Bone

Bone is a dynamic mineralized connective tissue that serves simultaneously as the structural scaffold of the body, the primary reservoir for calcium and phosphate, and an endocrine organ producing hormones (particularly FGF23) with systemic effects on the kidney, cardiovascular system, and parathyroid gland. Bone houses the bone-marrow niche that sustains lifelong hematopoiesis, and its cellular composition — osteoblasts, osteocytes, osteoclasts, and bone marrow stromal cells — is continuously renewed through a tightly coupled remodeling cycle.

With aging, this coupling breaks down. Net bone loss begins in the third to fourth decade, accelerates post-menopause in women, and proceeds more slowly but inexorably in men. Concurrent with structural degradation, senescent osteocytes accumulate and secrete SASP factors that propagate local bone destruction and impair the bone marrow niche. Bone also participates in the bone-vascular axis paradox: as bone loses mineral, arteries paradoxically gain it — a shared mechanism involving ectopic runx2 activity in vascular smooth-muscle cells and reduced inhibitory signals including matrix-gla-protein and osteopontin. These interconnected aging trajectories make bone a central tissue node in the hallmarks-of-aging framework.

Anatomy at a glance

Macrostructure: cortical vs trabecular

Adult bone contains two macroscopic compartments with distinct turnover kinetics and aging trajectories:

Compartment	Location	Fraction of bone mass	Annual turnover	Aging trajectory
Cortical (compact)	Shafts of long bones, outer shell of all bones	~80%	~3%/yr	Thinning + porosity increase; trabecular bone lost first but cortical loss drives fracture risk most in men
Trabecular (cancellous)	Vertebral bodies, femoral neck, ends of long bones, inside flat bones	~20%	~25%/yr	Faster loss; connectivity deteriorates — trabecular plates perforate and disconnect, irreversibly weakening load-bearing structure

Trabecular bone loss generally precedes and outpaces cortical loss in the early post-menopausal period, which is why vertebral fractures are common early. Hip fractures, driven primarily by cortical compromise, become more prevalent in later decades. unsourced — quantitative rates of compartment-specific loss in longitudinal human DXA studies need primary citation.

Bone matrix composition

Bone matrix is a composite of:

~65% mineral — predominantly hydroxyapatite (Ca₁₀(PO₄)₆(OH)₂) crystals; provides compressive stiffness
~25% organic — ~90% Type I collagen (tensile toughness) + ~10% non-collagenous proteins (NCP) including osteopontin, matrix-gla-protein, osteocalcin, fibronectin, and bone sialoprotein
~10% water — within the lacunocanalicular network; functions in ion exchange and mechanosensory fluid flow

With aging, advanced glycation end-products (AGEs) accumulate in the collagen scaffold — cross-linking collagen fibrils abnormally, reducing post-yield toughness and increasing brittleness. AGE accumulation in the lacunocanalicular network also impairs osteocyte mechanosensing ¹. Hydroxyapatite crystal size increases and mineral substitution patterns shift, further altering mechanical properties.

Cellular composition

Cell type	Origin	Function	Aging notes
osteocytes	Osteoblast differentiation / entombment	Mechanosensing; SOST/sclerostin secretion (Wnt inhibitor); FGF23 secretion; LCN maintenance	95% of bone cells; apoptosis accumulates; senescent subset drives SASP; micropetrosis (empty lacunae) increases — see osteocytes
Osteoblasts	BMSCs via runx2 + bmp-2 signaling	Collagenous matrix secretion; mineralization; transition to osteocyte or bone-lining cell	Reduced number and activity with age; BMSC fate drift toward adipogenesis (see below) — see osteoblasts
Osteoclasts	Monocyte/macrophage lineage via RANKL signaling	Bone resorption; releases growth factors during resorption lacuna	Relative hyperactivity vs osteoblasts in aging — net resorption; regulated by RANKL/OPG ratio — see osteoclasts
BMSCs (marrow stromal)	Mesenchymal lineage	Osteoblast + adipocyte + chondrocyte precursors; niche support	Age-associated lineage drift: PPAR-γ elevation + Wnt/β-catenin decline → adipogenesis predominance → marrow fat expansion — see [[mesenchymal-stem-cells
hematopoietic-stem-cells	Hematopoietic niche in marrow	Blood cell production; reside in endosteal + perivascular niches	HSC pool aging documented on bone-marrow and hematopoietic-stem-cells; cross-link for full detail

