Preclinical versus clinical evidence: how to weigh published research on peptides

Much of what is written online about research peptides blurs a line that scientists treat as fundamental: the line between preclinical evidence (cells and animals) and clinical evidence (controlled studies in humans). Every compound profile on this website keeps the two apart. This article explains why, and how to read each.

The evidence hierarchy, briefly

From weakest to strongest, for questions about effects in humans:

1. Mechanistic and in-vitro studies — what a molecule does to cells, receptors or enzymes in a dish.
2. Animal studies — effects in a living organism, usually rodents, often in artificially induced injury or disease models.
3. Case reports and uncontrolled human studies — observations without a comparison group.
4. Controlled clinical trials — humans randomised to the compound or a control, ideally blinded.
5. Systematic reviews and meta-analyses of controlled trials.

Steps 1 and 2 are preclinical. They generate hypotheses and characterise mechanisms. They do not establish that a compound is safe or effective in people. The historical attrition rate between promising animal data and approved medicines is very high; most candidates fail.

Why animal data does not transfer

• Models are not diseases. A surgically transected rat tendon is a controlled wound, not a chronic human tendinopathy.
• Dosing and exposure differ. Animal studies often use exposures far beyond anything relevant to humans.
• Species differ in receptors, metabolism and physiology.
• Publication bias favours positive animal results.

None of this makes preclinical research unimportant. It is where mechanisms are discovered. It is simply the wrong kind of evidence for a claim about human effects.

Reading a human study critically

If a compound has human data, ask:

• Was it randomised and controlled? Against placebo or an active comparator?
• Was it blinded? Open-label studies of subjective endpoints are unreliable.
• How many participants, and for how long?
• Was the endpoint pre-registered and clinically meaningful?
• Who funded it, and do the authors have interests in the result?
• Has it been replicated by an independent group?

A single phase 1 study in 10 healthy volunteers is human evidence, but it answers a narrow question (tolerability and pharmacokinetics), not whether the compound works.

Licensed medicines versus research compounds

Several compounds on this website — semaglutide and tirzepatide, for example — are licensed medicines with large phase 3 programmes. That evidence concerns a pharmaceutical product, manufactured to pharmacopoeial standards, prescribed and monitored by clinicians, in a defined population. It says nothing about research-grade material, and nothing on this website should be read as suggesting otherwise.

How we apply this

On each compound page you will find:

• A preclinical findings section, clearly labelled.
• A separate human clinical evidence section, which states plainly when no such evidence exists.
• A limitations section listing unresolved questions.
• A reference list in which every citation is tagged preclinical, clinical or review and carries a verification status.

See, for example, how the BPC-157 profile separates an extensive rodent literature from the absence of controlled human trials.

This article is for scientific information only and does not constitute medical advice.