Hayter 2016

Download as pdf or txt
Download as pdf or txt
You are on page 1of 16

1

British Journal of Clinical Psychology (2016)


© 2016 The British Psychological Society
www.wileyonlinelibrary.com

The impact of health anxiety in patients with


relapsing remitting multiple sclerosis:
Misperception, misattribution and quality of life
Aimee L. Hayter1*, Paul M. Salkovskis2, Eli Silber3 and
Robin G. Morris1
1
Department of Psychology, King’s College, Institute of Psychiatry, London, UK
2
Department of Psychology, Bath University, UK
3
King’s College Hospital NHS Foundation Trust, London, UK
Objectives. Multiple sclerosis (MS) is a progressive disease with an unpredictable
prognosis. Previous studies have reported health anxiety within the MS population. This
study examines the effect of health anxiety on MS patients’ quality of life (QoL) and
evaluates the potential contribution of cognitive factors in maintaining health anxiety.
Methods. A total of 84 patients with relapsing remitting multiple sclerosis (RRMS) were
screened for health anxiety. From this sample, a group with relatively high and another
group with low anxiety (n = 21 in each group) were identified. A further 21 healthy controls
were recruited for comparison. A measure of QoL was then completed. Cognitive biases
were investigated by measuring perception and attribution of common bodily symptoms as
well as appraisal of performance on neuropsychological and physical fatigue tests.
Results. The high health anxiety group reported poorer QoL relative to the other
groups, independent of level of disability. They were also more likely to misattribute
common bodily changes to MS, and perceive their (objectively intact) performance on
tests of cognition and fatigue as being impaired, attributing the cause of impairment to MS.
Conclusion. Health anxiety may be a factor in mediating the psychosocial impact of MS.
Skilled psychological treatment which changes misperception and misattribution may
significantly benefit patients with MS and elevated health anxiety.

Practitioner Points
Clinical implications
 Health anxiety impacts on quality of life in patients with MS even when disability and other measures of
psychological distress are taken into account.
 High levels of health anxiety distort perceptions of symptoms in patients with MS in line with the
predictions made by the cognitive model of health anxiety.

Limitations of study
 This study is limited to patients with RRMS within the relatively early stages of their disease and is based
on a small sample size.
 Health anxiety is correlated with measures of generalized anxiety, depression, and worry, although it is
found to have a unique impact on quality of life in patients with MS.

*Correspondence should be addressed to Aimee L. Hayter, Institute of Psychiatry, Psychology & Neuroscience, King’s College
London, 16 De Crespigny Park, London SE5 8AF, UK (email: [email protected]).

DOI:10.1111/bjc.12106
2 Aimee L. Hayter et al.

Relapsing remitting multiple sclerosis (RRMS) is a degenerative neurological condition.


Symptoms can be extremely unpredictable over multiple domains, including sensory,
cognitive, and physical abilities. It has been suggested previously that the inability to
predict what symptoms will emerge as the condition progresses may contribute to mental
health problems such as depression and anxiety, thus lowering QoL (Arnett, Barwich, &
Beeney, 2008; Dennison, Moss-Morris, & Chalder, 2009). Negative illness perceptions and
cognitive biases (such as catastrophizing) have been found to be associated with poor
psychological outcomes in patients with MS, including general distress, anxiety, and
depression (Dennison, Moss-Morris, Silber, Galea, & Chalder, 2010; Jopson & Moss-
Morris, 2003; Osborne, Jensen, Ehde, Hanely, & Draft, 2007; Skerrett & Moss-Morris, 2006;
Taillefer, Kirmayer, Robbins, & Lasry, 2002).
Reduced QoL in patients with MS is reported, relative to the general population
(Morales-Gonzalez, Benito-Leon, Rivera-Navarro, & Mitchell, 2004; Nortvedt, Riise, &
Myhr, 1999) and in comparison with patients suffering from other chronic diseases
(Sprangers, de Regt, & Andries, 2000). This cannot be explained by the impact of physical
disability alone, as the association between physical disability and QoL ranges from absent
to slight (Correlation range: 0.02–0.29; Benedict, Wahlig, & Bakshic, 2005; Merkelbach,
Sittinger, & Koenig, 2002; Patti, Caopardo, & Palermo, 2003; Solari, Palmisano, &
Mendozzi, 1999). It has been found that physical disability is not associated with QoL after
adjusting for anxiety and depression within the earlier stages of RRMS (Jassens et al.,
2003). At this early stage of the disease, when patients experience a lower level of physical
disability, symptoms may be vague, with a variety of manifestations. The cause of
symptoms can be ambiguous, possibly prompting patients to question their source.
Hence, psychological factors may account for more of the variance in QoL scores than
physical disability (Benedict et al., 2005; Jassens et al., 2003; Kern et al., 2009;
Merkelbach et al., 2002).
Although it has been previously suggested that levels of anxiety, depression, and
distress mediate this correlation (Dennison et al., 2009), what is unclear is the
mechanism by which these difficulties become elevated in patients with MS and hence
impact QoL. The cognitive model of health anxiety (Salkovskis, 1996) suggests that
psychological distress may be triggered by the tendency to misinterpret ambiguous
stimuli with respect to illness. Specifically, the presence of relatively enduring beliefs
about health interacts with health-relevant information (such as perceived bodily
variations) to produce a sense of threat (Salkovskis & Warwick, 1986; Warwick &
Salkovskis, 1990). Vulnerable individuals have relatively low thresholds for making
threat-related (mis)interpretations and misattributions. These cognitive biases result in
anxiety and drive a range of behavioural, attentional, physiological, and/or affective
responses, which can maintain threat relevant appraisals and intensify the experience of
bodily variations and symptoms.
This process has been found to occur in anxious patients with physical health
problems. For example, anxious patients with cardiac problems have been found to make
catastrophic somatic attributions (e.g., that they are having a heart attack) for bodily
changes/sensations (e.g., shortness of breath) rather than normalizing attributions (e.g.,
exertion) for sensations congruent with their underlying condition (Ratcliffe, MacLeod, &
Sensky, 2006).
In patients with MS, this is also likely to be the case. Previous studies have found clinical
levels of health anxiety in 25% of patients with MS (Kehler & Hadjistavropoulos, 2009). A
relationship between maladaptive illness perceptions and levels of general anxiety has
been demonstrated in this population (Jopson & Moss-Morris, 2003; Vaughan, Morrison,
Health anxiety and its impact on QOL in RRMS 3

