Sonic Balance™ Pilot Study Hypothesis
By Duane Czyzewski, Founder — Sonic Balance™ Neuroacoustic Therapy
Sonic Balance™ Pilot Study Hypothesis
Draft framework for internal review — not yet submitted for ethics review or peer input
This document translates the three-tier Mechanobiology Model into a testable research plan. It's scoped deliberately narrow: a single-practitioner pilot can establish whether the personalization mechanism itself is real and reliable (Tier 2), which is the load-bearing claim underneath everything else in the model. It is not scoped to test Tier 3 (cellular/gene-expression effects) — that would require lab partnership, biomarker assay capability, and a larger sample than one practice can generate. Being explicit about that boundary up front is what keeps this pilot honest and fundable later, rather than overreaching now.
---
1. Background
The Mechanobiology Model established a hypothesis (Tier 3) that individualized vibroacoustic stimulation, guided by real-time HRV feedback, may connect autonomic-level regulation to downstream cellular mechanotransduction effects documented in the broader literature. That hypothesis cannot be tested directly yet. What can be tested now, with existing equipment and client roster, is the foundational claim it depends on:
> That an individual has a measurable, reproducible HRV-Derived Resonant Response Frequency — and that stimulating at that frequency produces a different autonomic and subjective outcome than stimulating at a non-personalized frequency.
If that foundational claim doesn't hold up under controlled measurement, the rest of the model has nothing to stand on. If it does hold up, it becomes a legitimate, citable basis for pursuing the more ambitious cellular-level research later — and a much stronger position to approach a research or funding partner from than the current single-client documentation allows.
---
2. Central Research Question
Does an individualized, HRV-identified resonant frequency produce a measurably different autonomic and subjective response than a non-personalized (sham or generic) frequency, and is that individual frequency stable across repeated sessions?
This is really two linked questions — reliability (is the frequency stable within a person over time?) and specificity (does that frequency do something a randomly chosen or generic one doesn't?). Both need to be answered before the personalization claim can be called established rather than plausible.
---
3. Hypotheses
H1 (Reliability): An individual's HRV-Derived Resonant Response Frequency, identified via the frequency-sweep crossover method, will show significant test-retest stability across sessions conducted under similar conditions (accounting for known drift on high-load or compromised days, as already observed in the Artie/Olimpia session data).
H2 (Specificity): Sessions delivered at an individual's identified resonant frequency will produce a significantly greater shift toward autonomic balance (dual-axis SNS/PNS crossover, not sum) than sessions delivered at a matched-duration sham frequency for the same individual.
H3 (Subjective correlation, exploratory): Client-reported subjective state (VAS Stressed/Relaxed subscales, cross-checked periodically against STAI-6) will correlate with objective HRV shift more consistently under corrected measurement (post beat-correction) than has been observed in the currently flagged subjective-objective divergence pattern — testing whether that divergence is a genuine physiological/experiential dissociation or partly a measurement artifact.
These are stated so that a negative result on H1 or H2 is a real, informative finding — not something the design quietly can't detect. A pilot that can't fail isn't a pilot.
---
4. Design
Type: Single-site, within-subject, repeated-measures pilot. Each client serves as their own control — this fits the current one-practitioner roster and avoids needing a large between-subjects sample to detect an effect.
Arms (within-subject, order randomized per client):
· Personalized condition — stimulation delivered at that client's HRV-identified resonant frequency, determined via the existing Settling → Frequency Sweep → Treatment protocol.
· Sham/control condition — stimulation delivered at a frequency selected without reference to that individual's crossover data (e.g., a fixed reference tone, or another client's resonant frequency, matched for duration and volume/amplitude).
Randomizing which condition a given client receives first, across sessions, controls for order effects and habituation — the same logic already flagged in the protocol notes about avoiding monotonic frequency sweeps that confound frequency with elapsed session time.
Blinding: Full blinding isn't achievable (frequency content may be perceptible), but the client can be kept naive to which condition is "personalized" vs. "sham" on a given visit, and outcome scoring should be done by someone not involved in real-time session delivery where feasible — even partial blinding meaningfully reduces expectation-driven bias in the subjective measures.
---
5. Participants
Existing client roster plus ongoing self-sessions, expanded as feasible. Given known small-N constraints, this is explicitly a pilot for feasibility and effect-size estimation — not a powered confirmatory study. Its job is to determine whether the effect is large and consistent enough to justify a bigger, external study, and to surface design problems (measurement, protocol, dropout) before scaling.
Consent: Participants should be informed this is research-track data collection distinct from their regular clinical sessions, with a plain-language description of what's being tested and that no diagnostic or treatment claims are being made. This is worth formalizing as a written consent process even at pilot scale — it protects both the client and the eventual credibility of the data if this moves toward publication or partnership.
