Supplements for Bones, Joints and Connective Tissue
Glucosamine failed its large trials, calcium supplements carry a cardiovascular question, and the compound with the most interesting recent data needs vitamin C and a loading stimulus to do anything.
The Short Answer
This category contains one of the clearest failures in supplement science and one of the more interesting recent findings. Glucosamine and chondroitin were tested in large well-designed trials and did not work for most patients. Collagen peptides, taken with vitamin C and combined with loading, have a plausible mechanism and a growing body of small positive trials. Between them sits calcium, where supplementation carries a cardiovascular question that dietary calcium does not.
The Glucosamine and Chondroitin Story
These were among the most widely used supplements in the world, with a mechanistic rationale based on providing cartilage building blocks.
The large trials were mostly negative. A major multi-centre trial found no significant benefit over placebo for knee osteoarthritis pain in the overall population, with a possible signal in a subgroup with moderate to severe pain that did not survive as a primary finding. Subsequent meta-analyses have generally found effects small enough to be clinically unimportant, and industry-funded studies report larger effects than independent ones.
Major osteoarthritis guidelines now recommend against their routine use, with some regional variation.
Two caveats keep this honest. Glucosamine sulphate and glucosamine hydrochloride are different salts and some argue the sulphate form has better data, and observational analyses have reported associations between glucosamine use and lower cardiovascular and all-cause mortality, which is intriguing, confounded and not a reason to take it for joints.
The general lesson is worth more than the specific compound: a compelling mechanism, wide use and enormous sales are not evidence, and this category was tested properly and found wanting.
Collagen Peptides: The Interesting Case
Collagen is the principal structural protein in tendon, ligament, bone and skin. The intuitive objection, that eating collagen simply supplies amino acids like any protein, is reasonable and possibly incomplete.
The proposed mechanism involves specific di- and tripeptides, particularly prolyl-hydroxyproline, which survive digestion, appear in plasma and may act as signals to fibroblasts rather than only as substrate. That is a mechanistically distinct claim and it has some support.
The evidence. Small trials report improvements in joint pain in athletes and in people with activity-related knee pain, improvements in tendon properties when combined with loading, and effects on skin elasticity and hydration in a reasonably consistent set of trials. Bone density trial data in postmenopausal women exist and are limited.
The important qualifier. Several tendon trials combined collagen with a loading protocol, and the loading is doing substantial work. Collagen without loading is unlikely to reproduce those results, and the practical protocol involves taking it around an hour before loading, with vitamin C, which is required as a cofactor for collagen synthesis.
Dose. Trials generally use 10 to 15 g of hydrolysed collagen peptides for joint and skin outcomes, and around 15 g with vitamin C for tendon work.
Bone: Calcium, Vitamin D and K2
| Nutrient | Position |
|---|---|
| Dietary calcium | Preferred route; food sources not associated with the cardiovascular signal |
| Calcium supplements | Some analyses suggest increased cardiovascular risk; reserve for genuine intake shortfall |
| Vitamin D | Required for calcium absorption; correction where status is low is well established. High intermittent doses have been associated with increased falls in some trials |
| Vitamin K2 | Activates matrix Gla protein and osteocalcin; some trial evidence on bone and vascular calcification markers |
| Magnesium | Required for bone matrix and vitamin D metabolism; intake commonly low |
| Protein | Higher intake associated with better bone density; the old concern about protein harming bone has not held up |
| Boron, silicon, strontium | Thin evidence; strontium supplements are not the prescription form and carry cardiovascular concerns |
The calcium position deserves care. Adequate calcium intake matters for bone, and the route matters: meta-analyses have reported a cardiovascular signal with supplements that is not seen with dietary intake, and the mechanism proposed involves acute rises in serum calcium after a supplement dose that food does not produce. The reasonable position is food first, supplement only to close a genuine gap, and split doses if supplementing.
The protein point is worth stating because the opposite was believed for years. Higher protein intake is associated with better bone density, and the acid-load hypothesis suggesting protein leaches calcium from bone has not been supported.
The Rest of the Joint Aisle
Curcumin, bioavailable forms. Trials in knee osteoarthritis report symptomatic benefit, with some studies comparing favourably to conventional options. Among the better-supported options in this category, and formulation-dependent.
Boswellia. Reasonable osteoarthritis symptom data through 5-lipoxygenase inhibition. Standardisation varies.
Omega-3. Better evidence in inflammatory arthritis than in osteoarthritis, and reasonable as general anti-inflammatory support.
MSM. Some small positive trials in osteoarthritis symptoms, modest effects.
Undenatured type II collagen. A different mechanism from collagen peptides, proposed to act through oral tolerance. Small trials, early.
Hyaluronic acid, oral. Limited evidence; the injectable form is a separate clinical question with its own mixed evidence.
Creatine. Not marketed for this Pillar and belongs in it. It supports strength and lean mass gains alongside resistance training, and has trial data in older adults, which makes it more relevant to Pillar 7 outcomes than most joint supplements.
