evipedia.ai
← Back

Cathelicidin LL-37 for Health & Longevity

Evidence Review created on 11/05/2026 using AI4L / Opus 5.5

Also known as: LL-37, LL37, FALL-39, hCAP18, hCAP-18, Human Cationic Antimicrobial Protein 18, Cathelicidin Antimicrobial Peptide, CAMP, Ropocamptide

Motivation

Cathelicidin LL-37 is a short protein fragment that the human body makes itself, mainly in white blood cells and in the surfaces that meet the outside world — skin, saliva, gut and airways. Its basic job is to break open the outer membranes of microbes while also acting as a signal to the immune system.

The fragment was first described in the mid-1990s as a possible human-made antibiotic, and the later discovery that vitamin D switches on the gene behind it drew wide attention. Synthetic copies have since been applied directly to long-standing leg and foot wounds and injected into tumours, and the molecule has more recently appeared in the market for peptides sold for self-experimentation. In parallel, researchers have studied what happens where the body makes unusually large amounts of it.

This review examines the human evidence on cathelicidin LL-37 in the context of health and longevity: how it is thought to act, which outcomes have actually been measured in people, what harms have been described, the regimens used in studies, how material quality is judged, and which questions the evidence leaves open.

Benefits - Risks - Protocol - Conclusion

This section lists in-depth, high-level treatments of cathelicidin LL-37 from sources outside the dedicated reference, systematic-review and database sections below.

Of the six priority platforms, only Lifespan.io offered listable content. Peter Attia’s two LL-37 articles (October 2026) are members-only and could not be verified, so they are not listed. Life Extension names the peptide only in passing, in two sentences of its 2010 vitamin D and skin article Optimal Skin Protection with Vitamin D, and FoundMyFitness only in single-sentence news notes; neither provides a high-level overview of the topic. Searches of Rhonda Patrick’s, Andrew Huberman’s and Chris Kresser’s sites returned nothing further.

Grokipedia

  • Cathelicidin antimicrobial peptide

    The primary dedicated page, covering gene regulation, proteolytic processing, the human LL-37 ortholog and therapeutic challenges; useful for how the peptide differs across species.

Examine

  • Cathelicidin

    Examine’s outcome page on the body’s own peptide, framing it as an outcome that diet, lifestyle and supplements may influence; it does not cover applied synthetic peptide.

ConsumerLab

No ConsumerLab article or product review exists for cathelicidin LL-37; searches for both “LL-37” and “cathelicidin” returned no relevant results. ConsumerLab tests retail supplements and foods, and no consumer product containing this peptide is sold through those channels, which is the likely reason for the absence.

Systematic Reviews

These are the systematic reviews and meta-analyses that bear most directly on cathelicidin LL-37, covering both its claimed effects and its principal risks.

The literature provides systematic reviews on both sides of the trade-off — infection defence on the claimed-effect side, inflammatory skin disease and skin cancer on the risk side — but none of them pools trials of the applied peptide itself, so its wound-repair effect remains unrepresented at this evidence level.

Mechanism of Action

The CAMP gene (cathelicidin antimicrobial peptide, the only human cathelicidin gene) encodes an 18-kilodalton precursor, hCAP18, stored in neutrophils (first-responding white blood cells), skin, gut, airway and saliva. Proteinase 3 in neutrophils and kallikrein 5 (KLK5, a protein-cutting enzyme of the outer skin layer) cut it to the 37-residue, positively charged peptide, per a review (Svensson 2025). Transcription is driven by the vitamin D receptor (VDR, which active vitamin D binds) (Gombart 2005, Liu 2006), and by histone deacetylase inhibitors (agents that unpack DNA so a gene is read more easily) such as butyrate (Raqib 2012).

Two explanations compete. The direct one: positive charge draws it to negatively charged microbial membranes, which it permeabilises. The host-directed one, per reviews: it neutralises lipopolysaccharide (LPS, the endotoxin of gram-negative bacteria), draws in immune cells and signals via FPR2 (a receptor guiding immune cells) and EGFR (a receptor driving cell growth) (Svensson 2025, Kuroda 2015); wound-trial authors propose macrophage VEGF (vascular endothelial growth factor, driving new blood vessels) aids repair (Miranda 2023). Better granulation (new repair tissue) without extra fall in bacterial counts favours the host-directed route.

No human pharmacokinetic study exists: half-life, distribution and clearance are unknown. Breakdown is by host and microbial proteases, not cytochrome P450 enzymes (the liver’s main drug-metabolising system); glycosaminoglycans (long sugar chains of connective tissue) strongly inhibit activity (Barańska-Rybak 2006). Selectivity is broad, engaging scavenger receptors SR-B1 and CD36 (which pull fats into immune cells) and the LDL receptor (low-density lipoprotein, the main cholesterol carrier) (Nakamura 2024).

Historical Context & Evolution

The peptide entered the literature in 1995 as FALL-39, a cysteine-free sequence predicted from bone-marrow and testis messenger RNA, proposed as a human counterpart to antibiotic peptides known in pigs (Agerberth 1995). The original intent was pharmaceutical: an endogenous antibiotic bacteria might struggle to resist.

