Story
Glucosamine + Chondroitin
Last updated
In one pass Glucosamine and chondroitin are the two supplements middle-aged and older people buy most often to protect their knees and ease joint pain. Not this — Glucosamine cures knee pain — In the large GAIT trial (2006, NEJM) the main result was negative — no benefit was shown. A signal in people with moderate-to-severe pain did not pass the analysis set in advance.
Educational content, not medical advice — consult a clinician.
Story path
Chapter 1
Building blocks of cartilage
Glucosamine and chondroitin are the two supplements middle-aged and older people buy most often to protect their knees and ease joint pain. They really are raw materials of joint cartilage, but whether swallowing them gets anything to the knee is another matter.
Start with what they are inside a joint. Cartilage contains long-chain molecules called glycosaminoglycans (GAGs), which hold water like a sponge and give cartilage its springiness. Glucosamine is the basic building unit of those chains, and chondroitin (chondroitin sulfate) is a whole chain hung with sulfate groups. These chains wind together with proteins to form the base material that fills cartilage. Calling them cartilage's raw materials is correct.
But between being a raw material and reaching that piece of cartilage after you swallow it lies a whole supply line: first through the gut and the liver, then into the blood, and finally seeping into a tissue that has no blood vessels. Some is lost at every step, which is why, from the mechanism alone, this class of supplement is bound to have a small effect.
Start with what they are inside a joint. Cartilage contains long-chain molecules called glycosaminoglycans (GAGs), which hold water like a sponge and give cartilage its springiness. Glucosamine is the basic building unit of those chains, and chondroitin (chondroitin sulfate) is a whole chain hung with sulfate groups. These chains wind together with proteins to form the base material that fills cartilage. Calling them cartilage's raw materials is correct.
But between being a raw material and reaching that piece of cartilage after you swallow it lies a whole supply line: first through the gut and the liver, then into the blood, and finally seeping into a tissue that has no blood vessels. Some is lost at every step, which is why, from the mechanism alone, this class of supplement is bound to have a small effect.
Mechanism · Where the swallowed pill ends up
Put the opening question in terms of the body: when a glucosamine pill goes down your throat, how much of it ends up at the cartilage in your knee? There are three gates on the way, and each one takes a cut.Gate 1 · The gut and the liver: it is folded into the body's sugar pool
Glucosamine is an amino sugar, which is to say a glucose molecule with an amino group attached. The small intestine can absorb it into the blood, but once there it does not know the way: blood first carries it through the portal vein into the liver, where liver cells take most of it apart on the spot for energy or turn it into other sugars, and only the rest travels on through the body. More important, the body already makes it itself (starting from glucose plus glutamine, and almost every cell can do this). Food contains a little too, mainly in animal cartilage such as chicken feet, chicken cartilage and beef bone broth, while supplements are mostly extracted from shrimp and crab shells or cow cartilage. So the dose you swallow is water poured into a pool that is being topped up all the time, not a dedicated line to the knee.
As a result, in pharmacokinetic studies the concentration measured in the blood after an oral dose is two to three orders of magnitude lower than the concentration used when cartilage cells are studied in a dish. The lab experiments showing that glucosamine stimulates cartilage cells to make more matrix used concentrations the body never reaches. That alone is enough to explain why it looks good in a test tube and so often shows no measurable difference in people.
Gate 2 · Chondroitin is a large molecule, and the whole chain cannot get in
Chondroitin is not a small molecule but a long chain of dozens of sugar units, loaded with sulfate groups. The gut wall absorbs nutrients through transporters built for small molecules, and a chain this long basically cannot pass. Most of it is cut up inside the gut by digestive enzymes and gut bacteria into short pieces and single sugars, which are absorbed as fragments. In other words, the chondroitin you eat is mostly no longer chondroitin by the time it reaches the blood, and only a small share enters the circulation as an intact long chain.
Gate 3 · Cartilage has no blood vessels, so even what the blood delivers cannot get in
This gate is the easiest to overlook and the most important. Nearly every tissue in your body has capillaries running right alongside it to deliver supplies, but joint cartilage does not; it is one of the thickest tissues in the body without blood vessels. Cartilage cells sit buried in the matrix they themselves made, and the only way they get nutrients is by diffusion: nutrients first dissolve into the synovial fluid in the joint cavity, then seep into the matrix little by little.
What pushes the nutrients in? Your movement. When you walk, climb stairs, or bend and straighten your knee, the cartilage is squeezed flat and springs back again and again, pushing old fluid out and drawing new fluid in like a sponge. That is why a joint left still for a long time actually degenerates faster, and why exercise is ranked ahead of every supplement later on: movement itself is the cartilage's only delivery pump.
The three gates together
The gut cuts some away, the liver holds some back, what is left spreads through the whole body, and whatever remains still has to seep, bit by bit and only with joint movement, into a tissue that has no blood vessels. Every step on this chain dilutes the dose, and not one of them is aimed at the knee. So calling them cartilage's raw materials is not wrong. What is wrong is the half-sentence silently attached to it: so eating them rebuilds cartilage. A whole supply line sits between the raw material and the finished tissue, and the pill you swallowed does not run that line.
That does not mean it must be useless (what the human trials show is in the chapter What the largest trial found), but it explains one thing: why the effect of this class of supplement is bound, by the mechanism itself, to be small.
