Kindling and Oxidative Stress as Contributors to Myalgic Encephalomyelitis/Chronic Fatigue Syndrome

Abstract

Myalgic Encephalomyelitis/chronic fatigue syndrome (ME/CFS) is one of the more complex illnesses involving multiple systems within the body. Onset of ME/CFS frequently occurs quickly, and many patients report a prior exposure to a viral infection. This debilitating illness can affect the immune, neuroendocrine, autonomic, and neurologic systems. Abnormal biological findings among some patients have included aberrant ion transport and ion channel activity, cortisol deficiency, sympathetic nervous system hyperactivity, EEG spike waves, left ventricular dysfunction in the heart, low natural killer cell cytotoxicity, and a shift from Th1 to Th2 cytokines. We propose that the kindling and oxidative stress theories provide a heuristic template for better understanding the at times conflicting findings regarding the etiology and pathophysiology of this illness.

Chronic fatigue syndrome (CFS), also known as Myalgic Encephalomyelitis or Myalgic Encephalopathy (ME), is a highly incapacitating illness with an annual value of lost productivity in the USA estimated to be $9.1 billion (Reynolds, Vernon, Bouchery, & Reeves, 2004). Moreover, total direct and indirect costs due to Myalgic Encephalomyelitis/chronic fatigue syndrome (ME/CFS) ranges from $19 to $24 billion (Jason, Benton, Johnson, & Valentine, 2008). Patients with ME/CFS are more functionally impaired than those suffering from type II diabetes mellitus, congestive heart failure, multiple sclerosis, and end-stage renal disease (Anderson & Ferrans, 1997; Buchwald, Pearlman, Umali, Schmaling, & Katon, 1996). Andersson and Ferrans found that the scores for quality of life were lower for ME/CFS than any other chronic illness group. It is estimated that over 800,000 individuals have this illness in the USA (Jason et al., 1999). Given the prevalence and impact of this illness on patients, there is a need to identify theoretical frameworks for better understanding the etiology and pathophysiology of this complex illness. Below, we review two promising theories involving kindling and oxidative stress.

Kindling Theory

According to the kindling theory, repeated exposure to an initially subthreshold stimulus can eventually exceed threshold limits, resulting in persistent hypersensitivity to the stimulus and ultimately, spontaneous behavioral manifestations. Kindling among patients with ME/CFS might appear after prolonged stimulation of the limbic-hypothalamic-pituitary axis, either by high-intensity stimulation (e.g., brain trauma) or by chronically repeated low-intensity stimulation (e.g., an infectious illness). In support of kindling theory, patients with ME/CFS often report prior exposure to a viral infection. Viruses increase activation of macrophages, which produce a release of interleukin-1beta, causing an alteration in the electrical activity of the brain (Maier, Watkins, & Fleshner, 1994).

Kindling was first discovered by Goddard (1967) while studying the effects of electrical stimulation of the amygdaloid complex on learning. Initially, he electrically stimulated brains of rats at a very low intensity, which was below the threshold for eliciting seizure activity. When this stimulation was applied over a period of weeks, the rats experienced epileptic convulsions. In other words, after repeated exposure to small electric shocks, the rats began to have spontaneous seizure-like electrical events. Goddard and others also found it was possible to induce kindling via chemical stimulation. Evidence indicates that kindling involves afterdischarge of cell populations that continue to fire after the initiating stimulation has ceased (Loescher & Ebert, 1996). The stimulus is followed by a growing EEG 3Hz spiking, which increases and decreases in amplitude many times. At times the 3Hz spiking can vanish, only to return seconds later. This seizure activity often spreads to adjacent structures in the brain.

Gellhorn (1970) postulated that under prolonged stimulation of the limbic-hypothalamic-pituitary axis, a lowered threshold for activation can occur. Once this system is charged, either by high-intensity stimulation (e.g., due to an acute viral infection) or by chronically repeated low-intensity stimulation (e.g., through repeated chemical exposure), it can sustain a high level of arousal with little or no external stimulus. Girdano, Everly, and Dusek (1990) suggest that the excessive arousal can lead to an increase in the dendrites of the limbic system, which can further increase limbic stimulation. The limbic system might sprout more excitatory postsynaptic receptors and decrease its inhibitory presynaptic receptors. Subsequently, people with kindling may experience excitatory neurotoxicity. Brouwer and Packer (1994) have conducted research indicating that people with ME/CFS might have “unstable cortical excitability associated with sustained muscle activity resulting in varied magnitudes of descending volleys” (p. 1212). This is an indication that kindling might be occurring within the brains of patients with ME/CFS.

Two receptors residing on the cell surface membranes of neurons are GABA (gamma aminobutyric acid), which inhibits neuronal firing, and NMDA (N-methyl-D-aspartate), which excites neuronal firing. The GABA and NMDA receptors should be balanced, but after an injury or viral attack, NMDA fires more than GABA. Minor and Hunter (2002) have proposed that prolonged exposure to inescapable stressors will eventually deplete GABA, thus reducing an important form of inhibition on excitatory glutamate transmission. Doi, Ueda, Nagatomo, and Willmore (2009) studied hippocampal glutamate and GABA transporters (i.e., GLAST, GLT-1, and EAAC1) by injecting rats three times a week with pentylenetetrazol, which can induce kindling. Levels of EAAC1 and GAT-1 in easily-kindled rats were decreased by 30% compared to levels in rats that were resistant to being kindled. This suggests that decreased EAAC1 and GAT-1, which diminish GABA function, are associated with the convulsive threshold at the beginning of kindling development.

The importance of GABA to sleep difficulties, one of the primary problems of patients with ME/CFS, was highlighted in a recent study by Winkelman et al. (2008). This research team found brain GABA levels were nearly 30% lower in patients with primary insomnia and GABA levels were negatively correlated with wake after sleep onset. Recent findings indicate that the glial cell or astrocyte produces adenosine, which appears to be implicated in controlling wake to sleep transitions (Halassa et al., 2009). In response to cellular damage such as inflammation, concentrations of adenosine become quickly elevated from 300 nM to up to 600-1,200 nM, which probably promotes sleep and rest. When levels of adenosine increase, brain nerve impulses are suppressed. High adenosine levels can suppress epileptic seizures, and it appears that adenosine acts through the A1 receptor to produce sleep pressure. The adenosine A(2A) receptor (A(2A)R) plays a crucial role in the regulation of sleep, and in rats, Hong et al. (2005) found the A(2A)R agonist induced sleep by inhibiting the histaminergic system through increasing GABA release. It is possible that kindling could also result in low levels of adenosine over time, and thus the development of sleep difficulties. It should be noted that adenosine is only one of many substances that promotes sleep, and others include tumor necrosis factor, interleukin-1, interleukin-6, and growth releasing hormone, whereas sleep-inhibiting substances include corticotropin releasing hormone, substance P, and interleukin-10 (Krueger, 2009).

