Cognitive and Physical Fatigue in Multiple Sclerosis:Relations Between Sel-report and Objective Performance

Robert H. Paul, William W. Beatty, Department of Psychiatry and Behavioral Sciences, University of Oklahoma Health, Sciences Center, Oklahoma City, Oklahoma; Ronni Schneider, Heinrich Heine University, Institute of General Psychology, Düsseldorf, Germany; Carlos R. Blanco, Karen A. Hames, Department of Psychiatry and Behavioral Sciences, University of Oklahoma Health, Sciences Center, Oklahoma City, Oklahoma.

Fatigue is one of the most disabling symptoms of multiple sclerosis (MS), but little is known about patients' perceptions of fatigue and changes in their performance on cognitive tasks. To study these relations, 39 patients with clinically definite MS and 19 matched healthy control participants completed baseline self-reports of physical and cognitive fatigue and measurements of grip strength, word list learning, and vigilance. Following 30 min of testing on a cognitive work battery (verbal fluency, Wechsler Adult Intelligence Scale-Revised [WAIS-R] Comprehension, WAIS-R Vocabulary), the baseline measures were readministered. At baseline, MS patients reported more physical and cognitive fatigue than did controls and the MS patients performed more poorly on the grip strength, word list learning, and vigilance tasks. Following cognitive work, the patients reported increased physical and cognitive fatigue, but their objective performance on grip strength, learning, and vigilance was unchanged from baseline. Controls showed no change in fatigue ratings or performance. [Applied Neuropsychology 5(3):143-148, 1998. © 1998 Lawrence Erlbaum Associates, Inc.]

Many individuals with multiple sclerosis (MS) report experiencing fatigue that is characterized by a lack of energy and a need for rest (Freal, Kraft, & Coryell, 1984). Not all MS patients are affected by fatigue, but most report it to be a daily symptom (Krupp, Alvarez, LaRocca, & Scheinberg, 1988; Krupp, LaRocca, Muir-Nash, & Steinberg, 1989), and nearly one third report that fatigue is their most disabling symptom (Krupp et al., 1988).
When compared to healthy controls, MS patients report more frequent and severe fatigue, which is easily exacerbated by physical exertion and heat. Further, MS patients report that fatigue develops quickly, worsens throughout the course of the day, prevents further physical functioning, and interferes with their daily activities (Freal et al., 1984; Krupp et al., 1988). Not surprisingly, fatigue is a leading cause of early retirement in the MS population (Edgley, Sullivan, & Dehoux, 1991). Phenomenologically, fatigue has both physical and cognitive components. However, in the MS literature most studies have only focused on physical fatigue, typically defined as a state of physical exhaustion, or included both physical and cognitive components of fatigue in a generalized measure. Physical fatigue in MS is associated with greater physical disability, a progressive disease course, and older age (independent of disease severity; Colosimo et al., 1995), but the correlates of cognitive fatigue in MS are not known.
Prolonged mental effort is known to produce a state of cognitive fatigue in healthy individuals (Bartley & Chute, 1947; Finkelman, 1994; Fiske & Schneider, 1981), but whether fatigue increases more rapidly for MS patients who are performing a cognitive task is not known. In one study (Jennekins-Schinkel, Sanders, Lanser, & Van der Velde, 1988), MS patients exhibited longer reaction times than did healthy controls before and after the administration of a neuropsychological battery, but the difference score between pre- and post-test battery performance for the MS group was not statistically different from that of the controls.
By contrast, Kujala, Portin, Revonsuo, and Ruutianen (1995) found that MS patients developed cognitive fatigue over the course of a 15-min vigilance task as evidenced by poorer performance in the last two 5-min blocks compared to the first 5-min block. Moreover, this pattern of performance was evident in cognitively intact MS patients and in patients with mildly deteriorated cognitive capacities, but not in healthy controls.
In a recent study, Johnson, Lange, DeLuca, Korn, and Natelson (1997) compared performance on the Paced Auditory Serial Addition Test (PASAT; Gronwall, 1977) by patients with MS, major depression of psychiatric origin, or chronic fatigue syndrome (CFS). During the course of a 3-hr-long test battery, the participants rated their overall level of fatigue and performed the PASAT on four occasions. Despite elevated depression scores, all of the patient groups exhibited the same degree of improvement with practice as the healthy controls, but only the MS patients reported increasing levels of fatigue during the course of the test battery. This observation raises the possibility that in MS, changes in subjective ratings of fatigue might not parallel objective measures of performance.
To better understand the nature of cognitive fatigue and its relation to physical fatigue and cognitive performance in MS, we recorded subjective ratings of cognitive and physical fatigue before and after the administration of a comprehensive test battery intended to induce cognitive fatigue. In addition, we administered tests of memory, vigilance, and grip strength before and after the test battery to examine the influence of mental exertion on objective measures of cognitive and physical performance. Assuming that participants could distinguish between cognitive and physical components of fatigue, we predicted that (a) MS patients would experience greater cognitive fatigue than normal control participants following prolonged mental effort and (b) performing cognitive work would impair subsequent cognitive but not physical performance to a greater degree in MS patients than in control participants.

