For years, a traditional healer in KwaZulu-Natal, South Africa, has been treating stubborn wounds with a dark herbal paste made from six plants — half of it Cannabis sativa. Patients reportedly got better. Nobody had ever checked whether that was the plants working, or simply time and luck. Now a team of South African and Zimbabwean scientists has put the remedy through a full laboratory workup, and the results are far more interesting than a straightforward win or loss.
The study, published on 29 July 2026 in the International Journal of Molecular Sciences — Evaluation of the Antioxidant, Anti-Inflammatory, and Antimicrobial Activity of a Cannabis sativa-Infused African Product Used for Wound Healing — was led by researchers at the University of KwaZulu-Natal, working with the traditional healer himself, Mr Shezi, who supplied the formulation and disclosed how it is made. The product is referred to in the paper simply as Product Shezi. The verdict, in short: it is a genuinely capable antioxidant and a credible anti-inflammatory, and it is a poor antibacterial that may actively make wound infections worse.
Key Takeaways
Product Shezi is a polyherbal traditional medicine in which cannabis makes up half the mixture; the other half comes from five well-known South African medicinal plants. In laboratory tests it neutralised up to 70% of free radicals and fully protected human-type skin cells from oxidative damage, and it blocked prostaglandin E2 — the molecule behind wound pain and swelling — about as effectively as diclofenac at a fraction of the dose. Against bacteria, though, the picture reversed sharply. Only two of seven wound pathogens responded at all, and only at concentrations hundreds of times higher than a standard antibiotic requires — concentrations that were already toxic to skin cells. Most striking of all, instead of breaking up the bacterial biofilms that keep chronic wounds from closing, the extract encouraged them to grow in every single strain tested. The researchers were also unable to say how much CBD or THC the product contains, because the chemistry was screened only qualitatively. Everything here happened in laboratory dishes, not on human skin.
Why Chronic Wounds Are Such a Brutal Medical Problem
A wound that heals properly moves through four overlapping stages: it stops bleeding, it inflames, it rebuilds, then it remodels. Inflammation is not the enemy in this process — it is the cleanup crew, clearing debris and killing invaders. The trouble starts when that stage refuses to end.
Two things keep it stuck. The first is oxidative stress: immune cells at the wound site release reactive oxygen species such as hydrogen peroxide to sterilise the area, but in excess these same molecules shred cell membranes, damage proteins, and dissolve the structural scaffolding that new tissue needs to grow on. The second is bacteria — and specifically, bacteria that stop swimming freely and settle into biofilms, dense communities wrapped in a slimy protective matrix. Inside a biofilm, bacteria can tolerate antimicrobial concentrations roughly a thousand times higher than their free-floating counterparts. This is the central reason diabetic foot ulcers and pressure sores can stay open for months or years.
Existing tools have real limits. Synthetic antioxidants tend to work poorly when smeared on a wound. Anti-inflammatory drugs carry gastrointestinal and kidney risks. Antibiotics are losing ground to resistance. Against that backdrop, over 80% of people on the African continent rely on traditional medicine as their first line of healthcare, wounds and skin infections included — which makes the near-total absence of laboratory data on those remedies a genuine gap rather than an academic curiosity.
What Is Actually in Product Shezi
The formulation is half Cannabis sativa by weight. The other half is an equal blend of five indigenous species: Drimia altissima, Albuca fastigiata, Bulbine latifolia, Hypoxis hemerocallidea (widely known as African potato), and Hypericum aethiopicum. All six were collected in the uMgungundlovu District, formally identified by a university botanist, and deposited as herbarium specimens — a level of documentation that many ethnopharmacology papers skip. The team also obtained a permit from the South African Health Products Regulatory Authority for the medicinal cannabis component.
Preparation followed the healer’s own method: sun-dry the plants, grind them, boil the mixture for two hours, filter, then freeze-dry into a powder. The scientists added a second version made with methanol rather than water, purely to test whether an organic solvent would pull out antibacterial compounds that boiling water misses.
Chemical screening confirmed the extract contains flavonoids, saponins, tannins, glycosides, terpenoids, steroids, and alkaloids. Cannabis is a plausible contributor here — human skin is densely populated with cannabinoid receptors, and the science of cannabinoids in dermatology has expanded rapidly over the past decade — but the five companion plants have antioxidant and anti-inflammatory track records of their own.
