asxan.ai Agents Hub

Joint & Bone Health

Evidence Stack

Bone mineralization · cartilage matrix · joint comfort

Evidence-first joint-and-bone support stack — what the human-evidence record actually shows for the ingredients most associated with lifelong mobility, including the baseline-status-dependent vitamin D story, the collagen joint-comfort signals, and the honest nulls (GAIT glucosamine-chondroitin, VITAL fractures in replete adults, MSM). This is mechanism and evidence mapping, not medical advice. Diagnosed musculoskeletal conditions (osteoarthritis, rheumatoid arthritis, osteoporosis) require physician-led management; discuss any supplementation with an orthopedist, rheumatologist, or endocrinologist. All PubMed identifiers are verified against PubMed before inclusion; cross-market regulatory claims appear verbatim per their authorising authority (FDA · EFSA · ANVISA · TGA).

Last reviewed · How we assess evidence →

Quick Summary

  • Vitamin D3 fracture reduction depends critically on baseline status — replete vs deficient. LeBoff 2022 VITAL fracture sub-study (PMID 35939577) NEJM in 25,871 adults reported vitamin D3 2,000 IU/day produced null fracture reduction in already-vitamin-D-replete generally healthy adults over a median 5.3 years. Kong 2022 (PMID 35504603) Osteoporos Int daily-dose meta-analysis, by contrast, supports D3 + calcium fracture reduction in deficient populations. The honest framing: target the deficient, do NOT empirically supplement the already-replete. See /ingredients/vitamin-d3/.
  • Vitamin K2 (MK-7) has 3-year RCT signal for postmenopausal BMD. Knapen 2013 (PMID 23525894) Osteoporos Int 3-year RCT in postmenopausal women supported lumbar spine and femoral neck BMD at 180 µg MK-7 per day. MK-7 routes calcium into bone matrix via osteocalcin γ-carboxylation — mechanism complement to D3 + calcium.
  • Collagen peptide signal for knee OA pain is robust but evidence quality is moderate. Lin 2023 (PMID 37717022) J Orthop Surg Res meta-analysis of RCTs reported pain reduction with standardized mean difference -0.58 (95% CI -0.98 to -0.18; p = 0.004) for collagen peptide vs placebo in knee OA. Safety profile comparable to placebo. Source-species, molecular-weight, and hydrolysis-method heterogeneity remain dominant interpretive caveats.
  • Undenatured type II collagen (UC-II) showed 180-day WOMAC improvement vs both placebo AND glucosamine-chondroitin. Lugo 2016 (PMID 26822714) Nutr J multicenter randomized double-blind placebo-controlled study in 191 OA patients reported significant WOMAC pain, stiffness, and physical function improvement at day 180 (p = 0.0003 vs placebo, p ≤ 0.044 vs glucosamine-chondroitin). Mechanism: oral immune-tolerance pathway (Peyer's patches), distinct from hydrolyzed collagen substrate provision.
  • Glucosamine + chondroitin and MSM are null/negative at the meta-analytic level. Clegg 2006 GAIT (PMID 16495392) NEJM primary intention-to-treat analysis at 24 weeks reported NO significant difference between glucosamine + chondroitin and placebo on OA pain; Wandel 2010 (PMID 20847017) BMJ network meta-analysis pooled the negative signal. Brien 2011 (PMID 19474240) concluded DMSO and MSM are "not clinically effective" for OA pain. These appear here as honest-negative context, not recommendations.
  • This is not medical advice. The vitamin D3, K2, collagen, and UC-II signals here are for daily resilience and joint-comfort in generally healthy or pre-disease adults — they are NOT evidence-supported treatments for diagnosed OA, RA, or osteoporosis. Anyone with a diagnosed condition should follow a specialist's treatment plan.

The Evidence Stack

The "evidence" column below describes the strength and direction of the outcome evidence in qualitative terms — well-established, robust, moderate/mixed, preliminary/emerging, or null/negative. The S/A/B/C tier that grades how extensively an ingredient is studied (its evidence volume) lives on each linked ingredient page, not here.

