Collagen Side Effects: What the Trials Recorded, the Kidney-Stone Mechanism, and an FDA Signal That Turns Out to Be Shampoo

Collagen peptides are one of the most widely used supplements on a gym shelf, and the usual answer to "what are the side effects?" is "none". The randomised trials mostly support that, but they were rarely designed to find side effects, and that matters. This page sets out what the trials recorded and the one mechanism with laboratory evidence behind it. It also covers the allergy question by source animal, and what FDA's adverse-event file holds once its largest signal is read properly. It gives no usage guidance, per our editorial standards. How much the trials used is on how much collagen per day.

What the trials recorded

Two caveats keep "no adverse events" from meaning "proven harmless". First, size: a trial of 50 people for 12 weeks can miss anything rarer than a few per cent. Second, method: a trial that does not collect side effects with a standard checklist will report "none" by default. The 2026 review's finding that most of its 25 trials carry a high risk of bias is the honest frame. The trials found no common, serious harm at 2.5 to 15 g a day, and nothing more than that.

The mechanism worth knowing: hydroxyproline and oxalate

Collagen is unusually rich in hydroxyproline, an amino acid found almost nowhere else in the diet. The body breaks hydroxyproline down to glyoxylate, which can become glycolate or oxalate. Oxalate is the main component of the most common kind of kidney stone.

Knight 2006 tested this directly on controlled oxalate diets (PMID 17021603):

• 30 g of gelatin versus 30 g of whey protein: urinary glycolate 5.3-fold higher, urinary oxalate 43% higher on gelatin.
• Single loads: 5 g and 10 g of gelatin produced significant changes in plasma hydroxyproline and urinary oxalate and glycolate; 1 g and 2 g did not.
• The authors estimated that hydroxyproline turnover accounts for 5–20% of the oxalate the body makes itself.

Rodgers 2019 ran a smaller test (PMID 31161522): five healthy men, 30 g of gelatin a day for three days. Glycolate and urine pH rose, but oxalate excretion did not change, and the calculated calcium oxalate supersaturation did not either.

So the pathway is real and measurable at gelatin loads in the range of collagen label servings, and its size in practice is unsettled. No trial found for this page has counted kidney stones in collagen users. The group to whom this applies is people with a history of calcium oxalate stones or a hyperoxaluria diagnosis, and for them the question belongs with their nephrologist or urologist. The parallel question for creatine is answered on creatine and kidneys.

Allergy, by source animal

Collagen is the main protein in gelatin, and gelatin allergy is documented: bovine type I collagen was identified as the major allergen behind gelatin reactions to some vaccines. For fish:

• Sakaguchi 2000 (PMID 10984381) found IgE antibodies to fish gelatin in 3 of 10 children with fish allergy, in both children allergic to fish and bovine gelatin, and in 5 of 15 children with atopic dermatitis sensitised to fish. Different fish gelatins cross-reacted with each other; fish and bovine gelatin showed little cross-reactivity.
• US labelling: fish is a major food allergen, and an ingredient that contains protein derived from fish falls under the same definition (21 U.S.C. 321(qq)). Marine collagen sold in the US must say so. Bovine and porcine collagen carry no comparable mandatory allergen statement.

The practical reading: the source animal printed on the label is the allergy information. A label reading only "collagen peptides" with no source stated leaves a fish-allergic or beef-allergic reader without what they need. Whether any particular product is tolerated is a question for an allergist.

FDA's adverse-event file, read properly

FDA's CAERS database collects voluntary reports of problems with foods, supplements and cosmetics. Queried through openFDA on 2026-10-02 (data last updated 2026-07-07), it holds 464 reports in which a product with "collagen" in its name appears, whether as the suspected product or as one taken alongside it.

The top line is not what it looks like. Breaking the 53 hair-loss reports down by product name, at least 13 name an OGX biotin-and-collagen shampoo (under three spellings), a cosmetic applied to the scalp. Other products named in the same reports include biotin, vitamin D and multivitamins. Read without that breakdown, the file seems to say collagen supplements cause hair loss, which no trial or report pattern supports.

