Autonomic function in chronic fatigue syndrome with and without painful temporomandibular disorder

Abstract

Background: Chronic fatigue syndrome (CFS) is heterogeneous in nature, yet no clear subclassifications currently exist. There is evidence of dysautonomia in almost 90% of patients and CFS is often co-morbid with conditions associated with autonomic nervous system (ANS) dysfunction, such as temporomandibular disorders (TMD). The present study examined the point prevalence of TMD in a sample of people with CFS and explored whether co-morbidity between the conditions is associated with greater ANS dysfunction than CFS alone. Method: Fifty-one patients and 10 controls underwent screening for TMD. They completed a self-report measure of ANS function (COMPASS-31) and objective assessment of heart rate variability during rest and standing (derived using spectral analysis). Frequency densities in the high-frequency (HF) and low-frequency (LF) band were calculated. Results: Patients with CFS were divided into those who screened positive for TMD (n = 16, 31%; CFS + TMD) and those who did not (n = 35, 69%; CFS − TMD). Both CFS groups had significantly higher self-rated ANS dysfunction than controls. CFS + TMD scored higher than CFS − TMD on the orthostatic and vasomotor subscales. The CFS + TMD group had significantly higher HF and significantly lower LF at rest than the other two groups. In discriminant function analysis, self-report orthostatic intolerance and HF units correctly classified 75% of participants. Conclusions: Almost one-third of CFS patients screened positive for TMD and this was associated with greater evidence of parasympathetic dysfunction. The presence of TMD shows potential as an effective screen for patients with CFS showing an autonomic profile and could help identify subgroups to target for treatment.

Full text

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=rftg20
Download by: [Newcastle University] Date: 01 December 2015, At: 03:01
Fatigue: Biomedicine, Health & Behavior
ISSN: 2164-1846 (Print) 2164-1862 (Online) Journal homepage: http://www.tandfonline.com/loi/rftg20
Autonomic function in chronic fatigue syndrome
with and without painful temporomandibular
disorder
Lucy J. Robinson, Justin Durham, Laura L. MacLachlan & Julia L. Newton
To cite this article: Lucy J. Robinson, Justin Durham, Laura L. MacLachlan & Julia L.
Newton (2015) Autonomic function in chronic fatigue syndrome with and without painful
temporomandibular disorder, Fatigue: Biomedicine, Health & Behavior, 3:4, 205-219, DOI:
10.1080/21641846.2015.1091152
To link to this article: http://dx.doi.org/10.1080/21641846.2015.1091152
Published online: 05 Oct 2015.
Submit your article to this journal
Article views: 18
View related articles
View Crossmark data

Autonomic function in chronic fatigue syndrome with and without
painful temporomandibular disorder
Lucy J. Robinson
a,b
, Justin Durham
c
, Laura L. MacLachlan
d
and Julia L. Newton
d,e
*
a
School of Psychology, Newcastle University, Newcastle upon Tyne, UK;
b
Northumbria
National Health Service (NHS) Foundation Trust, Northumberland, UK;
c
Centre for Oral
Health Research, Newcastle University, Newcastle upon Tyne, UK;
d
Institute of Cellular
Medicine, Newcastle University, Newcastle upon Tyne, UK;
e
Newcastle Hospitals NHS
Foundation Trust, Newcastle upon Tyne, UK
(Received 22 May 2015; accepted 28 August 2015)
Background: Chronic fatigue syndrome (CFS) is heterogeneous in nature, yet no
clear subclassifications currently exist. There is evidence of dysautonomia in almost
90% of patients and CFS is often co-morbid with conditions associated with
autonomic nervous system (ANS) dysfunction, such as temporomandibular
disorders (TMD). The present study examined the point prevalence of TMD in a
sample of people with CFS and explored whether co-morbidity between the
conditions is associated with greater ANS dysfunction than CFS alone. Method:
Fifty-one patients and 10 controls underwent screening for TMD. They
completed a self-report measure of ANS function (COMPASS-31) and objective
assessment of heart rate variability during rest and standing (derived using
spectral analysis). Frequency densities in the high-frequency (HF) and low-
frequency (LF) band were calculated. Results: Patients with CFS were divided
into those who screened positive for TMD (n= 16, 31%; CFS + TMD) and those
who did not (n= 35, 69%; CFS −TMD). Both CFS groups had significantly
higher self-rated ANS dysfunction than controls. CFS + TMD scored higher than
CFS −TMD on the orthostatic and vasomotor subscales. The CFS + TMD group
had significantly higher HF and significantly lower LF at rest than the other two
groups. In discriminant function analysis, self-report orthostatic intolerance and
HF units correctly classified 75% of participants. Conclusions: Almost one-third
of CFS patients screened positive for TMD and this was associated with greater
evidence of parasympathetic dysfunction. The presence of TMD shows potential
as an effective screen for patients with CFS showing an autonomic profile and
could help identify subgroups to target for treatment.
Keywords: chronic fatigue syndrome; temporomandibular disorder; autonomic
dysfunction; dysautonomia; phenotypes; heart rate variability; orthostatic
intolerance
Introduction
Chronic fatigue syndrome (CFS) is characterised by debilitating fatigue that has lasted
at least 6 months and is not satisfactorily explained by the presence of identified
© 2015 IACFS/ME
*Corresponding author. Email: julia.newton@ncl.ac.uk
Fatigue: Biomedicine, Health & Behavior, 2015
Vol. 3, No. 4, 205–219, http://dx.doi.org/10.1080/21641846.2015.1091152
Downloaded by [Newcastle University] at 03:01 01 December 2015

