Chronic fatigue syndrome (CFS), commonly known as myalgic encephalomyelitis (ME), is a debilitating disease of unknown etiology. CFS/ME is a heterogeneous disease associated with a myriad of symptoms but with severe, prolonged fatigue as the core symptom associated with the disease. There are currently no known biomarkers for the disease, largely due to the lack of knowledge surrounding the eitopathogenesis of CFS/ME. Numerous studies have been conducted in an attempt to identify potential biomarkers for the disease. This mini-review offers a brief summary of current research into the identification of metabolic abnormalities in CFS/ME which may represent potential biomarkers for the disease. The progress of research into key areas including immune dysregulation, mitochondrial dysfunction, 5′-adenosine monophosphate-activated protein kinase activation, skeletal muscle cell acidosis, and metabolomics are presented here. Studies outlined in this mini-review show many potential causes for the pathogenesis of CFS/ME and identify many potential metabolic biomarkers for the disease from the aforementioned research areas. The future of CFS/ME research should focus on building on the potential biomarkers for the disease using multi-disciplinary techniques at multiple research sites in order to produce robust data sets. Whether the metabolic changes identified in this mini-review occur as a cause or a consequence of the disease must also be established.
Chronic fatigue syndrome/myalgic encephalomyelitis (CFS/ME) is a debilitating disease centered on the symptoms of fatigue and post-exertional malaise (PEM). There are currently no definitive biomarkers for the disease resulting in a lack of effective treatment options. This mini-review outlines some of the progress being made regarding the identification of potential metabolic biomarkers in CFS/ME.
Data suggest that CFS/ME patients have a low level of chronic activation of the immune system. Numerous markers of immune dysregulation have previously been reported in CFS/ME patients including, but not limited to, increased levels of pro-inflammatory cytokines, natural killer cell function deficiency, CD8+ cytotoxic T-cell numbers, increased expression of activation markers on T-lymphocyte cell surfaces, and altered activity in mitochondrial function [1,2]. Abnormalities in processes such as these may explain symptoms typically found in CFS/ME such as fatigue, PEM, and autonomic symptoms .
Many inflammatory cytokines have been investigated in terms of CFS/ME with some of the most commonly reported being tumor necrosis factor α, C-reactive protein, interleukin (IL)-1, and IL-6. A relatively large 2017 study (186 patients; 388 controls) investigating cytokines in CFS/ME found only two cytokines to differ between the CFS/ME and healthy control cohorts . Transforming growth factor beta was shown to be elevated in CFS/ME patients, while resistin was shown to be decreased in the patient cohort. The same study investigated correlations between 51 cytokines and disease severity. Seventeen cytokines were shown to positively correlate with disease severity including 13 pro-inflammatory cytokines, and only one cytokine (CXCL9) was shown to have an inverse correlation with disease severity. However, inconsistent results between studies looking at the role of cytokines in CFS/ME have been reported [4–8]. The effect of cytokine response to exercise has been investigated as one of the key symptoms of CFS/ME is PEM [7,8]. There were conflicting results from two of the most recent studies looking at the effect of exercise on cytokine production with one study showing no significant associations between cytokines and post-exertional symptoms or perceived effort , while another recent study was able to discriminate between CFS/ME patients and healthy controls post-exercise using cytokine profiling .
The most consistently reported immune irregularity in CFS/ME is a reduction in the number or the function of natural killer (NK) cells [9,10]. The proliferation, maturation, and activation of NK cells are stimulated by specific cytokines including IL-2, IL-12, IL-15, and IL-18 . Activated NK cells proceed to release cytokines to activate other NK cells in addition to a general activation of the immune system as a whole. CD69 acts as an early marker of NK cell activation; therefore, a reduction in the number of NK cells and/or expression of CD69 may open the door to allowing infections to become chronic due to the lack of challenge by the host immune system. Consistent with this hypothesis, it has previously been shown that there is a reduction in the numbers of NK- and CD69-positive cells in CFS/ME patients [12,13]. Previously recorded lower cytotoxic capacity of circulating NK cells in CFS/ME patients than in healthy controls means that the bodies of patients are less capable of eliminating viruses and may provide an insight into the cause of the disease . Other studies have also found there to be an increase in the expression of adhesion and activation cytokines by NK cells among CFS/ME patients [15–17].
