Cissus for Bone and Fracture Healing: Examining the Claim

Most traditional herbs arrive with a vague indication — "for strength", "for the blood", "cooling". Cissus quadrangularis arrives with a surgical one: it is supposed to make a broken bone knit faster. Its Sanskrit name asthisamharaka and its Hindi name hadjod both say so outright, and the bone-setters who used it were treating fractures, not general debility.

That specificity is why this page is harder to write than most. On almost every herb the honest verdict is "nobody has tested this". Here somebody did — in India from the 1960s, in animals and then in patients — and reported that it worked. The trouble is that those studies would not pass a modern journal's methods desk, and the trial that would settle the question has never been run. Cissus therefore sits in an uncomfortable position: the fracture-healing claim is neither unsupported nor established. What follows tries to describe that position precisely rather than rounding it up into "clinically proven" or down into "no evidence".

What Is Actually Being Claimed
The Name Is the Claim
The Traditional Record, and Its Limits
The Older Indian Clinical Literature
Old and Weak Is Not Absent, and Not Refuted
Reading the Animal Work Before Reporting It
The Osteoblast and Stem-Cell Studies
Proposed Mechanisms and How Far They Go
The Maxillofacial Strand
Bone Density Is a Different Question
What a Trial That Settled This Would Look Like
Numbers This Page Refuses to Give
Evidence Ledger
If You Have a Fracture Right Now
Key Research Papers
Connections

What Is Actually Being Claimed

Four claims get sold as one. They need different proof and are not equally supported.

Faster union of an ordinary fracture — a break that would knit in twelve weeks knits in nine. The traditional claim, and testable.
Better quality of the healed bone — more mineral, better trabecular architecture, higher breaking load. Mostly an animal endpoint.
Prevention of bone loss — slowing osteoporosis in someone who has not broken anything. A different question on a different timescale, and where human evidence is thinnest.
Repair of ligament, tendon and cartilage — extrapolated from the bone claim by the fitness market. See Cissus for Joint and Tendon Pain.

A bottle labelled "bone and joint support" trades on the strength of claim 1 to sell claims 3 and 4. Claim 1 has an old, flawed, genuinely positive human literature. Claim 3 has ovariectomised rats. Claim 4 has one small pilot study and a lot of copywriting.

The Name Is the Claim

A general rule worth applying to any herb: check whether the common name smuggles in a clinical assertion. Usually a name describes the plant — how it grips, its colour, where it grows. Here it is the opposite, and unusually direct. Asthisamharaka is asthi, bone, plus a root meaning to hold together or arrest destruction. Hadjod is blunter: haddi, bone, plus jodna, to join. Vajravalli, "thunderbolt creeper", is a strength metaphor pointing the same way. The English trade names — "adamant creeper", "devil's backbone" — reference hardness or a spine.

Two consequences follow, pulling opposite ways. The tradition is unambiguous about what the plant was for, which is useful: nobody can reinterpret hadjod as a general tonic to make a weak result look like a hit. But the name also pre-loads the conclusion for every unblinded study ever run on it. A surgeon in 1970 reading a radiograph of a patient he knows is taking the bone-joining herb is not a neutral instrument. Naming a plant after its outcome is exactly the circumstance in which blinding stops being a formality.

Note what the name does not claim. Unlike names that imply safety in a particular group, hadjod asserts nothing about pregnancy, children or long-term use, and nothing in the etymology should be read as reassurance there. That check comes back clear — and a clear check is still information.

The Traditional Record, and Its Limits

The traditional practice is consistently described. Crushed fresh stem was applied as a paste over a splinted limb; stem juice or dried powder was given internally, often in ghee or milk, through the weeks of immobilisation. The same material was extended to sprains, ligament injuries, back pain and gout. In Thai traditional medicine the plant took a different road entirely and became a dispensed remedy for haemorrhoids — a reminder that "traditional use" is plural and regional, not one verdict.

That record establishes that generations of practitioners believed the plant helped fractures and used it that way. It cannot establish that it did. Fractures heal on their own; that is the default, and any remedy given during the natural course of healing collects the credit. Traditional practice had no way to separate the two — no untreated comparison group, no radiographs, no standardised fracture types, and no register of the patients whose bones did not knit. That is not a dismissal of the tradition. It is why the tradition was worth testing, which is what happened next.

The Older Indian Clinical Literature

Evidence tier: human, positive, old, methodologically weak, and geographically concentrated.

Modern interest begins with work at Banaras Hindu University in the 1960s and 1970s associated with K. N. Udupa, G. C. Prasad and colleagues, followed by a group including S. S. Chopra examining experimental fracture repair, its blood chemistry and its histology. Taken at face value the findings were consistent: earlier callus, earlier mineral deposition, earlier recovery of mechanical strength. Later Indian work extended the claim to patients.