physical illness. In order to make a diagnosis, four from a possible eight symptoms are
required: cognitive difficulties, sore throat, tender lymph nodes, muscle pain, joint pain
without swelling, headaches, unrefreshing sleep, and post-exertional malaise.[1] This
diagnostic approach permits a large number of permutations to meet the criteria,
thereby potentially resulting in significant heterogeneity in those identified as having
CFS; individuals may meet differing criteria yet still be described as experiencing
the same illness.
This diagnostic-system-induced heterogeneity may partly explain the difficulty in
identifying the underlying aetiology and pathophysiology of CFS. Heterogeneous
samples drawn by utilising the ‘gold standard’diagnostic method [1] may generate con-
tradictory results depending on the relative proportions of symptom clusters in any
given sample. Several putative pathophysiological processes have been identified as
being implicated in CFS,[2] but they often only apply to subgroups and there are cur-
rently no clearly replicable subclassifications. Improving the phenotyping and genotyp-
ing of CFS would be a significant step forward to guide both clinical decision-making
and ongoing research.
There is evidence of autonomic nervous system (ANS) dysfunction (‘dysautono-
mia’) in nearly 90% of those with CFS.[3–11] This suggests that autonomic dysfunc-
tion may play a role in the genesis of fatigue. Dysautonomia is not unique to CFS.
There is evidence of dysautonomia in other unexplained clinical conditions such as
fibromyalgia, irritable bowel syndrome, and interstitial cystitis.[12–14] It is therefore
of interest to explore whether the presence of additional clinical complaints can be
used to subclassify CFS patients and elucidate pathophysiological processes involved
in some presentations.
One group of disorders of potential relevance are temporomandibular disorders
(TMD), a family of conditions often characterised by facial pain related to the muscles
of mastication or the temporomandibular joint.[15,16] On its own, there is evidence of
dysautonomia and genotypic differences in TMD that determine pain sensitivity.[17–
21] TMD is associated with marked impairment of quality of life,[22–24] and in conjunc-
tion with other dysautonomias, this effect appears to be multiplicative.[25]
TMD and CFS can present co-morbidly, with between 21% and 32% of CFS
patients reporting TMD.[26–28] The primary aim of the present study was to
examine the point prevalence of painful TMD in a sample of people with CFS,
whose primary complaint was not facial pain. The secondary aim was to use painful
TMD as a categorical exploratory variable to examine whether co-morbidity between
CFS and TMD is associated with greater subjective and objective ANS dysfunction
compared to CFS without TMD. The tertiary aim was to explore whether autonomic
function could be used effectively to discriminate between CFS with and without
TMD and controls.
Methods
Participants
Participants were recruited as part of a Medical Research Council-funded observational
study aimed at understanding the pathogenesis of autonomic dysfunction in patients
with CFS. Their primary complaint was chronic fatigue and not facial pain. Participants
were recruited via the Newcastle and North Tyneside National Health Service (NHS)
Clinical CFS Service and fulfilled the diagnostic criteria for CFS as outlined in the
206 L.J. Robinson et al.
Downloaded by [Newcastle University] at 03:01 01 December 2015

National Institute for Clinical Excellence guidelines.[1] Participants underwent a medical
assessment at the time of recruitment that identified that they did not have other diag-
noses. The study was reviewed and approved by the Newcastle NHS Ethics Committee
(REC 12/NE/0146, CLRN ID 97805). Consecutive patients attending the clinic were
provided with a Patient Information Sheet and invited to contact the research team
if they were willing to be involved. Participants were not selected positively or
negatively according to any criteria other than the fact that they were attending the
clinical service and had a Fukuda diagnosis of CFS. However, potential participants
were excluded if they screened positive for a major depressive episode as assessed
by a trained medic using the Structured Clinical Interview for the Diagnostic and Stat-
istical Manual for Mental Disorders (version IV; SCID-IV [29]).
Controls were recruited via notices provided in the hospital and university together
with a distribution of posters via the local myalgic encephalomyelitis Patient Support
Group where we invited relatives of those with CFS to participate. Controls were con-
sidered to be community controls rather than healthy controls; that is, they were not
positively or negatively recruited according to fatigue severity or the presence or
absence of particular symptoms. Participants attended the Clinical Research Facility
at Newcastle upon Tyne Hospitals NHS Foundation Trust, and on their first assessment
were invited to complete a series of symptom assessment tools. At the same visit they
completed the other assessments as outlined below.
Measures
Screening questionnaire for painful TMD: Painful TMD was screened for using the
long form of the validated instrument developed by Gonzalez et al.[30] This self-
report instrument contains six items and has a sensitivity of 99% and a specificity of
97%.[30] The first item (‘In the last 30 days, on average, how long did any pain in
your jaw or temple area on either side last?’) is scored on a three-point unipolar
ordinal scale: ‘no pain’[0]; ‘from very brief pain to more than a week, but it does
stop’[1]; and ‘continuous’.[2] The remaining items are scored on a dichotomous
score. This includes the question: ‘In the last 30 days, have you had pain or stiffness
in your jaw on awakening’, followed by four items asking whether these specific activi-
ties have changed the pain in the jaw or temple area: chewing hard food, opening the
mouth or moving it from side to side, holding the teeth together/clenching, and ‘talking,
kissing or yawning’, all scored Yes [1]/No [0]. Response codes are summed giving a
summary score whereby higher scores indicate higher likelihood of a painful TMD,
with the threshold value for a positive screening being greater than or equal to 3.
Composite Autonomic Symptom Scale-31 item (COMPASS-31): Scores for the
COMPASS-31 were derived from the long form of the instrument [31] using the abbre-
viated scoring criteria.[32] The questionnaire involves self-reported information about
the presence, frequency, and severity of symptoms associated with ANS dysfunction. It
has six subscales pertaining to different symptom areas: orthostatic intolerance, vaso-
motor, secretomotor, gastrointestinal, bladder, and pupillomotor function. The short
form has demonstrated good to excellent internal consistency on each of the subscales
(Cronbach’sα≥0.71).
Objective ANS measurement: ANS function was measured during a 10-minute
supine rest and in response to standing (from the point of standing with continuous
recording for a further 2 minutes whilst standing) using the Taskforce Monitor (CN
Systems; Gratz, Austria). This system is a clinically applicable validated assessment
Fatigue: Biomedicine, Health & Behavior 207
Downloaded by [Newcastle University] at 03:01 01 December 2015

