Laura Anne Wortinger
Neurocognitive correlates of adolescent chronic fatigue syndrome
Dissertation for the Degree of Philosophiae Doctor
Oslo, 2017
Department of Paediatrics and Adolescent Health Akershus University Hospital
&
Institute for Clinical Medicine Faculty of Medicine
University of Oslo
© Laura Anne Wortinger, 2017
Series of dissertations submitted to the Faculty of Medicine, University of Oslo
ISBN 978-82-8377-053-7
All rights reserved. No part of this publication may be
reproduced or transmitted, in any form or by any means, without permission.
Cover: Hanne Baadsgaard Utigard.
Print production: Reprosentralen, University of Oslo.
ii
ACKNOWLEDGEMENTS 1
LIST OF ARTICLES 3
ABBREVIATIONS 4
SUMMARY 5
INTRODUCTION 7
Fatigue 7
Adolescent Chronic Fatigue Syndrome (CFS) 7
Pain 9
Depression 9
Development of adolescent CFS 10
Sustained arousal model 11
Cognition in adolescent CFS 11
Emotion and cognition 12
Pain and cognition 12
CFS and the brain 13
Brain morphology in CFS 13
Brain physiology in CFS 14
Therapies for adolescent CFS 16
NorCAPITAL project 17
AIMS OF THE DISSERTATION 19
MATERIALS AND METHODS 21
Participants 21
Data Collection 24
Statistical analysis 25
Additional analyses specific to the PhD dissertation 26
SYNOPSIS OF STUDIES 27
Study 1 27
Study 2 28
Study 3 29
Additional results specific to the PhD dissertation 30
iii
DISCUSSION 33
Main findings 33
Neurocognitive network dynamics 33
Affective and Cognitive processing 36
Regional right Anterior Insula Connectivity 37
Pain processing 39
Physical activity 40
CFS brain connectivity 41
Sustained Arousal 42
CFS neural mechanisms 42
Specificity of neural changes 44
Future perspectives 46
Methodological issues 49
Participants 49
Functional magnetic resonance imaging 50
Emotional conflict processing 51
CONCLUSION 52
REFERENCES 53
ARTICLES 68
1
Acknowledgements
The work for this dissertation was conducted at the Department of Paediatrics, Oslo University Hospital; Intervention Center, Oslo University Hospital; Department of Paediatrics and Adolescent Health, Akershus University Hospital; Institute for Clinical Medicine, Faculty of Medicine, University of Oslo; and Department of Psychology, University of Oslo.
This dissertation would not have been possible without funding from the Research Council of Norway, South-Eastern Norway Regional Health Authority, and the University of Oslo.
I would like to thank the participants and their families for their contribution to these studies. I express gratitude to personnel for their time and effort invested in the data collection and all referring hospital units.
I would like to gratefully acknowledge the guidance, support, and encouragement of my doctoral supervisor, Dr. Vegard Bruun Wyller, and co-supervisors, Dr. Tor Endestad and Dr. Merete G. Øie, for their mentorship and collaboration.
My gratitude extends to Annika M. D. Melinder, Dag Sulheim, Even Fagermoen, and Andre Sevenius for their enthusiasm and willingness to provide both data and clinical expertise, to Chinh Bkrong Thi Thuy Nguyen, Maria Pedersen, Tarjei Tørre Asprusten, Linn Nilsen Rødevand, Julie Mangersnes, Kristine Jacobsen, Line Sletner, Jakob Seth Fiskå, Sadaf Malik, Betty Van Roy, and Stine Andersen for insightful discussions about adolescent chronic fatigue syndrome, and to Cecilie Sol Skaftnes at the Department of
2 Psychology for many hours of mutual support, complaining, and conversations in our office.
Finally, a heartfelt appreciation goes to the strongest women I know: my mom, sister, and daughter. They give unconditional love and support in every juncture of my life.
Gratefulness continues to my children, Natalie, Austin, and Max, who keep me motivated to be a better mom. Lastly, my boyfriend, Nils Inge, thank you for sharing my enthusiasm for research and inspiring me every day.
3
List of Articles
1 Wortinger, L. A., Endestad, T., Melinder, A. M. D., Øie, M. G., Sevenius, A., & Wyller, V. B. (2016). Aberrant resting-state functional connectivity in the salience
network of adolescent chronic fatigue syndrome. PLoS One, 11(7), e0159351.
2 Wortinger, L. A., Endestad, T., Melinder, A. M. D., Øie, M. G., Sulheim, D.,
Fagermoen, E., & Wyller, V. B. (2016). Emotional conflict processing in adolescent chronic fatigue syndrome: A pilot study using functional magnetic resonance imaging. Journal of Clinical and Experimental Neuropsychology, 1-14.
3 Wortinger, L. A., Øie, M. G., Endestad, T., & Wyller, V. B. (submitted). Altered right anterior insular connectivity and loss of associated functions in adolescent Chronic Fatigue Syndrome. Clinical Physiology and Functional Imaging.
4
Abbreviations
BOLD Blood oxygen level dependent
CBT Cognitive behavioral therapy
CEN Central executive network
CFS Chronic fatigue syndrome
dAI Dorsal anterior insula
DMN Default mode network
FC Functional connectivity
fMRI Functional magnetic resonance imaging
ICA Independent component analysis
ICN Intrinsic connectivity networks
lPFC Lateral prefrontal cortex
mpINS Midposterior insula
PPT Pressure pain threshold
PPC Posterior parietal cortex
SN Salience network
5
Summary
Disabling fatigue, cognitive complaints, pain sensitivity, and exertion intolerance are common among adolescent patients with chronic fatigue syndrome (CFS), yet these symptoms remain poorly understood at the neurobiological level. It has been proposed that neural alterations generate specific symptoms and identifying these changes might lead to improved treatment for patients. This dissertation presents three investigations with a neurocognitive perspective on adolescent CFS symptomology.
The first paper describes a functional magnetic resonance imaging (fMRI) study on neurocognitive brain-network dynamics of adolescent CFS patients and healthy
comparison participants. The study found salience network (SN) functional connectivity (FC) decreases to the right insula, which were related to fatigue severity in patients. The second paper reports on adolescent CFS patients and healthy participants who
underwent fMRI while performing an emotional conflict task. We found non-responsive linear relationships between brain regions and behavioral measures, which indicate impaired salience detection in patients. The third paper describes an investigation on the regional FC patterns of a central hub of the SN, the right dorsal anterior insula (dAI), in patients and healthy counterparts. Adolescent CFS patients demonstrated a decrease in FC to a central node important for the maturation of top-down cognitive control. The inferred top-down cognitive control deficiency in patients also lacked associations with known functions that require efficient cognitive control: cognition, pain regulation, and physical activity.
Our results suggest neurocognitive network dynamics in adolescent CFS are different from those found in healthy participants. These findings might have broader
6 implications on how the symptomology of CFS arises from a complex interplay of brain networks in adolescent patients.
7
Introduction
Fatigue
Fatigue is an enigmatic feeling from the body. The exact function and pathophysiology of fatigue are unclear despite being a leading clinical complaint (Kroenke, Stump, Clark, Callahan, & McDonald, 1999) that can continue long after serious medical conditions are resolved (Dantzer, Heijnen, Kavelaars, Laye, & Capuron, 2014). Acute fatigue is normally relieved with rest, while persistent fatigue is unrefreshed with sleep.
