Layout

Primary and Nonprimary Fatigue in
Multiple Sclerosis
Susan J. Forwell, PhD, (OT)C, FCAOT; Sandra Brunham, BSR(PT), MSc(Rehab); Helen Tremlett, MRPharmS, PhD; Wendy Morrison, RN, CNN(C); Joel Oger, MD, FRCPC, FAAN Using the fatigue algorithm described in the Clinical Practice Guidelines for Fatigue and Multiple
Sclerosis (MS), this study aimed to determine the frequency of fatigue related directly to the MS dis-
ease process (primary fatigue) versus that related to disease symptoms such as depression (nonprimary
fatigue), differentiate primary from nonprimary MS fatigue, and identify any characteristics unique
to primary MS fatigue. Consecutive clinically definite MS patients from the University of British
Columbia MS clinic were invited to participate. The screening assessment included standardized scales
and questionnaires. In total, 50 patients completed the study. Nonprimary fatigue was present in 36
individuals (72%), of whom all but 2 had multiple factors contributing to fatigue. The most common
factors contributing to nonprimary fatigue were sleep problems (58%), mobility limitations (52%),
and depression (40%). Compared with patients with nonprimary fatigue, those with isolated primary
fatigue had lower fatigue scores and reported fatigue onset at midday (P < .05). Any attempt to study
or manage MS fatigue should be preceded by identification and amelioration of treatable nonprimary
fatigue factors before focusing on primary fatigue in MS. Int J MS Care. 2008;10:14–20.
Fatigue is reported by 75–90% of people with with depression in some studies and is independent in
multiple sclerosis (MS), and up to 50% of others.14,16,18,19 This conflicting evidence highlights the patients indicate fatigue as one of their worst need for a systematic and thorough evaluation and problems.1,2 Fatigue can negatively affect a range of treatment for fatigue.4 The Clinical Practice Guidelines issues, including quality of life and employment.3-5 for Fatigue and Multiple Sclerosis addresses this issue Despite its frequency, fatigue in MS remains elusive, by using a sequential approach, described in a fatigue partly because of the complexity of factors contributing to it and the lack of defined biological markers.2 Definitions
Although identifying the presence of fatigue in indi- viduals with MS is not a problem, methods of measure- Fatigue in MS can be described in two categories: ment are highly subjective, so determining the under- primary fatigue (PF) and nonprimary fatigue (NPF).
lying cause is challenging.4,6 Possible contributory NPF results from other symptoms of MS such as factors to fatigue in MS include lesion site, nerve fiber depression, mobility inefficiency, or respiratory prob- conduction loss, sleep disturbances, physical disability, lems. NPF may be caused by acute situations such as comorbidity, and iatrogenic causes.7-11 Some of these flu, cold, or stress or chronic conditions such as chronic factors, such as sleep disturbances, may be secondary to pain or cardiac problems. NPF may be of short dura- MS-related problems, including bladder dysfunction and leg spasms, that confound the assessment.10,12,13 PF is related directly to the MS disease process, Some studies show fatigue to be present in patients when no NPF factors can be recognized. Because no with early disease and mild disability,14 but others sug- laboratory, scanning, or other procedures that would gest that fatigue is worse with advanced disease or has isolate or identify PF are currently available, the diag- no relationship with disability.2,15-17 Fatigue correlates nosis is one of exclusion. PF persists in MS even whenNPF factors are eliminated or managed and lasts >6months.20 Although no characteristics specific to PF From the MS Clinic, University of British Columbia, Vancouver, BC,Canada.
have been systematically identified, scholars have Primary and Nonprimary Fatigue
observed that heat sensitivity,1,22 time of fatigue onset,3 perature changes, and appear in the absence of other fac- and severity of fatigue6 warrant further investigation.
