Research Study Records

A compound page summarizes an evidence set. A study record does something narrower: it examines one registered study or primary paper closely enough to show what was tested, who was studied, which outcome was primary, what the result was, and where the inference stops.

This primer explains how to read a study record and provides a checklist for interpreting its evidence. For a broader introduction to paper structure, see how to read a research study.

Published study records

Browse the peptide and compound study directory for study records organized by research topic.

• Acetyl-L-carnitine during taxane therapy: primary versus secondary neuropathy findings, visit-specific participant counts, and same-trial long-term follow-up.

• Growth hormone in prolonged critical illness, excess mortality in two randomized trials and the limits of an anabolic rationale.
• L-carnitine and carotid plaque in metabolic syndrome, neutral plaque volume, exploratory stenosis concerns, missing scans and multiplicity.

• Retatrutide phase 2 obesity trial: NCT04881760, Weight-change findings, analysis populations, adverse events, and limitations.
• Ipamorelin phase 2 postoperative-ileus trial: NCT00672074, Null efficacy findings, adverse events, attrition, and source limits.
• Elamipretide MMPOWER-3 trial: NCT03323749, Null primary endpoints, fatigue-reporting discrepancy, safety, and evidence limits.
• Tirzepatide SURMOUNT-1 trial: NCT04184622, Weight-change findings, estimands, attrition, adverse events, and evidence limits.
• Cagrilintide phase 2 monotherapy trial: NCT03856047, Modeled weight changes, nominal comparator findings, missing data, and harms.
• CagriSema phase 2 combination trial: NCT04982575, Null primary comparison, secondary findings, estimands, attrition, and harms.
• Bremelanotide RECONNECT phase 3 trials, Desire and distress findings, null satisfying-event outcome, endpoint changes, attrition, and harms.
• Tesamorelin first phase 3 trial: NCT00123253, Visceral-fat findings, analysis differences, attrition, harms, and formulation limits.
• Kisspeptin-10 IHH case series: PMID 30430143, Seven-person LH findings, incomplete paired visits, pituitary priming, and missing safety data.
• Thymosin alpha-1 TESTS trial: NCT02867267, Corrected mortality results, subgroup limits, adverse events, and missing data.
• Afamelanotide EPP phase 3 trials: NCT01605136, Sunlight findings, FDA missing-data analyses, adverse events, and product-specific limits.
• Cibinetide sarcoidosis phase 2b trial: NCT02039687, Corneal imaging, inconclusive clinical outcomes, serious events, and conflicting analysis populations.
• BPC-157 knee-pain chart review: PMID 34324435, Telephone-recalled pain reports, mixed exposures, no control, safety gaps, and causal limits.
• GHK-Cu Lamin Gel ulcer trial: PMID 17147644, Named formulation, wound-care context, participant-flow discrepancies, endpoint timing, uncertainty, and regulatory limits.
• Melanotan II crossover study: PMID 9679884, Ten-person physiological findings, later methodological clarification, adverse effects, formulation identity, and evidence limits.
• Six-month oral glutathione trial: NCT01044277, Participant flow, mixed biomarkers, immune-subset limitations, adverse-event reporting, formulation debate, and evidence boundaries.
• Cerebrolysin CASTA trial: NCT00868283, Neutral primary result, exploratory subgroup signal, late registration, enrollment mismatch, and abstract-only evidence limits.
• Levocarnitine cancer-fatigue trial: NCT00091169, Null fatigue comparison, biomarker changes, analysis-count discrepancies, adverse-event reporting, and evidence limits.
• Full-length thymosin beta-4 dry-eye trial: NCT01387347, Two null primary outcomes, selected analyses, safety-reporting discrepancies, missing data, and material-identity limits.
• Four-week oral glutathione biomarker trial: PMID 21875351, Null biomarker comparisons, missing data, reporting inconsistencies, harms, funding, and material limits.
• CJC-1295 with DAC studies: PMID 16352683, Hormone responses, accumulation, adverse events, comparison limits, and the DAC material-identity boundary.
• DSIP chronic-insomnia study: PMID 1299794, Objective sleep findings, unchanged subjective quality, comparison limits, and unresolved safety and material identity.
• Hexarelin physiology study: PMID 7957536, Acute growth-hormone responses, repeated-participant design, fitted-model limits, and unresolved safety and material identity.
• Direct NAD+ metabolome pilot: PMID 31572171, Plasma and urinary measurements, randomized-pilot limits, short safety follow-up, and material-identity boundaries.
• ARA 290/Cibinetide type 2 diabetes trial: NTR3858, Neuropathy and metabolic outcomes, small-sample limits, adverse events, and material-identity boundaries.
• Full-length thymosin beta-4 venous-ulcer phase 2 trial: NCT00832091, Wound-closure outcomes, analysis limits, safety reporting, and full-length-peptide identity.
• Intranasal glutathione phase 2b Parkinson trial: NCT02424708, Clinical and biomarker findings, missing data, adverse events, and formulation limits.
• Kisspeptin-54 second-dose IVF phase 2 trial: NCT01667406, Oocyte-maturation outcomes, trial design, harms, and product-transfer limits.
• KRP-116D 50% DMSO phase 3 bladder trial: jRCT2080223511, Symptom outcomes, comparator context, adverse events, and formulation-specific boundaries.
• Retatrutide TRANSCEND-T2D-1 phase 3 trial: NCT06354660, Registered design, prespecified outcomes, status limits, and no assumption of favorable results.
• Tesamorelin HIV-associated NAFLD randomized trial: NCT02196831, Liver-fat outcomes, analysis populations, adverse events, and formulation boundaries.
• Tirzepatide SURMOUNT-OSA phase 3 trials: NCT05412004, Sleep-apnea and weight outcomes, estimands, adverse events, and approved-product boundaries.