Clinical · Why the X-ray and the pain disagree
Why this class of supplement is worth taking apart carefullyArthritis is very common in middle and older age: one estimate is that about half of people in China aged 65 and over show cartilage degeneration on an X-ray. Add that glucosamine and chondroitin are themselves a market worth billions of dollars, bought by enormous numbers of people, and they deserve to have their evidence read line by line rather than being dismissed with useless or miracle cure.
First, which kind of arthritis
Primary osteoarthritis (OA): worn in slowly by aging plus decades of chronic wear and tear. This is mainly what this story is aboutSecondary osteoarthritis: something else came first, such as a torn meniscus, a ruptured ligament, rheumatoid arthritis or gout, and joint degeneration followed. The cause is different and so is the management, so do not paste this story's conclusions onto it
A counter-intuitive fact worth remembering: pain and the degeneration on an X-ray do not match one to one.
Someone whose X-ray looks terrible may hardly hurt, while someone with only mild degeneration may hurt too much to walk downstairs; the differences between people are huge. This is not a flaw in the imaging. It is because wear and pain are simply not the same thing. Which tissue the pain actually comes from is explained in the chapter What the largest trial found.
This fact has two very practical consequences:
Whether a treatment works cannot be judged from imaging alone. An unchanged X-ray in a person who no longer hurts is a real improvement; a nicer X-ray in a person who still hurts badly is still real painWhen you trial a supplement yourself, what you should record is whether it hurts, whether you can climb stairs, and how far you can walk, not whether the X-ray has improved; the X-ray will not show any change within a few months
In practice · Three names that look alike
On pharmacy shelves, and in the trials covered later, three names that look almost the same keep turning up. Tell them apart before you buy.Glucosamine sulfate: Europe regulates it as a prescription drug (the originator product is called Dona), and most of the trials with positive results used it.
Glucosamine hydrochloride: the main form in US pharmacies, and the one used in the famous negative trial covered later, GAIT.
The glucosamine in these two is actually the same molecule; the only difference is the ion it is paired with, sulfate in one and chloride in the other. In the stomach the ions separate, and what gets absorbed is the same glucosamine either way; the absorbed fractions differ slightly, and which is better is still unsettled. But which salt was used and whether the trial came out positive or negative overlap heavily, so the question keeps coming up; the accounting is done in the chapter What the largest trial found.
N-acetylglucosamine (NAG): the name differs by only a few letters, but it is a different thing, glucosamine with an extra acetyl group attached, and research on it is mainly about the gut lining, not joints. Read the label so you do not buy the wrong one.
Chapter 2
What the largest trial found
The trial that most needs to be told honestly about glucosamine and chondroitin is a large randomized trial called GAIT (published in the *NEJM* in 2006). It enrolled more than fifteen hundred people with knee osteoarthritis (OA) and randomly assigned them to five groups for 24 weeks: glucosamine alone, chondroitin alone, the two together, a real anti-inflammatory drug, celecoxib, as a comparison, and placebo.
The result was disappointing: glucosamine and chondroitin, alone or together, did no better than placebo. Only the celecoxib group beat placebo, with a response rate 10 percentage points higher.
One detail kept it from being written off completely: among the three hundred or so people with more severe (moderate-to-severe) pain, the combination had a clearly higher response rate than placebo (79% vs 54%). But that was an exploratory analysis of a subgroup, and GAIT2, which set out two years later to test it, did not reproduce it.
So the position here is this: once all the trials are pooled, the effect is too small to matter clinically. If you want to use it, you can give it a serious three-month trial, record your pain and function, and see for yourself whether you respond; but do not expect it to regrow your joints, and it cannot replace weight loss, strength training, physiotherapy, anti-inflammatory drugs, or the doctor you should be seeing.
The result was disappointing: glucosamine and chondroitin, alone or together, did no better than placebo. Only the celecoxib group beat placebo, with a response rate 10 percentage points higher.
One detail kept it from being written off completely: among the three hundred or so people with more severe (moderate-to-severe) pain, the combination had a clearly higher response rate than placebo (79% vs 54%). But that was an exploratory analysis of a subgroup, and GAIT2, which set out two years later to test it, did not reproduce it.
So the position here is this: once all the trials are pooled, the effect is too small to matter clinically. If you want to use it, you can give it a serious three-month trial, record your pain and function, and see for yourself whether you respond; but do not expect it to regrow your joints, and it cannot replace weight loss, strength training, physiotherapy, anti-inflammatory drugs, or the doctor you should be seeing.
Evidence · The two-year follow-up and other trials
GAIT2: two more years of follow-upThe same investigators extended the follow-up as GAIT2 (Sawitzke 2010). That paper measured clinical outcomes only: not one image was taken and no cartilage was measured. The conclusion did not reverse. On the primary outcome, a 20% drop over 24 months in the WOMAC pain score (the most widely used pain and function questionnaire in osteoarthritis), no group beat placebo. The investigators specifically tested the interaction terms for the 2006 moderate-to-severe pain subgroup, and none was significant: that signal did not replicate.
Structure belongs to a different paper (Sawitzke 2008): joint-space width measured by X-ray over 24 months, with no significant difference between groups. Note that this paper used X-ray, not MRI, and it is not the 2010 paper.
Those two sentences are about two different things: whether a little less cartilage wore away (a structural outcome), and whether you actually hurt (a symptom outcome). Why the two can go separate ways is taken apart in the section The cartilage is not what hurts.
