Peptides for inflammation: what the evidence supports and what it doesn't

Thymosin alpha-1 has the best human evidence of any anti-inflammatory peptide, and Americans cannot legally buy it. What is for sale here, KPV and BPC-157, is sold as research chemicals not approved for human use, and neither has ever been tested in a person, for inflammation or anything else. What is proven and what is purchasable barely overlap, so the useful next step is a number from your own blood rather than a product page.

What peptides are good for inflammation?

The phrase "anti-inflammatory peptides" covers three product worlds that share nothing but the word peptide. First, a real drug: thymosin alpha-1, a fragment of a thymus protein that modulates T-cell function (T cells are the immune system's coordinators). It is approved by national regulators in more than 30 countries as thymalfasin (brand name Zadaxin), mostly for chronic hepatitis B ⁹, and it is not FDA-approved in the US. Second, research chemicals: KPV, a three-amino-acid fragment (lysine-proline-valine) of the hormone alpha-MSH, and BPC-157, a synthetic gastric-protein fragment, both sold online with mouse evidence behind them and nothing else. Third, groceries: collagen peptides, which are digested food protein, not signaling drugs. LL-37, the body's own antimicrobial peptide, also shows up on vendor lists; whether it calms or fuels inflammation depends on the tissue, and that split record is covered on its own page.

The evidence, compound by compound

Thymosin alpha-1: the only one with human trials

Thymosin alpha-1 has roughly three decades of prescription use abroad as thymalfasin ⁹, but its randomized trials sit in hepatitis and sepsis, not chronic low-grade inflammation. The trial that made its reputation was ETASS, which randomized 361 patients with severe sepsis: 28-day mortality was 26.0% on thymosin alpha-1 versus 35.0% in the control group, a relative risk of 0.74 (treated patients had roughly three-quarters the death risk of controls, though the uncertainty range, 0.54 to 1.02, crosses no effect) that missed the conventional significance threshold at P=0.062, meaning the 9-point gap was promising but small enough that chance could not be ruled out. Two secondary readouts did clear that threshold: the log-rank survival comparison at P=0.049, and in-hospital mortality at 28.7% versus 39.4% (RR 0.73, 0.54 to 0.98, P=0.032) ¹. The authors stated their conclusion carefully:

The use of Tα1 therapy in combination with conventional medical therapies may be effective in improving clinical outcomes in a targeted population of severe sepsis.Wu et al., Critical Care 2013 (the ETASS trial)

Then the signal was tested properly and did not survive. TESTS, published in BMJ in 2025, randomized sepsis patients across 22 centers, double-blind and placebo-controlled, and analyzed 1089 of them: 28-day mortality was 23.4% on thymosin alpha-1 versus 24.1% on placebo, hazard ratio 0.99 (0.77 to 1.27), P=0.93, with no secondary or safety outcome separating the groups either ⁷. So the state of the best-evidenced anti-inflammatory peptide is one borderline result in sepsis, one larger and better-designed trial that did not reproduce it, and no trial at all in chronic inflammation, which is what people are actually searching for. In the US it exists only inside clinical trials or on the gray market; a prescription in Milan does not make a vial from an unregulated US vendor the same product.

KPV and BPC-157 have only mouse evidence

KPV's paper trail starts with Hiltz and Lipton, who showed in 1989 that the tail fragment of alpha-MSH had anti-inflammatory activity in an animal inflammation assay ². The modern anchor is a 2008 mouse study: oral KPV reduced colitis severity, inflammatory cytokines (TNF and IL-6, the immune system's alarm signals), and tissue damage in mice, after entering gut-lining cells through PepT1, a transporter those cells normally use to absorb protein fragments from food ³. That is the entire case: no human study of KPV has been published, for efficacy or for safety. If your interest is gut-specific, the gut healing page covers that territory in depth.

BPC-157 is usually sold for healing rather than inflammation; its anti-inflammatory reputation comes from rodent studies where lower inflammation rides along with faster tissue repair. Every one of those studies is in animals. Rather than re-argue the record here, the injury recovery page carries it:

Both 2026 moves are worth reading precisely, because vendors quote them selectively. FDA removed BPC-157 from Category 2 of its interim 503A list in April 2026, which is a procedural reclassification rather than an authorization, and on 23 to 24 July 2026 the agency's Pharmacy Compounding Advisory Committee voted to recommend adding BPC-157 and KPV, along with four other peptides, to the 503A Bulks List ¹⁰. Those votes bind nobody: HHS has to sign off and FDA has to finish notice-and-comment rulemaking first, so no US pharmacy can legally compound either compound today, and the only supply is gray market.

1

KPVThe gut anti-inflammatory2

LL-37The antimicrobial peptide3

Thymosin Alpha-1The immune modulator

Browse the full reference library →

Doses in circulation vs doses in trials

The figures below are what circulates in practitioner use and vendor copy, reported here so you can see where each number comes from. None is a dosing instruction.

Timeline expectations deserve the same honesty. The mouse colitis experiments ran days; ETASS and TESTS measured mortality at 28 days ^{1 7}; and the week-8 biomarker retest that practitioner protocols use is convention, not a milestone any trial produced.

Injectable peptides ship as freeze-dried powder that has to be mixed with bacteriostatic water and then refrigerated (how reconstitution works), and anything injected demands sterile technique (the hygiene guide covers the red flags). Those guides exist because people do this unsupervised. They are educational references, not instructions to use compounds that are not approved for human use.

Are collagen peptides anti-inflammatory?

