Does Cartilage Grow Back? Evidence For and Against
Can cartilage grow back? Adult human cartilage does not regrow the way skin does, but it is not inert either. Two Duke University studies measured ongoing protein turnover in adult knee, hip, and ankle cartilage and identified a regenerative microRNA, miR-21, in the tissue. Whether that turnover can rebuild a worn joint in a living person has not been shown, and other laboratories have published findings that pull in the opposite direction. This article lays out both sides.
Key Takeaways
- Duke researchers measured ongoing protein turnover in adult cartilage, highest in the ankle, lower in the knee, lowest in the hip.
- A 2025 follow-up identified microRNA-21 as a candidate regulator shared with limb-regenerating animals.
- Both studies used donor and surgical tissue. No study has shown a worn joint rebuilding in a living adult.
- Other labs report miR-21 rising in osteoarthritic cartilage and suppressing cartilage formation, so its role is unsettled.
- The metabolic explanation involving thyroid function is a hypothesis drawn from Ray Peat's writing, not a tested treatment.
The content in this article is for educational purposes only and is not medical advice, diagnosis, or treatment. Always consult a qualified healthcare provider before making changes to your health routine.
Why Cartilage Is Slow to Heal
Evidence: established anatomy and centuries of surgical observation.
In 1743 the surgeon William Hunter wrote in Philosophical Transactions that ulcerated cartilage, once destroyed, is not repaired.1 Nearly three centuries later that sentence still opens most orthopedic discussions of the topic. Hyaline cartilage has no blood vessels, no nerves, and no lymphatic drainage. The cells that maintain it, chondrocytes, sit locked inside a dense matrix and divide rarely in adults. When a defect goes all the way through the cartilage layer, the body fills it with fibrocartilage, a scar tissue with less type II collagen that wears down faster than the original surface.
This is why the standard answer to the question in the title has been no. The Duke work asks a narrower question. Is the cartilage that remains being rebuilt at all, and if so, what controls it?
The 2019 Study: Cartilage Proteins Turn Over
Evidence: peer-reviewed human tissue study using donor cartilage from hip, knee, and ankle.
A Duke team led by Ming-Feng Hsueh and Virginia Kraus measured the age of cartilage proteins with a molecular clock.2 Certain amino acids change their chemical form at a predictable rate after a protein is built, so a tissue full of unchanged amino acids is a tissue that has been rebuilding itself. The method allowed them to compare turnover across joints without waiting years to watch it happen.
They reported three findings. Turnover depends on location, with the ankle highest, the knee in the middle, and the hip lowest. That pattern matches how often each joint develops primary osteoarthritis, which is rare in the ankle despite the load it carries. Surface layers near the joint fluid renewed more than deep layers. And microRNAs known to regulate limb regeneration in salamanders and zebrafish were present in human cartilage at higher levels in the joints with more turnover. Hsueh called it an inner salamander capacity, a phrase that traveled far beyond the paper.
The 2025 Follow-Up: A Candidate Regeneration Switch
Evidence: peer-reviewed human tissue study with cell culture experiments.
In July 2025 the same team published a second Science Advances paper that moved from observation toward mechanism.3 Using mass spectrometry and RNA sequencing on human cartilage, they identified 69 small RNAs that tracked with signs of matrix rebuilding in osteoarthritic tissue. Six of them were regulated the same way across humans, axolotls, zebrafish, and bichir. MicroRNA-21 stood out among the six.
In cell experiments, miR-21 raised the expression of cartilage-forming genes, reduced matrix breakdown, and lowered the release of inflammatory cytokines from human cartilage cells. The paper also reports that ankle cartilage kept building new matrix under osteoarthritic conditions, while knee and hip cartilage did not. The authors point to documented fingertip regrowth in young children and cartilage repair after joint distraction surgery as signs that humans retain some latent regenerative capacity, and their stated goal is a therapy that activates these pathways in knees and hips.
Evidence Against, and What the Duke Studies Do Not Show
Evidence: the studies' own limitations, plus conflicting peer-reviewed work on miR-21.