Bone remodeling cycle

The remodeling cycle maintains bone mass and repairs micro-damage. It proceeds in four phases and takes ~4–6 months in young adults:

Activation — osteocyte apoptosis or mechano-signal → RANKL elevation → osteoclast recruitment
Resorption — osteoclasts dissolve mineral and digest collagen; release of TGF-β, IGF-1, BMPs from matrix
Reversal — TRAP+ mononuclear cells prepare the resorption pit; coupling signals recruited
Formation — osteoblasts recruited by released growth factors; deposit new collagenous osteoid; mineralize

Coupling (resorption → formation) is critical: the growth factors released during osteoclast resorption (TGF-β, IGF-1, bmp-2) recruit and activate osteoblasts. Uncoupling with age — driven by SASP, reduced osteoblast numbers, and impaired BMP/Wnt signaling — produces the net bone loss of aging.

Key molecular regulators:

RANKL (TNFSF11) → RANK on osteoclast precursors → osteoclastogenesis. RANKL produced by osteoblasts and osteocytes.
Osteoprotegerin (OPG) — decoy receptor produced by osteoblasts; blocks RANKL → RANK; the OPG/RANKL ratio is the master remodeling switch.
SOST / sclerostin — secreted by osteocytes; inhibits Wnt/β-catenin in osteoblasts → reduces bone formation. The pharmacological target of romosozumab (anti-sclerostin mAb, FDA 2019).
bmp-2 — drives BMSC → osteoblast differentiation via SMAD1/5/8 → runx2 expression.
runx2 — master osteoblast TF; required for osteoblast commitment and osteocyte differentiation; paradoxically upregulated in VSMCs in calcifying arteries (see bone-vascular paradox below).

Mineral metabolism axis

Bone is the primary systemic reservoir for calcium (99% of body calcium) and phosphate (85% of body phosphate). The mineral homeostasis hormonal axis involves:

PTH (parathyroid hormone) — elevated in response to hypocalcemia → stimulates osteoclast activity → liberates calcium; stimulates renal calcium reabsorption + phosphate wasting; stimulates 1,25(OH)₂D synthesis
Vitamin D (1,25-dihydroxyvitamin D₃ / calcitriol) — promotes intestinal calcium and phosphate absorption; stimulates osteoblast activity
FGF23 — produced predominantly by osteocytes; suppresses renal phosphate reabsorption (phosphaturia) and suppresses 1,25(OH)₂D synthesis; the counter-regulatory phosphaturic hormone. Requires klotho as obligate co-receptor. FGF23 rises with age and CKD, contributing to low calcitriol, secondary hyperparathyroidism, and downstream cardiovascular risk.

Aging-relevant changes

1. Bone mineral density decline

Peak BMD is attained in the late second to early third decade, varies substantially by sex, ethnicity, and habitual loading. Thereafter:

Women: ~0.5–1%/yr loss; accelerates markedly in the first 5–10 years post-menopause (estrogen loss removes the osteoclast-suppressive effect of estrogen → OPG downregulation → RANKL-driven resorption surge)
Men: ~0.3–0.5%/yr loss, continuous but unaccelerated, driven partly by age-associated testosterone decline

The clinical threshold for osteoporosis is T-score ≤ −2.5 (DXA, WHO criteria).

