Psych

Explaining fatigue in multiple sclerosis: cross-validation of a biopsychosocial model

Melloney L. M. Wijenberg1,3 • Sven Z. Stapert1,3 • Sebastian Köhler2 • Yvonne Bol3

Received: December 14, 2015 / Accepted: May 20, 2016 / Published online: May 28, 2016

� Springer Science+Business Media New York 2016

Abstract Fatigue is a common and disabling symptom in

patients with multiple sclerosis (MS), but its pathogenesis is

still poorly understood and consequently evidence-based

treatment options are limited. Bol et al. (J Behav Med

33(5):355–363, 2010) suggested a new model, which explains

fatigue in MS from a biopsychosocial perspective, including

cognitive-behavioral factors. For purposes of generalization to

clinical practice, cross-validation of this model in another

sample of 218 patients with MS was performed using structural

equation modeling. Path analysis indicated a close and ade-

quate global fit (RMSEA = 0.053 and CFI = 0.992). The

cross-validated model indicates a significant role for disease

severity, depression and a fear-avoidance cycle in explaining

MS-related fatigue. Modifiable factors, such as depression and

catastrophizing thoughts, propose targets for treatment options.

Our findings are in line with recent evidence for the effec-

tiveness of a new generation of cognitive behavioral therapy,

including acceptance and mindfulness-based interventions,

and provide a theoretical framework for treating fatigue in MS.

Keywords Multiple sclerosis � Fatigue � Catastrophizing � Physical disability � Structural equation modelling � Biopsychosocial model

Introduction

Multiple sclerosis (MS) is characterized by a chronic

inflammation of the central nervous system, which results

in demyelination and atrophy, but has an unknown patho-

genesis and an unpredictable course. It is one of the most

common neurological disorder in young adults (Compston

& Coles, 2008) with a prevalence of 0.9 per 1000 (Hirtz

et al., 2007). Patients with MS report a variety of physical

and neuropsychiatric symptoms, with fatigue being the

most frequent and disabling symptom reported: 80–92 %

of patients with MS report fatigue, and 40–69 % rate

fatigue as their most disabling symptom (Brañas et al.,

2000; Giovannoni, 2006; Minden et al., 2006). Fatigue is a

major reason for decreased societal participation and is also

related to disability and poor quality of life.

Unfortunately, the multifactorial pathogenesis of fatigue

in MS is not completely understood, and evidence-based

treatment options remain scarce (Asano et al., 2014; Bol

et al., 2009; Kos et al., 2008; Pucci et al., 2007). Bol et al.

(2010) examined its multifactorial pathogenesis by fitting a

biomedical and a cognitive behavioral model in a sample

of 262 patients with MS using structural equation mod-

elling (SEM). Results showed that both models poorly

explained fatigue in MS, and based on previous research

and the results of their SEM analyses, they formulated a

new model. This final model was an integration of the first

two models, including both biomedical and cognitive-be-

havioral factors, and can be considered as the fatigue

equivalent of the fear-avoidance model of chronic muscu-

loskeletal pain (Crombez et al., 2012; Vlaeyen et al., 1995).

In this integrated model, catastrophizing about fatigue has

a central role: being fueled by depression, it mediated the

relationship between fatigue and fatigue related fear and

avoidance behavior (Bol et al., 2010).

& Yvonne Bol y.bol@zuyderland.nl

1 Faculty of Psychology and Neuroscience, Maastricht

University, Maastricht, The Netherlands

2 Faculty of Health, Medicine and Life Sciences, School for

Mental Health and Neuroscience, Maastricht University,

Maastricht, The Netherlands

3 Department of Medical Psychology/Academic MS Center

Limburg, Zuyderland Medical Center, PO Box 5500,

6130 MB Sittard-Geleen, The Netherlands

123

J Behav Med (2016) 39:815–822

DOI 10.1007/s10865-016-9749-3

Catastrophizing about fatigue is defined as a fearful

interpretation of the meaning of fatigue by exaggerated

negative thinking, magnification of symptoms, and help-

lessness (e.g. ‘fatigue is terrible and I think it can never

improve’ or ‘when I feel tired, there is nothing I can do to

decrease its intensity’) (Lukkahatai & Saligan, 2013). If

fatigue is erroneously interpreted as a sign of pathology

over which one has little or no control, this could gradually

extend to a fear and avoidance of physical activities and

subsequently decreased physical abilities. According to the

fear-avoidance model, this would then lead to an increase

in fatigue concluding its cyclic pattern. Lukkahatai and

Saligan (2013) showed in their systematic review a con-

sistent strong positive correlation between catastrophizing

and fatigue severity in several clinical conditions that share

fatigue as one of their core symptoms, such as multiple

sclerosis, chronic fatigue syndrome, fibromyalgia and

cancer.