tool that measures heart rate and blood pressure continuously beat to beat. From these
measures heart rate variability (HRV), blood pressure variability, and baroreflex sensi-
tivity were derived using spectral analysis and the autoregressive model. Of the avail-
able indices, the following variables were examined: average heart rate, average
systolic and diastolic blood pressure, stroke index, total peripheral resistance (TPRI),
and spectral power densities in the low-frequency (LF) and high-frequency (HF)
bands reported as normalised units for HRV (low-frequency normalised units, LFnu,
and high-frequency normalised units, HFnu, respectively), as well as the LF/HF
ratio. These variability indices are taken to reflect autonomic heart rate control, with
greater HF values (when expressed as normalised units) reflecting greater vagal (para-
sympathetic) modulation and higher LF values (when expressed as normalised units)
indicating greater sympathetic modulation.[33] The LF/HF ratio has been argued to
capture ‘sympathovagal balance’and higher values suggest greater dominance of the
sympathetic nervous system (SNS).[34]
Procedure
All procedures were performed in the same order in all participants. After completing
the symptom assessment tools, the Taskforce Monitor was applied and participants
rested supine in a quiet room with standard temperature and lighting. All measurements
were performed at the same time of day and after a light breakfast. After 10 minutes of
unrecorded rest, measurement of heart rate and blood pressure was recorded over 10
minutes of supine rest. At the end of this procedure, participants were asked to stand
as quickly as possible with support if necessary. Recording continued for 2 minutes
of active standing.
Data analysis
Data were analysed using SPSS 21 (IBM, Release 21.0.0.0). Comparisons between the
three groups were conducted using one-way ANOVA with post-hoc Tukey’s Honestly
Significant Difference (HSD) tests. A significance level of p≤.05 was used for both
ANOVA and the post-hoc tests. Before employing parametric analyses, distributions
of all continuous data were checked. Any distribution with skewness or kurtosis ≥1
was explored further using visual analysis with a histogram and boxplot. Following rec-
ommended best practice,[35] distributions showing evidence of outliers were win-
sorised. Distributions showing evidence of skew or kurtosis were transformed using
square root transformation or log 10 transformation as appropriate until the values
were within the criterion. Distributions showing evidence of both outliers and skew
or kurtosis were first winsorised and then transformed as appropriate.
Effect sizes between patients and controls were calculated using Cohen’sdwith the
following formula: (mean control group −mean patient group)/pooled standard devi-
ation. Effect sizes between the patient groups were calculated by subtracting the patients
who screened negative for TMD (−TMD) from those who screened positive for TMD
(+TMD) such that a positive effect size indicated a higher value in the +TMD group.
Conventions for Cohen’sdare that small = 0.2–0.5, medium = 0.5–0.8, and large >
0.8.[36] Analyses and effect sizes were conducted on transformed data, but untrans-
formed means and standard deviations are reported in the Tables 2 and 3(transformed
data are available on request). Pearson’s correlations were conducted between objective
and subjective measures of ANS function. To identify whether the autonomic measures
208 L.J. Robinson et al.
Downloaded by [Newcastle University] at 03:01 01 December 2015

were effective discriminators between the groups, a discriminant function analysis
(DFA) was conducted. The (transformed) resting and standing autonomic data and
the COMPASS subscale scores were entered in a stepwise manner (cut-off for inclusion
p< .05, cut-off for removal p> .1). The DFA was estimated using Wilks’Lambda.
Results
A total of 51 patients were recruited to the study (38 females) with a mean age of 46.0
(SD 11.9) years (see Table 1). They had a mean duration of illness of 13.7 (SD 10.1)
years (range 1–45 years). Ten controls were also recruited (7 females) with a mean age
of 49.4 (SD 15.3) years. There were no significant differences in age between the three
groups (F
2,60
= 0.62, p= .54) or gender (χ
2
= 4.55, p= .10). The two patient groups did
not differ significantly in age (t
49
= 0.82, p= .42), but there were significantly more
females in the CFS + TMD group than in the CFS −TMD group (χ
2
= 4.33, p= .04),
which is consistent with the higher prevalence of TMD in women than men.[37]
Prevalence of TMD
Sixteen patients screened positive for TMD (31%). These patients were identified as the
‘CFS + TMD’group. Thirty-five patients did not meet the threshold for TMD and were
identified as the ‘CFS −TMD’group. None of the control participants scored above the
threshold for TMD.
Subjective autonomic measures
The CFS + TMD group scored significantly higher than the control group on all sub-
scales of the COMPASS-31 (the self-report subjective measure of ANS dysfunction,
all p= .04; Table 2). Effect sizes ranged from medium (d=−0.68) for the pupillomotor
subscale to very large (d=−3.32) for orthostatic intolerance (Table 4). The CFS −TMD
group scored significantly higher than the control group on four of the subscales (ortho-
static intolerance, secretomotor, gastrointestinal, and pupillomotor; all p= .02) with
effect sizes ranging from large (d=−0.85) for the secretomotor subscale to very large
(d=−1.8) for the orthostatic intolerance subscale. The CFS + TMD group scored signifi-
cantly higher than the CFS −TMD group on two subscales (orthostatic intolerance and
vasomotor) (p= .02) with medium effect sizes (d= 0.78 and d= 0.75, respectively).
Objective autonomic measures
Significant differences were noted between the three groups in univariate ANOVA
for three of the autonomic measures at rest: HFnu, LFnu, and the LF/HF ratio
(Table 3). Follow-up Tukey’s HSD tests indicated that the CFS + TMD group had
Table 1. Demographics.
CFS −TMD CFS + TMD Control
Mean age (SD) 46.9 (11.9) 44.0 (12.0) 49.4 (15.3)
Gender (M:F) 12:23 1:15 3:7
Mean length of illness (SD) 14.5 (10.0) 12.1 (10.4) –
Fatigue: Biomedicine, Health & Behavior 209
Downloaded by [Newcastle University] at 03:01 01 December 2015

significantly higher values for HFnu and significantly lower values for LFnu and LF/
HF ratio than both the other two groups (p< .05) (see Figure 1). There were no stat-
istically significant differences between the control group and the CFS −TMD group.
Similar differences were noted while standing with significant differences in uni-
variate ANOVA between the three groups in HFnu, LFnu, and LF/HF ratio
(Table 3). Follow-up Tukey’s HSD tests indicated that the CFS + TMD group had sig-
nificantly higher values than the CFS −TMD group for the HFnu and significantly
lower values for the LFnu (both p= .02). Differences for the LF/HF ratio did not
reach statistical significance (p= .07). Differences between the CFS + TMD and the
control group were not statistically significant (all p> .15) and there were no significant
differences between the CFS −TMD and the control group (all p> .39).
The effect sizes (Cohen’sd) for these statistically significant differences indicate
very large effects between the two patient groups (d> 1.3) and large effects between
the CFS + TMD and the control group (d> 1.0) for the resting measures and medium
effects (d> 0.7) for the standing measures (Table 4).
Relationship between subjective and objective measures
To identify whether self-report ANS measures were correlated with the objective
measures, the six subscales of the COMPASS were correlated with HF and LF HRV
indices. In the sample as a whole, the only significant relationships were between
self-report orthostatic intolerance and both HFnu and LFnu (respectively, r=−0.34,
p= .01; r= 0.30, p= .02; n= 59). Due to the small sample sizes, these relationships
were not statistically significant in any of the individual groups.
Table 2. Untransformed mean COMPASS-31 subscale scores.
CFS −
TMD
a
(n= 35)
CFS +
TMD
b
(n= 16)
Control
c
(n= 10) F
(2,68)
1
pPost-hoc
2
Mean SD Mean SD Mean SD
Orthostatic
intolerance
3.61 1.95 5.06 1.65 0.40 0.84 34.67 <.001 b > c***, a*;
a > c***
Vasomotor 1.18 1.45 2.38 1.89 0.40 1.26 8.13 .001 b > c***; b
>a*
Secretomotor
3
1.97 1.98 2.38 1.78 0.40 1.26 5.48 .006 b > c*; a >
c*
Gastrointestinal 8.97 4.02 11.00 5.11 4.00 2.71 16.65 <.001 b > c***; a
> c***
Bladder 0.91 1.35 1.63 1.45 0.70 1.25 3.26 .045 b > c*
Pupillomotor 7.18 3.14 9.19 3.73 2.20 2.49 19.80 <.001 b > c***; a
> c***
Notes: ‡p< .1.
*p< .05.
**p< .01.
***p< .001.
1
F- and p-values reported from analyses on transformed data where appropriate.
2
Tukey’s Honestly Significant Difference (HSD) test.
3
Untransformed values of this variable were used in analyses, although kurtosis exceeded the permitted value
(kurtosis = 1.1). A variety of different transformations failed to improve the degree of kurtosis and the marginal
violation of the criterion meant the raw values gave the best distributional statistics.
210 L.J. Robinson et al.
Downloaded by [Newcastle University] at 03:01 01 December 2015