Thirty-nine patients (33 female and 6 male) who met criteria for clinically definite MS (Poser et al., 1983) were recruited from local support groups and community neurologists. The patients averaged 45.5 ± 6.7 years of age, 14.9 ± 2.5 years of education, 12.2 ± 4.8 years since disease diagnosis, and 4.1 ± 2.5 on the Ambulation Index (AI; Hauser et al., 1983). The AI is a measure of overall physical disability that is based on a 0 (asymptomatic) to 9 (wheelchair-dependent; unable to transfer independently) point scale. The AI has been shown in previous research (Beatty, Goodkin, Hertsgaard, & Monson, 1990) to be highly correlated with the summary score from the Expanded Disability Status Scale (Kurtzke, 1983).
Nineteen community-based healthy controls (13 female and 6 male) were recruited from the community through various sources (e.g., University of Oklahoma Health Sciences Center [OUHSC] personnel and patients' relatives). The controls averaged 42.4 ± 16.6 years of age and 15.1 ± 2.1 years of education. Patients and controls with a history of neurologic disease (other than MS), severe psychiatric illness (e.g., schizophrenia, bipolar illness), alcohol or drug abuse, serious head injury (loss of consciousness for more than 1 hr), gross visual impairments (corrected vision worse than 20/70 in the better eye, or ophthalmoplegia), or serious medical illness (e.g., recent or complicated heart attack) were excluded from the study.
The protocol was approved by the OUHSC Institutional Review Board. All participants provided written informed consent after a thorough explanation of the study. To minimize participants' fatigue prior to testing, all were given the option of being tested at home or in the laboratory; nearly all participants elected to be tested at home. No participants were paid for their participation. No patients were tested until at least 1 month elapsed following an exacerbation.
All participants were administered the Multiscale Depression Inventory (Nyenhuis et al., 1995). This 50- item self-report inventory measures affective, self-perceptive, and functional symptoms of depression separately as the Mood, Evaluative, and Vegetative scales. Mean scores for the controls were 19.8, 17.2, and 25.4 for the Mood, Evaluative, and Vegetative scales, respectively. Patients averaged 26.7, 25.3, and 36.4 on the same scales. Between-group differences were significant for the Evaluative and Vegetative scales, Fs(1, 56) > 6.91, ps < .05, but not for the Mood scale, F(1, 56) = 3.80, p > .05.
Fatigue and Performance Variables
The following tests were administered twice -- once at the beginning and again at the conclusion of the test session.
Fatigue scales. Existing rating scales (Krupp et al., 1989; Smets, Garssen, Bonke, & Haes, 1995) for measuring subjective fatigue combine elements of physical and mental fatigue and are poorly suited to measuring changes in fatigue over short intervals. We developed separate scales for physical and cognitive fatigue (see Appendix).
Ratings. For each question for the initial rating, participants selected a number from 1 (not at all)to 5 (a great deal) that best described how they felt at that moment. For the second rating, after completing the cognitive work battery, participants rated their perceived change in fatigue since their first rating of the fatigue scale, from 1 (much less)to 3 (no change)to 5 (much more). The wording of each question was altered slightly (e.g., "Compared to your first rating are you having trouble concentrating?"). Apilot study indicated that it was necessary to have participants rate change in fatigue because many patients produced maximum ratings on several items (i.e., 5, or a great deal of fatigue) on the initial ratings (Paul, 1997).
Grip strength. Participants were instructed to squeeze a hand dynamometer (Lafayette Instrument Co., model 78010) once with their dominant hand as hard as they could. The dependent measure was force recorded by the dynamometer.
Word List Learning 1 (Beatty, Goodkin, Monson, Beatty, & Hertsgaard, 1988). Participants were read a 14-word list and instructed to recall as many words as possible. The procedure was repeated until four learning trials were completed. Two different but equally difficult lists were used for the tests at the beginning and end of the session. The lists are known from previous studies to be sensitive to the memory impairments associated with MS (Beatty, Goodkin, Monson, & Beatty, 1989; Beatty et al., 1988).
Vigilance. The Gordon Distractibility Task (Gordon Systems Inc., DeWitt, NY) was administered as a measure of vigilance. On this test, participants viewed a three-position screen on which numbers flashed rapidly and randomly in all three positions. The participants were instructed to rest their hand on a response button and press the button whenever the sequence 1-9 appeared in the center position. Participants were given a short practice trial to become familiar with the test. Forty-five targets occurred at unpredictable intervals during the 9-min-long test. Hits and false positives recorded in three 3-min intervals served as the primary measures.
Cognitive Work Battery
Once the aforementioned tests were completed for the first time, participants were administered a battery consisting of letter and category fluency tests and the Famous Faces test (Beatty et al., 1988) as well as the Vocabulary and Comprehension tests from the Wechsler Adult Intelligence Test-Revised (Wechsler, 1981). After 30 min of work on the cognitive battery, participants were administered the subjective measures of fatigue and the grip strength, word list learning, and vigilance tests a second time.
On the initial ratings, the patients indicated greater feelings of physical (M = 2.65 ± 1.24 vs. 1.49 ± 0.56) and cognitive (M = 2.56 ± 1.07 vs. 1.60 ± 0.67) fatigue than did controls, Fs(1, 56) > 12.59, ps < .001. On the ratings of change in fatigue after performing the cognitive work battery, patients reported increases in physical (Mdn = 3.75) and cognitive (Mdn = 3.57) fatigue, whereas controls experienced no increase in fatigue (i.e., no significant change from the expected value of 3.0) for either physical or cognitive fatigue (Mdns = 3.00 and 3.14, respectively). Differences between groups in the magnitude of change were significant by medians tests, which were corrected for continuity: chi 2 (1, N = 58) = 13.93, p < .001 for physical fatigue, and chi 2 (1, N = 58) = 5.01, p < .05 for cognitive fatigue.
The similar results for the initial ratings of cognitive and physical fatigue as well as for changes in perceived physical and cognitive fatigue over the session arose because both of the measures were highly correlated for both patients and controls (rs > .66, ps < .001). Thus, our measures failed to discriminate cognitive from physical fatigue.
Four patients were incapable of completing the vigilance task because they experienced extreme distractibility. On the objective measures of physical and cognitive performances, the pattern was quite different, as shown in Table 1. At both Time 1 and Time 2, controls had higher grip strength; recalled more words, Fs(1, 56) > 19.09, ps < .001; and detected more targets on the vigilance task, F(1, 52) = 7.47, p < .01, than the patients, but neither the main effects of time (all Fs < 2.85, ps > .05, ns) nor the Group [lozenge] Time interactions (Fs < 3.24, ps > .05) were significant. Thus, neither controls nor patients with MS showed changes in physical or cognitive performance as a consequence of the programmed cognitive work, although the patients showed the expected deficits in initial and final performance on all tasks. Additional analyses indicated that the use of prescription medications for MS or nicotine and caffeine did not alter the results.

