Many psychometric instruments have been developed to measure various components of fatigue. In this study the factor structure of the scales included in nine psychometric instruments developed to measure different components of fatigue were investigated. One hundred and thirty-four research participants completed the nine instruments in groups of approximately 30 each. Factor analysis of the scores obtained on these scales yielded a twofactor solution with 11 scales loading 0.40 or greater on the first factor and 10 scales loading 0.40 or greater on the second factor. These two factors, Distress-Fatigue and Vigor, accounted for 48 percent of the total variance and approximated simple structure. The results reveal that the different psychometric measures of fatigue are really measuring only two different factors rather than the many different factors specified by the different instruments.
Fatigue is a phenomenon that is experienced by most, if not all, individuals and is accepted as a normal part of everyday life. However, a substantial number of individuals complain of persistent tiredness or listlessness. According to the National Ambulatory Medical Care Survey (National Center for Health Statistics, 1978), fatigue is the seventh most frequent initial complaint in U.S. medical offices. Studies on fatigue in primary care show that prevalence rates vary between 7 and 45 percent (Lewis & Wessely, 1992). Epidemiological studies conducted in the United States have revealed that between 14 and 41 percent of individuals report being fatigued (Chen, 1986; Manu et al., 1989). Feelings of fatigue are prevalent among individuals with physical or emotional illness such as individuals with depression, cancer, heart disease, gynecological disorders, and of course, chronic fatigue syndrome (Kroenke et al., 1988; Krupp, LaRocca, Muir-Nash, & Steinberg, 1989; Sugarman & Berg, 1984).
During the past decade there have been a number of attempts to develop psychometrically sound instruments that will provide a quantitative index of the subjective state of fatigue. These have consisted of both unidimensional (e.g., Krupp et al., 1989) and multidimensional (e.g., Schwartz, Jandorf, & Krupp, 1993) instruments. The implication, particularly of the multidimensional instruments, is that there are different dimensions of fatigue. However, the different dimensions supposedly assessed are characterized differently. For example, the instrument developed by Schwartz et al. identified four different dimensions consisting of fatigue severity, situation-specific fatigue, consequences of fatigue, and responsiveness of fatigue to rest/sleep. Chalder et al. (1993) developed an instrument which measured the two dimensions of physical and mental fatigue. There is, however, some evidence to suggest that the unidimensional measures of fatigue are as good at measuring fatigue as are the multidimensional measures. Given the numerous measures of fatigue and the different dimensions they supposedly measure, it would be important to know what similarities and differences exist among the different measures of fatigue. The present study was conducted to identify the factor structure of nine different measures of fatigue.
The research participants were 81 female and 53 male undergraduate students enrolled in an undergraduate introductory psychology course. The participants ranged from 18-54 years of age with 81% of the participants being 18-25 years of age. Participants completed the study in groups of approximately 30 each.
Nine psychometric inventories specifically developed to measure fatigue were selected for assessment in this study. The Multidimensional Fatigue Inventory (MFI; Smets, Garssen, Bronke, & DeHaes, 1995) is a 20-item self-report instrument consisting of five scales: General Fatigue, Physical Fatigue, Reduced Activity, Reduced Motivation, and Mental Fatigue. Each scale contains four items rated on a scale of O to 5 with the scale score of O having the anchor of entirely true and the scale score of 5 having the anchor of no, not true. The five scales were identified through factor analysis and are assumed to measure different aspects of fatigue. The MFI has a mean internal consistency (Cronbach's alpha) coefficient of .84 and good construct validity demonstrated by its ability to discriminate between cancer patients, junior physicians, army recruits, and Chronic Fatigue Syndrome patients.
The Fatigue Scale (Chalder et al., 1993) is a self-rating measure of fatigue severity for both physical and mental symptoms. The scale consists of eight questions relating to physical symptoms and six questions relating to symptoms of mental fatigue. Items are scored in a Likert format with response options ranging from better than usual to much worse than usual. The Fatigue Scale has an internal consistency reliability coefficient (Cronbach's alpha) of .845 for the physical symptoms scale and .821 for the mental symptoms scale.