2. Osteocyte apoptosis and micropetrosis

Individual osteocytes can survive for decades in stable cortical bone, but apoptosis accumulates with age, particularly in cortical bone. Dead osteocytes leave behind empty lacunae that progressively fill with hydroxyapatite (micropetrosis) — increasing local mineral density while eliminating the living mechanosensory cell. Micropetrosis correlates with cortical porosity and impairs fluid flow through the lacunocanalicular network, diminishing the mechanosensory signal that drives Wnt-dependent bone formation. needs-replication — quantitative human cortical micropetrosis rate with age needs dedicated citation.

3. Senescent osteocyte accumulation and Bone-SASP

A subset of aging osteocytes enter cellular senescence (p16^INK4a+^ / p21^CIP1+^) rather than dying by apoptosis. Senescent osteocytes are disproportionately pro-resorptive: they produce SASP factors (IL-6, IL-8, MMP-3, MMP-13, RANKL) that recruit osteoclasts, suppress adjacent osteoblast function, and propagate senescence to neighboring osteocytes ².

The “Bone-SASP” concept (Fang et al. 2023) formalizes the idea that osteocyte-derived SASP factors act both locally (within bone) and systemically (on the marrow niche) to amplify bone loss beyond what stochastic osteocyte death alone would produce ³. Critically, the senescent signal propagates: Farr et al. 2023 (JCI) showed that in aged mice, local (osteocyte-specific, DMP1-Cre × p16-LOX-ATTAC) senolysis improved lumbar spine trabecular bone formation but had no significant effect on femur trabecular or cortical parameters, bone resorption, or marrow adiposity. Systemic senolysis (p16-INK-ATTAC, pan-tissue p16⁺ clearance) improved both spine and femur parameters and additionally reduced bone resorption and marrow adipocyte numbers — benefits absent with local senolysis alone. Furthermore, only systemic senolysis significantly reduced Rankl expression in whole bone, implying that non-osteocyte senescent cells are important RANKL sources in aging bone. Transplantation of senescent fibroblasts (~10⁶ cells i.p.) into young adult (4-month-old) male C57BL/6 mice (n=11/group) induced senescence in distant host osteocytes at 60 days post-transplant (confirmed by TAF assay), establishing that circulating SASP factors from non-skeletal senescent cells can propagate senescence to osteocytes in vivo ⁴.

Osteocyte ablation experiments (Ding et al. 2022, eLife) confirmed the causal role of osteocyte presence for bone and marrow homeostasis: partial ablation of DMP1-positive osteocytes (DMP1cre × Rosa26-DTA heterozygotes; complete ablation is perinatally lethal) caused severe osteoporosis, sarcopenia, degenerative kyphosis, and shortened lifespan (~20–40 weeks vs >2 years in wild-type), accompanied by SASP induction in both mesenchymal (osteoblast-lineage) and hematopoietic (myeloid-lineage) cells as shown by single-cell RNA sequencing ⁵. Myeloid-biased hematopoiesis and impaired lymphopoiesis (reduced B cells, expanded neutrophils and monocytes) co-occurred, demonstrating that osteocyte depletion reshapes the marrow niche at the lineage-commitment level. This cross-tissue consequence of osteocyte dysfunction mechanistically links bone aging to the broader marrow niche dysfunction documented on bone-marrow.

4. BMSC adipogenic lineage drift

With age, bone marrow stromal cells (BMSCs) — the osteoblast precursor pool — shift toward adipogenic differentiation. Mechanistic drivers include:

Elevated PPAR-γ expression (pro-adipogenic) vs reduced Wnt/β-catenin (pro-osteogenic)
Reduced runx2 activity (key osteoblast TF) in BMSCs
SASP-mediated inhibitory cytokines (IL-6, TNF-α) that suppress RUNX2 and BMP signaling

The result is marrow fat expansion (yellow marrow replacement), reduced osteoblast output, and impaired bone formation. BMSC-derived marrow adipocytes also modify the hematopoietic niche. unsourced — quantitative human BMSC adipogenic shift data need primary citation.

