Peptides for Long COVID: What Is Being Studied—and What Is Only Being Marketed

No peptide has been shown in a completed, controlled clinical trial to treat Long COVID. Some peptide-based medicines and experimental peptides are being studied, but the evidence ranges from registered human trials with no results yet to animal experiments, anecdotes, and advertising. Those categories are not interchangeable.

That distinction matters because people with Long COVID face real disability, few proven treatment options, and a medical system that often fails them. The unmet need explains why people search for possibilities. It does not turn an interesting mechanism, an online testimonial, or a product listing into evidence of benefit or safety.

Peptides promoted for Long COVID sit in very different evidence buckets. A marketed product, a registered human trial, and an animal experiment answer different questions.

Quick answer: As of July 29, 2026, registered trials of tirzepatide and larazotide are active but not recruiting and have no results posted. NIH RECOVER-TLC has announced a semaglutide study, but a public registry record with verifiable endpoints was not located at drafting time. SPIKENET remains preclinical. MOTS-c, KPV, BPC-157, and TB-500 do not have controlled clinical evidence establishing benefit for Long COVID.

Why “peptides for Long COVID” describes four different evidence buckets

The word peptide describes a type of molecule, not a level of evidence. Insulin is a peptide hormone. Semaglutide and tirzepatide are regulated peptide-based medicines. Larazotide is an investigational peptide in a controlled trial. Other compounds sold online as “research peptides” may have little human evidence and uncertain product identity.

For Long COVID, the useful comparison is not “peptide versus non-peptide.” It is:

| Evidence bucket | Examples | What is actually known for Long COVID | |---|---|---| | Unapproved wellness or research peptides | MOTS-c, KPV, BPC-157, TB-500 | No controlled clinical evidence establishes efficacy; product quality may also be uncertain | | Approved medicines studied off-label | Tirzepatide; planned RECOVER-TLC semaglutide study | Approved for other indications, but Long COVID benefit remains unproven | | Investigational human peptide trial | Larazotide (AT1001) | Randomized study is active, with no results posted | | Preclinical candidate | SPIKENET | Animal findings only; no human trial or clinical benefit data |

Marketing often flattens this table into one vague story: peptides may affect inflammation, gut permeability, mitochondria, or viral processes, therefore peptides may help Long COVID. The first half may describe a testable hypothesis. It does not prove the second half.

What is known about MOTS-c, KPV, BPC-157, and TB-500 for Long COVID?

MOTS-c, KPV, BPC-157, and TB-500 are being promoted ahead of Long COVID efficacy and safety evidence. None has a completed randomized controlled trial showing that it improves Long COVID symptoms, function, or quality of life.

The proposed rationales vary:

• MOTS-c is discussed in relation to mitochondrial signaling and metabolism.
• KPV is discussed for anti-inflammatory activity and gut-barrier biology.
• BPC-157 is marketed around tissue repair and gastrointestinal effects.
• TB-500 is marketed around wound healing and tissue recovery.

These are mechanism claims or extrapolations—not demonstrated Long COVID outcomes. A compound can alter a pathway in cells or animals and still fail because it does not reach the right tissue, has the wrong effect in humans, produces harms, or does not meaningfully change how patients feel or function.

The regulatory context also matters. At its July 23, 2026 Pharmacy Compounding Advisory Committee meeting, FDA presented reviews of BPC-157-, KPV-, TB-500-, and MOTS-c-related bulk substances for possible inclusion on the 503A Bulks List. The uses FDA evaluated were ulcerative colitis, wound healing or inflammatory conditions, obesity, and osteoporosis—not Long COVID. An advisory committee discussion is not evidence that a substance treats Long COVID, and committee recommendations are not themselves FDA approval.

FDA has separately identified significant safety concerns for some bulk substances used in compounding. Those concerns include limited safety information, possible immune reactions, peptide-related impurities, and difficulty characterizing active ingredients. Even when a biological idea is plausible, the identity, purity, sterility, and consistency of an unapproved or research-labeled product remain separate questions.

For broader context, see how to evaluate peptide claims online and what a “research use only” label does—and does not—mean.

Are GLP-1 medicines being studied for Long COVID?

Yes, tirzepatide is in a registered controlled trial, and NIH RECOVER-TLC has announced a semaglutide study—but neither establishes benefit yet. These medicines are different from unapproved research peptides because specific tirzepatide and semaglutide products have FDA-reviewed manufacturing, labeling, and evidence for approved indications. Long COVID is not one of those established indications.

Tirzepatide: NCT07128082

The Long COVID Treatment Trial is a randomized, placebo-controlled study of tirzepatide. ClinicalTrials.gov listed it as active, not recruiting when checked on July 29, 2026. The record reports 1,000 enrolled participants, an estimated completion in December 2026, and no posted results.

Its primary outcome is the difference in Fatigue Severity Scale scores between groups at three or 12 months. Secondary outcomes include overall health measured with EQ-5D-5L, post-exertional malaise measured with the DSQ-PEM, and functional capacity measured with FUNCAP.

