Skip to main content
NCBI home page
Human Brain Mapping logoLink to Human Brain Mapping
. 2016 Aug 29;38(1):283–292. doi: 10.1002/hbm.23360

Fatigue in Parkinson's disease: The contribution of cerebral metabolic changes

Sang Soo Cho 1,2, Kelly Aminian 1,2, Crystal Li 2, Anthony E Lang 3, Sylvain Houle 2, Antonio P Strafella 1,2,3,✉
PMCID: PMC6867035  PMID: 27571419

Abstract

Fatigue is a common and disabling non‐motor symptom in Parkinson's disease associated with a feeling of overwhelming lack of energy. The aim of this study was to identify the neural substrates that may contribute to the development of fatigue in Parkinson's disease. Twenty‐three Parkinson's disease patients meeting UK Brain Bank criteria for the diagnosis of idiopathic Parkinson's disease were recruited and completed the 2‐[18F]fluoro‐2‐deoxy‐D‐glucose (FDG)‐PET scan. The metabolic activities of Parkinson's disease patients with fatigue were compared to those without fatigue using statistical parametric mapping analysis. The Parkinson's disease group exhibiting higher level of fatigue showed anti‐correlated metabolic changes in cortical regions associated with the salience (i.e., right insular region) and default (i.e., bilateral posterior cingulate cortex) networks. The metabolic abnormalities detected in these brain regions displayed a significant correlation with level of fatigue and were associated with a disruption of the functional correlations with different cortical areas. These observations suggest that fatigue in Parkinson's disease may be the expression of metabolic abnormalities and impaired functional interactions between brain regions linked to the salience network and other neural networks. Hum Brain Mapp 38:283–292, 2017. © 2016 Wiley Periodicals, Inc.

Keywords: Parkinson's disease, fatigue, regional glucose metabolism, FDG PET, brain network

Abbreviations

ACC

Anterior cingulated cortex

BDI

Beck depression inventory

CEN

Central executive network

DLPFC

Dorsolateral prefrontal cortex

DMN

Default mode network

ESS

Epworth Sleepiness scale

FDG

2‐[18F]fluoro‐2‐deoxy‐D‐glucose

FSS

Fatigue Severity Scale

FWHM

Full width at half‐maximum

LED

Levodopa equivalent dose

MNI

Montreal Neurologic Institute

MoCA

Montreal Cognitive Assessment

PCC

Posterior cingulate cortex

PDSS

Parkinson's Disease Sleep Scale

PET

Positron emission tomography

SN

Salience network

TOF

Time‐of‐flight

UPDRS

Unified Parkinson Disease Rating Scale motor score

INTRODUCTION

Fatigue is a state of profound physical and/or mental weakness, associated with a feeling of abnormal and overwhelming lack of energy, distinct both qualitatively and quantitatively from general tiredness [Brown et al., 2005]. In Parkinson's disease, this is a common and disabling non‐motor symptom affecting up to 70% of patients, and often related to disease progression [Alves et al., 2004; Friedman et al., 2007], however, fatigue has generally been found to be independent of disease severity [Friedman et al., 2011; Friedman et al., 2016].

In healthy subjects, fatigue has often been shown to be associated with various cognitive changes. Individuals with fatigue may display perseveration on the card sorting task and prolonged planning time on the tower of London task [van der Linden et al., 2003a], while others can make more systematic errors on a complex computer task [van der Linden et al., 2003b]. A relationship between fatigue and cognitive function has also been found in Parkinson's disease [Goldman et al., 2014].

To date, only a handful of neuroimaging studies have examined the neural substrates that underpin fatigue. According to some of these studies, prefrontal areas play an important role in the pathophysiological mechanism of fatigue. In particular, activity in the medial orbitofrontal cortex (Brodmann's area 10/11) seems often to be associated with the subjective sensation of fatigue [Tajima et al., 2010]. The fatigue sensation may involve as well a system composed of the insula and posterior cingulate cortex (PCC), where activation of the insula was related with the evaluation of mental effort [Otto et al., 2014] and during high‐demand, mental‐fatigue‐inducing tasks, the increase activation in the PCC was positively associated with the level of fatigue sensation [Cook et al., 2007] Others studies in Parkinson's disease patients with fatigue have shown also a reduction in serotonin transporter binding in the cingulate and amygdala and in dopamine storage capacity in the caudate and insular region [Pavese et al., 2010]. The activation abnormalities along with dysregulation of both serotonin and dopamine in cognitive and limbic regions are not surprising, given that fatigue is a complex symptom [Honig et al., 2009] strongly associated with the cognitive and limbic systems.

Within the limbic network, the insular cortex in particular, part of the salience network (SN), is a key region involved in the processing of relevant information related to the state of the body [Williamson et al., 2003]. This cortical area bridges the default mode network (DMN) into the central executive network (CEN) guiding cognitive/emotional behavior [Goulden et al., 2014; Sridharan et al., 2008]. Dysfunction in the insular cortex plays an important role in the complex symptomology of Parkinson's disease [Criaud et al., 2016; Gomez‐Esteban et al., 2011].

With this background in mind, we hypothesized that parkinsonian individuals with fatigue would likely show metabolic changes in brain regions associated with the three cognitive networks highlighted above (i.e., SN, DMN, and CEN) which may as well correlated with fatigue level and possibly influenced connectivity. Thus, our primary objective was to study the metabolism and potential role of the regions linked to these cortical neural networks (i.e., SN, DMN, and CEN) in Parkinson's disease patients with fatigue, while controlling for potential confounding effects like sleep, disease severity and depression which commonly occur in Parkinson's disease.