More recent immune studies in CFS/ME have focused on the role of B-cells in the pathogenesis of disease and as a potential target for therapies. A recent study in Norway has investigated the effect of the B-cell depleting drug rituximab on CFS/ME patients and has shown it to significantly improve symptoms associated with the disease . The group conducted a double-blind placebo-controlled study as well as an open-label phase II study and showed rituximab to be effective in alleviating symptoms in a subset of CFS/ME patients [19,20]. Another area of interest in the role of B-cells in the pathogenesis of CFS/ME is the presence of autoantibodies to neuroendocrine receptors [21–24]. One study has shown altered autoantibody receptor expression on B-cells which could cause the clinical presentation of symptoms associated with CFS/ME, although this has yet to be shown, as the functional consequences of these changes are still to be determined . Other groups have shown altered proportions of B-cell subsets in CFS/ME patients such as elevated CD20- and CD21-positive B-cells, and increased transitional and naive B-cell subsets [15,17]. However, these results have not been confirmed by results from other research teams [26,27].
While immune dysregulation appears to be a promising area of research with regard to CFS/ME, studies have shown that there is no correlation between changes in lymphocytes — with their activation or the populations of subsets — and clinical improvement in CFS/ME [28,29]. Previous studies have also shown that stressors such as sleep disturbance, which is commonly observed in CFS/ME patients, can both initiate and perpetuate changes in immune function. Therefore, any changes in immune function may occur as a consequence of other symptoms of the disease as opposed to being the primary causing factor . Owing to the heterogeneous nature of the disease and the day-to-day variability within each patient, variations in proteins such as cytokines might be expected due to their short half-lives. Additionally, research into immune function within the CFS/ME population is made more difficult due to the naturally variable nature of immune cells throughout the day . This variability may explain the inconsistent results between studies. Additionally, inconsistent methodology and patient sampling between studies may contribute to the conflicting results between studies.
There is some evidence that CFS/ME may be caused, at least in part, by an acquired mitochondrial dysfunction. Studies into CFS/ME have shown key indicators of mitochondrial dysfunction such as abnormal adenosine triphosphate (ATP) production, increased mitochondrial damage, and an impairment of the oxidative phosphorylation pathway [31–36]. Additionally, crucial symptoms associated with CFS/ME, such as fatigue and myalgia, are also shared by diseases known to be caused by mitochondrial dysfunction . One group investigating mitochondrial dysfunction in CFS/ME neutrophils has suggested that ATP levels are significantly reduced in CFS/ME patients in such a way that can be detected and used as a diagnostic test [32–34]. These studies used a series of measures of ATP production compiled into a single value, termed the mitochondrial energy score, which is used as a marker of mitochondrial function. Not only did these studies show a significant reduction in mitochondrial function in the neutrophils of CFS/ME patients compared with healthy controls, but also showed a significant correlation between mitochondrial function and disease severity with those more severely affected by the disease having a lower mitochondrial function. A more recent study investigating ATP production in peripheral blood mononuclear cells (PBMCs) of CFS/ME patients showed ATP to be elevated in CFS/ME patients compared with healthy controls . However, while overall levels of ATP were shown to be increased in CFS/ME PBMCs, when mitochondrial ATP and non-mitochondrial ATP were looked at separately, mitochondrial ATP was shown to be comparable between the CFS/ME and control cohorts while the non-mitochondrial produced ATP was significantly elevated in the CFS/ME cohort. This suggests that the abnormality in PBMC ATP production does not occur as a result of mitochondrial abnormalities but rather occurs due to abnormalities in other ATP-producing pathways. A recent study using extracellular flux analysis to analyze the oxidative phosphorylation (OXPHOS) and glycolytic function of PBMCs from CFS/ME patients and healthy controls showed significantly lower OXPHOS in CFS/ME PBMCs . Seven parameters of OXPHOS were measured, and significant differences between the CFS/ME and control cohorts were reported. The key OXPHOS parameters for differences between the two cohorts were determined to be basal respiration and maximal respiration. Consistent impairment of maximal respiration in CFS/ME PBMCs was shown in all experimental conditions which suggest that CFS/ME PBMCs are unable to increase their OXPHOS when cells are put under stress, leaving the cells unable to meet the cellular demand for energy. The significantly lower basal respiration in CFS/ME PBMCs suggests that these cells may also be unable to meet cellular energy demands under basal conditions. Overall, results suggested that CFS/ME PBMCs were unable to utilize mitochondrial respiration (OXPHOS) to the same extent as control PBMCs. This impairment indicates that PBMCs may have a role to play in the disease pathway and should be utilized in future experiments into the identification of biomarkers in CFS/ME. The same study also reported comparable levels of glycolysis in PBMCs from CFS/ME and healthy control cohorts, which suggests normal functioning of the glycolysis pathway .