Interest in health optimisation arrived through vitamin D. In 2005 the gene was shown to be a direct vitamin D receptor target, strongly upregulated in myeloid cells (immune cells including neutrophils and macrophages) by active vitamin D (Gombart 2005), and in 2006 activation of Toll-like receptors (microbe-detecting immune sensors) on human macrophages was shown to drive the same vitamin D-dependent antimicrobial response, the effect failing in low-vitamin-D serum (Liu 2006). That pair linked a nutrient to infection resistance and is the main reason the peptide is discussed in longevity circles.

The picture complicated rather than resolved. In 2007, high cathelicidin and serine protease (a family of protein-cutting enzymes) activity were shown to produce rosacea-like inflammation, with abnormally processed forms in rosacea skin (Yamasaki 2007). Drug development ran in parallel: two wound trials run by the developers Pergamum and Promore Pharma, first-in-human (Grönberg 2014) and phase IIb (Mahlapuu 2021), and an intratumoral melanoma study opened in 2015 (NCT02225366). What changed was not that the early findings were overturned — the antimicrobial and repair activities remain reproducible in the laboratory — but that the same activities showed a damaging face at higher local concentrations, and that the clinical signal proved smaller than the mechanism suggested. Both readings remain live.

Expected Benefits

High 🟩 🟩 🟩

No benefit reaches High: the supporting human evidence consists of small single-centre randomized trials and a post hoc subgroup (an analysis chosen after the results were known) of one larger trial, with no outcome replicated by independent research groups.

Medium 🟩 🟩

No benefit reaches Medium: each positive human finding is contradicted by a larger randomized trial of the same intervention, uncontrolled, or from trials of agents that induce the body’s own peptide rather than of the peptide itself.

Low 🟩

Faster Healing of Hard-to-Heal Skin Ulcers ⚠️ Conflicted

Topical synthetic LL-37 is applied to chronic venous leg and diabetic foot ulcers. Two small trials were positive; the only adequately powered one was null. All enrolled older patients with vascular or diabetic disease, not healthy adults. Net reading: the strongest trial found no effect.

Magnitude: Confirmed wound closure 26.5% with 0.5 mg/mL versus 25.3% with placebo in 144 analysed patients (not significant); in the post hoc subgroup with wounds ≥10 cm², 28.1% versus 8.1% (odds ratio, the relative odds of closure, 4.45; p = 0.046, the probability of so large a difference arising by chance; no confidence interval, or CI, the range of plausible true values, reported) (Mahlapuu 2021, a trial designed, funded and authored by employees of Promore Pharma, which was developing the peptide). Granulation index change at day 21 (mean ± standard deviation, the typical spread around the mean): 0.64 ± 0.30 with LL-37 cream versus 0.30 ± 0.26 with placebo, p = 0.006 (Miranda 2023); healing-rate constant about six-fold placebo at 0.5 mg/mL, p = 0.003 (Grönberg 2014, authored from Pergamum AB, the peptide’s then-developer).

Reduction of Wound Bacterial Burden ⚠️ Conflicted

Direct microbial killing is the oldest claim, holding in laboratory assays against bacteria, fungi and viruses. The one randomized trial culturing wounds found no greater fall than placebo; the species present degrade the peptide or shelter in biofilm (a protective bacterial film). Net reading: no demonstrated effect on bacterial counts.

Magnitude: Median aerobic colony count over 28 days fell by 1,600 colony-forming units per swab with LL-37 cream versus 10,412 with placebo — a larger fall in the placebo arm, not statistically significant; target-ulcer infection affected 1 patient (2.1%) on 0.5 mg/mL, 2 (4.1%) on 1.6 mg/mL and 3 (5.9%) on placebo in the larger trial (Miranda 2023, Mahlapuu 2021).

Faster Recovery from Bacterial Infection When the Body’s Own Peptide Is Induced ⚠️ Conflicted

Vitamin D3 and butyrate-class histone deacetylase inhibitors raised peptide production in tuberculosis and shigellosis (bacterial dysentery) trials; inducers, not the peptide, were tested. One trial’s week-4 gain in clearing bacteria from sputum (coughed-up mucus) failed in eight pooled trials measuring time to clearance (Jolliffe 2019). Net reading: no consistent benefit.

Magnitude: Sputum culture conversion (no tuberculosis bacteria growing from the sample) at week 4 was 71% (46/65) with phenylbutyrate plus vitamin D3, 64.4% (38/62) with vitamin D3 alone and 43.7% (27/64) with placebo (full-text figures; the abstract gives 61.3% and 42.2%); adjusted odds ratios 3.42 (95% CI 1.64–7.15, p = 0.001) and 2.20 (95% CI 1.07–4.51, p = 0.032); phenylbutyrate alone was null at 46.6% (27/58), p = 0.62 (Mily 2015). Across 1,850 participants in eight randomized trials, vitamin D did not change time to culture conversion overall (adjusted hazard ratio, the relative rate of conversion, 1.06; 95% CI 0.91–1.23), speeding it only in multidrug-resistant disease (adjusted hazard ratio 13.44; 95% CI 2.96–60.90) (Jolliffe 2019). Butyrate enemas raised stool LL-37 at days 4 and 7 and reduced stool inflammatory cells against placebo (Raqib 2012).

Antitumour Immune Activation ⚠️ Conflicted

Intratumoral injection is proposed to recruit immune cells against tumours. Human data come from three patients with metastatic melanoma in an uncontrolled trial; one withdrew for lack of efficacy. A review describes tumour killing and promotion of invasion in cell and animal work (Kuroda 2015). Net reading: direction is tissue-dependent.