SUBJECTS AND METHODS

Patients

Twenty‐three Parkinson's disease patients (mean age 64.7 ± 7.2; six females) meeting UK Brain Bank criteria for the diagnosis of idiopathic Parkinson's disease patients were recruited and completed the 2‐[18F]fluoro‐2‐deoxy‐D‐glucose (FDG) positron emission tomography (PET) scan after having given written, informed consent. The study was approved by the University Health Network Research Ethics Board. Parkinsonian disability was assessed using the Hoehn and Yahr Rating Scale and the Unified Parkinson Disease Rating Scale motor score (UPDRS Ш), while being off medication for a minimum of 12 hours (overnight withdrawal). Levodopa equivalent dose (LED) for each patient was calculated based on Tomlinson et al. [2010]. To reduce possible confounding effects, Montreal Cognitive Assessment (MoCA) and Beck depression inventory (BDI) were obtained in all Parkinson's disease patients. Patients with a history of other neurological diseases, unstable psychiatric illness, sleep disorders or any medical condition that precluded them from PET imaging were excluded from the study. Fatigue Severity Scale (FSS), a self‐report instrument was used to assess levels of fatigue and its effect on daily functioning [Krupp et al., 1989]. This provides nine statements addressing fatigue's effects on daily functioning, querying its relationship to motivation, physical activity, work, family, and social life. Each item has seven response options from 1 (disagree) to 7 (strongly agree). The final FSS scoring is done by calculating the average response to the questions (FSS total score dividing by nine). A total score higher than 36 can be classified as fatigued. Patients were classified without fatigue (n = 12) or with fatigue (n = 11) on the basis of their total score. The Epworth Sleepiness scale (ESS) and Parkinson's Disease Sleep Scale (PDSS) were used to assess daytime sleepiness [Kumar et al., 2003] and nighttime sleep disturbances [Chaudhuri et al., 2002], respectively.

Image Acquisition

PET scans were obtained using a high resolution PET/CT, Siemens mCT (Siemens Medical Solutions USA, Inc.) operating in 3D mode with an in‐plane resolution of approximately 4.6 mm full width at half‐maximum (FWHM). The patients fasted at least 6 hours before the scan. After the intravenous bolus injection of 5 mCi of FDG, patients were allowed to rest in a dimly lit room for 45 min during the uptake phase. Five minutes of emission scan were acquired followed by a low dose (0.2 mSv) CT scan for attenuation correction. To minimize subjects' head movements in the PET scanner, we used a custom‐made thermoplastic facemask together with a head‐fixation system (ORFIT mask system). PET images were reconstructed with time‐of‐flight (TOF) and displayed in 256 × 256 matrix (pixel size = 1.59 × 1.59 mm). Patients were imaged after overnight withdrawal of their antiparkinsonian medications. This withdrawal time is standard practice in parkinsonian research [Defer et al., 1999] and ensures to standardize and minimize the effect of medication while patients are still functional.

PET Analysis

All PET images were preprocessed and analyzed using Statistical Parametric Mapping 8 (SPM 8; Wellcome Department of Imaging Neuroscience, London, UK, http://www.fil.ion.ucl.ac.uk/spm). Images were transformed into Montreal Neurologic Institute (MNI) template using a standardized PET template image to remove the individual anatomical variability. We used the normalization algorithm provided by SPM 8 that employs a 12‐parameter affine transformation followed by nonlinear deformations. Then, the spatially normalized images were smoothed with a FWHM 8‐mm Gaussian kernel to increase the signal‐to‐noise ratio and to remove subtle variations in anatomic structures. Individual voxel counts were normalized versus total brain counts (proportional scaling in SPM) to remove individual global brain metabolism differences.

After spatial and count normalization, significant differences in glucose metabolism between Parkinson's disease with and without fatigue were estimated at every voxel using t‐statistics. Differences of regional brain glucose metabolism between the two PD groups were investigated at the whole‐brain level using SPM 8, and significant clusters were tested if it was located within our a priori region of interests (ROIs) (i.e., SN: insula, anterior cingulate cortex, DMN: posterior cingulate/precuneus, medial prefrontal cortex, inferior parietal lobule, medial temporal lobe, lateral temporal cortex, CEN: dorsolateral prefrontal cortex (DLPFC) and posterior parietal cortex). Each ROI was defined using WFU‐PickAtlas toolbox (http://www.fmri.wfubmc.edu/cms/software) [Maldjian et al., 2003]. A height threshold of uncorrected P < 0.005 and an extent threshold of 100 voxels were used for generating the initial t‐map to determine the predicted peaks and visualization. Brain regions that showed P < 0.001 at voxel level were considered significant. A subsequent correlation analysis between FSS and brain glucose metabolism of the statistically significant regional cluster was conducted using SPSS 16.0 (SPSS Inc., Chicago, Illinois). Considering the possible comorbidity of depression shown in other studies [Defer et al., 1999; McDonald et al., 2003], BDI was used as a nuisance variable for the analysis.

In order to examine the whole‐brain metabolic connectivity in the SN and CEN, we selected seed regions (right insula for SN and PCC for DMN) from the significant clusters from the group analysis. Count from normalized FDG mean value of seed areas were extracted using SPM‐VOI tool, and a seed‐based connectivity analyses was performed to test function relationship between cognitive networks using SPM 8 (uncorrected P < 0.005, k = 100 voxels).