5. FGF23 dysregulation

Osteocyte stress and klotho decline elevate circulating FGF23 with age. Elevated FGF23:

Suppresses renal calcitriol synthesis → secondary hyperparathyroidism
Stimulates left ventricular hypertrophy via FGFR4 signaling (Klotho-independent)
Associates independently with cardiovascular mortality across community and CKD populations

Cross-reference fgf23 (verified) for the full quantitative evidence base. The aging-associated FGF23 rise represents a systemic endocrine signature of bone-mineral axis stress.

6. The bone-vascular axis paradox

A consistent clinical observation across populations: bone mineral density loss and vascular calcification are positively correlated — where one is worse, so is the other — despite being superficially antagonistic processes ⁶. This is the “calcification paradox.”

Shared mechanistic underpinnings include:

runx2 mislocalization — in young bone, RUNX2 drives osteoblast differentiation and is absent in vascular smooth muscle cells (VSMCs). In aging and CKD, VSMCs ectopically upregulate RUNX2 → undergo osteogenic trans-differentiation → vascular mineralization. Meanwhile, reduced RUNX2 function in BMSCs contributes to osteoblast insufficiency. The same TF drives opposing misregulation in two tissues simultaneously.
matrix-gla-protein (MGP) insufficiency — MGP is a potent inhibitor of tissue calcification; its activity requires vitamin K-dependent carboxylation. Deficiency (dietary or CKD-driven) simultaneously permits vascular calcification and impairs bone mineral regulation. See matrix-gla-protein.
osteopontin — secreted by osteocytes; inhibits hydroxyapatite nucleation in soft tissues; altered in aging.
Bone-derived extracellular vesicles — aged bone matrix releases EVs that carry crystalline mineral and calcification-promoting cargo; a 2022 Nature Communications study showed these EVs function as systemic messengers for the calcification paradox, promoting vascular calcification while reflecting ongoing bone demineralization. needs-replication — single study; mechanistic chain not fully characterized.

For full discussion of the vascular side, see vascular-calcification.

Key aging phenotypes

osteoporosis — T-score ≤ −2.5; clinically manifests as fragility fractures (hip > vertebral > wrist)
Osteopenia — T-score −1.0 to −2.5; subclinical bone loss; high population prevalence in 50–65 year olds
Fragility fracture — fracture from low-energy trauma (≤ fall from standing height); hip fractures carry ~20–30% one-year mortality in the elderly
sarcopenia — bone and muscle are mechanically coupled; bone loss and muscle loss co-occur in “osteosarcopenia”; mechanostimulus from muscle contraction is required to suppress osteocyte sclerostin and sustain Wnt-driven bone formation

Interventions (brief; depth on intervention and compound pages)

Intervention	Mechanism	Evidence level
Resistance / impact exercise	Mechanostimulus → osteocyte fluid shear → SOST suppression → Wnt activation → bone formation	Strong; multiple RCT meta-analyses in postmenopausal women ⁷
Vitamin D + calcium adequacy	Substrate; suppresses PTH-driven resorption	Strong for fracture prevention in deficient populations; weaker in replete
Bisphosphonates	Osteoclast apoptosis-induction → reduces resorption	Pharmacological standard-of-care; strong RCT evidence
Denosumab (anti-RANKL mAb)	Blocks RANKL → RANK → suppresses osteoclastogenesis	Phase 3+ evidence; approved; rebound bone loss on discontinuation
Romosozumab (anti-sclerostin mAb)	Blocks SOST → Wnt disinhibition → anabolic + antiresorptive dual mechanism	FDA approved 2019; cardiovascular safety signal (see osteocytes for mechanism)
Teriparatide (intermittent PTH 1-34)	Intermittent PTH → net anabolic (vs continuous PTH → catabolic)	Phase 3; significant BMD gain and fracture risk reduction
Senolytics (dasatinib + quercetin)	Senescent osteocyte clearance	Preclinical mouse: improved spine trabecular BV/TV ⁴; Phase 2 RCT (n=60 postmenopausal women): primary endpoint (CTx bone resorption) did not differ from control at 20 weeks (p=0.611); P1NP formation marker transiently increased at 2–4 weeks in the overall group; exploratory analysis in high-senescent-burden tertile showed P1NP ↑ (+34%, p=0.035) and CTx ↓ (−11%, p=0.049) at 2 weeks post-dosing, plus radius BMD ↑ (+2.7%, p=0.004) at 20 weeks ⁸ — supersession candidate: clinical result mixed; patient-stratified benefit hypothesis requires prospective testing
[[interventions/lifestyle/mediterranean-diet	Mediterranean diet]]	Adequate calcium, vitamin K2 (MGP carboxylation), polyphenols