& Miller, 2003). A concept related to health anxiety, namely ‘illness worry’, has been
found to be more strongly correlated with a tendency to attribute ambiguous bodily
changes to somatic causes (e.g., ‘a relapse in symptoms’) in patients with MS compared to
patients with other illnesses with an unpredictable course (chronic fatigue syndrome;
Taillefer et al., 2002). This result was found despite the MS group showing subclinical
levels of illness worry. Further, studies involving patients with MS have demonstrated that
distress is associated with catastrophic interpretations of common physical sensations
(Dennison et al., 2010; Osborne et al., 2007; Skerrett & Moss-Morris, 2006). However,
these previous studies have not demonstrated whether patients’ distress is associated
with changes over other domains which are commonly affected by MS, such as physical
fatigue and cognition.
This study examines the extent to which poor QoL is a function of health anxiety, here
adopting the strategy of comparing patients with MS with high levels of health anxiety
(HiHAMS) with those who have lower levels (LoHAMS). In addition, it evaluates whether a
HiHAMS group misperceive and misattribute bodily sensations with ambiguous causes
more than a LoHAMS group, and whether the tendency to make misinterpretations is
specific to MS congruent bodily sensations (e.g., feeling shaky), or more generalized to
include MS incongruent bodily sensations (e.g., a pounding heart). The study compares
HiHAMS and LoHAMS patients’ subjective perception (or misperception) of their
performance on objectively measured neuropsychological tasks and tests of physical
fatigue, and the extent to which they erroneously attribute their perceived performance
to their MS.

Method
Participants
Patients with RRMS were sought for this study from 13 weekly specialist MS neurology
outpatient clinics in the south-eastern region of England (south-east London, Kent, and
East Sussex). Of the 188 patients approached, 32 declined to participate and 72 did not
meet the inclusion criteria which were as follows: Aged over 18 years, diagnosed less
than 10 years ago, and low physical disability levels, defined as having a Hauser
Ambulation Index score of 5 or below (the Hauser Ambulation Index is a semi-
quantitative scale for ranking ambulation disability on a scale of 1–9, with higher scores
indicating more physical disability; Hauser et al., 1983). Applying these criteria resulted
in 84 patients entering the study. These participants were then divided into two
subgroups. The first group (LoHAMS) consisted of 21 patients with low levels of health
anxiety (a score of 10 or less on the Health Anxiety Inventory; HAI). The second group
(HiHAMS) consisted of 21 patients with high levels of health anxiety (a score of 18 or
more on the HAI). These groups were compared on various measures detailed below,
with additional comparisons (age, gender, education levels) with a healthy control
group (HC). The HC participants were recruited from the community in the same
region as the outpatient clinics and were not diagnosed with any serious conditions/
illnesses.
For all participants, written informed consent was obtained. Ethical approval was
sought and granted from the appropriate ethics committee (Wandsworth Research Ethics
Committee, REF: 10/H0803/5). No major ethical concerns were identified, and strategies
were discussed regarding the nature of advice given to participants who had high levels of
health anxiety.
4 Aimee L. Hayter et al.

Measures
Measures differed in their purpose, as follows: Some characterized the overall sample and
subsamples (health anxiety, anxiety, depression, and disability). QoL was specifically
tested as an outcome measure. Participants were also asked to report bodily sensations
and complete two neuropsychological tests and a measure of physical fatigue. These tests
were included to provoke cognitive biases such as misperception and misattribution. The
resulting cognitive biases were then captured by further self-report measures.

Health anxiety
A modified version of the HAI was used, this being a short self-report questionnaire. In
previous studies, a cut-off score of 18 or greater has been used to identify clinical levels of
health anxiety, with a score of between 15 and 17 identifying those with high, but not
clinical levels of health anxiety (Seivewright et al., 2004). Within the current study,
participants with a score of 10 or below were identified as having low levels of health
anxiety, as this score was one standard deviation below that of 15. The HAI is a reliable and
valid measure in the general population and in those with physical health problems
(Salkovskis, Rimes, Warwick, & Clark, 2002), and the modified HAI has been developed
for use with MS patients (Alberts, Sharpe, Kehler, & Hadjistavropoulos, 2011; Kehler &
Hadjistavropoulos, 2009).

Anxiety and depression


Anxiety, depression, and worry (Hospital Anxiety and Depression Scale – HADS; Penn
State Worry Questionnaire – PSWQ) were self-assessed by the MS participants (Molina &
Borkovec, 1994; Zigmond & Snaith, 1983). The HADS has been validated for use in the MS
population (Camp et al., 2005). The PSWQ is a 15-item self-report questionnaire, with
scores over 62 indicating the clinical cut-off for generalized anxiety disorder. This
questionnaire has been used to study worry in patients with MS (Bruce & Arnett, 2009).

Disability
Patients’ level of disability was assessed to characterize the samples and ascertain whether
differences in QoL between those with high and low levels of health anxiety might be
accounted for by differences in the extent of their disability. The Guys Neurological
Disability Scale (GNDS; Sharrack & Hughes, 1999) was chosen as a widely used
comprehensive measure assessing 12 different domains of disability associated with MS
including cognition, mood, vision, communication, swallowing, mobility, bladder,
bowels, sexual functioning, and fatigue. The severity of disability in each domain is
scored from 0 to 5, a higher score indicating greater disability.

