Collagen Science Update – October 2026
A Comparison of Clinical Outcomes Among Undenatured Collagen Type II, Glucosamine Sulfate, and Diacerein on Osteoarthritic Knee Treatment: A Randomized Clinical Trial
In this randomized controlled trial, the authors compared 3 common knee osteoarthritis options against a placebo over 6 months: undenatured type II collagen (UC-II, 40 mg a day), prescription-grade glucosamine sulfate, and diacerein, an anti-inflammatory drug. They enrolled 216 adults with mild to moderate knee OA, most of them women in their early 60s, with 96% staying in the study through the end. Every participant, including the placebo group, also followed the same program of exercise, lifestyle changes and education on managing their knee health. By the 3-month mark, scores for pain, stiffness and day-to-day function (measured on the WOMAC scale) had improved by roughly half in every group, placebo included. With everyone improving that much, none of the 3 treatments pulled meaningfully ahead of placebo at 6 months. UC-II finished about 5 points better than placebo on a scale where around 12 points is the change patients tend to notice, and that gap wasn’t statistically significant. Stiffness was the closest call for UC-II, landing right on the edge of significance, and safety was good across the board, with no serious side effects in any group.
So what should the industry take from this? On average, people taking UC-II improved about 5 points more than those on placebo, but the study wasn’t large enough to tell whether that small effect was real, or whether UC-II’s true effect might be big enough for patients to feel, which is why the authors call for bigger and longer trials. When good exercise and lifestyle care is already helping people with mostly mild OA this much, there isn’t a lot of room left for any additional interventions to show an extra effect. Collagen also works gradually over months of daily use, and cartilage changes even slower. The authors list the 6-month length as their study’s first limitation, noting that it was likely too short to pick up structural changes in the joint, and didn’t measure any markers of cartilage breakdown. Finally, UC-II is a milligram-level ingredient thought to work through the immune system in the gut, which makes it very different from the gram-level collagen peptides most people think of.
Key Takeaways:
• In a 6-month trial of 216 adults with mild to moderate knee OA, UC-II (40 mg a day), glucosamine and diacerein all failed to beat placebo on the main WOMAC score, and the follow-up measures didn’t differ between groups either.
• Every group improved by about half within 3 months, largely due to the exercise and lifestyle program all participants received, which left little room for any add-on to stand out, especially in people with mild OA.
• The authors claim 6 months was probably too short to see changes in the joint itself, so larger and longer studies that also track cartilage markers are the logical next step. Access the study: https://doi.org/10.2106/JBJS.OA.26.00225
Narkbunnam, R., Awirotananon, K., Chareancholvanich, K., Achawakulthep, C., Adulkasem, N., Chatmaitri, S., Hirunyachoke, P., & Pornrattanamaneewong, C. (2026). A comparison of clinical outcomes among undenatured collagen type II, glucosamine sulfate, and diacerein on osteoarthritic knee treatment: a randomized clinical trial. JB & JS Open Access, 11(3), e26.00225. https://doi.org/10.2106/JBJS.OA.26.00225
Developing an Analytical Framework for the Molecular Weight Analysis of Fish-Derived Collagen Peptides in Beverages: A Multi-Method Comparative Study
Molecular weight is one of the first numbers people check on a collagen peptide spec sheet, since smaller peptides are generally associated with better absorption and activity. The standard methods for measuring it were written for pure peptide powders, though, and a finished collagen drink is a much messier sample, with vitamins, acids, sweeteners and botanicals all mixed in. Researchers from China Agricultural University and Beijing Qingyan Boshi Health Management set out to see how well 3 established Chinese testing methods hold up, and whether they give the same answer.
They tested 2 of the company’s commercial drinks, each containing 5 to 6 grams of fish collagen peptides and a small amount of elastin peptides. One was paired with vitamin C, and the other with green tea extract (EGCG), hyaluronic acid and a few other ingredients. As a comparison, they also tested the plain peptide powder from the same production batches. All 3 methods gave consistent results when repeated, but the results also raised real concerns. First, other ingredients in the drinks showed up in the readings and could be mistaken for peptides; once the team set a cutoff to screen those signals out, the drinks’ peptide profiles closely matched the plain powder. Second, the standard way of converting test results into molecular weights uses a straight-line calculation, and
under that approach the 3 methods disagreed noticeably in several size ranges, including the under-1 kDa fraction that matters most for bioactivity. Switching to a curved (cubic) calculation, which better fits how this kind of test actually behaves, brought all 3 methods into close agreement, with less than 5% variation in the under-1 kDa range.
The authors propose that the industry adopt the curved calculation as a shared approach so results from different labs and methods can be compared directly. Until that happens, they suggest one of the 3 methods (an enterprise standard originally written for collagen tripeptides) as the most dependable option. This was a single-lab study on one instrument, using 2 drinks from one brand and testing methods based on Chinese standards, so it needs confirming on other equipment and in other products. For brands and suppliers, the practical message is that a molecular weight claim is only as meaningful as the method and calculation behind it, and those details are worth asking about.
Key Takeaways:
· Testing 2 commercial fish collagen drinks against their source peptide powder, researchers found that ingredients like vitamin C and green tea extract can interfere with molecular weight testing, and that screening out those signals made the drinks’ peptide profiles match the original powder.
· The 3 established testing methods disagreed in key size ranges when results were calculated the standard straight-line way, and a curved calculation brought them within 5% of each other in the under-1 kDa range most linked to bioactivity.
