Energy metabolism in sickness behavior versus ME/CFS

Acute inflammatory conditions consume large amounts of energy, whereby adipose and muscle tissues and glycogen are used to provide energy to combat pathogens [48]. This causes a negative energy balance; that is, the expended usage of calories is greater than the intake. Lowered synthesis of muscle proteins, proteolysis, and lipolysis and loss of tissue proteins and fats are typically induced by proinflammatory cytokines produced in response to invading pathogens in injured tissues [48]. The objective of the sickness response from an evolutionary perspective is energy conservation [3, 7, 15]. Proinflammatory cytokines, such as TNFα, not only modulate a negative energy balance to compensate for the increased energy demands but also reduce food intake by causing anorexia and decreased voluntary energy utilization [48]. Characteristic sickness behavior symptoms, such as lethargy, sleepiness, fatigue, listlessness, malaise, hyperalgesia, psychomotor retardation, loss of libido and cognitive deficits aim to limit highly energy-consuming processes, such as motor, sexual and brain activities [3]. As such, energy that otherwise would be consumed for locomotor, reproductive and neurocognitive activities is withdrawn from the brain, muscles and other peripheral tissues and redirected to fuel the calorie-dependent activities of immune cells [2, 3, 7, 15]. Thus, the typical sickness behaviors save energy and contribute to redirecting the conserved energy to immune cells as well as to pyrogenic and other inflammatory processes. As such, sickness behaviors, in shifting resources to a patterned immune response, help to combat acute infection or injury.

Moreover, by inducing anorexia (and thus restricting calorie intake), intracellular signaling pathways associated with inflammation are attenuated. This includes those leading to IL-6 production [49]. Calorie restriction additionally attenuates lipopolysaccharide (LPS)-induced sickness behavior in a dose-dependent manner [50]. Therefore, we have argued that sickness behavior is part of the compensatory (anti)-inflammatory reflex system (CIRS), which is induced in response to inflammatory processes [3]. Thus, sickness behavior augments the beneficial effects of inflammation and at the same time limits an overzealous inflammatory response [3, 50]. The increased energy expenditure and anorexia leading to the negative energy balance eventually causes loss of body fat, protein mass and thus lean body mass and therefore may lead to inflammation-associated weight loss [3]. Weight loss results in anti-inflammatory effects attenuating the production of, for example, IL-6 and TNFα and increasing that of endogenous anti-inflammatory compounds, for example, adiponectin [3]. Thus, not only proinflammatory cytokine-induced anorexia but also the consequent weight loss is a homeostatic adaptive response, which should be considered as part of the CIRS [3].

Finally, energy saving sickness behaviors, such as fatigue, listlessness, loss of libido, and neurocognitive disorders, not only play a role in the resolution of inflammation but are crucial in preventing the transition from acute to chronic inflammation. Thus, by saving energy and compensating for the negative energy balance sickness behaviors prevent inflammatory sequelae, such as depletion of energy stores, cachexia, anemia, osteopenia and insulin resistance, which determine the transition towards chronic inflammation [51]. The transition from acute to chronic inflammation occurs around 19 to 43 days after the onset of the acute phase of inflammation [51]. Chronic inflammation ensues when the acute inflammatory response and the CIRS, including sickness behavior, were not able to eradicate the primary infectious agent or heal the injury, for example, pyogenic bacteria, viral infections, fungi, sarcoidosis, and autoimmune responses [52].

ME/CFS, however, is a chronic disease, which is accompanied by an inability to generate energy on demand [12]. Mitochondrial dysfunctions and abnormally low ATP and high lactate levels play a role in ME/CFS [12, 53–61]. ME/CFS patients display exercise-induced exhaustion much earlier than healthy controls. When reaching exhaustion, ME/CFS patients display diminished intracellular ATP and dysregulated oxidative metabolism, coupled to increased glycolysis in the exercising striated muscles [12, 58]. ME/CFS patients display reduced rates of ATP resynthesis in the aftermath of exercise versus controls resulting from impaired oxidative phosphorylation [56]. Patients with ME/CFS show prolonged elevations of lactate, returning extremely slowly to normal levels [53, 59]. Behan et al. [55] found histopathological abnormalities in the mitochondria of skeletal muscles in ME/CFS. During exercise, the latter display decreased voluntary muscle contractions, which worsen subsequently [59]. Thus, in ME/CFS patients reduced intracellular ATP, oxidative metabolism and accelerated glycolysis in skeletal muscles determine early exhaustion [12, 58, 61]. Several studies report significantly increased levels of ventricular lactate in ME/CFS patients, suggesting central energy dysregulation [62, 63].