Read the design rather than the abstract, and these problems are present — individually or together — across most of that literature.

Small samples. Dozens of subjects, not hundreds. Union time varies enormously with age, site, comminution, smoking, diabetes and fixation, so a small study can be swamped by allocation luck alone.
Absent or inadequate blinding. Patient and assessor generally both knew who was taking the herb — and the endpoint was a judgement about callus on a plain film.
No placebo in several. Comparators were untreated patients or historical controls, which controls for none of the attention and immobilisation compliance that come with being in the treatment arm.
Soft endpoints. "Time to clinical union" and callus grading are assessor-dependent. Standardised radiographic union scores did not exist yet; functional outcome scores were rarely reported.
Concentration in a few centres. Replication from the same institution and tradition as the original is confirmation, not independent replication.
Retrieval difficulty. Much of this work appeared in Indian journals whose back issues are poorly digitised, and several frequently cited items circulate mainly as second-hand citations inside reviews. That cuts both ways: hard to verify the claims, and equally hard to check whether the people citing them reported them accurately.

Where the citation trail is the only trail, this page says so. For the 1960s and 1970s items, volume, page and even title as commonly reproduced should be treated as unverified secondary metadata. Every reference below is therefore a PubMed search built from author names and distinctive words, so a reader can see the actual record — or see that there is not one.

Old and Weak Is Not Absent, and Not Refuted

Three verdicts get collapsed into each other in supplement writing, and keeping them apart is the single most useful thing this page can do.

Absent. No adequate study has ever been run. The honest statement is "unknown". Most traditional herb claims live here.
Refuted. Adequate studies were run and found no effect, or found harm. The honest statement is "tested and failed".
Old, weak and positive. Studies exist, they point the same way, and they are not good enough to rely on. The honest statement is "suggestive and unresolved".

Cissus and fracture healing is in the third category, uncomfortable precisely because it is not one of the easy two. Treating it as absent is unfair to a real if flawed body of work and to a tradition specific enough to have generated it; treating it as established is the mistake on the label. The Sawangjit systematic review and meta-analysis in Phytotherapy Research (2017) reached essentially this conclusion across all indications: effects are possible, the trials are few, and their quality limits what can be concluded. That is not a euphemism for "no" — it is a statement that the question is open.

A distinct sub-verdict is worth logging. Some newer maxillofacial work is randomised but multi-arm and multi-agent, making it positive evidence for a protocol rather than for the plant — a different defect from being old and unblinded, and one that adds little strength to the older literature.

Reading the Animal Work Before Reporting It

The animal literature is where the claim looks strongest, so it deserves the closest reading. Three design questions matter more than the reported result.

Was the extract given before or after the injury?

Dosing an animal for two weeks and then breaking its bone tests prophylaxis — whether the animal was better prepared to heal — which is not the question a person in a cast is asking. Post-fracture dosing tests treatment. Fracture-repair studies fare better here than, say, rodent hepatoprotection work, because the injury is deliberately created and dosing usually follows it. But the timing must be read, not assumed; where methods leave it vague, the result is not evidence that taking the herb after a break helps.

Does the model model human fracture healing?

A rodent long-bone osteotomy heals fast, on a reliable schedule, in a young healthy animal with intact blood supply and no comorbidity. Human non-union — the outcome anyone actually fears — clusters in smokers, diabetics, the elderly, open injuries and comminuted fractures. A model in which every animal heals cannot show that a treatment rescues the ones that would not have. The dog radius-ulna study by Deka and colleagues (Indian Journal of Pharmacology, 1994) is closer to a clinical fracture than a rat osteotomy, and its authors called it preliminary.

What was the comparator arm for?

A positive control validates an assay; it does not rank treatments. Where an extract matches or beats a reference agent in a rodent, the first question is how well that agent performs in that species — often badly, for dose-scaling and metabolic reasons unrelated to human efficacy. "Comparable to standard treatment in rats" is a statement about rats.

The fair summary of the Cissus animal work: multiple groups, multiple species, consistent direction, some hard endpoints (mineral content, mechanical strength) — in models that systematically favour finding an effect. Consistency across labs is worth something. It is not worth a clinical recommendation.