What the other evidence says
Further trials and national guidelines have assessed it again and again. Wandel 2010, a network published in the *BMJ*, pooled the large trials: glucosamine, chondroitin and the two together all had effects on joint pain too small to matter clinically, and none slowed the narrowing of joint space. The guidelines do not agree with one another: most do not list it as a recommended treatment; the Osteoarthritis Research Society International (OARSI) 2019 guideline strongly recommends against every glucosamine and chondroitin product, pharmaceutical grade included; and Europe's ESCEO is one of the few to give prescription-grade crystalline glucosamine sulfate a strong recommendation (that accounting is in the section Sulfate versus hydrochloride). So the practical reading is that you should treat it neither as first line nor as a scam.
That is the position taken here: neither a scam nor a good medicine, but something that once flashed a signal in one subgroup and is very weak once the trials are pooled.
Mechanism · The cartilage is not what hurts
Put the pain outcome next to the structural outcome and something looks odd: the joint space on X-ray was not preserved, yet some people report a little less pain. The cartilage did not get better, but the pain might have: how can both be true at once?The answer: the pain never came from the cartilage in the first place.
Cartilage has no nerves and no blood vessels
Joint cartilage is a white, slippery layer on the end of a bone, and its job is to let two bones slide over each other with almost no friction. To do that it has to be both smooth and dense, so it grows neither blood vessels nor nerve endings. That means if half the cartilage wears away, the cartilage itself does not hurt, because it has no line to send the message on.
So where does the pain come from? From the places around the cartilage that do have nerves:
The synovium: the thin membrane lining the inside of the joint that produces synovial fluid. Bits of worn cartilage that drop into the joint cavity are treated as foreign by immune cells in the synovium, which becomes inflamed, congested and swollen. It is packed with nerve endings, so once it swells it hurts. This pain feels swollen and warm, and rest does not ease it muchThe subchondral bone: the layer of bone beneath the cartilage, which is densely supplied with nerves. Once the cartilage thins, the impact of each step is no longer absorbed by that shock pad and lands directly on the bone; over time tiny fractures and bone-marrow edema appear. This is a deep, dull ache that gets worse as soon as you put weight on the jointThe joint capsule and ligaments: when the joint swells they are stretched, and that tight feeling and the stiffness when you get up in the morning come from hereThe surrounding muscles: held tense and compensating for a long time because of the pain, they become a new source of pain themselves
So both things can be true at once
If something can calm the synovium's inflammation a little, you will feel less pain, and it does not need to thicken the cartilage at all. Conversely, something that truly preserves cartilage may not make today's walk any easier in the short run. The pain outcome and the structural outcome are two independent lines: trials have to measure them separately, and so should you when you judge for yourself.
This mechanism also explains the counter-intuitive fact from the chapter Building blocks of cartilage: why someone with a terrible X-ray may not hurt while someone with a mild one may hurt badly. The X-ray measures how thick the cartilage is, while whether it hurts depends on whether the synovium is inflamed and whether the subchondral bone can take the load, and those two are only loosely related to cartilage thickness.
It also gives you a more practical use: when you judge any treatment for a joint, first ask which end it is trying to move. Is it trying to make the cartilage wear a little less (slow, hard, and years before you can see it), or trying to make the synovium less inflamed (fast, something you can feel, and known within weeks)? The two run on completely different timescales, and a three-month trial was never going to test the first.
Evidence · Why the argument has lasted so long
If something were truly useless, the argument usually would not last this many years. Glucosamine is still being argued over because several things landed on top of one another, and each is enough to make different trials come out differently.The salt is not the same: the trials with positive results mostly used the European prescription-grade sulfate, while GAIT used the hydrochloride. Whether the two are truly equivalent is still something nobody can say for sure (the section Sulfate versus hydrochloride works through that accounting in detail).
How much is actually in the product: the real content of cheap pharmacy products often does not match the label. A product with half the labeled amount will of course test negative, but that measures the product, not the molecule.
The people who benefit may be a small group: in GAIT's subgroup, people with moderate-to-severe pain seemed to respond while those with mild pain did not. Average the two groups together and a small effect that is really there gets diluted until you cannot see it. That is also why negative overall and some people really did get better can both be true. Remember, though, that this subgroup signal was not reproduced in the two-year follow-up.
Pain is extremely easy for a placebo to move: osteoarthritis pain is subjective and it fluctuates. In GAIT, 60.1% of the placebo group had their pain drop by at least 20%. Measuring a small effect against that much background noise takes a very large sample, and most trials are not that large.
Put these four together and you get a very practical judgment: for this class of supplement, the conclusion of any single trial does not mean much. What matters is how much effect is left once the trials are pooled. What is left after pooling is very weak.
Evidence · Sulfate versus hydrochloride
Why are some randomized trials positive and others negative? To a large extent, it depends on which salt was used.Glucosamine sulfate and glucosamine hydrochloride:
The chemical difference: the paired ion differs, the glucosamine itself is the sameContent: 1500 mg of the sulfate contains about 1180 mg of glucosamine; 1500 mg of the hydrochloride contains about 1240 mg (slightly more)Bioavailability: most studies find the two similar (about 25-44% absorbed and available after an oral dose, mainly in the small intestine)
Why do people think the sulfate is better? Europe's prescription-grade Dona (the originator product of the Italian company Rotta) is crystalline glucosamine sulfate (CGS), made under strict quality control. Most of the positive randomized trials (Reginster 2001, published in *The Lancet*; Pavelka 2002) used Dona, while GAIT 2006 (*NEJM*) used the hydrochloride and came out negative. So the difference may lie not in a different salt but in Dona's quality: a 1500 mg label means 1500 mg in the product, purity above 99%, 1500 mg taken once a day rather than in divided doses, plus 3 or more years of sustained research.
