Green-Lipped Mussel vs Glucosamine: What the Force-Plate Trials Actually Found

Buy a green-lipped mussel lipid extract — not a mussel-powder chew — if your dog has diagnosed hip arthritis, six measured weeks on a glucosamine product changed nothing, and roughly $1,320 a year is money you can carry indefinitely. Buy the glucosamine chew if you want the cheap, safe, widely stocked option and you accept that the two best objective trials could not tell it apart from placebo. Buy neither if your dog is overweight or visibly sore: weight loss and a proper pain-relief conversation with your vet both have better evidence than anything on this page, and in both mussel trials it was the NSAID arm that moved the force plate too. We earn the same rate on all three products below. Nothing changes for us whichever you pick.

The distinction that reorganises this entire aisle#

There is one sentence you need before any of the numbers make sense, and almost nobody selling in this category will say it out loud.

The green-lipped mussel products that performed well in blinded trials were concentrated oil extracts. The green-lipped mussel products on the shelf at your feed store are dried whole-mussel powder. They are not the same thing, and the powder does not inherit the extract's results.

Here is the scale of the difference (and if you want the two shelf products this trial actually pits against each other, Antinol vs Dasuquin runs that head-to-head in full). The trials dosed 50 mg capsules of a marine lipid concentrate.^[1] YuMOVE's Advanced Vet Strength chew delivers 660 mg of ActivEase green-lipped mussel powder across a three-chew serving. GlycoFlex 3 — renamed "Extra Strength Healthy Hip & Joint" on its Amazon listing, verified 14 August 2026 — delivers 600 mg of Perna canaliculus per tablet. A shopper comparing "660 mg" against "50 mg" will conclude the chew is thirteen times stronger. That comparison is meaningless, because one number counts dried shellfish and the other counts an oil fraction pulled out of it.

We have no quarrel with a brand saying its chew contains green-lipped mussel. It does. What no brand may honestly do is point at the force-plate trials while selling the powder, and as of today we know of no published blinded trial in dogs that tested the powder at all.

So this page is really about three products, not two: the extract that was tested, the glucosamine chew that was tested, and the mussel powder that was not.

dogorthopedic.com

What the two trials actually did#

Both trials came out of Khon Kaen University in Thailand with B. Duncan X. Lascelles of North Carolina State University as senior author. Both were prospective, block-randomised, double-blinded and placebo-controlled, and both used peak vertical force — the load a dog actually puts through the sore leg as he trots across a plate in the floor, expressed as a percentage of his body weight. It is measured by equipment, not by opinion.

Sources: the 2023 paper^[1] and the 2024 paper.^[2]

The 2023 result, in the paper's own numbers#

Change in peak vertical force from baseline, mean ± standard deviation:

At week four, the three bolded arms were significant and significantly greater than both placebo and glucosamine. At week six the glucosamine group, in the authors' words, "lay between the placebo group and the other treatment groups." Their conclusion, verbatim: "PCSO-524 and EAB-277, but not glucosamine/chondroitin, resulted in significant improvements in PVF from baseline after 4 weeks, and 6 weeks, and to a similar degree to that seen with carprofen."^[1]

One detail from that trial deserves its own line, because it is uncomfortable for us and we would rather you heard it here.

The glucosamine arm was not generic glucosamine. It was Dasuquin with MSM, named in the paper, at 900 mg glucosamine hydrochloride, 350 mg chondroitin sulfate, 800 mg MSM and 90 mg ASU per chew — the large-dog formula, out of the same tub you can buy on Amazon.^[1] It is our second-best-earning recommendation on this site. We are telling you anyway, because a comparison page that hides the result that costs it money is not worth reading.

The 2024 result, and the number that ends the argument#

The second trial dropped glucosamine and tested EAB-277 against a plant extract, an NSAID and placebo in 101 dogs. At six weeks, EAB-277 and meloxicam both beat placebo on force plate, and 4CYTE did not differ from placebo.^[2]

But the group averages are not the useful part. The authors also counted individual dogs, defining a responder as one whose peak vertical force improved by more than 2%:

P < 0.001 across the groups.^[2]

Roughly seven dogs in ten improved measurably on the mussel extract. Eight in ten on the anti-inflammatory drug. Fewer than one in ten on the placebo. Nobody in the room — not the owners, not the vets collecting the data — knew which dog was on which.