This design can estimate effects more reliably than testimonials because it includes a comparison group and prespecified outcomes. But “being studied” is not a softer synonym for “works.” Until results are available and analyzed, the trial is an open question.

Semaglutide: RECOVER-TLC

NIH’s RECOVER-TLC initiative has publicly described plans to study semaglutide for Long COVID. At drafting time, however, PeptideBase could not locate a matching public ClinicalTrials.gov record that allowed the study’s current recruitment status, enrollment, endpoints, and results status to be independently verified.

That is an important evidence-literacy point: a program announcement is not a trial result, and it should not be padded with details that are not yet public. The RECOVER-TLC website describes the initiative’s goal of testing candidate therapies to improve symptoms and quality of life. The registry should be checked again when a specific protocol is posted.

Approved product quality also does not prove a new off-label use. It narrows one uncertainty—whether a regulated product meets its approved manufacturing and labeling standards—but does not answer whether the medicine helps Long COVID or which patients might be harmed.

Read more about GLP-1 medicines versus “research peptides”.

What is the larazotide Long COVID trial testing?

Larazotide is being tested in a randomized human trial, but no results are available. Larazotide (AT1001) is an investigational, locally acting synthetic octapeptide designed to influence tight-junction regulation in the gut.

ClinicalTrials.gov record NCT05747534 was active, not recruiting when checked on July 29, 2026. It reports actual enrollment of 107 people ages 7 to 50 and an estimated study completion date of September 30, 2026.

The primary outcome combines adverse-event monitoring with symptom assessment using the Symptom Burden Questionnaire for Long COVID over eight weeks. A secondary outcome examines antigen testing, cytokine profiles, and immune responses in participants ages 7 to 21 over 21 days.

The study is often described as pediatric because its mechanistic secondary analysis focuses on children and young adults, although the registered eligibility range extends through age 50. Calling it simply a pediatric trial can therefore hide part of the protocol.

The gut-barrier hypothesis is biologically interesting: investigators propose that altered intestinal permeability could allow viral antigens to enter circulation and contribute to inflammation. The trial exists because that hypothesis still needs to be tested against placebo on safety and patient outcomes. An ongoing trial is evidence of scientific interest, not evidence of efficacy.

What does the SPIKENET study show?

SPIKENET has preclinical animal evidence, not human Long COVID evidence. A 2024 paper in Viruses studied the peptide in mice infected with mouse hepatitis virus-1, a coronavirus model. Researchers reported changes in inflammation, oxidative stress, tissue edema, and related biological measures.

The animals did not have human Long COVID, and the experiment did not measure whether people experience less fatigue, post-exertional malaise, cognitive dysfunction, pain, or disability. The candidate still requires pharmacokinetic and safety work before human testing can answer clinical questions.

Preclinical studies are valuable for deciding what may deserve further investigation. They are poor tools for predicting whether a therapy will ultimately help patients. Our explainer on what “preclinical” actually means covers that translation gap in more detail.

Why anecdotes and retrospective surveys cannot isolate a peptide effect

Anecdotes can identify possibilities and harms, but they cannot reliably show what caused a change. Long COVID symptoms can fluctuate. People often start several medications, supplements, dietary changes, rehabilitation approaches, or pacing strategies around the same time. Expectations, selection bias, and natural symptom variation further complicate interpretation.

A retrospective survey has similar limits. It can describe what respondents remember and report, but it usually cannot determine:

• whether improvement would have happened without the intervention;
• whether people who responded differ from those who did not;
• whether the product contained the stated ingredient;
• whether another simultaneous intervention caused the change;
• whether worsened symptoms or dropouts were fully captured; or
• whether the measured change was large enough to improve daily function.

These limits do not mean patients are mistaken about their experiences. They mean individual experiences cannot isolate treatment effects. Controlled trials use randomization, comparison groups, defined products, prespecified endpoints, and systematic adverse-event collection precisely because human illness is too messy for testimonials to settle causation.

Evidence check: “It helped me” is meaningful to the person reporting it. It is not the same claim as “this peptide improves Long COVID outcomes more than placebo in a defined population.”

Why a plausible mechanism is not proof of clinical benefit

Mechanism explains why a study might be worth doing; clinical outcomes show whether the idea survives contact with patients. Long COVID likely includes multiple biological subtypes, and inflammation, gut dysfunction, metabolic changes, immune dysregulation, vascular effects, and viral persistence may not play the same role in every person.

Four common marketing shortcuts are:

1. Anti-inflammatory: A laboratory anti-inflammatory signal does not establish the right dose, tissue effect, duration, or net clinical benefit.
2. Gut barrier: Improving a biomarker of permeability does not automatically improve fatigue, cognition, post-exertional malaise, or function.
3. Mitochondrial: Affecting metabolic signaling does not prove that energy limitation in Long COVID is corrected.
4. Antiviral: Interfering with a virus or receptor in a model does not show that a candidate clears clinically relevant viral material or improves established Long COVID.