Three rules for the rest of the table:

• A report is not a cause. Reports are unverified, often name several products, and do not establish that any one caused the reaction.
• Counts are not rates. Nobody knows how many people took collagen, so 39 nausea reports cannot be turned into a percentage.
• The serious-outcome columns overlap. 109 reports are marked hospitalisation and 38 life-threatening. Those labels come from the reporter and include reports where collagen was a concomitant product.

What the file does show is consistent with the mechanism and allergy sections above. Allergic-type reactions (hypersensitivity, itching, rash, hives) and digestive complaints dominate, and kidney stones appear but rarely. Reports by year peaked in 2021 (74) and have fallen since (31 in 2025).

Digestive tolerance and dose

Digestive complaints are the most common category in the reports, and they are the usual complaint with any large protein serving. Collagen labels range up to about 20 g a serving, as counted on how much collagen per day, while the trials used 2.5 to 15 g. Collagen also scores zero on protein quality because it lacks tryptophan, a point made on collagen vs whey protein. That page explains why collagen is not a replacement for a complete protein, which is a nutrition question rather than a side effect.

What is not established

• Kidney stones in users: no trial or cohort has counted them; the evidence is an oxalate mechanism in short feeding studies.
• Pregnancy and breastfeeding: a PubMed search on 2026-10-02 combining collagen peptides or hydrolysate with pregnancy, lactation or breastfeeding, limited to trials, returned 17 records, and none tests a collagen supplement in pregnant or breastfeeding women (they are bone-marker, calcium and infant-formula studies). That is an absence, not evidence either way, and the decision sits with the reader's clinician.
• Long-term use: the longest trials cited on this site ran about 12 months.
• Contaminants: heavy-metal content varies by source and manufacturer and was not measured for this page.
• Drug interactions: none documented in the sources read; that reflects how little has been studied.

The honest summary

• Trials: few or no adverse events recorded at 2.5–15 g/day, in mostly small, short, often high-bias studies.
• The one mechanism with data: hydroxyproline → oxalate, seen at 5–30 g gelatin loads, inconsistent across two studies, no stone counts.
• Allergy follows the source animal; fish collagen must declare fish in the US.
• FDA's file: 464 reports naming a collagen product; its top signal, hair loss, is partly shampoo.

Limits of this page

The openFDA query matched product names containing "collagen" and cannot separate supplements from cosmetics or suspect from concomitant products without reading each report; the shampoo count is a floor from the product-name breakdown, not a full audit. Systematic-review conclusions are reported as their authors state them. The oxalate studies used gelatin, which has the same amino-acid composition as collagen peptides but is not identical in processing.

Sources and dates

• Choi FD, Sung CT, Juhasz ML, Mesinkovsk NA. Oral collagen supplementation: a systematic review of dermatological applications. J Drugs Dermatol 2019;18:9–16 — PMID 30681787
• Bassila C, Bassila JC, Slim M. Efficacy and safety of hydrolyzed collagen supplementation on skin health outcomes: a systematic literature review of randomized controlled trials. Eur J Clin Nutr 2026;80:867–85 — PMID 42342959
• Knight J, Jiang J, Assimos DG, Holmes RP. Hydroxyproline ingestion and urinary oxalate and glycolate excretion. Kidney Int 2006;70:1929–34 — PMID 17021603
• Rodgers A, Cele P, Ravenscroft N, et al. Theoretical and laboratory investigations of the effects of hydroxyproline ingestion on the metabolic and physicochemical risk factors for calcium oxalate kidney stone formation in a small group of healthy subjects. Int Urol Nephrol 2019;51:1121–7 — PMID 31161522
• Sakaguchi M, Toda M, Ebihara T, et al. IgE antibody to fish gelatin (type I collagen) in patients with fish allergy. J Allergy Clin Immunol 2000;106:579–84 — PMID 10984381
• 21 U.S.C. 321(qq), definition of "major food allergen", read on law.cornell.edu 2026-10-02.
• PubMed, searched 2026-10-02: (collagen peptide* OR collagen hydrolysate OR hydrolyzed collagen) AND (pregnan* OR lactation OR breastfeeding), trial publication types → 17 records, all titles read, 0 collagen-supplement trials in pregnancy or lactation.
• FDA CAERS via openFDA food/event, queried 2026-10-02 (last updated 2026-07-07): products.name_brand:collagen → 464 reports; counts by reaction, outcome, product name and year.