We previously proposed that the defects in energy production and mitochondrial functions in ME/CFS are probably caused by chronic inflammatory and oxidative and nitrosative stress (O&NS) processes [12, 58]. Thus, raised levels of proinflammatory cytokines resulting from systemic inflammation disable oxidative phosphorylation within mitochondria. This is reflected in the increased lactate and mitochondrial dysfunction commonly seen in chronic inflammatory states [64]. For example, TNFα causes a marked decrease in mitochondrial membrane potential, which may increase reactive oxygen species (ROS) [65] and increases mitochondrial membrane permeability leading to membrane depolarization [66]. Increased ROS damages the electron transport chain leading to depleted ATP production, which in turn causes a deficiency in oxidative phosphorylation leading to overt mitochondrial disease [67]. Defects in oxidative phosphorylation leads to increased ROS production, which acts to create self-propagating mitochondrial dysfunction [68] contributing to increased IL-1β and IL-18 via inflammasome induction [69]. Mitochondrial shutdown induced by proinflammatory cytokines is ultimately engineered by nitric oxide (NO) and is reversible [70]. NO signaling regulates mitochondrial number and function [71] and inhibits mitochondrial respiration [72] by competing with oxygen at complex 1 and cytochrome oxidase [73] resulting in diminished ATP production. Peroxynitrite inhibits mitochondrial respiration by modulating electron transport complexes I and III [74]. Attenuation of mitochondrial respiration by for example NO or its derivatives activates ROS and reactive nitrogen species (RNS) produced by mitochondria [75]. In addition, the increased levels of nuclear factor (NF)κB observed in ME/CFS [76] contribute to a shift towards aerobic glycolysis (the Warburg effect) as observed in cancer cells [58].

Kennedy et al. [54] reported elevated surrogate biomarkers of O&NS in ME/CFS, which positively associated with symptom exacerbation following energy expenditure. Fatigue results from ROS accumulation and diminished availability of ATP in muscle cells [77]. As to how this relates to the putative subtypes of Booth et al. [78], where ME/CFS patients are subdivided on the basis of decreased ADP uptake by mitochondria versus decreased ATP output requires further investigation. These two subgroups are predicted to differentially show increased lactate production and altered ability for repeat exercise. It is not unlikely that changes in oxidative status, lipid peroxidation and mitochondrial membrane rigidity are impacting on the activity of the ADP-ATP translocator (TL), perhaps having differential effects on TL uptake or output functions. Around 20% of mitochondria in cells at any given point in time are in the process of transport, either exhausted and being removed from sites of high-energy need or fresh mitochondria are being imported into such sites. Further investigation is required as to whether the energetic changes in mitochondria in ME/CFS contribute to dysregulated signaling for transport. This may overlap with the genetic susceptibility to ME/CFS mediated by changes in the disrupted-in-schizophrenia 1 (DISC1) gene, which is associated with mitochondrial transport [79]. ATP is also an important neuronal and glia transmitter. As to how altered mitochondrial ATP regulation modulates such wider ATP functions is unknown, but would be expected to contribute to cognitive deficits, especially in high energy demand activities of the ventral lateral prefrontal cortex that are required to unhook specific memory exemplars.

In summary, energy metabolism plays a key role in sickness behavior and ME/CFS and also differentiates sickness from ME/CFS. Table 1 and Figure 1 show the differences in energy metabolism between both conditions. Thus, sickness behaviors aim to conserve and redirect energy to immune cells, preventing transition from acute to chronic inflammation. ME/CFS, however, shows chronic dysfunctions in ATP and lactate production, caused by peripheral and central mitochondrial dysfunctions, driven, in part, by chronic inflammation and O&NS.