Besides the role of catastrophizing and fear-avoidance

behavior, previous research has shown a significant asso-

ciation between depression and fatigue in patients with MS,

independent of physical disability (Bakshi et al., 2000).

With regard to the direction of influence, a longitudinal

study of Patrick et al. (2009), including 2768 patients with

MS, showed that depression was one of the most important

predictors of fatigue at 1-year follow-up. With regard to

disease severity, Hadjimichael et al. (2008) showed a sig-

nificant positive correlation between disease severity and

fatigue in patients with MS, explaining that more physical

disability and neurological impairment are associated with

higher levels of fatigue.

This biopsychosocial model of Bol et al. (2010) inte-

grates these individual observations in a single model of

fatigue in MS, however cross-validation is necessary to

make a valid generalization and application to everyday

clinical practice possible. In the present study, we

hypothesize that the associations between fatigue, depres-

sion, catastrophizing and disease severity described by the

biopsychosocial model will explain fatigue in another large

group of MS patients. This cross-validation is important for

the understanding of the origin and perpetuating of fatigue

in patients with MS and will provide a theoretical frame-

work for treating fatigue in patients with MS.

Methods

Participants

Participants were recruited from hospital databases of the

department of Neurology of the Zuyderland Medical

Center in Sittard-Geleen, the Netherlands. A total of 621

Dutch-speaking patients with clinically definite MS

according to McDonald criteria (Polman et al., 2005), aged

between 18 and 65 years, were eligible for inclusion. Their

treating neurologist sent the initial letters to secure confi-

dentiality. A total of 403 patients were interested in par-

ticipating and responded (65 % response rate). These

patients were sent an information letter, an informed con-

sent and questionnaires. A total of 312 participants returned

the forms (77 % response rate). Questionnaires were filled

in between May 2011 and September 2011. Participants

who previously participated in the study of Bol et al.

(2010) (N = 86) were excluded. Informed consent was

obtained from all participants included in the study.

Patients did not receive any financial compensation for

their participation.

Measures

Basic demographic information

Age, gender, level of education, employment status, mar-

ital status and use of psychopharmacological drugs were

obtained by a demographic inventory filled in by the

patients. The level of education was based on the highest

completed level of education and divided into three cate-

gories: primary school (low level of education); junior

vocational training (middle level of education); senior

vocational training or academic training (high level of

education). Medical data, such as disease duration, disease

course, MS subtype and disease severity were collected

from the hospital databases.

Disease severity

Disease severity was assessed with the Expanded Disability

Status Scale (EDSS) (Kurtzke, 1983). This scale comprises

the evaluation of 8 functioning systems (pyramidal, cere-

bellar, brainstem, mental, bowel and bladder, visual-optic,

sensory and other). The EDSS score, based on the evalu-

ation of an experienced neurologist, ranges from 0 to 10,

where 0 indicates a normal neurological examination and

10 indicates death due to MS. Recent EDSS scores

(\3 months) were extracted from the hospital database.

Physical disability

Physical disability was assessed with the physical dimen-

sion of the SF-36, a Dutch translation of the Short Form

Health Survey developed and validated by Aaronson et al.

(1998). Bol et al. (2010) showed a high reliability of this

measure in patients with MS. It consists out of four sub-

scales; physical functioning, role limitations due to physi-

cal health problems, bodily pain, and general health. Each

816 J Behav Med (2016) 39:815–822

123

standardized subscore of the physical dimension ranges

from 0 to 100, where a total score of 400 resembles optimal

physical health and no physical disability.

Fear avoidance

Fear avoidance was assessed with the fatigue version of the

Tampa Scale for Kinesiophobia (TSK-F) (Silver et al.,

2002), which is an adjusted version of the TSK for chronic

pain (Miller et al., 1991; Vlaeyen et al., 1995). Silver et al.

(2002) replaced in all 17 items the word ‘pain’ by the word

‘fatigue’ to make the questionnaire suitable for investiga-

tion of fatigue-related fear and avoidance behavior. The

score ranges from 17 to 68, where a higher score indicates

a higher level of fear-avoidance behavior. This instrument

is found to be valid (Silver et al., 2002) and reliable in

patients with MS (Bol et al., 2010; Silver et al., 2002).

