How the evidence tiers work

• The three tiers
• ✗ is a safety flag
• No letter grades
• “No human trials”
• Community dosing logs
• Soviet clinical series
• Khavinson bioregulators
• Predicted from mechanism
• How to use this

Every compound, supplement and topical in this Vault carries an evidence tier. This page is what the tier means — and, just as importantly, what it does not mean.

The three tiers

Everything in the Vault sits in one of three places, decided by what kind of evidence exists behind the specific claim being made — not by how promising the compound is, how popular it is, or whether I like it.

✗ is a safety flag, not a grade

Where you see a cross on this site, the concern is harm. It never means a disappointing trial.

That distinction was Coach Cam's call, and the reasoning behind it is worth stating in full: a compound is usually tested for one specific indication, in one specific population, at one specific dose. Missing that endpoint tells you about that question. It does not tell you the molecule does nothing. AOD-9604 was judged on scale weight; a lipolytic agent that does not suppress appetite was never going to win that contest, and the lipolysis data did not stop existing when the trial read out.

So a negative trial is written up as a negative trial — what was measured, what moved, and why the miss happened — and it stays in its tier. The cross is reserved for things that can hurt you, it sits alongside the tier rather than replacing it (something can be both well-studied and genuinely risky), and it stays rare enough to mean something.

There are no letter grades on this site

A–D grades are a protocol device and they live inside Skool, where each one is attached to a specific claim in a specific protocol and explained where it sits. They are not a Vault feature. If you see a bare letter on a Vault page, it is a defect — there is a build gate whose entire job is to catch one.

The situations that come up again and again

Five evidence patterns recur across hundreds of items in this Vault. Each one is explained once, here, and linked from the pages it applies to.

“No human trials” — which is three different facts

An empty clinical tier is never just an absence. It always has a reason, and the reasons are not interchangeable:

• No route to a return. The compound is sold for laboratory use and nobody has funded a trial, because there is no patent at the end of it to pay for one. The correlative and theoretical tiers are the actual evidence base here, not a consolation prize.
• It never left preclinical. Development stopped, so the ceiling on what anyone can claim is a rodent model — and rodent models transfer to humans poorly in most of the endpoints people care about.
• Not yet. Human work is underway and the position may change. That is a statement about the calendar, not about the molecule.

Read which one a page says. They lead to different decisions.

Community dosing logs

For a large part of this space, the human record beyond the trials is self-reported: forum logs, coaching notes, what people say worked.

That is real information about tolerability and almost nothing about efficacy. The reason is selection, not honesty. Nobody posts the cycle where they felt no different. What survives is a filtered sample that will always look better than the truth, and the filter gets stronger the more enthusiastic the community is.

Where it is worth something: side effects. People report what went wrong far more reliably than they report an absence of benefit, so a large body of logs with no reports of a particular problem is weak evidence that the problem is uncommon at those doses.

Soviet and post-Soviet clinical series

Several families of compounds — the bioregulators especially, but also a number of nootropics — have a human record that is genuinely clinical and almost entirely Russian-language. Real patients, real endpoints, published work, decades of it.

It was also largely conducted without the blinding, randomization, pre-registration or independent replication a Western regulator would require, and very little of it has been repeated outside the originating institutes.

That is more than nothing and less than a trial, and the honest read sits between the two. Dismissing it because it is Russian is lazy. Citing it as equivalent to a Western phase III is dishonest. Where the Vault relies on this literature, it says so on the page.

The Khavinson peptide bioregulators

The short peptide bioregulators — Epitalon, Thymalin, Cerluten, Pinealon and the rest of the series — share one mechanistic claim rather than having a separate rationale each. Stating it once is more honest than paraphrasing it thirty times as though each peptide arrived at it independently.

The claim: peptides this short are proposed to pass into the cell nucleus and bind directly to specific promoter regions of DNA, switching tissue-specific genes back toward a younger expression pattern. Khavinson's group has published binding and chromatin-decondensation work supporting it.

Two things to hold at once. The mechanism is a real, testable hypothesis with published cell-level work behind it — it is not marketing invention. But essentially all of that work comes from one research group, none of it has been independently replicated at scale, and sequence-specific gene targeting by a tetrapeptide is an extraordinary claim that would reshape molecular biology if it held broadly.

Most of the series is also sold as oral capsules, which adds a second unproven step on top of the first: surviving digestion intact. Run these as an experiment you measure, not as a protocol you trust.

Mechanism-predicted safety and interference

Much of this compound class has never been through a human safety trial, and almost none of it has been through an interaction study. There are two ways to handle that. One is to print a generic warning and say nothing useful. The other is to reason forward from what the molecule actually does.

The Vault does the second, and labels it. Where a page says predicted from mechanism, it means exactly that: no trial of this combination exists, and what follows is what the biology implies. Treat it as a reason to watch something, not as a finding.

The same applies to the bloodwork rows. Those are the markers a compound is expected to move and in which direction — and knowing that in advance is mostly about not panicking, because some of those markers moving is the compound working.

How to actually use this

The tier is the first question, not the last one. A sensible read runs:

1. What kind of evidence is behind this specific claim? That is the tier.
2. How big is the effect, and in whom? A validated compound with a small effect in a population unlike you may matter less than a mechanistic one that targets exactly your problem.
3. What could go wrong, and would I see it? Safety content on this site is free and always outside the paywall, for that reason.
4. What would tell me it is working? If the answer is “how I feel”, expect to fool yourself. If it is a marker, get the baseline first — a baseline you didn't take is one you can never go back for.

My job is to tell you which one you are looking at and let you make the call. Grading something low is not me dismissing it — it is me refusing to oversell it. This is the research space, and being able to reason forward from a mechanism matters as much as waiting for the trial.

← Open the interactive Vault