Immunoinflammatory pathways in sickness versus ME/CFS

Elevated levels of proinflammatory cytokines drive most if not all aspects of the sickness response either directly or indirectly [80]. Proinflammatory cytokines, including IL-1, IL-6, and TNFα, activate the production and/or release of secondary inflammatory mediators, such as prostaglandins (PGs) and NO [81]. Proinflammatory cytokines directly stimulate numerous neurohormonal systems. A variety of mechanisms allow proinflammatory cytokine signals to circumnavigate of the blood brain barrier [82]. Blood borne proinflammatory cytokines can diffuse through relatively permeable areas of the blood brain barrier via the circumventricular organs [6], allowing proinflammatory cytokines to directly interact with microglia and astrocytes in the glia limitans. Another crucial pathway involves a complex series of interactions between proinflammatory cytokines and brain endothelial cells [3]. The third pathway involves the activation of neurons by proinflammatory cytokines via the vagus nerve and catecholaminergic circuits of the sympathetic nervous system (SNS) [12]. (For a more detailed examination of immune to brain communication, see [83].) Interestingly, the production of TNFα persists in the brain much longer than in the periphery [84–86], while TNFα additionally passes both ways across the blood-brain barrier [87]. This is concordant with chronic neuroinflammation and low-grade peripheral inflammation being interconnected phenomena. Within the brain, proinflammatory cytokines, prostaglandins and NO invoke brain responses to infection in the periphery [88]. Once microglia are activated, astrocytes are recruited leading to further activation of neuroinflammatory signals [89]. Thus, both cell types collaborate to propagate neuroinflammation and the physiological changes directed at diminishing the replication of pathogens increasing the metabolism of carbohydrate fat and protein [90] and generating energy-conserving behaviors [3].

The central nervous system (CNS) modulates the immune response via two pathways: (a) the hypothalamic-pituitary-adrenal (HPA) axis and (b) the release of catecholamines and acetylcholine [91]. Communication of proinflammatory cytokine signaling to the brain via the vagus provokes a rapid anti-inflammatory response through increased HPA axis and cholinergic nerve activity [92]. This forms the basis of the CIRS response including the HPA axis, cholinergic and anti-inflammatory responses [3]. This CIRS response in the CNS and autonomous nervous system generates its own characteristic symptom pattern (for example, autonomic symptoms) and is thus part of the sickness response. The SNS modulates the immune response at systemic, local and regional levels [93]. Catecholamines inhibit production of proinflammatory and T helper (Th)1 cytokines, while enhancing the synthesis of Th2 cytokines together with IL-10 and transforming growth factor (TGF) 1 [94]. Acetylcholine release through activation of the parasympathetic nervous system attenuates the production of IL-1, TNFα and IL-6 [91].

Such mechanisms underpinning the sickness response reveals it as an automated irresistibly integrated response conserved by evolution and which is induced by proinflammatory cytokines, including IL-1 and TNFα. The relation between elevated proinflammatory cytokines and fatigue and fatiguability is well documented [95]. Elevated proinflammatory cytokines are largely responsible for the severe fatigue seen in multiple sclerosis [96], cancer-related fatigue and cognitive symptoms [97], fatigued breast cancer survivors [98], and the debilitating fatigue reported by underperforming athletes [99]. Inhibition of IL-1β is associated with a 50% reduction in fatigue in patients with Sjögren's syndrome [100]. Proinflammatory cytokines and prostaglandin E2 (PGE2) mediate inflammatory hyperalgesia [95] as well as neurocognitive abnormalities leading to problems with concentration and memory [95, 101]. Many of these adverse effects especially in the CNS are mediated by IL-1β [102]. The hippocampus is the hub of processes involving learning and memory and is also the region of the brain with the highest levels of IL-1 receptors and is especially vulnerable to the effects of neuroinflammation [103]. While physiological levels of IL-1 are essential in the development of memory and learning, elevated endogenous levels results in neurocognitive dysfunctions and abnormal behaviors [104, 105]. IL-1β also mediates the suppression of feeding behavior and appears to regulate lipid metabolism by antagonizing the performance of lipoprotein lipase, which is the principal regulator of lipid store mobilization in the body [106]. IL-1β and NFκB are essential mediators of the anhedonic effects of sickness behaviors [107]. Moreover, IL-1β in concert with TNFα inhibits sexual behavior [108]. TNFα and IL-1β suppress the clock genes that regulate circadian rhythm and hence could influence a veritable myriad of biological functions. The net effect of the action of these cytokines on clock genes is to reduce the metabolic rate and the demand for energy in the form of ATP [15].

