Is 'leaky gut' a real clinical diagnosis?

Increased intestinal permeability is a measurable physiological state with well-characterised tight-junction biology. 'Leaky gut syndrome' as a catch-all wellness term for non-specific symptoms is viewed sceptically by mainstream gastroenterology — the critique is that linking permeability to a long list of systemic symptoms outruns the evidence. Both positions are defensible. In celiac disease, inflammatory bowel disease, and MIS-C, the mechanistic link between permeability and systemic consequence is established; in unexplained fatigue or generalised bloating, labelling it 'leaky gut' is a framing with limited clinical usefulness.

How strong is the evidence for BPC-157 in humans?

Weak by pharmaceutical standards. Three pilot studies have been published (knee pain, interstitial cystitis, IV safety in two healthy adults). A multicentre Phase II in ulcerative colitis has been cited but the published record is thin; an unpublished Phase I in IBD is sometimes mentioned but its data is not accessible to independent review. The overwhelming majority of the dossier is preclinical (rodent and canine). Dosing recommendations circulating online are based on anecdotal reports rather than human pharmacokinetic or efficacy trials. If you are deciding whether to use BPC-157, the honest reading is: you are making that decision on preclinical inference, not human trial evidence.

What happened with Larazotide's Phase III?

Larazotide acetate made it further toward approval than any other peptide in this class. Its 0.5 mg dose hit the Phase IIb primary endpoint for residual celiac symptoms in patients already on a gluten-free diet. The Phase III 'CedLara' trial was discontinued in 2022 when interim analysis showed the effect size was small enough that the sample size needed for statistical significance was not feasibly recruitable. Higher doses (1 mg, 2 mg) were no better than placebo — a dose-response inversion attributed to in-lumen cleavage producing fragments that competitively inhibit the parent peptide. The drug is not approved for any indication.

Can I get these compounded in the US?

BPC-157 specifically: no. In 2023 the FDA classified BPC-157 as a Category 2 bulk drug substance, which prohibits US compounding pharmacies from producing it. Larazotide has no marketed or compounded form since the CedLara discontinuation. KPV and LL-37 are research compounds without pharmaceutical supply routes. Every practical acquisition in the US currently passes through grey-market suppliers — the same supply chain that the contamination estimates (12–58% of 'research only' supplements) are drawn from.

Will NSAIDs interfere with BPC-157?

Mechanistically yes — BPC-157's repair pathway is downstream of VEGFR2 and involves prostaglandin-dependent steps. NSAIDs inhibit prostaglandin synthesis, which appears to counteract BPC-157's healing effect in animal models. It is not a safety-critical interaction in the acute sense, but if you are taking BPC-157 for an injury or inflammatory condition and simultaneously taking ibuprofen or naproxen for pain control, the peptide is working against an active pharmacological brake. For elective pain control, paracetamol does not have the same prostaglandin-inhibiting effect.

Why does LL-37 need cycling when the others don't?

LL-37 is immunostimulatory by design — it is a host-defence peptide that enhances macrophage microbicidal function. Prolonged administration (beyond 2–4 weeks) has been associated with immune fatigue and chronic low-grade inflammation, essentially the opposite of the acute-phase antimicrobial response it is trying to achieve. The community convention is 2–4 weeks on followed by 2 weeks off, capped at 4-week cycles. BPC-157, Larazotide, and KPV do not have the same immunostimulatory profile and do not require equivalent cycling — though BPC-157 is anecdotally cycled for a different reason (6–12 weeks on / 4–8 weeks off as a general safety conservatism, not a pharmacologically specific one).

