Adenosine Accumulation and Pathological Sleep Pressure
Adenosine is the primary homeostatic sleep signal: it accumulates during wakefulness and is cleared during sleep via A1 and A2A receptor-mediated processes. This section examines evidence that ME/CFS patients exhibit aberrant adenosine dynamics—elevated basal levels, impaired clearance, or heightened receptor sensitivity—producing a state of chronic excessive sleep pressure that manifests as unrefreshing sleep, daytime somnolence, and post-exertional fatigue amplification.
1 Adenosine as Metabolic Waste Signal
Adenosine is generated extracellularly through two converging routes: (1) enzymatic hydrolysis of released ATP and AMP via the ecto-nucleotidase cascade (CD39 \(\to\) CD73), and (2) intracellular export of adenosine formed from AMP dephosphorylation during periods of high energetic demand (Chang et al. 2021). In healthy tissue, extracellular adenosine remains low during wakefulness and accumulates progressively as a function of metabolic activity, creating the homeostatic sleep signal (Huang et al. 2024). In ME/CFS, the mitochondrial ATP synthesis deficiency documented in Chapters 6–8 would predictably elevate AMP/ADP ratios even at baseline, driving constitutive adenosine generation independent of the normal wakefulness duration signal.
2 Glial Adenosine Clearance and Its Failure in ME/CFS
Extracellular adenosine is cleared primarily by astrocyte-mediated uptake via equilibrative nucleoside transporters (ENT1 and ENT2) and enzymatic degradation by adenosine deaminase (Chang et al. 2021). Astrocytes thus act as the principal regulators of the extracellular adenosine tone that determines sleep pressure (Rábago-Monzón et al. 2025). Reactive astrogliosis, which is implicated in ME/CFS neuroinflammation (see Section Microglia Activation and Neuroinflammatory Fatigue), disrupts this clearance machinery: reactive astrocytes show altered ENT expression and adenosine deaminase activity, impairing the buffering of extracellular adenosine surges. The predicted consequence is a chronically elevated basal adenosine tone producing pathological sleep pressure even without extended prior wakefulness.
3 A2A Receptor Upregulation in Neuroinflammation
During microglial activation, A2A adenosine receptors are upregulated coincident with P2Y12 downregulation (Orr et al. 2009). This receptor shift has been termed chemotactic reversal: activated microglia respond to extracellular adenosine (the breakdown product of ATP) by retracting processes and adopting amoeboid morphology, rather than extending toward the signal as resting microglia do. In ME/CFS, Nakatomi et al. (Nakatomi et al. 2014) demonstrated 45–199% elevated TSPO binding in cingulate cortex, hippocampus, amygdala, thalamus, and brainstem (n=9 patients vs. n=10 controls), establishing sustained microglial activation as a feature of the disease. The hypothesis is that this persistent neuroinflammation maintains elevated A2A receptor density across the sleep-regulatory regions, enhancing the sensitivity of the system to adenosine and producing excessive sleep pressure responses to normal or modestly elevated adenosine concentrations. (Study quality: Medium; direct ME/CFS A2A receptor expression data absent; supported by mechanistic extrapolation from Nakatomi 2014 and Orr 2009.)
A2A receptor blockade using selective antagonists reduces neuroinflammation in mixed glial cell models, with A2A antagonists outperforming A1 agonists in anti-inflammatory and antioxidant efficacy (Navia et al. 2020). This mechanistic evidence supports the rationale for adenosine receptor-targeted interventions in inflammatory fatigue conditions, though no ME/CFS-specific clinical trials have been conducted.
4 Adenosine, Caffeine Sensitivity, and ADORA2A
Caffeine exerts its wake-promoting and fatigue-opposing effects exclusively through adenosine receptor blockade, with the A2A subtype being the primary target for sleep-regulatory effects (Huang et al. 2024). Individual sensitivity to caffeine is substantially determined by the ADORA2A c.1083T>C single nucleotide polymorphism: carriers of specific genotypes show caffeine-induced brain electrical changes that closely resemble insomnia, while other genotypes show minimal caffeine-sleep interaction (Rétey et al. 2007).