Quality of life
Quality of life was measured by the Quality of Life Index (QLI; Ferrans & Powers, 1992).
This assesses participants’ satisfaction in different life areas, and also their importance.
The ratings of importance were used to weight ratings of satisfaction, with higher scores
indicating higher QoL. The QLI has been developed for use with various physical health
conditions including MS and has good levels of reliability and validity in this population
(Stuifbergen, 1995).
Health anxiety and its impact on QOL in RRMS 5

Misperception and misattribution


The study aimed to measure subjective perception and attribution of ‘symptoms’ over
three domains commonly affected by MS: Bodily sensations, cognition, and fatigue. The
tasks used to provoke misperception and misattribution over these domains are detailed
below. Once these tasks had been completed, the participants recorded their perceptions
of performance and attributions for performance on a separate measure.

Bodily sensations. The modified Symptom Interpretation Task (SIT; modified from
Robbins & Kirmayer, 1991; see below for modifications made by Ratcliffe et al., 2006) was
used to measure misattribution of common bodily sensations to underlying pathology.
Robbins and Kirmayer (1991) developed the Symptom Interpretation Questionnaire (SIQ)
to measure the type and number of attributions given for common bodily sensations,
reporting good reliability and validity. In this study, this measure was adapted using the
Ratcliffe et al. (2006) method; participants generate causal attributions to six bodily
sensations selected from the Anxiety Sensitivity Index (ASI; Taylor & Cox, 1998), three
congruent with MS symptoms and three incongruent. Participants listed as many causal
attributions as they could for each symptom within 60 s.
The attributions were then categorized to one of four attribution types: Normalizing
(e.g., ‘due to a cold’), psychological (e.g., ‘because I’m anxious’), somatic/illness (e.g.,
‘because I have a life threatening illness’; as per Robbins & Kirmayer, 1991), or a more
specific somatic/illness category – that of MS-related attributions (e.g., ‘because of my
MS’). This category was collapsed into an overall ‘somatic/illness’ category for analysis.
A scoring manual was derived from the literature (Robbins & Kirmayer, 1991; Sensky,
MacLeod, & Rigby, 1996) to aid in categorization of the attributions into the first three
categories. The fourth category was described in terms of published research on MS
symptoms. Categorization was performed by a clinical psychologist unaffiliated to the
authors and blind to participants’ group allocations. To establish inter-rater reliability, 25
randomly chosen SIT questionnaires (40%) consisting of 432 separate causal attributions
were rated by an independent rater, blind to categorizations made by the first rater, but
with access to the scoring manual. The inter-rater reliability was high (j = .93 (p < .01),
95% CI (0.897, 0.960)).

Cognition. The Brixton Spatial Anticipation Test (Brixton; Burgess & Shallice, 1997) and
the Symbol Digit Modality Test (SDMT; Smith, 1982) were used, both having a number of
demand characteristics which may indicate to the participant that they are failing when in
fact they have intact performance. The Brixton test measures executive functioning,
performance requiring planning, problem-solving, and mental flexibility. The total
number of task errors was used to measure performance. The SDMT measures information
processing speed with higher scores indicating faster information processing speed. It
was not expected that the participants with MS would show a deficit on either test, as
patients with RRMS tend to show intact performance if within 10 years of diagnosis
(Summers et al., 2007).

Physical fatigue. A hand grip dynamometer, which measures physical fatigue/


endurance, was used as it has a number of demand characteristics which may indicate
to the participant that they are failing when in fact they have intact performance. The
6 Aimee L. Hayter et al.

protocol applied was based on one used in a study investigating attributions made for
performance using the hand grip dynamometer in people with chronic pain (Rode,
Salkovskis, Dowd, & Hanna, 2006). Participants were asked to grip the dynamometer as
hard as they could, and then to hold a third of this grip for as long as they could (hence
calibrating the dynamometer for the strength of each participant). The length of time the
grip was held was a measure of physical endurance on the task.

Perception and attribution of task performance. After each of the Brixton, SDMT, and
hand grip dynamometer tasks, participants completed two self-report measures. First,
they rated their performance in comparison with that of others on a scale of 50
(extremely badly in comparison with others) to 50 (extremely well in comparison with
others). Second, they attributed their performance to MS by rating how much better they
would have done on the task if they did not have MS, using a scale of 0 (no better) to 100
(very much better). Only the patients with MS completed this second measure.

Procedure
Participants completed three self-report questionnaires (HAI, HADS, and PSWQ) as they
waited for their neurology appointment. The remaining self-report measures (QLI) were
completed not more than 2 weeks later. During this session, the GNDS was completed by
the lead author. The patients then performed the SIT, SDMT, Brixton, and hand grip
dynamometer tasks in a pseudorandom and counterbalanced order. After each task,
participants completed measures of performance evaluation, worry about perceived
performance, and attribution of performance to MS.
Healthy control participants completed both studies in one session and completed all
questionnaires and tasks except those pertaining to MS.

Statistical analysis
The HiHAMS, LoHAMS, and (where appropriate) HC groups were compared using
analysis of variance (ANOVA) and analysis of covariance (ANCOVA) for three-group
analyses, and two-sided t-tests for two-group comparisons. Where within-subject
elements were included in analyses, mixed-model ANOVAs were used, with simple main
effects being used to interpret interactions when present. Where the epsilon coefficient
was significant, Greenhouse–Geisser corrections were used. Alpha level was set to <.05
for statistical significance for all tests. Where multiple comparisons were used for the
three-group analyses, corrections were by means of Bonferroni tests. All two-group
comparisons were two-sided.