That last clause is the whole reason these trials matter, and the next section explains why.

dogorthopedic.com

Axis one: why measurement changes the answer#

If you have ever been sure a supplement was helping and then wondered whether you were fooling yourself, the literature has an unusually precise answer for you, and it is not an insult.

Fifty-eight dogs with osteoarthritic lameness were enrolled in the placebo arm of an FDA-approved study with force platform analysis running throughout. Owners judged their dogs improved 39.7% of the time. Veterinarians did it 44.8% of the time at a walk, 44.8% at a trot, and 43.1% when palpating the joint. The mean ground reaction forces did not move at all, and the effect grew stronger the longer the trial ran.^[3]

Four in ten. On a sugar pill. Watched by people who love the dog and, in the vets' case, by people trained to assess lameness for a living.

This is not a story about gullible owners. It is a story about a hard measurement problem: arthritis fluctuates, dogs have good days, weather changes, and hope is a lens. It is the reason the joint supplement aisle can be full of products that are genuinely believed in and have never been shown to work. And it is why an uncontrolled, open-label study — the kind most supplement brands cite, if they cite anything — tells you close to nothing here.

The Kampa trials are worth your attention because they removed that lens twice over: nobody knew who got what, and the outcome was read off a machine.

Axis two: what the glucosamine result does and does not establish#

This is the section where a page with an agenda would either bury the caveats or pretend the case is closed. Here is the honest accounting, in both directions.

What makes the glucosamine result stronger than one trial#

The usual defence of glucosamine is that generic products do not inherit the trial record, because the published canine studies used Nutramax's specific patented forms — FCHG49 glucosamine and TRH122 chondroitin. That argument is real and we have made it ourselves on Dasuquin vs Cosequin. It cannot be used against this trial, because this trial used the branded Nutramax product containing exactly those forms.

And there is independent replication of the null, which is more than the mussel side can currently say. A separate group at a separate institution ran a double-blind, randomised, placebo-controlled trial of a glucosamine/chondroitin product in 60 client-owned dogs — 30 treated, 30 on placebo — for 97 days, with activity monitors worn continuously rather than a force plate. No difference in daily activity counts between the groups. Owner-reported pain scores improved over 90 days in both arms, with no difference between them, which the authors described as suggesting a caregiver placebo effect.^[4] Different research group, different country, different objective instrument, different duration, same answer.

That second study is by the same authors who measured the 39.7% placebo rate. They went looking for the effect they had already characterised, and they found it sitting in both arms of their own trial.

What the glucosamine result does not establish#

Five things, and the exacting reader will hold us to all of them.

1. Fifteen dogs per arm cannot rule out a modest benefit. The 2023 sample size was calculated to detect the PCSO-524 effect, not to exclude a smaller glucosamine one. A null result in a study powered for someone else's effect is weak evidence of absence.
2. The dose was maintenance, not loading. The trial gave about 30 mg of glucosamine per kg of body weight daily for the full six weeks. For a 32 kg dog that is roughly one large-dog chew. Nutramax's own label prescribes two chews a day for dogs over 60 lb during exactly that first four-to-six-week window — about 57 mg/kg. The trial therefore ran maintenance dosing through the period the manufacturer says to double. That is a fair criticism and it belongs in the record.
3. Six weeks is short for a product whose advocates argue for slow structural effects.
4. Peak vertical force is one narrow question. It measures limb loading. It does not measure whether the dog is more comfortable lying down, sleeping better, or willing to do stairs. In the 2023 trial the subjective orthopaedic scores did not change significantly in any group — including the carprofen group, which the force plate showed clearly working. An instrument that misses an NSAID's effect will miss other things too.
5. The older literature is mixed, not uniformly negative. A 2017 review in Open Veterinary Journal concluded that despite limited and conflicting evidence, glucosamine and chondroitin are commonly recommended, and that "there is a paucity of well-designed clinical veterinary studies investigating the true treatment effect."^[5] A 2025 review in the Journal of Veterinary Medical Science asks the question in its own title — "do they work?" — and its summary table sets directly conflicting canine results side by side, with McCarthy 2007 reporting significant improvement and Moreau 2003 reporting that the nutraceutical did not improve pain or functional disability.^[6]

The honest summary: glucosamine and chondroitin have not been shown to work when anyone measured properly, twice, by two groups. That is not the same as proof they do nothing, and anyone who tells you it is has overreached in the opposite direction. What the evidence takes away is the confidence, not the product.