The Positive Affect Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) consists of two scales which were developed to measure Positive and Negative Affect. The Positive Affect scale (PA) reflects the extent to which a person feels enthusiastic, active, and alert. High PA is characterized by high energy and full concentration, whereas low PA scores typically reflect sadness and lethargy. The Negative Affect scale (NA) subsumes a variety of aversive mood states, including guilt, nervousness, anger, and fear. Low NA reflects a state of calmness and serenity. The PANAS includes 10 positive and 10 negative items and respondents are to rate each item in terms of the extent to which that item communicates the way they have felt during the past few weeks on a 5-point scale ranging from very slightly or not at all to extremely. The internal consistency reliability coefficient for the PANAS ranges from .86 to .90 for PA and from .84 to .87 for NA. The correlation between the PA and NA scales ranges from -.12 to -.23 providing evidence of discriminant validity.
The Fatigue Assessment Inventory (FAI; Schwartz, Jandorf, & Krupp, 1993) is a 29-item instrument consisting of four scales: Fatigue Severity, Situation Specific, Psychological Consequences, and Responds to Rest/Sleep. The Fatigue Severity scale is used to measure overall fatigue level. The Situation Specific scale measures the degree to which fatigue symptoms are pervasive, or are influenced by circumstances such as stress, heat, or cold. The Psychological Consequences scale measures the consequences of fatigue such as loss of patience or motivation. The Responds to Rest/Sleep scale measures whether a person's fatigue responds to rest or sleep. Internal consistency for the FAI (Cronbach's alpha) ranges from .70 to .92.
The Mental Fatigue Inventory (Bentall, Wood, Marrinan, Deans, & Edwards, 1993) is a nine-item inventory developed to assess the cognitive consequences of fatigue. Items are responded to on a 5-point scale ranging from not at all to very much. The inventory discriminates effectively between Chronic Fatigue Syndrome patients, depressed patients, muscle-diseased patients, recovered Chronic Fatigue Syndrome patients, and normal controls. Internal consistency for the Mental Fatigue Inventory is .93 (Cronbach's alpha) and test-retest reliability (two-day interval) is .88.
The Profile of Fatigue-Related Symptoms (PFRS; Ray, Weir, Phillips, & Cullen, 1992) is a 54-item questionnaire developed to assess symptom pattern and severity in patients suffering from Chronic Fatigue Syndrome. The PFRS consists of four scales, Emotional Distress, Cognitive Difficulty, Fatigue, and Somatic Symptoms, which are responded to on a 7-point scale with response options ranging from not at all to extremely. The PFRS discriminates effectively between patients and students and has good convergent validity (.65 to .88) and internal consistency (.86 to .97) and test-retest reliability (.88 to .96). Subscale intercorrelations range from .44 to .64 indicating that there is some overlap between them.
The Visual Analogue Scale for Fatigue (VAS-F; Lee, Hicks, & Nino-Murcia, 1991) is a visual analogue instrument developed to assess fatigue and energy levels in patient settings and normal populations. The VAS-F consists of a fatigue and energy subscale comprising 18 items with bipolar descriptors on either end of a 100 mm line. Respondents are asked to mark the point on this line that indicates how they feel in relation to the bipolar descriptor. Internal consistency is .96 (Cronbach's alpha) for both the Fatigue scale and the Energy scale. Correlations with the Profile of Mood States and the Stanford Sleepiness Scale provide convergent validity.
The Profile of Mood States (POMS; McNair, Lorr & Droppleman, 1992) fatigue scale is one of the six scales comprising the POMS. This is a 7-item selfreport scale responded to on a 5-point scale from not at all to extremely. Internal consistency coefficients range from .84 to .95 and test-retest reliability coefficients range from .65 to .74. Predictive and construct validity have been demonstrated in studies ranging from brief psychotherapy studies to studies investigating response to emotion-inducing conditions (McNair et al.).