The local maxima of each cluster in the Montreal Neurological Institute (MNI) coordinates were transformed using Talairach Daemon (http://www.talairach.org.), expressed in Talairach coordinates.

For the demographic comparisons, student t‐test was used, and for psychometric analysis, Pearson's partial correlation analysis (two‐tailed) was performed using SPSS 16.0. P < 0.05 was considered significant for all analysis. All data were expressed as mean ± SD.

RESULTS

The clinical features of each group are summarized in Table 1. There were no significant differences between groups in age (t = 0.94, P = 0.36), year education (t = 0.92, P = 0.37), MoCA (t = 1.22, P = 0.24), BDI (t = 1.23, P = 0.23), UPDRS Ш (t = 0.70, P = 0.49), LED (t = 0.90, P = 0.38) and H‐Y score (t = 0.54, P = 0.60). None of PD patients in this study was taking antidepressant medication. Controlling for the effect of sleep, there were no significant differences in sleep quality between Parkinson's disease with and without fatigue when measured with PDSS (t = 0.59, P = 0.57) and ESS (t = 0.19, P = 0.85). We did not observe any relationship between fatigue level and disease severity index (UPDRS III: r = −0.02, P = 0.45; H‐Y: r = 0.15, P = 0.23) nor daily pharmacological therapy (LED: r = −0.24, P = 0.27; levodopa: r = −0.11, P = 0.63; dopamine‐agonist: r = −0.16, P = 0.46) indicating that neither disease severity nor pharmacological therapy was directly related with fatigue in Parkinson's disease patients. We also tested the relationship between the level of fatigue in each patient and overall cognitive function and mood status. There was no relationship with MoCA score (r = −0.21, P = 0.33), indicating that the level of fatigue in our cohort of patients was not directly related with the overall cognitive function. BDI as well did not show significant level of correlation with fatigue but showed trend level of relationship (r = 0.39, P = 0.06), and for this reason BDI was used as nuisance variance for analysis.

Table 1.

Demographics of Parkinson's disease patient without (PD‐Fatigue−) and with fatigue (PD‐Fatigue+)

PD‐fatigue− PD‐fatigue+ P value
N (F:M) 12 (2:10) 11 (4:7) 0.28a
Mean age year (SD, range) 63.3 (8.4, 52–76) 66.2 (5.6, 54–72) 0.36
UPDRS Ш (SD) 37.1 (12.2) 33.5 (12.8) 0.49
Hoehn and Yahr Scale (SD) 2.1 (0.5) 2.0 (0.2) 0.60
LED mg (SD) 776.6 (326.0) 665.9 (259.1) 0.38
Levodopa 518.8 (362.6) 506.8 (215.1) 0.93
DA agonist 145.8 (137.7) 86.4 (119.0) 0.23
BDI (SD) 8.0 (4.0) 9.9 (3.4) 0.23
MoCA (SD) 27.5 (1.9) 26.2 (3.2) 0.24
ESS (SD) 9.2 (4.3) 9.5 (5.4) 0.85
PDSS (SD) 107.0 (16.4) 102.7 (18.6) 0.57
FSS (SD) 3.2 (0.8) 5.0 (0.9) P < 0.0001
Open in a new tab

Chi‐square analysis; SD: Standard Deviation; N.S: not significant; UPDRS: unified Parkinson's disease rating scale LED: levodopa equivalent dose; BDI: Beck depression Inventory; MoCA: Montreal Cognitive Assessment; ESS: Epworth sleepiness scale; PDSS: Parkinson's Disease Sleep Scale; FSS: Fatigue Severity Scale (total score dividing by nine).

To test the first hypothesis, we conducted a voxel‐wise group comparison analysis to test changes in regional glucose metabolism. The results are shown in Figure 1 and Table 2. The imaging analysis showed that the Parkinson's disease group exhibiting higher level of fatigue presented lower in regional glucose metabolism in the right insular cortex (BA 13) region of the SN, in the left superior (BA 41) and right middle (BA 21) temporal gyri. Greater regional glucose metabolism was found in the bilateral posterior cingulate cortex (PCC, BA 29), a region known to be part of the DMN.

Figure 1.

Figure 1

Open in a new tab

Brain regions showing a significant decrease (cold color scale) and increase (hot color scale) in cerebral glucose metabolism on resting state FDG PET scans in Parkinson's disease patient with fatigue, controlled for depression. Significant voxels in statistical T‐maps were overlaid on “ch256” brain template using MRIcroGL (http://www.mccauslandcenter.sc.edu/mricrogl/).

Table 2.

Brain regions showing hypo/hyper metabolism in Parkinson's disease patient with fatigue (PD‐Fatigue−) and Parkinson's disease patient without fatigue (PD‐Fatigue+)

Region BA Coordinatea T‐value Cluster size
X Y Z
PD‐fatigue− > PD‐fatigue+
Lt Superior Temporal Gyrus BA 41 −50 −25 7 4.82 212
Rt Insular BA 13 44 12 −4 3.88 137
Rt Middle Temporal Gyrus BA 21 53 −24 −7 3.72 103
PD‐fatigue− < PD‐fatigue+
Bi Posterior Cingulate BA 29 −4 −44 13 4.29 126
Open in a new tab

Lt: left, Rt: right, Bi: bilateral, BA: Brodmann's area.

Talairach coordinate (mm).