· The study was done in one lab using one brand’s products, but it’s a useful reminder that molecular weight claims depend on how they’re measured, and the authors call for a shared industry approach so results can be compared across labs.
Access the study: https://doi.org/10.1002/fsn3.72386
Feng, M., Li, Y., Luo, Y., Jiang, Y., & Zhao, C. (2026). Developing an analytical framework for the molecular weight analysis of fish-derived collagen peptides in beverages: a multi-method comparative study. Food Science & Nutrition, 14(9), e72386. https://doi.org/10.1002/fsn3.72386
Specific Collagen Peptides Support Postprandial Metabolic Health by Modulating Incretin Hormones, Gastric Transit, and Glycemic Control in Participants with Normoglycemia and Prediabetes
GLP-1 is front and center in conversations about metabolic health right now, and many people are asking whether foods and ingredients can support the body’s own GLP-1 response. This study tested that idea with collagen. It enrolled 30 adults, 12 with normal blood sugar and 18 with prediabetes, who drank either 10 grams of a specific collagen peptide (Nextida GC, made from pork gelatin) or a placebo half an hour before a carb-heavy meal. Each person tried both drinks on separate days, in random order, without knowing which was which. The researchers tracked blood sugar, insulin, and 2 gut hormones that help manage blood sugar and satiety, GLP-1 and GIP, and had participants swallow a small tracking capsule to see how long food stayed in the stomach.
In the half hour before the meal, the collagen drink increased insulin, GLP-1 and GIP compared with placebo. After the meal, blood sugar rose about 37% less with collagen, and there wasn’t any additional insulin release that was seen as well. In people with prediabetes, food also stayed in the stomach longer, which suggests slower digestion plays a part alongside the hormone response. A complication is that when examining the two groups separately, the decrease in blood sugar was statistically significant in individuals with normal blood sugar, but this significance was not observed in the prediabetes group alone.
The study also looked at a single drink before a single meal, so we don’t know yet whether taking it every day leads to lasting changes, and the findings are specific to this one peptide ingredient.
Key Takeaways:
• Taken 30 minutes before a carb-heavy meal, a 10-gram collagen peptide drink raised GLP-1, GIP and insulin and cut the after-meal rise in blood sugar by about 37% compared with placebo, without increasing total insulin.
• People with prediabetes also digested the meal slower, though the blood sugar benefit in that group on its own wasn’t statistically significant.
• This was a single-dose study funded by the ingredient’s maker, so longer, independent trials are needed to see whether daily use leads to lasting metabolic benefits. Access the study: https://doi.org/10.1016/j.cdnut.2026.109512
Virgilio, N., Schön, C., Steiner, N., Schlageter, V., Wilhelm, M., Silva, C. I., Gevaert, E., & Prawitt, J. (2026). Specific collagen peptides support postprandial metabolic health by modulating incretin hormones, gastric transit, and glycemic control in participants with normoglycemia and prediabetes. Current Developments in Nutrition, 10(9), 109512. https://doi.org/10.1016/j.cdnut.2026.109512
Oral Recombinant Human Collagen III Protects Gastric Mucosa via G-Quadruplex-Mediated Activation of the Pla2g4a-Prostaglandin Axis
Most collagen on the market comes from animals, usually cow hides, pig skin or fish skin and scales. Recombinant collagen is made a different way: engineered cells carrying the human collagen gene produce it, and it’s then purified, which gives you collagen that matches the human version without any animal tissue. This study from China Agricultural University looked at this unique collagen ingredient’s effectiveness in protecting the stomach lining.
The researchers gave rats recombinant human type III collagen by mouth every day for 32 days at 3 different doses, along with a group that got commercial fish collagen at the highest dose for comparison, and then gave the rats alcohol, which damages the stomach lining. The recombinant collagen cut the damaged area by 78% and protected about 1.5 times better than the fish collagen. When the team traced how it worked, they found that digestion in the stomach breaks the collagen down into a small peptide called GPAEF, which shows up within 30 minutes. In lab tests, that peptide switched on a gene the stomach uses to make prostaglandins, the compounds that help keep its lining protected. The collagen also supported the stomach’s antioxidant defenses and helped repair its protective barrier.
This is early animal research, and the doses were far higher than anything a person would take. Still, it’s an interesting first look at an animal-free collagen ingredient, and human studies will be needed to show whether it holds up.
Key Takeaways:
• Recombinant human type III collagen, made by engineered cells from the human collagen gene with no animal tissue involved, cut alcohol-related damage to the stomach lining by 78% in rats after 32 days of daily dosing.
• At the highest dose it protected about 1.5 times better than a commercial fish collagen, and a small peptide released during digestion appears to switch on the stomach’s own protective prostaglandin pathway.
• The work is early, so human studies are needed before it has any bearing on products. Access the study: https://doi.org/10.1002/advs.77964
Lan, X., Dai, Q., Wang, Y., Gu, J., Wang, C., Xu, W., Gao, X., Cheng, J., Geng, W., Wu, X., Li, H., He, X., Fu, S., Zhu, L., & Xu, W. (2026). Oral recombinant human collagen III protects gastric mucosa via G-quadruplex-mediated activation of the Pla2g4a-prostaglandin axis. Advanced Science, e77964. https://doi.org/10.1002/advs.77964