Fatigue perception might be defined by its subjective sensory qualities that motivate behavior to refrain from activity to avoid tissue-damaging processes, similar to pain (Merskey & Bogduk, 1994). Fatigue might provide a teaching signal that enables
individuals to avoid future harm and gives rise to a negative reinforcement. Additionally, fatigue might be viewed as a primordial drive, like thirst, hunger, and temperature, which is generated by specific sensory pathways (Craig, 2003b) in order to maintain homeostasis in the body for the purpose of survival (Craig, 2011).
Adolescent Chronic Fatigue Syndrome (CFS)
CFS is characterized by disabling and abiding fatigue accompanied by cognitive dysfunction, sleep abnormalities, autonomic manifestations, and pain, wherein symptoms worsen with mental, emotional, and physical exertion (IOM, 2015). The incidence of CFS has two distinct age peaks, at 10–19 years and 30–39 years, and an incidence rate ratio of about 7 for both groups (Bakken et al., 2014). It is unclear
whether the same vulnerability for the development of CFS is present during these two age peaks, or if different etiological factors might explain this pattern onset. Even though
8 the cause of CFS is unknown, differences on how CFS presents in adolescence and
adulthood have emerged (Collin, Nuevo, van de Putte, Nijhof, & Crawley, 2015). Whereas adults have longer and greater disability and fatigue, adolescents have more headaches and fewer symptoms of tender lymph nodes, palpitations, dizziness, general malaise and pain (Collin et al., 2015). Additionally, adolescents have greater comorbid depression, but not anxiety compared to adults (Collin et al., 2015). The prognosis for improvement of symptoms or making a full recovery in adolescent CFS is between 54–94% (Joyce, Hotopf, & Wessely, 1997; S. L. Nijhof, Bleijenberg, Uiterwaal, Kimpen, & van de Putte, 2012), while adult improvement is not more than 22% (White et al., 2011) and full recovery from untreated CFS is rare (Cairns & Hotopf, 2005).
The interference fatigue has on the quality of life in adolescent CFS patients might be related to the additive effects of comorbid symptoms. Health related quality of life for adolescent CFS was reported as being worse than adolescents who received renal transplantation (with a median age since transplantation at 4.9 years) and children in remission after acute lymphoblastic leukemia (Winger et al., 2015). A registry linkage study found children with CFS were more often registered with a diagnosis of
depressive disorder, migraine, muscle pain, and infections than children with type 1 diabetes mellitus (Bakken et al., 2016). In a population birth cohort study, adolescent patients with exclusive CFS reported a higher category of pain-related daily interference than adolescents with exclusive chronic widespread pain (Norris, Deere, Tobias, &
Crawley, 2016). Additionally, higher levels of depressive symptoms were specific to the exclusive CFS group (Norris et al., 2016).
9 Pain
The relationship between pain symptoms and neural function might provide useful information on CFS pathology. Hyperalgesia is a prominent feature of adolescent CFS (Bakken et al., 2016; S. L. Nijhof, Priesterbach, Bleijenberg, Engelbert, & van de Putte, 2013; Norris et al., 2016; Winger et al., 2014) and increased pain ratings and lower pain thresholds are commonly observed (S. L. Nijhof, Priesterbach, Bleijenberg, et al., 2013;
Winger et al., 2014). In an adolescent CFS treatment study, pain diminished after successful treatment, even though treatment was not targeted for pain symptoms (S. L.
Nijhof, Priesterbach, Bleijenberg, et al., 2013). In those patients who did not recover at follow-up, they still reported more pain symptoms and lower pain thresholds, which might be an indication of the intrinsic feature of pain in adolescent CFS (S. L. Nijhof, Priesterbach, Bleijenberg, et al., 2013). Depressive symptoms, on the other hand, influenced neither pain thresholds nor pain ratings throughout the study (S. L. Nijhof, Priesterbach, Bleijenberg, et al., 2013).
Depression
The degree to which depressive symptoms influence neurocognitive function is unknown in adolescent CFS. A higher rate of comorbidity with depression disorder is more common in adolescent CFS than in healthy controls or chronic illness control groups, like cystic fibrosis, chronic widespread pain, and type 1 diabetes mellitus
(Bakken et al., 2016; Lievesley, Rimes, & Chalder, 2014; Loades, Sheils, & Crawley, 2016;
Norris et al., 2016; Walford, Nelson, & McCluskey, 1993). Not only are depression and depressive symptoms more common in adolescent CFS compared to other chronic conditions, they are more common in adolescent CFS compared to adult CFS (Collin et al., 2015). To meet a diagnosis of CFS, persistent fatigue cannot be explained by other
10 conditions, including depression, but a diagnosis can be made in the presence of
comorbid depression, which does not fully explain the disabling fatigue (Reeves et al., 2005). It is estimated that 30% of children with CFS have symptoms that reveal a
depression diagnosis (Bould, Collin, Lewis, Rimes, & Crawley, 2013; Knight et al., 2014).
Since those with significant depression tend to be excluded from treatment studies, the relationship between depression and depressive symptoms and adolescent CFS is unclear, which has a grave impact on effective treatment for depression in CFS and the influence of depression on treatment success (Loades et al., 2016).
Development of adolescent CFS
The pathophysiology of CFS remains elusive. The lack of biomarkers and understanding of the disease hinders the development of targeted treatments. Recent findings report enhanced sympathetic nervous activity, pain intolerance, low-grade systemic
inflammation, attenuated hypothalamus-pituitary adrenal axis function, altered neuroendocrine control, cognitive impairment, genetic vulnerabilities, and large physical activity reductions (Hall et al., 2016; Meyer et al., 2015; Sulheim et al., 2015;
Sulheim et al., 2014; Winger et al., 2014; Wyller et al., 2016) that seem to indicate a pattern of stress responses across bodily systems in adolescent CFS. A multisystem perspective might be important in understanding CFS pathology.
The sustained arousal model of disease mechanisms applies a multisystem approach to CFS and considers predisposing, precipitating, and perpetuating risk factors that
contribute to aberrant stress responses (Wyller, Eriksen, & Malterud, 2009). Sustained arousal might be driven by abnormalities in brain function, but how they relate to cognition, symptomology, and maintenance of adolescent CFS is unclear.
11 Sustained arousal model
Figure 1: The sustained arousal model of disease mechanisms in adolescent CFS, modified from Wyller et al. (2009).
Cognition in adolescent CFS
Neurocognitive studies point to specific deviations in cognitive control that might play an important role in adolescent CFS. Deficits in inhibition, working memory, verbal learning (Sulheim et al., 2015), interference control (van de Putte et al., 2008), and attention (Haig-Ferguson, Tucker, Eaton, Hunt, & Crawley, 2009; Kawatani et al., 2011) have been observed in adolescent patients with CFS. An event-related potential (ERP) study found that working memory impairment was associated with frontal lobe alterations (Tomoda et al., 2007). An fMRI study found increasing poor task
performance in patients resulted in a less efficient use of neural resources in frontal regions, where greater mental effort afforded costly energy requirements (Mizuno et al.,
12 2015). Data suggests that adolescent CFS is accompanied by a decline in general
cognitive functioning (L. N. Nijhof et al., 2016).