The Osteoblast and Stem-Cell Studies

The cell-culture layer is the most mechanistically coherent part of the file. Potu and colleagues, in Clinics (2009), reported that a petroleum-ether stem extract increased bone-marrow mesenchymal stem cell proliferation and pushed those cells toward the osteoblast lineage. Other groups report rises in alkaline phosphatase and osteocalcin and more mineralised nodule formation in osteoblast-like lines. In ovariectomised rats — the standard post-menopausal bone-loss model — ethanol extracts have been reported to limit loss, including work by Shirwaikar and colleagues in the Journal of Ethnopharmacology (2003).

Three caveats travel with that paragraph and should not be separated from it.

A petroleum-ether extract is not a capsule and not a decoction. Petroleum ether pulls the lipophilic fraction — sterols, triterpenes — and leaves behind most of what water or ethanol would deliver. The traditional preparation was fresh stem, juice, or powder in ghee. A result from one solvent does not transfer to a product made with another.
Osteoblast differentiation in a dish is upstream of everything that matters. Between a mineralised nodule in a well plate and a united fracture lie absorption, first-pass metabolism, concentration at the fracture site, and the coordinated resorption-and-formation sequence of real callus. Agents that look excellent in osteoblast culture have done nothing in people.
Nobody has shown the responsible compounds reach human bone. There is no adequate human pharmacokinetic study of the relevant constituents, which leaves the cell work as a plausibility argument.

Proposed Mechanisms and How Far They Go

Four mechanisms are offered, with descending amounts of support.

Steroidal and ketosteroid constituents

The plant contains phytosterols and, by the supplement industry's own standardisation, a fraction marketed as "ketosterones", to which anabolic and cortisol-related effects are attributed. What is established is that the plant contains sterols. The leap to a defined anabolic mechanism in human bone is not: the assay for the fraction is not standardised across laboratories, and no human study has shown a hormonal effect adequate to explain accelerated union. This page treats the ketosterone story as marketing pharmacology — a hypothesis with a trade name.

Stilbenes

Adesanya and colleagues characterised stilbene derivatives from the stem in the Journal of Natural Products (1999), including resveratrol-family compounds and dimers such as the quadrangularins. Real chemistry — the plant is in the grape family, and grape-family plants make these. It is also the doorway to a common error: research on grapes, wine, grape-seed extract or high-dose resveratrol does not transfer to Cissus. Different amounts, different mixture, different plant part. A label that cites resveratrol literature for Cissus is borrowing evidence from another plant.

Vitamin C and collagen

The fresh stem is a good source of ascorbic acid, and vitamin C is required for the hydroxylation steps in collagen cross-linking — part of bone matrix and soft-tissue repair alike. This is the most mundane and most defensible mechanism on the list, and it argues against the capsule: an extract standardised to sterols carries essentially none of the vitamin C. Anyone reasoning from the traditional vegetable to the modern product is switching interventions midway. See Vitamin C and Collagen.

Anti-inflammatory and antioxidant activity

Extracts scavenge radicals in chemical assays and reduce inflammatory mediators in cultured cells. So does nearly every plant extract. Too non-specific to support a bone claim, and an in-vitro anti-inflammatory result must never be reported as though it were a clinical effect.

The Maxillofacial Strand

The newest human bone work comes largely from Indian oral and maxillofacial surgery, where mandibular fractures make an attractive test bed: relatively standardised fracture type, standardised fixation, routine follow-up radiography. Pilot and clinical evaluations there — including Singh and colleagues in the National Journal of Maxillofacial Surgery (2011) and Brahmkshatriya and colleagues in Ayu (2015) — have reported faster union, better bone density at the fracture line, or lower pain scores with Cissus alongside standard care.

This is the most encouraging part of the modern record, and it carries two specific problems.

Several of these studies are multi-arm and multi-agent. Where a design compares Cissus against another botanical and against a proprietary osteogenic product, a positive overall result attributes nothing cleanly to Cissus, and small per-arm samples make the between-arm comparison weaker still. Before crediting the plant with a formula's result, check what was actually administered — a rule that becomes decisive on the weight-loss trials, where the most-cited studies used a multi-ingredient product.
Pilot studies are permission to run a bigger trial, not a result. Their own authors say so. The pattern across this literature is a series of pilots, published over decades, whose confirmatory successors never appear.

Bone Density Is a Different Question

Supplement labels merge fracture healing with osteoporosis; they should not. Healing a break is a weeks-to-months remodelling of one site. Osteoporosis is a decades-long systemic shift in the balance between resorption and formation, and its only endpoint that matters is fracture incidence, not a laboratory marker.

For Cissus the human osteoporosis evidence amounts to small studies reporting movement in bone turnover markers. That is a weak surrogate: bone medicine includes agents that shifted markers convincingly and did not reduce fractures, which is why regulators demand fracture endpoints. No trial has shown that Cissus prevents fractures or produces a clinically meaningful bone-density change in humans over the years-long horizon osteoporosis treatment requires.