Gut Healing Peptides

Peptide-driven gut healing splits across four mechanistically distinct compounds, each acting on a different structural layer of the gastrointestinal barrier:

BPC-157 (Body Protection Compound-157) — a 15-amino-acid synthetic pentadecapeptide derived from human gastric juice. Acts on VEGFR2, activating the PI3K–Akt–eNOS axis to drive angiogenesis and tissue repair. The deepest mechanistic dossier in the class, almost entirely preclinical.
Larazotide acetate (AT-1001) — an 8-amino-acid zonulin receptor antagonist. Inhibits MLC-2 phosphorylation through the ROCK pathway, blocking the disassembly of tight junctions and reducing paracellular permeability. The peptide that got the closest to approval — and then lost its Phase III.
KPV (Lysine-Proline-Valine) — a tripeptide fragment of α-MSH, internalised via the PepT1 transporter. Suppresses TNF-α, IL-6, and IL-1β by inhibiting NF-κB. A small molecule acting locally on the inflammatory layer.
LL-37 (Cathelicidin) — a host-defence peptide acting on FPR2 and EGFR to stimulate mucus synthesis and macrophage microbicidal function. Reinforces the outer defence layer; doses above ~300 mcg/day can paradoxically disrupt the flora it is protecting.

The mechanism map is coherent. The human evidence map is not. Almost everything in the animal literature has struggled to produce a positive Phase III trial — Larazotide's CedLara discontinuation in 2022 is the cautionary example the whole class is sitting with.

The structural integrity of the gut barrier depends on tight junctions — protein complexes between epithelial cells that regulate paracellular permeability. Zonulin is the signalling protein that unlocks them. When zonulin is elevated, tight junctions loosen; larger molecules pass into circulation; immune activation follows.

The MIS-C case — mechanism in public view

In Multisystem Inflammatory Syndrome in Children (MIS-C), a severe post-COVID complication, SARS-CoV-2 antigen lingered in the gut for weeks after the respiratory infection cleared. The prolonged antigen exposure drove zonulin release, loosened the intestinal barrier, and allowed inflammatory viral particles into the bloodstream — the hyperinflammatory cascade that defines MIS-C.

In the same case series, Larazotide was trialled as an adjuvant. Patients treated with Larazotide showed faster resolution of GI symptoms and quicker clearance of viral antigens from the blood relative to standard-treatment controls. A clean mechanistic demonstration — in a small, selected paediatric population, with a specific trigger — that closing the tight junctions shortens the inflammatory storm.

The zonulin identity problem

Before leaning too heavily on the zonulin framework: a 2025 review raised the uncomfortable question of what commercial 'zonulin ELISA' kits are actually detecting. Some evidence suggests Larazotide may derive from an immunoglobulin sequence rather than zonulin biology directly. The clinical effect of barrier re-sealing is real; the molecular identity of the 'zonulin' being antagonised is less settled than most reviews imply.

BPC-157 is the peptide with the widest claimed therapeutic range in this class — and the widest gap between what animal studies report and what humans have been shown. The compound is stable in gastric juice, survives oral dosing in animal models, and in preclinical work has been reported to accelerate repair in skin, muscle, tendon, ligament, bone, peripheral and central nerve, and GI fistula.

Mechanism — VEGFR2, angiogenesis, and blood-flow shunting

The Zagreb cluster's mechanistic model centres on VEGFR2 activation, which in turn triggers the PI3K–Akt–eNOS axis and upregulates the transcription factors Egr-1 and NAB2. The downstream effect is angiogenesis and tissue repair. A separate and clinically interesting finding is that BPC-157 can recruit existing blood vessels to create bypass shunts — in ischaemic colitis models, adjacent arcade vessels have been shown to open and restore circulation around the damaged area. The 'vascular engineering' claim is more specific than standard angiogenesis.

The evidence gap

The overwhelming majority of BPC-157's dossier is preclinical — rodent and canine. Human data: three pilot studies have been cited (knee pain, interstitial cystitis, intravenous safety), plus an unpublished Phase I trial in IBD whose data has not been peer-reviewed and cannot be independently scrutinised. A 2025 systematic review stated plainly that 'rigorous large-scale trials are lacking.' For users, the practical consequence is that dosing recommendations circulating online are based on anecdotal reports, not pharmacokinetic data.