Vitamin C. Required for collagen synthesis, so relevant as a cofactor rather than as an intervention.
What Outperforms All of It
Resistance training. Builds muscle and stimulates bone, with effects in people in their eighties and nineties. No supplement approaches it.
Progressive tendon loading. The best-evidenced intervention in tendinopathy management by a wide margin, and it is what the collagen trials were built around.
Impact loading for bone. Bone responds to strain rate and magnitude. Jumping and hopping protocols have trial evidence for bone density in postmenopausal women.
Adequate protein, distributed across meals. The nutritional intervention that matters most.
Weight management where joints are symptomatic. Improves knee symptoms through both mechanical and inflammatory routes.
Exercise therapy for osteoarthritis. First-line in every guideline, and consistently more effective than patients expect.
The pattern in this Pillar is unusually stark. The interventions that work require effort and cannot be sold, and the products that can be sold have mostly failed their trials.
A Defensible Stack
Core: adequate protein at 1.2 to 1.6 g per kg for older adults, distributed at 25 to 40 g per meal. Vitamin D where status is low. Magnesium where intake is low. Calcium from food, supplemented only to close a genuine gap.
Well evidenced and cheap: creatine 3 to 5 g daily, for its effect on strength and lean mass alongside training.
Reasonable for a specific complaint: collagen peptides 10 to 15 g with vitamin C, taken around an hour before loading, if tendon or joint symptoms are the target and a loading protocol is in place. Bioavailable curcumin or boswellia for osteoarthritis symptoms.
Consider: vitamin K2 alongside vitamin D.
Skip: glucosamine and chondroitin, which were tested and failed; strontium supplements; and proprietary joint blends combining several of the above at undisclosed doses.
Escalate clinically for: a fragility fracture at any age, which warrants bone density assessment; joint swelling, morning stiffness over an hour or symmetrical small joint involvement, which suggests inflammatory arthritis; unexplained bone pain; or any fracture from a low-impact fall. Osteoporosis responds well to available care, and it is under-identified and under-addressed, particularly in men.
The AEONNN Perspective
Pillar 7 is where AEONNN's Evidence layer records the most decisive negative finding in consumer supplementation. Glucosamine and chondroitin were tested in large well-designed trials, effects were small enough to be clinically unimportant, and major guidelines recommend against routine use. The platform does not recommend them.
Collagen peptides are the interesting counterpart and carry an important qualifier the marketing omits: several of the positive tendon trials combined collagen with a loading protocol, and the loading is doing substantial work. AEONNN recommends it as an adjunct to loading with vitamin C as a cofactor, never as a substitute for the loading.
The Safety layer holds the calcium position, where route matters: a cardiovascular signal appears with supplements and not with dietary intake, so the platform prefers food and supplements only to close a genuine gap. And creatine belongs in this Pillar even though it is marketed elsewhere, because its effect on strength and lean mass alongside training is more relevant to Pillar 7 outcomes than any joint product.
Pillar Matrix mapping
Database Matrix layers
- Evidence Layer (PubMed, Cochrane, ClinicalTrials.gov)
- Safety Layer (DrugBank, FAERS)
- Mechanistic Layer (KEGG, Reactome, UniProt)
- Quality / Formulation Layer (ConsumerLab, Labdoor)
Frequently Asked
Does glucosamine work for joints?
Large well-designed trials were largely negative, meta-analyses find effects too small to be clinically important, and major osteoarthritis guidelines recommend against routine use.
Do collagen supplements work?
Small trials report benefits for joint pain, tendon properties and skin measures. Several tendon trials combined collagen with a loading protocol, and the loading is doing substantial work.
How should collagen be taken?
Trials generally use 10 to 15 g of hydrolysed peptides with vitamin C, taken around an hour before a loading session. Without loading it is unlikely to reproduce the trial results.
Are calcium supplements safe?
Some meta-analyses report a cardiovascular signal with supplements that is not seen with dietary calcium. Food first, supplement only to close a genuine gap, and split doses if supplementing.
Does protein harm bone?
No. Higher protein intake is associated with better bone density, and the older acid-load hypothesis suggesting protein leaches calcium from bone has not been supported.
What is the best supplement for this Pillar?
Adequate protein, then creatine at 3 to 5 g daily for its effect on strength and lean mass alongside training. Vitamin D where status is low.
When should bone health be assessed clinically?
After any fragility fracture at any age, after a fracture from a low-impact fall, or with unexplained bone pain. Osteoporosis responds well to available therapy and is under-identified, particularly in men.
Evidence and review
Any dosage ranges cited here reflect the ranges used in published human trials, not personal recommendations. Evidence in this field moves, so this article is reviewed quarterly and carries its last-updated date above. Nothing here is intended as medical advice, and supplementation should be discussed with a qualified clinician, particularly alongside prescribed medication or an existing condition.