Ingredient Bone / joint evidence (qualitative) Key Trial / Meta-analysis asxan.ai page
Vitamin D3 Robust for fracture reduction in deficient adults; null/negative in already-replete adults (VITAL fracture RCT) LeBoff 2022 VITAL PMID 35939577 (NEJM · 25,871 adults · vitamin D 2,000 IU/d · null fracture in replete · 5.3 y); Kong 2022 PMID 35504603 (Osteoporos Int · daily D + Ca fracture meta in deficient) /ingredients/vitamin-d3/
Vitamin K2 (MK-7) Moderate / emerging — 3-year postmenopausal BMD RCT signal Knapen 2013 PMID 23525894 (Osteoporos Int · 3-y RCT · 180 µg MK-7/d · lumbar spine + femoral neck BMD) /ingredients/vitamin-k2/
Collagen Peptides (hydrolyzed) Robust signal for knee OA pain; evidence quality moderate (heterogeneity caveats) Lin 2023 PMID 37717022 (J Orthop Surg Res · meta · SMD -0.58 / 95% CI -0.98 to -0.18); de Miranda 2021 PMID 33742704 (skin parameters meta · cross-context substrate-provision rationale) /ingredients/collagen-peptides/
Type II Collagen (UC-II) Moderate / robust — 180-day multicenter RCT vs placebo AND glucosamine-chondroitin Lugo 2016 PMID 26822714 (Nutr J · 180-day multicenter DBPC · 191 OA · WOMAC pain p=0.0003 vs placebo · oral immune-tolerance mechanism) /ingredients/collagen-type-ii/
Glucosamine + Chondroitin Null / negative — primary ITT null + network meta null Clegg 2006 GAIT PMID 16495392 (NEJM · primary ITT null in OA pain · 24 wk); Wandel 2010 PMID 20847017 (BMJ network meta · null pooled) /ingredients/glucosamine-chondroitin/

How It Works

Each ingredient engages bone or cartilage biology by a different route — vitamin D3 through intestinal calcium-phosphate absorption, vitamin K2 through osteocalcin calcium-routing, hydrolyzed collagen through di-/tri-peptide substrate provision, and UC-II through an oral immune-tolerance pathway. Glucosamine + chondroitin's substrate-precursor logic is plausible but does not translate into a positive primary endpoint.

Vitamin D3 — calcium-phosphate absorption and the LeBoff baseline-status story. Vitamin D3 (cholecalciferol) hydroxylated to 25-OH-D and then 1,25(OH)2D regulates intestinal calcium / phosphate absorption and supports bone mineralization. LeBoff 2022 VITAL fracture sub-study (PMID 35939577) NEJM enrolled 25,871 generally healthy adults aged 50+ and reported vitamin D3 2,000 IU/day produced null effect on total, non-vertebral, or hip fractures over a median 5.3 years — in this already-replete cohort. Kong 2022 (PMID 35504603) Osteoporos Int daily-dose meta-analysis, in contrast, supports D3 + calcium fracture reduction in deficient and vulnerable populations. The right framing is baseline-status-driven: target the deficient (serum 25-OH-D < 20 ng/mL, frail elderly, institutionalized, limited sun exposure), do NOT empirically supplement the already-replete (serum 25-OH-D ≥ 30 ng/mL) without clinical rationale. See /ingredients/vitamin-d3/.

Vitamin K2 (MK-7) — osteocalcin γ-carboxylation and the calcium-routing story. Menaquinone-7 (MK-7) is the cofactor for γ-glutamyl carboxylase, which γ-carboxylates osteocalcin (bone-Gla protein) and matrix-Gla protein. Carboxylated osteocalcin binds calcium hydroxyapatite in bone matrix; undercarboxylated osteocalcin (ucOC) is a functional K2 deficiency marker. Knapen 2013 (PMID 23525894) Osteoporos Int 3-year RCT in postmenopausal women supported lumbar spine and femoral neck BMD preservation at 180 µg MK-7 per day. The mechanism complements (does NOT replace) D3 + calcium; the framing is calcium-routing into bone matrix rather than absorption.

Hydrolyzed collagen peptides — di-/tri-peptide substrate provision and the knee OA pain signal. Hydrolyzed collagen supplies small hydroxyproline-containing peptides (Pro-Hyp, Gly-Pro-Hyp) that are measurable in plasma after oral dosing and reach connective tissue (cartilage and dermis). Lin 2023 (PMID 37717022) J Orthop Surg Res meta-analysis of RCTs reported knee OA pain reduction at standardized mean difference -0.58 (95% CI -0.98 to -0.18; p = 0.004; moderate-quality evidence). de Miranda 2021 (PMID 33742704) skin parameters meta extends the substrate-provision rationale into dermal matrix context. Source species (bovine / porcine / marine), molecular-weight distribution, and hydrolysis method materially affect outcome — cross-product extrapolation is NOT warranted.