Results
Overview
First, the effect of health anxiety on quality of life was analysed, taking into account the
level of disability in the sample. Second, actual and perceived cognitive and physical task
performance was analysed for low and high health anxiety MS groups, and compared to
that of the healthy control group. Secondly, attributions in relation to MS congruent and
incongruent symptoms were analysed.
Health anxiety and its impact on QOL in RRMS 7

Health anxiety levels and demographics


The total sample of 84 patients with MS had a mean score of 13.7 on the HAI (SD = 5.38)
(Table 1). Within this sample, 29% of patients scored 18 and above on the HAI, this cut-off
score indicating clinically significant levels of health anxiety (Seivewright et al., 2004).
The second part of the study was then completed by 21 patients with MS who scored 18
and above on the HAI (HiHAMS), and 21 who scored 10 and below on the HAI (LoHAMS).
These groups were not significantly different in age, t(40) = 0.11, p > .1, time since
diagnosis, t(40) = 0.00, p > .1, gender, or disability status on the GNDS, t(40) = 0.60,
p > .1. Differences were found between the groups on the HADS-A, t(40) = 6.25, p < .01,
HADS-D, t(40) = 3.15, p < .01, and PSWQ, t(40) = 4.16, p < .01.

Effect of health anxiety on quality of life


A one-way ANOVA comparing all three groups on QLI measured QoL showed a significant
main effect, F(2,60) = 20.34, p < .001. Bonferroni comparisons showed that HiHAMS
participants reported significantly lower QoL than LoHAMS, t(60) = 5.16, p < .001,
d = 1.56, and HC, t(60) = 5.83, p < .001, d = 1.73, whereas LoHAMS and HC
participants did not differ, t(60) = 0.67, p > .05, d = 0.22.
It is possible that the group differences in QoL reflected differences in levels of
objective disability experienced by the two groups, with high health anxiety reflecting
greater impairment. To evaluate whether health anxiety was associated with poorer QoL
over and above level of disability, the GNDS (as an objective measure of disability) was
used as a covariate in the group comparison. The group effect remained significant in this
ANCOVA, F(1,39) = 24.74, p < .001. Planned contrasts indicated that the HiHAMS group
still reported significantly lower QoL, t(40) = 4.97, p < .001, d = 1.49; See Table 1.
ANCOVAs showed that group differences in QoL were still found when controlling for
general level of depression, general anxiety, and worry; HADS-A, F(1,39) = 4.11, p < .01,
HADS-D, F(1,39) = 11.95, p < .01, and PSWQ, F(1,39) = 9.44, p < .01.

Objective performance on cognitive and physical endurance tasks


Cognitive performance
SDMT and Brixton scores were used in a 2 (SDMT scores, Brixton scores) 9 3 (HiHAMS,
LoHAMS, HC) mixed-model ANOVA. There was no main effect of group on the cognitive
performance scores, F(2,60) = 1.00, p > .1. There was a main effect of test on cognitive
performance, F(1,60) = 1263.86, p < .01, accounted for by the different score ranges for
SDMT and Brixton. There was no interaction effect, group X test; F(2,60) = 0.51, p > .05.
These results indicate that the groups were comparable in actual cognitive performance.

Physical fatigue
A one-way ANOVA indicated no significant effect of groups on length of time (in seconds)
the participants held their grip on the dynamometer, F(2,60) = 0.21, p > .1; see Table 2).

Perceived performance on cognitive and physical fatigue tasks


A 3 (HiHAMS, LoHAMS, HC) by 3 (performance ratings on Brixton, SDMT, hand grip
dynamometer) mixed-model ANOVA indicated that there was a significant main effect of
8
Aimee L. Hayter et al.

Table 1. Descriptive statistics for QLI score (standard deviations)

QLI (adjusted
Gender Years since HADS-A HADS-D PSWQ GNDS QLI for GNDS)
Group Female, Male Age diagnosis Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD)

HiHAMS 17,4 42.90 (11.24) 5.09 (3.14) 10.76 (3.11) 6.14 (3.64) 60.62 (11.80) 13.29 (5.22) 14.61 (5.05) 14.73 (4.48)
LoHAMS 17,4 42.56 (9.55) 5.10 (3.33) 4.24 (3.63) 2.76 (3.32) 41.86 (16.94) 12.29 (5.62) 21.75 (4.13) 21.63 (4.48)
HC 17,4 40.52 (10.69) 22.67 (4.22)
Health anxiety and its impact on QOL in RRMS 9

Table 2. Descriptive statistics for performance on SDMT, Brixton, and Hand Grip scores

SDMT Brixton Hand grip


Group Mean (SD) Mean (SD) Mean (SD)

HiHAMS 51.57 (8.62) 15.14 (3.37) 40.76 (51.39)


LoHAMS 54.00 (6.24) 14.95 (3.68) 48.52 (47.32)
HC 53.62 (4.55) 15.71 (3.27) 41.47 (25.83)

group on perceived task performance, F(2,60) = 18.22, p < .01. There was also a
significant main effect of task type, F(2,120) = 5.35, p < .01. The interaction between
group and task was not significant, F(4,120) = 1.17, p > .05. Multiple comparisons by
group indicated a significant difference between the HiHAMS group and both the LoHAMS
and HC groups (Bonferroni p < .01). Perceived task performance rated by LoHAMS and
HC participants was not significantly different (p > .05), indicating that the HiHAMS
group rated their performance as being poorer relative to the other groups (see Table 3).

Attribution of performance on cognitive and physical fatigue tasks to MS


Only those participants with MS rated how much better they would have performed the
three tasks if they did not have MS. A 2 (HiHAMS, LoHAMS) by 3 (MS attribution ratings for
performance on SDMT, Brixton, hand grip dynamometer) mixed-model ANOVA showed a
significant main effect of attribution to MS, F(1,40) = 20.65, p < .01. There was also a
significant main effect of task on MS attribution, F(1.62,64.94) = 5.28, Greenhouse–
Geisser p < .05. The task and group interaction was not significant, F(1.62,64.94) = 0.05,
Greenhouse–Geisser p > .05. These results suggest that the HiHAMS group was more
likely to attribute their perceived poorer task performance to MS than the LoHAMS group
(see Table 3).