In summary, there is some evidence that shared inflammatory pathways (that is, increased proinflammatory cytokine levels) underpin sickness behavior and determine part of the symptoms in ME/CFS (see also Table 1). However, while sickness behavior is a beneficial response induced by proinflammatory cytokines to conserve energy and combat pathogens thus preventing transition to chronic inflammation, ME/CFS is a chronic inflammatory disorder accompanied by a combined Th1 and Th2 response and multiple signs of a seriously dysregulated immune system.

O&NS pathways in sickness versus ME/CFS

Other pathways that discriminate sickness behavior from chronic inflammatory disorders are reduced antioxidant levels and increased O&NS, driving O&NS damage to lipids, DNA, proteins and mitochondria (see above). Inflammatory processes, CMI activation and O&NS processes are inseparably associated. For example, phagocytes and activated M1 macrophages produce large amounts of ROS and RNS [123]. CMI activation produces neopterin, increasing ROS and RNS [124]. O&NS processes also activate the production of proinflammatory cytokines and T cells [125]. Therefore sickness behavior may be accompanied by elevated levels of ROS/RNS, which help to eradicate invading pathogens, that then normalize with the resolution of inflammation.

ME/CFS, however, shows chronically increased ROS/RNS, O&NS processes and O&NS damage, including increased malondialdehyde (MDA), isoprostane, 2,3 diphosphoglyceric acid, 8-OH-deoxyguanosine, protein carbonyls and thiobutyric acid [126–130]. iNOS production is significantly higher in ME/CFS patients versus controls [131]. As discussed in the next section, ME/CFS is accompanied by a chronic hyperproduction of NO [95, 126]. Raised oxidative stress levels also occur in response to exercise in ME/CFS [129] potentially explaining one of the mechanisms underlying post-exertion malaise [12]. Skeletal muscle oxidative imbalance contributes to increased muscle fatiguability [130].

Lowered levels of coenzyme Q10, zinc and glutathione have been reported in ME/CFS [126, 127, 131] with amelioration of oxidative stress occurring during remission [128]. As well as decreased coenzyme Q10 [127], decreased vitamin C and E will impact mitochondrial function, increasing lipid peroxidation. Such effects are potentially prevented by melatonin [132], as shown after strenuous exercise in humans [133]. Maes et al. [134] reported significantly increased ω6 lineolic and arachidonic acids in ME/CFS, driving a reduced ω3 to ω6 polyunsaturated fatty acids (PUFAs) ratio versus controls, contributing to inflammation in ME/CFS [135].

Such alterations can generate pathology via a number of mechanisms. Isoprostane, a prostaglandin (PG)F2-like compound, correlates positively and significantly with general ME/CFS disease severity as well as the magnitude of exercise induced disease exacerbation [54]. Elevated MDA promotes phospholipids rigidity, altering membrane fluidity, permeability and transport mechanisms [136]. Importantly, elevated MDA decreases mitochondrial membrane fluidity [137]. A decrease in melatonin, found in some ME/CFS studies, would contribute to increased mitochondrial membrane rigidity [138], as well as reduced natural killer cell activity in ME/CFS [119]. Melatonin also modulates aspects of sickness behaviors [139], inhibiting NFκB induction and microglia activation [140], suggesting that variations in melatonin may confound the comparison of processes in ME/CFS and sickness behavior.

In summary, while sickness behavior is probably accompanied by increased ROS/RNS, which should normalize upon the resolution of inflammation, ME/CFS is accompanied by chronically activated O&NS processes causing chronic damage to lipids, DNA, proteins and mitochondria. While in sickness behavior the putative increase in ROS/RNS would be an adaptive response, increased O&NS largely underpins the pathophysiology of ME/CFS.

Autoimmune responses in ME/CFS

Autoimmune processes also frequently occur in ME/CFS and are not germane to a beneficial short-term response as sickness behavior. IgM autoimmune responses directed against oxidatively damaged lipid membrane components (for example, oleic acid), anchorage molecules (for example, palmitic and myristic acid and S-farnesyl-L-cysteine), residue molecules of lipid peroxidation (for example, azelaic acid and MDA), and amino acids or proteins modified by nitrosylating species (for example, NO-tyrosine, NO-phenylalanine, NO-tryptophan, NO-arginine, and NO-cysteine) have all been reported [12, 126]. These reactions are directed against neoantigenic determinants (neoepitopes), which are created as a result of damage to lipids and proteins by O&NS [12, 126, 141]. These antigenic structures are normally invisible to the immune system but become targeted because structural modifications induced by elevated O&NS have rendered them immunogenic. The levels of these corrupted entities correlate positively and significantly with the severity of selected ME/CFS symptoms, such as fatigue, muscle pain, a flu-like malaise, and so on [12, 126].