Magnitude: Immune-related complete or partial response recorded in 3 of 3 dosed patients (2 at 250 µg, 1 at 500 µg per tumour) in the posted registry results, no control group, so no comparison rate exists (NCT02225366).

Speculative 🟨

Neutralisation of Bacterial Endotoxin and Damping of Inflammatory Signalling

The peptide binds lipopolysaccharide and alters cytokine output, the proposed basis for claims about chronic low-grade inflammation. Basis is mechanistic only: cell and biochemical assays in a review (Svensson 2025); no human outcome data.

Higher Circulating Peptide Levels with Vitamin D Repletion ⚠️ Conflicted

Blood peptide level is an unvalidated biomarker. A randomized sepsis trial raised it with cholecalciferol (Quraishi 2015); a randomized kidney-disease trial found no change (Alvarez 2013). Net reading: inconsistent.

Benefit-Modifying Factors

  • Species-restricted gene regulation: The vitamin D response element in the human promoter is primate-specific, so rodent findings on vitamin D-driven induction do not transfer; human-specific cholesterol handling behaves the same way (Gombart 2005, Nakamura 2024).

  • Vitamin D receptor and HLA variants: Receptor-pathway variation plausibly alters induction, and the peptide binds the psoriasis risk allele HLA-C*06:02 (a tissue-type gene variant), which may shape the immune response to it (Mabuchi 2016).

  • Baseline vitamin D status: Among 19 adults, plasma hCAP18 tracked 25-hydroxyvitamin D at or below 32 ng/mL (r = 0.81, where 1 is perfect) but not above (r = 0.19), suggesting deficiency matters most; cross-sectional only (Dixon 2012, from supplement maker USANA).

  • Baseline peptide level in the target tissue: Wound-fluid LL-37 was uniformly low in diabetic ulcers before treatment, and higher serum LL-37 predicted faster venous-ulcer healing, suggesting the deficient tissue is where supplementation could matter (Miranda 2023, Krejner 2017).

  • Wound size as an effect modifier: Significant healing differences appeared only where baseline wound area was at least 10 cm²; below that threshold no difference was seen, and measurement error is larger (Mahlapuu 2021).

  • Sex: No sex difference has been reported. The phase IIb trial analysed 83 women and 65 men without separating outcomes, and premenopausal women were not enrolled, so female reproductive-age data are absent (Mahlapuu 2021).

  • Pre-existing conditions: Diabetic foot ulcers start with low local peptide, leaving a deficit to correct (Miranda 2023); rosacea skin already carries excess, abnormally processed peptide, leaving none (Yamasaki 2007).

  • Age: The venous-ulcer trial’s mean age was 67.8 years (full text; abstract: 67.6), covering the older end of the target range; the diabetic-foot trial capped entry at 60 (mean 53). No trial has tested healthy adults (Mahlapuu 2021, Miranda 2023).

Potential Risks & Side Effects

High 🟥 🟥 🟥

No risk reaches High: the only adverse-event signal for the applied peptide, more local reactions at the highest strength, comes from one placebo-controlled trial and has not been replicated by an independent group; every other concern rests on observational, animal or in-vitro data.

Medium 🟥 🟥

No risk reaches Medium: the one controlled adverse-event signal, local reactions at the highest strength, is outweighed by a larger placebo-controlled trial, and every other concern rests on indirect, animal or in-vitro data.

Low 🟥

Application-Site Reactions ⚠️ Conflicted

Redness, swelling and scaling are the commonest reactions. 3.2 mg/mL caused more frequent, more severe local reactions than lower strengths (Grönberg 2014). The larger phase IIb trial found no consistent excess at 0.5 and 1.6 mg/mL (Mahlapuu 2021). Net reading: irritation is confined to the abandoned highest strength.

Magnitude: Redness at the ulcer in 42.6% and on adjacent skin in 54.1% of 148 treated patients, pooled across both peptide arms and placebo; the report gives no arm-specific rates and states no consistent between-arm difference, and severe reactions affected 5 patients (3.4%) or fewer at any visit. Serious adverse events numbered 4 with 0.5 mg/mL (n = 48), 7 with 1.6 mg/mL (n = 49) and 1 with placebo (n = 51), none judged related, with no deaths (Mahlapuu 2021). One of 13 patients on the cream had mild irritant dermatitis attributed to wound exudate (fluid seeping from the wound) rather than the cream; the report gives no placebo-arm figure (Miranda 2023).

Amplification of Inflammatory and Autoimmune Skin Disease

Excess skin cathelicidin is tied to disease: abnormally processed fragments drive rosacea-like inflammation, and peptide–DNA complexes become an immune target in psoriasis and cutaneous lupus (an autoimmune skin disease). Evidence is indirect: it concerns the body’s own peptide in diseased skin, not peptide applied by choice.

Magnitude: Direction only, under disease conditions: lesional cathelicidin expression falls as rosacea improves on treatment and rises with disease activity, and the cathelicidin pathway was among the four pathways with the greatest support across 38 rosacea studies reporting 119 biomarkers (Di Nardo 2016, a trial with Galderma Laboratories employees among its authors and sponsors; Geng 2024). The literature reports no outcome figure for disease triggered by applying the peptide.