Mitochondrial function in the skeletal muscle cells of CFS/ME patients have also been investigated [35,39]. In vivo and in vitro studies investigating bioenergetics in skeletal muscle of CFS/ME patients have reported abnormalities in CFS/ME. Phosphorus magnetic resonance spectroscopy (PMRS) has been used to show decreased ATP production rate and slower re-synthesis of ATP following exercise [40,41]. Results from one study implied that there was an increase in lactate and a decrease in mitochondrial ATP production in patients when compared with controls . However, plasma creatine kinase levels from the same study suggested that the muscular mitochondrial OXPHOS was normal in patients and therefore postulated that the decrease in mitochondrial ATP synthesis must be due to another factor and not caused by a defect in the enzymes of the OXPHOS pathway. This conclusion was supported by other studies , one of which showed no significant difference in the activity of mitochondrial respiratory chain complexes in the skeletal muscle of CFS/ME patients and healthy controls . However, other studies have shown that bioenergetic abnormalities in skeletal muscle only exist in a subset of the patient population , while some have found no differences in ATP production among CFS/ME patients [43,45]. Some studies have suggested that the primary cause of CFS/ME is due to oxidative damage in the muscle of patients and not as a direct result of abnormalities in ATP production [41,46]. Studies have shown there to be a strong correlation between the degree of mitochondrial dysfunction and the severity of illness , although other studies have produced conflicting data [44,47–49].
Owing to the sensitivity of mitochondria to cellular changes, and their role as the energy production center of the cell, any small changes in mitochondrial function may have a large effect on patients such as causing symptoms such as severe fatigue and PEM. However, studies showing changes in mitochondrial function in CFS/ME should be interpreted with caution as the mitochondrial changes may be due to dysregulation of upstream signaling pathways, such as the 5′-adenosine monophosphate-activated protein kinase (AMPK) pathway, rather than a defect in the mitochondria themselves. The area of mitochondrial dysfunction in CFS/ME requires further study in order to determine the exact role for mitochondria in the disease.
5′-Adenosine monophosphate-activated protein kinase
AMPK is a heterotrimeric enzyme that plays a role in maintaining intracellular metabolic homeostasis. It is a central component of cellular metabolism playing a key part in regulating growth and metabolism in response to a decrease in intracellular ATP levels .
In a study by Brown et al. , electrical pulse stimulation was used to investigate the effect of exercise on the phenotype of skeletal muscle cells. Results showed an impairment of activation of AMPK in CFS/ME cells. Additionally, there was shown to be impaired stimulation of glucose uptake and a diminished release of IL-6. Results from the present study are indicative of a primary muscular abnormality in CFS/ME and may be responsible for the PEM experienced by patients. Owing to the retention of these differences in CFS/ME muscle cells that have been cultured in vitro, it has been postulated that muscular abnormalities in CFS/ME may arise as a consequence of genetic or epigenetic variations .
AMPK has been shown to be impaired in fibroblasts taken from fibromyalgia (a CFS/ME-like disease with high levels of fatigue) in response to moderate oxidative stress. AMPK activation by metformin or incubation with serum from caloric-restricted mice improved the response to moderate oxidative stress and mitochondrial metabolism in fibromyalgia fibroblasts . This may provide a potential therapy option for the amelioration of fatigue not only in CFS/ME but also other diseases in which fatigue plays a large role.
Skeletal muscle cell acidosis
Previous experiments have shown there to be an over-depletion of energy stores in the skeletal muscles of CFS/ME patients . It has been postulated that this occurs as a result of an overutilization of the anaerobic respiratory pathways within the cells thereby causing a decrease in the capacity of the patient to exercise and an increase in PEM experienced by the patient . This hypothesis is supported by studies which have shown CFS/ME patients to have significantly reduced anaerobic thresholds when compared with sedentary controls [39,53]. One byproduct of anaerobic respiration is protons . The ability of skeletal muscle cells to manage the protons released via anaerobic respiration can influence the exercise performance of the individual and may explain why CFS/ME patients suffer from PEM and feel unable to perform relatively low-level exercise tasks due to a lack of energy [56,57]. Investigation into this avenue of CFS/ME research is vital as even minimal changes in pH, such as those caused by an inability to manage protons correctly, can have widespread effects on multiple vital cellular processes such as enzyme function. Increased lactate production is not unique to skeletal muscle cells in CFS/ME, with previous research showing an increase in ventricular cerebrospinal fluid lactate .
Studies have shown there to be abnormalities in the peripheral bioenergetics of CFS/ME patients. Many studies have used PMRS with conflicting results. One study by Jones et al.  used PMRS to record differences in intramuscular acid handling between CFS/ME patients and sedentary controls after a standardized exercise protocol. Results showed patients to have a significantly higher suppression of proton efflux immediately post-exercise in addition to patients having significantly longer times recorded before maximum proton efflux was reached. These results imply that abnormalities in intramuscular pH recovery exist in CFS/ME patients after exercise . Proton efflux is essential for the normalization of acidosis which builds up during exercise; therefore, the suggestion that CFS/ME patients are unable to deal with the acid build up as well as control subjects may provide an explanation as to why patients feel pain and fatigue in skeletal muscles following even low-level exercise. A small in vitro study looking at mitochondrial respiratory chain complex activity in skeletal muscle cells from CFS/ME patients and healthy controls showed no significant differences between the two cohorts, although a lower mitochondrial cohort was indicated . This suggests that the skeletal muscle cell acidosis and resulting PEM observed in CFS/ME patients do not occur as a result of altered mitochondrial function. As with other previously discussed hypotheses surrounding CFS/ME, there is conflicting data on the role of intracellular acidosis in the disease with other studies reporting no significant differences in intramuscular pH or lactate threshold when patients were subjected to a graded exercise test until exhaustion [49,59].