Europe's ESCEO 2019 guideline (Bruyère 2019) gives prescription-grade crystalline glucosamine sulfate (pCGS, that is, Dona) a strong recommendation, and at the same time explicitly does not recommend non-prescription glucosamine (the hydrochloride, or sulfate that is not Dona). ESCEO uses the strong and weak levels of the international GRADE system, not letter grades. This is a clear example of the gap between drug regulatory standards and food supplement standards. Do not forget the other side: OARSI 2019 strongly recommends against every product, pharmaceutical grade included, so the two societies take directly opposite positions here.
What it means in practice:
In Europe: prescription Dona (roughly ¥500 or more a month) costs more than non-prescription products but has stronger randomized-trial supportIn the US and China: with no Dona prescription available, a non-prescription product with third-party certification (USP, NSF) is the next-best optionAvoid: cheap products without third-party certification (the labeled content may be inflated, with low actual content and low purity)
Chondroitin varies just as much in quality: European prescription-grade chondroitin is highly pure (above 95%) with a standardized molecular weight, while non-prescription products are often found to contain cheaper polysaccharides mixed in, with actual content that does not match the label. Adulteration is common and documented, but there is currently no verified quantitative figure for what the average purity is.
In practice · How to choose and how to trial it
A checklist for choosing a product:1. Look for third-party certification: USP, NSF, ConsumerLab Approved
2. Trust the certification, not the brand: common overseas brands such as Doctor's Best, Jarrow, Now Foods, Pure Encapsulations and Thorne have certified products, but check that the bottle in your hand carries the certification mark
3. A clearly stated form: the label says glucosamine sulfate 1500 mg, not glucosamine complex
4. Avoid: miracle joint formulas with 20 ingredients and the like, which are mostly dilution plus marketing
5. Price versus content: ¥100-300 a month is reasonable; ¥30-50 a month very likely means too little active ingredient; ¥500 or more a month is not necessarily better
Turning this into action:
If you want to use glucosamine plus chondroitin and can get Dona (in Europe, or bought from abroad), use DonaIf you cannot get Dona: use a USP-certified non-prescription glucosamine sulfate at 1500 mg a day, plus chondroitin at 1200 mg a dayTrial it for 3 months, score yourself on the WOMAC questionnaire before and after, then decide whether to continue or stopDo not choose an uncertified product just because it is cheaper; the money wasted often exceeds the money saved
Chapter 3
What actually helps knee pain
If the goal is a knee that actually feels better, rank the options by strength of evidence and glucosamine sits fairly far down the list.
The real first line is not a drug at all; it is weight loss plus exercise. In IDEA, a randomized trial of people with knee osteoarthritis who were overweight or obese, the target was to lose 10% or more of body weight. After a year and a half, the diet-plus-exercise group had lost a little over a tenth of their weight on average, and their knee pain scores had fallen by nearly half from the start, more than in the group that only exercised. Strength training and aerobic exercise reduce pain and improve function, with almost no side effects.
Why does weight loss work so well? When you walk, your knee bears a force several times your body weight, so every pound you lose takes several pounds off the joint. Fat tissue also releases inflammatory signals into the blood; in this trial, the two weight-loss groups had lower blood levels of the inflammatory marker than the exercise-only group. Losing weight pushes down on both routes, load and inflammation, at once.
The real place of glucosamine plus chondroitin on this list is as an add-on, not first line: for people who have already done the weight loss, the training and the anti-inflammatory drugs and are still left with persistent moderate pain. The way to use it is to take it while recording pain and function, look at the results objectively after three months, keep going if things are clearly better, and stop and save the money if nothing has changed.
The real first line is not a drug at all; it is weight loss plus exercise. In IDEA, a randomized trial of people with knee osteoarthritis who were overweight or obese, the target was to lose 10% or more of body weight. After a year and a half, the diet-plus-exercise group had lost a little over a tenth of their weight on average, and their knee pain scores had fallen by nearly half from the start, more than in the group that only exercised. Strength training and aerobic exercise reduce pain and improve function, with almost no side effects.
Why does weight loss work so well? When you walk, your knee bears a force several times your body weight, so every pound you lose takes several pounds off the joint. Fat tissue also releases inflammatory signals into the blood; in this trial, the two weight-loss groups had lower blood levels of the inflammatory marker than the exercise-only group. Losing weight pushes down on both routes, load and inflammation, at once.
The real place of glucosamine plus chondroitin on this list is as an add-on, not first line: for people who have already done the weight loss, the training and the anti-inflammatory drugs and are still left with persistent moderate pain. The way to use it is to take it while recording pain and function, look at the results objectively after three months, keep going if things are clearly better, and stop and save the money if nothing has changed.