Tumour Promotion ⚠️ Conflicted

Neutralising the peptide slowed ovarian tumour growth in mice (Coffelt 2009); the FDA cites nonclinical protumorigenic findings (FDA compounding safety-risk list); a review describes tumour-promoting receptor routes but anticancer effects in other cell lines (Kuroda 2015). Human data: three injected melanoma patients (NCT02225366). Net reading: direction is tissue-dependent.

Magnitude: Posted results of that single-arm trial record squamous cell carcinoma (a skin cancer of the outer skin cells) in 1 of 3 dosed patients and actinic keratosis (a rough, sun-damaged skin patch that can become cancerous) in 1 of 3; the trial had no control group, so no comparison rate exists and attribution to the peptide cannot be separated from the patients’ metastatic melanoma and sun-damaged skin (NCT02225366). No controlled study has measured tumour incidence under the peptide.

Speculative 🟨

Pro-Atherogenic Cholesterol Loading in Macrophages ⚠️ Conflicted

The peptide loads macrophages with LDL; peptide-expressing mice grew larger plaques (Nakamura 2024), but T cells reactive to the mouse peptide reduced plaque (Chernomordik 2020). Basis: animal and unvalidated biomarker data. Net reading: unresolved.

Immune Response Against the Administered Peptide

Repeated synthetic-peptide administration can provoke neutralising antibodies or reactions. Compounded preparations may pose immunogenicity risk for certain routes (FDA compounding safety-risk list). Basis is mechanistic and regulatory only; no human immunogenicity data exist.

Male Reproductive Toxicity Signal

Nonclinical research findings suggest detrimental effects on male reproduction (FDA compounding safety-risk list). The agency names no species, dose, route or endpoint, with no human data; the basis is this regulatory statement alone.

Impurities and Mischaracterised Material in Non-Pharmaceutical-Grade Peptide

Compounded preparations may carry complexities regarding peptide-related impurities and active-ingredient characterisation (FDA compounding safety-risk list). Research-grade peptide faces no equivalent control. Basis is regulatory assessment; no harm from a specific impurity is documented.

Risk-Modifying Factors

  • Applied concentration: Efficacy follows a bell-shaped dose-response while irritation rises with strength. Local reactions were more frequent and severe at 3.2 mg/mL, and 1.6 mg/mL healed worse than 0.5 mg/mL, so higher strengths add irritation without benefit (Grönberg 2014).

  • Route of administration: The FDA names immunogenicity risk specifically for certain routes (FDA compounding safety-risk list); trials of the synthetic peptide used topical application or intratumoral injection, and one oral trial used peptide-secreting bacteria (Zhao 2023), so no systemic exposure data bound the risk.

  • HLA-C*06:02 carriage: The peptide binds this psoriasis risk allele, providing a mechanistic route by which carriers could mount a stronger autoreactive T-cell response; no clinical data stratify risk by genotype (Mabuchi 2016).

  • Pre-existing inflammatory skin disease: Psoriasis, rosacea and cutaneous lupus already involve excess lesional peptide and interferon-driven inflammation, so added peptide acts on an amplified rather than a deficient pathway, as a review describes (Pahar 2020).

  • Baseline lipids and cardiovascular status: The cholesterol-loading mechanism scales with available LDL, and self-reactive T-cell responses were detected in acute coronary syndrome (sudden blockage of heart blood flow), making established atherosclerosis the plausible high-risk context (Nakamura 2024, Chernomordik 2020).

  • Sex: The one reproductive signal on record is male-specific and nonclinical; no trial has reported sex-stratified adverse events, and premenopausal women were absent from the phase IIb trial (FDA compounding safety-risk list, Mahlapuu 2021).

  • Age: Safety data come almost entirely from patients aged 33–94 years with vascular disease, where background rates of erysipelas (a spreading bacterial skin infection), thrombosis and fracture are high and confound attribution; younger adults are unstudied (Mahlapuu 2021).

  • Skin cancer history at the application site: A systematic review found cathelicidin elevated in basal cell carcinoma, so sun-damaged or tumour-bearing skin is where tumour promotion is least ignorable; no study has tested applying the peptide to such skin (Fijałkowska 2023).

Key Interactions & Contraindications

Documented interactions:

  • High-dose cholecalciferol (vitamin D3, available over the counter): plasma LL-37 rose 30% on 400,000 IU versus a 17% fall on placebo by day 5 in sepsis (p = 0.04) (Quraishi 2015); no change in early kidney disease (Alvarez 2013). Consequence: inconsistent increase in endogenous peptide.

  • 4-Phenylbutyrate (2 × 500 mg daily): raised peptide content in monocyte-derived macrophages against placebo at week 12 (p = 0.034) and accelerated intracellular bacterial clearance. Consequence: additive induction through histone deacetylase inhibition (Mily 2015).

  • Sodium butyrate, rectal (80 mM twice daily): induced peptide expression in rectal lining and raised stool levels at days 4 and 7 versus saline. Consequence: additive mucosal induction; the same route as fermentable-fibre metabolites (Raqib 2012).

  • Doxycycline (40 mg modified-release, a tetracycline-class antibiotic): cathelicidin expression and protein fell over 12 weeks versus placebo as rosacea improved. Consequence: opposing effect, reducing endogenous peptide (Di Nardo 2016).