This distinction matters because a study headline can hide the details that determine its meaning. A favorable secondary outcome does not replace a null primary outcome. A registered study is not necessarily completed or published. A paper about a named investigational or approved drug does not authenticate a separately sold catalog material.

The Serum Research Study Library is educational. It does not provide medical advice, treatment recommendations, dosing, preparation, administration, reconstitution, or directions for human or veterinary use. No study record is a product endorsement or a quality record for a Serum Research item.

What to check in a study record

Read the question before the result

A study can measure many outcomes. The study record starts with the prespecified primary outcome because that is the question the design and sample-size plan were built to answer. Secondary outcomes can add useful information, but they remain secondary. Exploratory, subgroup, and post-hoc analyses are labeled so they do not silently become the main conclusion.

The record also separates the outcome measure from the broader claim. A change in a laboratory value, imaging measurement, hormone concentration, walking test, or questionnaire score is evidence about that measurement under the protocol. It is not automatically evidence of longer life, fewer hospitalizations, improved daily functioning, or another outcome the study did not test.

The model and population also set the boundary: a cell experiment, animal study, and human trial answer different kinds of questions. The guide to in-vitro, animal, and human evidence explains those distinctions.

Registration, results posting, and publication are different records

PubMed is a National Library of Medicine resource for finding biomedical literature.[1] Finding a citation is a starting point for reading the evidence, not a substitute for assessing its quality.

ClinicalTrials.gov defines protocol fields for items such as study design, arms, outcomes, enrollment, sponsor, and dates.[2] Its results definitions separately describe participant flow, baseline characteristics, outcome measures, adverse events, and related fields.[3]

Those records can be compared with a journal article, but none should be treated as infallible. ClinicalTrials.gov states that sponsors and investigators submit study information and that the U.S. government does not review or approve the safety and science of every listed study.[4] A completed status does not mean results were posted. Posted results do not mean a journal article exists. A publication can also emphasize a different analysis from the registry’s original outcome hierarchy.

Prospective registration improves transparency by establishing a dated protocol record, but registration is not an evidence grade, a favorable result, or FDA approval.[5]

Use the ClinicalTrials.gov and FDA status guide to keep trial status, posted results, publication, and product approval separate.

Null, negative, and mixed findings remain visible

A study record does not discard a trial because its primary result was not statistically significant. It reports the estimate, uncertainty, sample size, missing data, and design limits so readers can distinguish “no statistically significant difference was shown” from the stronger and often unsupported statement “there was no effect.”

The same rule applies in the other direction. A statistically significant result does not establish that the effect was large, clinically important, durable, independently replicated, or applicable outside the tested population and intervention. Reporting standards such as CONSORT help identify the design and result details readers need, but checklist completion is not a validity score.[6]

Prespecification does not by itself make an analysis confirmatory. Check whether the analysis plan controlled for multiple comparisons and whether a confidence interval or p-value is nominal or adjusted. Also distinguish a directly observed outcome from an estimate that depends on assumptions about missing data.[6] The negative and null results guide explains why an inconclusive result should remain visible.