Ultimately, chronic stress sensitizes neural processes and this over-activation might lead to fatigue. The limbic system plays a regulatory role pertaining to symptoms of fatigue, pain, memory, and cognition. In part, it plays this role with the use of dopamine to inhibit NMDA receptor-mediated nociception (Wood et al., 2007). Chronic stress has been shown to attenuate dopaminergic activity (Wood, 2004), and this disruption in dopamine function might also lead to some of the cognitive dysfunction experienced by many ME/CFS patients (Nieoullon, 2002). It should be noted that one study did not find abnormalities in regional amino acid neurotransmitter function in patients with ME/CFS (Mathew et al., 2009).

Kindling can develop through a number of ways for patients with ME/CFS. Zalcman, Savina, and Wise (1999) have found that immunogenic stimuli can alter brain circuitry, changing its sensitivity to seemingly unrelated subsequent stimuli. In addition, stress might be a conditioned stimulus that leads to an impaired immune response (Gupta, 2002). Exposure to stress can induce long-term potentiation, such that the brain cells react more strongly in response to future exposures to a drug or stress (Saal et al., 2003).

Secondary lesions can also occur in the brain, which are not located at the site of the original kindling. These secondary sites are exposed to cortical excitability through normal synaptic pathways and intercellular communication. Therefore, different regions of the brain may become kindled in a secondary manner, separate from the initial kindling. These secondary sites could then affect different organ systems in a top-down fashion, leading to the diverse symptom patters found in patients with ME/CFS.

Gupta (2002), borrowing on the work from LeDoux (1996), has suggested that an infection, chemical or physiological stressor can act to create a cell assembly within the unconscious amygdala, which can create sympathetic stimulation through the hypothalamus and other brain pathways involving the flight or fight response. The amygdala first determines whether a stimuli poses a threat, and if so, then the amygdala initiates autonomic and endocrine responses to help the organism survive. Areas of the prefrontal cortex and anterior cingulate (which will be referred to below) are involved in attention to dangerous or negative stimuli, which ultimately influence the amygdala. In long-term potentiation, synaptic strength does increase between co-firing neurons after brief but repetitive stimulation, and this has many similarities to kindling. LeDoux (1996) refers to these cell assemblies as being particularly resistant to extinction; so for some patients this hard-wiring may only be regulated rather than extinguished. Gupta (2002) believes that activation of the amygdala causes continuous sympathetic stimulation that is a predominantly unconscious process over which patients have little control, but it eventually leads to mental and physical exhaustion as well as glandular depletion. Wyller, Eriksen, and Malterud (2009) proposed compatible theories that sustained arousal is the primary mechanism of CFS, and they also reviewed findings on how arousal responses can be modified by sensitization.

This kindling theory has been used to explain secondary generalized temporal lobe epilepsy, and Bell et al. (1997) have reviewed evidence that kindling is implicated in multiple chemical sensitivity disorders. This kindling theory could be extended to help explain the etiology of ME/CFS.

Oxidative Stress Theory

Martin Pall (2007) has suggested that oxidative stress might help explain the pathophysiology among patients with ME/CFS. According to Pall’s theory, when NMDA receptors on neurons are activated by a virus, bacteria, mold, toxin, microbe or allergy, they trigger nitric oxide production. In addition, the mitochondria in cells take in oxygen and nutrition and output carbon dioxide, water and ATP (energy). For every molecule of ATP generated, one molecule of superoxide is generated. The superoxide in the mitochondria sometimes leaks out, where it combines with nitric oxide. One molecule of nitric oxide combines with one molecule of superoxide to make one molecule of peroxynitrite. Peroxynitrite will break down to release hydroxyl radical, and that will cause genetic damage. Peroxynitrite and free radical formulation has been implicated in coronary artery disease, cancer, Parkinson’s and Alzheimer’s diseases, Multiple Sclerosis (MS), and autoimmune diseases. Peroxynitrite also acts to increase levels of both nitric oxide and superoxide, which then produce more peroxynitrite; thus producing a self-sustaining cycle. The enzyme superoxide dismutase, embedded in the mitochondria, can break superoxide down to hydrogen peroxide and then to water to prevent superoxide from leaking out of mitochondria, but this enzyme cannot do its job without proper amounts of selenium and glutathione (and the latter is depleted in patients with CFS, Van Konynenburg, 2007). During sleep, the brain produces more superoxide dismutase, and this works to neutralize the free radicals that have been generated during the day due to metabolism.

It is of interest that many patients with ME/CFS report that Coenzyme Q10 (CoQ10), Klonopin, and Neurontin are helpful in reducing symptoms. CoQ10 binds to excess superoxide so that it cannot couple with nitric oxide to produce peroxynitrite. Also of interest, Klonopin and Neurontin upregulates GABA and down regulates NMDA, which reduces nitric oxide. Women are reported to produce more nitric oxide than men (Forte et al. 1998), possibly contributing to the gender bias seen in ME/CFS. These findings support Pall’s (2007) theory, and suggest that oxidative stress might play an important role in ME/CFS. Kindling might be the mechanism to elicit excessive arousal and excitatory neurotoxicity, whereas oxidative stress might represent the end result of this kindling. If structural changes in the neurons have occurred, with a growth in excitatory postsynaptic receptors and a decrease in inhibitory presynaptic receptors, then it might be extremely difficult to recover from such an illness, which is confirmed by outcome studies indicating a poor prognosis for adult patients with ME/CFS (Friedberg & Jason, 1998).

There are several studies that have been published suggesting that oxidative stress does occur in patients with ME/CFS. For example, Kennedy et al. (2005) found that patients with ME/CFS had significantly increased levels of isoprostanes, and among patients who were not obese or hypertensive, ME/CFS symptoms correlated with isoprostane levels. In the group of normotensive and nonobese patients with ME/CFS, antioxidant activity, as measured by glutathione levels, was not reduced. This supports an interpretation that oxidative stress is due to excessive free radical formation and not depleted antioxidant reserves. Robinson et al. (2009) investigated the levels of IL-6, its soluble receptors (sIL-6R and sgp130) and F(2)-isoprostanes, at rest and during exercise, and they found that F(2)-isoprostanes were higher in CFS patients at rest and at 24 hours post-exercise.

Several animal studies have also provided supportive evidence. Kumar, Garg, and Kumar (2008) found that forced swimming for 7 days caused a chronic fatigue-like condition and oxidative damage in mice. Trazodone was administered each day 30 minutes before the forced swim test, and this pretreatment attenuated the oxidative damage. However, when L-arginine (a NO precursor) was administered 15 minutes before administration of trazodone, the L-arginine reversed the protective effect of trazodone. This study suggests involvement of the nitric oxide pathway in the neuroprotective potential of trazodone in a mouse model of ME/CFS. In another animal model, Gupta et al. (2009) found significant oxidative stress after immunological activation, and an antioxidant called curcumin significantly reduced both oxidative stress and serum tumor necrosis factor-alpha. Finally, Lyle et al. (2009) forced rats to swim in water for 15 minutes per day on 21 consecutive days, which lead to oxidative stress. Nardostachys jatamansi extract (NJE) given orally had an antioxidant effect as it tended to normalize augmented lipid peroxidation, nitrite, superoxide dismutase activities and catalase level.