Table 3. Untransformed mean autonomic function data recorded during 10-minute rest and while standing.
CFS −TMD
a
(n= 35)
CFS + TMD
b
(n= 16)
Control
c
(n= 10) F
1
pPost-hoc
2
Mean SD Mean SD Mean SD
10-minute rest
Heart rate (bpm) 74.5 7.9 74.1 14.2 73.6 7.1 0.42 .66
Systolic blood pressure (mmHg) 108.9 20.3 106.9 16.4 104.9 11.2 0.10 .91
Diastolic blood pressure (mmHg) 68.9 10.8 68.6 12.2 67.9 6.5 0.02 .98
Stroke index (ml/m
2
) 41.0 12.0 41.0 11.8 42.7 10.2 0.09 .92
Total peripheral resistance index (dyne*s/cm
5
) 2215.0 801.7 2308.6 859.7 2051.7 549.4 0.08 .92
Heart rate variability (ms
2
/Hz):
LFnu
^
65.3 11.2 45.4 17.4 62.4 12.9 12.49 <.001 b < a***, b < c**
HFnu
^^
34.1 11.0 54.6 17.4 37.6 12.9 13.45 <.001 b > a***, b > c**
LF/HF ratio 2.2 2.0 1.2 0.9 1.5 0.9 6.26 .003 b < a**
Standing
Heart rate (bpm) 89.2 9.6 86.1 14.4 84.7 7.8 1.42 .25
Systolic blood pressure (mmHg) 120.1 23.8 112.4 29.0 112.5 13.7 0.75 .48
Diastolic blood pressure (mmHg) 82.7 16.8 78.5 21.1 81.8 6.0 0.20 .82
Stroke index (ml/m
2
) 32.5 8.4 32.5 7.2 33.5 5.9 0.31 .74
Total peripheral resistance index (dyne*s/cm
5
) 2656.2 854.1 2725.0 1314.4 2571.8 599.2 0.03 .97
Heart rate variability (ms
2
/Hz):
LFnu
^
70.5 13.9 57.6 21.1 69.4 11.2 3.87 .026 b < a*
HFnu
^^
29.5 13.9 42.4 21.1 30.6 11.2 3.87 .026 b > a*
LF/HF ratio 3.8 2.7 2.4 2.3 2.9 1.8 2.59 .084
Notes: *p< .05.
**p< .01.
***p< .001.
1
F- and p-values reported from analyses on transformed data where appropriate.
2
Tukey’s HSD test.
^
Low-frequency normalised units.
^^
High-frequency normalised units.
Fatigue: Biomedicine, Health & Behavior 211
Downloaded by [Newcastle University] at 03:01 01 December 2015

Discriminant function analysis
The DFA identified two independent discriminant functions. The first explained 73.1%
of the variance (canonical R
2
= 0.46), whereas the second explained 26.9% (canonical
R
2
= 0.24). The correlations between the variables and the discriminant functions
suggested the first function captured self-report orthostatic intolerance (r= 0.95
between COMPASS orthostatic intolerance subscale and function 1, compared to
r=−0.32 with function 2) and the second captured parasympathetic activity at rest
(r= 0.87 between HFnu and function 2, compared to r= 0.50 with function 1).
The two functions significantly discriminated between the groups (Λ= 0.41, χ
2
(4)
= 49.47, p< .001). The horizontal spacing of the group centroids on the discriminant
function plot (Figure 1) indicates that increasing self-report orthostatic intolerance dis-
criminates each of the groups from one another (with controls reporting the lowest and
CFS + TMD the highest). The vertical spacing (with the CFS + TMD and control
groups closer together in the vertical plane) indicates that lower parasympathetic
activity at rest discriminated the CFS −TMD group from the other two groups.
Overall, the analysis classified 74.6% of participants correctly (80% of controls, 79%
Table 4. Effect sizes (Cohen’sd) for between-group differences (calculated on transformed
values where appropriate).
Effect size (Cohen’sd
a
)
Resting Standing
CFS −
TMD
vs CFS
+ TMD
Control
vs.
CFS −
TMD
Control
vs.
CFS +
TMD
CFS −
TMD
vs. CFS
+ TMD
Control
vs.
CFS −
TMD
Control
vs.
CFS +
TMD
Mean heart rate −0.27 0.10 0.17 −0.40 0.50 0.04
Systolic blood
pressure
−0.03 0.16 0.14 −0.30 0.34 0.00
Diastolic blood
pressure
−0.05 0.05 0.01 −0.18 0.06 −0.15
Stroke index 0.00 −0.15 −0.15 0.10 −0.30 −0.16
TPRI 0.08 0.09 0.16 0.00 0.10 0.08
Heart rate variability
Low-frequency
(nu
a
)
−1.48 0.25 −1.07 −0.79 0.08 −0.66
High-frequency
(nu)
1.55 −0.31 1.07 0.79 −0.08 0.66
Low-/High-
frequency ratio
−1.19 0.31 −1.07 −0.65 0.32 −0.39
COMPASS-31
Orthostatic
intolerance
0.78 1.80 3.32 –––
Vasomotor 0.75 0.55 1.17 –––
Secretomotor 0.21 0.85 1.23 –––
Gastrointestinal 0.46 1.31 1.60 –––
Bladder 0.60 0.19 0.68 –––
Pupillomotor 0.60 1.65 2.11 –––
Note:
a
For control versus patient group comparisons, positive values indicate higher values in the patient group.
For the patient group comparisons, positive values indicate higher values in the CFS + TMD patients than the CFS
−TMD patients.
212 L.J. Robinson et al.
Downloaded by [Newcastle University] at 03:01 01 December 2015