Worth naming the alternatives so the comparison is concrete: adequate calcium, vitamin D sufficiency, vitamin K2, magnesium, resistance and impact loading, not smoking, and prescription therapy where a bone-density scan indicates it. See Osteoporosis. Cissus substitutes for none of them.

What a Trial That Settled This Would Look Like

The strongest honest move available on a claim like this is not to hedge but to specify the study that would resolve it — which also disposes of the argument that traditional remedies cannot be trialled. Fracture healing is one of the more tractable outcomes in medicine. The tools exist. They have simply never been pointed at this plant properly.

One standardised fracture type — closed tibial shaft with intramedullary nailing, say, or mandibular angle with defined fixation. Mixing sites destroys the comparison.
Randomised allocation with concealment, so neither surgeon nor patient influences which arm a fracture lands in.
A defined, assayed intervention with a matched placebo. A named extract with documented constituent content, so the trial is repeatable. "Cissus extract" is not a specification.
Blinded assessment against a validated union score. Radiographic union scoring systems exist (RUST for tibial fractures, for instance) precisely because "the surgeon thought it looked united" is unreliable. Two blinded readers, with an inter-rater reliability statistic.
Hard co-primary endpoints: time to radiographic union and time to weight-bearing or functional recovery, plus non-union rate.
Adequate power — hundreds of patients across multiple centres, not dozens at one.
Pre-registration and full outcome reporting, so a negative result cannot quietly disappear.
Independent funding, or at minimum an independent statistician and full disclosure. Given how much of the existing Cissus literature is industry-adjacent, this is not a formality.

None of that is exotic. It is how fracture-healing adjuncts such as bone morphogenetic proteins and ultrasound stimulation were assessed — sometimes to their disadvantage. The absence of such a trial for a remedy this widely sold, sixty years after the first positive reports, is itself a finding.

Numbers This Page Refuses to Give

Several figures circulate confidently in Cissus writing and this page will not repeat them, because they cannot be sourced to a record that can be checked.

A percentage reduction in healing time. Figures like "one third faster" appear in supplement copy and secondary reviews, attributed to the poorly retrievable 1960s and 1970s papers, and the number shifts depending on who repeats it. A confident figure here would be a guess.
A milligram dose for fracture healing. No dose-finding study exists for this indication. What exists is practitioner convention in grams of dried stem, and a much smaller milligram figure from the weight-loss trials of a concentrated extract. Neither was chosen for bone.
A ketosterone content. Labels state percentages; the assay is not standardised between laboratories, so the same printed number does not mean the same content in two bottles.
A calcium or vitamin C figure per dose. Composition figures for the fresh stem vary widely with maturity, region and drying, and most of the calcium is bound as poorly absorbed oxalate. A per-capsule nutrient claim derived from a per-100-g plant analysis would mislead by a large factor.
Metadata for the oldest fracture papers. Where this page names a 1960s or 1970s study, treat volume and page details as unverified secondary reporting.

Evidence Ledger

Ranked against the strength of the evidence, not against the strength of the marketing.

Best supported (human, but weak): faster reported union alongside standard fixation in small Indian trials, including recent maxillofacial work. Tier: multiple positive human studies of low methodological quality.
Mechanistically coherent (preclinical): osteoblast differentiation in culture, reduced bone loss in ovariectomised rats, earlier callus in animal fracture models. Tier: consistent across labs, in models that favour a positive finding.
Plausible and mundane: vitamin C in the fresh stem contributing to collagen synthesis. Tier: established biochemistry — but it applies to the vegetable, not to a sterol-standardised extract.
Asserted, not established: a "ketosterone" anabolic or cortisol-modulating mechanism. Tier: marketing hypothesis.
Absent in humans: prevention of osteoporotic fracture, or meaningful bone-density gain. Rat data plus surrogate markers only.
The three most solid facts here are all limitations: no adequate blinded modern trial; no established dose for bone healing; and no shelf product can be matched to a study unless it names the tested extract.

If You Have a Fracture Right Now

The practical advice does not depend on resolving the evidence question, which is convenient.

A broken bone needs imaging, reduction where indicated, immobilisation and follow-up. No herb substitutes for any of that, and the most serious harm around Cissus is not toxicological — it is a delayed orthopaedic assessment.
Fix what demonstrably impairs union. Smoking above all; then protein intake, vitamin D status, and glycaemic control in diabetes. Each has more evidence than this plant.
Tell the surgeon if you take it — reported blood-glucose effects matter around surgery, and an unrecorded supplement is a variable nobody can interpret later. See Cissus: Dose, Formulations and Safety.
If you try it anyway, you are betting on an old, weak, positive literature. That is a defensible bet with a cheap, generally well-tolerated supplement. It is not the same as a treatment.

