Enter the 6-digit PIN.
UARS phenotype Β· epiglottic/laryngeal-dominant Β· MAD partial responder Β· predicted Oura / SnoreLab / v3 deltas per stack
Each toggle adds its expected effect on top of baseline. Effects are mid-range estimates; 95% CI bands shown in the grid below. All predictions are hypotheses translated from RCT/MA effect sizes into consumer-metric space β Oura restless is a proprietary proxy, not PSG-validated. Use these as testable predictions against your own pipeline, not promises.
Raw deltas from INTERVENTION-STACK.md Β§7. Ranges = 95% CI or qualitative confidence. Tier color shows evidence strength.
| Stack | ΞRestless | ΞDeep % | ΞHRV (ms) | ΞLowest HR | ΞSnoreLab | Ξv3 (Ο) | Tier |
|---|
Not PSG-validated against arousal index. Oura publishes 79% stage-classification agreement with PSG but no published mapping of restless periods β AI. Changes <30/night within noise; >50/night meaningful.
Mechanism-based mapping: MFT's 20β30% arousal index reduction in combined MAs should translate proportionally if Oura restless tracks arousal index. PΓ©pin 2025 dose-response for respiratory effort is the most reliable anchor for MAD titration.
Oura deep-sleep detection has 79.5% sensitivity vs PSG. Trend reliable, absolutes less so. Target 18%+ achievable if fragmentation resolves.
Expected gains are modest because deep-sleep is ceiling-bounded (~20%). Stacking typically yields 1β4 percentage points at most without CPAP.
Conceptual link to arousal burden is strong (lower sympathetic activity β higher HRV) but no direct Oura-HRV vs PSG-arousal-index validation published. Robust RERA treatment should lift HRV by 20-40%.
Acoustic OSA diagnosis AUC 0.94 (meta). Reasonable proxy for OSA severity trend. But silent RERAs don't show here β Pierre's phenotype is partly below SnoreLab's detection threshold.
No PSG validation. Research metric derived from envelope-modulation 3-10 Hz palatal-flutter band, IBI CV, crescendo-silence arousal pattern. Within-patient trend only.
Generated 2026-04-16 Β· companion to pipelines/panthera-titration/INTERVENTION-STACK.md Β· not medical advice Β· use as decision-support not prescription.