of CFS −TMD, and 62.5% of CFS + TMD), suggesting these variables have potential
to discriminate effectively between groups with and without ANS dysfunction.
Discussion
Almost one-third of patients with CFS (31%) screened positive for painful TMD and
those with the co-morbidity showed evidence of greater subjective and objective
ANS dysfunction compared to both community controls and people with CFS
without painful TMD. This is the first time the COMPASS-31 has been used in CFS
and further use in future studies may help reconcile inconsistencies in the literature
to date. Differences in the autonomic measures were significant discriminators
between the three groups and showed promising potential for identifying subgroups.
These preliminary results suggest that screening for painful TMD in those with CFS
could form the basis of a strategy to identify those with an ‘autonomic phenotype’of
the condition.
Patients with CFS + TMD showed evidence of greater dominance of the parasym-
pathetic nervous system (lower LF/HF ratio), which was reflected in both lower sym-
pathetic activity (lower LF values) and greater parasympathetic activity (higher HF
values). The differences were greatest when participants were at rest. Those without
painful TMD showed no significant differences from the control group on any of the
HRV measures.
Previous studies of HRV in CFS have generally reported no significant differences
between patients and controls at rest. In a systematic review, Meeus et al. [38] cited four
studies that measured HRV when supine in patients with CFS and none reported any
significant differences.[39–42] This is consistent with the findings in the present
study in the CFS −TMD group.
Figure 1. Combined-groups centroid plot from the discriminant function analysis.
Fatigue: Biomedicine, Health & Behavior 213
Downloaded by [Newcastle University] at 03:01 01 December 2015

Studies of HRV in patients with TMD have noted differences from controls reflec-
tive of greater sympathetic activity.[43,19,44] Schmidt and Carlson [44] reported
higher LF and lower HF in 22 patients with masticatory muscle pain while sitting,
although the differences were on the border of statistical significance. Eze-Nliam
et al. [43] recorded HRV indices during sleep in 37 patients with TMD and reported
lower HF and a higher LF/HF ratio, both consistent with greater sympathetic activity.
Maixner et al. [19] reported lower LF and HF in over 160 TMD patients assessed when
supine, which the authors suggested reflected a reduction in parasympathetic tone and a
bias towards greater sympathetic activity.
Consistent with Maixner et al. [19], the present sample showed evidence of lower
LF; however, the effect size is much larger (d= 1.07 versus d≈0.30). While both of
these studies are in contrast with the elevated LF values reported by Schmidt and
Carlson,[44] the latter used a seated rather than a supine assessment protocol, which
may explain these differences. In contrast with Maixner et al. [19], the present study
reported higher HF and a lower LF/HF ratio, both consistent with greater parasympa-
thetic dominance.
There are two important methodological differences that might relate to the discre-
pancy in the findings. Unlike the two previously cited studies,[19,43] the present study
reported spectral power in normalised units, which subtracts the very low frequency
(VLF) component from total power and reports HF and LF as a proportion of what
remains.[34] This makes an important adjustment for differences between samples in
total power [34]–for example, the absolute level of HF may be lower in patients
with painful TMD compared to controls, but if total power is also lower in patients,
HF may still be a relatively greater proportion of total power (once VLF is discounted).
It is an important methodological point for future studies, as Maixner et al. [19] reported
significantly lower total power in their TMD sample, suggesting this should be appro-
priately adjusted for in reporting HF and LF values. Additionally, consistent reporting
of HRV indices would increase comparability of studies.
The second main difference is the present study employed rigorous screening cri-
teria for those with depression, who were excluded from the sample. Depressive symp-
toms are often elevated in CFS and TMD [45–53] and depression itself is associated
with lower LF, HF, and higher LF/HF,[54] suggesting greater sympathetic activity in
this population. By excluding those with current depression, the present study may
have removed a potential confound in the assessment of HRV in the CFS population.
So while methodological reasons for the differences cannot be ruled out, it nonethe-
less remains feasible that the present findings could still be consistent with aberrant
function of the SNS –increased parasympathetic activity may arise as a compensatory
response to chronically elevated sympathetic activity. Recent studies have identified
haplotypes of the gene encoding catecholamine-O-methyltransferase (COMT) that
are associated with lower enzymatic activity and are linked to greater incidence of
TMD.[17,55–57]D
eficiency in the metabolism of catecholamines could conceivably
result in difficulty modulating the SNS and it is possible parasympathetic activity
may then increase to reciprocally inhibit SNS activity. Taking the reported (subjective)
increased difficulties with orthostatic intolerance in those with CFS + TMD on the
COMPASS and the (objective) HRV data as a whole, it seems reasonable to hypoth-
esise that there are problems with SNS function in this group.
If this or other candidate mechanisms of dysautonomia play a key role in the genesis
of symptoms, it would be anticipated that interventions that modulate ANS function
would be effective in the management of these conditions. Various candidate therapies
214 L.J. Robinson et al.
Downloaded by [Newcastle University] at 03:01 01 December 2015