Emotion and cognition
Because depressive symptoms are common in adolescent CFS, an emotional cognitive study might shed light on brain alterations and give an indication on how central emotional attention abnormalities are in CFS. In cognitive tasks where there is an emotional element, a picture of a fearful face or a threatening word (e.g. RAPE),
attention to the emotional stimuli is known to alter cognitive processing. In depression, abnormal deployment of attention to negative emotional stimuli is considered a central mechanism of the disease (Browning, Holmes, & Harmer, 2010; Gotlib, Krasnoperova, Yue, & Joormann, 2004). It is also understood to underlie clinical observations, where patients interpret, focus, and remember information in a more negative manner than healthy individuals (Mathews & MacLeod, 2005). This negative attentional bias in depression is associated with a distinct pattern of abnormal brain activations, increased amygdalae activations representing an enhanced detection of saliency and biological relevance (Adolphs, 2008; Adolphs, Tranel, Damasio, & Damasio, 1995; Davis & Whalen, 2001) and decreased anterior cingulate cortex (ACC) and lateral prefrontal cortex (lPFC) activations representing a deficit in top-down cognitive control (Bishop, Duncan, Brett,
& Lawrence, 2004; MacDonald, Cohen, Stenger, & Carter, 2000).
Pain and cognition
The intrinsic feature of pain in adolescent CFS might be related to cognitive
performance. The experience of pain constitutes a complex interplay of many factors, including cognitive functions (Apkarian, Baliki, & Geha, 2009). The ability to tolerate
13 pain is commonly linked to performance on neuropsychological tests, as observed in healthy participants (Oosterman, Dijkerman, Kessels, & Scherder, 2010), patients with chronic nonmalignant pain syndromes (Landro et al., 2013), and chronic pain
development after surgery (Attal et al., 2014). In an fMRI study on chronic back pain patients, when spontaneous pain was sustained at a higher level, the medial prefrontal cortex (mPFC) was increasingly active and negatively correlated with lPFC activity, indicating decreasing top-down cognitive control. Decreased gray matter density in the lPFC was also observed in these patients, and together, the results suggest a tight
interplay between brain activity, neuronal death, and cognitive abnormalities in chronic back pain (Apkarian et al., 2009).
CFS and the brain
Brain morphology in CFS
Structural MRI studies on adults with CFS have shown a reduction in global gray-matter volume, which correlated with decreases in physical activity (de Lange et al., 2005).
Another study found reduced gray-matter volume in bilateral prefrontal cortices, where a region in the right lPFC decreased with increasing fatigue severity in CFS patients (Okada, Tanaka, Kuratsune, Watanabe, & Sadato, 2004). Reductions in gray-matter volume were reported, as well, in occipital lobes, right posterior parietal cortex, and left parahippocampal gyrus (Puri et al., 2012). Even though global gray-matter volume did not differ between groups, this study found more pain was associated with reductions in gray-matter volume in the left lPFC, and no associations were found for fatigue,
depressive symptoms, physical activity, and psychomotor speed (van der Schaaf et al., 2016). Finally, a cognitive behavioral therapy intervention study found CFS patients had lower global gray-matter volume than healthy participants at baseline, and after CBT
14 intervention, CFS patients not only improved in health status, physical activity and cognitive performance, but showed a significant increase in gray-matter volume in the lPFC (de Lange et al., 2008).
In a white-matter study, global white matter volume reductions were observed in CFS patients (Zeineh et al., 2015). Greater diffusion restrictions in white-matter fibers were observed in the right inferior longitudinal fasciculus (ILF) and the right arcuate
fasciculus in patients with CFS, wherein the right anterior arcuate diffusion restrictions correlated positively with disease severity. Right hemispheric alterations in white- matter fibers could suggest degeneration of crossing fibers or strengthening of short- range fibers, and authors suggest the right anterior arcuate diffusion properties may serve as a biomarker for CFS (Zeineh et al., 2015). A longitudinal CFS study found decreased white-matter volume in the inferior fronto-occipital fasciculus (IFOF), which suggests an abnormal rate of deterioration of the IFOF in patients (Shan et al., 2016).
Anatomical decreases in both gray and white matter are apparent in adult CFS, and whether these global and regional reductions in brain structure carry external validity for adolescent CFS remains to be seen.
Brain physiology in CFS
Functional MRI studies on adults with CFS have shown differences in brain activations compared to healthy participants that suggest greater effort and less efficient use of neural resources during cognitive tasks. Verbal working memory performance in CFS patients was associated with greater activations across regions commonly recruited for this task (Lange et al., 2005). Another verbal working memory study found greater activations in mPFC/ACC and temporal cortex, but decreased activations in in
15 dorsolateral PFC and parietal cortex (Caseras et al., 2006). During fatiguing tasks,
increased activation in the occipito-parietal cortex, posterior cingulate gyrus and parahippocampal gyrus, and decreased activation in dorsolateral and dorsomedial prefrontal cortices were observed in patients (Caseras et al., 2008). Additionally, the cerebellar, temporal, cingulate and frontal regions were increased in CFS patients, but the posterior parietal cortex was decreased compared to healthy participants (Cook, O'Connor, Lange, & Steffener, 2007).
The differential increases and decreases across brain regions in adult CFS led
researchers to look at subcortical structures to see if basic functions, like motivation, might be influencing cognition. In fact, adults with CFS exhibited significantly decreased activations in the basal ganglia during a monetary gambling task, which correlated with increased mental fatigue, general fatigue, and reduced physical activity (Miller et al., 2014). Likewise, a childhood CFS study found lower activity in the basal ganglia, which negatively correlated with severity of fatigue in patients (Mizuno et al., 2016).
Consistent decreases in regions of the basal ganglia were found in both adult and childhood CFS studies and suggest altered underlying dopaminergic function, which is closely related to reward and motivation. Investigations on the reward neural system might provide a better indication of neural dysfunction in CFS.
Resting-state fMRI investigations have not been undertaken previously and might provide useful information on neurocognitive network dynamics in adolescent CFS.
Changes in brain blood flow create a blood-oxygen-level dependent (BOLD) signal that provides a method to estimate functional connectivity (FC) between brain regions that share functional properties or networks. Resting-state network FC is understood to play a major role in shaping goal-directed processes, with a smaller amount of neural
16 variance contributed to the actual cognitive task (Cole, Bassett, Power, Braver, &
Petersen, 2014; Williams, 2016).
Prominent features of several major psychiatric and neurological disorders are often found to be related to an aberration of three core intrinsic connectivity networks (ICN) of the brain (Menon, 2011). The three ICN are the default mode network (DMN), central executive network (CEN), and salience network (SN). Previous resting-state FC studies have found SN abnormalities in adult CFS (Boissoneault, Letzen, Lai, O'Shea, et al., 2016;
Boissoneault, Letzen, Lai, Robinson, & Staud, 2016; Gay et al., 2015); particularly, regional FC decreases observed in the right insula (Boissoneault, Letzen, Lai, O'Shea, et al., 2016; Boissoneault, Letzen, Lai, Robinson, et al., 2016). Alterations of the DMN and CEN have been reported in adult CFS, as well (Boissoneault, Letzen, Lai, O'Shea, et al., 2016; Boissoneault, Letzen, Lai, Robinson, et al., 2016; Gay et al., 2015; Kim et al., 2015).