Clinical · The steps of pain relief and their costs
Weight loss plus exercise is the foundation, but a foundation works slowly, and the pain is happening now. The steps below are ordered from light to heavy by effect plus cost.Anti-inflammatory drugs (, non-steroidal anti-inflammatory drugs)
When the pain is acute and the joint is inflamed, anti-inflammatory drugs really do work: they press on the very pathway that keeps the synovium inflamed and makes pain nerves more and more sensitive. The costs:
Oral ones (ibuprofen, naproxen and similar): effective, but hard on the stomach, the cardiovascular system and the kidneys, so long-term use calls for particular careDiclofenac gel rubbed on the knee: the drug mostly stays where it is applied and whole-body side effects are small, so it is worth trying first; the OARSI 2019 guideline strongly recommends topical anti-inflammatories for knee osteoarthritis. A joint that sits close to the surface, like the knee, suits it especially wellCelecoxib: kinder to the stomach and gut; as with other oral anti-inflammatories, people with heart or blood-vessel disease should weigh it up with their doctorAcetaminophen (paracetamol, Tylenol): weaker but safer; current guidelines have moved it down from first choice, though it is still useful
Injections into the joint
Steroids: calm the synovium's inflammation directly, with good short-term (3-6 weeks) pain relief. But repeated injections may actually speed up cartilage degeneration, so this is a bridge, not a long-term planHyaluronic acid: aims to restore the lubricating quality of the joint fluid; the evidence is mixedPlatelet-rich plasma (PRP): your own blood is drawn, the platelets concentrated and injected back into the joint. It is commercially very popular, the evidence is equally mixed, and some guidelines do not recommend it
(The details, costs and guideline disagreements for these three injections are set out in full in the chapter Should you try it for knee pain?.)
Surgery
Only at a late stage, when every conservative option has failed, does surgery come in, from meniscus repair to total knee or hip replacement. Surgery can replace a worn joint surface, but it cannot replace muscle or habits, and recovery afterward still depends on strength training.
Put the steps back in order: the earlier ones are cheaper and safer and ask more of your own time; the later ones are faster, more expensive and harder to undo. Glucosamine plus chondroitin belongs in the slot marked foundation done, still a little short, not ahead of any of these steps.
mcalindon-2017-jama-corticosteroid
Evidence · Other joint supplements at a glance
The joint aisle of the supplement shelf is crowded, so here is a quick ranking of the ones you hear about most, by strength of evidence only:MSM (methylsulfonylmethane): supported only by scattered small trials; large trials are missingCurcumin: shows an anti-inflammatory signal, but it is inherently very poorly absorbed, so how much of what you swallow reaches the blood is a big problemOmega-3 fatty acids: more a whole-body anti-inflammatory than something aimed at the jointsCollagen peptides: a few small randomized trials have seen joint pain ease slightly, for example Clark 2008, a 24-week trial in athletes with activity-related joint pain; but most participants did not have osteoarthritis, and the results stop at subjective symptoms, with nothing showing cartilage being rebuilt (collagen peptides have a story of their own with the details)Boswellia extract: supported by some small trials; the evidence is limited (S-adenosylmethionine): of some use, but expensive, and it calls for caution when combined with antidepressantsVitamin D: people who were already deficient may find joint pain eases somewhat once the deficiency is corrected; people who are not deficient get no extra benefit from more
There is a general way to read this list: the effects of these things are almost all small, while the opponents they have to beat (weight loss, strength training) have large effects. So the real question was never does this supplement have a little use, but where does it rank for your limited time and money.
clark-2008-collagen-joint
Myth · Does eating cartilage rebuild cartilage
Chicken feet, pig trotters and bone broth nourish the joints: this is one of the stickiest claims in the Chinese-speaking world. It does have a shadow of a reason behind it, since these foods really do contain collagen and a little glucosamine. But the reasoning breaks once they reach the stomach.Why it breaks in the stomach
Proteins and polysaccharides are large molecules, and the gut wall does not let them through whole. Stomach acid first unravels collagen's triple helix, and protein-cutting enzymes from the pancreas and the gut wall then cut it into individual amino acids and short peptides; the sugar chains are likewise cut into single sugars. What actually enters the blood is amino acids, short peptides and single sugars, generic parts that carry no label saying I came from cartilage, please deliver me to the knee.
Once in the blood, these parts join the body's pools of amino acids and sugars and go wherever something is being built: they may go to repair muscle, to make enzymes, or simply be burned for energy. The share that goes to joint cartilage depends on whether the cartilage cells happen to be making matrix at the time, and has little to do with whether you drank bone broth or soy milk that morning.
This is the simplified thinking of like feeds like. The same logic applies everywhere: eating pig brain does not feed your brain, eating collagen does not turn directly into collagen in your skin, and eating cartilage does not grow directly into cartilage. The body's step of taking things apart and rebuilding them erases any trace of where the raw material came from.
Stack on top of that the earlier gate: cartilage has no blood vessels. Even if the blood really does carry a little more raw material, there is no pipe running straight to the cartilage cells; it still has to seep in, bit by bit, with joint movement.
So are these soups and dishes a waste?
Not that either. The protein and zinc in them really are raw materials for building cartilage matrix; they are just generic raw materials. Eggs, lean meat and soy foods supply the same things, more cheaply and with less fat. So drinking bone broth is not wrong; what is wrong is treating it as a way to target the joints.
Chapter 4
Should you try it for knee pain?
If your knee hurts, should you try glucosamine or not? Walk through the steps below and the answer becomes clear.