Undenatured type II collagen (UC-II) — oral immune-tolerance pathway and the Lugo 180-day RCT. UC-II is undenatured (intact triple-helix structure) type II collagen at ~40 mg/day; the proposed mechanism involves oral immune-tolerance induction via Peyer's patch presentation, suppressing autoimmune-like cartilage destruction. Lugo 2016 (PMID 26822714) Nutr J multicenter randomized double-blind placebo-controlled study in 191 OA patients over 180 days reported significant WOMAC pain (p = 0.0003 vs placebo), stiffness (p = 0.004 vs placebo), and physical function (p = 0.007 vs placebo) improvement; UC-II also outperformed glucosamine + chondroitin (p ≤ 0.044 across WOMAC subscales). Mechanism is distinct from hydrolyzed collagen — UC-II works at low milligram doses via immune-tolerance, not gram-doses for substrate provision.

Glucosamine + chondroitin and the GAIT primary null. Glucosamine sulfate / hydrochloride and chondroitin sulfate are cartilage matrix substrate / glycosaminoglycan precursors. The NIH-funded GAIT trial (Clegg 2006 PMID 16495392) NEJM — the largest single OA RCT — reported the primary intention-to-treat (ITT) analysis at 24 weeks was null: no significant difference between glucosamine + chondroitin and placebo on the primary OA pain endpoint. Wandel 2010 (PMID 20847017) BMJ network meta-analysis pooled the negative signal. Subgroup analyses in the moderate-severe pain subgroup of GAIT trended positive, but the primary ITT result is the rigorous reading. Glucosamine + chondroitin is included on this page as honest-negative context — not a recommendation.

Body systems engaged: Musculoskeletal. Mechanism tags: Mineral metabolism · Collagen synthesis · Immune tolerance modulation.

What the Trials Show — Including the Nulls

This page does NOT support claims of curing, treating, or diagnosing osteoarthritis, rheumatoid arthritis, osteoporosis, or any other diagnosed musculoskeletal disease. The vitamin D3, K2, collagen peptide, and UC-II signals here are for daily resilience and joint-comfort support in generally healthy or pre-disease adults — they are NOT evidence-supported as treatments for diagnosed OA, RA, or osteoporosis. Anyone with a diagnosed musculoskeletal condition should follow an orthopedist / rheumatologist / endocrinologist's treatment plan.

Vitamin K2 is contraindicated for anyone on warfarin or other vitamin-K antagonist anticoagulants. MK-7 directly antagonizes warfarin's mechanism. Anyone on warfarin should NOT supplement MK-7 without explicit prescriber direction and INR monitoring. DOACs (apixaban, rivaroxaban) are not vitamin-K-antagonist class and the interaction is mechanism-irrelevant — but clinical consultation is still warranted.

Collagen peptide products are NOT interchangeable. Lin 2023 (PMID 37717022) and de Miranda 2021 (PMID 33742704) meta-analyses pool RCTs using heterogeneous collagen peptide preparations differing in molecular weight, source species (bovine / porcine / marine / chicken), hydrolysis method, and peptide profile. Cross-product extrapolation from one commercial RCT signal to another differently-prepared product is methodologically unsound.

Anti-resorptive prescription therapy is the foundational pharmacology for diagnosed osteoporosis — supplements layer on top, NOT substitute. Bisphosphonates, denosumab, anabolic agents (teriparatide, romosozumab), and (in select cohorts) hormone replacement therapy are the foundational pharmacologic pillars for diagnosed osteoporosis. Vitamin D3, calcium, and MK-7 layer on top — they do NOT substitute for indicated antiresorptive therapy.

Falls prevention, strength training, and adequate dietary protein are the largest fracture-risk levers. ≥1.6 g/kg/day protein (Morton 2018 PMID 28698222 protein meta-regression — see Athletic Performance), structured resistance training, balance / falls-prevention programs, and home-hazard modification are the most evidence-anchored fracture-prevention levers — particularly in older adults. Supplements are layered context.

Stacking & Timeline

Mechanistic pairings are plausible but rarely backed by head-to-head synergy trials; realistic timelines run from weeks (collagen joint-comfort) to years (BMD preservation and the multi-year fracture-trial horizon).

Mechanistic pairs

Vitamin D3 + Calcium + MK-7 · the bone-mineralization triad. D3 + calcium (per Kong 2022 PMID 35504603 in deficient populations) anchor absorption and substrate; MK-7 (Knapen 2013 PMID 23525894) routes calcium into bone matrix via osteocalcin γ-carboxylation. The three mechanisms are non-redundant. Honest caveat: in already-replete adults the marginal benefit of additional D3 is null (LeBoff 2022 PMID 35939577).

