01Identify the claim that crosses the evidence boundary

NCATS describes preclinical research as connecting basic disease science with human medicine through models that can include cells, animals, tissues and simulations. That research setting is distinct from studies directly investigating outcomes in people.NCATS — Translational Science Spectrum (opens in a new tab)

In an original hypothetical example, Peptide T changes a laboratory score in an animal model. A later headline says it improves recovery in humans. The headline introduces both a new population and a broader outcome; those changes require evidence of their own.

Start by writing the original finding without promotional shorthand. Include which model was used and what the score represented. Then write the human claim separately, so the difference between the two becomes visible.

A proposed biological explanation can make the connection plausible. Plausibility is useful for forming a research question, but it does not supply an observed human result.

02Compare more than the species

For the hypothetical score, ask whether it measures a mechanism, a visible behaviour or a functional outcome. A change in one of these does not automatically establish a change in the others.

The experimental time window also belongs in the comparison. A short observation made at a selected stage cannot, without further evidence, establish that an effect persists or applies at a different stage.

Do not fill these gaps using a numerical conversion between animal and human quantities. Such arithmetic does not establish comparable exposure, benefit or safety, and this article provides no dosing or administration guidance.

03Expect the strength of the connection to vary

Perel and colleagues compared animal and clinical evidence for six selected interventions in a 2007 study. Agreement varied, and the authors discussed methodological bias and how well models represented clinical circumstances. They explicitly cautioned against generalising about all animal research from that small selection.Perel et al. — Comparison of treatment effects between animal experiments and clinical trials (opens in a new tab)

That study is a methodological example, not a current verdict on any peptide sold here. It shows why the relevant question is how well a particular body of evidence connects to a particular claim.

Multiple animal studies may strengthen confidence that an observation occurs under the tested conditions. If they all leave the same human-relevance question unresolved, repeating that observation does not by itself answer the missing question.

04Write a conclusion that keeps the gap visible

For the original example, a bounded summary would say that the study reported a change in a specified score in an animal model, then explain whether evidence connecting that score to the proposed human outcome was available.

If human studies exist, examine their material, population and outcomes independently. A clinical paper mentioning the same peptide family may not have tested the same molecule or the same claim.

If the relevant human outcome was not studied, say so directly. Avoid presenting an animal result as a human benefit with a small qualifier added later; the evidence setting belongs in the main sentence.

This approach leaves room for promising hypotheses while making clear what is known, what has actually been measured and which inference remains to be tested.