The tested material is part of the conclusion

Every result belongs to the exact test article used in the study. Sequence, modification, salt or counterion, formulation, manufacturing controls, presentation, route, schedule, and storage conditions may matter to comparability. A shared short name does not prove that a paper, reference standard, certificate, and catalog vial describe the same material.

FDA describes an approval review of a drug's data and proposed labeling for its intended use and points readers to Drugs@FDA to identify approved drug products.[7] An ingredient-name match is not evidence that another preparation has the same approved application, formulation, manufacturing controls, or labeling. The approved-drug versus research-material guide explains that boundary.

This primer keeps study evidence separate from product links and purchase controls. The scientific record remains useful even when no commercial relationship can be established.

How studies are selected

Our editorial selection criteria favor a study that adds a distinct and useful reading task. Priorities include:

• pivotal or widely searched randomized trials;
• trials with null or mixed primary outcomes that are often omitted from summaries;
• studies where a surrogate, subgroup, extension, or post-hoc result needs careful labeling;
• trials whose intervention is commonly confused with a different analogue, form, blend, or finished product; and
• records that teach a reusable lesson about design, registration, missing data, uncertainty, or evidence transfer.

A registry listing with no posted results and no primary publication can support a status update, but not a result summary. A useful study record needs enough source material to explain what the evidence can and cannot show.

Review and correction policy

Before an individual study summary is published, the editorial check should compare the live registry, primary paper, protocol and analysis plan when available, corrections or retractions, results-posting state, and any FDA source used for an approval statement. The page should state when its sources were checked and who is responsible for maintaining it. Source checking is not independent scientific or clinical peer review.

If a registry, publication, or official record changes materially, the visible page, source registry, and change log should be updated together. A corrected result does not disappear into an undated rewrite.

Frequently asked questions

Does a ClinicalTrials.gov listing mean the study worked?

No. A listing shows that a record exists. It does not prove completion, posted results, publication, a favorable outcome, independent verification, or FDA approval.

Is a randomized trial automatically reliable?

No. Randomization can reduce important biases, but interpretation still depends on allocation, masking, adherence, missing data, outcome selection, analysis, precision, reporting, and applicability.

What if a secondary outcome is positive but the primary outcome is null?

Both should be reported, with the primary result leading. The secondary finding may be useful, but it does not replace the prespecified primary test and may require multiplicity or exploratory-analysis caution.

Does statistical significance mean the result is important?

Not by itself. Magnitude, confidence intervals, baseline risk, outcome relevance, duration, adverse events, missing data, and replication all affect interpretation.

What if a paper and registry report different results?

First compare the outcome, time point, analysis population, grouping of treatment arms, and version dates. Different estimates may answer different questions. Consult the protocol, analysis plan, and any explanation of changes; if the difference remains unresolved, report the discrepancy rather than choosing the more favorable result.[2][3][6]

Can a trial result be applied to a research vial with the same ingredient name?

Not without a defensible identity and comparability record. A name match does not establish sequence, form, formulation, content, purity, stability, sterility, or finished-product equivalence.

Why are study records not linked to products?

Because a study page documents evidence, not inventory. Keeping it separate from price, availability, ratings, and purchase controls prevents a clinical result from becoming an implied sales claim.

Primary and official sources

1. National Library of Medicine, About PubMed
2. ClinicalTrials.gov, Protocol Registration Data Element Definitions
3. ClinicalTrials.gov, Results Data Element Definitions
4. ClinicalTrials.gov, Disclaimer
5. ICMJE, Clinical Trials Registration
6. SPIRIT–CONSORT Group, CONSORT 2025 expanded checklist
7. FDA, Approved Drugs: Questions and Answers

Review and corrections

Sources checked: September 3, 2026. Please report a correction with the affected statement and supporting source.

Editorial responsibility: Serum Research prepared and maintains this educational primer. No independent scientific, statistical, regulatory, or clinical review is claimed. This guide does not provide medical advice or establish the identity, quality, safety, effectiveness, or equivalence of any catalog material or lot. See how Serum Research reviews evidence or return to the Research Peptide and Compound Library.