  • Azelaic acid 15% gel: reduced cathelicidin and kallikrein 5 messenger RNA in treated rosacea skin and inhibited the protease directly. Consequence: opposing effect, reducing local peptide generation (Coda 2013).

Theoretical interactions (inferred from mechanism, not studied in humans):

  • Heparin and heparin-like anticoagulants (blood-thinning drugs; unfractionated heparin, enoxaparin): inferred from electrostatic sequestration — anionic sugar chains bind the cationic peptide. Antibacterial activity was strongly inhibited in heparin plasma and glycosaminoglycan-rich wound fluid in vitro (Barańska-Rybak 2006).

  • Anionic wound dressings and fillers (calcium alginate, carboxymethylcellulose, hyaluronate gels): inferred from the same charge-binding mechanism as glycosaminoglycans. Consequence: loss of applied activity. Separating application from dressing contact is untested.

  • Topical and systemic corticosteroids (anti-inflammatory steroid drugs; hydrocortisone cream over the counter, prednisone, clobetasol): inferred from corticosteroid suppression of epidermal cathelicidin production, shown in mice with worse cutaneous infection (Aberg 2007).

  • Co-administered antibiotics (polymyxin B, colistin, daptomycin): inferred from shared membrane targeting; a review (Voronko 2025) calls combining cathelicidins with conventional antibiotics a promising strategy against resistance. No human study has tested the pairing, and no antagonism has been shown.

  • Proteolytic wound environments (infection with Pseudomonas aeruginosa or Staphylococcus aureus): inferred from low proteolytic stability noted in a review (Voronko 2025) — bacterial and host proteases degrade the applied peptide, plausibly abolishing effect in the wounds that most need it.

  • Statins and other LDL-lowering agents (atorvastatin, rosuvastatin, ezetimibe): inferred from the cholesterol-loading mechanism, where uptake scales with available LDL. Consequence: possibly less peptide-driven cholesterol loading of artery-wall macrophages; whether lowering LDL mitigates the effect has not been studied (Nakamura 2024).

Populations who should avoid Cathelicidin LL-37:

  • Active psoriasis, rosacea or cutaneous lupus erythematosus on or near the application site (active lesions, not history alone): inferred from excess lesional peptide, near 300 µM in psoriasis per a review (Svensson 2025), and rosacea-type fragments inflaming mouse skin (Yamasaki 2007); no human data.

  • Skin cancer at the application site, current or treated (basal cell carcinoma, squamous cell carcinoma, melanoma), or field actinic damage: FDA cites nonclinical protumorigenic findings; a systematic review finds cathelicidin elevated in basal cell carcinoma (FDA compounding safety-risk list, Fijałkowska 2023).

  • Men attempting conception: inferred from the FDA’s statement that nonclinical research findings suggest detrimental effects on male reproduction; the agency names no species, dose or endpoint, and there are no human data (FDA compounding safety-risk list).

  • Pregnancy and breastfeeding: inferred from cytotoxicity to human cells at 1–10 µM noted in a review, a risk to developing tissue if absorbed, and the FDA’s nonclinical male reproductive signal; no female, fetal or lactation data exist (FDA compounding safety-risk list, Svensson 2025).

Risk Mitigation Strategies

Doses, timings and thresholds below follow common practice unless cited.

  • Lower of the tested concentrations: Protocols stay at 0.5 mg/mL rather than 1.6 or 3.2 mg/mL; the highest strength caused more frequent and more severe local reactions and healed no better, so escalation adds irritation without benefit (Grönberg 2014).

  • Topical route only: Published synthetic-peptide use is topical or intratumoral; injected and systemic routes carry the immunogenicity risk the FDA names for certain routes, with no systemic human exposure data to bound it (FDA compounding safety-risk list).

  • Dosing by measured area: Trial dosing was 25 µL per cm² of treated area, measured rather than estimated; area errors drove both over- and under-dosing in the phase IIb trial (Mahlapuu 2021).

  • Single-site trial application first: Treating one site for one to two applications before extending to others brings out redness and scaling — the commonest reactions in the trials — before they affect a large area.

  • Distance from inflamed and sun-damaged skin: Application away from active psoriasis, rosacea and actinic (sun-damaged) or previously cancerous skin avoids sites where added peptide acts on an already amplified inflammatory or proliferative pathway (Yamasaki 2007).

  • Stopping threshold for spreading redness: Redness extending beyond the treated margin is the point at which use is stopped and medical assessment sought, since erysipelas was among the more common serious events in the trial population (Mahlapuu 2021).

  • Certificate of analysis: The trial material’s specification — identity by liquid chromatography-mass spectrometry, content within 90–110% of label, total impurities near 0.5% — is the benchmark that limits impurity exposure (Mahlapuu 2021).

  • Course length within the studied range: Courses ran 4 to 13 weeks in the trials. Beyond that, neither efficacy nor safety has been observed, and antibody formation against the peptide is unmeasured.

  • Separation from anionic dressings: A cleansed, dried surface under a non-anionic film dressing rather than alginate keeps the peptide active, since anionic glycosaminoglycans bind it and strongly inhibit its activity in vitro (Barańska-Rybak 2006).

Therapeutic Protocol

No approved product or clinical guideline exists; the regimens below are those used in published human trials. Parameters without a citation — timing within the day, cycling and titration steps — reflect common practice rather than tested protocol.