Catastrophizing

Catastrophizing about fatigue was assessed with the Fati-

gue Catastrophizing Scale (FCS), which is an adapted

version of the Pain Catastrophizing Scale (PCS) (Sullivan

et al., 1995). Psychometric properties of the PCS are ade-

quate (Osman et al., 2000). The PCS consists out of 13

items measuring the self-reported frequency of catastro-

phizing thoughts about experienced pain. As with the TSK

adaptation, Bol et al. (2010) adapted all the PCS items by

replacing the word ‘pain’ by the word ‘fatigue’. Scoring

alternatives ranged from ‘strongly disagree’ to ‘strongly

agree’. As in the study of Bol et al. (2010), three MS-

related items were added (‘When I am tired, this is a signal

there is something wrong in my brain’, ‘When I am tired,

this is a warning for physical decline’, ‘When I am tired,

this is a sign that my MS is getting worse’). In total 16

items were administered and the score ranges from 0 to 64

with higher scores indicating higher intensity of catastro-

phizing. Bol et al. (2010) showed a high reliability of this

measure in patients with MS. In the current sample the

reliability was excellent (a = 0.94).

Fatigue

Fatigue was assessed with the Abbreviated Fatigue Ques-

tionnaire (AFQ), a valid and reliable instrument (Alberts

et al., 1997). Administration to patients with MS also

revealed its reliability (Bol et al., 2010). This questionnaire

is a selection of four items of the Checklist Individual

Strength (CIS-20) developed by Vercoulen et al. (1999).

Items are rated on a 7-point Likert scale with scoring

alternatives ranging from ‘Yes, that is true’ to ‘No, that is

not true’. The final score ranges from 4 till 28, with higher

scores indicating a higher severity of physical fatigue.

Depression

Depression was assessed with the subscale depression of

the Hospital Anxiety and Depression Scale (HADS) (Zig-

mond & Snaith, 1983), a valid and reliable screening

instrument for patients with MS (Honarmand & Feinstein,

2009). The total score ranges from 0 to 21 with a higher

score indicating a higher intensity of depression. Honar-

mand and Feinstein (2009) showed that patients with MS

with a score of 8 or higher are likely depressed.

Statistical analyses

Data analyses were performed using SPSS 22.0.0.0 for

Windows (SPSS Inc., Chicago, IL). If less than 25 % of the

items of questionnaires, or more than 50 % if a question-

naire consisted of four items, were missing, missing values

were imputed by the mean of the remaining non-missing

items of the scale (27 values across 24 participants).

Descriptive statistics were used to describe the sample. No

variable was significantly skewed (skewness \-1 or [1) nor were there any significant outliers (all cases were

within 1.5 interquartile ranges from the upper or lower

quartile). Cronbach’s alpha was used to test reliability of

all questionnaires. Relations between all variables were

analyzed by Pearson-correlations. An alpha level of .05

was used for all statistical tests.

Cross-validation was analyzed with structural equation

modeling in Mplus 7 (Muthén & Muthén, 1998–2012). The

biopsychosocial model of Bol et al. (2010) was specified in a

path analysis using manifest variables only (no measurement

model). Error terms were assumed to be uncorrelated and left

free. The Root Mean Square Error of Approximation

(RMSEA) was used as a global fit index, because parsimony

and sample size are taken into account. RMSEA represents

the lack of fit in comparison with a perfect fit and should

therefore be low. RMSEA values up to 0.05 indicate a close

fit, values between 0.05 and 0.08 indicate an acceptable fit,

values between 0.08 and 0.10 indicate a mediocre fit, and

those greater than 0.10 indicate a poor fit. Furthermore, the

comparative fit index (CFI) was used, because it represents

the relative improvement of the model in comparison with a

baseline model, usually a model in which all observed

variables are uncorrelated. Values larger than 0.95 indicate a

good fit and values between 0.90 and 0.95 indicate an

acceptable fit. Furthermore, the Chi square test of model fit,

Standardized Root Mean Square Residual (SRMR) and

Tucker–Lewis Index (TLI) were also reviewed as fit indexes.

A non-significant Chi square test of model fit indicates a

J Behav Med (2016) 39:815–822 817

123

good fit. SRMR values smaller than .08 indicate an accept-

able fit, whereas values smaller than 0.05 indicate a good fit.

TLI values higher than .90 are acceptable and values higher

than .95 represent a good fit. To control for possible nor-

mality assumption violation, a robust maximum likelihood

estimator for standard errors, also known as the ‘Huber

Sandwich Estimator’, was used (Huber, 1967). Modification

indices were inspected to consider further fine-tuning of the

model to the data-at-hand in an exploratory fashion. Finally,

direct and total effects of the significant variables were cal-

culated.