Collagen Peptides + Resistance Training · the joint-cartilage pair. Lin 2023 (PMID 37717022) collagen peptide meta + structured resistance training (and adequate dietary protein per Morton 2018 PMID 28698222) layer together for joint-comfort and musculoskeletal aging. Direct collagen + RT head-to-head RCTs are limited but mechanism rationale supports layering.

UC-II vs Glucosamine-Chondroitin · the OA pain head-to-head. Lugo 2016 (PMID 26822714) Nutr J directly compared UC-II vs glucosamine + chondroitin in 191 OA patients over 180 days; UC-II outperformed both placebo AND glucosamine + chondroitin across WOMAC pain, stiffness, and physical function. This is one of the most directly informative head-to-head trials in the joint-supplement space.

Whole-Foundation Mobility Stack · sleep, strength, protein, falls prevention. Bauer 2015 PROVIDE (PMID 26170041) sarcopenic 65+ whey + leucine + D3 RCT anchors the protein-leucine-D3 sarcopenia prevention rationale. Layering bone + joint supplements on top of structured strength training, ≥1.6 g/kg/day protein, balance work, and home-hazard mitigation is the actual evidence-anchored mobility stack.

When to see results — realistic timeframes

12 weeks · collagen peptide knee OA pain reduction window. Lin 2023 (PMID 37717022) pooled trials predominantly 12-week durations for meaningful OA pain reduction.

180 days (6 months) · UC-II WOMAC improvement window. Lugo 2016 (PMID 26822714) reported significant WOMAC improvements at day 180 — the trial-anchored time horizon. Earlier interim results were less robust.

3 years · MK-7 BMD preservation window. Knapen 2013 (PMID 23525894) is a 3-year RCT — BMD changes are slow and require sustained intervention. Short-term BMD measurement on supplements is NOT informative.

5.3 years · VITAL fracture trial null window. LeBoff 2022 (PMID 35939577) median follow-up was 5.3 years; the null result on fractures in already-replete adults represents a sustained, large-cohort, long-duration negative signal — NOT a short-trial artifact.

Lifetime · cumulative bone density accrual through age ~30, then preservation. Peak bone mass is reached around age 25–30 in most adults. Calcium intake, vitamin D status, weight-bearing activity, body weight, and avoidance of smoking / excessive alcohol across decades determine peak bone mass; thereafter preservation of bone mass is the goal. Supplements in middle and older adulthood are layered on the lifetime substrate, NOT a substitute.

The shared anchor across these pages is the bone-mineralization and musculoskeletal-preservation logic — vitamin D status, MK-7 calcium-routing, and protein + resistance training reappear in each.

Frequently Asked Questions

1. Should I take vitamin D for my bones?

It depends on baseline status. LeBoff 2022 VITAL fracture sub-study (PMID 35939577) showed that in 25,871 already-vitamin-D-replete generally healthy adults, vitamin D3 2,000 IU/day for a median 5.3 years produced NO reduction in fractures. Kong 2022 (PMID 35504603) meta-analysis, by contrast, supports D + Ca fracture reduction in deficient populations. The right framework is target the deficient (serum 25-OH-D < 20 ng/mL, frail elderly, institutionalized, limited sun exposure) — do NOT empirically supplement the already-replete (serum 25-OH-D ≥ 30 ng/mL).

2. Does glucosamine + chondroitin actually work for knee pain?

The rigorous meta-analytic answer is NO. The NIH-funded GAIT trial (Clegg 2006 PMID 16495392) NEJM — the largest single OA RCT — reported the primary intention-to-treat analysis at 24 weeks showed no significant difference vs placebo on OA pain. Wandel 2010 (PMID 20847017) BMJ network meta-analysis pooled the negative signal. Some moderate-severe pain subgroups in GAIT trended positive — but the primary ITT remains negative. If you find personal benefit from glucosamine + chondroitin, that is fine; but the meta-analytic literature does NOT robustly support it.

3. Is UC-II better than glucosamine-chondroitin for OA?

In Lugo 2016 (PMID 26822714) Nutr J multicenter randomized double-blind placebo-controlled study of 191 OA patients over 180 days, UC-II outperformed BOTH placebo (WOMAC pain p = 0.0003) AND glucosamine + chondroitin (WOMAC pain p = 0.016, stiffness p = 0.044, physical function comparable). The mechanism difference is also informative: UC-II works via oral immune-tolerance (Peyer's patches, low milligram dose around 40 mg/day) rather than substrate provision (glucosamine + chondroitin at gram doses). For OA pain, UC-II has stronger direct head-to-head evidence than glucosamine + chondroitin.

