Relationship Between Perceived Sleep Problems
and Thalamic Size in Patients with Chronic Fatigue
Syndrome Compared to Non-Fatigued Controls:
A Preliminary Study
Stefan Begré1, Tobias Lütgert1, Luca Remonda2, Roland Wiest2, Claus Kiefer2 and
Roland von Känel1
1Department of General Internal Medicine, Division of Psychosomatic Medicine, University Hospital/
Inselspital, Bern, Switzerland. 2Department of Neuroradiology, University Hospital/Inselspital, Bern,
Abstract: Chronic fatigue syndrome (CFS) is characterized by disabling fatigue of unknown etiology. The thalamus
is a key subcortical structure in sleep disorders and certain cognitive functions previously shown to be impaired in CFS
patients. We investigated the association between subjective sleep quality and thalamic size in CFS. Twelve right-handed
CFS patients and 12 age-, gender-, and handedness-matched healthy controls completed the Jenkins Sleep Question-
naire in order to assess subjective sleep problems. Thalamic size was determined by MR-based volumetry. More sleep
problems correlated with greater total thalamic volume in patients (rP = 0.62, 95% CI 0.07–0.88, p = 0.032) but not in con-
trols (rP = −0.034, p = 0.30). In post hoc analysis, more sleep problems correlated with right thalamic size in patients
(rP = 0.70, 95% CI 0.21−0.91, p = 0.012) but not in controls (rP = −0.080, p = 0.81). Our preliminary results provide a basis
for further studies on a possible role of the thalamus in sleep complaints and fatigue of patients with CFS.
Categories: Psychiatry, Neuroscience
Keywords: Chronic Fatigue Syndrome, Sleep, Thalamus
Chronic Fatigue Syndrome (CFS) is a symptom complex characterized by disabling fatigue of unknown
etiology, self-reported impairments in concentration and short-term memory, sleep disturbances, and
musculoskeletal pain. People tend to report “fatigue” as a perceived difﬁ culty, or even inability, to initiate
and sustain voluntary activities (Chaudhuri and Behan, 2004). This feeling is commonly communicated
as a lack of drive or action initiation (van der Linden et al. 2003). Nonetheless, fatigue mainly represents a
subjective state of mind. Therefore, it remains difﬁ cult to disentangle complaints of tiredness, lack of
energy related to depression, and fatigue as a consequence of chronic pain or as a symptom of a physical
disease. Particularly, it has to be mentioned that depression co-occurs at a relatively high prevalence with
CFS (Fukuda et al. 1994). Accordingly, any deﬁ nition of fatigue must account for subjective perception
and overlaps of the various clinical entities associated with fatigue (Moncrieff and Fletcher, 2007). The
deﬁ nition of CFS requires that the fatigue state must last for 6 or more consecutive months without any
active medical condition to explain chronic fatigue (Fukuda et al. 1994).
The subjective feeling of “fatigue” is a common symptom in neurology and occurs in diseases of
the central and peripheral nervous system. Fatigue can be distinguished as symptoms of peripheral
neuromuscular fatigue on the one hand and symptoms of physical and mental fatigue on the other.
Central fatigue may occur because of an integration failure of limbic input and motor functions within
the basal ganglia affecting the striatal-thalamic-frontal cortical system (Chaudhuri and Behan, 2004).
Whether these pathophysiologic mechanisms contribute to central fatigue in CFS is unknown. However,
for a better understanding of fatigue in many diseases and in order to develop new treatment options,
it seems important to learn more about these mechanisms.
Correspondence: Stefan Begré, M.D, Department of General Internal Medicine, Division of Psychosomatic
Medicine, University Hospital (Inselspital) Berne, CH-3010 Berne, Switzerland (Tel: +41 31 632 83 61;
Fax: +41 31 382 11 84; E-mail: email@example.com).
Copyright in this article, its metadata, and any supplementary data is held by its author or authors. It is published under the
Creative Commons Attribution By licence. For further information go to: http://creativecommons.org/licenses/by/3.0/.
Clinical Medicine: Psychiatry 2008:1 7–15
Begré et al
The thalamus has a strategic position in the with relatively more sleep problems. A parallel line
central autonomic network, running from the lim-
of research on neuropsychiatric diseases other than
bic cortical regions to the lower brain stem such CFS, such as schizophrenia (Geuze et al. 2005;
as to regulate the body’s homeostasis in an inte-
Danos et al. 2003) and parkinsonism (Byne et al.
grated fashion. In addition, the thalamus is a key 2002), demonstrated an association between reduc-
subcortical structure for certain cognitive functions tion of number of neurons and a decrease in tha-
such as attention, speed of information processing, lamic volume with disturbances in vigilance and
and working memory, all of which may show attention. In this preliminary study we therefore
functional disturbance in CFS patients (Busichio investigated whether subjective sleep quality could
et al. 2004). Besides fatigue, unrefreshing sleep is be related to thalamic size in patients suffering
the most commonly reported case-deﬁ ning symp-
from CFS. Supported by the above literature, we
tom of CFS. A substantial proportion of CFS assumed that the smaller the thalamic size, the
patients indicate sleep complaints starting after the more the vigilance and attention would be per-
onset of illness (Reeves et al. 2006). In healthy turbed, and, as a consequence, the higher the sleep
subjects the thalamus is deeply involved in the pressure would be. Therefore, we hypothesized
pathophysiology of sleep and has been shown to vice versa that the more subjectively impaired the
play a primary role in the organization of the wake-
sleep, the greater the thalamic size in CFS patients,
sleep rhythm (Reeves et al. 2006; Nofzinger, with no such relation seen in non-CFS controls.