Therapies for adolescent CFS
For 48% of children with CFS, they waited one year or longer to receive a diagnosis; this long time span is an indication that treatment for childhood CFS is less than optimal (Bakken et al., 2016). Even though CFS etiology is unknown, it is known that adolescence is a critical age for cognitive development and a vulnerable period for the onset of
disabling mental diseases, wherein if left untreated a mental disorder can lead to a more severe, more difficult to treat illness, and to the development of co-occurring illnesses (Merikangas et al., 2010). A timely diagnosis followed by appropriate treatment is necessary for adolescent patients. However, no pharmacological intervention exists, and the only effective forms of treatment are cognitive behavioral therapy (S. L. Nijhof et al., 2012; S. L. Nijhof, Priesterbach, Bleijenberg, et al., 2013; S. L. Nijhof, Priesterbach,
17 Uiterwaal, et al., 2013) and graded exercise therapy (Gordon, Knapman, & Lubitz, 2010).
CFS during adolescence might represent an age group where it is easier to identify real disease mechanisms, as opposed to secondary phenomena associated with years of chronicity in adults, and a better understanding of these complex mechanisms might lead to the development of targeted treatments.
NorCAPITAL project
This dissertation is from data collected in the NorCAPITAL-project (The Norwegian Study of Chronic Fatigue Syndrome in Adolescents: Pathophysiology and Intervention Trial; Clinical Trials ID: NCT01040429), which explores disease mechanisms, low-dose clonidine treatment effects, and patients’ experiences in adolescent CFS. NorCAPITAL rests upon the sustained arousal model of CFS disease mechanisms (Wyller et al., 2009), and only data from the cross-sectional part of the project were used in these studies.
CFS patients included (n=120)
Randomized 1:1 ra.o
8 weeks of treatment
Healthy controls (n=39)
Clonidine intervention
(n=60) Placebo intervention
(n=60)
Follow-up clonidine group, week 8, during treatment (n=55)
Follow-up placebo group, week 8, during treatment (n=51)
Discon.nua.on of treatment week 9
Follow-up clonidine group, week 30, after treatment (n=54)
Follow-up placebo group, week 30, after treatment (n=49)
Cross-sectional part
Intervention part
18 Figure 2: Overview of the NorCAPITAL project. This dissertation contains data from the cross-sectional part of the design.
At baseline, all CFS patients underwent assessment of genetics, microbiology,
endocrinology, autonomic function, immune function, cognition, symptoms, and physical activity. In addition, a computer-based randomization procedure allocated a subset of the CFS patients to extended baseline assessment of autonomic function, immune function, cognition, and fMRI of the brain. Similarly, all healthy comparison participants underwent baseline assessments, and a subset was subjected to the extended part that included fMRI.
19
Aims of the Dissertation
The goal of this dissertation was to investigate neurocognitive network dynamics and emotional cognition using fMRI, areas unexplored in adolescent CFS. This dissertation includes three separate studies on an adolescent CFS patient group and a healthy comparison group.
In the first study, intrinsic connectivity networks (ICN) of the brain provide a systematic framework for understanding fundamental aspects of human brain organization and function. We hypothesized that intrinsic functional connectivity within the DMN, CEN and SN would be altered in adolescent CFS patients and that these alterations would be related to clinical symptoms of fatigue and pain.
The aim of the second study was to explore and link emotional conflict processing to underlying neural mechanisms in adolescent CFS. To gauge emotional conflict processing, we measured the amount of interference (response time slowing and decrease in accuracy) observed on behavioral measures. Firstly, we hypothesized that adolescent CFS patients would show reduced behavioral conflict interferences. Secondly, we hypothesized that adolescent CFS patients would exhibit decreased responses in conflict detection regions of the SN—ACC, amygdalae, and insula—and that these alterations would be related to behavioral interferences. Finally, we explored
associations between depressive and anxiety symptoms and neural function and tested whether adolescent CFS patients would show a negative affect bias, a common marker in mood and anxiety disorders.
The aim of the third study was to investigate regional connectivity of the right dorsal anterior insula (dAI) in adolescent CFS patients compared to a healthy comparison
20 group. Secondly, we explored the relationship between right dAI functional connectivity and three functional domains: cognition, pain, and physical activity. The right dAI is a central hub of the brain’s SN and is associated with interoception, biological and cognitive salience detection, and top-down cognitive control engagement.
21
Materials and Methods
The NorCAPITAL-project was conducted at the Department of Pediatrics, Oslo
University Hospital, Norway, which is a national referral center for young CFS patients.
The current studies are based on cross-sectional data collected during the first and second clinical in-hospital day of NorCAPITAL, from March 2010 to May 2012. All participants received a gift-card worth NOK 200. Informed, written consent was obtained from all participants and from parents/next of kin if required. The study was conducted in accordance with the Helsinki Declaration of the World Medical Assembly and approved by the Norwegian National Committee for Ethics in Medical Research.
Participants
All referring units were equipped with written information for distribution to potential study participants and their parents/next of kin. If consent was given, a standard form required the referral unit to confirm the result of clinical investigations considered compulsory to diagnose pediatric CFS (pediatric specialist assessment, comprehensive hematology and biochemistry analyses, chest x-ray, abdominal ultrasound, and brain magnetic resonance imaging; Royal College of Paediatrics and Child Health, 2004). Also, the referring units were required to confirm that the patient (a) was unable to follow normal school routines due to fatigue; (b) was not permanently bedridden; (c) did not have any concurrent medical or psychiatric disorder that might explain the fatigue; (d) did not experience any concurrent demanding life event (such as parents’ divorce) that might explain the fatigue; (e) did not use pharmaceuticals (including hormone
contraceptives) regularly. If medical history or current health status indicated a psychiatric condition, physicians were required to refer patients to a psychiatrist for
22 evaluation. If a comorbid psychiatric disorder was found, those patients were removed from the study (Sulheim et al., 2014). Completed forms were consecutively conveyed to the study center and carefully evaluated. Patients considered eligible for this study were summoned to a clinical meeting at our study center, after which a final inclusion
decision was made.
In agreement with National Institute for Health and Clinical Excellence (NICE) clinical guidelines (National Institute for Health and Clinical Excellence, 2007; Royal College of Paediatrics and Child Health, 2004), we applied a “broad” case definition of CFS,
requiring 3 months of unexplained, disabling chronic/relapsing fatigue of new onset. We did not require that patients meet any other accompanying symptom criteria, in contrast to the case definition from the International Chronic Fatigue Syndrome Study Group at the Centers for Disease Control and Prevention (commonly referred to as the Fukuda- definition), which appears to be most frequently used in the scientific community (Fukuda et al., 1994).
