 = Somewhat disagree, 4 = Neither agree nor disagree, 5 = Somewhat agree, 6 = Agree, 7 = Strongly agree)
• Reverse-Scored Items: None. All items are positively keyed toward higher fatigue severity.
• Scoring Options:
  • Mean Score (Standard Practice): Calculated by summing all 9 items and dividing by 9 (score range: 1.0 to 7.0).
  • Total Sum Score: Calculated by summing the raw responses of all 9 items directly (score range: 9 to 63).
• Clinical Interpretation and Cutoff Scores:
  • Mean FSS < 4.0 (Total < 36): Normal or non-pathological fatigue range (characteristic of healthy populations).
  • Mean FSS ≥ 4.0 (Total ≥ 36): Clinically significant, pathological fatigue causing noticeable functional impairment.
  • Mean FSS ≥ 5.0 (Total ≥ 45): Severe fatigue indicating substantial, pervasive disability.

11. Permissions & Fee and Test Year

The Fatigue Severity Scale was first published in 1989 in the Archives of Neurology (now JAMA Neurology). The seminal publication is: Krupp, L. B., LaRocca, N. G., Muir-Nash, J., & Steinberg, A. D. (1989). The fatigue severity scale: Application to patients with multiple sclerosis and systemic lupus erythematosus. Archives of Neurology, 46(10), 1121–1123.

Regarding licensing, permissions, and fees:

• Academic and Non-Commercial Research: The Fatigue Severity Scale is widely regarded in the international academic and clinical research community as an open-access psychometric instrument. Researchers and clinicians typically administer the 9-item scale without licensing fees, provided proper bibliographic citation is accorded to Krupp et al. (1989).
• Commercial and Pharmaceutical Clinical Trials: While the instrument has been placed in the public domain of clinical research through its publication, commercial clinical trials and digital platform developers seeking certified licensing, official linguistic translations, or digital electronic clinical outcome assessment (eCOA) implementations frequently consult the original authors or academic copyright holders (via the American Medical Association / JAMA Network) to ensure regulatory compliance.
• Translations: Validated cross-cultural adaptations exist in over 30 languages (including French, German, Italian, Spanish, Turkish, Brazilian Portuguese, Japanese, Chinese, and Norwegian), each published under academic fair-use guidelines within regional validation papers.

12. References

Below is a comprehensive list of primary foundational and psychometric references formatted according to the American Psychological Association (APA 7th edition):

• Amtmann, D., Bamer, A. M., Noonan, V., Chung, H., & Cook, K. F. (2012). Comparison of the psychometric properties of two fatigue scales in multiple sclerosis. Rehabilitation Psychology, 57(2), 159–166. https://doi.org/10.1037/a0027890
• Flachenecker, P., Kümpfel, T., Kallmann, B., Gottschalk, M., Grauer, O., Rieckmann, P., Trenkwalder, C., & Toyka, K. V. (2002). Fatigue in multiple sclerosis: A comparison of different assessment tools. Neurological Sciences, 23(6), 283–288. https://doi.org/10.1007/s100720200062
• Krupp, L. B., LaRocca, N. G., Muir-Nash, J., & Steinberg, A. D. (1989). The fatigue severity scale: Application to patients with multiple sclerosis and systemic lupus erythematosus. Archives of Neurology, 46(10), 1121–1123. https://doi.org/10.1001/archneur.1989.00520460115022
• Lerdal, A., Wahl, A., Rustøen, T., Hanestad, B. R., & Moum, T. (2005). Fatigue in the general population: A translation and test of the psychometric properties of the Norwegian version of the fatigue severity scale. Scandinavian Journal of Public Health, 33(2), 123–130. https://doi.org/10.1080/14034940410028406
• Mills, R. J., Young, C. A., Pallant, J. F., & Tennant, A. (2009). Clarifying the fatigue severity scale: Rasch analysis of the Swedish version. Journal of Rehabilitation Medicine, 41(9), 744–749. https://doi.org/10.2340/16501977-0407
• Taylor, R. R., Jason, L. A., & Torres, A. (2000). Fatigue rating scales: An exploration of completeness, reliability, and validity. Work, 15(2), 97–107. https://pubmed.ncbi.nlm.nih.gov/12441492/
• Valko, P. O., Bassetti, C. L., Bloch, K. E., Held, U., & Baumann, C. R. (2008). Validation of the fatigue severity scale in a Swiss cohort. Sleep, 31(11), 1601–1607. https://doi.org/10.1093/sleep/31.11.1601
• Wessely, S., Hotopf, M., & Sharpe, M. (1998). Chronic fatigue and its syndromes. Oxford University Press. https://doi.org/10.1093/med/9780198523314.001.0001

13. Items of the Scale