The Vitality Inventory (VI; Christensen & Ouellette, 1995) is a 22-item factor analytically derived self-report inventory providing a measure of Exhaustion and Vigor. Internal consistency, coefficient alpha, is .92 for the Exhaustion scale and .91 for the Vigor scale. Construct validity is demonstrated by the VFs correlation with similar measures and by demonstrating that individuals with chronic fatigue and fibromyalgia score significantly lower on the Vigor scale and significantly higher on the Exhaustion scale of this inventory.
RESULTS AND DISCUSSION
The responses research participants provided to each of the inventories were scored and these scores were analyzed by principal components factor analysis to identify the factor structure existing across these different measures of fatigue. Using the dual criteria of factors with eigenvalues greater than 1.0 and a sharp slope change in the magnitude separating the second factor from the residual error scree (Cattell, 1966) led to the retention of two factors. The varimax rotated factor matrix is presented in Table 1. The varimax rotated factor matrix resulted in a simple structure with 11 scales - PANAS negative affect, FAI Fatigue Severity, MFI, PFRS Emotional Distress, Cognitive Difficulty, Fatigue, and Somatic Symptoms scales, VAS-F Fatigue scale, POMS Fatigue scale, and the VI Exhaustion scale - loading 0.40 or greater on the first factor. Ten scales - MFI General Fatigue, Physical Fatigue, Reduced Activity, Reduced Motivation, and Mental Fatigue, The Fatigue Physical Symptoms and Mental Symptoms scale, PANAS Positive Affect, VAS-F Energy scale, and the VI Vigor scale - loaded .40 or greater on the second factor. These two factors accounted for 37.55 and 10.28 percent of the total variance respectively, and approximated simple structure (Kerlinger, 1986).
Inspection of the items included in the scales loading on the first factor indicates that they provide an assessment of the degree of fatigue and distressful experiences ranging from depression, difficulties in concentration, and confusion to pessimism. This factor is, therefore, labeled the Distress-Fatigue factor. The second factor, labeled Vigor, comprises a group of scales that reflect the extent to which a person feels vigorous, active, excited, and motivated.
Three of the scales of the FAI - Situation Specific, Fatigue Consequences, and Responds to Rest/Sleep - did not load on either of the two factors which were extracted. Inspection of the items comprising these two factors reveals that these scales are measuring very specific components of fatigue - such as the variables that either bring on fatigue or reduce fatigue. Therefore, it is not surprising that they did not load on either of the two factors that were extracted. It is also interesting that these scales did not result in the extraction of a third factor. Apparently, these three scales are measuring factors that have little interrelationship or relationship with the two factors that were extracted.
The fact that two factors were extracted that accounted for about half of the total variance suggests that the scales loading on each of these two factors are measuring similar behaviors. This suggests that, rather than measuring a variety of different fatigue-related components, most of the instruments are really measuring either a vigorous state or a distressed-fatigue state. For example, the various scales of the MFI, and the Fatigue Scale all loaded on factor 2 suggesting that all scales included in the instruments are providing a measure of the extent to which a person feels vigorous. Similarly, the four scales of the PFRS all loaded on factor 1 which is measuring the extent to which a person feels distressed-fatigued.
Of the nine psychometric instruments included in this study, only the PANAS, VAS-F, and VI provided a measure of both the Vigor and Distress-Fatigue factors. Therefore, it would seem that any one of these three scales would be most appropriate ones to use as they all provide an assessment of both factors.
The present study has revealed that the scales comprising most fatigue inventories can be reduced to two factors. This conclusion has been reached through an assessment of responses of a sample of individuals in the general population. Having a good measure of fatigue that can be used with the general population is seen as very important because a significant percentage of the population reports being fatigued (Manu et al., 1989). The present study has identified three scales that seem to be the best ones to use with this population. However, there are also various medical conditions, such as Chronic Fatigue Syndrome and Fibromyalgia that are characterized by symptoms of fatigue. While it would seem logical that the scales identified in this study as the most appropriate ones for measuring fatigue in a general population would also be appropriate for measuring fatigue in these individuals, it is an empirical question as to whether this generalization can be made. If the results of the present study hold up under replication with other samples, a strong case would exist for using only these two factors as measures of fatigue and vigor.