Tissue in a lab, not joints in a person
Both studies used cartilage removed during joint replacement surgery or taken from donors. Measuring new protein in a sample tells you maintenance is happening in that sample. It does not tell you a defect can fill in, or that a thinned surface can thicken again. No clinical trial has tracked cartilage rebuilding in living adults through these mechanisms, and the Duke authors do not claim one has.
Turnover is not the same as repair
The 2019 gradient shows renewal concentrated in the surface layers that touch joint fluid. Deep layers, where structural damage does the most harm, renewed least. The hip, the joint replaced most often in older adults, showed the least turnover of the three. The data describes which joints maintain themselves best, and it says nothing yet about reversing loss that has already occurred.
miR-21 cuts both ways
A 2014 study from Huazhong University measured miR-21 in knee cartilage from 10 people with osteoarthritis and 10 trauma amputees.4 The osteoarthritic samples carried more miR-21, and in a chondrocyte cell line the molecule suppressed cartilage formation by blocking growth differentiation factor 5, a protein needed for cartilage maturation. A 2020 mouse study reached a similar conclusion, reporting that deleting miR-21 protected the jaw joint from experimentally induced osteoarthritis.5 The same molecule appears as a repair signal in one lab and a degeneration signal in another. Which joint, which stage of disease, and which target gene is measured probably decides the answer, and that has not been sorted out.
The Metabolism Hypothesis
Evidence: hypothesis drawn from Ray Peat's writing on thyroid and tissue renewal; not tested in a cartilage trial.
Ray Peat's framework holds that tissue renewal depends on cellular energy production, and that low thyroid function, chronic stress hormones, and calorie restriction slow protein synthesis throughout the body. Georgi Dinkov, posting as Haidut on the Ray Peat Forum, applied this reading to the 2019 Duke study and argued that a high metabolic rate is a regenerative factor in its own right, citing animal work on limb regrowth.6 Under this view, the joints with more turnover may be the ones with better local conditions for energy metabolism, and keeping the whole system warm and well fed would support whatever renewal capacity the cartilage has.
A complication for the thyroid argument
Thyroid signaling has its own connection to osteoarthritis, and it runs the other direction. A common variant of DIO2, the enzyme that converts T4 into active T3 inside cells, was identified in 2008 as a susceptibility locus for symptomatic hip osteoarthritis.7 DIO2 is also expressed at higher levels in osteoarthritic cartilage, and in rat cartilage T3 pushes chondrocytes toward a hypertrophic state and increases the enzymes that break down matrix.8 Systemic thyroid status and local T3 activation inside a single chondrocyte are different measurements, and the rat work applied T3 directly to cartilage rather than restoring a low thyroid state. Even so, these findings mean that a simple more-thyroid-more-repair claim cannot be made from current evidence.
What I Do With This Information
Evidence: established nutritional biochemistry for raw materials; clinical experience and the Peat framework for the rest, labeled where they apply.
I read the Duke work as a reason to keep the conditions for repair in place, not as a promise that a worn joint will rebuild. If cartilage maintains itself at all, it needs the same things any tissue needs, and none of them are exotic.
Supply the raw materials
Collagen synthesis requires vitamin C to hydroxylate proline and lysine, which is why scurvy causes connective tissue to fail. Copper runs lysyl oxidase, the enzyme that crosslinks collagen into a stable matrix. Glycine and proline are the amino acids collagen is built from, and gelatin supplies them at roughly a third glycine by weight. In practice this means fruit and potatoes for vitamin C, liver and shellfish for copper, and bone broth, gelatin, and slow-cooked cuts for glycine and proline. Food sources carry these nutrients with their cofactors, which is why I reach for them before supplements.
Move the joint
Cartilage has no blood supply, so nutrients reach chondrocytes through joint fluid as the joint loads and unloads. The Duke finding that surface layers bathed in joint fluid renewed fastest fits this. Regular, gentle, pain-free movement keeps that exchange going. Sunlight and warmth help circulation to the tissues around the joint, which is where cartilage draws its supply.
Keep metabolic rate up
Following Peat and Broda Barnes, I favor adequate protein, sugars from fruit and milk, saturated fat, and staying warm, because protein synthesis in any tissue depends on the energy available to the cell. This is a framework rather than a tested cartilage protocol, and I label it as one.