In the sections below, we will examine how the kindling and oxidative stress theories might be helpful in better understanding findings from different domains within the ME/CFS literature.

Genetics

Gow et al. (2005) reported gene expression data suggesting that ME/CFS may involve ion transport and ion channel activity, which is necessary for the generation of action potentials and the release of neurotransmitters at synaptic terminals. Muscle fatigue and post-exertional malaise may relate to a shift of membrane hypopolarization potential (Chaudhuri, Behan, & Behan, 2005). Sustained changes in cell membrane function may follow exposure to infections and neurotoxins as when ciguatera toxin irreversibly inactivates sodium channels in an open mode, and these can cause delayed symptoms of chronic fatigue. Whistler, Jones, Unger, and Vernon (2005) examined differences in gene expression before and after exercise for those with ME/CFS and matched controls. They also found differences in ion transport and ion channel activity at baseline, and these differences were exaggerated after exercise. Saiki et al. (2008) in a study of CFS associated markers, used quantitative real-time polymerase chain reaction to validate nine genes encoding granzyme in activated T or natural killer cells (GZMA), energy regulators (ATP5J2, COX5B, and DBI), proteasome subunits (PSMA3 and PSMA4), putative protein kinase c inhibitor (HINT), GTPase (ARHC), and signal transducers and activators of transcription 5A (STAT5A).

Discrepant findings have occurred in the gene studies, and Light, White, Hughen, and Light (2009) suggest that this is due to the fact that it is only with exercise that reliable gene differences emerge between controls and patients, and the studies above did not employ an exercise challenge. In selecting genes to examine before and after exercise challenges, it is critical to focus on genes that contribute to the primary symptoms of CFS including fatigue, postexertional malaise, and muscle pain. Light et al. (2008) conducted mouse experiments and identified two classes of sensory neurons that were capable of sending signals interpreted as physical fatigue and muscle pain. They found the following molecular receptors detected the metabolites produced by muscle contraction: an acid sensing ion channel (also called amiloride sensitive ion channel) or ASIC (ASIC3), 2 Purinergic X type receptors (P2X5 and/or P2X4), and transient receptor potential vanilloid type 1 (TRPV1). Based on this mouse model, Light et al. (2008) suggested that increased expression of these molecular receptors encoding metabolites could be a marker of enhanced fatigue and/or muscle pain.

Light and colleagues (2009) also maintain that genes from the sympathetic nervous system have also been implicated in CFS. In response to the metabolite buildup in working muscles, the sympathetic nervous system regulates regional blood flow (e.g., b-Adrenergic receptors play a major role in maintaining sufficient blood flow to skeletal muscles during exercise preventing excessive accumulation of metabolites). In addition, genes from the immune system have also been implicated in CFS. Moving from animal models to studies with normal humans, Light and colleagues found that mRNA for metabolite detecting (ASIC3, P2X4, P2X5, TRPV1), sympathetic nervous system (adrenergic a-2A, b-1, and b-2, as well as COMT), and immune (IL6, IL10, TNF-a, TLR4, CD14) were upregulated after strenuous exercise (Light et al., 2007). In a later study, Light et al. (2009) examined patients with CFS versus controls after exercise. Light et al. found patients with CFS demonstrated increases after exercise that reliably exceeded responses of control subjects in mRNA for genes that can detect increases in muscle produced metabolites (ASIC3, P2X4, P2X5), genes that are essential for sympathetic nervous system processes (adrenergic a-2A, b-1, and b-2, as well as COMT), and immune function genes (IL10, and TLR4). Of interest, at baseline before the exercise challenge, there were no significant differences between the CFS and control groups. Significant correlations were found between post exercise physical and mental fatigue and increases in mRNA of the genes. Finally, approximately 90% of the CFS patients could be distinguished from control subjects using just 4 of the genes measured (i.e., P2X4, adrenergic b-1, adrenergic b-2, IL10). Light et al. (2009) concluded that ME/CFS patients might have enhanced sensory signal for fatigue that is increased after exercise. These findings all indicate persistent changes in cell membrane function, which are compatible with a kindling theory.

The situation might even be more complicated, as Kerr et al. (2008) clustered quantitative polymerase chain reaction data from patients with ME/CFS and found seven subtypes with distinct differences in clinical phenotypes. Still, these data do implicate a number of genes that are related to oxidative stress and changes in ion transport and ion channel activity that might be affected by kindling.

Infectious Factors

The onset of ME/CFS has sometimes been linked with the presence of an infection. For example, some cases of ME/CFS have been reported as following acute mononucleosis, Lyme disease, and Q fever (Komaroff, 2000). Certain viruses [e.g., HSV-1, HHV-6, Epstein Barr virus (EBV), and cytomegalovirus] may influence the relapsing and remitting pathogenesis of ME/CFS (Englebienne & Meirleir, 2002). Lombardi et al. (2009) identified xenobiotic murine retrovirus (XMRV), a gammaretrovirus associated with a subset of prostate cancer, in the blood of 67% of ME/CFS patients but only 3.7% in controls. Retroviruses like XMRV activate a number of other latent viruses like the EBV, and this could explain why so many different viruses have been associated with ME/CFS. Neurotropic viral infections that replicate within and subsequently damage the central nervous system could be responsible for the appearance of lesions and the presence of focal epileptiform seizure activity in an ME/CFS viral onset subgroup. Magnetic resonance (MR) studies of encephalopathy and encephalomyelitis associated with acute EBV infection have found T2 prolongation over gray and white matter, brain atrophy, and periventricular leukomalacia (Shian & Chi, 1996). A MR study examining a pediatric population of patients suffering from chronic EBV infection has shown evidence for the presence of lesions in the hippocampal region (Hausler et al. 2002). In some cases, cortical lesions caused by herpes viridae infections fade before MR documentation can take place. Lesions can then reappear under specific conditions of environmental stimuli, a process that fits well with the relapsing and remitting hypothesis of ME/CFS.

Hickie et al. (2006) followed up with people who had cases of mononucleosis (glandular fever), Q fever, and Ross River virus, respectively, who later met the criteria for ME/CFS. The authors found that the percentage who went on to have ME/CFS was the same for the three infectious diseases (11% at 6 months), suggesting that the reason these people develop ME/CFS is not associated with the particular pathogen, but rather with their host response. The syndrome was predicted largely by the severity of the acute illness rather than by demographic, psychological, or microbiological factors. In other words, it is the severity of the host response that determines the injury, and this would also be compatible with a kindling explanation. In the same cohort study, Vollmer-Conna et al. (2008) later found that individuals with high levels of IFN gamma and low levels of IL10 were significantly more likely to experience severe acute illness following infection and were more likely to be symptomatic for a longer time. IFN gamma is one of several pro-inflammatory cytokines, and IL10 is one of several anti-inflammatory cytokines.