Attributions of congruent and incongruent bodily sensations to MS


To compare differences in the number of normalizing, psychological, and somatic
attributions given by the three groups for MS congruent and MS incongruent symptoms, a
3 9 3 9 2 mixed-model ANOVA was conducted, with the following factors: Group,
attributions (normalizing, psychological, and somatic), and congruence (MS congruent
and MS incongruent). There was no main effect of group, F(1,60) = 2.50, p > .05. There
was a significant main effect of attribution type, F(2,120) = 19.24, p < .001, and
congruence, F(1,60) = 20.32, p < .001. There were also significant second-order
interactions between group and attribution type, F(3.63,109.10) = 8.25, p < .001, and
symptom type and attribution, F(1.61, 96.33) = 40.29, p < .001. However, all of these
effects were modified by a significant third-order interaction effect, between group,
attribution, and congruence, F(3.21, 96.33) = 5.91, p < .01.
Having identified the predicted third-order interaction, planned comparisons were
made. It had been hypothesized that the HiHAMS participants would generate more
somatic attributions than psychological and normalizing attributions. Therefore, these
data were summarized into the ratio of somatic: Psychological + normalizing attributions.
The attribution ratio generated for congruent symptoms was then compared to the
attribution ratio generated for incongruent symptoms. A 3 9 2 mixed-model ANOVA was
conducted on this transformed data, with group and congruence (congruent attribution
ratio, incongruent attribution ratio) as factors. There were significant main effects of
10
Aimee L. Hayter et al.

Table 3. Descriptive statistics for perceived performance and attribution to MS on cognitive and physical fatigue tests

Perceived performance (scale 50 to 50) Attributed to MS (scale 0 to 100)

SDMT Brixton Handgrip Mean across Brixton Handgrip Mean across


Group Mean (SD) Mean (SD) Mean (SD) tasks (SD) SDMT Mean (SD) Mean (SD) Mean (SD) tasks Mean (SD)

HiHAMS 15.47 (16.95) 12.38 (14.11) 24.29 (21.81) 17.3 (11.85) 53.57 (31.39) 47.85 (28.40) 63.09 (38.42) 54.84 (23.36)
LoHAMS 2.86 (15.94) 3.81 (9.21) 0.00 (19.24) 2.2 (14.8) 22.14 (27.95) 13.33 (21.06) 30.76 (30.21) 22.08 (23.36)
HC 4.05 (11.58) 0.95 (7.68) 1.43 (14.59) 1.2 (13.4) – – – –
Health anxiety and its impact on QOL in RRMS 11

4
HiHAMS
3.5
LoHAMS
3
HC
2.5

Means
2

1.5

0.5

0
Incongruent Congruent
Symptoms

Figure 1. Attribution ratio for MS congruent and incongruent symptoms by group.

groups, F(1,60) = 22.53, p < .001, and congruence, F(1,60) = 27.91, p < .001. These
effects were modified by a significant interaction between attribution and symptom type,
F(2,60) = 18.00, p < .001. Figure 1 shows this interaction. A simple main effects analysis
for groups, followed by within-subject paired t-tests, was carried out to clarify further.
One-way ANOVAs were significant for both MS congruent symptoms, F(2,62) = 20.54,
p < .001, and MS incongruent symptoms, F(2,62) = 11.56, p < .001. Multiple compar-
isons between groups for congruent symptoms showed that the attribution ratio was
significantly higher for the HiHAMS group, relative to both LoHAMS, t(41) = 5.16,
p < .01, and HC, t(41) = 5.87, p < .001; Bonferroni p < .05. LoHAMS and HC groups did
not differ from each other on the attribution ratio for congruent symptoms (p > .05). This
same pattern of significance was found for the MS incongruent symptoms between the
groups. Finally, three within-group paired t-tests were conducted to contrast the
attribution ratios for congruent and incongruent symptoms, using a Bonferroni-corrected
alpha level of .016. The HiHAMS group gave a significantly higher attribution ratio for
congruent symptoms relative to incongruent symptoms, t(20) = 4.73, p < .001. LoHAMS
and HC groups were not significantly different in this comparison (p > .016).
These findings show that high health anxious MS patients interpret bodily sensations
negatively compared to those with lower health anxiety and healthy controls; they are
particularly likely to misinterpret MS congruent bodily sensations in this way.

Discussion
This study found that health anxiety makes an important contribution to the psychosocial
impact of MS. Clinically significant health anxiety was found in 29% of the sample of
patients with RRMS, a rate comparable to that found in previous studies of patients with
MS (Kehler & Hadjistavropoulos, 2009). When relatively high and relatively low health
anxiety groups were compared, there was no evidence of ‘objective’ differences in
impairment and disability. However, there was clear evidence that the high health anxious
group misperceived their objectively intact cognitive and physical fatigue performance as
impaired due to MS. This group were also experiencing significantly poorer QoL. In terms
of the psychological mechanisms believed to be important in the experience of health
anxiety, it was found that, relative to comparison groups, patients with MS and high levels
of health anxiety were more likely to make somatic attributions of ambiguous symptoms
12 Aimee L. Hayter et al.