2006). Altered thalamic function was also observed
in sleep disorders showing both increase and
decrease of thalamic activity. In stroke patients, Methods
paramedian thalamic neuronal loss is accompanied The local ethical committee approved the study
by deﬁ cient arousal during the day, leading to protocol and all participants gave written informed
hypersomnia and insufﬁ cient spindling and slow-
consent. Patients were recruited from our Psycho-
wave sleep production at night (Lugaresi, 1992). somatic Division in the Department of General
SPECT revealed a decreased blood ﬂ ow in the left Internal Medicine, University Hospital Bern,
thalamus during hypersomnolent periods of recur-
Switzerland. Eligible patients were approached by
rent hypersomnia (Nose et al. 2002). An fMRI the physician responsible for the hospital ward and
study revealed increased bilateral thalamic activa-
asked whether they would volunteer to participate
tion in healthy patients performing cognitive tasks in a study on brain imaging in CFS. All measure-
after 24h of sleep deprivation (Bassetti et al. 1996). ments were made in the outpatient psychosomatic
PET showed pronounced thalamic hypometabo-
clinic. The study was conducted on 12 right-handed
lism in familial fatal insomnia in which severe patients (9 women and 3 men) suffering from CFS
neuronal depletion in the mediodorsal (MD) and as diagnosed by previously defined research
anteroventral nuclei of the thalamus were demon-
criteria (Fukuda et al. 1994). Speciﬁ cally, patients
strated (Nose et al. 2002). These nuclei constitute had no somatic condition which could possibly
the limbic part of the thalamus, interconnecting explain fatigue as per a thorough medical history,
limbic and paralimbic regions of the cortex and physical examination, and laboratory work-up
other subcortical structures in the limbic system following the recommendations by the Centers for
including the hypothalamus.
Disease Control and Prevention (Danos, 2004;
To the best of our knowledge investigations of Henderson et al. 2000). Four patients had received
the association between thalamic size and sleep psychotropic medication within the last 4 months;
have not been performed in healthy persons. one had had paroxetine and methylphenidate, one
Volume measurement of brain structures has amitriptyline, one doxepine, and one continuous
become an increasingly useful tool to better under-
release morphine, respectively.
stand neuropsychiatric disorders (Chee and Choo,
Most psychiatric disorders are not exclusion
2004; Montagna et al. 2003). Variability in the criteria for a diagnosis of CFS (Danos, 2004;
relationship between thalamic volume and sleep Henderson et al. 2000). Therefore, only the fol-
function may be based on differences in the number lowing life-time and current psychiatric disorders
of thalamic neurons, neuronal or axonal density, which are speciﬁ ed exclusion criteria for CFS were
size of myelin sheaths, glia, and ﬂ uid content. Such identiﬁ ed in a clinical interview performed by an
mechanisms might be more prominent in subjects experienced physician: bipolar affective disorder,
Clinical Medicine: Psychiatry 2008:1
A preliminary Study
severe depressive disorder with melancholic maximum score of 20 points. Higher scores
features, any psychotic disorder, any eating indicate more sleep problems and lower sleep
disorder, and alcohol or illicit drug abuse. Spe-
ciﬁ cally, patients were asked whether a physician
(incl. a psychiatrist) had ever mentioned their Depression
having any of these disorders.
Symptoms of depression were assessed using the
The sample of healthy controls was recruited German version (Herrmann et al. 1995) of the
by word-of-mouth, and comprised mainly employ-
7-item depression subscale of the Hospital Anxiety
ees of the university hospital. The 12 controls who and Depression Scale (HADS) (Kudielka et al.
volunteered to participate were all carefully case-
2004) rendering a severity score between 0 and 21.
matched in terms of handedness, sex and age Depressive symptom levels do not allow for diag-
(± 1 year) with CFS patients. None of the controls nosing a clinical depression, but are suggestive of
took any psychopharmacological drug. Moreover, mild depression (8–10 points), moderate depression
controls had been required to indicate no signiﬁ cant (11–14 points), and severe depression ( 15 points),
mood disturbance and a negative history for any respectively. In studies comparing the HADS
life-time and current psychiatric disorders as veri-
depression subscale score with gold standard
ﬁ ed by an experienced psychiatrist in a clinical clinical assessment of depression in medically ill
patients, sensitivity estimates ranged from 56% to
100% and speciﬁ city estimates were between 73%
and 94% (Zigmond and Snaith, 1983; Goldberg,
1985). In patients with CFS, a cutoff 8 points on
the HADS depression subscale has been proved
CFS diagnosis and severity
valid and efﬁ cient for the screening of a depressive
The symptomatology of CFS was assessed using disorder as diagnosed in a structured clinical inter-
the semi-structured clinical interview asking for view (Silverstone et al. 1994; Henderson, 2005).
duration and number of symptoms as deﬁ ned by Because controls had been required to indicate no
the International CFS Study Group (Fukuda et al. signiﬁ cant mood disturbance they did not complete
1994). The average intensity of the nine CFS the depression scale.