No. Collagen peptides are hydrolyzed food protein, and eating protein is not an anti-inflammatory intervention. The best human evidence is a 24-week randomized trial in athletes, 147 randomized and 97 evaluable, where 10 g of collagen hydrolysate daily reduced activity-related joint pain ⁴. That is a real but modest joint-comfort finding, and the joint pain page works through what it does and does not support. No trial shows collagen lowering hs-CRP or any other inflammatory marker the way a drug would. If collagen helps your knees, take the win. It is not treating inflammation.

How KPV works, as far as mice can tell

Track one number instead

An hs-CRP draw runs roughly $25 to $70 direct-to-consumer in most states, before a blood-draw fee, which makes it the cheapest useful thing on this page. Chronic low-grade inflammation is measurable, which is what separates it from the rest of this category. The AHA and CDC set the working bands: below 1 mg/L is low, 1 to 3 is average, above 3 is high, and anything above 10 mg/L is treated as an acute event to diagnose and retest rather than interpret ⁸. Furman and colleagues tie sustained low-grade inflammation to cardiovascular, neurodegenerative, and metabolic disease across the life span ⁵. What nobody has done is measure hs-CRP change on thymosin alpha-1, KPV, or BPC-157 in adults with chronically elevated markers. Reported reductions come from uncontrolled clinic observation, with no published dataset behind any figure.

That absence is the strongest argument for generating your own number, with one caveat nobody selling you a retest mentions: hs-CRP is noisy in the same person from week to week, so a single before-and-after pair can move on its own. The usual clinical fix is two draws about two weeks apart, averaged, as your baseline ⁸. One retest is a hint. A retest that a second draw agrees with is a result.

1. 1Draw a baselinehs-CRP through a clinician, often alongside IL-6 and ESR (erythrocyte sedimentation rate, another inflammation marker). Two draws about two weeks apart, averaged: one number is too noisy to be a baseline. If a value comes back above 10 mg/L, see the acute-cause warning below before interpreting anything.
2. 2Run one change for eight weeksWhatever you and the clinician agreed on: sleep repair, weight loss, treating a hidden dental or gut infection, or a supervised compound. One variable, or the retest is unreadable.
3. 3Retest at week 8Same lab, same conditions. Skip hard training for 48 hours beforehand and retest later if you are sick, since acute anything spikes CRP. A small move is not a move: confirm it with a second draw before you credit it to anything.
4. 4Decide on numbers, not feelSubjective improvement arrives first and can be pure expectation. If the marker has not moved by week 8, the driver has not been addressed, and more of the same is not the answer.

Inflammation genetics: what your variants do and do not predict

Baseline inflammatory tone is partly genetic: genome analyses of more than 200,000 people identified 58 loci (genome locations) associated with circulating CRP ⁶. Partly genetic does not mean peptide-predictive. No study has tested whether any inflammation variant changes response to thymosin alpha-1, KPV, or any other peptide, and any product implying otherwise is ahead of the science. What the variants describe is your baseline, which is useful for one thing: knowing what to measure and how to read it.

• IL6 rs1800795: a promoter variant reported to associate with differences in baseline IL-6 output. The IL-6 axis does appear in the genetics of circulating CRP ⁶, but rs1800795 itself has never been linked to peptide response.
• TNF rs1800629 (the -308 variant): associated with differences in TNF output. It has not been established as a predictor of response even to anti-TNF biologic drugs, where the research money is, and for peptides nothing has been tested.
• CRP rs1205: sits in the CRP gene itself, one of the 58 loci that shift measured C-reactive protein ⁶, so the same inflammatory state can read higher or lower on an hs-CRP test depending on genotype. It changes how your number should be interpreted, not whether any compound will work.

The practical use of a genotype here is measurement strategy. The one variant that changes anything you do this week is CRP rs1205, because it moves the test rather than the disease: a carrier's 2.8 mg/L and a non-carrier's 2.8 mg/L are not the same body. Someone carrying higher-output variants at IL6 and TNF has more reason to establish an hs-CRP and IL-6 baseline before changing anything, because those blood numbers, not the genotype, will say whether an intervention worked. A DNA report can tell you which variants you carry and what to watch; it cannot tell you a peptide will work, and ours does not claim to.

Side effects and unknowns

• Thymosin alpha-1: about three decades of supervised use at licensed doses abroad show mild injection-site reactions and occasional flu-like symptoms in the first weeks, and the 1089-patient placebo-controlled trial found no safety outcome that separated it from placebo ⁷. That record covers supervised patients on licensed product, not gray-market self-use.
• KPV: no human safety data exists, because no human study exists. The rodent work reports no adverse effects, but those were efficacy experiments in mice and were not designed to detect harm. A conservative reading avoids KPV in active or recent cancer until human data arrive, since alpha-MSH biology touches cell-growth signaling.
• BPC-157: animal studies report no serious adverse events, but they were efficacy studies, not safety trials, and nobody collects adverse events from gray-market use. Absence of reports is not evidence of safety.
• Anything injected carries infection risk on top of compound risk; sterile technique is non-negotiable and the injection hygiene guide lists the red flags.

If you are researching vendors regardless, our comparison page lists sources with third-party testing status. Disclosure: links on that page are affiliate links, and we may earn a commission if you buy through them. It changes nothing above; the evidence gaps are what they are.

This article is educational and is not medical advice, diagnosis, or treatment. None of the compounds discussed is FDA-approved to treat inflammation, and several cannot legally be sold for human use in the US. Any decision about testing, treatment, or any compound belongs with a licensed clinician who knows your history.