Bentall, R. P., Wood, G. C., Marrinan, T., Deans, C., & Edwards, R. H. T. (1993). A brief mental fatigue questionnaire. British Journal of Clinical Psychology, 32, 375-379.
Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276.
Chalder, T., Berelowitz, G., Pawlikowska, T., Watts, L., Wessely, S., Wright, D., & Wallace, E. P. (1993). Development of a fatigue scale. Journal of Psychosomatic Research, 37, 147-153.
Chen, M. K. (1986). The epidemiology of self-perceived fatigue in adults. Preventive Medicine, 15, 74-81.
Christensen, L., & Ouellette, J. (1995). Development of a measure of energetic arousal. Paper presented at the annual convention of the Southeastern Psychological Association. Savannah, GA.
Kerlinger, F. N. (1986). Foundations of behavioral research. New York: Holt, Rinehart, and Winston.
Kroenke, K., Wood, D. R., Mangelsdorff, A. D., Meier, N. J., & Powell, J. B. (1988). Chronic fatigue in primary care: Prevalence, patient characteristics, and outcome. Journal of the American Medical Association, 260, 929-934.
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, 1121-1123.
Lee, K. A., Hicks, G., & Nino-Murcia, G. (1991). Validity and reliability of a scale to assess fatigue. Psychiatric Research, 36, 291-298.
Lewis, G., & Wessely, S. (1992). The epidemiology of fatigue: More questions than answers. Journal of Epidemiology and Community Health, 46, 92-97.
Manu, P., Matthews, D. A., Lane, T. J., Tennen, H. Hesselbrock, V., Mendola, R., & Affleck, G. (1989). Depression among patients with chief complaint of chronic fatigue. Journal of Affective Disorders, 17, 165-172.
McNair, D. M., Lorr, M., & Droppleman, L. F. (1992). Profile of mood states. San Diego, California: Educational and Industrial Testing Service.
National Center for Health Statistics. (1978). National survey data on food consumption: Uses and recommendations. Washington, DC: National Academy Press.
Ray, C., Weir, W. R. C., Phillips, S., & Cullen, S. (1992). Development of a measure of symptoms in Chronic Fatigue Syndrome: The Profile of Fatigue-Related Symptoms (PFRS). Psychology and Health, 7, 272-43.
Schwartz, J. E., Jandorf, L., & Krupp, L. B. (1993). The measurement of fatigue: A new instrument. Journal of Psychosomatic Research, 37, 753-762.
Smets, E. M. A., Garssen, B., Bronke, B., & DeHaes, J. C. J. M. (1995). The Multidimensional Fatigue Inventory (MFI): Psychometric qualities of an instrument to assess fatigue. Journal of Psychosomatic Research, 39, 315-325.
Sugarman, R. R., & Berg, A. O. (1984). Evaluation of fatigue in a family practice. Journal of Family Practice, 19, 643-647.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect. The PANAS scales. Journal of Personality and Social Psychology, 54, 1063-1070.
LARRY CHRISTENSEN AND MARCI PIPER-TERRY
University of South Alabama, AL, USA
Larry Christensen and Marci Piper-Terry, Department of Psychology, University of South Alabama, Alabama, USA.
Appreciation is due to reviewers including: Kamel Gana, PhD, Department of Psychology, University of Tours, 3, rue des tanneurs, 37041 Tours, France. Email:
Please address correspondence and reprint requests to: Larry Christensen, Department of Pyschology, Life Sciences Building, Room 324, University of South Alabama, Mobile, AL 36688, USA. Phone: 251-460-6371; Fax: 251-460-6320; Email:
Copyright Society for Personality Research, Incorporated 2004
Provided by ProQuest Information and Learning Company. All rights Reserved