Gut–Brain–Immune Axis
Gut microbiome alterations in ME/CFS (Gastrointestinal and Microbiome Dysfunction) affect both immune function and CNS signaling through multiple pathways. The integrated model includes three gut-derived signals:
- Short-chain fatty acids (SCFAs): Butyrate, produced by commensal bacteria, maintains intestinal barrier integrity and exerts anti-inflammatory effects (Folkerts et al. 2020). Reduced butyrate-producing bacteria in ME/CFS (Hsu et al. 2025) is modeled as decreased \([\text{butyrate}]\), which increases intestinal permeability and permits translocation of bacterial products (lipopolysaccharide, LPS) into the circulation.
- LPS-driven immune activation: Translocated LPS activates monocytes/macrophages through TLR4, increasing pro-inflammatory cytokine production. This is an additional input to the cytokine network model (cytokine general).
- Microbiome-derived neurotransmitter precursors: Gut bacteria produce tryptophan metabolites, GABA, and other neuroactive compounds that reach the CNS via the vagus nerve or systemic circulation, modulating the neurotransmitter models in Neuroendocrine and Autonomic Models.
The gut–brain–immune axis coupling is modeled as:
\[ J_\text{LPS} = k_\text{perm} \cdot \frac{K_\text{butyrate}}{K_\text{butyrate} + [\text{butyrate}]} \cdot [\text{LPS}_\text{gut}] \tag{1}\]
where intestinal permeability increases as butyrate decreases, allowing more LPS translocation. The circulating LPS then drives immune activation, which via the neuroimmune pathway affects CNS function. This chain provides a mechanistic link between gut dysbiosis and neurological symptoms in ME/CFS.
1 GI Motility and SIBO Dynamics
GI dysmotility is a core component of the Septad framework (Prospective Phenotyping as Harm Reduction). Gastric emptying scintigraphy studies report delayed emptying in approximately 60% of tested ME/CFS patients (Vasquez-Marti et al. 2023), while breath testing identifies SIBO in 48% of those evaluated (Karhu et al. 2023). Gastroparesis and impaired migrating motor complex (MMC) function predispose to small intestinal bacterial overgrowth (SIBO), which in turn drives immune activation and nutrient malabsorption (Gastrointestinal and Microbiome Dysfunction, Gastrointestinal Dysfunction). The preceding gut–brain–immune model captures dysbiosis \(->\) permeability \(->\) LPS translocation but omits motility entirely. This subsection extends the gut compartment to formalize the autonomic \(->\) motility \(->\) SIBO \(->\) immune \(+\) energy cascade.
The model tracks three additional state variables: gut motility index \(\mathcal{G} \in [0,1]\) (where 1 is normal MMC cycling and 0 is complete gastroparesis), small intestinal bacterial load \(B_\text{SI}\) (colony-forming units, normalized to healthy baseline \(B_{\text{SI,0}}\)), and nutrient absorption efficiency \(\eta \in [0,1]\):
\[ \begin{aligned} \frac{d \mathcal{G}}{d t} &= \frac{1}{\tau_{\mathcal{G}}} [\mathcal{G}_\text{set}(V, \text{MC}_d) + \mathcal{G}_\text{pro} - \mathcal{G}] \\ \frac{d B_\text{SI}}{d t} &= r_B \cdot B_\text{SI} (1 - \frac{B_\text{SI}}{K_B(\mathcal{G})}) - \delta_\text{MMC} \cdot \mathcal{G}^2 \cdot B_\text{SI} - \delta_\text{Abx} \cdot B_\text{SI} \\ \frac{d \eta}{d t} &= \frac{1}{\tau_\eta} [\eta_0 \cdot \frac{K_\eta}{K_\eta + B_\text{SI} \\/ B_{\text{SI,0}}} - \eta] \end{aligned} \tag{2}\]
In gi motility, \(\mathcal{G}_\text{pro} >= 0\) represents the prokinetic drug effect (e.g., prucalopride, low-dose erythromycin), which directly augments motility independently of vagal tone; \(\mathcal{G}_\text{pro} = 0\) in the absence of prokinetic treatment. The motility set point \(\mathcal{G}_\text{set}\) depends on vagal efferent tone \(V\) (from the ANS model, ans balance) and mast cell degranulation state \(\text{MC}_d\) (from the mast cell model, mast cell dynamics):
\[ \mathcal{G}_\text{set}(V, \text{MC}_d) = underbrace(\frac{V^{n_V}}{K_V^{n_V} + V^{n_V}}, \text{vagal drive}) \cdot underbrace(\frac{K_{\text{MC,gut}}}{K_{\text{MC,gut}} + \text{MC}_d}, \text{mast cell inhibition}) \tag{3}\]
Vagal tone drives motility through the enteric nervous system: reduced vagal output (as in ME/CFS autonomic dysfunction) lowers the motility set point, impairing MMC cycling. Simultaneously, degranulated mast cells in the gut mucosa release histamine and prostaglandins that can disrupt coordinated peristalsis (Gastrointestinal and Microbiome Dysfunction). The Hill coefficient \(n_V\) determines the steepness of the vagal drive—a steep response means that modest vagal impairment has disproportionate effects on motility, consistent with the clinical observation that many ME/CFS patients develop gastroparesis despite only moderate autonomic dysfunction.