  • Standard topical regimen: 0.5 mg/mL in 10.5% polyvinyl alcohol gel, 25 µL per cm² of wound, applied twice weekly every 3 ± 1 days for up to 13 weeks, on cleansed dry tissue under compression (Mahlapuu 2021).

  • Cream alternative: 0.5 mg/g cream, 0.025 mL per cm², twice weekly for 4 weeks after saline cleansing and sharp debridement (cutting away dead tissue), covered with a transparent film dressing (Miranda 2023).

  • Dose-finding range: The first-in-human trial compared 0.5, 1.6 and 3.2 mg/mL twice weekly for 4 weeks after a 3-week placebo run-in; 0.5 mg/mL performed best and 3.2 mg/mL worst (Grönberg 2014).

  • Intratumoral regimen (oncology research only): 250 µg per tumour, escalating to 500 µg, injected into two to four lesions once weekly for up to 8 weeks (NCT02225366).

  • Endogenous-induction alternative: 5,000 IU vitamin D3 daily, with or without 2 × 500 mg 4-phenylbutyrate daily, raised peptide levels in tuberculosis; a single 400,000 IU cholecalciferol dose raised plasma levels by day 5, 200,000 IU not clearly (Mily 2015, Quraishi 2015).

  • Competing approaches are not ranked: Compression therapy alone remains standard care for venous ulcers and healed a quarter of patients in the phase IIb trial; the peptide was tested as an addition to it, never against it (Mahlapuu 2021).

  • Sources differ on concentration: The gel trials used 0.5 mg/mL; the cream trial used 0.5 mg/g; the dose-finding trial also tested 1.6 and 3.2 mg/mL, with the highest strength abandoned (Grönberg 2014, Miranda 2023).

  • Originating groups: The topical wound approach was developed by Pergamum AB and its successor Promore Pharma with Karolinska and Linköping investigators; the intratumoral approach came from MD Anderson Cancer Center (Grönberg 2014, NCT02225366).

  • Best time of day: Untested. Every trial tied application to the routine dressing change rather than to a time of day, so no circadian preference can be derived.

  • Half-life and dose frequency: No human half-life has been published. The twice-weekly interval was chosen to match dressing-change practice, not peptide kinetics, and the peptide is not given as a single dose (Mahlapuu 2021).

  • Genetic considerations: No pharmacogenetic variant guides dosing. The vitamin D receptor pathway governs the endogenous route, and HLA-C*06:02 binds the peptide, but neither has been used to select a dose (Mabuchi 2016).

  • Sex-based differences: None reported. The phase IIb trial pooled 83 women and 65 men and excluded premenopausal women, so no sex-specific regimen can be derived (Mahlapuu 2021).

  • Age-based differences: No age adjustment was used across ages 33–94 years, including the older end of the target range; dosing was by treated area alone (Mahlapuu 2021).

  • Baseline biomarkers: Low local peptide levels and low vitamin D mark the groups with room to respond; serum peptide level predicted healing rate and has been proposed for selecting candidates (Krejner 2017).

  • Pre-existing conditions: Trials required adequate arterial supply, excluded deep or systemically infected wounds and excluded end-stage kidney failure, so the tested regimen applies only to superficial, uninfected or mildly infected tissue (Miranda 2023).

Discontinuation & Cycling

  • Short course, not lifelong: Every human regimen was a fixed course of 4 to 13 weeks ending at wound closure or at the protocol end; no trial has administered the peptide for longer or continuously (Mahlapuu 2021).

  • No withdrawal effects reported: Across the 16-week post-treatment observation period, no safety signal and no between-arm difference emerged after stopping, and no rebound phenomenon was described (Mahlapuu 2021).

  • No taper used or needed: Treatment stopped abruptly at the final application in all trials. Because, as a review notes, the peptide is degraded locally by proteases rather than accumulating, a taper has no mechanistic rationale (Voronko 2025).

  • Relapse after stopping is possible: Of 39 patients whose ulcers had closed and who attended follow-up, 6 had reopened wounds — 2 on the higher dose, 4 on placebo — attendance being too low for comparison (Mahlapuu 2021).

  • Post-stop care continued: After closure, four visits over two weeks maintained standard wound care and compression without further peptide, to stop the wound reopening (Mahlapuu 2021).

  • Cycling untested: No trial has interrupted and resumed treatment, so whether cycling preserves effect or reduces irritation is unknown; the bell-shaped dose-response offers no basis for inferring it (Grönberg 2014).

Sourcing and Quality

  • No approved product exists: The peptide is investigational, with no FDA-approved product. Trial material was described as not approved by the FDA and not commercially available, used for research only (NCT02225366).

  • Trial-grade specification: Drug substance was made by AmbioPharm using solid-phase synthesis with a reverse-phase chromatography purification step; batches met 90–110% content with total impurities of 0.5% and positive identity by chromatography-mass spectrometry (Mahlapuu 2021).

  • What to require of any material: A certificate of analysis showing identity by mass spectrometry, purity by chromatography, the impurity total, and sterility for anything applied to broken skin — the four attributes the trial material was released against (Mahlapuu 2021).

  • Compounding pharmacies are not a route: The FDA placed cathelicidin LL-37 in category 2 of its interim compounding policies, citing immunogenicity and characterisation concerns, before the nomination was withdrawn; the agency now lists it among nominations withdrawn from category 2 (FDA compounding safety-risk list).