tors.21 Environmental heat has been suggested to worsenfatigue in MS.1,15,22 The experience and perception of fatigue in MS is unique to each individual and is opti- The primary objective of this study was to use the mally evaluated in the context of daily life.6 Clinical Practice Guidelines to systematically screen a Based on the fatigue algorithm, evaluation and treat- consecutive cohort of patients with MS to evaluate the ment for fatigue in MS initially targets the NPF factors incidence of NPF and PF and the frequency of factors (columns 3–5); when ameliorated, the algorithm that contribute to NPF. The secondary objective was to addresses PF. Unfortunately, the frequency of the vari- determine whether the fatigue experienced was greater ous NPF factors is unknown. As new pharmacological in PF. The final objective was to determine whether PF treatments are developed for MS fatigue,23-26 differentia- had any unique characteristics or descriptors such as tion and quantification of these types of fatigue is cru- heat sensitivity, time of fatigue onset, or relativity to cial before beginning clinical trials. Because this prelimi- nary work has not yet been done, therapeutic trials, Fatigue Algorithm for MS
to date, have included heterogeneous groups of MSpatients (with regard to fatigue), suggesting that caution The fatigue algorithm is a flow chart that represents is needed when generalizing the results of these stud- a process of identification, assessment, and intervention ies.27 Thus, understanding the frequency of NPF factors for the multiple variables associated with fatigue in MS.20 An assumption embedded in the algorithm isthat fatigue as experienced in MS may be associated with several factors. The cause-and-effect relationship Patients were recruited through a posting in the between these factors and MS fatigue is tenuous at best University of British Columbia (UBC) MS clinic, by because of the subjective outcome measures used and invitation from their neurologist, and through MS the inability to rigorously isolate each factor. The Society newsletters. Inclusion criteria for the study were approach of the fatigue algorithm is to identify the age ≥19 years; a clinically or laboratory definite diagno- presence of these factors and account for their possible sis of MS27; clinically stable MS, defined as no relapse contribution to and relationship with fatigue in MS within 30 days; an Extended Disability Status Scale (EDSS) score of <7.5; willingness to provide informed The fatigue algorithm is organized into seven consent; and identification of fatigue as a problem (by columns (Figure 1). Columns 1 and 2 identify fatigue answering the question: Is fatigue a problem for you?).
as a problem and consider any relevant history. Col- Patients were excluded from the study if they had initi- umn 3 confirms and treats or alternatively discounts ated disease-modifying therapy within the past 6 potential acute onset of fatigue that is not related to months, used oral or systemic steroids within the past 4 MS. Column 4 considers any chronic fatigue factors weeks, or been diagnosed with a serious psychiatric or that may or may not be related to MS (eg, sleep prob- medical condition that, in the opinion of their clinic lems, depression, medications, deconditioning), con- neurologist, would preclude reliable participation in comitant medical problems (eg, anemia or thyroid the study. This study was approved by the UBC Behav- disease), and significant chronic stressors. Column 5 identifies any factors that are secondary to MS symp- For this cross-sectional study, the coordinator toms that may result in excessive energy expenditure to screened each potential participant. Eligible patients compensate for reduced mobility or respiratory func- were booked for a 1-hour assessment visit. Six self- tion. Collectively, columns 3–5 deal with NPF as report questionnaires—demographic form, two fatigue defined in this study. In columns 6 and 7, PF of MS is scales, sleep scale, fatigue questionnaire, and sleep ques- addressed, focusing on pharmacological and rehabilita- tionnaire—were provided to participants and complet- ed in advance of the assessment. The visit included a PF has been anecdotally observed to be at its worst in structured interview, review of completed question- early and midafternoon, follow the body’s diurnal tem- naires, evaluation of mobility, and a screen for depres- Fo
F r
o we
w l
e l
l
l e
t
e
t a
l
a .
l
. Fa
1
re
u
Fig
Primary and Nonprimary Fatigue
sion, respiratory function, and pharmaceutical and Fifty-nine individuals with MS initially volunteered The fatigue algorithm was used to guide the screen- to participate in this study. Nine were excluded because ing for PF and NPF.20 The screening tools and of recent relapses (n = 2), presence of a psychiatric questionnaires recommended in the Clinical Practice condition (n = 1), recent initiation of an immune- Guidelines were used.20 Where no measure was recom- modulating therapy (n = 1), and lack of clinically defi- mended, a standardized scale or method was sought or nite MS (n = 1). Three individuals declined involve- developed. The measures used in this study were the ment in the study after initial contact, and one person’s Fatigue Severity Scale (FSS),18 the Modified Fatigue study visit could not be scheduled. Therefore, 50 par- Impact Scale (MFIS),2,20 a fatigue questionnaire,20 ticipants completed the study and formed the cohort.
