Peptides and the evidence grade you need
BPC-157, TB-500, ipamorelin, CJC-1295. The marketing is enormous; the human evidence is mostly thin. Here is the calm, honest reading.
Peptides have become the supplement aisle of the longevity world. Influencers talk about them like statins; doctors mostly haven't heard of them; the evidence is somewhere quieter than either group will admit.
We are not here to tell you whether to use them. We are here to teach you how to read what's said about them, and how to measure your own response if you do.
the evidence hierarchy, applied to peptides
Use three tiers, in order. Strong: replicated human RCTs with meaningful endpoints. Promising: smaller human trials, mechanistic plausibility, consistent cohort signal. Anecdotal: animal models, in-vitro work, expert opinion, single-clinic case series, influencer claims.
For most peptides being marketed today — BPC-157, TB-500, the GHRH/GHRP family — the bulk of available data is anecdotal or animal. Some entries (e.g. specific tendon-healing or wound contexts for BPC-157) are promising but small (BMJ Open, 2023). Almost none rise to strong (Cochrane review, 2024).
why this matters before you spend a euro
Anecdotal evidence is not zero — it is a starting hypothesis. The honest move is to run the protocol like an experiment, not like a prescription. If you treat anecdote like fact, you confuse training, sleep, season, and placebo for the peptide every single time.
how AI helps you triage what you're reading
- Paste the article, post, or product page into a sourced AI. Ask it to separate strong, promising, and anecdotal claims explicitly.
- Ask for the primary sources behind the strongest claim. If they're missing or animal-only, weaken the claim in your head.
- Ask what would make you change your mind — what trial, what endpoint, what duration would move the claim up a tier.
- Save the brief in your ledger before you start anything, so you can compare reality to expectation later.
and how it helps you measure honestly if you do run one
A clean baseline. A single variable. A defined outcome. Enough weeks. A stopping rule. That is the entire backbone of a defensible personal n=1 (Hashimoto et al., 2025). AI can hold the structure for you so the protocol survives the noise of normal life.
- Baseline 4 weeks of the outcome you actually care about — sleep quality, recovery, soreness, body composition, training load.
- Run the peptide as a single variable for 6–12 weeks.
- Compare the change to your normal week-to-week variability, not to the influencer's before/after photo.
- Stop on side effects. Stop on no signal. Stop on doubt.
where the line is
Most peptides discussed online are off-label, research-use-only, or unregulated in the EU. Sourcing, dosing, sterility, and interactions are real problems with real consequences. None of that should be outsourced to a chatbot, and none of that is what we teach.
What we do teach: read claims with grade-aware honesty, measure your response with discipline, and let evidence — your own and the literature's — do more of the deciding than the algorithm does.
Recommended next