  • Research-peptide suppliers carry no equivalent control: Material in the second trial came from the research supplier Isca Biochemicals and was validated by the investigators’ own chromatography before use — a step no retail purchaser performs (Miranda 2023).

  • No third-party testing programme covers it: ConsumerLab’s searches return nothing for this peptide, and no independent verification scheme tests peptides sold for self-experimentation, so label claims are unverified.

  • Formulation and storage matter: Cream at 0.5 mg/g had a projected 99-month shelf life at 4 °C and 75 months at 28 °C (12-week accelerated test) but melted at 40 °C (Miranda 2023); trial gel was mixed with diluent just before use (Mahlapuu 2021).

Practical Considerations

  • Time to effect: Tissue-level change was measurable at the first weekly assessment, with granulation differences from day 7 and the treated group reaching full granulation by day 14. Complete wound closure took roughly 83–90 days (Miranda 2023, Mahlapuu 2021).

  • Pitfall — assuming more is better: The dose-response is inverted. The lowest of three concentrations healed best and the highest healed worst while causing the most irritation, so escalating is counterproductive (Grönberg 2014).

  • Pitfall — expecting an antibiotic: The marketed antimicrobial property did not translate: colony counts fell no more than with placebo in the only trial that cultured treated wounds, and common wound species degrade or evade the peptide (Miranda 2023).

  • Pitfall — treating small lesions: No healing difference appeared in wounds under 10 cm², where measurement error is also largest, so small-area use has no supporting signal (Mahlapuu 2021).

  • Pitfall — injecting it: No human study has given the peptide under the skin or into a vein. Subcutaneous use, common in peptide self-experimentation, carries the immunogenicity concern the FDA names and no exposure data (FDA compounding safety-risk list).

  • Regulatory status: Not FDA-approved for any indication (NCT02225366), and not available through legitimate compounding after its category 2 placement and withdrawn nomination (FDA compounding safety-risk list). Clinical use occurs only inside trials.

  • Cost and access: No priced product exists; synthesis cost is itself cited in a review as a barrier to development, and access outside trials means research-grade material of unverified quality (Voronko 2025).

  • Payer incentives shape the comparison: Compression is standard care for venous ulcers (Mahlapuu 2021) and costs far less than a synthetic peptide, so insurers and health systems have a systematic interest in favouring it — an inferred asymmetry that may bias trial funding.

Interaction with Foundational Habits

  • Sleep: No direct interaction demonstrated. One night of total sleep deprivation changed neither salivary peptide concentration nor secretion rate, before or after exercise, so acute sleep loss does not appear to suppress this arm of mucosal defence (Gillum 2015). Whether chronic disruption or sleep apnoea does is being tested.

  • Nutrition: Direct and potentiating. Vitamin D status governs transcription of the gene (Gombart 2005); rectal butyrate, the main fibre-fermentation product, induced it in a trial, but dietary fibre is untested (Raqib 2012). Blood expression was higher in people with alcohol use disorder or after acute intoxication in a small cross-sectional study (Rojas-Pirela 2026).

  • Exercise: Potentiating but duration-dependent. Salivary peptide concentration rose from 15.5 to 22.3 ng/mL after 45 minutes of running at 75% of peak oxygen uptake, and secretion rates stayed elevated an hour later (Gillum 2015). In a crossover trial, cycling intensity alone changed nothing; only 60 and 90 minutes raised it (Ito 2024).

  • Stress management: Indirect and blunting, on animal evidence. Psychological stress reduced epidermal antimicrobial peptide production and worsened skin infection in mice, an effect attributed to stress-hormone (glucocorticoid) signalling and reversed by blocking it (Aberg 2007). No human study has tested stress reduction here; corticosteroids act on the same pathway.

Monitoring Protocol & Defining Success

Before starting, the measures worth having are a serum 25-hydroxyvitamin D level, a baseline immunoassay (an antibody-based laboratory test) for the circulating peptide, and — where a wound is the target — a photographic area and granulation measurement plus a swab culture. The venous-ulcer trial (Mahlapuu 2021) added monthly haematology and clinical chemistry, the only published safety-laboratory schedule for this intervention. Ongoing, wound area was measured weekly — for 4 weeks, with granulation, in the cream trial (Miranda 2023) and 13 weeks in the venous-ulcer trial — with blood safety tests monthly and a review at 8 and 16 weeks after the last application. For the endogenous route, vitamin D and peptide levels moved within 5 days of a high dose (Quraishi 2015), so reassessment at 1 week, 4 weeks, then every 3–6 months matches the observed kinetics.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
Serum 25-hydroxyvitamin D ≥32 ng/mL (Dixon 2012) Marker the intervention is expected to change via the endogenous route: output tracks vitamin D only below this level Threshold from Dixon 2012; conventional reference range 30–100 ng/mL; fasting not required; above 32 ng/mL the cross-sectional correlation with circulating peptide disappears; whether further repletion raises it has not been tested
Plasma hCAP18 / LL-37 No established target — track change from the individual’s own baseline Marker the intervention is expected to change; the only direct readout of the endogenous route Measured by immunoassay; values are method- and kit-dependent, so serial samples should use one assay; moved within 5 days of high-dose vitamin D (Quraishi 2015)
Target wound area and granulation index No established target — track change from the individual’s own baseline Marker the intervention is expected to change; the primary efficacy readout in both wound trials Digital planimetry from a perpendicular photograph with a scale marker; error is proportionally larger in wounds under 10 cm², which is where trial signals vanished (Mahlapuu 2021)
Aerobic wound swab culture No established target — presence or absence of a pathogen Safety check: wound infection calls for antibiotics and for pausing an unproven topical agent Levine swab technique; colony counts fell no more with the applied peptide than with placebo, so a rising burden is not evidence of underdosing (Miranda 2023)
Full blood count with liver and kidney chemistry Laboratory’s own reference ranges (standard reference range) Safety check: an out-of-range result stops or changes use The monthly schedule used in the phase IIb trial, where the vast majority of values stayed within reference ranges with no between-arm trend, though four actively treated patients had clinically significant deviations alongside adverse events; includes alanine aminotransferase (ALT, a liver enzyme) and creatinine (Mahlapuu 2021)