the Beck Depression Inventory 18 (BDI-18),28,29 the Twenty-four patients (48%) were on disease-modifying Epworth Sleepiness Scale (ESS),30 a sleep question - therapies. Fatigue was ranked as the number one prob- naire,20 the ambulation section of the Functional Inde- lem for 35 participants (70%) and had been present for pendence Measure,31 the dyspnea section of the Chron- at least 6 months in 49 individuals (98%). Demo- ic Respiratory Diseases Questionnaire,32 and EDSS.33 graphic characteristics are shown in Table 1. A retro- All prescription and nonprescription medications spective database search of all patient visits during the were systematically recorded, including careful screen- period of the study (n = 1865) was completed to deter- ing of alcohol and caffeine intake, as well as alternative mine how representative the study cohort was of medications or therapies such as marijuana and herbal the MS clinic population (Table 1). Statistical testing remedies. Each person’s medications were reviewed for showed no significant difference in age, sex, or disease their possible contribution to fatigue. Medications that course between the study participants and patients seen were known to cause drowsiness but were taken atappropriate times (ie, at night), were not deemed to in the MS clinic. The cohort, however, had a lower affect fatigue adversely. Where the medication type, mean EDSS score (suggesting less disability) than the dosage, frequency, timing of dose, or potential interac- clinic population (P < .05).
tion with other medications was suspected to affect The scores from the four standardized question- fatigue, the case was reviewed by the study pharmacist naires (FSS, MFIS, ESS, BDI-18) and the sleep ques- or neurologist. The clinical opinion was used to code tionnaire are summarized in Table 2. Fatigue, as scored the medications as being contributory or noncontribu- on the FSS and MFIS, ranged from mild to severe in tory to fatigue. Note that the fatigue algorithm suggests impact and intensity for study participants. EDSS a longitudinal process of data collection, which was not Table 1. Demographic characteristics of study
clinically feasible when pilot tested. To address this cohort (n = 50)
issue for the study, the self-report measures (first fivelisted above) were completed in advance of the study Characteristic
MS clinic*
visit. The remaining measures were completed during The two principal assessors reviewed all question- naires and scales. Consistent with the Clinical Practice Guidelines algorithm,20 PF was identified and coded in the absence of other factors or when factors were not Statistical analyses were carried out with the Statisti- cal Package for the Social Sciences (version10). Statisti- cal significance was established at P < .05. Pearson χ2 or Mann-Whitney test was used to analyze the differ- ences between groups. Correlation between nominal ordered and interval nonnormal or ordinal variables SD, standard deviation; EDSS, Expanded Disability Status Scale*Patient visits during study. was analyzed using Kendall’s τb correlation.34,35 †Significantly different at P < .05.
Forwell et al.
Table 2. Cohort scores on questionnaires
Table 3. Frequency of factors contributing to
fatigue in multiple sclerosis (MS) study subjects
Questionnaire
n (%)
Mean (SD) Range
Contributing factors
n
Acute factors (eg, heat, infection, temporary routine SD, standard deviation; FSS, Fatigue Severity Scale; MFIS, Modi-fied Fatigue Impact Scale; ESS, Epworth Sleepiness Scale; BDI-18, scores showed no correlation with fatigue scores on Objective 2: Unique Characteristics of PF
either the FSS or MFIS (P = .30 and .24, respectively) Three characteristics were investigated as possible or with sleep problems or depression as measured on discriminators of PF, as suggested in previous literature the ESS and BDI-18 (P = .90 and .25, respectively.) about fatigue and MS1,15,20-22: heat sensitivity, time ofday of fatigue onset, and impact of fatigue. Forty sub- Objective 1: Frequency of PF and NPF
jects (80%) reported fatigue to be exacerbated by heat.