exist, including vagal nerve stimulation, β-blockers, or interventions focusing on self-
regulation training,[58,59,56,60] and they would be of great interest to investigate
further in this population of patients.
Limitations
There are a number of limitations of the current study. The sample size is relatively
small, which may have limited power in some of the statistical analyses. Replication
in a bigger sample would rule out false-negative findings on the other indices of auto-
nomic function. A self-report instrument was used to identify painful TMD, and despite
its excellent psychometric properties, a standardised clinical assessment such as the
diagnostic criteria/TMD [15] would enable a more informed clinical diagnosis to be
made which subclassified the type of TMD. No information was recorded on previous
experience of TMD or previous treatment for TMD. Some of the CFS −TMD group
conceivably could have suffered from TMD previously and it is unknown what
impact this might have on ANS function once the condition has been successfully
treated or has resolved over time. Although participants did not have any other
medical diagnoses indicated by the assessments undertaken for the study, measures
of other co-morbidities that are common in this group (e.g. fibromyalgia and irritable
bowel syndrome) were not included, leaving it unclear whether it is simply the presence
of any co-morbidity that is associated with altered ANS function or whether the differ-
ences are specific to TMD.
It is also possible that a third broader factor, of which TMD may be an expression,
may explain the differences. It has been demonstrated that joint hypermobility and
Ehlers–Danlos Syndrome (EDS) may be a risk factor for TMD,[61,62] and EDS in
itself is associated with both self-report and objectively assessed autonomic dysfunc-
tion.[63,64] Joint hypermobility was not assessed in this study, and it –or other poten-
tial factors that were not measured –may relate to both the autonomic dysfunction and
the presence of TMD in the subgroup of participants showing the most aberrant HRV.
Relatedly, future studies should look to include a TMD-only group, which would help
address the extent to which TMD is the important factor for altered ANS function rather
than any other morbidity.
Conclusion
To further research into the aetiology and management of CFS, robust and validated
methods of identifying phenotypic subgroups of patients need to be established. The
present study suggests that assessing for co-morbid painful TMD shows potential
promise as a means to identify those with an ‘autonomic profile’. If this proves to be
the case, the ability to readily identify patients with an autonomic profile opens up
the possibility of testing treatments that modulate ANS function in this population,
which will ultimately clarify the role of dysautonomia in the genesis of symptoms
and potentially identify therapeutic management options.
Acknowledgements
LR analysed the data and wrote the manuscript. JD contributed to study design, was involved in
data analysis & interpretation, and wrote sections of the manuscript pertaining to TMDs. LM
was involved in the design of the study and conducted the data collection. JN secured the
Fatigue: Biomedicine, Health & Behavior 215
Downloaded by [Newcastle University] at 03:01 01 December 2015

funding and was instrumental in study design, wrote sections of the manuscript pertaining to
methodology, and was involved in data analysis and interpretation. All authors read and
approved the final manuscript.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This research was funded by the Medical Research Council [grant number MR/
J002712/1] to JN.
Notes on contributors
Lucy J. Robinson is a Clinical Psychologist and Clinical Academic Fellow at Newcastle
University. Her research interests include persistent physical symptoms, interoception
and the interaction between life events and physiology in the experience of distressing
physical symptoms.
Justin Durham is a Senior Lecturer and honorary Consultant Oral Surgeon with a spe-
cialism in orofacial pain.
Laura L. MacLachlan is a medic who was involved with studies into Chronic Fatigue
Syndrome as part of her Doctor of Medicine degree.
Julia L. Newton is Clinical Professor of Aging and Medicine at Newcastle University as
well as Dean of Clinical Medicine. She has expertise in Chronic Fatigue Syndrome and
autonomic function.
References
[1] Fukuda K, Straus SE, Hickie I, Sharpe MC, Dobbins JG, Komaroff A. The chronic fatigue syn-
drome: a comprehensive approach to its definition and study. International Chronic Fatigue
Syndrome Study Group. Ann Intern Med. 1994;121(12):953–959.
[2] Excellence NIfHaC. Chronic fatigue syndrome/myalgic encephalomyelitis (or encephalopathy):
Diagnosis and management of CFS/ME in adults and children. CG53. London: National
Institute for Health and Care Excellence; 2007.
[3] Freeman R, Komaroff AL. Does the chronic fatigue syndrome involve the autonomic nervous
system? Am J Med. 1997;102(4):357–364.
[4] Newton JL, Okonkwo O, Sutcliffe K, Seth A, Shin J, Jones DE. Symptoms of autonomic dys-
function in chronic fatigue syndrome. QJM: Mon J Assoc Phys. 2007;100(8):519–526. doi:10.
1093/qjmed/hcm057.
[5] Newton JL, Sheth A, Shin J, et al. Lower ambulatory blood pressure in chronic fatigue syn-
drome. Psychosom Med. 2009;71(3):361–365. doi:10.1097/PSY.0b013e31819ccd2a.
[6] LaManca JJ, Peckerman A, Walker J, et al. Cardiovascular response during head-up tilt in
chronic fatigue syndrome. Clin Physiol (Oxford, England). 1999;19(2):111–120.
[7] Hollingsworth KG, Jones DE, Taylor R, Blamire AM, Newton JL. Impaired cardiovascular
response to standing in chronic fatigue syndrome. Eur J Clin Invest. 2010;40(7):608–615.
doi:10.1111/j.1365-2362.2010.02310.x.
[8] Naschitz JE, Rosner I, Rozenbaum M, et al. The capnography head-up tilt test for evaluation of
chronic fatigue syndrome. Semin Arthritis Rheum. 2000;30(2):79–86. doi:10.1053/sarh.2000.
9201.
216 L.J. Robinson et al.
Downloaded by [Newcastle University] at 03:01 01 December 2015