deﬁ ning symptoms was interviewer-rated on a
Importantly, the questionnaires applied to mea-
Likert scale (0–4) to yield a maximum CFS severity sure sleep quality (Jenkins et al. 1988) and depres-
score of 36 points. At the time of the interview all sive symptoms (Zigmond and Snaith, 1983) have
patients suffered from CFS for longer than been widely used. In spite of their development in
6 months as per deﬁ nition. Patients were speciﬁ -
the 1980s, they are seen as state-of-the art to assess
cally asked when they ﬁ rst perceived a symptom the domains of subjective sleep complaints (Steptoe
they retrospectively would relate to CFS. This et al. 2008) and depressed mood (Henderson et al.
provided a proxy measure of duration the condition 2005), respectively.
had possibly last.
MRI recording and analysis
All MRI examinations were performed with a Sonata
Subjective sleep quality of patients and controls 1.5T scanner (Siemens Erlangen, Germany) with a
was assessed by the Jenkins Sleep Questionnaire 40 mT/m (200 mT/m-ms) gradient system and a CP
(JSQ) (Jenkins et al. 1988). The JSQ has been standard head coil, using the scanner software Syngo
designed to track common sleep problems in MR 2002B (VA21B). For anatomical imaging, a
clinical populations and has been successfully T1-weighted, sagittal oriented 3D-MPRAGE
applied in a German speaking population. The JSQ sequence (TR/TE/TI 2000/3.93/590 ms, matrix
comprises 4 items asking into a) difficulty in 256 × 256, FOV 256 × 256 mm, ﬂ ip angle 15°, slab
initiating sleep, b) awakening during the night, 160 mm) with a 1 mm3 isovoxel resolution was
c) awakening during sleep with difficulty obtained. Volumetric evaluations were performed by
maintaining sleep, and d) awakening exhausted in manually tracing the thalamus in three planes on
the morning despite having slept as usual. All items high-resolution T1-weighted scans using the
are rated on a Likert scale (0–5) yielding a commercially available BrainVoyager software
Clinical Medicine: Psychiatry 2008:1
Begré et al
(Version 4.9.6). The 3-dimensional on-line ROI investigate signiﬁ cant associations between two
tracing simultaneously in 3 planes allows the exact continuous variables. Because this was a case-
delineation of the thalamic anatomical borders, control design, conditional logistic regression was
which were deﬁ ned in accordance with the T1-signal performed to test whether continuous measures
changes at the grey/white matter interface at the (e.g. sleep score, thalamic volume) predict CFS
boundaries of the thalamus. The anterior boundary status. Our small sample size precluded, however,
of the thalamus was deﬁ ned as the posterior point reliable computation of logistic regression analysis
of the interventricular foramen and the posterior with the use of covariates and their interaction
boundary coincided with the section in which pul-
terms (Greenland et al. 2000). Alternatively, we
vinar thalami were seen. The lateral boundary of tested whether there would be a signiﬁ cant bivar-
the thalamus was deﬁ ned at the plane, where the iate correlation between sleep problem score and
posterior limb of the internal capsule was seen. The the total thalamic volume in CFS patients and
superior and inferior boundaries of thalamus was controls. Post hoc comparisons applied the con-
deﬁ ned at the level of body of the fornix and hypo-
servative Bonferroni correction for multiple
thalamic sulcus respectively. The manual ROI testings. That is, in case of a signiﬁ cant correlation
tracing was performed by a rater (TL) with experi-
between sleep problems and total thalamic volume
ence in neuroanatomy and blind to diagnosis. in CFS patients, a p-value of 0.025 would be
A spheric water ﬁ lled dummy model was used for required to state statistical signiﬁ cance of a cor-
calibration of the realistic thalamic volumes.
relation between the sleep problem score and
thalamic size of either side.
Statistical analyses were performed using SPSS Results
12.0 software package (SPSS Inc. Chicago, IL). Table (1) demonstrates that matching in terms of
The level of signiﬁ cance was set at p 0.05 and age, gender, and handedness between patients and
all tests were two-tailed. All data showed a normal controls was achieved. Thalamic size of either
distribution as per the Kolmogorov-Smirnov test. hemisphere and aggregate volume of thalamic
Pearson’s correlation test (rP) was used to nuclei did not signiﬁ cantly predict CFS status.
Table 1. Subjects’ characteristics.
(n = 12)
(n = 12)
Sleep quality score
a) Sleep initiation
b) Awakening during night
c) Sleep maintenance
d) Waking up exhausted
Duration since ﬁ rst CFS-related
CFS severity score
Thalamic size (mm3)
Clinical Medicine: Psychiatry 2008:1
A preliminary Study
In CFS patients, there was a positive correlation
between the sleep score and total thalamic volume
(i.e. volumes of the right and left thalamus
P = 0.62, 95% CI 0.07–0.88, p = 0.032;
Fig. 1a), which was not seen in non-CFS controls
P = −0.34, p = 0.30). Even though the total sleep
score (p = 0.84) and total thalamic volume (p = 0.95)
showed a normal distribution (Kolmogorov-
Smirnov test), the orientation of the regression line
between the sleep problem score and total thalamic
volume in patients with CFS was suggested to be
inﬂ uenced by one subject with comparably good
sleep (i.e. a score of 5 on the JSQ) and small
thalamic volume (cf. Fig. 1a). Therefore, we ana-
lyzed our data computing standardized residuals
Sleep problem score [points]
and Cook’s distance D to detect potential outliers Figure 1b. Sleep quality and right thalamic size in CFS patients.
and inﬂ uential cases in our regression model. We The ﬁ gure depicts the positive relationships (p’s 0.05) with ﬁ t line
between the sleep problem score and total thalamic size (Fig. 1a)
found that all values were within acceptable limits and right thalamic size (Fig. 1b), respectively, in the 12 patients with
(statistical analyses not shown in detail).
chronic fatigue syndrome (CFS).