Step one: find out what the problem is. See a doctor (an orthopedic specialist or a rheumatologist) to rule out other causes such as rheumatoid arthritis, gout or a torn meniscus, get an X-ray to see how far the degeneration has gone, and have your weight and waist measured while you are there.
Step two: build the foundation that helps most, which always comes ahead of any supplement. If you are overweight, lose 5-10% first; train your quadriceps, hips and core to take load off the knee; for aerobic exercise choose low-impact options such as swimming, cycling or an elliptical trainer, and do not run until the pain has settled; add physiotherapy; and during a flare, start with a topical anti-inflammatory gel.
Step three: only now does glucosamine plus chondroitin come in. Who is worth a trial, how much to take, how to score it after three months, and who should avoid it are all covered next.
Two warnings that matter: do not just grit your teeth through joint pain, because left alone for years it distorts the way you walk and in turn drags down your hips, lower back and the other knee; and no pill can put back cartilage that has already worn away, so losing weight and training early to slow the degeneration will always do more than any supplement taken at a late stage.
Glucosamine plus chondroitin is the kind of supplement that has a hint of a signal, but not a strong one. No scare tactics and no hype: whether to spend the money is your call.
Step one: find out what the problem is. See a doctor (an orthopedic specialist or a rheumatologist) to rule out other causes such as rheumatoid arthritis, gout or a torn meniscus, get an X-ray to see how far the degeneration has gone, and have your weight and waist measured while you are there.
Step two: build the foundation that helps most, which always comes ahead of any supplement. If you are overweight, lose 5-10% first; train your quadriceps, hips and core to take load off the knee; for aerobic exercise choose low-impact options such as swimming, cycling or an elliptical trainer, and do not run until the pain has settled; add physiotherapy; and during a flare, start with a topical anti-inflammatory gel.
Step three: only now does glucosamine plus chondroitin come in. Who is worth a trial, how much to take, how to score it after three months, and who should avoid it are all covered next.
Two warnings that matter: do not just grit your teeth through joint pain, because left alone for years it distorts the way you walk and in turn drags down your hips, lower back and the other knee; and no pill can put back cartilage that has already worn away, so losing weight and training early to slow the degeneration will always do more than any supplement taken at a late stage.
Glucosamine plus chondroitin is the kind of supplement that has a hint of a signal, but not a strong one. No scare tactics and no hype: whether to spend the money is your call.
In practice · Who should try it and how much
Who is worth a trialYou have done the basic measures for 6-12 weeks and moderate-to-severe pain remainsYou do not want to take oral anti-inflammatory drugs (because of worries about your stomach or kidneys)Spending a few tens to a hundred-odd yuan a month is acceptable to youAnd you are willing to trial it seriously for three months before judging
Who should skip it
People with only mild pain: the placebo effect is too large, and you can easily fool yourselfPeople on a tight budget: it is not good value, and the same money spent on a pair of well-fitting shoes or a training class returns morePeople who have not started the basics: get weight loss and strength training going first, since that is the slot with the biggest effect
Form and dose
Glucosamine sulfate 1500 mg a day (the European prescription dose), or split into 500 mg three times a dayIf you want to add chondroitin, 1200 mg a day (the combination GAIT used)Stick to products with USP or NSF third-party certification, and do not buy cheap products whose labeled content cannot be trusted
How to score it after three months
Find a WOMAC pain and function questionnaire (you can find one online) and put your starting score next to your score after three months:
Clearly better (30% or more): it may be working for you (part of it may also be natural ups and downs and placebo), and continuing is reasonableOnly a little better (10-20%): most likely the placebo effect at work; it depends on whether you want to keep payingNo change at all: stop and save the money
Why write the scores down before comparing? Because pain goes up and down anyway, and memories of I think it has been a bit better lately are extremely unreliable: if today happens to be a good day, you will file it as the pill working. A written score will not cover for you.
One more point about time: a three-month window tests the pain end, not the cartilage end. Changes in cartilage are measured in years, and no three-month trial will see them. So a verdict of no improvement means only that your pain did not improve.
Safety · Who should skip it or watch closely
Overall it is very mild; after all, it is something the body makes every day, and the main issue with large oral doses is not toxicity. But a few things call for care:Shellfish allergy: glucosamine is mostly extracted from shrimp and crab shells, so people allergic to shellfish should not use itDiabetes: it is itself an amino sugar, and at large doses it may raise fasting blood glucose slightly (1-3%). The effect is small, but people whose medication is being adjusted should monitor itPeople taking warfarin: there are case reports of the (the lab test that shows how strongly warfarin is thinning the blood) rising after the two were combined, so keep an eye on the INR. This is the most important point, because if something goes wrong the result is bleeding, not just discomfortPregnancy and breastfeeding: there are no human safety data, so skip it
The two earlier warnings, spelled out
Do not grit your teeth through joint pain. When something hurts, people unconsciously change the way they walk to avoid the painful spot: less weight on the sore side, shorter steps, a slightly tilted pelvis. That compensation shifts the load onto the hips, the lower back and the other knee, which then compensate in turn, so the number of joints that finally need treating goes from one to three, and the two later ones were fine to begin with.
No pill can put back cartilage that has already worn away. Cartilage loss seen on imaging is mostly irreversible: cartilage has no blood vessels, so the cells and raw materials needed for repair cannot get in, and cartilage cells themselves divide very poorly and after adulthood basically stop rebuilding that matrix. So the real lever was always earlier: losing weight and training early to slow the degeneration will always do more than any supplement taken at a late stage.