Reports of frank autoimmune reactions are commonplace in people with ME/CFS. Antibodies to cardiolipin, nuclear envelope antigens and neuronal antigens have all been reported [12, 142–145]. Several teams have reported elevated titers of autoantibodies directed towards gangliosides, serotonin, phospholipids, anti-68/48K and microtubule associated proteome as well as anti-lamin single stranded DNA [145–147]. Autoantibodies against muscarinic cholinergic receptors, mu-opioid receptors and dopamine receptors have been detected [148]. Autoantibodies directed against ganglioside M1, for example, play a role in neuroimmune disorders, correlating with neurocognitive dysfunctions as observed in neuropsychiatric systemic lupus erythematosus [149, 150].

HPA axis function in sickness and ME/CFS

Sickness behaviors induced by different challenges and in different species are accompanied by marginally increased to significantly increased plasma corticosterone levels [151]. Moreover, corticosterone causes sensitization of LPS-induced pyrexia and pain, but not lethargy, and leads to enhanced sickness-induced neuroinflammation [152]. IL-1, via upregulation of prostaglandin and COX enzymes, enhances HPA axis activation [153] and hypothalamic norepinephrine synthesis [154]. IL-1β also increases brain tryptophan concentrations and the rate of 5- hydroxytryptamine (5-HT) metabolism, which may further activate the HPA axis [154]. Given glucocorticoid anti-inflammatory effects, HPA axis activation is part of the CIRS response [3]. Glucocorticoids also significantly regulate immune genes and cell functions, for example, decreasing the Th1/Th2 ratio [155]. Moreover, glucocorticoids participate in programmed cell death and energy-related processes, for example, glucose, lipid, protein and carbohydrate homeostasis [155, 156].

The activation of the HPA axis in sickness behavior contrasts the findings in ME/CFS. HPA axis hypoactivity is a characteristic feature in some people with ME/CFS [157, 158], for example, low baseline levels of HPA axis hormones, aberrant diurnal hormone levels, reduced HPA axis responses to provocation by corticotropin-releasing hormone (CRH) or adrenocorticotropic hormone (ACTH), blunted HPA axis responses to physical and psychological stress, and enhanced sensitivity to glucocorticoids [159–165]. This lowered HPA activity may be explained by prolonged stimulation of O&NS and immunoinflammatory pathways in ME/CFS. For example, chronic elevations of IL-6 may blunt the release of ACTH [166]. TNFα may inhibit the stimulation of CRH, ACTH-induced cortisol release and adrenal gland function [167, 168].

In summary, while sickness behavior is accompanied by HPA axis activation as part of a CIRS, ME/CFS is accompanied by lowered HPA axis activity, which may be secondary to activation of immunoinflammatory and O&NS pathways (see Table 1).

Leptin: sickness versus ME/CFS

Leptin is an important mediator of infection-induced inflammation and sickness behaviors [169, 170]. However, leptin also accelerates the recovery from hypoxia-induced sickness behavior via an increase in IL-1 receptor antagonist (IL-1RA) [171]. Obese rodents show an altered inflammatory and behavioral response to infection [172].

Treatment of ME/CFS with low dose hydrocortisone increases leptin, leading to a favorable treatment response [173]. Interestingly, recent data shows leptin as a risk factor for, and displaying efficacy in the treatment of depression, as well as Alzheimer's disease [174–176]. Leptin, like melatonin, is a wide immune regulator, increasing natural killer cell activity and Th1 responses, while inhibiting the cAMP induction of tryptophan 2,3-dioxygenase (TDO) in astrocytes, thereby inhibiting cognitive deficits mediated by TDO induction of kynurenic acid. The severity of ME/CFS is associated with increased indicants of metabolic syndrome [177], suggesting that wider metabolic dysregulation associated with obesity and leptin resistance will modulate the course and perhaps etiology of ME/CFS. Thus, leptin may have significant and differential regulatory effects in ME/CFS and sickness behaviors.