In a post hoc analysis we explored which tha-
lamic size would possibly make a greater contribu-
tion to the relationship between more sleep would have been n = 116. In contrast, a sample size
problems and a larger total thalamic volume. There of n = 13 would have been required to detect such
was a signiﬁ cant correlation between the sleep a signiﬁ cant correlation between the sleep score
score and right thalamic size in patients (r
and right thalamic size suggesting our sample size
P = 0.70,
95% CI 0.21–0.91, p = 0.012; Fig. 1b) but not in of n = 12 CFS patients sufficed to deem this
relationship “truly” signiﬁ cant.
P = −0.08, p = 0.81). In contrast, the cor-
relation coefﬁ cients between the sleep score and
Total thalamic volume and thalamic size of
left thalamic size was not signiﬁ cant in patients either side did not signiﬁ cantly correlate with CFS
severity and duration since the ﬁ rst CFS-related
P = 0.23, p = 0.48) and in controls (rP = −0.41,
p = 0.19). In CFS patients the number of cases symptom occurred.
required to discover a statistically signiﬁ cant cor-
In addition, we found six patients who scored
relation between the sleep score and left thalamic above the cut-off level suggestive of clinical
size (p = 0.05, two-sided) with a power of 80% depression on the HADS depression subscale.
More precisely, one patient was severely depressed,
three patients were moderately depressed, and two
patients were mildly depressed. However, com-
pared to patients scoring in the normal range of
depressive symptom levels, depressed patients
showed no signiﬁ cant difference in right thalamic
size, left thalamic size or total volume of thalamic
nuclei, respectively. Consistently, depressive
symptom scores were not signiﬁ cantly associated
with any of the thalamic volume measures across
all patients. A stratiﬁ ed analysis in terms of gender
revealed no signiﬁ cant correlation between sleep
otal thalamic volume [mm3]
problem scores and any thalamic volume measure
in patients and controls.
Sleep problem score [points]
We used MR-based volumetry to investigate the
Figure 1a. Sleep quality and total thalamic size in CFS patients.
relationship between thalamic size and sleep
Clinical Medicine: Psychiatry 2008:1
Begré et al
quality in 12 patients with CFS and 12 age-, and compare patient groups with sleep complaints with
gender-matched healthy, non-CFS controls. Clearly, relatively longer and shorter total sleep times or to
our study must be viewed as preliminary and as assess objective measures of sleep quality (e.g. by
such may primarily serve as a basis for future, more actigraphy) are needed to resolve this. Neverthe-
extensive investigations on possibly perturbed less, as yet, there is not much evidence for a major
neurofunctional networks in CFS patients. We role of abnormalities in sleep architecture and
therefore ﬁ rst acknowledge three limitations of our objective measures of sleepiness in CFS (Ball et al.
study which imply that following interpretation and 2004; Watson et al. 2004). Third, our psychiatric
discussion of ﬁ ndings must apply caution pending and psychometric assessment was limited by the
replication in larger samples of main ﬁ ndings with fact that we did not assess life-time and current
more sophisticated statistical analyses.
psychiatric disorders in a structured clinical inter-
First, our sample size did not allow us to apply view. In line with other studies (Wessely et al.
complicated statistical regression modelling 1999), however, half of our CFS patients had
because this can be unstable when controlling for depressive symptom scores above a cut-off level
too many covariates (Greenland et al. 2000; previously shown to perform well in identifying
Papoulis, 1990). As a consequence we could not CFS patients with clinically diagnosed depression
control our analyses for depressed mood and CFS (Henderson and Tannock, 2005).