The Fukuda definition requires at least 6 months of unexplained chronic or relapsing fatigue of new onset, severely affecting daily activities, as well as four or more of eight specific accompanying symptoms (headache, muscle pain, joint pain, sore throat, tender lymph nodes, impaired memory or concentration, unrefreshing sleep, and malaise after exertion). However, the validity of this definition has not been established (Brurberg, Fonhus, Larun, Flottorp, & Malterud, 2014). In fact, several empirical findings raise concerns, particularly among adolescents: A formal factor analysis of symptoms in a broadly defined group of chronic fatigued patients did not show a strong
correspondence with the Fukuda-definition of accompanying symptoms (Nisenbaum, Reyes, Unger, & Reeves, 2004). A study based upon the Swedish twin registry concluded
23 that there was no empirical support for the requirement of four out of eight Fukuda defined accompanying symptoms (Sullivan, Pedersen, Jacks, & Evengard, 2005).
A report on a broadly defined population of adolescent CFS patients concluded that the subgroup adhering to the Fukuda criteria was not characterized by a certain level of disability, nor was this subgroup specifically related to characteristics of underlying pathophysiology (Wyller & Helland, 2013). Accordingly, subgrouping based upon the Fukuda criteria did not influence the cross-sectional comparisons or the intervention effects in previously reported results from the NorCAPITAL project (Sulheim et al., 2014). Thus, the inclusion criteria in this study are wider than the Fukuda criteria. The main reason for not adhering to the Fukuda case definition was too few accompanying symptoms.
In NorCAPITAL, a total of 120 CFS patients were included. The studies in this
dissertation were based upon a subset of patients generated from a computer-based randomization procedure, where one fourth of the patients were randomized to be included in the present study (Sulheim et al., 2014). Disease mechanisms in CFS might change over time; in addition, disease duration might be a marker of prognosis. Thus, 18 months disease duration (median disease duration in the NorCAPITAL cohort) served as stratification criterion for the randomization procedure (Sulheim et al., 2014). The randomization procedure allocated 30 patients to fMRI assessment. A group of 24 healthy adolescents having a comparable distribution of gender and age were recruited from local schools. No chronic disease and no regular use of pharmaceuticals were allowed.
24 Of the 30 CFS patients allocated to fMRI, there were extenuating circumstances that led to a reduced number of patients in the final dataset. In study 1 and 3, five patients did not want to participate in the study, four patients were excluded due to orthodontic treatment, two were excluded due to scanning error, and one was excluded due to excessive movement > 3 mm in either of the three translation or three rotation parameters. In study 2, five patients did not want to participate in the study, four patients were excluded due to orthodontic treatment, two were excluded due to scanning error, and three patients were excluded due to excessive movement
parameters >3 mm in either of the three translation or three rotation parameters, and one was excluded due to poor performance (<50% accuracy).
Data Collection
A one-day in-hospital assessment included clinical examination, blood sampling,
autonomic testing, and cognitive tests. All participants were instructed to fast overnight and abstain from tobacco products and caffeine at least 48 hours. All procedures were carried out in a quiet room, in a fixed sequence, and by three researchers. Participants that were allocated to extended baseline assessment underwent brain fMRI and extended cognitive testing the next day.
Details of the procedures relevant for the present studies are outlined below in Table 1.
TABLE 1: Data collection methods and analysis methods used in the articles.
Method Articles
Symptom questionnaires
Fatiguea 1, 2, and 3
Depressionb 1, 2, and 3
25
Anxietyc 1 and 2
Paind 1
Neuropsychological tests
Intelligence quotient (IQ)e 1, 2, and 3
Emotional Conflict task during fMRI scanning 2
Working memory performancef 3
Neural measurement
Resting-state BOLD fMRI 1 and 3
Event-related BOLD fMRI 2
Independent component analysis (ICA) 1
Seed-to-voxel connectivity analysis 3
Pain and activity measures
Pressure pain threshold (PPT)g 3
Physical activity (activPAL)h 3
aChalder Fatigue Questionnaire (Chalder et al., 1993)
bMoods and Feelings Questionnaire (Sund, Larsson, & Wichstrom, 2001)
cSpielberger State Anxiety Inventory (Spielberger, Gorsuch, & Lushene, 1973)
dBrief Pain Inventory (Daut, Cleeland, & Flanery, 1983)
eWASI estimated full IQ(Wechsler, 2007)
fWISC-IV Digit Span Forward and Backward(Wechsler, 2004)
gPPT (Nie, Arendt-Nielsen, Andersen, & Graven-Nielsen, 2005)
hactivPAL (Dowd, Harrington, & Donnelly, 2012)
Statistical analysis
Analysis of the resting-state fMRI data was performed using the software package FSL (available from the FMRIB Software Library at www.fmrib.ox.ac.uk/fsl). Functional images of the emotional conflict task were analyzed using SPM (ver.8; www.fil.ion.
ucl.ac.uk/spm). Regional FC of right AI was determined by bivariate correlation using the Connectivity (CONN) FC Toolbox (ver.15.p; www.nitrc.org/ projects/conn). Neural
26 measurements, symptom questionnaires, neuropsychological tests, and demographic data were evaluated using SPSS, version 22, (IBM Inc.; Chicago, IL). Extra correlation analyses were performed using SPSS, version 24, (IBM Inc.; Chicago, IL). Symptom data were missing at random for two of the patients, and the group mean was used for their lost data.
Additional analyses specific to the PhD dissertation
In the writing the dissertation, additional analyses were performed on the whole patient sample and comparison participants to further assess the relationships between
working memory performance, fatigue, pain tolerance, and physical activity. Our group previously reported data on working memory (Sulheim et al., 2015), pain (Winger et al., 2014), and physical activity performance (Sulheim et al., 2014), but a comparison of how working memory relates to these factors has not been reported.
27
Synopsis of studies
Study 1 Background
Neural network investigations were absent in adolescent CFS. In this study, we examined whether the core intrinsic connectivity networks (ICN) were altered in adolescent CFS patients.
Methods
Eighteen adolescent patients with CFS and 18 aged matched healthy comparison participants underwent resting-state fMRI. Data was analyzed using dual-regression ICA, which is a data-driven approach for the identification of independent brain networks. Intrinsic connectivity was evaluated in the DMN, SN, and CEN. Associations between network characteristics and symptoms of CFS were also explored.
Results
Adolescent CFS patients displayed a significant decrease in SN functional connectivity to the right posterior insula compared to healthy comparison participants, which was related to fatigue symptoms.
Conclusion
Our findings of insula dysfunction and its association with fatigue severity in adolescent CFS demonstrate an aberration of the salience network, which might play a role in CFS pathophysiology.
28 Study 2
Background
Studies of neurocognition suggest that abnormalities in cognitive control contribute to the pathophysiology of CFS in adolescence, yet these abnormalities remain poorly understood at the neurobiological level. In this study, we examined whether affective and cognitive processing was altered on behavioral and neural levels in adolescents with CFS and a healthy comparison group.
Methods
Fifteen adolescent patients with CFS and 24 healthy adolescent participants underwent fMRI while performing an emotional conflict task that involved categorizing facial affect while ignoring overlaid affect labeled words.
Results
Adolescent CFS patients were less able to engage the left amygdala and left midposterior insula (mpINS) in response to conflict than the healthy comparison group. Associations between behavioral measures and conflict-related reactivity in the amygdala and mpINS were reported. Neural responses in the amygdala and mpINS were specific to fatigue severity.