Biomarkers may help explain a result, but symptom burden, function, quality of life, and harms matter to patients. A pathway diagram is not a patient outcome.

Why approved and unapproved products are not equivalent

Approval status does not answer every safety question, but it changes what is known about the product. An FDA-approved medicine is reviewed for specified indications, manufacturing controls, labeling, and benefit-risk evidence. Its warnings and adverse reactions are documented, and regulators can inspect manufacturing and monitor safety reports.

An unapproved or research-labeled product may create two uncertainties at once:

1. Clinical uncertainty: Does the compound help Long COVID, and what harms can it cause?
2. Product uncertainty: Is the vial accurately labeled, correctly concentrated, sterile, stable, and free of contaminants or related impurities?

Compounding is another distinct category. A compounded drug may be prepared for a patient under applicable law, but it is not an FDA-approved finished product and FDA does not verify it in the same way before marketing. A familiar compound name on a label does not transfer the evidence or quality controls of an approved branded product to an unrelated vial.

“Available online” describes access, not quality. “Research use only” describes a seller’s label, not suitability for human use. Desperation explains risk-taking; it does not make the risk smaller.

What would change the evidence?

Credible evidence would require defined products, appropriate comparison groups, meaningful endpoints, transparent safety reporting, and results that other researchers can scrutinize. The strongest next steps would include:

• completed randomized trials with prespecified Long COVID outcomes;
• clear reporting of participant subgroups and baseline severity;
• measurement of post-exertional malaise and function, not biomarkers alone;
• enough follow-up to detect sustained benefit and delayed harms;
• documentation of the exact product and manufacturing standard used; and
• replication by independent research groups.

Negative and inconclusive results matter too. They prevent plausible mechanisms from quietly becoming permanent marketing claims.

Bottom line

Marketing has outrun the evidence for peptides for Long COVID. That conclusion is not a dismissal of patients or of research. It is a demand that claims match the actual stage of evidence.

Tirzepatide and larazotide are in controlled human studies with no posted results. A RECOVER-TLC semaglutide study has been announced, but public registry details were not verifiable at drafting time. SPIKENET remains preclinical. MOTS-c, KPV, BPC-157, and TB-500 are promoted without controlled Long COVID efficacy evidence, while unapproved products add identity and contamination risks that mechanism claims cannot resolve.

The honest answer today is not “peptides work” or “peptides can never work.” It is that different candidates are at different stages, and none has yet crossed the line from hypothesis to proven Long COVID treatment.

Frequently asked questions

Are any peptides approved specifically for Long COVID?

No. No peptide or peptide-based medicine has FDA approval specifically for Long COVID. Some approved medicines are being investigated off-label, while other candidates remain investigational or preclinical.

Does an ongoing trial mean a peptide probably works?

No. A trial means researchers are testing a question. Benefit, lack of benefit, and harm all remain possible until results are completed, analyzed, and reported.

Are GLP-1 medicines the same as research peptides?

No. GLP-1 medicines such as specific approved semaglutide and tirzepatide products have regulated manufacturing and evidence for approved indications. That does not prove they treat Long COVID. Products sold as research peptides may lack both clinical evidence and reliable product-quality information.

Can a patient survey prove that a peptide helps?

No. Surveys can detect patterns worth investigating, but retrospective, self-selected reports cannot reliably separate the intervention from symptom fluctuation, concurrent treatments, expectations, or selection bias.

Why study larazotide if benefit is unproven?

Its gut-barrier mechanism and earlier observations provide a testable rationale. The randomized trial is needed to determine whether that rationale translates into acceptable safety and meaningful symptom improvement.

Is SPIKENET a Long COVID treatment?

Not at present. SPIKENET is a preclinical research candidate with animal data. It has not been shown to be safe or effective for Long COVID in humans.

---

This article is for general education and evidence literacy. It does not provide medical advice, treatment instructions, dosing, monitoring, sourcing, or purchasing guidance.

Sources

1. ClinicalTrials.gov. NCT07128082: The Long COVID Treatment Trial. Registry checked July 29, 2026.
2. ClinicalTrials.gov. NCT05747534: AT1001 for the Treatment of Long COVID. Registry checked July 29, 2026.
3. NIH RECOVER-TLC. RECOVER-TLC initiative. Checked July 29, 2026.
4. Hayn M, et al. SPIKENET, a Novel Peptide, Attenuates Acute Lung Injury in a Mouse Coronavirus Model. Viruses. 2024;16(6):838.
5. U.S. Food and Drug Administration. Certain Bulk Drug Substances for Use in Compounding that May Present Significant Safety Risks. Checked July 29, 2026.
6. U.S. Food and Drug Administration. July 23–24, 2026 Pharmacy Compounding Advisory Committee meeting. Checked July 29, 2026.
7. Griffis M. Experimental peptides are being advertised for Long COVID. But the evidence is trailing behind. The Sick Times. July 14, 2026.