Clinical · The options for joint injections
If glucosamine does not work or is not enough, what comes next? Further down the line are stronger measures such as joint injections and surgery. Before reaching this step, weight loss plus exercise (the approach with the largest effect in trials such as IDEA) should already have been done; and glucosamine itself had only a small effect when the trials were pooled in a network (Wandel 2010), the same conclusion GAIT reached.Injections into the joint
Steroids (glucocorticoids):
Common drugs: methylprednisolone, triamcinolone, betamethasoneEffect: fairly clear pain relief within 3-6 weeks, with a commonly quoted response rate of 70-80%Downsides:Repeated injections may speed up cartilage degeneration. In a double-blind randomized trial (McAlindon 2017, published in *JAMA*), people with knee osteoarthritis and synovitis were given triamcinolone every 3 months (4 times a year) for 2 years: their cartilage thinned more than in the group given saline, and their pain was no betterBlood glucose rises briefly (people with diabetes should take note)The risk of infection inside the joint is about 1/3000, rare but seriousClinical role: a short-term bridge for an acute flare or for moderate-to-late osteoarthritis
Hyaluronic acid (HA):
A component of joint fluid; injecting it adds lubricationUsually given as injections once a week for 3-5 weeks, with relief lasting 4-6 monthsThe guidelines disagree:The American Academy of Orthopaedic Surgeons (AAOS) 2021 guideline: not recommended for routine use (a strength of recommendation AAOS rates as moderate)OARSI 2019: a conditional recommendationThe latest version of the Cochrane review of these injections (viscosupplementation) is Bellamy 2006Brands include Synvisc and Hyalgan, at ¥3000-8000 per injectionIt varies from person to person: some people feel a clear benefit, others none
Platelet-rich plasma (PRP):
Platelets are taken from your own blood, concentrated and injected into the jointIt contains growth factors such as PDGF, TGF-β, (insulin-like growth factor 1) and VEGF, which in theory drive repairEvidence: in the RESTORE trial (Bennell 2021, published in *JAMA*, 288 people with knee osteoarthritis), PRP compared with a placebo injection made no significant difference to pain at 12 months, nor to cartilage volume. Small early trials were positive and the large randomized trial turned negative, much like the GAIT storyThe two guidelines disagree here: OARSI 2019 strongly recommends against it, while AAOS 2021 gives it a limited-strength recommendation, saying it may reduce pain and improve function. Guidelines contradicting each other is itself a sign that the evidence is not there yetYet it is commercially very popular and heavily promoted by private clinics, at ¥3000-8000 per injectionThe evidence is weak and the cost high; on current evidence it does not deserve a place near the front
Stem cells and mesenchymal stem cells (MSC): plenty of hype, very weak clinical evidence. Some hospitals in China and stem-cell centers abroad charge tens of thousands of yuan, up to ¥150,000. The US FDA has warned that most unapproved stem-cell therapies do not work and carry risks; they are not recommended.
mcalindon-2017-jama-corticosteroidbennell-2021-jama-restore-prp
Clinical · When to consider surgery
Surgery: when to consider itSurgical assessment comes in only when conservative treatment has gone on for 12 months or more and none of the measures above has worked, when the imaging shows late-stage disease (Kellgren-Lawrence grade 3-4, meaning clearly narrowed joint space and prominent bone spurs), and when quality of life is seriously affected.
Arthroscopic debridement: in a randomized trial with a sham-surgery control (Moseley 2002, published in the *NEJM*), it made no difference compared with sham surgery; not recommendedPartial meniscectomy: indicated only for an acute tear with symptoms of the joint catching or lockingOsteotomy: for younger, active people with osteoarthritis in only one compartment of the knee, to redistribute the line of force through the legUnicompartmental knee replacement (partial knee replacement): for late-stage disease in one compartment only; it preserves the ligaments and recovery is quickTotal knee replacement: the standard procedure. A commonly quoted figure is that about 95% of people improve substantially over 10 years; it suits people over 60 with severe osteoarthritis whose conservative treatment has failed; implants last 15-20 years; the risks include infection (1-2%), venous blood clots (VTE, 1-3%), loosening of the implant and needing a revision operation
Thinking about timing:
Too early: an implant lasts 15-20 years, so one put in at 50 may need revising by 70, and second operations succeed less oftenToo late: long-term chronic pain worsens gait, cardiovascular health and mood (depression)Closer to ideal: over 65, late-stage imaging, failed conservative treatment, and quality of life already seriously affected
The real balance between keeping the joint and replacing it:
Moderate osteoarthritis at 45-60: weight loss, training, physiotherapy, occasional anti-inflammatory drugs and steroid injections as a bridge; hold off and maintain for as long as possibleLate-stage osteoarthritis at 60-70: assess for surgery (partial or total knee replacement)Active people over 70: weigh the risks of surgery against the benefits
After surgery, recovery depends on slowly rebuilding muscle strength, and the measures that guard against sarcopenia (strength training, enough protein) apply just as well; the risk of the operation itself is tied to cardiovascular health; and weight loss is a precondition from start to finish.