Results

Patient sample

A total of two participants were excluded due to too many

missing values ([25 % of items of questionnaires missing). Finally, six participants were excluded due to a missing

value in the single exogenous variable, EDSS, which was

necessary for proper structural equation modeling (SEM)

analysis. This resulted in a final sample of 218 outpatients

(53 men, 165 women) with an average age of 48.0 years

(SD = 10.5, range 19–65). Most of them had a relapsing

remitting disease course (n = 153), while 43 patients had a

secondary progressive disease course and 21 patients had a

primary progressive disease course (1 missing value). The

mean disease duration was 8.8 years (SD = 7.5, range

0–30 years) with an average EDSS score of 3.6 (SD = 1.9,

range 0.5–8.0), which resembles a moderate disease

severity. Around 24 % of the sample showed high levels of

catastrophizing, using the cutoff score of 30 as suggested

by Sullivan et al. (1995) for patients with pain. Around

34 % of the sample showed high levels of fear avoidance,

using the cutoff score of 37 as suggested by Vlaeyen et al.

(1995) for patients with pain. See Table 1 for a summary of

all patient characteristics.

Reliability and correlations

Table 2 resembles means, standard deviations, ranges,

reliability indexes (Cronbach’s alphas) for all measures and

their intercorrelations (Pearson). All questionnaires had a

satisfactory internal consistency (range 0.69–0.94). All

intercorrelations were statistically significant (p \ 0.01) with the strongest correlation between depression and

physical disability. Higher levels of depression were

associated with lower levels of physical ability (r = -0.58,

p \ 0.001). The weakest correlation was found between disease severity and catastrophizing about fatigue

(r = 0.21, p \ 0.01).

Structural equation modeling analyses

Figure 1 shows the results of the path analysis of the new

model proposed by Bol et al. (2010). The RMSEA value

was 0.053 (90 % CI 0.000–0.112), which indicates an

acceptable fit. The SRMR, CFI and TLI value were

respectively 0.023, 0.992 and 0.979, indicating a good fit.

The Chi square test of model fit was non-significant

(p = 0.138) also indicating a good fit. Furthermore, all

hypothesized relationships were statistically significant.

The total explained variance of fatigue measured with the

AFQ was 44 %. All variables provided a significant con-

tribution to this explained variance. Both depression

(b = .27) and physical disability (b = -.45) were directly associated with fatigue. There were no modification

indexes given, suggesting that no alternative specification

of relationships between the variables were identified

which could improve the model. We added a relationship

from disease severity to depression, due to its significance

in the second model postulated by Bol et al. (2010), but this

worsened the global fit of our model and was subsequently

removed. Moreover, we ran an additional post hoc analysis

to study the variance in fatigue explained by the fear

avoidance cycle. For this, we omitted the paths to and from

depression and disease severity (see Fig. 1) from the

model. This showed that physical disability, fear-avoid-

ance, catastrophizing and their underlying associations

explain 39 % of the variance in fatigue, compared with

Table 1 Patient characteristics (n = 218)

Variable Value

Gender % female (n) 76 (165)

Age in years [mean (SD)] 48.0 (10.5) range 19.6–65.6

Disease duration in years [mean

(SD)]

8.8 (7.5) range 0.1–30.2

Disease course

Relapsing remitting (%) 71

Secondary progressive (%) 20

Primary progressive (%) 9

Use of disease modifying drugs

(% yes, % no)

61/39

Use of psychopharmaca (% yes,

% no)

25/75

Level of education (% low, %

middle, % high)

24/37/39

Marital status (% partner, % no

partner)

82/28

Employment status (% working,

% not working)

32/68

818 J Behav Med (2016) 39:815–822

123

44 % of the total model. See Table 3 for an overview of the

standardized direct, indirect and total effects on fatigue.

Discussion

Due to the high prevalence of fatigue in patients with MS

and its disabling impact on everyday activities and quality

of life, understanding its pathogenesis and identifying its

modifiable contributing factors are crucial. Bol et al. (2010)

showed that neither a biomedical nor a cognitive-behav-

ioral model explained fatigue in 262 patients with MS, but

suggested a new biopsychosocial model integrating ele-

ments of the previously tested models, i.e. disease severity,

depression and fear-avoidance cycle. To generalize and

apply this model to everyday clinical practice, cross-vali-

dation of this integrated model in another sample was

needed. We hypothesized that the biopsychosocial model

of Bol et al. (2010) can explain fatigue in MS in another

large sample.

