Hypothesis

The rate at which glucose returns to baseline after a standardized mixed‑meal challenge, quantified as the postprandial glucose disposal rate (PGDR), combined with a time‑domain heart‑rate‑variability index (RMSSD) during the same window, will explain a significantly larger proportion of variance in clamp‑derived ΔRQ than traditional CGM variability metrics (MAGE, TIR) in non‑diabetic adults.

Mechanistic Rationale

PGDR reflects the integrated action of insulin‑mediated hepatic glucose suppression and peripheral muscle uptake during the absorptive state. Faster glucose clearance indicates greater hepatic insulin sensitivity and more efficient switching from fat to glucose oxidation, which is precisely what ΔRQ measures during a clamp. Heart‑rate‑variability, particularly RMSSD, indexes parasympathetic tone that modulates hepatic glycogenolysis and insulin signaling via the vagal‑cholinergic anti‑inflammatory pathway. When parasympathetic activity is high, insulin’s suppression of hepatic glucose production is amplified, sharpening the postprandial glucose decline. Thus, PGDR captures peripheral disposal while RMSSD captures the hepatic regulatory lever; together they should mirror the dynamic fuel‑shift capacity assessed by ΔRQ.

Testable Predictions

1. In a regression model, PGDR alone will predict ΔRQ with an R² comparable to HOMA‑IR (≈0.30).
2. Adding RMSSD to PGDR will increase the model’s R² by at least 0.15 relative to PGDR alone.
3. The combined PGDR + RMSSD model will outperform models using MAGE or TIR (ΔR² ≥ 0.10 improvement).
4. The predictive advantage will be strongest when the meal contains 50 % carbohydrate, 30 % fat, 20 % protein, a composition that reliably engages both hepatic and peripheral pathways.

Experimental Approach

Recruit 60 healthy adults (age 20‑40, BMI 18‑25 kg/m²) with normal fasting glucose. Each participant completes:

• A 2‑hour standardized mixed‑meal test while wearing a CGM (14‑day factory‑calibrated device) and a chest‑strap HRV recorder.
• Extraction of PGDR as the negative slope (mg/dL/min) from peak glucose to baseline (defined as pre‑meal + 2 SD).
• Computation of RMSSD over the same postprandial interval.
• Within one week, a hyperinsulinemic‑euglycemic clamp with indirect calorimetry to obtain ΔRQ (difference between basal and insulin‑stimulated RQ). Perform hierarchical linear regression: Model 1 (PGDR), Model 2 (PGDR + RMSSD), Model 3 (MAGE/TIR), Model 4 (HOMA‑IR). Compare adjusted R² and use likelihood‑ratio tests.

Expected Outcomes

If the hypothesis holds, Model 2 will show a statistically significant increase in explanatory power over Model 1 (p < 0.01) and will surpass both CGM‑variability‑only models and HOMA‑IR in predicting ΔRQ. This would validate a low‑cost, wearable‑derived surrogate for metabolic flexibility, redirecting biohacker focus from passive glucose traces to active challenge‑response kinetics coupled with autonomic indices. Failure to observe the predicted improvement would suggest that hepatic‑vagal interactions contribute less to postprandial glucose dynamics than anticipated, prompting refinement of the mechanistic model.

[1] https://diabetesjournals.org/care/article/49/Supplement_1/S150/163922/7-Diabetes-Technology-Standards-of-Care-in [2] https://www.intelmarketresearch.com/continuous-glucose-monitors-market-36291 [3] https://pubmed.ncbi.nlm.nih.gov/36570168/ [4] https://www.explorationpub.com/Journals/edd/Article/1005115 [5] https://pmc.ncbi.nlm.nih.gov/articles/PMC2584808/ [6] https://arxiv.org/html/2505.03784v1 [7] https://pmc.ncbi.nlm.nih.gov/articles/PMC5017919/ [8] https://pmc.ncbi.nlm.nih.gov/articles/PMC6093334/ [9] https://onlinelibrary.wiley.com/doi/10.1111/sms.70113 }