The Evidence Hierarchy: How to Tell a Real Claim From a Good Story

Every health claim you read is backed by some kind of evidence. The question is never simply "is there evidence?" but "how strong is it, and how much should it move my confidence?" A claim supported by a dozen large human trials and a claim supported by one experiment in mice are worlds apart, even though both can be described with the same three words: "studies show."

This article explains the tool researchers use to keep those worlds separate: the evidence hierarchy. Once you understand it, marketing that once sounded authoritative starts to reveal exactly how thin, or how solid, it really is. At the end we explain how the Peptivis rating system turns this hierarchy into a simple label you can use at a glance.

The evidence pyramid

Picture a pyramid. At the bottom sit the weakest, most abundant forms of evidence. As you climb, the evidence gets stronger, harder to produce, and rarer. Each level answers a slightly different question and carries a different risk of being wrong.

Level 1: Anecdotes and case reports

At the base are individual stories. A case report is a physician writing up one patient's experience; an anecdote is someone online describing what happened to them. These are the easiest to generate and the least reliable for drawing general conclusions.

The problem is not that anecdotes are lies. It is that a single experience cannot separate the effect of the compound from everything else that was happening: the placebo effect, natural recovery, other lifestyle changes, or simple coincidence. Case reports are valuable for one thing, spotting something unexpected worth investigating, but they cannot tell you whether an effect is real or how often it happens.

Level 2: Preclinical research (cells and animals)

The next level up is laboratory work: experiments on cells in a dish (in vitro) and on animals such as mice and rats (in vivo). This is where a huge share of peptide science lives, and it is genuinely important. Preclinical research is how scientists explore mechanisms, screen for toxicity, and decide whether a compound is worth testing in humans at all.

But preclinical results come with a giant asterisk. A mouse is not a small human. Doses that work in a dish may be impossible or unsafe to reach in a person. The failure rate from "worked in animals" to "works in people" is notoriously high, most compounds that look promising in rodents never make it as human therapies. Preclinical evidence generates hypotheses. It does not confirm them.

Level 3: Observational human studies

Now we reach actual humans. Observational studies watch what happens to people without assigning a treatment: they follow groups over time (cohort studies) or compare people who did and did not have some exposure (case-control studies). Because these involve real people over real timeframes, they are a big step up from animal work.

Their core weakness is confounding. If people who take a supplement are also wealthier, more health-conscious, and more active, you cannot easily tell whether the supplement caused the better outcome or simply came along for the ride. Observational studies are excellent for spotting associations and generating strong hypotheses, but association is not causation, and good researchers are careful never to confuse the two.

Level 4: Randomized controlled trials (RCTs)

The randomized controlled trial is the workhorse of causal evidence. Participants are randomly assigned to receive either the intervention or a control (often a placebo), ideally without anyone, participant or researcher, knowing who got what until the end. Randomization is the magic: it spreads confounding factors evenly across groups, so a difference in outcomes can be attributed to the treatment itself.

A well-run RCT is the first level where you can say "this compound caused this effect" with real confidence. Not all RCTs are equal, though. A tiny trial of 15 people, or one that measures a surrogate marker instead of an outcome that matters, is far weaker than a large, long, well-designed trial.

Level 5: Systematic reviews and meta-analyses

At the top sit systematic reviews and meta-analyses. Rather than running a new experiment, these methodically gather all the relevant studies on a question, assess their quality, and, in a meta-analysis, statistically combine their results. By pooling many trials, they smooth out the noise of any single study and reveal the overall direction and size of an effect.

This is the strongest everyday evidence available. When a high-quality systematic review of multiple good RCTs points clearly in one direction, that is about as confident as science gets outside of overwhelming consensus. Even here, a review is only as good as the studies feeding it, garbage in, garbage out remains true.

Why "studies show" is so often misleading

Notice that every level of the pyramid can honestly be called "a study." That is exactly why the phrase "studies show" is nearly meaningless on its own. It tells you evidence exists somewhere; it tells you nothing about which floor of the pyramid that evidence sits on.