congruent with MS. In this high-anxiety group, somatic misinterpretations were found for
illness congruent symptoms to a substantial extent. It should be noted that they were also
found for incongruent symptoms, although to a lesser extent than for congruent ones.
Whilst it can be expected that RRMS will eventually result in a high burden of disability
which may limit work and social roles, leading to understandable distress and poorer QoL
(Dennison et al., 2009), the present study may indicate that such outcomes can emerge
ahead of any significant disability in patients experiencing high levels of health anxiety.
Similar results have been noted in some previous studies concerning the negative impact
of anxiety and depression on QoL in patients with MS (Jassens et al., 2003; Lerdal, Celius,
& Moum, 2009; Spain, Turbridy, Kilpatrick, Adams, & Holmes, 2007; Wynia, Middel, van
Dijk, De Keyser, & Reijneveld, 2008). Lerdal et al. (2009) and Spain et al. (2007) found
that patients with negative views about their illness and prognosis (e.g., ‘illness identity’)
reported poorer QoL, particularly in the domains of role and social function.
Consistent with previous studies (Taillefer et al., 2002), the current study found that
health anxious MS patients tended to attribute ambiguous bodily sensations to threatening
somatic causes (particularly those with bodily sensations congruent with symptoms of
MS). These findings went beyond attribution of verbally reported bodily sensations,
demonstrating that MS patients with high health anxiety also misinterpret ‘behavioural’
signs. Patients with high levels of health anxiety not only perceived themselves to be
impaired in terms of cognitive and physical fatigue during experimental tasks where their
performance was normal, but attributed these perceived ‘deficits’ to MS. People with a
diagnosis of MS and high health anxiety may catastrophize around MS congruent
symptoms because they could herald deterioration.
There are some limitations of the current study. Firstly, it consisted of a relatively small
sample size in each of the three groups and the findings should be treated as preliminary
until replicated. Secondly, although this current study suggests that health anxiety has a
unique impact on quality of life in patients with MS, this differs from the predictions of
other models of adjustment to long-term health conditions. These suggest a wide range of
cognitive, behavioural, and social/environmental factors which may interact to produce
what they describe as ‘poor adjustment’ in patients with MS (Dennison et al., 2010). In
line with these models, the current study found health anxiety was correlated with levels
of generalized anxiety, depression, and worry, although health anxiety was correlated
with poorer quality of life in patients with MS over and above these other factors.
Potentially, the most important implication of the present findings lies in the
possibility for targeted treatment of the cognitive biases shown to be associated with
health anxiety in patients with MS. There is accumulating evidence that cognitive–
behavioural approaches are effective for the treatment of these cognitive biases reported
by patients with physical health difficulties and comorbid health anxiety (Seivewright
et al., 2004; Tyrer et al., 2011). Cognitive–behavioural therapy (CBT) may be adapted for
use with MS patients by allowing for the diagnosis of MS to be incorporated into the
formulation. For example, patients may be engaged in treatment by exploring their
reaction to this unpredictable disease as understandably amplified by anxiety. It may also
be important to fully assess patients’ beliefs about illness causation, as well as their
experience and assumptions about the emotional, physical, and social consequences of
their condition, prognosis, and treatment prospects (Dennison et al., 2009). This may
indicate some generalized dysfunctional assumptions, about illness in general and MS
specifically, which may drive the rest of the health anxiety response. Within this context,
an alternative explanation could be presented – that the difficulties reported are caused by
worry about MS worsening, rather than the MS itself. This explanation may provide a less
Health anxiety and its impact on QOL in RRMS 13

threatening account for patients to further explore and evaluate in the later stages of
therapy. Taking account of specific misperception and misattribution, particularly for MS
congruent changes, in the way previously described for chest pain patients, may also be
appropriate (Jonsbu, Dammen, Morken, Moum, & Martinsen, 2011; Salkovskis, 1992).
Adapting CBT in this way for people with MS has been shown to reduce distress,
depression, and anxiety (Askey-Jones, David, Silber, Shaw, & Chalder, 2013; Knoop, van
Kessel, & Moss-Morris, 2012; Moss-Morris et al., 2013). Specifically, cognitive variables,
such as symptom focusing and believing symptoms are a sign of damage, have been found
to change significantly after CBT in patients with MS and high levels of fatigue (Knoop
et al., 2012). The other possibility is that the cognitive and physical tasks used here could
be deployed as behavioural experiments, in which patients are provided feedback about
actual impairment, similar to the use of actigraphy feedback in insomnia (Tang & Harvey,
2006), or use of biofeedback in fatigued patients with MS (Mackay et al., 2015).
In summary, the findings of this study indicate that some patients with MS experience
high levels of health anxiety and this may impact on QoL. In particular, this health anxiety
is associated with cognitive misperception and misattribution, in line with the cognitive
model of health anxiety (Salkovskis, 1996). There is some evidence that these factors are
highly amenable to CBT in both physically healthy and MS samples, offering patients with
RRMS the chance of improving their QoL even in the face of an unpredictable neurological
condition with a relatively poor prognosis.

References
Alberts, N. M., Sharpe, D., Kehler, M. D., & Hadjistavropoulos, H. D. (2011). Health anxiety:
Comparison of the latent structure in medical and non-medical samples. Journal of Anxiety
Disorders, 25, 612–614. doi:10.1016/j.janxdis.2011.01.011
Arnett, P. A., Barwich, F. H., & Beeney, J. E. (2008). Depression in multiple sclerosis: Review and
theoretical proposal. Journal of International Neuropsychological Society, 14, 691–724.
doi:10.1080/13803395.2015.1063591
Askey-Jones, S., David, A. S., Silber, E., Shaw, P., & Chalder, T. (2013). Cognitive behavioural therapy
for common mental disorders in people with multiple sclerosis: A bench marking study.
Behaviour Research and Therapy, 51, 648–655. doi:10.1016/j.brat.2013.04.001
Benedict, R. H. B., Wahlig, E., & Bakshic, R. (2005). Predicting quality of life in multiple sclerosis:
Accounting for physical disability, fatigue, cognition, mood disorder, personality and behaviour
change. Journal of the Neurological Sciences, 231, 29–34. doi:10.1016/j.jns.2004.12.009
Bruce, J. M., & Arnett, P. A. (2009). Clinical correlates of generalized worry in multiple sclerosis.
Journal of Clinical and Experimental Neuropsychology, 31, 698–705. doi:10.1080/
13803390802484789
Burgess, P., & Shallice, T. (1997). The Hayling and Brixton Tests. Test manual. Bury St Edmunds,
UK: Thames Valley Test Company.
Camp, S. J., Stevenson, V. L., Thompson, A. J., Ingle, G. T., Miller, D. H., & Borras, C. (2005). A
longitudinal study of cognition in primary progressive multiple sclerosis. Brain, 128, 2891–
2898. doi:10.1093/brain/awh602
Dennison, L., Moss-Morris, R., & Chalder, T. (2009). A review of psychological correlates of
adjustment in patients with multiple sclerosis. Clinical Psychology Review, 9(2), 141–153.
doi:10.1016/j.cpr.2008.12.001
Dennison, L., Moss-Morris, R., Silber, E., Galea, I., & Chalder, T. (2010). Cognitive and behavioural
correlates of different domains of psychological adjustment in early-stage multiple sclerosis.
Journal of Psychosomatic Research, 69, 353–361. doi:10.1016/j.jpsychores.2010.04.009
Ferrans, C. E., & Powers, M. J. (1992). Psychometric assessment of the Quality of Life Index.
Research in Nursing and Health, 15, 29–38. doi:10.1002/nur.4770150106
14 Aimee L. Hayter et al.

