Editorial Methodology

How we decide what to publish

Evidence-based peptide content is rare because rigor is expensive. We publish our method publicly so you can judge the rigor yourself and replicate any claim from its primary source.

Last reviewed April 2026 · Updated as rules and process change, with every revision logged.

The short version

  1. Every piece of content is screened against 23 automated compliance rules covering FDA, FTC, CAN-SPAM, Tennessee state law, social-syndication standards, and general editorial guardrails.
  2. Every research-derived claim is graded through a 12-point bias framework before it reaches a draft.
  3. Every outbound citation links to a primary source — not a secondary summary.
  4. Every regulatory or product-safety change is logged to a public changelog, never overwritten.
  5. We take no advertising and no compounding-pharmacy affiliate money. Ever. Financial relationships we do hold (lab testing, wearables, books) are disclosed inline at the point of mention.

The 12-point bias framework

Every study we summarize is screened on these twelve dimensions. When a study fails a screen, we still report the study — but the limitation is named in the summary, not omitted.

BIAS-01

Study design hierarchy

RCT > prospective cohort > case-control > case series > animal > in-vitro. Every claim is tagged with the strongest study tier we can find for it.

BIAS-02

Sample size adequacy

Studies powered below pre-registered thresholds are flagged; small-sample findings are reported as preliminary.

BIAS-03

Funding source disclosure

Industry-funded, government-funded, and independently-funded studies are labeled separately in every summary.

BIAS-04

Conflict of interest screen

Declared author COIs are surfaced alongside the finding, not buried in a footnote.

BIAS-05

Registration and pre-specification

Trials registered on ClinicalTrials.gov / WHO ICTRP before enrollment are weighted higher than post-hoc analyses.

BIAS-06

Outcome reporting completeness

Primary endpoints must be reported. Outcome-switching between registered and published protocol is flagged.

BIAS-07

Blinding and allocation concealment

Open-label data is reported as open-label, not as a blinded result.

BIAS-08

Replication status

Single-study findings are labeled as unreplicated. Replicated findings cite every replication.

BIAS-09

Effect size vs statistical significance

A p-value is not a claim. Effect sizes (absolute and relative) are reported when available.

BIAS-10

Publication venue and peer review

Predatory journals, non-indexed venues, and preprints are labeled as such.

BIAS-11

Species and translation gap

Rodent, cell-culture, and in-silico findings are never described as proof of human effect.

BIAS-12

Geographic and regulatory context

International approvals (e.g. Russia, China, EU) are reported as international approvals, not as FDA approvals.

The 23-rule compliance ruleset

Before publication, every newsletter, blog post, social thread, and product page runs through this ruleset. A draft cannot ship while any critical-severity violation is open. Rules and severity thresholds are versioned in our repo; this page is regenerated when they change.

FDA

Rule ID What it checks Citation
C-FDA-01 Drug claim language detection 21 USC 321(g)(1)
C-FDA-02 Therapeutic instruction detection 21 CFR 201.128
C-FDA-03 Dosing protocol detection (practice-of-medicine risk) —
C-FDA-04 Sourcing / vendor referral detection —
C-FDA-05 Missing substance disclaimer for mentioned peptides —
C-FDA-06 Compounding misinformation detection —

FTC

Rule ID What it checks Citation
C-FTC-01 Missing affiliate disclosure 16 CFR 255.5
C-FTC-02 Unsubstantiated health claims —
C-FTC-03 Missing sponsored content label —
C-FTC-04 Testimonial without typical-results disclosure —

CAN-SPAM

Rule ID What it checks Citation
C-SPAM-01 Missing physical postal address 15 USC 7704(a)(5)
C-SPAM-02 Missing unsubscribe mechanism —
C-SPAM-03 Deceptive subject line indicators —
C-SPAM-04 Missing sender identification —

Tennessee state

Rule ID What it checks Citation
C-TN-01 Practice of medicine language TCA § 63-6-204
C-TN-02 Deceptive trade practice indicators TCA § 47-18-104
C-TN-03 Credential misrepresentation —

Social / syndication

Rule ID What it checks Citation
C-SOC-01 Missing fair balance (benefit without risk) —
C-SOC-02 Missing educational disclaimer —
C-SOC-03 Character limit compliance —

General editorial

Rule ID What it checks Citation
C-GEN-01 Missing top-of-content medical disclaimer —
C-GEN-02 Age restriction notice —
C-GEN-03 Provider-patient relationship language —

Editorial process

  1. 1
    Source acquisition. We ingest peer-reviewed literature from PubMed, ClinicalTrials.gov, bioRxiv, medRxiv, the Federal Register, openFDA, and international databases across more than 20 sources.
  2. 2
    Bias screen. Every study passes the 12-point framework. Findings that fail a screen are still reported, with the limitation named in the summary.
  3. 3
    Voice normalization. Draft content runs through an automated voice enforcer that strips hype language, adds hedging where warranted, and rejects drug-claim phrasing.
  4. 4
    Compliance check. The 23 automated rules run before publication. Critical violations block publication; lower-severity flags are reviewed by a human editor.
  5. 5
    Human editorial review. Every piece is reviewed by a licensed clinician editor before publication.
  6. 6
    Publication and archival. Every outbound citation is archived to the Wayback Machine on publish; link-rot is swept weekly.

Financial independence

  • We do not take pharmaceutical advertising.
  • We do not accept compounding-pharmacy sponsorships or affiliate commissions.
  • We do not accept paid reviews or paid inclusion on the reclassification tracker or any rating framework we publish.
  • Affiliate relationships we do maintain — lab-testing services, wearables, books — are disclosed inline at the point of mention per 16 CFR 255.5.
  • Our revenue comes from reader-supported products (Gumroad, newsletter premium tier) and non-pharma affiliates only.

Corrections policy

If we publish something wrong, we say so publicly. Corrections are logged on the originating page and surfaced in the next newsletter issue. We do not silently edit published material.

Report a factual error to corrections@thepeptidecolumn.com with a primary-source citation. Valid corrections are typically processed within 72 hours.

Not medical advice. Methodology transparency does not make educational content a treatment recommendation. See our Medical Disclaimer and No Provider-Patient Relationship notices.