[9] Streeten DH, Thomas D, Bell DS. The roles of orthostatic hypotension, orthostatic tachycardia,
and subnormal erythrocyte volume in the pathogenesis of the chronic fatigue syndrome. Am J
Med Sci. 2000;320(1):1–8.
[10] Schondorf R, Benoit J, Wein T, Phaneuf D. Orthostatic intolerance in the chronic fatigue syn-
drome. J Auton Nerv Syst. 1999;75(2–3):192–201.
[11] Rowe PC, Bou-Holaigah I, Kan JS, Calkins H. Is neurally mediated hypotension an unrecog-
nised cause of chronic fatigue? Lancet. 1995;345(8950):623–624.
[12] Petzke F, Clauw DJ. Sympathetic nervous system function in fibromyalgia. Curr. Rheumatol.
Rep. 2000;2(2):116–123.
[13] Mazurak N, Seredyuk N, Sauer H, Teufel M, Enck P. Heart rate variability in the irritable bowel
syndrome: a review of the literature. Neurogastroenterol Motil. 2012;24(3):206–216. doi:10.
1111/j.1365-2982.2011.01866.x.
[14] Martínez-Martínez L, Mora T, Vargas A, Fuentes-Iniestra M, Martínez-Lavín M. Sympathetic
nervous system dysfunction in fibromyalgia, chronic fatigue syndrome, irritable bowel syn-
drome, and interstitial cystitis: a review of case-control studies. J Clin Rheumatol. 2014;20
(3):145–150. doi:10.1097/RHU.0000000000000089.
[15] Schiffman E, Ohrbach R, Truelove E, et al. Diagnostic criteria for temporomandibular disorders
(DC/TMD) for clinical and research applications: recommendations of the International RDC/
TMD Consortium Network* and Orofacial Pain Special Interest Groupdagger. J Oral Facial
Pain Headache. 2014;28(1):6–27. doi:10.11607/jop.1151.
[16] Dworkin SF, LeResche L. Research diagnostic criteria for temporomandibular disorders:
review, criteria, examinations and specifications, critique. J Craniomand Disord: Facial Oral
Pain. 1992;6(4):301–355.
[17] Diatchenko L, Slade GD, Nackley AG, et al. Genetic basis for individual variations in pain per-
ception and the development of a chronic pain condition. Hum Mol Genet. 2005;14(1):135–143.
doi:10.1093/hmg/ddi013.
[18] Nackley AG, Tan KS, Fecho K, Flood P, Diatchenko L, Maixner W. Catechol-O-methyltrans-
ferase inhibition increases pain sensitivity through activation of both beta2- and beta3-adrener-
gic receptors. Pain. 2007;128(3):199–208. doi:10.1016/j.pain.2006.09.022.
[19] Maixner W, Greenspan JD, Dubner R, et al. Potential autonomic risk factors for chronic TMD:
descriptive data and empirically identified domains from the OPPERA case-control study. J
Pain. 2011;12(11 Suppl):T75–T91. doi:10.1016/j.jpain.2011.09.002.
[20] Chen H, Nackley A, Miller V, Diatchenko L, Maixner W. Multisystem dysregulation in painful
temporomandibular disorders. J Pain. 2013;14(9):983–996.
[21] Tchivileva IE, Lim PF, Smith SB, et al. Effect of catechol-O-methyltransferase polymorphism
on response to propranolol therapy in chronic musculoskeletal pain: a randomized, double-blind,
placebo-controlled, crossover pilot study. Pharmacogenet Genom. 2010;20(4):239–248. doi:10.
1097/FPC.0b013e328337f9ab.
[22] Dahlstrom L, Carlsson GE. Temporomandibular disorders and oral health-related quality of life. A
systematic review. Acta Odontol Scand. 2010;68(2):80–85. doi:10.3109/00016350903431118.
[23] Rakib A, White PD, Pinching AJ, et al. Subjective quality of life in patients with chronic fatigue
syndrome. Qual Life Res. 2005;14(1):11–19.
[24] Van Heck GL, De Vries J. Quality of life of patients with Chronic Fatigue Syndrome. J Chronic
Fatigue Syndr. 2002;10(1):17–35.
[25] Durham J, McDonald C, Hutchinson L, Newton JL. Painful Temporomandibular Disorders
(TMDs) are common in patients with Postural Orthostatic Tachycardia Syndrome (PoTS) and
impacts significantly on quality of life. J Oral Facial Pain Headache2015;29(2):152–157.
doi:10.11607/ofph.1396.
[26] Aaron LA, Burke MM, Buchwald D. Overlapping conditions among patients with chronic
fatigue syndrome, fibromyalgia, and temporomandibular disorder. Arch Int Med. 2000;160
(2):221–227.
[27] Korszun A, Papadopoulos E, Demitrack M, Engleberg C, Crofford L. The relationship between
temporomandibular disorders and stress-associated syndromes. Oral Surg Oral Med Oral Pathol
Oral Radiol Endod. 1998;86(4):416–420.
[28] Metcalf LN, McGregor NR, Roberts TK. Membrane damaging toxins from coagulase-negative
Staphylococcus are associated with self-reported temporomandibular disorder (TMD) in patients
with chronic fatigue syndrome. J Chronic Fatigue Syndr. 2004;12(3):25–43.
Fatigue: Biomedicine, Health & Behavior 217
Downloaded by [Newcastle University] at 03:01 01 December 2015

[29] First MB, Spitzer RL, Williams JBW, Gibbon M. Structured clinical interview for DSM-IV
(SCID-IV). New York (NY): Biometrics Research Department, New York State Psychiatric
Institute; 1995.
[30] Gonzalez YM, Schiffman E, Gordon SM, et al. Development of a brief and effective temporo-
mandibular disorder pain screening questionnaire: reliability and validity. J Am Dental Assoc
(1939). 2011;142(10):1183–1191.
[31] Suarez GA, Opfer-Gehrking TL, Offord KP, Atkinson EJ, O’Brien PC, Low PA. The autonomic
symptom profile: a new instrument to assess autonomic symptoms. Neurology. 1999;52(3):523–528.
[32] Sletten DM, Suarez GA, Low PA, Mandrekar J, Singer W. COMPASS 31: a refined and abbre-
viated Composite Autonomic Symptom Score. Mayo Clin Proc. 2012;87(12):1196–1201.
doi:10.1016/j.mayocp.2012.10.013.
[33] Malpas SC. Neural influences on cardiovascular variability: possibilities and pitfalls. Am J
Physiol Heart Circ. Physiol. 2002;282(1):H6–H20.
[34] Electrophysiology TFoTESoCaTNASoPa. Heart rate variability: standards of measurement,
physiological interpretation, and clinical use. Eur Heart J. 1996;17:354–381.
[35] Osborne JW, editor. Best practices in quantitative methods. Thousand Oaks (CA): Sage
Publications Inc; 2008.
[36] Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale (NJ): Lawrence
Erlbaum Associates; 1988.
[37] Slade GD, Bair E, By K, et al. Study methods, recruitment, sociodemographic findings, and
demographic representativeness in the OPPERA study. J Pain. 2011;12(11, Suppl):T12–T26.
doi:10.1016/j.jpain.2011.08.001.
[38] Meeus M, Goubert D, De Backer F, et al. Heart rate variability in patients with fibromyalgia and
patients with chronic fatigue syndrome: a systematic review. Semin. Arthritis Rheum. 2013;43
(2):279–287. doi:10.1016/j.semarthrit.2013.03.004.
[39] De Becker P, Dendale P, De Meirleir K, Campine I, Vandenborne K, Hagers Y. Autonomic
testing in patients with chronic fatigue syndrome. Am. J Med. 1998;105(3, Suppl 1):22S–
26S. doi:10.1016/S0002-9343(98)00168-5.
[40] Duprez DA, De Buyzere ML, Drieghe B, et al. Long- and short-term blood pressure and RR-
interval variability and psychosomatic distress in chronic fatigue syndrome. Clin Sci. 1998;94
(1):57–63.
[41] Yamamoto Y, LaManca JJ, Natelson BH. A measure of heart rate variability is sensitive to
orthostatic challenge in women with chronic fatigue syndrome. Exp Biol Med. 2003;228
(2):167–174.
[42] Yataco A, Talo H, Rowe P, Kass DA, Berger RD, Calkins H. Comparison of heart rate varia-
bility in patients with chronic fatigue syndrome and controls. Clin Auton Res. 1997;7
(6):293–297. doi:10.1007/BF02267720.
[43] Eze-Nliam CM, Quartana PJ, Quain AM, Smith MT. Nocturnal heart rate variability is lower in
temporomandibular disorder patients than in healthy, pain-free individuals. J Orofac Pain.
2011;25(3):232–239.
[44] Schmidt JE, Carlson CR. A controlled comparison of emotional reactivity and physiological
response in masticatory muscle pain patients. J Orofac Pain. 2009;23(3):230–242.
[45] McInnis OA, Matheson K, Anisman H. Living with the unexplained: coping, distress, and
depression among women with chronic fatigue syndrome and/or fibromyalgia compared to an
autoimmune disorder. Anxiety Stress Coping. 2014;27(6):601–618. doi:10.1080/10615806.
2014.888060.
[46] Carlson CR, Reid KI, Curran SL, et al. Psychological and physiological parameters of mastica-
tory muscle pain. Pain. 1998;76(3):297–307.
[47] Gatchel RJ, Garofalo JP, Ellis E, Holt C. Major psychological disorders in acute and chronic
TMD: an initial examination. J Am Dent Assoc (1939). 1996;127(9):1365–1370, 1372, 1374.
[48] McCreary CP, Clark GT, Merril RL, Flack V, Oakley ME. Psychological distress and diagnostic
subgroups of temporomandibular disorder patients. Pain. 1991;44(1):29–34.
[49] List T, Dworkin SF. Comparing TMD diagnoses and clinical findings at Swedish and US TMD
centers using research diagnostic criteria for temporomandibular disorders. J Orofac Pain.
1996;10(3):240–253.
[50] Gallagher RM, Marbach JJ, Raphael KG, Dohrenwend BP, Cloitre M. Is major depression
comorbid with temporomandibular pain and dysfunction syndrome? A pilot study. Clin J
Pain. 1991;7(3):219–225.
218 L.J. Robinson et al.
Downloaded by [Newcastle University] at 03:01 01 December 2015