Of the 50 participants in the study, 36 (72%) were However, no difference in fatigue scores was found identified as having fatigue factors that may be related between those reporting that fatigue worsened with to NPF, including acute issues, chronic conditions, and heat and those indicating that heat was not a factor problems secondary to MS symptoms. In the absence (P = .52 and .11, MFIS and FSS, respectively, Mann- of other factors, the remaining 14 patients (28%) were Whitney test). In addition, no difference was found in reported heat sensitivity between the PF (11 of 14 Of those with NPF, 34 experienced more than one [80%]) and NPF (29 of 36 [80%]; P < .05) groups.
factor possibly contributing to their fatigue. The most These results suggest that heat sensitivity is not a frequent NPF factors reported were sleep problems (58% of patients experienced more than two interrup- Statistical analysis revealed that time of fatigue onset tions per night), unmanaged mobility impairments was significantly different between the PF and NPF (52%), and depression (40%). Table 3 shows the fre- groups. Individuals with PF were significantly more quency of each factor that may contribute to NPF, and likely to report onset of fatigue at midday, ie, beginning Table 4 gives details of medications deemed to con- in late morning and early afternoon (Pearson χ2 = 4.77; tribute to NPF. Note that PF may have been present in these subjects but could not be identified in the MFIS fatigue scores were significantly lower in the absence of known specific markers for PF.
PF group than in the NPF group (P = .03, Mann- Table 4. Medications deemed to contribute to fatigue taken by MS study subjects
Medication class
Medication (n)
Amitriptyline (1), citalopram (2), fluvoxamine (1), trazodone (1), venlafaxine (3), St. John’s wort (Hypericum perforatum) (1) Butalbital-codeine compound (2), codeine (1) Benzodiazepines and other hypnotics Oxazepam (1), nitrazepam (1), triazolam (1), zopiclone (1)Skeletal muscle relaxants Gabapentin (2), marijuana (2), oxybutynin (2) Note: All subjects in group (n = 10) were taking multiple medications deemed to contribute to fatigue.
Primary and Nonprimary Fatigue
Whitney test), suggesting that the negative impact of recently, heat was shown to magnify fatigue, with the fatigue is greater for those who experience fatigue from effect more pronounced among those who experienced multiple concurrent NPF factors. No significant differ- severe fatigue and a higher EDSS score.15 This variabili- ences were found between groups for FSS scores (P < ty across studies may be related to the subjective meas- ures used to assess the relationship between heat andfatigue. Alternatively, results may be confounded by the Discussion
inconsistent identification of other NPF factors, shown By using the fatigue algorithm to identify the fre- quency of factors that may contribute to fatigue in a More than 90% of individuals in the NPF group cohort of MS patients, contrary to what has often been experienced multiple factors that could be identified as reported, most of our patients were experiencing fac- a possible cause for their fatigue. Although these factors tors that fit the definition of NPF. This finding is semi- are not considered part of PF, many are associated with nal in addressing fatigue in MS for two reasons: many MS. Patients with MS have a 50% lifetime risk of of the NPF factors can be treated, and, in the absence experiencing depression,36 with point prevalence esti- of correction of such factors, evaluation, treatment, or mates ranging from 14% to 57%.29 This was similar in even study of PF will not be possible. Our study found our group, where 40% of the cohort experienced that less than one third of the cohort had PF in isola- depression. Sleep problems have been reported to be tion; however, PF may have been masked by the pres- three times higher in people with MS than in healthy ence of NPF factors. The true proportion of patients control subjects, which is consistent with our find- experiencing PF remains elusive until factors defining ings.11-13 Mobility problems are common in MS.20 More NPF are treated. Another explanation for the low fre- than half of the patients in our study experienced quency of PF in this study group could be selection mobility problems that, when associated with spastici- bias on the part of neurologists or patients. However, a ty, may have accompanied increased fatigue.37 In con- comparison with the general UBC MS clinic popula- trast to this, others have found no relationship between tion did not indicate any significant clinical or demo- gait and perceived fatigue.38 The tension in the litera- ture suggests that, when various factors contributing to With the definitions and criteria used in this study, a MS fatigue are isolated systematically, their contribu- diagnosis for PF can only result once the NPF factors tion may be identified and targeted for treatment.
have been identified, treated, and resolved. Ideally, a The fact that many of the causes of NPF are treat- positive, discerning definition for PF that includes its able may be the most compelling argument for a thor- own set of characteristics and biological markers will ough systematic screening of these potential contribu- evolve. The results from our study extend the possible tory factors. Such an approach would reduce their list of discriminators for PF. We were able to show that effect on the increased fatigue levels experienced by time of fatigue onset and lower levels of fatigue were patients with NPF.16 Indeed, when depression was unique characteristics of isolated PF. Time of onset was managed effectively, fatigue was reduced.39 This is an consistent with Herndon’s21 observation that fatigue important and urgent issue for clinical practitioners.