The Good News: It Mops Up the Chemistry That Keeps Wounds Open
This is where Product Shezi performed genuinely well.
In the standard DPPH free-radical test, the extract neutralised up to 70% of radicals. Vitamin C, the benchmark, managed 94% — so the remedy is not superior to a pure antioxidant, but 70% from a crude boiled plant mixture is a respectable number.
The cell-based test was more impressive. Researchers deliberately poisoned fibroblasts — the cells that rebuild damaged skin — with hydrogen peroxide, then measured survival. Product Shezi protected them significantly, whether applied at the same time as the peroxide or twenty-four hours beforehand. Above 3 µg/mL in the pre-treatment model, protection was complete. Pre-treatment mattering as much as co-treatment suggests the extract is not merely quenching radicals on contact but priming cells to withstand oxidative attack — the kind of dual action that would be useful in a wound dressing.
It Calmed Inflammation, But Only Half of It
Inflammation runs through several separate molecular channels, and the extract only engaged one of them.
Nitric oxide, one of the main inflammatory signals, was essentially untouched. Ibuprofen brought it down significantly after twenty-four hours; Product Shezi did not, at any concentration or time point.
Prostaglandin E2 told a different story. PGE2 is the molecule responsible for a good deal of the pain and swelling around an injury, and after twenty-four hours every concentration of the extract reduced it significantly compared with inflamed control cells. Nothing happened at twelve hours, which points to a slower mechanism than a conventional drug.
The third test was the most eye-catching. Heat-damaged protein is a hallmark of inflamed tissue, so researchers measured how well the extract prevented bovine serum albumin from unravelling. At 10 µg/mL, Product Shezi inhibited protein denaturation by 66%. Diclofenac, the reference drug, managed 70% — at 50 µg/mL, five times the dose. Statistically, the two were indistinguishable. It is a laboratory model rather than a clinical outcome, but as a signal of anti-inflammatory potency it is a strong one, and it sits comfortably alongside the growing clinical literature on cannabinoids and post-injury inflammation, including work on CBD and pain at surgical wound sites.
Where It Falls Apart: The Antibacterial Numbers Simply Don’t Work
Seven bacterial species known to colonise wounds went into the antimicrobial assays. Five walked away untouched: E. coli, both classical and hypervirulent Klebsiella pneumoniae, Pseudomonas aeruginosa, and Bacillus subtilis showed no measurable inhibition at any tested concentration.
Two responded — Staphylococcus aureus and Staphylococcus epidermidis — but the numbers deserve a hard look. The boiled water extract needed 5,000 µg/mL to inhibit S. aureus and 625 µg/mL for S. epidermidis. Doxycycline handled every strain in the panel at 2 to 16 µg/mL. The traditional preparation therefore required somewhere between 40 and 2,500 times more material to achieve less.
The decisive figure is the selectivity index — a simple ratio of the concentration that kills your cells against the concentration that kills the bacteria. Anything below 1 means the treatment damages you before it damages the infection, and screening convention wants a value of 10 or more before a compound is worth pursuing. Every extract-and-pathogen combination in this study came in at roughly 0.005 to 0.05. Not one reached 1. In plain language: at any concentration high enough to slow these bacteria down, the extract was already harming skin cells and immune cells. There is no safe window.
The Result Nobody Wanted: It Fed the Biofilms
Then came the finding that reframes everything else.
Biofilms are the reason chronic wounds are chronic, so the team tested whether Product Shezi could prevent them forming and whether it could dismantle established ones. It did neither. Across all seven bacterial species, in both the prevention and the eradication models, biofilm mass increased relative to untreated controls. The reference antibiotics reduced it. The traditional remedy expanded it.
The researchers offer a plausible explanation, and it is a familiar one in plant-extract microbiology. Bacteria build biofilms as a stress response. An agent that irritates them without killing them can trigger exactly that defensive reflex — and compounds rich in phenolics have been documented doing this before. There is also the possibility that specific metabolites in the mixture simply make a more hospitable surface for bacteria to settle on.
Whatever the mechanism, the practical implication is uncomfortable. A wound dressing that reduces inflammation while encouraging biofilm growth could plausibly make a wound feel better while making it heal worse. The authors’ own conclusion is that Product Shezi should be considered as an adjunct alongside antibiotics rather than as a standalone treatment — and on this evidence, that framing is not diplomatic hedging. It is the only defensible reading.