Table 2 Means, standard deviations (SD), ranges, Cronbach’s alphas (a) and Pearson-correlations of all measures

Mean (SD) Range a 2 3 4 5 6

1. Disease severity (EDSS) 3.6 (1.9) 0.5–8 – .23** .21* .22** .29** -.48**

2. Fatigue (AFQ) 19.7 (6.8) 4–28 0.90 – .55** .34** .54** -.63**

3. Catastrophizing about fatigue (FCS) 19.9 (14.1) 0–56 0.94 – – .58** .57** -.55**

4. Fatigue-related fear and avoidance (TSK-F) 34.3 (8.3) 20–68 0.73 – – – .41** -.42**

5. Depression (HADS-D) 6.0 (4.0) 0–17 0.82 – – – – -.58**

6. Physical disability (SF-physical) 208.5 (92.1) 25–400 0.69 – – – – –

EDSS Expanded Disability Status Scale, AFQ Abbreviated Fatigue Questionnaire, FCS Fatigue Catastrophizing Scale, TSK-F Fatigue Version of

the Tampa Scale for Kinesiophobia, HADS-D depression subscale of the Hospital Anxiety and Depression Scale, SF-physical Physical scale of

the Short Form Health Survey

* p \ 0.01; ** p \ 0.001

Fig. 1 Path analysis of the biopsychosocial model of fatigue

in multiple sclerosis (n = 218).

Note Values shown are

standardized regression

coefficients and based on cross-

sectional data. Light blue

variables and its relationships

represent the fear-avoidance cycle

within the model. Explained

variances are provided in

parentheses. Please note that the

scale of physical disability is

inverted. *p \ 0.05; **p \ 0.01; ***p \ 0.001 (Color figure online)

J Behav Med (2016) 39:815–822 819

123

The SEM analyses presented in this study, explaining

fatigue in a new sample of 218 patients with MS, showed

good support of the biopsychosocial model of Bol et al.

(2010). Catastrophizing, depression, physical disability,

disease severity and fear avoidance all contribute signifi-

cantly to fatigue, either directly or indirectly. Comparing

the results to that of the original publication, the global fit

indices RMSEA and CFI even slightly improved respec-

tively from 0.085 towards 0.053 and from 0.983 towards

0.992. This implies an increase in fit from mediocre to

acceptable (RMSEA) or even good (CFI).

The biopsychosocial model indicates a significant role

for disease severity, depression and an adapted fear

avoidance model in explaining MS-related fatigue. This

integrated model partly overlaps with a recently formulated

model by Wu et al. (2015) explaining post-stroke fatigue.

They suggest also an integration of biological and psy-

chological variables, including depressive symptoms,

coping and behavioral factors. Also in stroke patients, an

intervention including CBT elements showed a long term

reduction in fatigue (Zedlitz et al., 2012). Moreover,

Zedlitz et al. (2012) stated that the addition of graded

activity to the cognitive elements, which focuses on

improvement of physical disability, resulted in a longer

endurance of the fatigue reducing effects.

Translating the biopsychosocial model of Bol et al.

(2010) to clinical practice in MS, the model indicates

several modifiable factors, such as the fatigue-enhancing

cycle of fear avoidance and depression, which form

important targets for interventions. Diagnosing and treating

depression could be a first step to treat MS related fatigue.

Depression is with a life-time prevalence of approximately

50 % very prevalent in MS and probably underdiagnosed

and untreated (Feinstein, 2011; Maier et al., 2015). When

depression is treated, for instance with cognitive behavioral

therapy (CBT) (Hind et al., 2014), it is likely that fatigue is

also reduced. Next, CBT focusing on changing catastro-

phizing thoughts about fatigue could help fatigued MS

patients (Knoop et al., 2011; Moss-Morris et al., 2012; van

Kessel et al., 2008). Knoop et al. (2011) concluded that

changes in thoughts about fatigue play a crucial role in

CBT for fatigue in MS. Hoogerwerf et al. (submitted)

showed that also the third generation CBT, Mindfulness

Based Cognitive Therapy (MBCT) is an effective inter-

vention for severely fatigued MS patients. Patients were

not only less fatigued after MBCT, but also less depressed

and less catastrophizing about fatigue. This suggests that

catastrophizing can be reduced not only by altering the

content of thoughts such as in regular CBT, but even by

disengaging from the maladaptive thoughts about fatigue.