duration. Although depression and the time since
Owing to the preliminary nature of our study, a
the ﬁ rst CFS-related symptoms occurred were not series of further limitations also deserve consider-
individually related to thalamic size, these vari-
ation. Fellow researchers may want to address
ables could, when aggregated and applied to a these drawbacks in future study designs. Although
larger sample, account for some of the relationship the orientation of the regression line between the
between sleep quality and thalamic size observed sleep problem score and total thalamic volume in
in CFS patients. Some symptom domains of patients with CFS was not evidently inﬂ uenced by
depression and CFS may overlap and relate to outliers in statistical terms, more data is needed to
cognitive, limbic, and autonomic functions in all conﬁ rm the validity of the ﬁ ndings. Notably, our
of which the thalamus plays a role. However, the data do not preclude the possibility that there was
lack of association between depressive symptoms also a clinically relevant direct association between
and thalamic volumes in our patients is in accor-
subjective sleep problems and left thalamic size
dance with some of previous post-mortem studies since power analysis showed that the sample was
showing that depressed patients (without CFS) and clearly too low to detect a statistically signiﬁ cant
non-depressed subjects had similar volumes of relationship in our patients. The relationship
subcortical nuclei and of the thalamus in particular between sleep and thalamic size in CFS patients is
(Vasic et al. 2008; Bielau et al. 2005; Caetano et al. difﬁ cult to interpret as there was no difference in
2001). In contrast, a recent study found that patients thalamic size between patients and controls which
with a major depressive disorder, and thus consid-
observation might also reﬂ ect a power issue. More-
erably higher levels of depressed mood than over, conﬁ dence intervals are rather wide in rela-
observed in our CFS patients, had enlarged total tion to the small sample size and therefore difﬁ cult
thalamus volume. Moreover, neuroendocrine func-
to interpret. The questionnaire to assess sleep
tion is altered with longer duration of CFS (Gaab problems is widely applied; however, to the best
et al. 2004) such that it could be theorized that of our knowledge, its validity and reliability has
alteration in cortical structures might also occur not been tested in patients with CFS. Our control
with longer duration of CFS symptoms. However, subjects were mainly recruited among hospital
there was great variability in the duration since ﬁ rst employees, which are a selected population,
CFS-related symptoms occurred making it difﬁ cult thereby potentially limiting the generalizability of
to detect such an association.
our data. We did not relate thalamic volume to total
Second, our study does not allow the distinction brain volume. However, gender may be viewed to
of whether thalamic size correlated with sleep some extent as a proxy measure of total brain
problems per se, total sleep time, or some other volume that was indirectly considered by the nature
complaints that are unrelated to sleep problems but of our case-control design.
similarly prevalent in CFS patients (e.g. anxiety).
We found a positive correlation between more
More elaborated study designs, for instance, to sleep complaints and total and right thalamic size
Clinical Medicine: Psychiatry 2008:1
A preliminary Study
in CFS patients but not in controls. The observation afferent thalamic inputs implicated in arousal and
suggests that the more sleep disturbance CFS attention (Sherman, 2005) suggesting the thalamus
patients perceived, the greater was their right tha-
could be a major region in the processing of fatigue.
lamic size. We may interpret that the size of the For instance neurodegeneration in thalamic gray
thalamus could be associated with the pathology matter, which can also be found in neuropsychiatric
of sleep processing in CFS. In contrast, the CFS diseases other than CFS, might contribute to the
symptom score was not associated with thalamic genesis of “fatigue”. Neuropathological ﬁ ndings
size. However, besides the core symptom fatigue, in patients with fatal insomnia (Nose et al. 2002)
the CFS symptom scale considers eight additional and Creutzfeldt-Jakob disease (Taratuto et al.
symptoms of which only one relates to subjectively 2002) showed thalamic neuronal loss, spongiform
perturbed sleep (i.e. “unrefreshing sleep”). There-
changes and prominent gliosis, and altered tha-
fore, we may speculate that subjectively perturbed lamic function in sleep regulation, all of which can
sleep is perhaps the core dimension of CFS relating also change thalamic size. Aside from any focal
to thalamic size, whereas thalamus volume might lesion or inﬂ ammation, MRI, MRS and postmor-
play less of a role in the relationship with the other tem histopathology studies in multiple sclerosis
CFS symptom dimensions. Alternatively, the lack revealed reduced neuronal thalamic density by
of a statistical association between the CFS severity N-acetylaspartate (NAA) concentration and
score and thalamic size might also indicate that reduced thalamic volume (Cifelli et al. 2002;
other central structures than the thalamus underlie Wylezinska al. 2003). In poliomyelitis, postmortem
total CFS symptom severity.
histopathology demonstrated consistent presence
Consistent with previous ﬁ ndings, we found no of poliovirus lesions also in thalamic nuclei (Bruno
hemispheric asymmetry in the thalamic size of et al. 1998). Given thalamic size in both hemi-
either side in healthy subjects (Szabo, 2003). In spheres did not differ between our CFS patients
this ﬁ rst volumetry study of the thalamus in CFS, and controls, one preliminary interpretation could
we also found that right and left thalamic size were be that neuronal loss had not occurred. Four of our
similar in CFS patients and that thalamic size was patients had been on psychotropic medication
not different in patients compared to controls. (i.e. paroxetine and methylphenidate, amitriptyline,
Besides a power issue as mentioned above, we doxepine, and morphine within the last 4 months,
offer three possible explanations for this observa-
all of which could possibly affect sleep subjectively
tion. First, we hypothesize that there could be and objectively. An inﬂ uence of antidepressants
antecedent microstructural thalamic differences in (Mayers and Baldwin, 2005), methylphenidate
subjects with sleep disturbances related to CFS (Banerjee et al. 2004) and morphine (Villablanca,
which are not present in healthy people. Second, 2004) on sleeping patterns and arousal is well
even though not detected by our macroscopic documented. However, retrospectively, our patients
volumetric issue there could be some differences reported no deﬁ nite effect on sleep quality between
on a microstructural level inﬂ uencing sleep pattern before and after they were prescribed these medi-
(Begré et al. 2006; Yamada et al. 2006). Further cations. We therefore feel conﬁ dent that psycho-
structural evaluation by diffusion tensor imaging tropics did not inﬂ ate a spurious result in terms of
could enlighten this suggestion. Third, augmented the relationship between perceived sleep problems
thalamic perfusion was previously shown in and right thalamic size in our CFS patients.