Axis three: who paid for this, and who has replicated it#

Independence is part of evidence strength, so here is the part of the story that cuts against the mussel extract.

Both trials were funded by Pharmalink International Limited, which owns the EAB-277 and Antinol trademarks.^[1][2] Both papers state that the funder was not involved in study design, data collection, analysis, interpretation, writing or the decision to publish. The 2023 paper additionally discloses that the senior author has been a paid consultant for VetzPetz Asia Company, while stating he received no payment in connection with that study.

None of that makes the trials wrong. Industry funds most veterinary trials; if we discounted every industry-funded study we would have almost no evidence base left, and the design here — blinded, placebo-controlled, objectively measured, with the negative results on 4CYTE and glucosamine published rather than buried — is more rigorous than most of what this category produces. A funder willing to publish "our competitor's category did nothing and so did another supplement we tested" has behaved better than the norm.

But two facts stand:

• Both trials come from one research group. No independent team has published a replication of the mussel-extract result.
• The glucosamine null has been replicated independently. The mussel-extract positive has not.

That asymmetry is the single most important thing to hold in your head, and it points in an awkward direction: the finding we are most confident about is a negative one.

Axis four: the money, per day, for a 70 lb dog#

Every price below was pulled through the Amazon Creators API on 13 August 2026. Doses are the manufacturers' own, at their own weight bands.

Read the first and third rows together, because that pairing is the actual decision on this page. The product with the best evidence costs about six times the product with the weakest, and the product in the middle on price has the least evidence of the three. YuMOVE is roughly twice Dasuquin's annual cost while carrying mussel in the form nobody has tested, and its own "clinically proven in 6 weeks" claim traces to an unpublished study at a UK veterinary college it declines to name — nothing indexed on Europe PMC as of 12 August 2026. We walked that claim back on our own pages and logged it in our corrections log.

Here are the three, with what we think of each.

Dosing and sizing: where the tested dose quietly disappears#

Getting the dose wrong is how people conclude a product failed when they never ran it. On this pair, both sides have a sizing trap, and they are different traps.

The extract's trap is that its label stops scaling. Antinol Plus gives two soft gels a day up to 44 lb and four a day at 45 lb and over, flat, with no further band. A soft gel is 50 mg of EAB-277, so four is 200 mg whether the dog is a Springer or a Newfoundland.

A 45 lb dog on this label sits close to the 2024 trial's dose. A 70 lb dog sits between the two trials. A 120 lb dog sits below both.^[1][2][7] So if your dog is large and six weeks produce nothing, "it did not work" is not the only conclusion available. Raise the dose question with your vet rather than deciding it yourself in the kitchen, and note that at four a day this is already $1,320 a year.

The glucosamine trap is the opposite: the loading dose people skip. Dasuquin's schedule for dogs over 60 lb is two chews daily for four to six weeks, then one. A 150-count tub is 75 days of loading, not five months. Half a loading phase judged by feel produces exactly the 39.7%-shaped answer the literature warns about. Our dosing-by-weight reference has the band-by-band arithmetic for every product we cover.

And the mussel powder's trap is that its maker contradicts itself. YuMOVE's own product page prescribes three chews a day at maintenance for a 66–100 lb dog while its Amazon listing says "just one chew a day at maintenance for the average dog." That is a threefold difference, and it decides whether the product costs you about $142 or about $426 a year.

Who should buy neither#

Most readers, honestly. In order, cheapest first.