Results of this study indicate that feelings of fatigue are easily aroused in patients with MS. At the start of the session, the patients reported more physical and cognitive fatigue than controls, confirming many previous reports (Freal et al., 1984; Krupp et al., 1988; Krupp et al., 1989). Following performance of tests that required cognitive but not physical effort, the MS patients reported further increases in fatigue, whereas controls did not report such changes. However, the patients' subjective appraisals of their cognitive and physical capabilities were not paralleled by changes from their initial levels of performance on objective tests of vigilance, verbal learning and memory, and grip strength.
The present findings are strikingly similar to the recent results of Johnson et al. (1997). These investigators showed that MS patients, like patients with depression or CFS, exhibited normal improvement with practice over a series of four trials on the PASAT, a difficult test of auditory attention. Unlike controls or depressed and CFS patients, the MS patients in their study reported steadily increasing levels of fatigue during the 3-hr-long battery.
The patients in the Johnson et al. (1997) study were younger and less severely disabled than the patients in our study. Further, on the single 5-point question used by Johnson et al. to measure fatigue, the MS patients reported no more fatigue than controls on the initial rating. Despite these and other procedural differences, both studies found that MS patients experienced rapidly increasing fatigue as a consequence of mental effort, yet the quality of their cognitive performance remained unchanged or actually improved over the test session. Further, in a recent study involving 139 MS patients, there was no correlation between ratings of fatigue and cognitive performance (Schwartz, Coulthard-Morris, & Zeng, 1996).
In our study, participants performed the cognitive work battery for a fixed time period, but at their own pace. As a result, the patients, on average, completed fewer test items than controls. Further, the MS patients in our study performed more poorly than controls on both the vigilance and the verbal learning tasks. Although it might be of interest to examine fatigue ratings under conditions where patients and controls work at the same rate and achieve the same level of performance, it is not likely that these variables are critical. Johnson et al. (1997) controlled the presentation of tests, and the MS patients in their study did not differ from controls on the PASAT. Yet their MS patients, like those in our study, reported steadily increasing fatigue without deterioration in objective performance.
Because MS patients are extremely sensitive to sensations of fatigue and they are unable to distinguish physical from mental fatigue, it seems possible that their perceptions of increased physical and mental fatigue over the day might cause patients with MS to underestimate their objective performance on cognitive tasks. Whether this phenomenon actually occurs in the workplace is not known, but clinicians may wish to alert their patients to this possibility.
This study was completed in partial fulfillment of the requirements for the PhD in Biological Psychology at the University of Oklahoma Health Sciences Center by Robert Paul. We thank J. Michael Banowetz, MD, and John Hastings, MD, who allowed us to test their patients.