Trigger factors in sickness behavior and ME/CFS

Sickness behavior is an adaptive CIRS response induced by acute injuries or bacterial and viral infections. These phenomenological experiences combined with pyrexia and neuroendocrine changes represent an integrated hierarchal system conserved by evolution to combat infection, conserve energy, play a role in the resolution of inflammation and limit an overzealous inflammatory response [3]. When resolution of inflammation is not induced or when the inflammatory response is overzealous, a chronic inflammatory state may emerge [52, 178]. The consequent chronic inflammatory state is then localized where the trigger, either infection or injury, was present leading to persistent infections and medical disorders related to the affected organs or systemic immunoinflammatory and O&NS responses [3, 179].

In contrast to the role of acute infection/injury in sickness behavior, different trigger factors such as acute and chronic infections, environmental factors and other medical disorders may play a role in ME/CFS. Depending on the applied case definitions, the presence of medical illnesses may sometimes be regarded as exclusion criteria for a diagnosis of CFS [16]. Thus, Fukuda's criteria consider that the fatigue should not be caused by any conditions that may be identified by specific tests or diagnoses [16]. Other criteria however make the presence of neuroendocrine symptoms and intolerance of mental or physical exercise mandatory [10]. The latter authors propose that a strategy be developed which could detect patients with a neuroimmune condition rather than focus on patient populations that have fatigue whose origin is not revealed by routine medical testing or explained by any psychiatric condition. Moreover, we have argued that ME/CFS may be caused by other inflammatory disorders, including multiple sclerosis, autoimmune disorders, and so on, and that in these conditions also the diagnosis ME/CFS or ME/CFS due to general medical condition should be made [12, 17].

Upper respiratory system and flu-like infections often precede the onset of ME/CFS [180]. ME/CFS often occurs in epidemics suggesting that infections may cause ME/CFS [181]. Bacterial and viral infections, including Epstein-Barr virus (EBV), Coxiella burnetii, Parvo B19 and Mycoplasma, are well known trigger factors associated with the onset of ME/CFS [182]. However, infections not only trigger ME/CFS but may also function as maintaining factors. Thus, many patients with ME/CFS have persistent, recurrent or opportunistic bacterial and viral infections [183, 184]. These infections may maintain ME/CFS or cause relapses [185]. Moreover, the number and severity of symptoms, including neurological symptoms, is correlated with the existence of concurrent infections [186, 187]. Gene expression data show latent viral or bacterial infections in ME/CFS, for example, Epstein-Barr virus, enteroviruses and C. burnetii [188–190]. Other infections that are associated with this disorder are, among others, human herpesvirus (HHV) 6 and 7, cytomegalovirus, enteroviruses, Borna disease virus, Chlamydia pneumoniae and Borrelia burgdorferi [182].

In summary, sickness behavior and ME/CFS are related to infections. One major difference, however, is that sickness behavior is a short-lasting, adaptive response to acute infection, whereas recurrent or opportunistic infections play a role in the severity and in the relapsing and chronic course of ME/CFS. While sickness behavior is caused by acute infections, the onset of ME/CFS may sometimes be associated with the long-term effects of an acute infection, such as long-standing neuroinflammation, O&NS processes or the onset of autoimmune reactions. Moreover, it is probable that when an initial infection is not cleared and/or there was no resolution of inflammation, a chronic infection with activation of immunoinflammatory and O&NS processes may ensue. Arguably these processes may be associated with the onset of ME/CFS. By inference, in some patients who initially showed sickness behaviors in response to an acute infection, ME/CFS may develop. From a clinical point of view, the transition from acute (sickness behavior) to chronic inflammation occurs between 19 to 43 days. According to the diagnostic criteria of ME and CFS [10, 16], the patient cannot be diagnosed as having ME/CFS until 6 months after the onset of the disease. This means that patients, even when experiencing chronic inflammatory processes and from ME/CFS symptoms, cannot be diagnosed as ME/CFS between days 19 to 43 and 6 months. Prolonged or persistent sickness behaviors for this period are not adequate labels because sickness behavior denotes an adaptive and acute inflammatory response, whereas the patients after 19 to 43 days are already in a chronic inflammatory state. Therefore, we propose that those patients should be categorized as having chronic fatigue (CF), a diagnosis which then should be changed into CFS or ME some months later [11].