There are several limitations to this study, which should

be taken into account when interpreting the results and

could be addressed in future studies. First of all, the design

is cross-sectional making it impossible to draw firm con-

clusions about causality and temporal relations in the dis-

ease process. More prospective and longitudinal studies are

needed to confirm the proposed causal relationships. Sec-

ondly, postal questionnaires were used which made us

unable to compare responders with non-responders. The

response rate was favorable (77 %), but lower in compar-

ison with Bol et al. (2010) (93 % response rate). A possible

explanation could be related to the fact that more ques-

tionnaires were included which demanded more time and

energy of the participants. As a result, we cannot exclude

the possibility of a selection bias. Thirdly, all data were

self-reported and are therefore sensitive to retrospective

bias and response styles. Fourthly, our main outcome

measure, the AFQ, is a questionnaire consisting out of four

items. Despite its sufficient validity and reliability, Hore-

mans et al. (2004) argued that the AFQ lacks precision at

the individual patient level. Future studies should include

fatigue questionnaires which are validated in MS patients,

such as the Fatigue Severity Scale or the Modified Fatigue

Impact Scale (Rietberg et al., 2010). Finally, other factors,

some even modifiable, such as sleep disorders, cognitive

impairments and maladaptive coping styles, were not

assessed and therefore lacking in the biopsychosocial

model. Their inclusion could increase the explained vari-

ance of the model due to their previously established

influences on fatigue in MS (Rabinowitz & Arnett, 2009;

Strober & Arnett, 2005; Ukueberuwa & Arnett, 2014).

Furthermore, the overall anxiety level and other distorted

Table 3 Standardized direct, indirect and total effects on fatigue

Variable Direct Indirect Total

Fear-avoidance (TSK-F) 0.000 0.103** 0.103**

Physical disability (SF-physical) -0.447*** -0.173*** -0.620***

Depression (HADS-D) 0.274*** 0.024* 0.298***

Disease severity (EDSS) 0.000 0.288*** 0.288***

Catastrophizing (FCS) 0.000 0.054* 0.054*

TSK-F Fatigue Version of the Tampa Scale for Kinesiophobia, SF-physical Physical scale of the Short Form Health Survey, HADS-D depression

subscale of the Hospital Anxiety and Depression Scale, EDSS Expanded Disability Status Scale, FCS Fatigue Catastrophizing Scale

* p \ 0.05; ** p \ 0.01; *** p \ 0.001

820 J Behav Med (2016) 39:815–822

123

cognitive thinking habits besides catastrophizing, in which

elements of rumination, magnification and helplessness are

embedded (Sullivan et al., 1995), could possibly be another

useful addition for future studies due its modifiable char-

acter and insight in effective therapeutic elements.

Despite these limitations, this cross-validation of the

biopsychosocial model of Bol et al. (2010) forms an

important next step in explaining MS-related fatigue and

highlights a promising role for CBT. The integrated model

supports the clinical practice guidelines that both biological

and psychological factors should be taken into account

during the clinical assessment and treatment of fatigue in

MS (CBO, 2013; Van Kessel & Moss-Morris, 2006). It is

expected that development and evaluation of targeted

psychological interventions will help improving the

biopsychosocial model of MS related fatigue.

Acknowledgments We would like to thank all the patients who took part in this study; the therapists, psychological assistants and MS

nurses of Zuyderland Medical Center; Dr. Myreen Moors for her

effort in gathering and monitoring the data acquisition; Prof. Dr.

Raymond Hupperts for his kind cooperation and time investment.

Compliance with ethical standards

Conflict of interest Melloney L. M. Wijenberg, Sven Z. Stapert, Sebastian Köhler and Yvonne Bol declare that they do not have any

conflict of interest.

Human and animal rights and Informed consent All procedures were approved by and in accordance with the ethical standard of the

medical ethics committee of Zuyderland Medical Center and with the

1964 Helsinki declaration and its later amendments. Informed consent

was obtained from all patients for being included in the study.

References

Aaronson, N. K., Muller, M., Cohen, P. D., Essink-Bot, M.-L.,

Fekkes, M., Sanderman, R., et al. (1998). Translation, validation,

and norming of the Dutch language version of the SF-36 Health

Survey in community and chronic disease populations. Journal

of Clinical Epidemiology, 51, 1055–1068.

Alberts, M., Smets, E., Vercoulen, J., Garssen, B., & Bleijenberg, G.

(1997). ‘Verkorte vermoeidheidsvragenlijst’: een practisch hulp-

middel bij het scoren van vermoeidheid. Nederlands Tijdschrift

voor Geneeskunde, 141, 1526–1530.

Asano, M., Berg, E., Johnson, K., Turpin, M., & Finlayson, M. L.

(2014). A scoping review of rehabilitation interventions that

reduce fatigue among adults with multiple sclerosis. Disability

and Rehabilitation, 37(9), 729–738.

Bakshi, R., Shaikh, Z., Miletich, R., Czarnecki, D., Dmochowski, J.,

Henschel, K., et al. (2000). Fatigue in multiple sclerosis and its

relationship to depression and neurologic disability. Multiple

Sclerosis, 6, 181–185.

Bol, Y., Duits, A. A., Hupperts, R. M., Vlaeyen, J. W., & Verhey, F. R.

(2009). The psychology of fatigue in patients with multiple

sclerosis: A review. Journal of Psychosomatic Research, 66, 3–11.