ME/CFS patients commonly report anomalous caffeine responses: either exaggerated sensitivity at low doses, or apparent inefficacy of caffeine at standard doses. If ME/CFS is associated with upregulated A2A receptor expression (as hypothesized from the neuroinflammation evidence A2A Receptor Upregulation Amplifies Sleep Pressure in ME/CFS Neuroinflammation), two opposing consequences are plausible. Heightened receptor density could increase binding sites for caffeine antagonism, potentially requiring higher doses for effect (apparent insensitivity). Conversely, if baseline adenosine is chronically elevated and receptors are tonically occupied, even low doses of caffeine may produce disproportionate displacement effects (exaggerated sensitivity). The tolerance and withdrawal phenomena are mechanistically consistent with adenosine-receptor adaptation to chronic caffeine antagonism: habitual caffeine use down-regulates adenosine receptors and produces withdrawal fatigue on discontinuation (Carbone et al. 2025). Two caveats bound the claim. First, in a large ME/CFS cohort (Nurses’ Health Study) caffeine showed an inverse association with severe fatigue but no association with ME/CFS itself (Palacios et al. 2023), so the caffeine-ME/CFS relationship is not established and may be null. Second, no ADORA2A pharmacogenetics study has been conducted in an ME/CFS cohort; this remains an untested prediction. An L-theanine-caffeine ADHD crossover showing the combination improved selective attention but not processing speed (Nawarathna et al. 2026) illustrates that caffeine’s attention effects can be selective (see the related dissociation discussion at Catecholamine vs Non-Dopaminergic Mechanisms: Attention Selection vs Processing Speed), but does not resolve the ME/CFS-specific receptor-density question. Falsifiability: falsified if ADORA2A genotyping or A2A-receptor-density measurement in an ME/CFS cohort shows no upregulation and no genotype–caffeine-response association. (Certainty: Low. No ME/CFS-specific adenosine receptor density or caffeine pharmacokinetics study found in literature search; caffeine-ME/CFS association null in one large cohort (Palacios et al. 2023).) Consequence: if confirmed, this would give clinicians a way to interpret patients’ idiosyncratic caffeine responses as a receptor-density signal rather than random variability — but the caffeine-ME/CFS relationship is currently null in the largest cohort, so this remains a speculative decoding framework, not a validated clinical test.
5 Therapeutic Implications: A2A Antagonism and Adenosine Modulation
The mechanistic model developed in this section suggests two therapeutic targets: (1) reducing adenosine overproduction by addressing its mitochondrial and neuroinflammatory drivers, and (2) modulating A2A receptor signaling directly. Non-selective adenosine receptor antagonists including theophylline and caffeine block both A1 and A2A receptors. A2A-selective antagonism is emerging as a neuroinflammation strategy with anti-inflammatory and antioxidant properties demonstrated in glial models (Navia et al. 2020).
Theophylline, a non-selective adenosine receptor antagonist with A1 and A2A activity, has been proposed clinically for ME/CFS given its profile of reducing adenosine-mediated sleep pressure and its established use in orthostatic intolerance via adenosine-mediated vascular tone effects. However, no published controlled trial of theophylline specifically targeting adenosine dysregulation in ME/CFS has been identified. Any clinical use should be considered highly experimental, with no evidence base beyond mechanistic plausibility. (Certainty: Very Low — mechanistic rationale only, no ME/CFS clinical evidence.)
6 Sleep Architecture Findings and Adenosine Interpretation
Systematic review of polysomnographic studies in ME/CFS identifies elevated microarousal index as the single most consistent objective abnormality: all five studies measuring microarousal index found significantly elevated values in ME/CFS patients vs. healthy controls, while 13 studies found no difference in sleep onset latency (Maksoud et al. 2021). The pattern — normal sleep initiation but pathological sleep fragmentation — is mechanistically consistent with an adenosine dysregulation hypothesis. Sleep onset depends on adenosine reaching a threshold level (preserved in ME/CFS, since sleep onset latency is normal), while sleep maintenance depends on continued adenosine-mediated suppression of arousal circuits across the night. Heightened A2A receptor sensitivity or elevated basal adenosine tone may paradoxically impair this maintenance function by saturating receptors and triggering compensatory arousal responses.
The subjective–objective discrepancy (91% of patients report non-restorative sleep despite near-normal aggregate polysomnography scores) may reflect microarchitectural disruption not captured by standard sleep staging: elevated microarousals interrupt slow-wave sleep restorative function without prolonging sleep onset or dramatically altering sleep stage proportions.
7 The Nap Paradox as Adenosine Illustration
The failure of naps in ME/CFS encapsulates the adenosine dysregulation hypothesis. In healthy fatigue, a 20-minute nap clears accumulated adenosine and restores alertness. In ME/CFS, naps fail to restore because adenosine clearance is futile: the metabolic deficit (impaired mitochondrial ATP production \(\to\) elevated AMP/ADP ratios \(\to\) constitutive adenosine generation) regenerates sleep pressure immediately after any transient clearance (Gotts et al. 2015). The underlying fuel deficit persists regardless of rest duration. Additionally, alpha-delta intrusion (Section Sleep Architecture Failure Hypothesis) operates during daytime naps as well as overnight sleep, preventing even brief naps from entering the restorative sleep stages where adenosine-mediated restoration would occur. This adenosine clearance deficit is one manifestation of a broader failure of physiological coordination during sleep, validated by SleepFM’s finding that cross-modal decoupling predicts disease onset (Thapa et al. 2026) (see Section Post-Exertional Malaise May Involve Inflammation-Induced Routing Disruption of Brain Clearance). The nap paradox illustrates a broader principle: in ME/CFS, normal recovery mechanisms exist but run on an energy substrate that is insufficient. Adenosine clearance works; adenosine regenerates immediately. Sleep spindles can theoretically be generated; thalamic circuits lack the metabolic support to sustain them. Glymphatic clearance activates during delta sleep; delta is contaminated by alpha intrusion. Each repair system is intact in principle and broken in practice. See Section Unrefreshing Sleep for clinical implications.