White (2007) summarized findings from five cohort studies involving postinfectious illness. These studies indicate that a postinfectious fatigue syndrome does exist, and that it is not a mood disorder. It appears that there are two postinfectious fatigue syndromes, one characterized by excessive sleep and the other by insomnia associated with muscle and joint pain. The risk of prolonged fatigue or ME/CFS following postinfectious fatigue syndromes is five to six times that of common upper respiratory tract infections, and there is a 10-12% risk of ME/CFS six months after infectious onset. Wessely et al. (1995) found that some people with ME/CFS had viral infections, while others had other medical illnesses, before they developed ME/CFS. Therefore, a viral infection probably represents only one of several possible routes to the development of ME/CFS. Kindling theory would also support the notion that a variety of stressors (e.g., viruses, chemicals, injury) can lower the threshold for activation, which would foster more excitatory postsynaptic receptors and decrease inhibitory presynaptic receptors.

Immune System

The body’s reaction to one or more bacterial or viral invaders might induce symptoms in patients with ME/CFS. For example, activation by macrophages due to a virus or bacteria produces a release of interleukin-1beta, which causes an alteration in the electrical activity of the brain and a number of behavioral changes (e.g., decreases in activity and social interaction, somnolence) designed to reduce unnecessary energy expenditure, so that available energy stores can be used to fight the infection (Maier, Watkins, & Fleshner, 1994). These electrical changes in the brain can cause kindling.

Sheng, Hu, Ding, Chao, and Peterson (2001) injected an immunological stimulus that elicited a sustained upregulation of cytokines in the cerebral cortex and subcortical structures in a mouse; this coincided with marked reduction in running distance for two weeks. In humans, Vollmer-Conna et al. (2004) have found that the production of pro-inflammatory cytokines (IL-1b and IL6) was correlated with acute sickness behavior (i.e., fever, malaise, pain, fatigue, and poor concentration). Prolonged exposure to these cytokines might induce a state of chronic activation, which may lead to a depletion of the stress hormone axis and to other neuroendocrine features associated with ME/CFS. Under these circumstances, viruses and bacteria that had previously been contained and controlled by the immune system might begin to replicate and ultimately cause symptoms for the patient.

Many patients with ME/CFS appear to have two basic problems with immune function: a) poor cellular function, with low natural killer cell cytotoxicity and frequent immunoglobulin deficiencies (most often IgG1 and IgG3), and b) elevations of activated T lymphocytes, including cytotoxic T cells, and elevations of circulating cytokines (Evengard, Schachterle, & Komaroff, 1999; Patarca-Montero, Mark, Fletcher, & Klimas, 2000). Several researchers have found a shift from Th1 to Th2 cytokines among patients with ME/CFS (Antoni et al., 2003; Skowera et al., 2004). In the Th1 response, the T-helper cell produces pro-inflammatory cytokines, which activate T-cytotoxic cells as well as natural killer cells (Segerstrom & Miller, 2004), and contribute to clearance of intracellular pathogens. In contrast, the Th2 pathway involves major anti-inflammatory cytokines, which promote humoral immunity by differentiation of B cells into antibody-secreting B cells and B cell immunoglobulin switching to IgE. These anti-inflammatory cytokines inhibit production of pro-inflammatory cytokine and T-cell proliferation. Whereas a highly anti-inflammatory response minimizes inflammation, it can allow existing intracellular infections to linger. Of interest is the above mentioned finding by Vollmer-Conna et al. (2008), where those with high levels of a pro-inflammatory cytokine (IFN gamma) were significantly more likely to experience severe acute illness following infection.

The administration of a pro-inflammatory cytokine (Tumor Necrosis Factor-alpha) not only increases seizure activity in an animal kindling model, it may also potentiate the development of future activity (Shandra et al., 2002). Kindling and related seizure activity in animal models is associated with increased signal transcription and production of pro-inflammatory proteins within the brain (Plata-Salaman et al, 2000). Kindling then not only potentiates the production of pro-inflammatory mediators which can then contribute to the development of lesion, it is in turn influenced by them.

Borish et al. (1998) found evidence of low level inflammation, similar to that found in allergies, in a subgroup of individuals with ME/CFS. Borish et al. suggested that there might be two subgroups of individuals with ME/CFS, those with immune activation (infectious or inflammatory) and those devoid of immune activation, with other illness processes. It is possible that there are two groups of people with ME/CFS, those with kindling and oxidative stress and those without it, and that these might correspond to the two groups identified by Borish et al. (i.e., those with and without immune activation, respectively). Cook, Lange, DeLuca, and Natelson (2001) found that individuals with an abnormal MRI and ongoing inflammatory processes scored significantly worse on measures of physical disability. These findings support the idea that there might be important subtypes among patients with ME/CFS (Corradi, Jason, & Torres-Harding, 2006). It is at least possible that centrally mediated kindling and oxidative stress might lead to different types of immune, autonomic, or neuroendocrine dysfunction in different patients.

Neuroendocrinology

According to Baram and Hatalski (1998) summarize data indicating that considerable in vitro and in vivo data suggest that corticotropin-releasing hormone (CRH) induces neuronal excitability, which can lead to epileptic output. Within seconds of exposure to stress, CRH , located in the peptidergic neurons in the paraventricular hypothalamic nucleus (PVN), is released from nerve terminals to influence hormonal secretion from ACTH in the pituitary and glucocorticoid secretions in the adrenal. This is the mechanism by which fevers and trauma can activate CRH receptors in the hippocampus and amygdala to induce seizures among children and rats. CRH increases the frequency of spontaneous excitatory postsynaptic currents by 252% (Baram & Hatalski, 1998). Administration of CRH in rats produces seizures in the amygdala and epileptiform discharges in the hippocampus but the earliest CRH induced epileptiform discharges are produced in the amydgala and propagate to the hippocampus. There might also be a reciprocal relationship as limbic seizure kindling results in increased levels of CRH in the hippocampus (Heinrichs & Koob, 2004). Once the kindling has occurred, the CRH might play a less critical role in maintaining the kindling, and after time, CRH levels might even become depleted.

Scott and Dinan (1999) found that patients with ME/CFS tend to have a reduced adrenal secretory reserve and smaller adrenal glands when compared to healthy subjects; in contrast to major depression, enlarged adrenal glands are found. Altemus et al. (2001) found that patients with ME/CFS had a reduced ACTH response and a more rapid cortisol response to a vasopressin infusion, which suggests reduced hypothalamic CRH secretion. Glucocorticoids can have an inhibitory effect on serotonin function, and CRH release is modulated by serotonin. In addition, the increased prolactin response to fenfluramine is due to elevated activity of presynaptic serotonin neurons (Vassallo et al., 2001). This might be the region affected by kindling, and several other studies have also found evidence of increased serotonergic activity in patients with ME/CFS (Bakheit et al, 1992; Cleare et al., 1995; Demitrack et al., 1992; Sharpe et al., 1997). One pilot study found that medications that block serotonin (5-HT3) receptors were followed by at least a 35% improvement in about one-third of patients (Spath et al., 2000). Still, there may be subtypes of patients, as one study found decreased brain serotonin levels in patients with ME/CFS (Badawy et al., 2005). In addition, there are eight different Serotonin receptors, and this might explain differential outcomes of Serotnin altering therapies.