Bol, Y., Duits, A. A., Lousberg, R., Hupperts, R. M., Lacroix, M. H.,

Verhey, F. R., et al. (2010). Fatigue and physical disability in

patients with multiple sclerosis: A structural equation modeling

approach. Journal of Behavioral Medicine, 33, 355–363.

Brañas, P., Jordan, R., Fry-Smith, A., Burls, A., & Hyde, C. (2000).

Treatments for fatigue in multiple sclerosis: A rapid and

systematic review. Health Technology Assessment (Winchester,

England), 4, 1.

CBO. (2013). Richtlijn multipele sclerose 2012. Houten: Bohn Stafleu

Van Loghum.

Compston, A., & Coles, A. (2008). Multiple sclerosis. The Lancet,

372, 1502–1517. doi:10.1016/S0140-6736(08)61620-7

Crombez, G., Eccleston, C., Van Damme, S., Vlaeyen, J. W., &

Karoly, P. (2012). Fear-avoidance model of chronic pain: The

next generation. The Clinical Journal of Pain, 28, 475–483.

Feinstein, A. (2011). Multiple sclerosis and depression. Multiple

Sclerosis Journal, 17, 1276–1281.

Giovannoni, G. (2006). Multiple sclerosis related fatigue. Journal of

Neurology, Neurosurgery and Psychiatry, 77, 2–3.

Hadjimichael, O., Vollmer, T., & Oleen-Burkey, M. (2008). Fatigue

characteristics in multiple sclerosis: The North American

Research Committee on Multiple Sclerosis (NARCOMS) sur-

vey. Health and Quality of Life Outcomes, 6, 100. doi:10.1186/

1477-7525-6-100

Hind, D., Cotter, J., Thake, A., Bradburn, M., Cooper, C., Isaac, C.,

et al. (2014). Cognitive behavioural therapy for the treatment of

depression in people with multiple sclerosis: A systematic

review and meta-analysis. BMC Psychiatry, 14, 5.

Hirtz, D., Thurman, D., Gwinn-Hardy, K., Mohamed, M., Chaudhuri,

A., & Zalutsky, R. (2007). How common are the ‘‘common’’

neurologic disorders? Neurology, 68, 326–337.

Honarmand, K., & Feinstein, A. (2009). Validation of the Hospital

Anxiety and Depression Scale for use with multiple sclerosis

patients. Multiple Sclerosis, 15(12), 1518–1524. doi:10.1177/

1352458509347150.

Hoogerwerf, A. E., Bol, Y., Lobbestael, J., Hupperts, R. M., & van

Heugten, C. M. Mindfulness based cognitive therapy is feasible

and effective in severely fatigued patients with Multiple

Sclerosis: A waiting list controlled study (submitted).

Horemans, H. L., Nollet, F., Beelen, A., & Lankhorst, G. J. (2004). A

comparison of 4 questionnaires to measure fatigue in postpo-

liomyelitis syndrome. Archives of Physical Medicine and

Rehabilitation, 85, 392–398.

Huber, P. J. (1967). The behavior of maximum likelihood estimates

under nonstandard conditions. Paper presented at the proceed-

ings of the fifth Berkeley symposium on mathematical statistics

and probability.

Knoop, H., Van Kessel, K., & Moss-Morris, R. (2011). Which

cognitions and behaviours mediate the positive effect of

cognitive behavioural therapy on fatigue in patients with

multiple sclerosis? Psychological Medicine, 42, 205–213.

Kos, D., Kerckhofs, E., Nagels, G., & D’Hooghe, M. (2008). Origin

of fatigue in multiple sclerosis: Review of the literature.

Neurorehabilitation and Neural Repair, 22, 91–100.

Kurtzke, J. F. (1983). Rating neurologic impairment in multiple

sclerosis an expanded disability status scale (EDSS). Neurology,

33, 1444.

Lukkahatai, N., & Saligan, L. N. (2013). Association of catastro-

phizing and fatigue: A systematic review. Journal of Psychoso-

matic Research, 74, 100–109.

Maier, S., Balasa, R., Buruian, M., Maier, A., & Bajko, Z. (2015).

Depression in multiple sclerosis—Review. Romanian Jounal of

Neurology, 14, 22.

Miller, R. P., Kori, S. H., & Todd, D. D. (1991). The tampa scale.

Unpublished report. Tampa, FL.