The regulatory contradiction

The Zagreb cluster describes 'excellent safety profile, no reported toxicity.' The FDA, in 2023, classified BPC-157 as a Category 2 bulk drug substance — the classification reserved for substances that raise 'significant safety concerns' — and effectively barred it from US compounding pharmacies. WADA prohibits it under S0 (Unapproved Substances) and S2 (Peptide Hormones) categories. Both positions can be read off the same underlying data: the Zagreb view weights preclinical benefit heavily; the regulatory view weights the absence of robust human safety data and the commercial incentive to misuse heavily.

Beyond the barrier-and-repair layer, two smaller peptides handle distinct functional layers of the gut wall — KPV acts on the inflammatory cascade, LL-37 reinforces the antimicrobial and mucus defence.

KPV — the tripeptide NF-κB brake

KPV is a three-amino-acid fragment (Lys-Pro-Val) of α-MSH, the melanocyte-stimulating hormone. It enters epithelial cells via the PepT1 transporter — the same transporter that handles dipeptide and tripeptide absorption — and suppresses NF-κB signalling, pulling down TNF-α, IL-6, and IL-1β. The mechanism is local and well-characterised in cell and animal models of colitis. Whether its anti-inflammatory effect depends on melanocortin receptor activation or works independently is disputed: some sources identify a receptor-mediated pathway, others have found KPV does not bind MC1R directly and acts through a different route.

LL-37 — mucus, macrophages, and the cycling requirement

LL-37 (cathelicidin) is a host-defence peptide — the body produces it as part of the innate immune response to pathogens. Exogenous LL-37 acts on FPR2 and EGFR to stimulate mucus synthesis and enhance macrophage microbicidal function. The use case is ill-defined dysbiotic or post-antibiotic states where the antimicrobial defensin layer needs rebuilding. Two cautions are specific to LL-37:

Dose ceiling — at or above ~300 mcg/day, LL-37's antimicrobial effect starts disrupting the commensal flora it is protecting. Nausea, bloating, and loose stools at those doses appear to reflect iatrogenic dysbiosis from the peptide itself.
Cycling requirement — prolonged use (beyond 2–4 weeks) has been associated with immune fatigue and low-grade chronic inflammation. The community convention is 2–4 weeks on with a 2-week off-cycle; longer cycles are not supported by the literature.
Autoimmune contraindication — LL-37 is immunostimulatory by design. In lupus, Hashimoto's, and rheumatoid arthritis its administration has been reported to aggravate symptoms via immune activation. Absolute contraindication in active autoimmune disease.

The dominant practical safety risk in this class is not the peptides themselves but the supply chain. Grey-market sourcing is the failure mode that has produced the most documented harm.

The contamination problem

Estimates for contamination of ergo-nutritional supplements sold as 'research only' peptides range from 12% to 58%. The contamination profile includes bacterial endotoxin (causing febrile reactions), peptide-related impurities (immunogenicity), wrong sequences sold under correct names, and in the worst-documented cases, adulteration with bioactive compounds the buyer did not request. 'Research only' means 'not tested for human use' — not 'research grade' in any pharmaceutical sense.

The regulatory picture

BPC-157 is a Category 2 bulk drug substance in the US — not available through compounding pharmacies. Larazotide is not approved for any indication; its Phase III was discontinued in 2022 and no marketed product exists. KPV and LL-37 are research compounds without approved human indications. The entire class lacks a pharmaceutical supply route. Every legal acquisition in most jurisdictions is ambiguous at best; every practical acquisition passes through grey-market channels.

Theoretical long-term risks

Two theoretical risks with BPC-157 warrant mention even where the animal dossier looks clean. Pathologic angiogenesis — VEGF upregulation could in principle accelerate proliferation of an undiagnosed micro-tumour; the Zagreb cluster defends the peptide as showing anti-tumour effects in melanoma and colon cancer models, but this is not settled. Systemic nitric oxide overstimulation — strong NO pathway activation could theoretically interfere with haem insertion into haemoglobin (anaemia risk) or with CYP enzymes (altered drug metabolism). Neither has been clinically demonstrated; neither has been rigorously excluded in humans.