Neuroendocrine dysfunction in ME/CFS involve hypocortisolism and increased serotonin neurotransmitter function, whereas in depression, hypercortisolism and decreased serotonin neurotransmitter function have been found (Cleare et al., 1995). Also of interest are the findings that four minutes after a maximal treadmill exercise test for patients with ME/CFS, stress responsive hormones (adrenocorticotropin, catecholamines, prolactin) were at less than half the level of controls (Ottenweller, LaManca, Sisto, Guo & Natelson, 1997).

When stress occurs for weeks or months, and glucocorticoid levels are maintained at a high level for long periods of time, the immune system is suppressed. In addition to the suppression of inflammation by long term stress, the Th1 immune response is also suppressed. The immune response is then shifted to a Th2 immune response mechanism. Clauw and Chrousos (1997) suggest that individuals who develop ME/CFS may be genetically predisposed to development of the condition, and that hormonal changes in people with ME/CFS are primary, while immune changes are secondary. Clauw and Chrousos suggest that once the individual develops ME/CFS, which can occur abruptly or slowly through viral infections or emotional stressors, there is a blunting of the human stress response.

Chaudhuri and Behan (2004) speculate that there might be different neuroendocrine processes in subgroups of individuals that ultimately lead to chronic fatigue. For example, enhanced negative feedback of the HPA axis could account for alterations in HPA functioning in patients with ME/CFS. Fries, Hesse, Hellhammer, and Hellhammer (2005) suggest that an increased sensitivity to the negative feedback of circulating corticosterone contributes to the hyporeactive HPA axis under stressful conditions. In other words, they suggest enhanced pituitary feedback is the primary mechanism underlying the hypocortisolemic stress response. Increased sensitivity of lymphocytes to glucocorticoids might lead to the Th2 shift previously found in ME/CFS patients (Skowera et al., 2004). Kindling might also account for this increased sensitivity.

Ben-Zvi, Vernon, and Broderick (2009) posit that chronic stress will lead to depressed cortisol concentrations, and when stress is removed, the cortisol will stay at this depressed value. Glucocorticoids exert negative feedback at the hypothalamus and pituitary to inhibit secretion of CRH and ACTH. However, glucocorticoid negative feedback causes a reduction in corticotroph receptor expression, which leads to a desensitization of the pituitary to the stimulatory effect of CRH on ACTH release. Ben-Zvi et al. suggest that if cortisol is reduced, it will force a buildup of ACTH, and when ACTH increases about 30% above baseline, the body’s own natural feedback control will restore cortisol levels to normal. However, if kindling is the mechanism for the reduction of cortisol, such a strategy is unlikely to be effective.

Oxidative stress is linked with glucocorticoid resistance by affecting several aspects of glucocorticoid receptor (GR) activation and function including reduced GR nuclear transport (Okamoto et al., 1999), reduced GR transcription via decreases in histone deacetylase C2 (HDAC2) activity (Adcock et al., 2005), and decreased expression of glucocorticoid regulatory genes. HDAC2 activity, which is important in GR function, is decreased by tyrosine nitration leading to reduced steroid sensitivity. Antioxidant treatment partially restores dexamethasone sensitivity in reactive oxygen species (ROS)-exposed cells by increasing HDAC activity (Ito et al., 2001). Several redox-sensitive transcription factors that are activated by intracellular oxidative stress attenuate GR function in steroid resistant states (Marshall et al., 2000). For instance, ROS activate nuclear factor kB (Mercurio & Manning, 1999), which interferes with GR expression and function (Nissen & Yamamoto, 2000) and has been indicated as a possible mediator between psychosocial and oxidative stress (Bierhaus et al., 2003). Induction of oxidative stress involving genes and signaling molecules such as stress-activated protein kinases and heat shock proteins is associated with stress-related disorders (Schett et al., 2001). Conversely, epigenetic changes resulting from interventions involving meditative practices and the relaxation response lead to increased gene expression of glutathione S-transferase, increased intracellular levels of glutathione and heat shock protein 70 and enhanced activity of glutathione peroxidase and superoxide dismutase (Dusek et al., 2008; Sharma et al., 2008). These latter interventions are also linked with improvements in the pituitary-adrenal activity that could be mediated by an improved sensitivity of GR, which is very susceptible to changes in redox status.

When a person with a certain genetic makeup is subjected to long term stressors, the HPA axis and sympathetic nervous system become upregulated. Kindling might represent the mechanism for this upregulation. Elevated secretions of glucocorticoids and catecholamines (adrenalin and noradrenalin) may subsequently cause a Th1 to Th2 immune response shift. Because of the shift to Th2, the body does not have an effective defense against viral or intracellular bacterial infections (e.g., the Epstein Barr virus can become active and produce infections). Van Houdenhove, Van Den Eede, and Luyten (2009) suggest that at an early stage of the illness, a switch takes place from HPA axis hyper- to hypo-functioning, and this observation is supported by some animal and human data. When the HPA axis becomes downregulated, there is still not an effective Th1 response to attack viral infections, however, now the immune system may cause inflammation (explaining elevated antinuclear antibody levels). The patient with ME/CFS now has ineffective protection from viruses, intracellular bacteria, and inflammation (Van Konynenburg, 2003). In a review article, Van Den Eede, Moorkens, Van Houdenhove, Cosyns, and Claes (2007) concluded that even if the HPA axis dysfunction is not the primary factor, it is probably a relevant factor in ME/CFS symptom propagation. Kindling and oxidative stress could also help explain these findings, as the areas of the brain where the kindling occurs could determine what aspect of the HPA system is implicated in the disorder.

Autonomic Nervous System

Acetylcholine is a primary neurotransmitter of the parasympathetic nervous system and is widely distributed throughout the brain and spinal cord. Chaudhuri, Majeed, Dinan, and Behan (1997) believe that ME/CFS entails a depletion of acetylcholine and increased sensitivity of the post-synaptic acetylcholine receptors. This could cause sympathetic nervous system hyperactivity, which may decrease serum cortisol and may be the common denominator for low levels of DHEA in both inflammatory and non-inflammatory diseases (Kizildere, Gluck, Zietz, Scholmerich, & Straub, 2003). Norepinephrine and epinephrine inhibit the production of type 1/proinflammatory cytokines, whereas they stimulate the production of type 2/anti-inflammatory cytokines. This causes a selective suppression of Th1 responses and cellular immunity and a Th2 shift toward dominance of humoral immunity (Elenkov, Wilder, Chrousos, & Vizi, 2000). Again, neural kindling might be implicated in the depletion of this acetylcholine and increased sensitivity of the post-synaptic acetylcholine receptors.

