Inside Peptides
Search
Menu

The Inside Peptides Evidence Scale

A named, versioned, claim-specific scale with five grades: Strong, Moderate, Early Human, Preclinical and Insufficient. Each grade has a fixed definition, decision rule and anchor you can cite.

Last updated · IPES version 1.0

A conceptual spectrum of laboratory, animal, early human, controlled human and clinical research. Bar heights are illustrative, not measured evidence scores.
Research setting and evidence certainty are related but different. This illustration is not a numerical scale. [1][2]Original illustration · Inside Peptides

A scale you can cite.

The Inside Peptides Evidence Scale (IPES) is the five-grade system every Inside Peptides profile, comparison and dataset uses to describe how much good human evidence supports a specific claim. It is published here as version 1.0 with fixed definitions and decision rules, so that a grade quoted elsewhere means the same thing as it does on this page. Each grade has a permanent anchor (for example #preclinical), and the scale is versioned: if a definition changes, the version number changes and the previous wording is kept in our corrections log.

To cite the scale: Inside Peptides. (2026). Inside Peptides Evidence Scale (IPES), version 1.0. https://insidepeptides.org/evidence/evidence-levels/. To cite one grade, link to its anchor on this page. Data files from Inside Peptides record the grade in the evidenceLevel field using the keys strong, moderate, early-human, preclinical and insufficient.

Read the claim before the badge.

A badge summarizes our assessment of a specific claim. It does not rank a peptide’s overall worth, tell you to use it, or certify safety. Ask: evidence for which outcome, in whom, and for which formulation?

Certainty should be attached to the outcome, and statistical findings require clinical interpretation. [3][2]

Research setting is not a score.

Laboratory, animal and human describe where a study took place. Our levels also reflect whether the evidence supports a dependable conclusion. Moving to human research does not automatically make evidence strong.

01

Strong Evidence

Strong Evidence

Substantial, high-quality human evidence supports a clearly defined conclusion for the specified outcome and population.

Typical research
Well-conducted, appropriately powered human studies with independent confirmation, or a rigorous synthesis of directly relevant studies.
Main limitations
Still bounded by the studied population, formulation, outcome and follow-up. A strong conclusion can be negative or show no clinically meaningful benefit.
How we describe a claim
High-quality human research supports [specific conclusion] in [studied population].
02

Moderate Evidence

Moderate Evidence

Meaningful human evidence supports a conclusion, but an important uncertainty remains.

Typical research
Relevant controlled human research with incomplete replication, limited precision or a material design or applicability concern.
Main limitations
The estimated effect or its relevance could change with better evidence. Moderate does not mean the treatment is moderately effective.
How we describe a claim
Human studies suggest [specific finding], although [key limitation] remains.
03

Early Human Evidence

Early Human Evidence

Research in people addresses the claim, but it cannot yet support a dependable clinical conclusion.

Typical research
Pilot studies, small uncontrolled studies, preliminary clinical trials or limited human outcome data.
Main limitations
Small samples, short follow-up or inadequate comparators may leave effectiveness and harms unresolved.
How we describe a claim
Early human research has investigated [specific outcome]; effectiveness is not established.
04

Preclinical Evidence

Preclinical Evidence

Relevant findings come from animal, cell, laboratory or other nonclinical models, without adequate human evidence for the claim.

Typical research
Animal experiments, cell cultures, tissue models and experimentally tested mechanisms.
Main limitations
Results in a model do not establish safety or effectiveness in people. A plausible mechanism is not a clinical result.
How we describe a claim
In preclinical models, researchers observed [finding]. Whether this applies to people is unknown.
05

Insufficient Evidence

Insufficient Evidence

Reliable, relevant evidence is too limited or too compromised to draw a meaningful conclusion about the claim.

Typical research
No usable completed studies; theoretical arguments; anecdotes; or research with flaws that prevent a defensible conclusion.
Main limitations
Uncertainty is not evidence that a compound works, and it is not proof that it has no effect.
How we describe a claim
There is insufficient reliable evidence to determine [specific outcome].

A spectrum, not an escalator.

Human trials address questions that preclinical studies cannot settle. Early research may raise new questions, fail to replicate or stop; multiple studies still need appraisal for consistency and quality. [4][5]

  1. 01

    In vitro

    Cells, tissues or laboratory systems.

  2. 02

    Animal / preclinical

    Biology and safety questions in nonhuman models.

  3. 03

    Early human

    Initial research in people.

  4. 04

    Controlled human trials

    Compare outcomes between study groups.

  5. 05

    Multiple high-quality human studies

    Assess consistency and applicability across studies.

  6. 06

    Regulatory evaluation

    Evaluate a specific medicine and use, where applicable.

Progression is not guaranteed. These are research settings and evaluation steps, not a promise of approval. Strong clinical conclusions require more than reaching a trial phase. [4][5]Original educational diagram · Inside Peptides

The five level definitions above are Inside Peptides’ own editorial framework. Read the decision rules for how a label is selected. They are not a conversion of GRADE’s four certainty categories.

Two labels. Two different questions.

Preclinical Evidence

What supports the claim?

The evidence label describes the basis and limits of a conclusion.

Approved Medicine

Is a specified medicine authorised?

The status label needs a product, use and jurisdiction. It is shown here as a label example, not assigned to any compound.

Being studied does not mean being approved. A registry entry does not endorse the intervention, and a regulator’s decision has a defined scope. [6][7]

Understand the regulatory labels →

Four questions worth asking.

  1. Was it measured in people? Identify the research setting, not just the headline.
  2. What was the comparison? Look for a credible control and how participants were assigned.
  3. Was the outcome relevant? A biomarker change and a clinical benefit are different claims.
  4. How uncertain is the result? Read the effect estimate, limitations and what remains unknown.

These questions follow the distinction between trial design and interpretation of outcomes. [8][2]

References

Official source destinations checked 2026-10-06. External links are marked ↗.

  1. NIH / NIEHS: Alternatives to animal testing (external website)

    Research models include cells, animals and computer models; human relevance can be uncertain.

    Back to citation ↑
  2. Cochrane Handbook, chapter 15: Interpreting results (external website)

    Effect size, confidence intervals, clinical relevance and applicability.

    Back to citation ↑
  3. Cochrane Handbook, chapter 14: Grading certainty of evidence (external website)

    Outcome-specific certainty, five uncertainty domains; formal GRADE has four levels.

    Back to citation ↑
  4. FDA: Step 3: Clinical research (external website)

    Human study design, trial phases and investigational research.

    Back to citation ↑
  5. Cochrane Handbook, chapter 10: Meta-analysis (external website)

    Evidence synthesis, heterogeneity and limitations of pooling.

    Back to citation ↑
  6. ClinicalTrials.gov: Official disclaimer (external website)

    Registration is not government endorsement; sponsors supply study records.

    Back to citation ↑
  7. FDA: Step 4: Drug review (external website)

    FDA reviews submitted evidence and labeling in approval decisions.

    Back to citation ↑
  8. Cochrane Handbook, chapter 8: Risk of bias in randomized trials (external website)

    Randomization, allocation concealment, missing outcomes and selective reporting.

    Back to citation ↑