The NSAID interaction

BPC-157's repair mechanism involves prostaglandin-dependent pathways downstream of VEGFR2 activation. NSAIDs inhibit prostaglandin synthesis. The combination is reported to counteract BPC-157's healing effect — not a safety-critical interaction in the acute sense, but a useful-to-know one if the user has an injury or inflammatory condition for which they are simultaneously taking ibuprofen or naproxen.

The reframing of the gut as a central regulatory hub rather than a standalone digestive organ matters for how we think about gut-healing protocols. The barrier is connected to distant organ systems through well-characterised axes:

Gut-brain axis — BPC-157 has been shown in rodent models to counter Parkinson's-like damage from MPTP, protect somatosensory neurons, and accelerate transected sciatic nerve regeneration. Peripheral administration producing central effect is the reproducible finding across studies.
Gut-lung axis — gut microbiome dysbiosis worsens pulmonary injuries including ARDS through systemic cytokine signalling. A compromised gut barrier is a pro-inflammatory amplifier for distant lung inflammation, not a localised problem.
Gut-bone axis — prolonged glucocorticoid therapy contributes to osteoporosis in part through microbiota alteration. Healthy gut microbiome is a pre-requisite for skeletal homeostasis; peptide-driven barrier restoration without attention to the microbial community is incomplete.

Why lifestyle is not optional

The peptide layer sits on top of a nutritional, microbial, and behavioural base. If the base is inadequate, the peptide's effect is degraded — re-sealing a barrier that is then continuously insulted by dietary and microbial drivers produces a cycle, not healing. The nutritional foundation for gut healing overlaps substantially with standard inflammatory-bowel and functional-GI guidance:

Fermented foods and low-FODMAP-aware diet — the standard starting point for restoring microbial diversity without aggravating sensitive GI substrates.
Targeted probiotics rather than shotgun — strain-specific for the underlying condition; generic multi-strain probiotics are variably effective and occasionally harmful in SIBO or post-infectious states.
Polyphenol-rich diet — berries, green tea, olive oil — supports microbial diversity and has independent anti-inflammatory effects.
Sleep and stress regulation — the gut-brain axis runs in both directions; chronic stress and short sleep degrade barrier function through cortisol and vagal pathways.

For users considering a peptide protocol, the honest framing is that the peptide is an accelerant, not a foundation. The foundation is the food environment, the microbiome, and the sleep/stress posture. A powerful accelerant on a weak foundation produces a cycle of relapse; on a solid foundation, it genuinely shortens the healing timeline.

Gut healing in the peptide literature is the clearest example of a class where the animal evidence is rich, the mechanisms are coherent, and the human trials are sparse or disappointing. BPC-157 has a decades-long preclinical dossier and the FDA's Category 2 classification is a live regulatory warning, not a paperwork technicality. Larazotide came closest to approval and then lost its Phase III in 2022. KPV is an elegant local anti-inflammatory with a small, underpowered human evidence base. LL-37 is useful in the right window and potentially harmful outside it. The practical posture is that this is a field with compelling mechanism and under-developed clinical evidence, not a toolbox of settled interventions. Grey-market sourcing — where the practical acquisitions happen — is the single largest source of documented harm. For users, the weight of decision-making moves to sourcing discipline, autoimmune contraindication screening (especially for LL-37), and the lifestyle foundation that the peptide layer accelerates rather than replaces. The gut is doing more than digestion; the peptide tools for healing it are more promising than proven.

Gut Healing

1.Four Peptides, Four Layers of the Gut Barrier

2.The 'Leaky Gut' Mechanism — Zonulin and Tight Junctions

The MIS-C case — mechanism in public view

The zonulin identity problem

Explore This Protocol

Safety Triggers

Research Gaps

Contrarian Views

Protocol Reference

Stay Ahead of the Research

3.BPC-157 — Tissue Repair Beyond the Gut