To test our second hypothesis, we extracted relative metabolism from the brain region that showed group differences and tested if there was any significant correlation with individual fatigue level. The level of regional glucose metabolism in those brain regions correlated with fatigue score when controlling for BDI score (Fig. 2). The right insular cortex showed a significant negative correlation (r = −0.49, P = 0.02) along with the left superior temporal gyrus (r = −0.53, P = 0.03), implying that lower metabolism in these regions was associated with a higher fatigue level. In contrast, the region associated with DMN (i.e., PCC) showed a positive correlation (r = 0.57, P = 0.006) with fatigue score.

Figure 2.

Figure 2

Open in a new tab

Correlation between regional glucose metabolism and FSS score in the right insular cortex, left STG, and PCC, controlled for depression. Open circle indicate PD with fatigue and closed circle indicate PD without fatigue. The correlation coefficients r and corresponding p (two‐tailed) values are shown for each correlation.

We did not find any significant metabolic differences between two groups in CEN related brain regions, thus the seed‐based connectivity analysis was conducted using the right insula and PCC as a seed to test the third hypothesis. The connectivity analysis using the right insular region within the SN as the seed showed a clear pattern of functional relationship with different cortical/subcortical regions (Fig. 3 and Table 3). In Parkinson's disease without fatigue, the metabolic activity in bilateral prefrontal regions (part of the CEN), right thalamus and putamen showed a positive correlation with the insular activity, while other brain regions associated with somatosensory regions (i.e., right post‐central gyrus, left superior parietal lobule), bilateral precuneus within the DMN, and parahippocampal gyrus were inversely related (Fig. 3A). In contrast, in Parkinson's disease patients with fatigue, several of these functional correlations with the insular activity did not reach a significant level (i.e., prefrontal, somatosensory) and new positive correlations emerged with metabolic activities in the contralateral insular, bilateral thalamus and inferior frontal gyri. A negative correlation was found only in the ipsilateral middle orbitofrontal cortex (Fig. 3B). Comparing of two patient groups, bilateral DLPFC and right orbitofrontal cortex showed less metabolic correlation with right insula in PD with fatigue than without fatigue (P < 0.001, k >100), these results reflect the disrupted functional connectivity between these two brain regions in patients with fatigue.

Figure 3.

Figure 3

Open in a new tab

Brain regions where significant functional correlations with the right insular seed were overlaid on “ch256” brain template (upper row) and its topographical connectivity map (bottom row), were shown in Parkinson's disease patient without fatigue (A) and Parkinson's disease patient with fatigue (B). Seed area marked on yellow, red node/path indicates positive connectivity and blue node/path indicate negative connectivity (bottom row).

Table 3.

Brain regions that showed significant functional correlate with right insular seed in each Parkinson's disease group (P < 0.005 uncorrected for multiple comparisons with 100 voxels of spatial extended threshold)

Region BA Coordinatea T‐value Cluster size
X Y Z
PD‐Fatigue−
<Positive>
Rt Inferior Frontal Gyrus BA 47 43 13 −4 9.74 1429
Rt Inferior Frontal Gyrus BA 9 51 19 22 5.31
Rt Inferior Frontal Gyrus BA 44 44 9 31 3.97
Rt Inferior Frontal Gyrus BA 45 54 24 17 3.90
Rt Middle Frontal Gyrus BA 10 39 51 5 4.49
Rt Middle Frontal Gyrus BA 10 28 49 2 3.85
Rt Middle Frontal Gyrus BA 9 30 36 30 3.27
Rt Superior Frontal Gyrus BA 10 17 53 7 4.53
Rt Superior Temporal Gyrus BA 22 53 11 −5 5.59
Rt Thalamus 8 −11 4 4.88 317
RT Putamen 15 0 5 4.87
Lt Inferior Frontal Gyrus BA 9 −50 16 23 4.54 137
<Negative>
Lt ParaHippocampal Gyrus −30 6 −27 4.89 211
Rt ParaHippocampal Gyrus 32 6 −27 4.45 101
Lt Superior Parietal Lobule BA 5 −24 −45 63 4.29 158
Rt Postcentral Gyrus BA 4 20 −26 69 3.87 126
Rt Precuneus BA 7 12 −46 58 3.86 120
Lt Precuneus BA 7 −10 −47 50 3.58
Lt Paracentral Lobule BA 6 −4 −23 49 3.57 166
PD‐Fatigue+
<Positive>
Rt Inferior orbitofrontal Gyrus BA 47 53 21 −4 5.74 734
Lt Inferior Frontal Gyrus BA 44 −55 7 18 5.91 2044
Lt Inferior Frontal Gyrus BA 47 −53 24 −3 5.91
Lt Superior Temporal Gyrus BA 22 −50 −10 −3 5.49
Lt Insular BA 13 −40 2 4 5.84
Lt Thalamus −4 −12 −2 5.17 801
Rt Thalamus 8 −12 2 4.88
Rt Caudate 6 17 −5 4.22 152
<Negative>
Rt Middle Orbitofrontal Cortex BA 11 26 44 −14 4.17 232
Open in a new tab

Lt: left, Rt: right, BA: Brodmann's area.

Talairach coordinate (mm).