One hypothesis suggests that any observed changes in intracellular pH in the muscle cells of CFS/ME patients may be due to an overutilization of the lactate dehydrogenase (LDH) pathway and therefore an increase in the conversion of pyruvate to lactate. This could be due to a down-regulation of the pyruvate dehydrogenase complex (PDC). PDC coverts pyruvate into acetyl-CoA; therefore, when PDC is inhibited, pyruvate is instead converted into lactate by LDH. Pyruvate dehydrogenase kinase (PDK) regulates PDC activity by phosphorylating PDC and preventing enzymatic activity, thereby preventing the conversion of pyruvate into acetyl-coA. It has been hypothesized that CFS/ME patients may experience an overutilization of the LDH pathway due to increased inhibition of PDC by overactive PDK, resulting in a build-up of lactic acid within the muscle cells. One study used amino acid detection and mRNA expression to indicate that a functional impairment of PDC in CFS/ME patient blood may have a role to play in the pathogenesis of the disease by showing increased mRNA expression of PDK 1, 2, and 4, and an altered amino acid profile . Preliminary, unpublished work investigating the biochemical basis of skeletal muscle cell dysfunction in CFS/ME explored the possibility of skeletal muscle cell acidosis caused by an overutilization of the LDH pathway as the cause of CFS/ME . This work included a comprehensive assessment of the role of the LDH pathway in CFS/ME skeletal muscle cells by measuring intracellular pH, lactate concentration, glycolysis, and mitochondrial function of skeletal muscle cells, as well as the effect of an in vitro exercise model on the intracellular pH and lactate concentration of the cells . This work showed no evidence of abnormalities in skeletal muscle cells of CFS/ME patients.
One of the most comprehensive studies of metabolomics in CFS/ME was published in 2016 by Naviaux et al. . This study assessed 612 metabolites, from 63 pathways, in the plasma of 45 CFS/ME patients and 39 healthy controls. Results support the hypothesis of a hypometabolic state in CFS/ME, with 20 different pathways shown to be altered, with 80% of those pathways being down-regulated in CFS/ME. The pathways from which abnormal levels of metabolites were identified varied, with affected pathways including mitochondrial metabolism, which supports some studies discussed in previous sections. The authors suggest that CFE/ME patients have a homogenous metabolic response and conclude that this study has identified a chemical signature for CFS/ME using plasma metabolomics which could be used a biomarker for the disease . However, this interpretation of the results should be treated with caution, given the small sample size involved and the large number of metabolites assessed in a single study with a relatively small sample size. Additionally, patients with diseases with fatigue as a core symptom, such as fibromyalgia, Primary Sjogren's syndrome, and type 2 diabetes, should be assessed to see if the metabolic response identified here is unique to CFS/ME. Sedentary controls should also be included in future studies to assess whether the metabolic profile identified is due to deconditioning of CFS/ME patients. Results from the present study have identified a potential biomarker for CFS/ME; however, further work is needed to validate the results.
The etiopathogenesis of CFS/ME remains elusive despite efforts to identify biomarkers for the disease. Studies outlined in this mini-review have shown many potential causes for the pathogenesis of CFS/ME and identified potential biomarkers for the disease including cytokines, mitochondrial changes, skeletal muscle cell abnormalities following exercise, AMPK activation, and metabolic changes. The conflicting data reported between studies make identifying biomarkers difficult due to the lack of repeatability of studies. To elucidate the mechanisms of the disease, further research into the potential biomarkers identified to date must be conducted by multiple research groups in order to produce robust data sets identifying the most promising areas of CFS/ME research.
5′-adenosine monophosphate-activated protein kinase
chronic fatigue syndrome
peripheral blood mononuclear cell
pyruvate dehydrogenase complex
pyruvate dehydrogenase kinase
phosphorus magnetic resonance spectroscopy
The authors thank the following funders: the Medical Research Council (MRC), Action for ME, ME Research U.K., and the ME Association.
The authors thank the researchers who conducted the studies presented in this manuscript and all the patients who participated in those studies.
The Authors declare that there are no competing interests associated with the manuscript.