patients with CFS compared to healthy matched
Moreover, a possible inﬂ uence of psychotropic
controls (Tomoda et al. 2000; MacHale et al. medication on thalamic volume was documented
2000). Similarly, functional MRI in fatigued solely for paroxetine in obsessive-compulsive
people with multiple sclerosis showed an inverse patients who experienced reduction in symptom
correlation between fatigue severity and activation severity after psychotropic treatment (Rosenberg
of ipsilateral thalamus during simple motor tasks et al. 2000; Gilbert et al. 2000), however, none of
(Filippi et al. 2002). We may thus hypothesize that, our patients reported obsessive-compulsive symp-
as a result of chronic sleep impairment, a compen-
toms. Also, metabolic thalamic function was nor-
satory increase of thalamic perfusion could have malized with paroxetine treatment in a PET-study
led to thalamic hypertrophy in our CFS patients.
on patients with major depression. A SPECT study
At least 90% of thalamic synapses arise from showed deactivation in the right thalamus with
brain regions making up the many non-primary citalopram (Carey et al. 2004), but it is unknown
Clinical Medicine: Psychiatry 2008:1
Begré et al
whether paroxetine could have a similar effect in Bassetti, C., Aldrich, M.S., Chervin, R.D. and Quint, D. 1996. Sleep apnea
our CFS patients. Increased regional cerebral blood
in patients with transient ischemic attack and stroke: a prospective
study of 59 patients. Neurology, 47:1167–73.
ﬂ ow in the right thalamus after methylphenidate Basso, M.A., Uhlrich, D. and Bickford, M.E. 2005. Cortical function: a view
prescription was shown using PET (Schweitzer
from the thalamus. Neuron., 45:485–8.
et al. 2004). Altogether, we acknowledge that our Begré, S., Federspiel, A., Kiefer, C., Schroth, G., Strik, W.K. and Dierks, T.
sample size did not allow us to perform meaningful
2006. Alterations of white matter connectivity in ﬁ rst episode schizo-
phrenia. Neurobiol. Dis., 22:702–9.
statistics on unmedicated patients alone.
Bielau, H., Trubner, K., Krell, D., Agelink, M.W., Bernstein, H.G., Stauch, R.,
Although our results are preliminary they may
Mawrin, C., Danos, P., Gerhard, L., Bogerts, B. and Baumann, B. 2005.
contribute to the growing body of research sug-
Volume deﬁ cits of subcortical nuclei in mood disorders A postmortem
study. Eur. Arch. Psychiatry Clin. Neurosci., 255(6):401–12.
gesting that the thalamus does not only accumulate Bruno, R.L., Creange, S.J. and Frick, N.M. 1998. Parallels between post-
sensory information before sending it to the cortex,
polio fatigue and chronic fatigue syndrome: a common pathophysiol-
but rather plays a major role as a work station
ogy? Am. J. Med., 105:66S–73S.
Busichio, K., Tiersky, L.A., DeLuca, J. and Natelson, B.H. 2004. Neuro-
dynamically regulating information processing of
psychological deﬁ cits in patients with chronic fatigue syndrome.
a complicated arrangement of neuronal connec-
J. Int. Neuropsychol. Soc., 10:278–85.
tions organized in feedforward, feedback, and loop Byne, W., Buchsbaum, M.S., Mattiace, L.A., Hazlett, E.A., Kemether, E.,
circuits (Basso et al. 2005). Therefore, the thalamus
Elhakem, S.L., Elhakem, S.L., Purohit, D.P., Haroutunian, V. and
Jones, L. 2002. Postmortem assessment of thalamic nuclear volumes
could critically be involved in explaining the
in subjects with schizophrenia. Am. J. Psychiatry, 159:59–65.
pathogenesis of the heterogenous symptom Caetano, S.C., Sassi, R., Brambilla, P., Harenski, K., Nicoletti,
complex of CFS.
M., Mallinger, A.G., Frank, E., Kupfer, D.J., Keshavan, M.S. and Soares,
J.C. 2001. MRI study of thalamic volumes in bipolar and unipolar
patients and healthy individuals. Psychiatry Res., 108:161–8.
Carey, P.D., Warwick, J., Niehaus, D.J., van der Linden, G., van Heerden, B.B.,
Harvey, B.H., Seedat, S. and Stein, D.J. 2004. Single photon emission
Using MR-based volumetry, we found a positive
computed tomography (SPECT) of anxiety disorders before and after
association between subjective sleep problems and
treatment with citalopram. BMC. Psychiatry, 4:30.
thalamic volume in patients with CFS suggesting Chaudhuri, A. and Behan, P.O. 2004. In vivo magnetic resonance spectros-
copy in chronic fatigue syndrome. Prostaglandins Leukot. Essent.
an association between impaired sleep quality and
Fatty Acids, 71:181–3.
thalamic functioning in CFS. This notion corre-
Chee, M.W. and Choo, W.C. 2004. Functional imaging of working memory
sponds to the literature on the importance of
after 24 hr of total sleep deprivation. J. Neurosci., 24:4560–7.