Qualitative markers worth tracking alongside the laboratory measures:

  • Pain at rest and at dressing change, rated on one consistent scale; the venous-ulcer trial saw improvement in all arms (Mahlapuu 2021).
  • Redness, scaling, papules or vesicles (small raised bumps or fluid-filled blisters) on skin adjacent to the treated area — the most frequent reactions recorded (Mahlapuu 2021).
  • Exudate volume and odour, since heavy exudate caused the one irritant dermatitis case in the cream trial (Miranda 2023).
  • Appearance of granulation tissue, meaning colour and coverage of the wound bed, which changed before wound area did (Miranda 2023).
  • Any spreading warmth or systemic symptoms, which point to infection rather than a treatment effect.

Emerging Research

  • Vitamin D and omega-3 trial, peptide ancillary (VITAL Infection): NCT01758081, 25,874 adults aged 50+, 2,000 IU vitamin D3 daily, co-primary endpoints respiratory infection and circulating hCAP-18, completion December 2026. A positive peptide result would make the endogenous route credible in healthy adults; a null one would confine it to deficiency.

  • The unfinished large-ulcer trial: The phase IIb trial’s own conclusion was that a study powered for wounds of at least 10 cm² is required; no such trial is registered on ClinicalTrials.gov, and the phase IIb trial itself has no record there (Mahlapuu 2021). Without it, the healing claim stays at a post hoc subgroup.

  • Melanoma intratumoral study, results posted: NCT02225366 completed November 2020 with 4 of 36 planned enrolled, 3 dosed (250 or 500 µg per tumour). Posted results: no dose-limiting toxicity or deaths; squamous cell carcinoma and actinic keratosis each in 1 of 3; all 3 listed as responders, yet one withdrew for lack of efficacy.

  • Diabetic foot ulcer cream study, registry status stale: NCT04098562 is listed as unknown with a completion date of January 2020 now long past; its results were published separately, with 25 of 40 planned participants (Miranda 2023). The registry record should not be read as an ongoing trial.

  • Cathelicidins and neutrophil traps in sleep apnoea: NCT07129486, 120 participants, primary endpoints include plasma LL-37 and neutrophil extracellular trap activity, completion estimated July 2028. A finding that intermittent low oxygen suppresses the peptide would add a sleep-related lever; a null result would remove one.

  • Synovial peptide as an infection marker: NCT04858217, 270 participants, measuring synovial cathelicidin among diagnostic markers for prosthetic joint infection; listed as recruiting, but its March 2026 completion date has passed with no results posted. Positive accuracy would establish use as a marker, not a treatment; a null result would not affect therapeutic claims.

  • Engineered analogues and delivery systems: Truncated and retro-analogues and nanocarriers aim to fix the proteolytic instability and cytotoxicity that stalled the native peptide, as a review describes (Voronko 2025). Success would make the native peptide obsolete rather than validate it.

  • The cardiovascular question: Whether the cholesterol-loading pathway raises cardiovascular risk in people is open, and the mechanism is human-specific, so animal reassurance cannot settle it (Nakamura 2024, Chernomordik 2020). A confirmed association would weaken the case for raising levels at all.

Conclusion

Cathelicidin LL-37 is a peptide the body already makes, and that double identity shapes both its appeal and its problems. As an applied treatment it has been tested almost only on long-standing leg and foot wounds in older patients with vascular or diabetic disease. There the signal is uneven: smaller studies pointed to faster tissue repair, while the largest one found no overall advantage over an inactive gel, with a benefit appearing only in a secondary look at the biggest wounds. Its reputation as a natural antibiotic is not borne out: bacterial counts in treated wounds fell no more than with an inactive cream. Evidence for raising the body’s own levels through vitamin D or fermentation products comes from serious infections, far from a healthy adult seeking resilience.

The harms are largely the mirror image of the claimed benefits. Where the body makes large amounts, this peptide is tied to inflammatory skin disease, and, mostly in laboratory and animal work, to processes that load arteries with cholesterol and can favour tumour growth. Regulators have flagged purity and immune-reaction concerns for pharmacy-made versions, plus laboratory signals of harm to male reproduction.

The evidence base is thin and conflicted, and most wound data were produced, funded and written up by the company developing the peptide; the main supporting rosacea trial was similarly sponsored. There is no approved product, no human data on the peptide reaching the bloodstream and no long-term follow-up, while material sold for self-experimentation carries no pharmaceutical quality control.

Top - Benefits - Risks - Protocol