Hauser, S. L., Dawson, D. M., Lehrich, J. R., Beal, M. F., Kevy, S. V., & Propper, R. D. (1983). Intensive
immunosuppression in progressive multiple-sclerosis - a randomized, 3-arm study of high-dose
intravenous cyclophosphamide, plasma-exchange. New England Journal of Medicine, 308,
173–180. doi:10.1056/NEJM198301273080401
Jassens, A. C. J. W., van Doom, P. A., de Boer, J. B., Kalkers, N. F., van der Meche, F. G. A., Passchier, J.,
& Hintzen, R. Q. (2003). Anxiety and depression influence the relation between disability status
and quality of life in multiple sclerosis. Multiple Sclerosis, 9, 397–403. doi:10.1191/
1352458503ms930oa
Jonsbu, E., Dammen, T., Morken, G., Moum, T., & Martinsen, E. W. (2011). Short-term cognitive
behavioral therapy for non-cardiac chest pain and benign palpitations: A randomized controlled
trial. Journal of Psychosomatic Research, 70, 117–123. doi:10.1016/j.jpsychores.2010.09.013
Jopson, N. M., & Moss-Morris, R. (2003). The role of illness severity and illness representations in
adjusting to multiple sclerosis. Journal of Psychosomatic Research, 54, 503–511. doi:10.1016/
S0022-3999(02)00455-5
Kehler, M. D., & Hadjistavropoulos, H. D. (2009). Is health anxiety a significant problem for
individuals with multiple sclerosis? Journal of Behavioural Medicine, 32(2), 150–161.
doi:10.1007/s10865-008-9186-z
Kern, S., Schrempf, W., Schneider, H., Schultheiss, T., Reichmann, H., & Ziemssen, T. (2009).
Neurological disability, psychological distress, and health-related quality of life in MS patients
within the first three years after diagnosis. Multiple Sclerosis, 15, 752–758. doi:10.1177/
1352458509103300
Knoop, H., van Kessel, K., & Moss-Morris, R. (2012). Which cognitions and behaviours mediate the
positive effect of cognitive behavioural therapy on fatigue in patients with multiple sclerosis?
Psychological Medicine, 42, 205–213. doi:10.1017/S0033291711000924
Lerdal, A., Celius, E. G., & Moum, T. (2009). Perceptions of illness and its development in
patients with multiple sclerosis: A prospective cohort study. Journal of Advances Nursing, 65
(1), 184–192. doi:10.1111/j.1365-2648.2008.04862.x
Mackay, A. M., Buckingham, R., Schwartz, R. S., Hodgkinson, S., Beran, R. G., & Cordato, D. J. (2015). The
effect of biofeedback as a psychological intervention in multiple sclerosis: A Randomized controlled
study. International Journal of MS Care, 17(3), 101–108. doi:10.7224/1537-2073.2014-006
Merkelbach, S., Sittinger, H., & Koenig, J. (2002). Is there a differential impact of fatigue and physical
disability on quality of life in multiple sclerosis? The Journal of Nervous and Mental Disease,
190, 388–393. doi:10.1097/00005053-200206000-00007
Molina, S., & Borkovec, T. D. (1994). The Penn State Worry Questionnaire: Psychometric properties
and associated characteristics. In G. C. L. Davey & F. Tallis (Eds.), Worrying: Perspectives on
theory, assessment and treatment (pp. 265–283). Chichester, UK: John Wiley.
Morales-Gonzalez, J. M., Benito-Leon, J., Rivera-Navarro, J., & Mitchell, A. J. (2004). A systematic
approach to analyse health-related quality of life in multiple sclerosis: The GEDMA study.
Multiple Sclerosis, 10, 47–54. doi:10.1191/1352458504ms967oa
Moss-Morris, R., Dennison, L., Landau, S., Yardley, L., Silber, E., & Chalder, T. (2013). A randomized
controlled trial of cognitive behavioural 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
Nortvedt, M. W., Riise, T., & Myhr, K. M. (1999). Quality of life in multiple sclerosis: Measuring the
disease effects more broadly. Neurology, 53, 1098–1103. doi:10.1212/WNL.53.5.1098
Osborne, T. L., Jensen, M. P., Ehde, D. M., Hanely, M. A., & Draft, C. (2007). Psychosocial factors
associated with pain intensity, pain-related interference, and psychological functioning in
persons with multiple sclerosis and pain. Pain, 127, 52–62. doi:10.1016/j.apmr.2010.09.021
Patti, F., Caopardo, M., & Palermo, F. (2003). Health-related quality of life and depression in an Italian
sample of multiple sclerosis patients. Journal of the Neurological Sciences, 211, 55–62.
doi:10.1016/S0022-510X(03)00040-6
Health anxiety and its impact on QOL in RRMS 15