[51] McCue P, Buchanan T, Martin CR. Screening for psychological distress using internet admin-
istration of the Hospital Anxiety and Depression Scale (HADS) in individuals with chronic
fatigue syndrome. Br J Clin Psychol/Br Psychol Soc. 2006;45(Pt 4):483–498. doi:10.1348/
014466505×82379.
[52] Sepede G, Racciatti D, Gorgoretti V, et al. Psychophysical distress and alexithymic traits in
chronic fatigue syndrome with and without comorbid depression. Int J Immunopathol
Pharmacol. 2011;24(4):1017–1025.
[53] Moss-Morris R, Petrie KJ. Discriminating between chronic fatigue syndrome and depression: a
cognitive analysis. Psychol Med. 2001;31(3):469–479.
[54] Kemp AH, Quintana DS, Gray MA, Felmingham KL, Brown K, Gatt JM. Impact of depression
and antidepressant treatment on heart rate variability: a review and meta-analysis. Biol
Psychiatr. 2010;67(11):1067–1074. doi:10.1016/j.biopsych.2009.12.012.
[55] Diatchenko L, Fillingim RB, Smith SB, Maixner W. The phenotypic and genetic signatures of
common musculoskeletal pain conditions. Nat Rev Rheumatol. 2013;9(6):340–350. doi:10.
1038/nrrheum.2013.43.
[56] Light KC, Bragdon EE, Grewen KM, Brownley KA, Girdler SS, Maixner W. Adrenergic dys-
regulation and pain with and without acute beta-blockade in women with fibromyalgia and tem-
poromandibular disorder. J Pain. 2009;10(5):542–552. doi:10.1016/j.jpain.2008.12.006.
[57] Diatchenko L, Anderson AD, Slade GD, et al. Three major haplotypes of the beta2 adrenergic
receptor define psychological profile, blood pressure, and the risk for development of a common
musculoskeletal pain disorder. Am J Med Genet Part B, Neuropsychiatr Genet: Official Pub Int
Soc Psychiatr Genet. 2006;141b(5):449–462. doi:10.1002/ajmg.b.30324.
[58] Sauer SE, Burris JL, Carlson CR. New directions in the management of chronic pain: self-regu-
lation theory as a model for integrative clinical psychology practice. Clin Psychol Rev. 2010;30
(6):805–814. doi:10.1016/j.cpr.2010.06.008.
[59] Carlson CR, Bertrand PM, Ehrlich AD, Maxwell AW, Burton RG. Physical self-regulation
training for the management of temporomandibular disorders. J Orofac Pain. 2001;15(1):47–55.
[60] Sutcliffe K, Gray J, Tan MP, et al. Home orthostatic training in chronic fatigue syndrome –a
randomized, placebo-controlled feasibility study. Eur J Clin Invest. 2010;40(1):18–24. doi:10.
1111/j.1365-2362.2009.02225.x.
[61] Pasinato F, Souza JA, Correa EC, Silva AM. Temporomandibular disorder and generalized joint
hypermobility: application of diagnostic criteria. Br J Otorhinolaryngol. 2011;77(4):418–425.
[62] De Coster PJ, Martens LC, De Paepe A. Oral health in prevalent types of Ehlers-Danlos syn-
dromes. J Oral Pathol Med: Official Publ Int Assoc Oral Pathologists Am Acad Oral Pathol.
2005;34(5):298–307. doi:10.1111/j.1600-0714.2004.00300.x.
[63] De Wandele I, Rombaut L, Leybaert L, et al. Dysautonomia and its underlying mechanisms in
the hypermobility type of Ehlers-Danlos syndrome. Semin Arthritis Rheum. 2014;44(1):93–
100. doi:10.1016/j.semarthrit.2013.12.006.
[64] De Wandele I, Calders P, Peersman W, et al. Autonomic symptom burden in the hypermobility
type of Ehlers–Danlos syndrome: a comparative study with two other EDS types, fibromyalgia,
and healthy controls. Semin Arthritis Rheum. 2014;44(3):353–361. doi:10.1016/j.semarthrit.
2014.05.013.
Fatigue: Biomedicine, Health & Behavior 219
Downloaded by [Newcastle University] at 03:01 01 December 2015