In the SIBO dynamics (gi motility), bacterial growth follows logistic kinetics with a carrying capacity \(K_B(\mathcal{G})\) that increases as motility decreases:
\[ K_B(\mathcal{G}) = K_{B,\text{min}} + (K_{B,\text{max}} - K_{B,\text{min}}) \cdot (1 - \mathcal{G}) \tag{4}\]
The MMC clearance term \(\delta_\text{MMC} \cdot \mathcal{G}^2\) is quadratic in motility, reflecting the fact that effective bacterial sweep requires coordinated phase III contractions—partial motility provides disproportionately less clearance. The antibiotic term \(\delta_\text{Abx}\) represents rifaximin or herbal antimicrobial treatment (Gastrointestinal Dysfunction).
1.1 Cross-System Coupling
The GI motility model couples to four other subsystems:
SIBO \(->\) Immune activation: Bacterial overgrowth increases luminal LPS concentration. The existing LPS translocation equation (gut immune) is extended with a SIBO-dependent source term: \[ [\text{LPS}_\text{gut}] = [\text{LPS}_{\text{gut,0}}] \cdot \frac{B_\text{SI}}{B_{\text{SI,0}}} \tag{5}\] This creates a positive feedback loop: autonomic dysfunction \(->\) reduced motility \(->\) SIBO \(->\) increased LPS \(->\) immune activation \(->\) sickness behavior \(->\) further autonomic impairment.
Malabsorption \(->\) Energy metabolism: Reduced nutrient absorption impairs substrate availability for mitochondrial respiration. The absorption efficiency \(\eta\) modifies the substrate input to the energy model: \[ J_\text{substrate} = \eta \cdot J_{\text{substrate,0}} \tag{6}\] where \(J_{\text{substrate,0}}\) is the dietary substrate flux at full absorption. This couples to the energy metabolism model (Energy Metabolism Models) by reducing the maximal glycolytic and fatty acid oxidation rates proportionally. SIBO-mediated fat-soluble vitamin malabsorption (via bile salt deconjugation) and bacterial consumption of B12 are both captured through \(\eta\) without requiring individual nutrient tracking.
SIBO \(->\) Butyrate depletion: The preceding model treated \([\text{butyrate}]\) as an exogenous parameter. Bacterial overgrowth in the small intestine competes with colonic commensals for fermentable substrates, making butyrate production endogenous. This replaces the static treatment with an explicit ODE: \[ \frac{d[\text{butyrate}]}{d t} = \sigma_\text{butyrate} \cdot \frac{K_\text{SIBO}}{K_\text{SIBO} + B_\text{SI} \\/ B_{\text{SI,0}}} - \delta_\text{butyrate} \cdot [\text{butyrate}] \tag{7}\] where \(\sigma_\text{butyrate}\) is the colonic production rate and the inhibitory Hill term captures substrate competition from SIBO. This links SIBO to gut barrier integrity via gut immune: SIBO \(->\) reduced butyrate \(->\) increased permeability \(->\) more LPS translocation, creating a second positive feedback loop that amplifies the immune consequences of dysmotility.
Mast cell \(<->\) Motility bidirectional loop: Histamine released by degranulated mast cells impairs motility (motility setpoint), while SIBO-derived LPS and bacterial metabolites can trigger further mast cell activation through TLR signaling. This is captured by the existing mast cell priming equation (mast cell dynamics) receiving LPS as an additional trigger input, creating a self-amplifying MCAS–dysmotility–SIBO triangle.
1.2 Treatment Predictions
The GI motility model generates testable treatment predictions:
- Prokinetics (prucalopride, low-dose erythromycin): increase effective motility via \(\mathcal{G}_\text{pro}\) (gi motility), bypassing the vagal impairment. The model predicts that prokinetics should reduce SIBO recurrence rate by maintaining MMC clearance (\(\delta_\text{MMC} \cdot \mathcal{G}^2\) term), consistent with established clinical practice of continuing prokinetics after SIBO eradication to prevent recurrence.
- Rifaximin: increases \(\delta_\text{Abx}\), rapidly reducing \(B_\text{SI}\). The model predicts transient improvement followed by relapse if motility remains impaired (\(\mathcal{G} < \mathcal{G}_\text{crit}\)), because the carrying capacity \(K_B(\mathcal{G})\) remains elevated. This formalizes the clinical observation that SIBO frequently recurs after antibiotic treatment unless underlying motility dysfunction is addressed.
- Mast cell stabilizers + prokinetics: the model predicts synergy because stabilizers reduce \(\text{MC}_d\) (improving motility set point) while prokinetics directly increase \(\mathcal{G}\), attacking the dysmotility from both upstream drivers simultaneously.
- Vagal nerve stimulation: increases \(V\) in the ANS model, which propagates to \(\mathcal{G}_\text{set}\) through motility setpoint. The model predicts that vagal stimulation should improve GI motility as a secondary benefit beyond its anti-inflammatory effects via the cholinergic pathway (vagal coupling).