References · 9
- Clegg, D. O., Reda, D. J., Harris, C. L., Klein, M. A., O'Dell, J. R., Hooper, M. M., et al. (2006). Glucosamine, chondroitin sulfate, and the two in combination for painful knee osteoarthritis. New England Journal of Medicine, 354(8), 795-808. 1,583 patients, 24 weeks: glucosamine 1,500 mg, chondroitin 1,200 mg, both, celecoxib 200 mg or placebo. Response (>= 20% pain reduction) with placebo 60.1%; glucosamine +3.9 points (P = 0.30), chondroitin +5.3 (P = 0.17), combination +6.5 (P = 0.09), celecoxib +10.0 (P = 0.008). Assignment was stratified by pain severity (mild n = 1,229; moderate to severe n = 354); in the moderate-to-severe stratum the combination responded 79.2% vs 54.3% (P = 0.002), which the authors call exploratory (abstract, PMID 16495392). 10.1056/NEJMoa052771
- Wandel, S., Jüni, P., Tendal, B., Nüesch, E., Villiger, P. M., Welton, N. J., Reichenbach, S., & Trelle, S. (2010). Effects of glucosamine, chondroitin, or placebo in patients with osteoarthritis of hip or knee: network meta-analysis. BMJ, 341, c4675. 10.1136/bmj.c4675
- Sawitzke, A. D., Shi, H., Finco, M. F., Dunlop, D. D., Harris, C. L., Singer, N. G., et al. (2010). Clinical efficacy and safety of glucosamine, chondroitin sulphate, their combination, celecoxib or placebo taken to treat osteoarthritis of the knee: 2-year results from GAIT. Annals of the Rheumatic Diseases, 69(8), 1459-1464. GAIT 2-year continuation, CLINICAL outcomes only — no imaging, no cartilage. No arm beat placebo on 20% WOMAC pain reduction, and the interaction terms for the moderate-severe pain subgroup were NOT significant: the 2006 subgroup signal did not replicate. 10.1136/ard.2009.120469
- Sawitzke, A. D., Shi, H., Finco, M. F., Dunlop, D. D., Bingham, C. O., Harris, C. L., et al. (2008). The effect of glucosamine and/or chondroitin sulfate on the progression of knee osteoarthritis: A report from the glucosamine/chondroitin arthritis intervention trial. Arthritis & Rheumatism, 58(10), 3183-3191. This is the GAIT structural arm: 24-month joint space width by X-RAY, no significant difference between groups. The story used to attribute "MRI cartilage thickness" to the 2010 clinical paper. 10.1002/art.23973
- Bruyère, O., Honvo, G., Veronese, N., Arden, N. K., Branco, J., Curtis, E. M., et al. (2019). An updated algorithm recommendation for the management of knee osteoarthritis from the European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis and Musculoskeletal Diseases (ESCEO). Seminars in Arthritis and Rheumatism, 49(3), 337-350. ESCEO uses GRADE strong/weak, not letter grades. It gives prescription-grade crystalline glucosamine sulfate a STRONG recommendation — "Level B" was never in this document. 10.1016/j.semarthrit.2019.04.008
- Bannuru, R. R., Osani, M. C., Vaysbrot, E. E., Arden, N. K., Bennell, K., Bierma-Zeinstra, S. M. A., Kraus, V. B., Lohmander, L. S., Abbott, J. H., Bhandari, M., Blanco, F. J., Espinosa, R., Haugen, I. K., Lin, J., Mandl, L. A., Moilanen, E., Nakamura, N., Snyder-Mackler, L., Trojian, T., … McAlindon, T. E. (2019). OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis. Osteoarthritis and Cartilage, 27(11), 1578–1589. 10.1016/j.joca.2019.06.011
- Messier, S. P., Mihalko, S. L., Legault, C., Miller, G. D., Nicklas, B. J., DeVita, P., et al. (2013). Effects of intensive diet and exercise on knee joint loads, inflammation, and clinical outcomes among overweight and obese adults with knee osteoarthritis (IDEA). JAMA, 310(12), 1263-1273. 18 months; the diet goal was a mean loss of at least 10% of baseline weight (10-15%). Weight change: exercise -1.8 kg, diet -8.9 kg, diet + exercise -10.6 kg. WOMAC pain (0-20), full-text Tables 3-4: diet + exercise 6.7 to 3.7 (-45%; adjusted 3.6), exercise 6.1 to 4.4 (-28%), diet 6.6 to 4.8 (-27%); diet + exercise had less pain than exercise (difference 1.02). IL-6 fell with diet (-16%) and diet + exercise (-15%) but not with exercise alone, and knee compressive force fell most with diet (full text, PMC4450354). 10.1001/jama.2013.277669
- Brophy, R. H., & Fillingham, Y. A. (2022). AAOS clinical practice guideline summary: management of osteoarthritis of the knee (nonarthroplasty), third edition. Journal of the American Academy of Orthopaedic Surgeons, 30(9), e721-e729. On PRP it gives a LIMITED-strength recommendation IN FAVOUR ('may reduce pain and improve function'), which is not the same as recommending against — OARSI is the body that recommends against. 10.5435/JAAOS-D-21-01233
- Bellamy, N., Campbell, J., Welch, V., Gee, T. L., Bourne, R., & Wells, G. A. (2006). Viscosupplementation for the treatment of osteoarthritis of the knee. Cochrane Database of Systematic Reviews, (2), CD005321. 10.1002/14651858.CD005321.pub2