In contrast to sickness behavior, the initial trigger in ME/CFS is not always well defined, while the chronic stage is associated with positive feedback loops between inflammatory processes and degenerative and autoimmune processes [95, 189]. Thus, CFS and ME show a similar pattern to other chronic degenerative disorders in which the initiating trigger is not well defined, for example, clinical depression, schizophrenia, multiple sclerosis, Parkinson's disease, inflammatory bowel disease, cancers, including breast cancer, and autoimmune disorders, lupus erythematosus, Sjögren's syndrome and rheumatoid arthritis, and so on. Moreover, patients with these disorders are primed to develop CF and CFS (and probably ME) through activated immunoinflammatory and O&NS pathways [12, 95]. For example, the prevalence of CF/CFS is very high in the above-mentioned autoimmune (41% to 81%) and inflammatory disorders, such as ankylosing spondylitis, biliary cirrhosis, post-poliomyelitis and psoriatic arthritis (48% to 50%) [191]. In post-stroke patients, fatigue is associated with inflammatory biomarkers [192]. Pascoe et al. [193] reviewed that post-stroke depression is caused by inflammatory and O&NS processes. By inference, similar processes may explain the onset of post-stroke fatigue. Research also showed a significant association between fatigue and cardiovascular disease (CVD), another inflammatory and O&NS disease [194]. There is also an increased incidence of chronic fatigue in hemodialysis patients ('hemodialysis fatigue') [195]. Since hemodialysis is accompanied by inflammatory and O&NS processes, the latter could also explain the increased incidence of fatigue during hemodialysis. In the postpartum period, another inflammatory state, increased levels of fatigue were found with a significant overlap between postnatal fatigue and depressive symptoms [196]. IFNα-based immunotherapy not only causes clinical depression through induction of the cytokine network, but also by an increased incidence of fatigue and CFS-like symptoms [197, 198]. Increased bacterial translocation is also associated with the onset of ME/CFS [199]. Loosening of the gut barrier may allow poorly invasive Gram-negative bacteria to translocate from the gut into the mesenteric lymph nodes and sometimes into the blood stream. Once translocated, the LPS is recognized by the Toll-like receptor 4 (TLR4) complex, which primes immune cells and consequently activates inflammatory and O&NS pathways [12, 199].

Psychological stressors are also associated with the onset of ME/CFS [200]. Psychosocial stress also increases the frequency of relapses or a general worsening of symptoms [17]. This may be explained by psychological stress leading to elevated levels of proinflammatory cytokines and O&NS processes that may well create the environment that fosters increased disease activity [201].

In contrast to sickness behavior, which is a CIRS response to infection/injury, a number of pathological predisposing factors may increase the vulnerability to develop ME/CFS. For example, IgG subclass deficiencies, vitamin D deficits, immune gene polymorphisms and a lowered ω3/ω6 PUFA ratio are observed in ME/CFS [12, 134]. IgG1 and IgG3 subclass deficiencies increase the risk to infections and autoimmune and inflammatory responses and thus to ME/CFS [12, 185]. A decrease in vitamin D is associated with ME/CFS, decreasing not only bone density, but also increasing the susceptibility to, and severity of, infections [202], in part via the regulation of NFκB [203]. Genetic polymorphisms in TNFα, IFNγ, IL-17, IL-10 and HLA genes are associated with the onset of ME/CFS [145]. The DISC1 gene is a susceptibility factor for ME/CFS, as well as for schizophrenia and depression [204]. DISC1 has a significant role in early development, including in the regulation of neuronal migration [205]. Given the relevance of altered mitochondrial function to ME/CFS, it is interesting that DISC1 is strongly associated with mitochondria, significantly regulating mitochondrial function and transport [79]. A possible role of early life immune insult has been proposed for ME/CFS, suggesting that the immune system may be primed by prenatal and postnatal immune regulatory events for an altered response to infection [206]. The lowered ω3/ω6 ratio detected in ME/CFS patients predisposes towards inflammatory and autoimmune responses as ω3 PUFAs are strongly anti-inflammatory and ω6 PUFAs proinflammatory [134].