Conclusion
These data demonstrate that adolescent CFS patients displayed deficits in emotional conflict processing. Our results suggest abnormalities in affective and cognitive functioning, which might underlie the pathophysiology of adolescent CFS.
29 Study 3
Background
Impairments in cognition, pain intolerance, and physical inactivity characterize adolescent CFS, yet little is known about its neurobiology. The right dorsal-anterior insular (dAI) connectivity provides a motivational context for biologically and
cognitively salient stimuli. In this study, we examined regional functional connectivity patterns of the right dAI in adolescent CFS patients and healthy participants.
Methods
Eighteen adolescent patients with CFS and 18 aged-matched healthy adolescent control participants underwent resting-state fMRI. The right dAI region of interest was
examined in a seed-to-voxel resting-state functional connectivity (FC) analysis using the CONN toolbox.
Results
Adolescent CFS patients demonstrated reduced right dAI FC to the right posterior parietal cortex (PPC) compared to the healthy group. Right dAI – PPC FC lacked associations with three functional domains in the CFS group; however, these known relationships were observed in the healthy group.
Conclusion
Our findings of dysfunctional connectivity of the right dAI and loss of functional associations with cognition, pain tolerance, and physical activity might represent a fundamental aspect in the neural architecture of adolescent CFS pathophysiology.
30 Additional results specific to the PhD dissertation
Simple correlation analyses showed evidence of a relationship between fatigue
symptoms and working memory (sum scores) performance in patients, r(120) = -.206, p
< .05, and poor evidence in healthy participants, r(39) = -.197, p = .23. However, there was evidence of a relationship between pain tolerance and working memory
performance in healthy participants, r(39) = .319, p < .05, and no evidence in patients, r(120) = .097, p = .29. There was good evidence of a relationship between physical
activity and working memory performance in healthy participants r(39) = .484, p < .01, and no evidence in patients, r(120) = .068, p = .46 (Table 2; Figure 3, 4 and 5).
TABLE 2: Pearson’s correlation coefficient values between working memory performance relationships with pain tolerance, physical activity, and fatigue.
Working memory (sum scores) CFS group Comparison group
PPT .097 .319*
Physical activity .068 .484**
Fatigue -.206* -.197
Note: * p < 0.05, **p < 0.01, ***p < 0.001 (2-tailed).
R² = 0.043
0 5 10 15 20 25 30 35
0 5 10 15 20 25 30
Fatigue Symptoms
Working memory performance
Fatigue and Cognition
CFS
Comparison CFS
Comparison
31 Figure 3: Scatter plot contains working memory performances (sum scores) and fatigue symptom scores from each group. Dark circles represent individual patients with CFS and lighter circles represent healthy comparison (HC) participants. As the degree of fatigue symptoms increased in the CFS group, working memory abilities decreased. A statistically significant relationship in the HC group was not found.
Figure 4: Scatter plot contains working memory performances (sum scores) and pain tolerance scores from each group. Dark circles represent individual patients with CFS and lighter circles represent HC participants. As working memory abilities increased so did the ability to tolerate pain in the HC group, but this relationship was not observed in the CFS group.
R² = 0.102
0 20 40 60 80 100 120 140 160 180
0 5 10 15 20 25 30
Pain tolerance
Woring memory performance
Pain and Cognition
CFS
Comparison CFS
Comparison
32 Figure 5: Scatter plot contains working memory performances (sum scores) and
physical activity scores from each group. Dark circles represent individual patients with CFS and lighter circles represent HC participants. As working memory abilities
increased so did physical activity in the HC group, but this relationship was not observed in the CFS group.
R² = 0.235
0 5000 10000 15000 20000 25000
0 5 10 15 20 25 30
Physical activity
Working memory performance
Physical activity and Cognition
CFS
Comparison CFS
Comparison
33
Discussion
Main findings
We found intrinsic functional connectivity (FC) of the salience network (SN)
distinguished adolescent CFS patients from their healthy counterparts. SN FC decreases to the right posterior insula were associated with fatigue severity. We reported
adolescent CFS patients were less able to engage the left amygdala and left midposterior insula (mpINS) in response to conflict than the healthy comparison group. Associations between behavioral interferences and conflict-related reactivity in the amygdala and mpINS were observed. Finally, we observed a reduction in the right dorsal-anterior insular (dAI) regional FC to the right posterior parietal cortex (PPC) in adolescent CFS patients compared to the healthy group. In the CFS group, right dAI – PPC FC decreases lacked associations with three functional domains: cognition, pain, and physical activity;
however, these known relationships were observed in the healthy group. Depressive symptoms did not influence neural alterations in the three studies.
Neurocognitive network dynamics
Resting-state brain network investigations have revealed interdependent and
distributed regions that engage and disengage depending on goal-directed behaviors (Raichle et al., 2001). The default mode network (DMN), central executive network (CEN), and salience network (SN) comprise the core intrinsic connectivity networks (ICN) (Menon, 2011), and efficiencies in the inter- and intra-connectedness of ICN underlie healthy neural development (Supekar & Menon, 2012). This ICN ensemble is understood to play a major role in shaping goal-directed processes—such as focused
34 attention and cognitive control, with a smaller amount of neural variance contributed to the actual cognitive task (Cole et al., 2014; Williams, 2016). The abnormal interplay of ICN dynamics is understood to underlie psychiatric and neurological disorders (Menon, 2011).
The neurocognitive network analyses of study 1 found only the SN distinguished the groups. Dysfunctional SN connectivity suggests an inefficiency in attending to
biologically and cognitively important information in patients. The DMN is anchored in the posterior cingulate cortex and ventromedial prefrontal cortex (vmPFC), and
alterations in the DMN are associated with deficits in self-referential mental activity (e.g.
rumination and hallucination) (Greicius et al., 2007; Whitfield-Gabrieli et al., 2009). The CEN encompasses dorsolateral prefrontal (dlPFC) and lateral posterior parietal cortices (PPC) and deficits in this network reflect impoverished cognition (e.g. working memory and executive control functions) (D'Esposito, 2007; Fuster, 2000; Goldman-Rakic, 1995;
E. K. Miller & Cohen, 2001; Smith & Jonides, 1998). The SN involves the cingulate-frontal operculum system, whose integral hub is the anterior insula (AI), and changes in the SN are related to attention deficits, alterations in the detection and selection of salient sensory, emotion, and cognitive stimuli (Menon, 2015).
Specifically, adolescent CFS patients illustrated a pattern of decreased FC to the right posterior, middle, and ventral AI within the SN, in which FC decreases to the posterior insular region correlated with increasing fatigue symptoms. The decreased SN FC to the right posterior-to-anterior insular sequence appear to signify inefficient signaling, representation, and integration of ascending sensory information that seems to engender a deficient model of fatigue awareness and subjectivity in patients. Even though we did not observe alterations within the DMN and CEN, deviations along the
35 posterior – anterior insular axis are common in disorders where there appears to be a disruption in the interpretation of salient biological and cognitive information (Craig, 2002, 2009, 2011; Menon, 2015).