Minden, S. L., Frankel, D., Hadden, L., Perloff, J., Srinath, K. P., &

Hoaglin, D. C. (2006). The Sonya Slifka longitudinal multiple

J Behav Med (2016) 39:815–822 821

123

sclerosis study: Methods and sample characteristics. Multiple

Sclerosis, 12, 24–38. doi:10.1191/135248506ms1262oa

Moss-Morris, R., Dennison, L., Landau, S., Yardley, L., Silber, E., &

Chalder, T. (2012). A randomized controlled trial of cognitive

behavioral therapy (CBT) for adjusting to multiple sclerosis (the

saMS trial): Does CBT work and for whom does it work?

Journal of Consulting and Clinical Psychology, 81, 251–262.

doi:10.1037/a0029132

Muthén, L. K., & Muthén, B. O. (1998–2012). Mplus user’s guide

(7th ed.). Los Angeles, CA: Muthén & Muthén.

Osman, A., Barrios, F. X., Gutierrez, P. M., Kopper, B. A., Merrifield,

T., & Grittmann, L. (2000). The Pain Catastrophizing Scale:

Further psychometric evaluation with adult samples. Journal of

Behavioral Medicine, 23, 351–365.

Patrick, E., Christodoulou, C., & Krupp, L. (2009). Longitudinal

correlates of fatigue in multiple sclerosis. Multiple Sclerosis, 15,

258–261.

Polman, C. H., Reingold, S. C., Edan, G., Filippi, M., Hartung, H. P.,

Kappos, L., et al. (2005). Diagnostic criteria for multiple

sclerosis: 2005 revisions to the ‘‘McDonald Criteria’’. Annals

of Neurology, 58, 840–846.

Pucci, E., Branas, P., D’Amico, R., Giuliani, G., Solari, A., & Taus,

C. (2007). Amantadine for fatigue in multiple sclerosis.

Cochrane Database of Systematic Reviews. doi:10.1002/

14651858.CD002818.pub2.

Rabinowitz, A. R., & Arnett, P. A. (2009). A longitudinal analysis of

cognitive dysfunction, coping, and depression in multiple

sclerosis. Neuropsychology, 23, 581.

Rietberg, M., Van Wegen, E., & Kwakkel, G. (2010). Measuring

fatigue in patients with multiple sclerosis: Reproducibility,

responsiveness and concurrent validity of three Dutch self-report

questionnaires. Disability and Rehabilitation, 32, 1870–1876.

Silver, A., Haeney, M., Vijayadurai, P., Wilks, D., Pattrick, M., &

Main, C. (2002). The role of fear of physical movement and

activity in chronic fatigue syndrome. Journal of Psychosomatic

Research, 52, 485–493.

Strober, L. B., & Arnett, P. A. (2005). An examination of four models

predicting fatigue in multiple sclerosis. Archives of Clinical

Neuropsychology, 20, 631–646.

Sullivan, M. J., Bishop, S. R., & Pivik, J. (1995). The pain

catastrophizing scale: Development and validation. Psycholog-

ical Assessment, 7, 524.

Ukueberuwa, D. M., & Arnett, P. A. (2014). Evaluating the role of

coping style as a moderator of fatigue and risk for future

cognitive impairment in multiple sclerosis. Journal of the

International Neuropsychological Society, 20, 751–755.

Van Kessel, K., & Moss-Morris, R. (2006). Understanding multiple

sclerosis fatigue: A synthesis of biological and psychological

factors. Journal of Psychosomatic Research, 61, 583–585.

van Kessel, K., Moss-Morris, R., Willoughby, E., Chalder, T.,

Johnson, M. H., & Robinson, E. (2008). A randomized

controlled trial of cognitive behavior therapy for multiple

sclerosis fatigue. Psychosomatic Medicine, 70, 205–213.

Vercoulen, J., Alberts, M., & Bleijenberg, G. (1999). De checklist

individuele spankracht (CIS). Gedragstherapie, 32, 131–136.

Vlaeyen, J. W., Kole-Snijders, A. M., Rotteveel, A. M., Ruesink, R.,

& Heuts, P. H. (1995). The role of fear of movement/(re) injury

in pain disability. Journal of Occupational Rehabilitation, 5,

235–252.

Wu, S., Mead, G., Macleod, M., & Chalder, T. (2015). Model of

understanding fatigue after stroke. Stroke, 46, 893–898. Zedlitz, A. M., Rietveld, T. C., Geurts, A. C., & Fasotti, L. (2012).

Cognitive and graded activity training can alleviate persistent

fatigue after stroke a randomized, controlled trial. Stroke, 43,

1046–1051.

Zigmond, A. S., & Snaith, R. P. (1983). The hospital anxiety and

depression scale. Acta Psychiatrica Scandnavica, 67, 361–370.

822 J Behav Med (2016) 39:815–822

123

Journal of Behavioral Medicine is a copyright of Springer, 2016. All Rights Reserved.