Ratcliffe, D., MacLeod, A., & Sensky, T. T. (2006). Anxiety in patients who have had a myocardial
infarction: The maintaining role of perceived physical sensations and causal attributions.
Behavioural and Cognitive Psychotherapy, 34, 201–217. doi:10.1017/S1352465806002773
Robbins, J. M., & Kirmayer, L. J. (1991). Attributions of common somatic symptoms. Psychological
Medicine, 21, 1029–1045. doi:10.1017/S0033291700030026
Rode, S., Salkovskis, P., Dowd, H., & Hanna, M. (2006). Health anxiety levels in chronic pain
clinic attenders. Journal of Psychosomatic Research, 60, 155–161. doi:10.1016/j.
jpsychores.2005.07.005
Salkovskis, P. M. (1992). Psychological treatment of noncardiac chest pain: The cognitive approach.
The American Journal of Medicine, 92, S114–SA121. doi:10.1016/0002-9343(92)80066-9
Salkovskis, P. M. (1996). The cognitive approach to anxiety: Threat beliefs, safety seeking behaviour,
and the special case of health anxiety and obsessions. In P. M. Salkovskis (Ed.), Frontiers of
cognitive therapy (pp. 48–74). New York, NY: Guildford.
Salkovskis, P. M., Rimes, K. A., Warwick, H. M. C., & Clark, D. M. (2002). The Health Anxiety
Inventory: Development and validation of scales for the measurement of health anxiety and
hypochondriasis. Psychological Medicine, 32, 843–853. doi:10.1017/S0033291702005822
Salkovskis, P. M., & Warwick, H. M. C. (1986). Morbid preoccupation, health-anxiety and
reassurance: A cognitive-behavioural approach to hypochondriasis. Behaviour Research and
Therapy, 24, 597–602. doi:10.1016/0005-7967(86)90041-0
Seivewright, H., Salkovskis, P., Green, J., Mullan, N., Behr, G., Carline, E., . . . Tyrer, P. (2004).
Prevalence and service implications of health anxiety in genitourinary medicine clinics.
International Journal of STD & AIDS, 15, 519–522. doi:10.1258/0956462041558122
Sensky, T., MacLeod, A. K., & Rigby, M. F. (1996). Causal attributions about common somatic
sensations among frequent general practice attenders. Psychological Medicine, 26, 641–646.
doi:10.1017/S0033291700035716
Sharrack, B., & Hughes, R. A. (1999). The Guy’s Neurological Disability Scale (GNDS): A new
disability measure for multiple sclerosis. Multiple Sclerosis, 5, 223–233. doi:10.1177/
135245859900500406
Skerrett, T. N., & Moss-Morris, R. (2006). Fatigue and social impairment in multiple sclerosis: The
role of patients’ cognitive and behavioural responses to their symptoms. Journal of
Psychosomatic Research, 61, 587–593. doi:10.1016/j.jpsychores.2006.04.018
Smith, A. (1982). The symbol digit modalities test manual: Revised. Los Angeles, CA: Western
Psychological Services.
Solari, A., Palmisano, L., & Mendozzi, L. (1999). The impact of multiple sclerosis on health-related
quality of life. Neurology, 52(S2), A141.
Spain, L. A., Turbridy, N., Kilpatrick, T. J., Adams, S. J., & Holmes, A. C. N. (2007). Illness perception
and health-related quality of life in multiple sclerosis. Acta Neurologica Scandinavica, 116,
293–299. doi:10.1111/j.1600-0404.2007.00895.x
Sprangers, M. A. G., de Regt, E. B., & Andries, F. (2000). Which chronic conditions are associated
with better or poorer quality of life? Journal of Clinical Epidemiology, 53, 895–907.
doi:10.1016/S0895-4356(00)00204-3
Stuifbergen, A. (1995). Health-promoting behaviors and quality of life among individuals with
multiple sclerosis. Scholarly Inquiry for Nursing Practice, 9, 31–55.
Summers, M. M., Fisniku, L. K., Anderson, V. M., Miller, D. H., Cipolotti, L., & Ron, M. A. (2007).
Cognitive impairment in relapsing-remitting multiple sclerosis can be predicted by imaging
performed several years earlier. Multiple Sclerosis, 13, 1–8. doi:10.1177/1352458507082353
Taillefer, S., Kirmayer, L. J., Robbins, J. M., & Lasry, J. C. (2002). Correlates of illness worry in chronic
fatigue syndrome. Journal of Psychosomatic Research, 54, 331–337. doi:10.1016/S0022-3999
(02)00332-X
Tang, N. K., & Harvey, A. G. (2006). Altering misperception of sleep in insomnia: Behavioral
experiment versus verbal feedback. Journal of Consulting and Clinical Psychology, 74, 767–
776. doi:10.1037/0022-006X.74.4.767
16 Aimee L. Hayter et al.

Taylor, S., & Cox, B. J. (1998). Anxiety sensitivity: Multiple dimensions and hierarchic structure.
Behaviour Research and Therapy, 36, 37–51. doi:10.1016/S0005-7967(97)00071-5
Tyrer, P., Cooper, S., Tyrer, H., Salkovskis, P., Crawford, M., Gree, J., . . . Barrett, B. (2011). CHAMP:
Cognitive behaviour therapy for health anxiety in medical patients, a randomised controlled
trial. BioMed Central Psychiatry, 14, 11–99. doi:10.1186/1471-244X-11-99
Vaughan, R., Morrison, L., & Miller, E. (2003). The illness representations of multiple sclerosis and
their relations to outcome. British Journal of Health Psychology, 8, 287–301. doi:10.1348/
135910703322370860
Warwick, H. M. C., & Salkovskis, P. M. (1990). Hypochondriasis. Behaviour Research and Therapy,
28, 105–117. doi:10.1016/0005-7967(90)90023-C
Wynia, K., Middel, B., van Dijk, J. P., De Keyser, J. H. A., & Reijneveld, S. A. (2008). The impact of
disabilities on quality of life in people with multiple sclerosis. Multiple Sclerosis, 14, 972–980.
doi:10.1177/1352458508091260
Zigmond, A. S., & Snaith, R. P. (1983). The hospital anxiety and depression scale. Acta Psychiatrica
Scandinavica, 67, 361–370. doi:10.1111/j.1600-0447.1983.tb09716.x

Received 5 May 2014; revised version received 11 December 2015

You might also like