Connectivity analysis within the DMN was completed using the PCC as the seed (Fig. 4 and Table 4). In Parkinson's disease without fatigue, the PCC seed showed positive correlations with bilateral occipital regions, angular gyrus and right post‐central gyrus, while an inverse, negative correlation was observed only with the right insular region (Fig. 4A). In Parkinson's disease patients with fatigue, several of these functional relationships with PCC again did not reach a significant level (i.e., somatosensory) with new positive correlations with left middle frontal, superior orbitofrontal gyri. Several negative correlations were instead observed between PCC and bilateral insular, anterior cingulate (ACC) (both part of the SN), middle/inferior frontal gyri (part of the executive network) and bilateral temporal areas (Fig. 4B). Comparing of two patient groups, right somatosensory area showed less metabolic correlation with right PCC in PD with fatigue than without fatigue (P = 0.001, k = 85).

Figure 4.

Figure 4

Open in a new tab

Brain regions where significant functional correlations with the PCC seed were overlaid on “ch256” brain template (upper row) and its topographical connectivity map (bottom row) were shown in Parkinson's disease patient without fatigue (A) and Parkinson's disease patient with fatigue (B). Seed area marked on yellow, red node/path indicates positive connectivity and blue node/path indicate negative connectivity.

Table 4.

Brain regions that showed significant functional correlate with PCC seed in each Parkinson's disease group (P < 0.005 uncorrected for multiple comparisons with 100 voxels of spatial extended threshold)

Region BA Coordinatea T‐value Cluster size
X Y Z
PD‐Fatigue−
<Positive>
Rt Superior Occipital Gyrus BA 19 30 −87 30 5.74 485
Rt Precuneus 14 −67 60 4.98 124
Lt Precuneus −5 −69 51 3.59
Rt Postcentral Gyrus 18 −29 77 4.60 135
Rt Angular Gyrus BA 7 40 −72 45 4.58 114
Lt Angular Gyrus BA 7 −32 −75 44 4.52 110
Lt Superior Occipital Gyrus BA 19 −18 −89 32 3.99 204
<Negative>
Rt Insular BA 13 41 0 −13 3.61 210
PD‐Fatigue+
<Positive>
Lt Middle Frontal Gyrus BA 6 −26 8 47 6.17 201
Rt Superior Orbitofrontal Gyrus BA 11 22 44 −16 5.61 179
Rt Middle Occipital Gyrus BA 18 14 −87 14 3.94 144
Rt Superior Occipital Gyrus BA 19 24 −86 34 3.88 171
<Negative>
Lt Middle Temporal Gyrus BA 21 −50 2 −34 5.80 172
Lt Insular BA 13 −32 19 −1 5.08 1387
Lt Superior Temporal Gyrus BA 22 −51 4 −2 4.58
Lt Inferior Frontal Gyrus BA 47 −53 27 0 4.28
Lt Inferior Frontal Gyrus BA 44 −58 6 16 3.76
Lt Inferior Frontal Gyrus BA 45 −50 20 6 3.24
Rt Insular BA 13 43 7 −5 4.83 464
Rt Superior Temporal Gyrus BA 38 50 4 −20 4.45
Lt Thalamus −4 −18 −3 4.42 193
Bi Anterior Cingulate BA 25 0 19 −8 4.05 105
Open in a new tab

Lt: left, Rt: right, Bi: bilateral, BA: Brodmann's area.

aTalairach coordinate (mm).

DISCUSSION

The results of this study provide evidence of significant cortical metabolic changes in Parkinson's disease patients with fatigue mainly in cerebral regions associated with the salience and default networks. In particular, within the SN, we observed a lower regional glucose metabolism in the right insular which was associated with higher level of fatigue. Further, the connectivity analysis produced evidence of a significant disruption of the functional correlational pattern between this region and different subcortical/cortical areas, prevalently involving prefrontal and somatosensory regions.

The insular cortex is a limbic region well known for processing relevant information related to the state of the body [Williamson et al., 2003] linking the DMN to the CEN in guiding cognitive/emotional behavior [Goulden et al., 2014; Sridharan et al., 2008]. This brain region plays a prominent role in parkinsonian disorders [Criaud et al., 2016; Gomez‐Esteban et al., 2011]. As a multi‐faceted brain structure, it is also involved in various other functions [Augustine, 1996], such as receiving important autonomic feedback [Chaudhuri et al., 2001] and being activated during tasks involving interoception (how the body ‘feels’) and subjective awareness of both positive and negative feelings [Craig, 2002]. Recent studies using large‐scale brain connectivity analyses have highlighted the role of the insular cortex in detecting salient events and enabling switching between the DMN and CEN, corroborating the hypothesis that this region operates as an “outflow hub” which coordinates other large‐scale networks [Menon and Uddin, 2010]. With this in mind, the impaired connectivity between the insular and prefrontal and somatosensory regions detected by our analysis may contribute to the development and maintenance of central fatigue possibly through a disrupted flow of critical interoceptive information, emotional and cognitive processes also described below.

Other imaging studies have suggested an impairment of both dopaminergic and serotonergic function contributing to the symptom of fatigue in Parkinson's disease [Pavese et al., 2010]. Further, according to Braak's staging hypothesis of Parkinson's disease [Braak et al., 2006], the deposition of alpha‐synuclein throughout the insular could also play an important role in alteration of synaptic activity and receptor functions in this region. The relevance of the insular to fatigue has also been reported in other diseases, e.g., stroke [Manes et al., 1999], as resulting from damage of this region with disconnection between the insular and the frontal lobe and cingulate cortex.