Inflammation and herbs
Turmeric, boswellia, and holy basil have human trial evidence for joint discomfort, of mixed quality and often with industry funding. I match herbs to a person's constitution and full health picture in consultation rather than handing out a list. A diet low in polyunsaturated seed oils fits both the Peat framework and the basic chemistry of fat oxidation, which produces the compounds that drive inflammatory signaling.
The recipes below feature gelatin, bone broth, and glycine-rich dishes that fit this approach.
Frequently Asked Questions
Does cartilage grow back?
Not in the way skin or bone does. Duke University research measured ongoing protein turnover in adult cartilage, highest in the ankle, and identified a regenerative microRNA in the tissue. Whether that turnover can rebuild a worn joint in a living person has not been demonstrated.
Does knee cartilage grow back?
Knee cartilage showed measurable turnover in the Duke tissue studies, less than the ankle and more than the hip. No study has shown a damaged knee surface regrowing on its own in an adult. Full-thickness defects fill with fibrocartilage, which is weaker than the original tissue.
How do you regenerate knee cartilage naturally?
No diet, supplement, or exercise has been shown to regenerate knee cartilage in adults. What is within reach is supplying the raw materials for collagen, moving the joint so nutrients reach the cartilage through joint fluid, and avoiding the metabolic conditions that slow tissue maintenance. These support whatever repair capacity exists. They do not replace an orthopedic evaluation for a damaged joint.
Why does ankle cartilage regenerate better than knee or hip cartilage?
Ankle cartilage carried higher levels of regeneration-associated microRNAs, including miR-21, and in the 2025 study it kept building new matrix under osteoarthritic conditions while knee and hip cartilage did not. The researchers describe the ankle as a regeneration-permissive environment and are studying what makes it different.
How long does cartilage take to heal?
No one has measured a timeline in a living adult. Cartilage heals slowly because it has no blood supply and its cells rarely divide. The Duke molecular-clock data shows turnover happening across years rather than weeks, and the amount of new protein needed to rebuild a worn surface is unknown.
Does eating gelatin rebuild cartilage?
Gelatin supplies glycine and proline, the amino acids collagen is built from, so it provides raw material the tissue would need if repair is happening. No trial has shown that eating gelatin rebuilds cartilage in a damaged joint. Bone broth, gelatin, and slow-cooked cuts are the best food sources, alongside liver and shellfish for copper and fruit and potatoes for vitamin C.
Notes and Sources
- Hunter W. Of the structure and diseases of articulating cartilages. Philosophical Transactions of the Royal Society. 1743;42:514–521. ↩
- Hsueh MF, Önnerfjord P, Bolognesi MP, Easley ME, Kraus VB. Analysis of "old" proteins unmasks dynamic gradient of cartilage turnover in human limbs. Science Advances. 2019;5(10):eaax3203. doi:10.1126/sciadv.aax3203 ↩
- Hsueh MF, Önnerfjord P, Kraus VB. Anabolic indices of matrix proteins identify regenerative small RNA intrinsic to human cartilage. Science Advances. 2025;11(28):eadu8440. doi:10.1126/sciadv.adu8440 ↩
- Zhang Y, Jia J, Yang S, Liu X, Ye S, Tian H. MicroRNA-21 controls the development of osteoarthritis by targeting GDF-5 in chondrocytes. Experimental & Molecular Medicine. 2014;46:e79. doi:10.1038/emm.2013.152 ↩
- MicroRNA-21-5p as a novel therapeutic target for osteoarthritis. Bone & Joint Research. 2020;9(10):689–700. ↩
- Dinkov G (Haidut). Humans have a 'salamander-like' ability to regrow cartilage. Ray Peat Forum, December 4, 2019. Forum thread. ↩
- Meulenbelt I, Min JL, Bos S, et al. Identification of DIO2 as a new susceptibility locus for symptomatic osteoarthritis. Human Molecular Genetics. 2008;17(12):1867–1875. doi:10.1093/hmg/ddn082 ↩
- Deiodinase 2 upregulation demonstrated in osteoarthritis patients cartilage causes cartilage destruction in tissue-specific transgenic rats. Osteoarthritis and Cartilage. 2013. Full text ↩