Other aspects of the circulatory system also seem to be involved in ME/CFS. In response to postural stress, 81% of patients with ME/CFS and no controls experienced ejection fraction decreases, suggesting left ventricular dysfunction in the heart. Those who had greater ejection fraction decreases experienced more severe ME/CFS symptoms in their daily lives (Peckerman, Chemitiganti, et al., 2003). Patients with ME/CFS might have lower cardiac output. The resulting low flow circulatory state may make it difficult for patients to meet the demands of everyday activity and lead to fatigue or other symptoms (Peckerman, LaManca, et al., 2003). In addition, Streeten and Bell (2000) found that the majority of patients with ME/CFS had striking decreases in circulating blood volume, which might be implicated in orthostatic hypotension. Additionally, it appears that the blood vessels in patients with ME/CFS are constricted dramatically. The heart and circulatory system’s response to standing upright is under the control of the sympathetic nervous system (Van Konynenburg, 2003), which appears to be overactive in patients with ME/CFS. Pall (2007) suggests that lowered cortisol levels can produce cardiac dysfunction, and that lowered cortisol production during and following exercise may be implicated in the cardiac dysfunction seen in many patients. Kindling could also be implicated in the overactive sympathetic nervous system.

Many patients with ME/CFS have co-morbid Fibromyalgia, and this latter disorder might also be associated with autonomic nervous system problems. Martinez-Lavin and Solano (2009) speculate that within the dorsal root ganglia, after a trauma or inflammation due to infection, sympathetic neurons sprout and connect to pain sensing neurons, and the dorsal root ganglia become hyperexcitable to painful inputs. It is interesting to note how kindling might also be involved in this process involving widespread pain.

Neurology

Gray and Robinson (2007) have studied brain networks using a physiological-based model of the brain’s electrical activity. They contend that a steady state of electrical activity in the brain is present in the absence of stimuli. A stimulus will change this activity, and if the brain is stable, when the stimulus is removed the electrical activity will return to a steady state. However, the same stimulus in an unstable brain will lead to a continual increase in electrical activity that can result in neurological disorders.

Tanaka and Watanabe (2007) used animal models to suggest that initially, in an acutely stressed stage, the serotonin and dopamine systems are activated in the central nervous system, however, in the later stage of severe fatigue, reduced neuronal activities and energy utilization induced by prolonged deprivation of rest elicit central fatigue and insufficient activation of these systems in the brain. This is a complex model and implicates infectious antecedents for the ultimate damage occurring in the brain, although evidence reviewed earlier suggests that infectious agents might be just one of many precipitators of kindling.

Kuratsune and Watanabe (2007) believe that brain dysfunction among patients with ME/CFS is caused by abnormal production of cytokines, which may be caused by reactivation of various herpes viral infections and/or chronic mycoplasma infection. Increases in transforming growth factor- β (TGF- β) have been observed during infection and stress in rats (Inoue & Fushiki, 2007). The increase in TGF-β inhibits the production of DHEA-S, which is related to the dysmetabolism of acetyl-L-carnitine, leading to deterioration of biosynthesis of glutamate in the anterior cingulum. This might cause autonomic imbalance and prolonged fatigue. Also, reactivation of various viruses can cause abnormal production of IFN-gamma in the brain, leading to elevations of 5-HTT mRNA. This can cause 5-HT deficiency in the synapses, leading to fatigue, pain, and depression. Abnormal production of IFN-gamma also triggers elevation of 2′,5′-oligoadenylate synthetase activity which leads to an abnormality of RNase-L pathway and CNS dysfunction. Oxidative stress may also be implicated in these processes.

Fatigue is also one of the more frequent symptoms of Multiple Sclerosis (MS). In a sample of patients with MS, Sepulcre et al. (2009) found that fatigue was associated with a disruption of brain networks involved in cognitive/attentional processes. More specifically, fatigue correlated with lesions in the right parietotemporal (periatrial area, juxtaventricular white matter deep in the parietal lobe and callosal forceps) and left frontal (middle-anterior corpus callosum, anterior cingulum and centrum semiovale of the superior and middle frontal gyri) white matter regions. In addition, fatigue scores significantly correlated with gray matter atrophy in frontal regions, specifically, the left superior frontal gyrus and bilateral middle frontal gyri.

Cook et al. (2001) found individuals with CFS had a larger number of brain abnormalities than healthy controls, and Lange et al. (1999) also report small, punctuate, subcortical white matter hyperintersities in the frontal lobes. Johnson and DeLuca (2005) summarized structural neuroimaging studies (MRI) and concluded that abnormalities among patients with ME/CFS have been inconsistent, but when they have been observed, they have been in the subcortical white matter (punctate areas of high signal intensity). Higher abnormalities appear to occur among patients with ME/CFS who do not have concurrent psychopathology, versus those who have concurrent psychopathology (Lange et al., 1999).

Tirelli et al. (1998) documented glucose hypometabolism in the frontal cortex and brain stem. Siessmeier et al. (2003) evaluated cerebral glucose metabolism (using 18-fluorodeoxyglucose positron emission tomography). Abnormalities were detectable in approximately half the patients with ME/CFS, and consistent with other research on infectious encephalitis due to multiple types of viruses, no specific pattern could be identified (some had hypometabolism bilaterally in the cingulate gyrus and the adjacent mesial cortical areas, decreased metabolism in the orbitofrontal cortex, or hypometabolism in the cuneus/praecuneus). Nestadt et al. (2007) found that those with ME/CFS had ventricular lactate levels that were elevated compared to those with Generalized Anxiety Disorder and healthy controls. This suggests mitochondria dysfunction and/or anaerobic energy conversion in the brain. Nitric oxide regulates mitochondrial respiration and cell function by inhibiting cytochrome oxidase, and mouse models suggest that mitochondria dysfunction may be a mechanism by which kindling and oxidative stress are virally mediated, causing post infectious break down of the blood brain barrier (Komatsu et al., 1999). In anaerobic energy conversion, lactic acid is produced by the cells as they use glucose for energy in the absence of adequate oxygen. Yamamoto et al., (2004) using PET, found that the density of the 5-HTT of the rostral subdivision of the anterior cingulate cortex was significantly reduced in patients with ME/CFS. These findings suggest that an alteration in the serotoninergic neurons in the anterior cingulate cortex might play a role in the pathophysiology of ME/CFS. Cleare, Messa, Rabiner, and Grasby (2005) found widespread reduction in 5-HT(1A) receptor binding potential, and this was particularly marked in the hippocampus bilaterally, where a 23% reduction was observed. Cook, O’Connor, Lange, and Steffener (2007) found that participants with ME/CFS did not differ from controls for either finger tapping or auditory monitoring tasks, but exhibited significantly greater activity in several cortical and subcortical regions during a fatiguing cognitive task. More specifically, mental fatigue was significantly related to brain activity during the fatiguing cognitive task, and significant positive relationships were found for cerebellar, temporal, cingulate and frontal regions, while a significant negative relationship was found for the left posterior parietal cortex. De Lange et al. (2005) observed significant reductions in grey matter volume in patients with ME/CFS. Johnson and DeLuca (2005) concluded that functional neuroimaging studies among patients with ME/CFS generally show hypometabolism in the frontal lobes and ganglia. Clearly, both kindling and oxidative stress could be implicated in these findings. Abnormal findings in different regions of the brain may be due to kindling that occurs in a secondary manner, separate from the initial kindling. These secondary sites could then affect different parts of the brain.