4. Do collagen peptides really help joint pain?

Lin 2023 (PMID 37717022) J Orthop Surg Res meta-analysis of randomized controlled trials reported knee OA pain reduction at standardized mean difference -0.58 (95% CI -0.98 to -0.18; p = 0.004; moderate-quality evidence) for collagen peptide vs placebo. Safety profile was comparable to placebo. The signal is reproducible but evidence quality is moderate — not high. Source species (bovine / porcine / marine / chicken), molecular weight, and hydrolysis method differ across products, so cross-product extrapolation is NOT warranted. Match the trial product when possible.

5. Should I take vitamin K2 with vitamin D?

Mechanism rationale supports the pairing — D3 + calcium drive absorption and substrate availability, MK-7 routes calcium into bone matrix via osteocalcin γ-carboxylation. Knapen 2013 (PMID 23525894) supported lumbar spine BMD at 180 µg MK-7 per day over 3 years in postmenopausal women. CRITICAL safety: vitamin K2 is contraindicated for anyone on warfarin or other vitamin-K antagonist anticoagulants — MK-7 directly antagonizes warfarin's mechanism. Talk to your clinician.

6. Should I take MSM for joint pain?

Brien 2011 (PMID 19474240) meta-analysis explicitly concluded "current evidence suggests DMSO and MSM are not clinically effective in the reduction of pain in the treatment of OA." Pooled pain reduction was "neither statistically nor clinically significant." MSM is not included on this page as an anchor.

References

All PMIDs verified against PubMed. Effect sizes are reported as published.

  1. PMID 35939577 · LeBoff 2022 · VITAL fracture sub-study · NEJM · 25,871 adults · vitamin D3 2,000 IU/d · null total / non-vertebral / hip fracture in already-replete adults · median 5.3 y
  2. PMID 35504603 · Kong 2022 · Osteoporos Int · daily-dose D + calcium fracture meta-analysis · supports fracture reduction in deficient populations
  3. PMID 23525894 · Knapen 2013 · Osteoporos Int · 3-year RCT · 180 µg MK-7/d · lumbar spine + femoral neck BMD preservation in postmenopausal women
  4. PMID 37717022 · Lin 2023 · J Orthop Surg Res · collagen peptide knee OA pain meta-analysis · SMD -0.58 (95% CI -0.98 to -0.18; p = 0.004; moderate-quality)
  5. PMID 26822714 · Lugo 2016 · Nutr J · UC-II 180-day multicenter DBPC RCT (n=191) · WOMAC pain p=0.0003 vs placebo · outperformed glucosamine-chondroitin
  6. PMID 33742704 · de Miranda 2021 · collagen peptide skin-parameters meta-analysis (cross-context substrate-provision rationale)
  7. PMID 16495392 · Clegg 2006 · GAIT · NEJM · primary ITT analysis null · no significant glucosamine + chondroitin vs placebo OA pain difference at 24 wk
  8. PMID 20847017 · Wandel 2010 · BMJ · glucosamine / chondroitin network meta-analysis · null pooled
  9. PMID 19474240 · Brien 2011 · Evid Based Complement Alternat Med · MSM / DMSO OA pain meta-analysis · "not clinically effective" (honest-negative anchor)
  10. PMID 28698222 · Morton 2018 · protein supplementation + resistance training meta-regression (cross-link context · falls / musculoskeletal preservation)
  11. PMID 26170041 · Bauer 2015 · PROVIDE · J Am Med Dir Assoc · sarcopenic 65+ whey + leucine + D3 RCT (cross-link context)

Coverage Notes

This Joint & Bone page draws from five linked ingredient pages on asxan.ai (vitamin D3, vitamin K2 / MK-7, collagen peptides, type II collagen / UC-II, glucosamine + chondroitin). The evidence stack above describes each ingredient's bone / joint outcome evidence qualitatively rather than asserting a volume tier. Glucosamine + chondroitin and MSM appear as honest-negative context — their meta-analytic primary endpoints are null — not as recommendations. Citations on this page are verified against the PubMed record by matching first author, year, title, and journal. Educational reference only — not medical advice. Anyone with a diagnosed musculoskeletal condition should follow an orthopedist / rheumatologist / endocrinologist's treatment plan.

← Home

Ask Agent Axor