Cifelli, A., Arridge, M. and Jezzard, P. 2002. Esiri MM, Palace J.,
thalamus in the regulation of sleep and to the dis-
Matthews PM. Thalamic neurodegeneration in multiple sclerosis.
tinctive features of sleep problems frequently
Ann. Neurol., 52:650–3.
reported by patients with CFS. Because this was Danos, P. 2004. Pathology of the thalamus and schizophrenia—an overview.
not a mechanistic study, more elaborated study
Fortschr. Neurol. Psychiatr., 72:621–34.
Danos, P., Baumann, B., Kramer, A., Bernstein, H.G., Stauch, R., Krell, D.,
designs are needed to investigate to what extent
Falkai, P. and Bogerts, B. 2003. Volumes of association thalamic
the thalamus could be part of the neurofunctional
nuclei in schizophrenia: a postmortem study. Schizophr. Res.,
network pertinent to CFS. Our ﬁ ndings must be
Filippi, M., Rocca, M.A., Colombo, B., Falini, A., Codella, M., Scotti, G.
understood as preliminary and more research in
and Scotti, G. 2002. Functional magnetic resonance imaging corre-
larger samples is clearly required to increase
lates of fatigue in multiple sclerosis. Neuroimage, 15:559–67.
knowledge of the interplay of brain structures, their Fukuda, K., Straus, S.E., Hickie, I., Sharpe, M.C., Dobbins, J.G. and
Komaroff, A. 1994. International Chronic Fatigue Syndrome Study
functioning, sleep, and the subjective state of
Group. In: The chronic fatigue syndrome: a comprehensive approach
fatigue in CFS patients. Such studies might advance
to its deﬁ nition and study. Ann. Intern. Med., 121:953–9.
our currently limited understanding of central Gaab, J., Engert, V., Heitz, V., Schad, T., Schürmeyer, T.H. and Ehlert, U.
processing of fatigue not only in CFS but also in
2004. Associations between neuroendocrine responses to the Insulin
Tolerance Test and patient characteristics in chronic fatigue syndrome.
a variety of other neuropsychiatric diseases and
J. Psychosom. Res., 56:419–24.
disorders in order to develop target interventions Geuze, E., Vermetten, E. and Bremner, J.D. 2005. MR-based in vivo hip-
in the future for the often disabled patient suffering
pocampal volumetrics: 2. Findings in neuropsychiatric disorders.
Mol. Psychiatry, 10:160–84.
Gilbert, A.R., Moore, G.J., Keshavan, M.S., Paulson, L.A., Narula, V., Mac
Master, F.P., Stewart, C.M. and Rosenberg, D.R. 2000. Decrease in
thalamic volumes of pediatric patients with obsessive-compulsive
disorder who are taking paroxetine. Arch. Gen. Psychiatry, 57:449–56.
Ball, N., Buchwald, D.S., Schmidt, D., Goldberg, J., Ashton, S. and Armitage, R.
Goldberg, D. 1985. Identifying psychiatric illness among general medical
2004. Monozygotic twins discordant for chronic fatigue syndrome:
patients. Br. Med. J. (Clin. Res. Ed), 291:161–2.
objective measures of sleep. J. Psychosom. Res., 56:207–12.
Greenland, S., Schwartzbaum, J.A. and Finkle, W.D. 2000. Problems due
Banerjee, D., Vitiello, M.V. and Grunstein, R.R. 2004. Pharmacotherapy
to small samples and sparse data in conditional logistic regression
for excessive daytime sleepiness. Sleep Med. Rev., 8:339–54.
analysis. Am. J. Epidemiol., 151:531–9.
Clinical Medicine: Psychiatry 2008:1
A preliminary Study
Henderson, J.M., Carpenter, K., Cartwright, H. and Halliday, G.M. 2000.
Sherman, S.M. 2005. Thalamic relays and cortical functioning. Prog. Brain
Loss of thalamic intralaminar nuclei in progressive supranuclear palsy
and Parkinson‘s disease: clinical and therapeutic implications. Brain,
Silverstone, P.H. 1994. Poor efﬁ cacy of the Hospital Anxiety and Depression
Scale in the diagnosis of major depressive disorder in both medical
Henderson, M. and Tannock, C. 2005. Use of depression rating scales in
and psychiatric patients. J. Psychosom. Res., 38:441–50.
chronic fatigue syndrome. J. Psychosom. Res., 59:181–4.
Steptoe, A., O’Donnell, K., Marmot, M. and Wardle, J. 2008. Positive affect,
Herrmann, C., Buss, U. and Snaith, R.P. 1995. HADS-D—Hospital Anxiety and
psychological well-being, and good sleep. J. Psychosom Res.,
Depression Scale. Deutsche Version: Ein Fragebogen zur Erfassung von
Angst und Depressivität in der somatischen Medizin. Berne: Huber
Szabo, C.A., Lancaster, J.L., Xiong, J., Cook, C. and Fox, P. 2003. MR
Jenkins, C.D., Stanton, B.A., Niemcryk, S.J. and Rose, R.M. 1988. A scale
imaging volumetry of subcortical structures and cerebellar hemi-
for the estimation of sleep problems in clinical research. J. Clin.
spheres in normal persons. AJNR. Am. J. Neuroradiol., 24:644–7.