Within the DMN, fatigue level was associated instead with greater cerebral glucose metabolism at the level of PCC with a disruption of normal connectivity and increased number of inverse, negative correlations with several cortical regions included within the salience and executive networks. The PCC within DMN is a highly connected and metabolically active brain region. Studies suggest that this complicated area through its anatomical connections plays an important role in cognitive and mental processing [Goldman and Holden, 2014; Leech and Sharp, 2014; Lou et al., 2004; Schilbach et al., 2012]. In particular, Leech and Sharp [2014] proposed that activity within the PCC is relatively high in the broad attentional, unfocused mental state allowing rapid transitions between different neural/cognitive states and becoming deactivated during specific attentional tasks. They suggested that the activation of the SN (particularly the right anterior insular cortex) is critically important in the control of task‐dependent deactivation of the PCC in response to unexpected events, and together form an anti‐correlated system regulating the attentional/mental focus. Thus, the impaired activation observed in our study of the insular region (either due to mesolimbic dopaminergic degeneration and/or alpha‐synuclein deposition), may lead to uncontrolled, high metabolic activity of the PCC that can be responsible for a persistent broad, unfocused mental state resulting in distracting, internally‐focused information contributing to mental fatigue. Both of these brain regions that showed functional differences correlated significantly with fatigue level.

In the present study, we found no prefrontal metabolic change in Parkinson's disease with fatigue. The DLPFC and orbitofrontal cortex may be involved in the self‐evaluation of mental fatigue, in fact both these brain regions may show fatigue related activation during performance of cognitively demanding task [Klaassen et al., 2016]. Patients with chronic fatigue syndrome have reduced gray‐matter volume in the bilateral prefrontal cortex [Okada et al., 2004], and a decrease in delta band power of MEG signal in the DLPFC is positively related to the daily level of fatigue sensation [Ishii et al., 2014]. Subjective sensation of fatigue has also been found to be associated with the medial orbitofrontal cortex activity [Tajima et al., 2010]. The lack of changes in regional glucose metabolism in our study does not necessarily indicate a lack of role of these prefrontal regions in Parkinson's disease patients with fatigue. The reduced connectivity from the insula to the prefrontal brain region in fatigued patients may indeed contribute to the disconnection from the SN core to CEN causing higher top‐down mental effort in the prefrontal cortex.

As already noted above, we cannot neglect the importance of the dopamine and serotonergic system. While serotonergic changes may be one of the pathophysiological contributors to chronic fatigue syndrome [Yamamoto et al., 2004], more recent study have shown no association between raphe serotonin transporter availability and fatigue in early PD patient [Qamhawi et al., 2015]. The potential role of dopaminergic‐related medications for the treatment of fatigue symptom have also been reported [Lou et al., 2009; Mendonça et al., 2007] nevertheless the evidence of their effect remain controversial [Franssen et al., 2014].

To avoid potential confounding factors influencing fatigue, we controlled for demographic and clinical profile, including age, gender, disease severity, mood, parkinsonian disability, medication as well as individual sleep quality. However, the small sample size still remains as a limitation of this study, and further investigations are needed to validate current findings. Less conservative statistical inferences can also be another factor limiting the generalization of our results. However, these can be used, when there are specific predictions to be tested. Another potential limitation of our study is the use of self‐reporting questionnaires for measuring individual fatigue, however, reliability and validity of FSS has been reported consistently in other diseases [Krupp et al., 1989] as well as in PD patients [Hagell et al., 2006]. In addition, mental fatigue in particular, may rely on a more personal perception of self‐status, difficult to measure with more objective scales.

In summary, fatigue in Parkinson's disease is a state of profound physical and/or mental weakness, associated with a feeling of abnormal and overwhelming lack of energy which can be associated with metabolic changes and impaired functional connectivity between brain regions linked to neural systems within the SN and other important cognitive networks such as DMN and CEN. This underappreciated clinical problem represents one of the major complaints of patients in the daily clinical practice [Friedman et al., 2016]. The current study provides important insights into the complicated neural mechanisms underlying this non‐motor symptom and opens the door for new research studies exploring the interactions between receptor abnormalities and cognitive networks and their causal role in the development of fatigue in Parkinson's disease.

ACKNOWLEDGMENT

Antonio Strafella is supported Canada Research Chair Program. Sang Soo Cho was supported by Parkinson Society Canada (Basic research fellowship).

Disclosure: Dr. Cho report no conflicts of interest related to this study that might bias this work. Ms. Aminian report no conflicts of interest related to this study that might bias this work. Ms. Li report no conflicts of interest related to this study that might bias this work. Dr. Lang report no conflicts of interest related to this study that might bias this work. Dr. Houle report no conflicts of interest related to this study that might bias this work. Dr. Strafella report no conflicts of interest related to this study that might bias this work.

S.S.C. and A.P.S. conceived and designed the study. S.S.C., K.A., and C.L. contributed to data collection and analysis of imaging data. S.S.C., A.E.L., S.H., and A.P.S. contributed to writing and drafting the manuscript.

Corrections added on 07 September 2016, after online publication.