scores were significantly higher for those experiencing The therapeutic team must identify and treat factors NPF, which was also supported by Bakshi.6 Our find- that magnify the disabling fatigue experience in MS.
ings suggest that as additional characteristics exclusive Finally, to study fatigue in MS, particularly in inter- to PF emerge, more precise screening measures and tar- vention research and therapeutic trials, the type of geted criteria for diagnosis can be developed. This will fatigue and factors being studied must be explicitly not occur, however, if the current confusion between defined. Although many MS fatigue studies include a screening and control for depression, our study showed Heat sensitivity was not found to be associated pref- that sleep problems and mobility impairment had a erentially with the PF versus the NPF group, nor did it higher frequency than depression, necessitating their relate to increased levels of fatigue. Early research inclusion for screening as confounding variables.15,23 showed conflicting results or that heat worsened fatigue Overlooking this step would weaken the interpretation in MS compared with a healthy control group.1,22 More and generalization of fatigue intervention studies.
Forwell et al.
The modest sample size used in this study warrants 18. Krupp L, LaRocca N, Muir-Nash J, Steinberg AD. The fatigue severi- ty scale: application to patients with multiple sclerosis and systemic cautious generalization of the findings. In addition, lupus erythematosus. Arch Neurol. 1989;46:1121–1123.
recruitment to the study was consecutive and based on 19. Vercoulen JH, Homes OR, Swanink CM, et al. The measurement of fatigue in patients with multiple sclerosis: a multidimensional com- participant response to posters and/or referral by a clin- parison with patients with chronic fatigue syndrome and healthy ical neurologist. This recruitment method may have subjects. Arch Neurol. 1996;53:642–649. 20. Multiple Sclerosis Council for Clinical Practice Guidelines. Clinical biased the study sample. However, the study cohort Practice Guidelines: Fatigue and Multiple Sclerosis: Evidence-BasedManagement Strategies for Fatigue. Washington, DC: Paralyzed was demographically and clinically representative of the 21. Herndon RM. Fatigue in multiple sclerosis. Int J MS Care. 1999; References
22. Krupp B, Alvarez LA, LaRocca NG, Scheinberg LC. Fatigue in MS.
Arch Neurol. 1988;45:435–437.
1. Freal JE, Kraft GH, Coryell JK. Symptomatic fatigue in multiple scle- 23. Rammohan KW, Rosenberg JH, Lynn DG, Blumenfeld AM, Pollak rosis. Arch Phys Med Rehabil. 1984;67:164–168.
CP, Nargaraja HN. Efficacy and safety of modafinil (Provigil) for the 2. Fisk JD, Pantefract A, Ritvo PG, Archibald CJ, Murray TJ. The impact treatment of fatigue in multiple sclerosis: a two centre phase 2 study.
of fatigue on patients with multiple sclerosis. Can J Neurol Sci. J Neurol Neurosurg Psychiatry. 2002;72:179–183.
24. Weinshenker BG, Penman M, Bass B, Ebers GC, Rice GPA. A dou- 3. Amato MP, Ponziani G, Rossi F, Liedl CL, Stefanile C, Rossi L. Quali- ble-blind, randomized, crossover trial of pemoline in fatigue associ- ty of Life in multiple sclerosis: the impact of depression, fatigue, and ated with multiple sclerosis. Neurology. 1992;42:1468–1471. disability. Mult Scler. 2001;7(5):340–344.
25. Krupp LB, Coyle PK, Doscher C, et al. Fatigue therapy in multiple 4. Janardhan V, Bakshi R. Quality of life in patients with multiple sclero- sclerosis: results of a double-blind, randomized, parallel trial of sis: the impact of fatigue and depression. J Neurol Sci. 2002; amantadine, pemoline and placebo. Neurology. 1995;45: 5. Edgley K, Sullivan MJL, Deboux E. A survey of multiple sclerosis.
26. Taus C, Giuliani G, Pucci E, D’Amico R, Solari A. Amantadine for Part 2: determinants of employment status. Can J Rehabil. fatigue in multiple sclerosis (Cochrane Review). Cochrane Database 6. Bakshi R. Fatigue associated with multiple sclerosis: diagnosis, 27. Poser CM, Paty DW, Scheinberg L, et al. New diagnostic criteria for impact and management. Mult Scler. 2003;9:219–227. multiple sclerosis: guidelines for research protocol. Ann Neurol.
