This protocol governs how HealthMaxxing turns a pair of logged signals into a personal wellness observation. It is intentionally conservative, publicly versioned, and designed to allow “not enough data” and “unclear” as valid outcomes.
1. Scope and exclusions
The protocol applies to nonmedical questions about a person’s own eligible wellness data. It excludes diagnosis, disease prediction, treatment selection, crisis support, medication decisions, unsafe experimentation, and claims about people who were not observed.
2. Predefine the question
- Name one exposure or earlier signal.
- Name one outcome or later signal.
- Choose a sensible same-day or lagged relationship.
- Write the direction neutrally; do not assume improvement.
- Identify the evidence class before analysis.
3. Check readiness
HealthMaxxing begins checking after seven logged days. Many questions need roughly 10–14 relevant overlapping days; some need more. The check must consider overlap, coverage, variability, timing, source consistency, and obvious change points. Calendar days alone are insufficient.
4. Analyze at the available level
The analysis should remain descriptive unless a stronger design is actually used. Report the window, paired observations, direction, uncertainty, and sensitivity to obvious outliers or missingness. Do not search many outcomes and present the most dramatic one as preplanned.
5. Name competing explanations
At minimum, review major schedule changes, illness, travel, device or permission changes, missing entries, medication changes, season, and events plausibly related to both signals. The relevant list varies by question.
6. Use constrained language
| Observation | Allowed form | Disallowed upgrade |
|---|---|---|
| Signals differed together | “On your logged days, X tended to coincide with Y.” | “X caused Y.” |
| One person’s record | “This may be worth watching in your routine.” | “This works for people.” |
| External topic research | “Published research makes this a plausible question.” | “Research proves HealthMaxxing works.” |
| Insufficient or mixed data | “No clear relationship appeared in this window.” | Silence, invented certainty, or selective reporting. |
7. Apply the safety boundary
Never recommend medication changes, restriction, hazardous exposure, ignoring symptoms, or substituting tracking for care. Escalate medical and urgent questions outside the product’s wellness scope.
8. Version and correct
Protocol version 1.0 was published July 21, 2026. Material changes to inputs, thresholds, missing-data rules, language, safety boundaries, or claim classes require a dated version update and review of affected public pages.
Sources and limitations
- N-of-1 and Novel Within-Subject Trial Methods — NIH Office of Disease Prevention. NIH introduction to rigorous within-person trial methods and the circumstances in which they are or are not useful.
- N-of-1 Trials, Their Reporting Guidelines, and the Advancement of Open Science Principles — PubMed Central. Peer-reviewed discussion of N-of-1 reporting, transparency, and open-science practices.
Sources are selected for the narrow statements on this page. A citation does not imply that the authors, journals, NIH, CDC, Apple, or any other source reviewed or endorse HealthMaxxing.