REFERENCES

  1. Alves G, Wentzel‐Larsen T, Larsen JP (2004): Is fatigue an independent and persistent symptom in patients with Parkinson disease? Neurology 63:1908–1911. [DOI] [PubMed] [Google Scholar]
  2. Augustine JR (1996): Circuitry and functional aspects of the insular lobe in primates including humans. Brain Res Brain Res Rev 22:229–244. [DOI] [PubMed] [Google Scholar]
  3. Braak H, Bohl JR, Muller CM, Rub U, de Vos RA, Del Tredici K (2006): Stanley Fahn Lecture 2005: The staging procedure for the inclusion body pathology associated with sporadic Parkinson's disease reconsidered. Mov Disord 21:2042–2051. [DOI] [PubMed] [Google Scholar]
  4. Brown RG, Dittner A, Findley L, Wessely SC (2005): The Parkinson fatigue scale. Parkinsonism Relat Disord 11:49–55. [DOI] [PubMed] [Google Scholar]
  5. Chaudhuri KR, Pal S, Bridgman K, Trenkwalder C (2001): Achieving 24‐hour control of Parkinson's disease symptoms: Use of objective measures to improve nocturnal disability. Eur Neurol 46:3–10. [DOI] [PubMed] [Google Scholar]
  6. Chaudhuri KR, Pal S, DiMarco A, Whately‐Smith C, Bridgman K, Mathew R, Pezzela FR, Forbes A, Hogl B, Trenkwalder C (2002): The Parkinson's disease sleep scale: A new instrument for assessing sleep and nocturnal disability in Parkinson's disease. J Neurol Neurosurg Psychiatry 73:629–635. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cook DB, O'Connor PJ, Lange G, Steffener J (2007): Functional neuroimaging correlates of mental fatigue induced by cognition among chronic fatigue syndrome patients and controls. Neuroimage 36:108–122. [DOI] [PubMed] [Google Scholar]
  8. Craig AD (2002): How do you feel? Interoception: The sense of the physiological condition of the body. Nat Rev Neurosci 3:655–666. [DOI] [PubMed] [Google Scholar]
  9. Criaud M, Christopher L, Boulinguez P, Ballanger B, Lang AE, Cho SS, Houle S, Strafella AP (2016): Contribution of insula in Parkinson's disease: A quantitative meta‐analysis study. Hum Brain Mapp 37:1375–1392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Defer GL, Widner H, Marie RM, Remy P, Levivier M (1999): Core assessment program for surgical interventional therapies in Parkinson's disease (CAPSIT‐PD). Mov Disord 14:572–584. [DOI] [PubMed] [Google Scholar]
  11. Franssen M, Winward C, Collett J, Wade D, Dawes H (2014): Interventions for fatigue in Parkinson's disease: A systematic review and meta‐analysis. Movement Disorders 29:1675–1678. [DOI] [PubMed] [Google Scholar]
  12. Friedman JH, Brown RG, Comella C, Garber CE, Krupp LB, Lou JS, Marsh L, Nail L, Shulman L, Taylor CB (2007): Fatigue in Parkinson's disease: A review. Mov Disord 22:297–308. [DOI] [PubMed] [Google Scholar]
  13. Friedman JH, Abrantes A, Sweet LH (2011): Fatigue in Parkinson's disease. Expert Opin Pharmacother 12:1999–2007. [DOI] [PubMed] [Google Scholar]
  14. Friedman JH, Beck JC, Chou KL, Clark G, Fagundes CP, Goetz CG, Herlofson K, Kluger B, Krupp LB, Lang AE, Lou JS, Marsh L, Newbould A, Weintraub D (2016): Fatigue in Parkinson's disease: Report from a mutidisciplinary symposium. Npj Parkinson's Disease 2:15025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Goldman JG, Holden S (2014): Treatment of psychosis and dementia in Parkinson's disease. Curr Treat Options Neurol 16:281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Goldman JG, Stebbins GT, Leung V, Tilley BC, Goetz CG (2014): Relationships among cognitive impairment, sleep, and fatigue in Parkinson's disease using the MDS‐UPDRS. Parkinsonism Relat Disord 20:1135–1139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gomez‐Esteban JC, Tijero B, Somme J, Ciordia R, Berganzo K, Rouco I, Bustos JL, Valle MA, Lezcano E, Zarranz JJ (2011): Impact of psychiatric symptoms and sleep disorders on the quality of life of patients with Parkinson's disease. J Neurol 258:494–499. [DOI] [PubMed] [Google Scholar]
  18. Goulden N, Khusnulina A, Davis NJ, Bracewell RM, Bokde AL, McNulty JP, Mullins PG (2014): The salience network is responsible for switching between the default mode network and the central executive network: Replication from DCM. Neuroimage 99:180–190. [DOI] [PubMed] [Google Scholar]
  19. Hagell P, Hoglund A, Reimer J, Eriksson B, Knutsson I, Widner H, Cella D (2006): Measuring fatigue in Parkinson's disease: A psychometric study of two brief generic fatigue questionnaires. J Pain Symptom Manage 32:420–432. [DOI] [PubMed] [Google Scholar]
  20. Honig H, Antonini A, Martinez‐Martin P, Forgacs I, Faye GC, Fox T, Fox K, Mancini F, Canesi M, Odin P, Chaudhuri KR (2009): Intrajejunal levodopa infusion in Parkinson's disease: A pilot multicenter study of effects on nonmotor symptoms and quality of life. Mov Disord 24:1468–1474. [DOI] [PubMed] [Google Scholar]