The alteration in FC to the right posterior insula and its correlation with fatigue severity in patients is particularly important because it parallels abnormalities found in pain research (Baliki & Apkarian, 2015; Baliki, Geha, Fields, & Apkarian, 2010; Baliki et al., 2012; Mansour et al., 2013). It is well established that sensory receptors in all tissue and organs in the body send information along homeostatic afferent pathways that is
integrated in the dorsal posterior insular cortex and provides an afferent representation of the physiological condition of the body – interoception (Craig, 2003b, 2009, 2011).
The interoceptive cortex integrates feelings from the body along the posterior – anterior insula. Re-representations are formed in the middle insula, and the right anterior insula provides a meta-representation of the state of the body that is associated with subjective awareness (Craig, 2002, 2003a, 2003b, 2009, 2011).
It could be that pain is an intrinsic feature of CFS because fatigue and pain are
primordial drives, like thirst, hunger, and temperature, which are generated by the same homeostatic afferent pathways that terminate in the posterior insular region of the interoceptive cortex (Craig, 2003b). These findings enter CFS into the heated debate within chronic pain research regarding the respective importance of peripheral
afferents versus the brain’s interpretation of afferent signals (Baliki & Apkarian, 2015).
The identification of common neurocognitive mechanisms provides an opportunity for future CFS research to combine knowledge discovered in chronic pain research and together advance treatment options for patients.
36 Affective and Cognitive processing
Within the adolescent CFS patients of this study, psychiatric disorders were ruled out prior to inclusion, but still they reported significantly more anxiety and depressive symptoms than the healthy group. Elevated anxiety and depressive symptoms are indicators of altered emotional processing. Additionally, CFS appears to be strongly associated with childhood adversities (Afari et al., 2014) and the underlying
pathophysiology is indicative of a sustained stress response (Wyller et al., 2009). The influence of life stress on depression is well studied (Colman et al., 2014) and adds to the usefulness of applying a cognitive task with an emotional component to investigate emotional and cognitive interactions in adolescent CFS.
In the emotional conflict task of study 2, there was no evidence of behavioral (response times or accuracy) differences between-groups, but neural activity showed the left amygdala and mpINS differentiated the groups. Even though the behavioral measures did not reveal between-groups effects, we found that the relationship between group differences and behavioral measures varied as a function of conflict-related neural activity in the left amygdala and mpINS, which better illustrates group differences in brain–behavior relations. Brain regions online during conflict detection and commonly recruited in conflict tasks—bilateral amygdalae, dorsal anterior cingulate cortex (dACC), and right AI were derived from prior work (Egner, Etkin, Gale, & Hirsch, 2008; Etkin, Egner, Peraza, Kandel, & Hirsch, 2006; Marusak, Etkin, & Thomason, 2015). The left midposterior insula (mpINS) region was included because of its association with stress in previous studies using conflict tasks (Bruce et al., 2012; Marusak, Etkin, et al., 2015).
Neural activity in regions responsible for detecting conflict should reflect the amount of behavioral interference, resulting in higher activity in those regions.
37 The linear association between conflict-related activity in the amygdala and behavioral interference measure shows the conflicting stimuli did not produce a significant
response in the amygdala. Even though amygdala reactivity was related to behavioral measure, it was not increased in response to emotional conflict and could be an indication of inefficiencies in the neural detection and appraisal systems in patients.
Reactivity of the amygdala has been associated with conflict detection, stimuli appraisal, and emotion regulation using emotional conflict tasks (Egner et al., 2008; Etkin et al., 2006).
The linear relationship between conflict reactivity in the mpINS and interference measure shows the conflicting stimuli did not produce a significant response in the mpINS. Even though the conflict-related reactivity of the mpINS was related to behavioral measure, it was not increased in response to emotional conflict and
displayed an opposite neural pattern to those that are stress related in anxiety disorder.
Interpersonal stress was associated with elevated activity of the mpINS using conflict tasks (Bruce et al., 2012; Marusak, Etkin, et al., 2015).
Regional right Anterior Insula Connectivity
In study 1, we found SN decreases in FC to the right insula, but the posterior to anterior pattern did not include the dorsal anterior insula (dAI), which corresponds most closely with the AI hub of the SN (Menon, 2015). The right dAI is not only associated with interoceptive awareness (Craig, 2002, 2009, 2011; Critchley, Wiens, Rotshtein, Ohman,
& Dolan, 2004), but is also important in the engagement of the CEN for top-down cognitive control (Menon & Uddin, 2010; Sridharan, Levitin, & Menon, 2008; Supekar &
Menon, 2012; Uddin, Supekar, Ryali, & Menon, 2011). Furthermore, maturation of the
38 functional coupling between the right dAI node of the SN and right posterior parietal cortex (PPC) node of the CEN is suggested to underlie cognitive control development (Supekar & Menon, 2012).
In study 3, immature FC of the right dAI - PPC and implied loss of cognitive control over associated functions of cognition, pain, and physical activity might be an indication of how prolonged fatigue potentially threatens normal neurocognitive network
development in adolescent CFS. Cognition (Supekar & Menon, 2012), pain (Craig, 2002, 2009), and physical activity (Erickson, Hillman, & Kramer, 2015; Erickson, Leckie, &
Weinstein, 2014; Khan & Hillman, 2014; Voss, Nagamatsu, Liu-Ambrose, & Kramer, 2011) are three functions associated with efficient right AI activity and cognitive control in studies on healthy groups and found in the healthy participants of this study. They happen to represent three clinical domains of impaired function in adolescent CFS (Sulheim et al., 2015; Sulheim et al., 2014; Winger et al., 2014).
Working memory performance is associated with connectivity efficiencies in the frontal- parietal regions of the CEN (D'Esposito, 2007; Fuster, 2000; E. K. Miller & Cohen, 2001), and SN deficits in engaging CEN contribute to impoverished cognitive control, as well (Menon & Uddin, 2010; Sridharan et al., 2008; Supekar & Menon, 2012; Uddin et al., 2011). For further support of the cognitive control mechanism, right dAI – PPC, found in study 3, we compared working memory performance with scores of fatigue symptoms, PPT, and physical activity to see if relationships were present in a larger sample of participants (Figure 3, 4, and 5). Again, we found a relationship between better working memory performance (indicating efficient cognitive control) and pain tolerance and physical activity in healthy participants, but not in the CFS group. Additionally, we found a negative relationship between working memory impairment and fatigue severity in
39 the patients. This relationship was not statistically significant in the health participants, but in a larger sample, fatigue might be related to a general decline in cognitive control.
These results suggest a tight interplay between, dysfunctional brain connectivity,
impoverished cognitive control, fatigue severity, pain intolerance, and physical inactivity in adolescent CFS.