---
6. Measures
Primary:
· RMSSD, SNS Index, PNS Index — derived only from beat-corrected RR data (Kubios reanalysis or equivalent artifact correction applied before any metric is calculated; the previously documented RMSSD inflation issue makes this non-negotiable for research-grade data, even though it's already standard practice going forward for clinical reports).
· Dual-axis crossover status (both SNS and PNS independently near zero) as the operational definition of equilibrium — not the SNS+PNS sum, per the existing defect-class note.
Secondary:
· Client subjective state report, via two paired instruments rather than one:
· Visual Analogue Scale (VAS), Stressed/Relaxed subscales — administered every session, immediately before and after stimulation. Two sliding-scale marks (0–100), roughly 15 seconds to complete. This is the same instrument used in published HRV-biofeedback research (e.g., VR-supported HRV-BF studies pairing VAS with SDNN change), which makes it a reasonable basis for comparing Sonic Balance's effect size against that literature later. A two-item VAS alone isn't psychometrically robust enough to carry H3 on its own, which is why it's paired below.
· STAI-6 (six-item State-Trait Anxiety Inventory, state form) — administered periodically (e.g., every third or fourth session) rather than every session, to keep per-session client burden low while still providing a validated check that the VAS trend reflects something real. Six items, 1–4 scale, ~60–90 seconds; correlates above .90 with the full 20-item STAI and has been independently re-validated multiple times as the standard short-form option for repeated-measures anxiety/relaxation research.
· Session duration to crossover, and stability of crossover once reached.
Exploratory / future phase (not in this pilot):
· Salivary cortisol or inflammatory marker sampling pre/post session — the most tractable bridge toward the Tier 3 hypothesis, but requires lab partnership and adds cost/complexity not appropriate for this initial pilot.
· Simultaneous EKG + respiration monitoring (Vernier GDX-EKG and respiration belt) once integrated, to add real-time coherence detection as a secondary outcome measure and help resolve the SD1/SD2 inversion documentation issue.
---
7. Data Integrity Requirements
Carried forward from existing protocol learnings, restated here because a research dataset has less tolerance for these than routine clinical documentation:
· All RMSSD and derived indices calculated only after RR-level artifact correction — never from raw or post-hoc filtered data.
· No substitution of 0.0 for missing values; missing data flagged and excluded, not treated as baseline.
· Frequency sweep order randomized/counterbalanced per session, not monotonic.
· Minimum dwell time per frequency step set according to literature-grounded RMSSD estimation windows, not shortened for session-time convenience.
· Track loop-boundary transitions explicitly as a potential confound and log them, given the prior observation that loop positions may correlate with ANS state shifts independent of frequency content.
· Electrode gel and secured strap placement standardized across sessions to minimize artifact-rate variance between conditions.
---
8. Analysis Plan (pilot-appropriate)
Given small N, this should prioritize descriptive and effect-size reporting over significance testing — the goal at this stage is "is there a signal worth pursuing," not "is this proven." Suggested approach:
· Test-retest reliability of resonant frequency (H1): intraclass correlation or simple session-to-session variance, with high/compromised-load days flagged as a moderating variable rather than excluded.
· Personalized vs. sham comparison (H2): within-subject paired comparison of crossover-quality metrics between conditions, reported with effect size and confidence intervals rather than a bare p-value, given the sample size.
· Subjective-objective correlation (H3): correlation coefficient between corrected HRV shift and subjective report, compared against the previously documented divergence pattern to see whether correction changes the picture.
---
9. Limitations (stated up front, not discovered later)
· Single practitioner, likely small N — results are hypothesis-generating, not confirmatory.
· No cellular or biochemical measurement in this phase — Tier 3 of the model remains untested by this design and should not be cited as supported by it.
· Sham condition can't be perfectly blinded given the perceptible nature of the stimulus; partial blinding is the realistic ceiling.
· Placebo/expectation effects are a genuine competing explanation for any personalized-vs-sham difference found, and should be discussed as such rather than assumed away.
· Existing client roster is not randomly sampled from any broader population — generalizability beyond this practice is not something this pilot can claim.
---
10. What This Pilot Would (and Wouldn't) Establish
If H1 and H2 hold: This would be a legitimate, citable basis for the claim that Sonic Balance's personalization mechanism produces a measurable, reproducible effect beyond a generic vibroacoustic session — the load-bearing claim for the whole protocol, and a credible foundation for approaching a research partner about the Tier 3 (cellular/biomarker) question.
If H1 or H2 don't hold: That's real information — it would mean either the resonant-frequency identification method needs refinement, or the personalization effect is smaller/noisier than session documentation has suggested, and either finding should shape the next iteration of the protocol rather than being set aside.
What it would not establish, under any outcome: Anything about gene expression, cytoskeletal dynamics, or cellular-level mechanotransduction in actual Sonic Balance clients. That remains Tier 3 — a separate, later study, requiring different instrumentation and likely an external lab or academic partner. (Future Collaboration)