1983;13(3):227–231.
7. Colombo B, Boneschi FM, Rossi P, et al. MRI and motor evoked 28. Mohr D, Goodkin DE, Likosky W, Beutler L, Gatto N, Langan MK.
potential findings in nondisabled MS patients with and without Identification of the Beck Depression Inventory items related to multi- symptoms of fatigue. J Neurol. 2000;247:506–509.
ple sclerosis. J Behav Med. 1997;20:407–414. 8. Iraite J, Subira ML, de Castro P. Modalities of fatigue in MS: correla- 29. Mohr D, Boudewyn AC, Goodkin DE, Bostrom A, Epstein L. Com- tion with clinical and biological factors. Mult Scler. 2000;6: parative outcomes for individual cognitive-behavioural therapy, sup- portive-emotional group psychotherapy and sertraline for treatment 9. Sandroni P, Walker C, Starr A. Fatigue in patients with multiple scle- of depression in MS. J Consult Clin Psychol. 2001;69(6):942–949.
rosis. Arch Neurol. 1992;49:517–524. 30. Johns MW. A new method for measuring daytime sleepiness: the 10. Schwid SR, Thorton CA, Paadya S, et al. Quantitative assessment of Epworth Sleepiness Scale. Sleep. 1991;14:540–545.
motor, fatigue and strength in MS. Neurology. 1999;53(4): 31. Hamilton BB, Granger CV, Sherwin FS, et al. A uniform national data system for medical rehabilitation. In: Fuhrer MJ, ed. Rehabilita- 11. Clark C, Fleming J, Li D, Oger J, Klonoff H, Paty D. Sleep distur- tion Outcomes: Analysis and Measurement. Baltimore, MD: bances, depression, and lesion site in patients with multiple sclero- sis. Arch Neurol. 1992;49:641–643. 32. Mahler DA, Horner A. Clinical measurement of dyspnea. In: Mahler 12. Saunders J, Whitham R, Schaumann B. Sleep disturbance, fatigue DA, ed. Dyspnea. Mt. Kisco, NY: Futura Publishing Co.; 1990. and depression in multiple sclerosis [abstract]. Neurology. 33. International Federation of MS Societies. Minimal Record of Disabil- ity for Multiple Sclerosis. New York: National MS Society; 1985. 13. Leo GJ, Rao SM, Bernardin L. Sleep disturbance in MS. Neurology. 34. Bland, M. An Introduction to Medical Statistics. Oxford, UK: Oxford 35. Rothman KJ. No adjustments are needed for multiple comparisons.
14. Schwartz CE, Coulthard-Morris L, Zeng Q. Psychosocial correlates Epidemiology. 1990;1:43–46.
of fatigue in multiple sclerosis. Arch Phys Med Rehabil. 1996; 36. Sadovnick AD, Remick RA, Allen J, et al. Depression and multiple sclerosis. Neurology. 1996;46(3):628–632.
15. Bergamaschi R, Romani A, Versino M, Poli R, Cosi V. Clinical 37. Miller RG, Green AT, Moussari RS, Carson PJ, Weiner MW. Exces- aspects of fatigue in multiple sclerosis. Funct Neurol. 1997; sive muscular fatigue in patients with spastic paraparesis. Neurolo- 16. Bakshi R, Shaikd Z, Miletich R, et al. Fatigue in multiple sclerosis 38. Morris ME, Cantwell C, Vowel SL, Dodd K. Changes in gait and and its relationship to depression and neurologic disability. Mult fatigue from morning to afternoon in people with MS. J Neurol Neu- rosurg Psychiatry. 2002;72(3):361–365.
17. Krupp LB, LaRocca NG, Goodman A, et al. Correlation of fatigue 39. Mohr DC, Hart SL, Goldberg A. Effects of treatment for depression with other measures of disease activity in chronic progressive multi- on fatigue in multiple sclerosis. Psychosom Med. 2003;65: ple sclerosis [abstract]. Neurology. 1988;38(suppl):A381.

Source: http://video.med.ubc.ca/videos/osot/faculty/sf/Primary_and_NonPrimary_Fatigue_in_MS.pdf