Key Research Papers

Every citation is a PubMed search built from author names and distinctive title words, never a numeric identifier, so a link cannot silently resolve to the wrong paper. Confirm the record before relying on anything here. Where metadata is uncertain — for the oldest Indian fracture literature, most of it — that is stated.

Sawangjit R, Puttarak P, Saokaew S, Chaiyakunapruk N. Efficacy and safety of Cissus quadrangularis L. in clinical use: a systematic review and meta-analysis of randomised controlled trials. Phytotherapy Research, 2017. Best entry point — it judges quality rather than counting positives. Find on PubMed.
Stohs SJ, Ray SD. A review and evaluation of the efficacy and safety of Cissus quadrangularis extracts. Phytotherapy Research, 2013. Useful compilation; read the competing-interests statement first. Find on PubMed.
Udupa KN, Prasad GC and colleagues, Banaras Hindu University, 1960s, on Cissus quadrangularis in fracture repair — the origin of the modern literature. Several related papers; the volume and page details as commonly reproduced are unverified secondary metadata. Search PubMed.
Chopra SS and colleagues on Cissus quadrangularis in experimental fracture repair, Indian literature of the 1970s. Metadata unverified; treat as a citation trail unless you retrieve it. Search PubMed.
Deka DK, Lahon LC, Saikia J, Mukit A. Effect of Cissus quadrangularis in accelerating healing process of experimentally fractured radius-ulna of dog. Indian Journal of Pharmacology, 1994. Closer to a clinical fracture than a rodent osteotomy; authors called it preliminary. Find on PubMed.
Potu BK and colleagues. Petroleum ether extract of Cissus quadrangularis enhances bone marrow mesenchymal stem cell proliferation and facilitates osteoblastogenesis. Clinics, 2009. Note the solvent — lipophilic fraction, not a decoction. Find on PubMed.
Shirwaikar A, Khan S, Malini S. Antiosteoporotic effect of ethanol extract of Cissus quadrangularis Linn. on ovariectomised rat. Journal of Ethnopharmacology, 2003. A prevention design, not a treatment design. Find on PubMed.
Muthusami and colleagues on Cissus quadrangularis and the proliferation, differentiation and matrix mineralisation of human osteoblast-like cells, around 2011. Confirm journal and year from the record. Search PubMed.
Singh V and colleagues. Clinical evaluation of Cissus quadrangularis and Moringa oleifera and osteoseal as osteogenic agents in mandibular fracture. National Journal of Maxillofacial Surgery, 2011. A multi-agent design — note what that does to attribution. Find on PubMed.
Brahmkshatriya HR and colleagues. Clinical evaluation of Cissus quadrangularis as an osteogenic agent in maxillofacial fracture: a pilot study. Ayu, 2015. Read the sample size and blinding statement before the conclusion. Find on PubMed.
Adesanya SA, Nia R, Martin MT and colleagues. Stilbene derivatives from Cissus quadrangularis. Journal of Natural Products, 1999. Where the resveratrol-family constituents were characterised — and why grape research keeps getting borrowed for this plant. Find on PubMed.

Live PubMed Searches

External Resources

NCCIH — how herbal evidence is graded.
NIH Bone Diseases Resource Center — patient-level bone health and fracture risk.
PubMed — the primary literature.

Connections

All Herbs
Cissus quadrangularis — botany, names, constituents, full cautions.
Cissus for Joint and Tendon Pain — where the bone claim gets extrapolated to soft tissue.
Cissus for Weight and Metabolic Outcomes — the modern trials and the multi-ingredient problem.
Cissus: Dose, Formulations and Safety — what a dose means when extracts differ by a factor of many.
Osteoporosis — the fracture-endpoint disease; why turnover markers are not enough.
Hip Fracture — the outcomes that matter most, where no herb has been tested.
Stress Fracture — the injury the fitness market most often invokes Cissus for.
Calcium — why high paper calcium does not make this a calcium supplement.
Vitamin D3 — a genuine determinant of bone healing, with real evidence.
Vitamin K2 — matrix carboxylation, better studied than any of this.
Magnesium — the third mineral leg of bone.
Boron — another trace element with bone claims and a thin human record.
Vitamin C — the collagen mechanism the stem supplies and the extract does not.
Collagen — the protein at the centre of bone matrix and tendon repair.
Kidney Stones — why oxalate limits how much fresh stem you can eat.

Back to Table of Contents