Hypothesis\n\nBaseline heart rate variability (HRV), specifically the time‑domain RMSSD index, predicts differential anxiety‑reducing efficacy of two Lactobacillus strains through distinct enteric‑vagal mechanisms.\n\n## Mechanistic rationale\n\n- L. plantarum produces γ‑aminobutyric acid (GABA) that activates GABA_B receptors on vagal afferent terminals in the gut wall, enhancing afferent firing and downstream nucleus tractus solitarius (NTS) activity, which in turn increases prefrontal‑amygdala inhibition and raises RMSSD【3】.\n- L. rhamnosus modulates the hypothalamic‑pituitary‑adrenal (HPA) axis via serotonergic and tryptophan metabolites that stimulate vagal afferents expressing 5‑HT₃ receptors, leading to vagal effent activation and HRV improvement independent of GABA signaling【4】.\n\nIndividuals with low baseline RMSSD have reduced vagal afferent sensitivity and therefore rely more on GABA‑mediated potentiation to restore vagal tone, whereas those with high baseline RMSSD already possess sufficient afferent responsiveness and benefit more from HPA‑modulating signals.\n\n## Testable predictions\n\n1. In a stratified RCT, participants with baseline RMSSD below the cohort median will show a significantly greater reduction in anxiety scores (e.g., GAD‑7) after 8 weeks of high‑dose L. plantarum (≥10¹¹ CFU/day) compared with placebo, while high‑RMSSD participants will not.\n2. Conversely, low‑RMSSD participants will show minimal anxiety change after L. rhamnosus (6×10¹⁰ CFU/day), whereas high‑RMSSD participants will exhibit a significant anxiety reduction.\n3. Changes in RMSSD over time will correlate with plasma GABA levels only in the L. plantarum low‑RMSSD subgroup, and with cortisol/ACTH levels only in the L. rhamnosus high‑RMSSD subgroup.\n\n## Falsifiability\n\nIf the interaction between baseline RMSSD, strain type, and anxiety outcome is absent (i.e., both strains improve anxiety equally across HRV strata, or no strain shows HRV‑dependent effects), the hypothesis is falsified.\n\n## Experimental design (brief)\n\n- Recruit 120 adults with mild‑to‑moderate anxiety (GAD‑7 ≥ 8).\n- Measure baseline RMSSD via 5‑min resting ECG.\n- Randomize to: L. plantarum, L. rhamnosus, or placebo (double‑blind).\n- Treat for 8 weeks; assess anxiety (GAD‑7, PSQI), HRV (RMSSD), plasma GABA, cortisol, and cytokines at weeks 0, 4, 8.\n- Use mixed‑effects models with baseline RMSSD, treatment, and their interaction as fixed effects.\n\n## Potential impact\n\nConfirming HRV as a predictive biomarker would enable precision psychobiotic therapy, matching strain to individual vagal phenotype and reducing trial‑and‑error in microbiome‑based mental‑health interventions.