  21. Ishii A, Tanaka M, Watanabe Y (2014): The neural substrates of self‐evaluation of mental fatigue: A magnetoencephalography study. PLoS One 9:e95763. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Klaassen EB, Plukaard S, Evers EA, de Groot RH, Backes WH, Veltman DJ, Jolles J (2016): Young and middle‐aged schoolteachers differ in the neural correlates of memory encoding and cognitive fatigue: A functional MRI study. Front Hum Neurosci 10:148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Krupp LB, LaRocca NG, Muir‐Nash J, Steinberg AD (1989): The fatigue severity scale: Application to patients with multiple sclerosis and systemic lupus erythematosus. Arch Neurol 46:1121–1123. [DOI] [PubMed] [Google Scholar]
  24. Kumar S, Bhatia M, Behari M (2003): Excessive daytime sleepiness in Parkinson's disease as assessed by Epworth Sleepiness Scale (ESS). Sleep Med 4:339–342. [DOI] [PubMed] [Google Scholar]
  25. Leech R, Sharp DJ (2014): The role of the posterior cingulate cortex in cognition and disease. Brain 137:12–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lou HC, Luber B, Crupain M, Keenan JP, Nowak M, Kjaer TW, Sackeim HA, Lisanby SH (2004): Parietal cortex and representation of the mental Self. Proc Natl Acad Sci U S A 101:6827–6832. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lou JS, Dimitrova DM, Park BS, Johnson SC, Eaton R, Arnold G, Nutt JG (2009): Using modafinil to treat fatigue in Parkinson disease: A double‐blind, placebo‐controlled pilot study. Clin Neuropharmacol 32:305–310. [DOI] [PubMed] [Google Scholar]
  28. Maldjian JA, Laurienti PJ, Kraft RA, Burdette JH (2003): An automated method for neuroanatomic and cytoarchitectonic atlas‐based interrogation of fMRI data sets. Neuroimage 19:1233–1239. [DOI] [PubMed] [Google Scholar]
  29. Manes F, Paradiso S, Robinson RG (1999): Neuropsychiatric effects of insular stroke. J Nerv Ment Dis 187:707–712. [DOI] [PubMed] [Google Scholar]
  30. McDonald WM, Richard IH, DeLong MR (2003): Prevalence, etiology, and treatment of depression in Parkinson's disease. Biol Psychiatry 54:363–375. [DOI] [PubMed] [Google Scholar]
  31. Mendonça DA, Menezes K, Jog MS (2007): Methylphenidate improves fatigue scores in Parkinson disease: A randomized controlled trial. Movement Disord 22:2070–2076. [DOI] [PubMed] [Google Scholar]
  32. Menon V, Uddin LQ (2010): Saliency, switching, attention and control: A network model of insula function. Brain Struct Funct 214:655–667. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Okada T, Tanaka M, Kuratsune H, Watanabe Y, Sadato N (2004): Mechanisms underlying fatigue: A voxel‐based morphometric study of chronic fatigue syndrome. BMC Neurol 4:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Otto T, Zijlstra FR, Goebel R (2014): Neural correlates of mental effort evaluation–involvement of structures related to self‐awareness. Soc Cogn Affect Neurosci 9:307–315. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Pavese N, Metta V, Bose SK, Chaudhuri KR, Brooks DJ (2010): Fatigue in Parkinson's disease is linked to striatal and limbic serotonergic dysfunction. Brain 133:3434–3443. [DOI] [PubMed] [Google Scholar]
  36. Qamhawi Z, Towey D, Shah B, Pagano G, Seibyl J, Marek K, Borghammer P, Brooks DJ, Pavese N (2015): Clinical correlates of raphe serotonergic dysfunction in early Parkinson's disease. Brain 138:2964–2973. [DOI] [PubMed] [Google Scholar]
  37. Schilbach L, Bzdok D, Timmermans B, Fox PT, Laird AR, Vogeley K, Eickhoff SB (2012): Introspective minds: Using ALE meta‐analyses to study commonalities in the neural correlates of emotional processing, social & unconstrained cognition. PLoS One 7:e30920. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Sridharan D, Levitin DJ, Menon V (2008): A critical role for the right fronto‐insular cortex in switching between central‐executive and default‐mode networks. Proc Natl Acad Sci U S A 105:12569–12574. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Tajima S, Yamamoto S, Tanaka M, Kataoka Y, Iwase M, Yoshikawa E, Okada H, Onoe H, Tsukada H, Kuratsune H, Ouchi Y, Watanabe Y (2010): Medial orbitofrontal cortex is associated with fatigue sensation. Neurol Res Int 2010:671421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Tomlinson CL, Stowe R, Patel S, Rick C, Gray R, Clarke CE (2010): Systematic review of levodopa dose equivalency reporting in Parkinson's disease. Mov Disord 25:2649–2653. [DOI] [PubMed] [Google Scholar]
  41. van der Linden D, Frese M, Meijman TF (2003a): Mental fatigue and the control of cognitive processes: Effects on perseveration and planning. Acta Psychol (Amst) 113:45–65. [DOI] [PubMed] [Google Scholar]
  42. van der Linden D, Frese M, Sonnentag S (2003b): The impact of mental fatigue on exploration in a complex computer task: Rigidity and loss of systematic strategies. Hum Factors 45:483–494. [DOI] [PubMed] [Google Scholar]
  43. Williamson JW, McColl R, Mathews D (2003): Evidence for central command activation of the human insular cortex during exercise. J Appl Physiol (1985) 94:1726–1734. [DOI] [PubMed] [Google Scholar]
  44. Yamamoto S, Ouchi Y, Onoe H, Yoshikawa E, Tsukada H, Takahashi H, Iwase M, Yamaguti K, Kuratsune H, Watanabe Y (2004): Reduction of serotonin transporters of patients with chronic fatigue syndrome. Neuroreport 15:2571–2574. [DOI] [PubMed] [Google Scholar]

Articles from Human Brain Mapping are provided here courtesy of Wiley

RESOURCES