The Number That Is Missing Entirely
Here is what the paper does not tell us: how much cannabis chemistry is actually in Product Shezi.
The phytochemical analysis was purely qualitative — a series of colour-change tests confirming that certain compound families are present, with no measurement of quantity. CBD content, THC content, terpene profile, cannabinoid ratios: none of it was quantified. The authors list this among their limitations and call for chromatography and mass spectrometry in future work, which is the right call, but it leaves a substantial hole. Half of this formulation is cannabis, and nobody knows what is in that half.
That matters for reproducibility, and it matters for safety. Plant material varies enormously between harvests, growing conditions, and preparation methods, which is why quality-control problems such as heavy metal contamination in cannabis keep surfacing even in regulated markets. Without quantitative chemistry, one batch of Product Shezi is not necessarily the same medicine as the next.
What This Study Actually Establishes
It would be easy to write this up as either a vindication of traditional medicine or a debunking of it. Neither is accurate.
What the researchers have shown is that a remedy used on real patients for years contains real, measurable bioactivity — the antioxidant and anti-inflammatory effects are not folklore, and the protein-denaturation result is genuinely competitive with a pharmaceutical. They have also shown that the antimicrobial reputation attached to the product does not survive scrutiny, and that the formulation carries a specific, previously invisible risk in the form of biofilm promotion.
Every one of these findings comes from cells and bacteria in plastic plates. None of it involved a living wound. In vitro results routinely fail to translate — a compound that protects fibroblasts in a dish may never reach the right skin layer at the right concentration in a person, and a biofilm effect observed in a well plate may behave differently in the messy environment of an actual ulcer. The authors are explicit that animal studies and in vivo wound-healing work must come next.
That caveat applies across this whole field. Topical cannabis research has produced encouraging controlled trials in conditions such as eczema, but even there the studies remain small and short. What this South African paper adds is a model for how traditional formulations should be evaluated: take the healer’s own preparation method seriously, document the plants properly, and then report the failures alongside the successes. The biofilm result is the least flattering thing in the paper, and publishing it is the most valuable thing the authors did.
FAQ
It offers partial support for a specific mechanism, not a conclusion. The cannabis-containing extract protected skin-repair cells from oxidative damage and reduced inflammatory prostaglandins in laboratory tests, both of which are relevant to healing. But cannabis was only half the formulation, its contribution was never isolated, and no wound was involved — these were cells in culture dishes. Direct wound-healing studies in animals or people have not yet been done with this product.
Not on the basis of this research. The study found that the extract was toxic to skin and immune cells at concentrations at or below those needed to inhibit bacteria, and that it increased bacterial biofilm growth in every organism tested. Biofilms are the primary reason wounds become chronic. Open wounds, and especially wounds that are not closing normally, need assessment by a qualified clinician.
Because the two effects work against each other in a real wound. Reduced inflammation might mean less pain and swelling, which feels like improvement. Meanwhile a thickening biofilm shields bacteria from both the immune system and antibiotics, which is the mechanism that turns a healing wound into a chronic one. A treatment that delivers both simultaneously could mask deterioration rather than reverse it.
Cannabis policy in South Africa has shifted substantially since the Constitutional Court’s 2018 ruling on private adult use, and medicinal cannabis is regulated by the South African Health Products Regulatory Authority. The research team obtained a SAHPRA permit for the cannabis material used in this study and full ethics clearance from their university’s Biomedical Research Ethics Committee. Rules governing traditional healers, cultivation, and commercial products vary and continue to evolve, and readers should verify current requirements in their own jurisdiction.
Disclaimer: This article is for informational and educational purposes only and does not constitute medical advice, diagnosis, or treatment recommendations. It reports on a peer-reviewed in vitro laboratory study; in vitro findings do not reliably predict effects in humans and should not be interpreted as evidence of safety or efficacy. Nothing here should be used to self-treat a wound, infection, or any other medical condition. Anyone with a wound that is not healing, showing signs of infection, or associated with diabetes or circulatory disease should seek care from a qualified healthcare professional. The Cannex does not endorse any product, preparation, or formulation described in this article, and cannabis laws differ substantially between countries and regions.