Pain processing
In pain theory, research has provided evidence on the unlikeliness that the cortex contains a pain matrix - neural tissue specifically responsive to nociceptive inputs or associated specifically to pain perception (Baliki & Apkarian, 2015). Instead, brain activity is interpreted as a process where nociceptor activity moves from the
subconscious to a conscious unpleasant perception, an experiential shift that is likely generalized across sensory modalities (Baliki & Apkarian, 2015). Therefore,
interpretations drawn from pain and resting-state data cannot adequately relate to a network in pain perception, as data is acquired during non-painful rest. The important features of resting-state activity are now understood as a representation of an
individual’s history of learning and memory that underlie perceptual variability (Lewis, Baldassarre, Committeri, Romani, & Corbetta, 2009). Pain theory also states that it is the same neural shift, as found in the subconscious to a conscious unpleasant nociceptive perception, that occurs in the neural transformations from acute to chronic pain (Baliki
& Apkarian, 2015; Baliki et al., 2010; Baliki et al., 2012; Mansour et al., 2013). It could be that similar neural shifts are also manifested in adolescent CFS, which is why pain is commonly comorbid.
40 Pain theory now suggests that the frontal cortical drives are embedded in corticostriatal circuits, which actively control the threshold for incorporating sensory afferent inputs into cortical conscious states, across sensory modalities (Baliki & Apkarian, 2015). Shifts in the threshold mechanisms of this circuitry influence synaptic learning-based
reorganization and lowers conscious perception of pain (Apkarian, 2008; Apkarian et al., 2009). The region best related to the consciousness of pain is the AI (Baliki, Geha, &
Apkarian, 2009), and top-down cognitive control regions modulate pain awareness of the AI (Lamm, Meltzoff, & Decety, 2010). Lowered PPT in the CFS patients of our study might be an indication of a shift in circuitry thresholds, and FC decreases of the dAI could indicate a loss of cognitive control in modulating conscious pain perception.
Physical activity
Physical activity in childhood has been shown to influence neural circuitry supporting efficiency in cognitive control (see Khan and Hillman (2014) for review). In
understanding the relationship between cognition and physical activity, neuroscience focuses on motivational processes of cognitive and physical effort. Initial studies on neural systems underlying cognitive effort-based decision making suggest they are similar to those that are related to physical effort (Botvinick & Braver, 2015; Schmidt, Lebreton, Clery-Melin, Daunizeau, & Pessiglione, 2012). In CFS reduced motivational neural circuitry was related to increased mental and general fatigue and reduced physical activity in adults (Miller et al., 2014) and increased fatigue symptoms in children (Mizuno et al., 2016).
Neural representations of the effort, and the valence of the outcomes they yield, might form the foundation of motivated behavior. An integration of cost and benefit outcomes
41 of effort might derive from a motivational context provided in the AI, where worse outcomes seem to have greater representation (Kurniawan, Guitart-Masip, Dayan, &
Dolan, 2013), and from the up-regulation of top-down control processes in response to motivationally salient cues (Padmala & Pessoa, 2011). It could be that the physical activity decreases in the CFS patients of our study yield a costly outcome representation in the dAI that corresponds with the failure to engage top-down cognitive control brain regions.
CFS brain connectivity
High-level attention and cognitive control processing require efficient interactions of the brain’s SN and CEN (Menon, 2015). In adult CFS, analyses of white matter pathways revealed right hemispheric alterations in the arcuate fasciculus that may reflect degeneration of crossing fibers or strengthening of short-range fibers (Zeineh et al., 2015). The arcuate fasciculus is a bundle of long and short fibers that runs laterally to connect frontal and parietal lobes (Catani & Thiebaut de Schotten, 2008). CFS alterations of the right arcuate fasciculus (Zeineh et al., 2015) might underlie the FC abnormalities of the right dAI – PPC found in study 3.
Consistent with study 1, SN alterations have been identified previously in CFS (Gay et al., 2015) and right insula FC decreases have been observed (Boissoneault, Letzen, Lai, O'Shea, et al., 2016; Boissoneault, Letzen, Lai, Robinson, et al., 2016). In adult CFS, decreased intrinsic connectivity of the CEN was observed (Gay et al., 2015) and both increases and decreases in regional FC patterns of the CEN have been reported
(Boissoneault, Letzen, Lai, O'Shea, et al., 2016; Boissoneault, Letzen, Lai, Robinson, et al., 2016). Even though study 1 did not find intrinsic CEN changes in adolescent patients,
42 the regional FC decreases between the SN node and CEN node in study 3 suggest
dysfunctional interactions between brain networks.
Sustained Arousal
The sustained arousal model of disease mechanisms in CFS suggests the central role of aberrant systemic stress responses (Wyller et al., 2009). Cumulative stress decreases right insular volume (Ansell, Rando, Tuit, Guarnaccia, & Sinha, 2012) and alters
underlying dopaminergic function (Pani, Porcella, & Gessa, 2000), which is important for the modulation of motivation and cognitive control interactions (Botvinick & Braver, 2015; Pani et al., 2000), sensory processing (Baliki & Apkarian, 2015), and self
awareness (Lou et al., 2016). We propose adolescent CFS is maintained by the inability to regulate stress that causes alterations in neural circuitry, which diminishes cognitive control and directly enhances ascending sensory neurotransmission. Although
speculative, sustained arousal might specifically explain the dysfunctional neural connectivity, fatigue severity, cognitive abnormalities, pain intolerance, and physical inactivity found in adolescent CFS patients. Within the frame of sustained arousal, future research should further address adolescent CFS symptomatology and their neural- mechanistic correlates with the goal to improve treatment options for patients.
CFS neural mechanisms
Extending the theory of brain mechanisms that underlie the chronification of pain (Baliki & Apkarian, 2015), they might also parallel those that occur in the move from acute to persistent fatigue. One such mechanism of brain reorganization was the decrease in FC and gray matter density of the right posterior insula, which diminishes cognitive control and directly enhances ascending transmission (Baliki & Apkarian,
43 2015; Baliki et al., 2010; Baliki et al., 2012). The ICN investigation of study 1 found decreased FC to the right posterior insula was associated with increased fatigue symptoms in patients. This could be an indication of the same right posterior insula reorganization found in chronic pain, but it contributes to increased fatigue perception and a decline in general cognitive functioning in adolescent CFS.
If we look at the results of studies 1, 2, and 3 in light of dysfunctional SN connectivity, which suggests abnormal processing of salient biological and cognitive information, it could be that major subcortical nodes of the SN – the amygdala, ventral striatum, and ventral tegmental area – weaken context-specific access to affective, motivational, and painful cues. The mesolimbic circuitry is a major dopamine pathway that begins in the ventral tegmental area, connects the nucleus accumbens (NAc), amygdala, hippocampus, and prefrontal cortex, and mediates motivated behavior. In fact synaptic learning-based reorganization of the NAc is implicated in the neural shift away from the posterior insula in chronic pain patients (Baliki & Apkarian, 2015; Baliki et al., 2010; Baliki et al., 2012;
Mansour et al., 2013). Even though the NAc was not a direct finding from studies 1, 2, and 3, alterations in the mesolimbic system can be inferred based on the findings of brain-behavior relations in study 2 and other childhood CFS research.
In response to emotional conflict, study 2 found a non-responsive linear relationship between the amygdala and mpINS and behavioral measures. It could be that the lack of response-time emotional conflict effect in patients was the failure of the NAc to respond to emotional cues during congruent and incongruent trials, which were not deemed motivationally significant. The competing emotional information in the stimuli should either aid or impair cognition, accordingly. Specifically, happy trials, as opposed to fearful trials, are regarded as low interference, in terms of arousal, and we found that it
