How this was actually run.

Cohort design, measurement, dropout, and the limits of what these numbers can and cannot say.

The three cohorts

Recruitment and consent

Participants were recruited informally through the author's professional and personal network, with bias toward over-35 adults living independently and able to commit to a 12–18 month protocol window. Each participant gave written, informed consent acknowledging that this was not a clinical trial, that there was no IRB review, that participation was voluntary at every step, and that they should continue working with their own physicians for any matter that ordinarily required one.

Intervention

The intervention is the same one documented in Chapter 5 · Protocols: a four-data-layer testing baseline (exome, focused bloodwork, IgE/IgA/GI inflammation panel, wearable), a personalized supplement and training prescription generated from those layers, and a structured set of training non-negotiables (strength, Zone 2, structured intervals, sleep). No individualized prescription deviated from the cookbook's published structure.

Measurement

• Heart rate variability: continuous from a wearable; reported as the change in moving-average HRV across protocol windows.
• Sleep: nightly minutes from the same wearable; reported as the change in average minutes-per-night.
• Weight: weekly self-report with sanity checks against bloodwork-window measurements.
• GI inflammatory markers: baseline + follow-up panels; reduction reported as a range across the cohort.
• Biological age (n=1 only): DunedinPACE.

Dropout

Cohort 3 dropout was 49% (17 of 35 did not complete). The cookbook treats dropout as the most important finding of the trial, not a footnote: compliance, not science, is the rate-limiting step in a consumer longevity program. Dropout reasons collected at exit clustered into three categories — schedule disruption (work or family events that displaced training), measurement fatigue (specifically wearable adherence), and supplement-plan complexity. Cohort 1 and 2 had similar but smaller dropout in absolute terms; the cookbook does not aggregate them into a single completion rate.

What the numbers can and cannot say

• Can say: the protocol is reproducible, the headline numbers (3–5× HRV improvement, >10% mean weight loss, 30–40 minutes additional sleep, 9–57% reduction in GI inflammatory markers, 10-year DunedinPACE drop on the author) survived three small completer cohorts and one personal n=1.
• Cannot say: that any individual recipe caused any individual outcome (the protocol is composed of layers); that the same effects would land in any specific person who tried the protocol; that the magnitude of effect reported is the magnitude an individual should expect; that any of this is medical advice or evidence at the level a clinical trial would produce.

Conflicts of interest and monetization

Sultan is the sole author. The cookbook was not commissioned by, sponsored by, or paid for by any clinic, supplement company, lab, wearable manufacturer, or pharmaceutical company. There are no affiliate links on any cookbook page. There are no sponsored placements. There is no paid course, no clinic, no membership tier, and no token sale. The work is published in part because there is no business model that depends on it staying behind a paywall.

If you have a methodological objection, email me. Substantive objections that change the cookbook's claims will be answered in the next revision.