Caseras et al. (2008) had patients with ME/CFS and healthy controls imagine fatigue-provoking events. Using fMRI, larger observed activity was found in the medial parietal cortex and precuneus in patients with ME/CFS compared with healthy controls during the fatigue-provocation task than a control task. De Lange, Knoop, Bleijenberg, and van der Meer (2008) suggest that larger activity in the precuneus may reflect the more vivid capability to imagine oneself in a fatiguing situation. Caseras et al. also found that lower cerebral activity was found in the dorsolateral prefrontal cortex, and this is the cortical area where Okada et al. (2004) have found reductions of grey matter among patients with ME/CFS. According to de Lange et al. (2008), the dorsolateral prefrontal cortex is essential for selecting and initiating behavior, and lesions in the lateral prefrontal cortex can reduce appropriate goal-directed voluntary behavior. Also of interest, de Lange et al. (2008) found that cognitive behavioral therapy could partly reverse the grey matter volume reduction in the lateral prefrontal cortex.

Billiot, Budzynski, and Andrasik (1997) found increased microvolt levels in lower frequencies (5-7 Hz) among patients with ME/CFS, and they suggested that the display of excess theta waves could be related to cognitive problems. Delta waves occurs from 0 to 4 Hz, theta from 4 to 8 Hz, alpha from 8 to 13 Hz and beta from 13 to 21 Hz. Peak alpha is the Hz value within the range of 8-12 Hz at which the most energy is generated. Theta-to-beta ratios were also calculated. The expected difference in the alpha range (8-12 Hz) was not found, but when counting backwards from 900 by 7s, the EEG microvolt activity was actually significantly lower than the non-patient comparison group. Research using low-resolution electromagnetic brain tomography by Sherlin et al. (2006) has found that twins with ME/CFS compared to their healthy co-twins had higher delta waves in the left uncus and parahippocampal gyrus and higher theta waves in the cingulate gyrus and right superior frontal gyrus. It appears that slowing of the deeper structures of the limbic system is associated with affect.

Donati, Fagioli, Komaroff, and Duffy (1994) examined quantified EEG (qEEG) data, which included standard EEGs and long latency evoked potentials (EP) with individuals diagnosed ME/CFS. Among 44% of the patients with ME/CFS, spike waves were observed compared to only 1.3% of all others (i.e., patients with depression who were medicated, patients with depression who were not medicated, as well as healthy controls). Spikes were most common in the temporal regions. Those in the ME/CFS group also had significantly more sharp waves, more frequent high amplitude alpha, and more frequent bursts of theta waves in the posterior regions. In the patients with ME/CFS, abnormalities were observed that involved high amplitude sharp alpha rhythm (10 Hz) that occurs in the occipital lobes upon closing the eyes. Also, discharges of the type one associates with epilepsy were seen in the temporal lobes. These are typically found after head injury and extreme sleep deprivation, also occurring after kindling. Temporal lobes have a predilection for infection by the herpes virus in acute herpes encephalopathy and encephalitis; therefore, the findings might be related to post-viral mild encephalopathy affecting primarily the temporal lobes, which could cause the self-reported memory and attention problems. In a later study, Duffy et al. (2009) found that factors derived from the EEG data were able to discriminate with nearly 90% accuracy patients with ME/CFS from healthy controls and from those with major depression.

Discussion

Findings reviewed in this article suggest that for some patients with ME/CFS, kindling might lead to oxidative stress through increased activation of NMDA receptors. Kindling might appear after prolonged stimulation of the limbic-hypothalamic-pituitary axis, either by high-intensity stimulation (e.g., brain trauma) or by chronically repeated low-intensity stimulation (e.g., an infectious illness). In support of kindling theory, there is evidence of changes in ion transport and ion channel activity (Gow et al., 2005). Several studies also suggest that oxidative stress does occur in patients with ME/CFS (Kennedy et al., 2005; Robinson et al., 2009).

Some individuals might be at higher risk of developing kindling and chronic activation, ultimately leading to oxidative stress. Vollmer-Conna et al. (2008) found that individuals with high levels of IFN-gamma (a pro-inflammatory cytokine) and low levels of IL10 (an anti-inflammatory cytokine) were significantly more likely to experience severe acute illness following infection. In addition, Glass et al. (2004) found that healthy individuals with certain biological patterns (i.e., lower cortisol, more heart rate variability, and NK attenuated response to stress) developed somatic symptoms when asked to stop exercising for a week. Individuals with diminished GABA functioning might also be more likely to develop kindling (Doi et al., 2009). These might be some of the predisposing neuroendocrine and immunologic irregularities of individuals who are at increased risk for developing ME/CFS. It would be of particular importance to study such high risk individuals in longitudinal designs.

Patients with ME/CFS often report prior exposure to a viral infection. Viruses increase activation of macrophages, which produce a release of interleukin-1beta, causing an alteration in the electrical activity of the brain (Maier, Watkins, & Fleshner, 1994). The production of pro-inflammatory cytokines (IL-1b and IL6) is correlated with acute sickness behavior (i.e., fever, malaise, pain, fatigue, and poor concentration; Vollmer-Conna et al., 2004), and prolonged exposure to these cytokines might induce a state of chronic activation and kindling. Kindling and related seizure activity in animal models is associated with increased signal transcription and production of pro-inflammatory proteins within the brain (Plata-Salaman et al., 2000), and the administration of TNF-alpha not only increases seizure activity in an animal kindling model, it may also potentiate the development of future activity (Shandra et al., 2002). Elevated levels of pro-inflammatory cytokines can also induce inducible nitric oxide synthase expression, which results in elevated levels of nitric oxide. The nitric oxide then reacts with superoxide to form the powerful oxidant peroxynitrite. For those individuals who are affected by this activation, oxidative stress might be a resultant consequence.

Viral exposure early in life could trigger an immunologic cascade with significant effects on kindling. The release of TNF-alpha and other mediators could contribute to immunologic sensitization through inflammation and corticosteroid mediation. This then might leave an individual primed to respond in an adverse fashion to a future stressor event through amygdala and hippocampal kindling. The response to a stressor event then might reintroduce an inflammatory response that could contribute to the development of lesions and symptomatology. This could help explain why viral exposure does not necessarily trigger immediate symptomatology. It is also more in line with the consensus opinion regarding MS, with early retroviral exposure and the development of disease later in life.

Areas of the prefrontal cortex and anterior cingulate influence the amygdala (Gupta, 2002), and kindling in these areas and others could cause continuous sympathetic stimulation that would eventually lead to mental and physical exhaustion as well as glandular depletion. Kindling could cause CRH to be released from the paraventricular nucleus of the hypothalamus. CRH causes ACTH to be released from the anterior pituitary, and ACTH in turn stimulates cortisol release from the adrenal cortex. Exposure to chronic stressors could eventually lead to hypocortisolism, which is frequently found among ME/CFS patients, along with increased serotonin neurotransmitter function (Cleare et al., 1995). Glucocorticoids can have an inhibitory effect on serotonin function, and CRH release is modulated by serotonin. There is evidence of elevated activity of presynaptic serotonin neurons (Vassallo et al., 2001), and this could be an area affected by kindling. Fries et al. (2005) suggest enhanced pituitary feedback is the primary mechanism underlying the hypocortisolemic stress response. Finally, chronic cortisol deficiency can cause an overproduction of interleukin-6 (Il-6), which has been associated with symptoms of ME/CFS (Arnold et al., 2002). Ejection fraction decreases and lower cardiac output (Peckerman, Chemitiganti, et al., 2003) could be due to lowered cortisol (Pall, 2007) as well as an overactive sympathetic nervous system.