Marketing exploits this relentlessly. A product page might cite "clinical research" that turns out to be a single in-vitro experiment. It might describe a compound as "shown to" do something, when the showing happened only in rodents. It might quote an impressive percentage from a study of eight people. None of this is necessarily fabricated. It is simply low-level evidence dressed in the language of high-level certainty.

The defense is a habit of asking three questions of any claim:

1. What level of the pyramid is this? Cells, animals, observational, RCT, or review?
2. In whom, and how many? Humans or animals; dozens or thousands.
3. Has it been repeated? One study is a signal; many consistent studies are evidence.

Preclinical vs. human data, and why so much peptide evidence is preclinical

For peptides specifically, this distinction is everything, because the field is unusually top-heavy with preclinical work.

There are structural reasons for that imbalance. Human trials are extraordinarily expensive and slow, often running many years and costing enormous sums, and they are typically funded by companies pursuing regulatory approval for a patentable drug. Many popular research peptides are not being developed toward approval by anyone, so the money that would fund large human trials never appears. The result is a literature full of intriguing animal findings and very little rigorous human data.

That is why compounds like BPC-157, TB-500, and the growth hormone secretagogues such as ipamorelin and CJC-1295 are so often discussed on the strength of rodent studies. It is not that the animal research is bad; it is that the human research that would confirm or refute it has largely not been done. Reading a stack of positive mouse studies and concluding "this works" is precisely the mistake the evidence hierarchy exists to prevent.

Contrast that with a compound like creatine, which has been tested in a large number of human trials across decades, or semaglutide, which went through full clinical development. Those compounds can support confident claims because the human evidence actually exists. The difference is not the type of molecule. It is the depth of the evidence behind it.

The Peptivis rating system

To make all of this usable at a glance, every compound profile on Peptivis carries an evidence rating. It compresses the pyramid into four honest tiers, and it describes how much we know, not whether we endorse anything.

Strong

Strong evidence means there is robust, consistent human evidence, typically multiple well-designed RCTs, often summarized in systematic reviews or meta-analyses, pointing the same way. This is the tier where confident statements are justified. Relatively few compounds in the peptide and performance space reach it.

Moderate

Moderate evidence means there is meaningful human evidence, but with real limitations: fewer trials, smaller samples, mixed results, or open questions about who benefits and by how much. The effect is plausible and human-supported, but not settled. Reasonable experts might still disagree.

Emerging

Emerging evidence means early signals exist, from small or preliminary human studies, or from strong preclinical work, but the evidence is not yet enough to draw firm conclusions. This is where a great deal of peptide research currently sits: interesting, actively studied, and genuinely unproven. Optimism here should be tempered with patience.

Insufficient

Insufficient evidence means the evidence is too thin, too low-quality, or too conflicting to say much of anything with confidence. There may be enthusiastic marketing and a pile of anecdotes, but the rigorous studies simply are not there. An Insufficient rating is not a verdict that something fails; it is an honest admission that the question is open.

The purpose of these labels is discipline. They force us to match the strength of a claim to the strength of its evidence, and they let you calibrate your own confidence without wading through every study yourself.

Putting it to work

The evidence hierarchy is not academic trivia. It is a practical filter for a world overflowing with confident health claims. The next time you read that a compound is "clinically proven" or that "research shows" a dramatic benefit, run it through the pyramid. Ask what level the evidence sits on, whether it was in humans, and whether anyone has repeated it.

Most of the time, that one habit will tell you more than the marketing ever intended to. And when the answer is "one mouse study" dressed up as a breakthrough, you will recognize the story for what it is: a hypothesis wearing the costume of a conclusion.

This article is educational and describes how to evaluate evidence. It is not medical advice, and it does not recommend using any specific compound. For decisions about your own health, consult a qualified professional.