Taratuto, A.L., Piccardo, P., Reich, E.G., Chen, S.G., Sevlever, G.,
Kudielka, B.M., von Kanel, R., Gander, M.L. and Fischer, J.E. 2004. Effort-
Schultz, M., Luzzi, A.A., Rugiero, M., Abecasis, G., Endelman, M.,
reward imbalance, overcommitment and sleep in a working popula-
Garcia, A.M., Capellari, S., Xie, Z., Lugaresi, E., Gambetti, P., Dlouhy,
tion. Work and Stress, 18:167–78.
S.R. and Ghettti, B. 2002. Insomnia associated with thalamic involve-
Lugaresi, E. 1992. The thalamus and insomnia. Neurology, 42:28–33.
ment in E200K Creutzfeldt-Jakob disease. Neurology, 58:362–7.
MacHale, S.M., Lawrie, S.M., Cavanagh, J.T., Glabus, M.F. and Murray,
Tomoda, A., Miike, T., Yamada, E., Honda, H., Moroi, T., Ogawa, M.,
C.L. 2000. Goodwin GM et al. Cerebral perfusion in chronic fatigue
Ohtani, Y. and Morishita, S. 2000. Chronic fatigue syndrome in
syndrome and depression. Br. J. Psychiatry, 176:550–6.
childhood. Brain Dev., 22:60–4.
Mayers, A.G. and Baldwin, D.S. 2005. Antidepressants and their effect on
van der Linden, D., Frese, M. and Meijman, T.F. 2003. Mental fatigue and
sleep. Hum. Psychopharmacol., 20(8):533–59.
the control of cognitive of cognitive processes: effects on persevera-
Moncrieff, G. and Fletcher, J. 2007. Tiredness. BMJ., 334:1221.
tion and planning. Acta. Psychol., 113:45–65.
Montagna, P., Gambetti, P., Cortelli, P. and Lugaresi, E. 2003. Familial and
Vasic, N., Walter, H., Höse, A. and Wolf, R.C. 2008. Gray matter reduction
sporadic fatal insomnia. Lancet Neurol., 2:167–76.
associated with psychopathology and cognitive dysfunction in uni-
Nofzinger, E.A. 2006. Neuroimaging of sleep and sleep disorders. Curr.
polar depression: A voxel-based morphometry study. J. Affect Disord.,
Neurol. Neurosci. Rep., 6:149–55.
Nose, I., Ookawa, T., Tanaka, J., Yamamoto, T., Uchimura, N., Maeda, H.
Villablanca, J.R. 2004. Counterpointing the functional role of the forebrain
and Kuwahara, H. 2002. Decreased blood ﬂ ow of the left thalamus
and of the brainstem in the control of the sleep-waking system.
during somnolent episodes in a case of recurrent hypersomnia.
J. Sleep Res., 13(3):179–208.
Psychiatry Clin. Neurosci., 56:277–8.
Watson, N.F., Jacobsen, C., Goldberg, J. and Kapur, V. 2004. Buchwald D.
Papoulis, A. 1990. Probability and Statistics. Prentence-Hall International
Subjective and objective sleepiness in monozygotic twins discordant
for chronic fatigue syndrome. Sleep, 27:973–7.
Reeves, W.C., Heim, C., Maloney, E.M., Youngblood, L.S., Unger, E.R.,
Wessely, S., Nimnuan, C. and Sharpe, M. 1999. Functional somatic
Decker, M.J., Jones, J.F. and Rye, D.B. 2006. Sleep characteristics
syndromes: one or many? Lancet, 354:936–9.
of persons with chronic fatigue syndrome and non-fatigued controls:
Wylezinska, M., Cifelli, A., Jezzard, P., Palace, J., Alecci, M. and
results from a population-based study. BMC. Neurol., 6:41.
Matthews, P.M. 2003. Thalamic neurodegeneration in relapsing-
Rosenberg, D.R., Benazon, N.R., Gilbert, A., Sullivan, A. and Moore, G.J.
remitting multiple sclerosis. Neurology, 60:1949–54.
2000. Thalamic volume in pediatric obsessive-compulsive disorder
Yamada, K., Matsuzawa, H., Uchiyama, M., Kwee, I.L. and Nakada, T.
patients before and after cognitive behavioral therapy. Biol.
2006. Brain developmental abnormalities in Prader-Willi syndrome
detected by diffusion tensor imaging. Pediatrics., 118:442–8.
Schweitzer, J.B., Lee, D.O., Hanford, R.B., Zink, C.F., Ely, T.D., Tagamets,
Young, K.A., Bonkale, W.L., Holcomb, L.A., Hicks, P.B. and German, D.C.
M.A., Hoffmann, J.M., Grafton, S.T. and Kilts, C.D. 2004. Effect of
2008. Major depression, 5HTTLPR genotype, suicide and antidepres-
methylphenidate on executive functioning in adults with attention-
sant inﬂ uences on thalamic volume. Br. J. Psychiatry, 192:285–9.
deﬁ cit/hyperactivity disorder: normalization of behavior but not
Zigmond, A.S. and Snaith, R.P. 1983. The hospital anxiety and depression
related brain activity. Biol. Psychiatry, 56:597–606.
scale. Acta. Psychiatr. Scand., 67:361–70.
Clinical Medicine: Psychiatry 2008:1