As previously mentioned, patients with ME/CFS appear to have a shift from Th1 to Th2 cytokines (Antoni et al., 2003). In an acutely stressed stage, animal models indicated that at first serotonin and dopamine systems are activated in the central nervous system (Tanaka & Watanabe, 2007). If kindling were to occur for weeks or months, glucocorticoid levels would initially be maintained at a high level, and the immune system would be suppressed, which would suppress both inflammation and the Th1 immune response, with a shift to a Th2 immune response. Norepinephrine and epinephrine inhibit the production of type 1/proinflammatory cytokines, whereas they stimulate the production of type 2/anti-inflammatory cytokines thereby causing a selective suppression of Th1 responses and cellular immunity and a Th2 shift (Elenkov, Wilder, Chrousos, & Vizi, 2000). Because of the Th2 shift, the body would not have an effective defense against viral or intracellular bacterial infections. Eventually, the HPA axis might switch from HPA axis hyper- to hypofunctioning (Van Houdenhove, Van Den Eede, & Luyten, 2009). But when the HPA axis has become downregulated, there would still not be an effective Th1 response to attack the viral infection; however, now the immune system may cause inflammation, explaining elevated antinuclear antibody levels (Van Konynenburg, 2003).

In addition, patients with ME/CFS might have depletion of acetylcholine and increased sensitivity of the post-synaptic acetylcholine receptors (Chaudhuri et al., 1997). Kindling might also be implicated in this process and it would also cause sympathetic nervous system hyperactivity, which may ultimately result in decreased serum cortisol (Kizildere, Gluck, Zietz, Scholmerich, & Straub, 2003). Kindling can then be associated with variation in the responsiveness of the HPA axis (Weiss, Castillo & Fernandez, 1993) and provide a new lens through which to view the frequent finding of hypocortisolism in those with CFS.

According to Light et al. (2009), there are constant interactions between the sympathetic nervous system, the immune system, and the sensory systems in CFS. If patients with CFS have enhancement of peripheral sensory signals, due to kindling, and these signals likely activate sympathetic nervous system reflexes that normally maintain blood flow to the brain and skeletal muscles. Long-term sensory receptor activation can lead to sensitization of spinal cord and brain systems that transmit fatigue signals, causing long-term fatigue enhancement within the central nervous system (Cook, O’Connor, Lange, & Steffener 2007), Because vascular smooth muscle adrenergic receptors desensitize to the constant release of catecholamines (Kaufman & Hayes, 2002), this dysregulation could lead to bouts of increased metabolites that would further activate sensory receptors. If there was a large number of ASIC3, P2X5 and/or P2X4, and TRPV1 receptors on sensory neurons (Light et al., 2009), resting levels of metabolites could activate sensory fatigue afferents, but in particular, exercise could send a continuous signal of muscle sensory fatigue to the central sympathetic nervous system causing dysregulation of sympathetic nervous system reflexes, and ultimately producing the recognition of enhanced fatigue.

The oxidative stress theory of Pall (2007) implicates nitric oxide production to ultimately produce peroxynitrite, and there is some evidence for this model (Kennedy et al., 2005). There is also supportive evidence by de Lange et al. (2005) of significant reductions in grey matter volume in patients with ME/CFS. But there are conflicting studies specifying where the damage is occurring. Johnson and DeLuca (2005) summarize findings indicating that some patients have hypometabolism in the frontal lobes and ganglia. Siessmeier et al. (2003) found abnormalities in about half the patients with ME/CFS, but no specific pattern could be identified (some had hypometabolism bilaterally in the cingulate gyrus and the adjacent mesial cortical areas, decreased metabolism in the orbitofrontal cortex, or hypometabolism in the cuneus/praecuneus). Others have found alteration in the serotoninergic neurons in the anterior cingulate cortex (Yamamoto et al., 2004), widespread reduction in 5-HT(1A) receptor binding potential in the hippocampus bilaterally (Cleare, Messa, Rabiner, & Grasby, 2005), problems in cerebellar, temporal, cingulate and frontal regions (Cook, O’Connor, Lange, & Steffener, 2007), and higher delta waves in the left uncus and parahippocampal gyrus and higher theta waves in the cingulate gyrus and right superior frontal gyrus (Sherlin et al., 2006). Additionally, larger observed activity was found in the medial parietal cortex and precuneus in fatigue provoking tasks and lower cerebral activity in the dorsolateral prefrontal cortex (Caseras et al., 2008), and abnormalities involving high amplitude sharp alpha rhythm (10 Hz) that occurs in the occipital lobes, and epileptic-like discharges in the temporal lobes have been found (Donati, Fagioli, Komaroff, & Duffy, 1994).

These discrepant findings could be due to different regions of the brain becoming kindled in a secondary way from the initial kindling, and these secondary sites could then affect different organ systems and even lead to mitochondrial dysfunction (Myhill, Booth, & McLaren-Howard, 2009). These findings support the idea that there might be important subtypes among patients with ME/CFS (Corradi, Jason, & Torres-Harding, 2006). Ultimately, there is a need for investigators to develop a CFS research network that can assemble large carefully defined data sets involving either natural history studies, investigations with individuals exposed to challenges (i.e., exercise, orthostatic, mental), or pharmacologic or non-pharmacologic interventions. Such large data sets might allow us to ultimately identify important subtypes that remain within our samples. Comparing small samples with patients possibly having different characteristics (e.g., some studies have all or almost all patients with post-exertional malaise whereas others have fewer with this classic CFS symptom) might complicate the search for biological markers. Standardization of the procedures to collect these types of large data sets represents one of the largest challenges to the field.

One limitation in this review is that we have not evaluated the quality of the work reviewed. Given the large number of studies mentioned, it would have been extremely difficult to review strengths and weaknesses of each scientific article. By focusing on those articles that are most theoretically supportive of the two hypotheses of this paper, we hope to encourage researchers to take next steps and focus their research in these areas. We hope that the large number of studies cited provide supportive data that might provide a better lens for understanding this complex disorder. It is also unclear whether the proposed kindling and oxidative stress mechanisms might be operative in all cases of ME/CFS, or that are involved singly or in combination, in specific phenotypes only. Future research will be needed to explore this topic in more detail.

It is at least possible that centrally mediated kindling and oxidative stress might lead to different types of immune, autonomic, or neuroendocrine dysfunction in different patients. Such a theory might also have implications for treatment by helping patients with normalization of neuroendocrine-immune functioning (Van Houdenhove et al., 2009; Jason, Benton, Torres-Harding, & Muldowney, 2009), external stimulation in the form of a delayed feedback (Kim, Roberts, & Robinson, 2009), or in the use of pharmacologic drugs, which attenuates central sympathetic outflow (Wyller et al., 2009).