1. The measurement, which is free. Before you spend anything, write down what you would need to see in six weeks to keep paying. Score one specific thing — the first three minutes after he gets up, the jump into the car, the stairs — or use our printable supplement trial scorecard. Without a baseline you are buying a product and a placebo effect at the same price, and you will never be able to separate them. The method for scoring fairly is in is my dog's limp getting better.
2. Weight, which is also free to ask about. Ask for a body condition score at the next appointment. For an overweight arthritic dog, taking weight off is the intervention with the least controversy behind it, and it is cheaper than every row in the price table above.
3. The pain-relief conversation with your vet. This is the finding people skip past: in both trials, the NSAID arm worked. Carprofen and meloxicam were not the control group that failed — they were the benchmark the good supplement matched. If your dog is sore now, the supplement question is not the urgent one.
4. Anyone who cannot sustain $1,320 a year indefinitely. Six weeks of benefit followed by stopping on cost is the worst outcome on this page: you pay for the trial and keep none of the result.
5. Anyone running three products at once. You will learn nothing about any of them. One product, six weeks, one measurement.

The channel note#

Normally this section compares the maker's own price against Amazon's, and we publish it even when it costs us the commission. On this pair we cannot, and the reason is itself worth knowing.

Antinol Plus has no live US direct store. antinolplus.com redirects to a Shopify storefront serving a password page titled "Opening Soon," and pethealthworks.com is a single page that sells nothing. Amazon is effectively the only US channel, so there is no direct price to compare and no independent returns route. The one real saving available is Amazon's own subscribe-and-save, which takes the 150-count from $135.99 to $122.39 — about $130 a year on a product at this dose.^[7]

Two count-size traps while you are there. The Antinol Plus 90-count is the worst value of the three, at $0.95 a soft gel against $0.91 for the 150-count — buying the middle size does not split the difference, it loses to both ends. And on Dasuquin, subscribe-and-save takes the 150-count from $84.99 to $80.74.^[8]

What we could not verify#

• Whether the mussel-extract result holds outside this research group. No independent replication is published. This is the single biggest open question on the page.
• Antinol Plus's "four double-blind, placebo-controlled clinical studies." The claim is on the Amazon listing. No Antinol or Pet HealthWorks property names those four studies. The brand's own sites give four different counts — four, "more than 10," "15+" and "over 40" — and the one reference page we found lists 13 papers on the older PCSO-524 extract and none on EAB-277.
• The mussel and krill split in the US. The US facts panel publishes only the 50 mg blend total; the 30 mg mussel / 20 mg krill breakdown appears on the brand's New Zealand and Australian sites but not for American buyers.
• Whether the glucosamine result would change at the loading dose. Nobody has run that trial. It is the obvious next study and we could not find it.
• What ASU contributes. The closest canine evidence is a 2025 randomised, double-blind trial of a combination supplement containing fish oil, ASU and plant extracts against an NSAID — which reported comparable efficacy, but enrolled 20 dogs, had no placebo arm, and cannot isolate what ASU did.^[9] Given a 39.7% caregiver placebo rate, a two-arm trial with no placebo cannot settle this. We corrected our own overstatement about ASU on 12 August 2026 and logged it.
• Calories per chew or soft gel for any product on this page. Not published by Nutramax, YuMOVE or Pet HealthWorks. Weight is the biggest modifiable factor in canine arthritis, and this category will not tell you what it is adding.

What to do this week#

Score something before you change anything. One specific, observable thing, written down tonight with today's date on it — that single act is worth more than the difference between any two products on this page, because without it you will be reading your own hope back to yourself in six weeks.

Then pick on the evidence rather than the shelf. If you are already on a glucosamine chew and have never run it properly, run it properly: the full loading dose, six weeks, judged against the number you just wrote down. If you have already done that and got nothing, the extract is the only thing in this aisle with a force plate behind it, and you now know exactly what it costs and exactly which questions about it are still open.

And if the dog is sore today, none of this is the priority. Book the appointment. In both of the trials on this page, the drug your vet can prescribe was the arm that worked.

The delay is the injury. A dog does not tell you the supplement stopped working, or never started. He just gets up more slowly, and everyone calls it age.