Hallmark connections

Hallmark	Bone mechanism
cellular-senescence	Senescent osteocyte accumulation → Bone-SASP → RANKL elevation + osteoblast suppression + marrow niche disruption
chronic-inflammation	Osteocyte SASP drives local and systemic inflammatory cytokines (IL-6, TNF-α, IL-1β) that suppress bone formation and promote resorption
stem-cell-exhaustion	BMSC adipogenic drift → reduced osteoblast pool; HSC niche degradation in marrow
deregulated-nutrient-sensing	FGF23-Klotho-PTH axis dysregulation; impaired calcium/phosphate sensing; mtor and insulin-igf1 signaling modulate osteoblast function
mitochondrial-dysfunction	Osteoblast and osteocyte energy demand is high; mitochondrial ROS drives osteoblast apoptosis; documented in aged rodent bone needs-human-replication

Limitations and gaps

#gap/needs-human-replication — senolytic clearance of senescent osteocytes for fracture prevention; mitochondrial dysfunction role in human osteocyte biology
#gap/needs-replication — bone-derived extracellular vesicles as systemic calcification paradox messengers (single Nature Comms 2022 study); quantitative micropetrosis rates in human cortical aging
#gap/unsourced — BMSC adipogenic-shift quantitative data in humans; cortical vs trabecular longitudinal loss rates in humans
#gap/dose-response-unclear — optimal vitamin K2 dose for MGP carboxylation and vascular/bone benefit
The osteocyte single-cell aging atlas does not yet exist; Tabula Muris Senis bone-marrow captures stromal/endosteal cells, not embedded osteocytes — see osteocytes for detail

Cross-references

osteocytes — terminally differentiated bone cells; FGF23 source; SOST source; senescence hub (sibling-seeded 2026-05-23)
runx2 — master osteoblast TF; ectopic VSMC expression drives vascular calcification (sibling-seeded 2026-05-23)
bmp-2 — drives BMSC → osteoblast commitment (sibling-seeded 2026-05-23)
osteopontin — non-collagenous bone matrix protein; soft-tissue calcification inhibitor (sibling-seeded 2026-05-23)
matrix-gla-protein — potent vascular and soft-tissue calcification inhibitor; vitamin K-dependent (sibling-seeded 2026-05-23)
fgf23 — bone-derived phosphaturic hormone; elevated in aging (verified 2026-05-23)
klotho — obligate FGF23 co-receptor; declines with age
bone-marrow — marrow niche (existing tissue page, verified)
hematopoietic-stem-cells — resides in bone marrow niche (existing, verified-partial)
mesenchymal-stem-cells — BMSC osteoblast precursors (existing cell-type page)
vascular-calcification — the arterial side of the bone-vascular axis paradox (sibling-seeded 2026-05-23)
arterial-stiffening — downstream cardiovascular phenotype; mechanistically linked via FGF23 + vascular calcification
cellular-senescence, chronic-inflammation, stem-cell-exhaustion, deregulated-nutrient-sensing — hallmark pages
exercise — resistance training as anabolic bone stimulus
mediterranean-diet — dietary pattern associated with bone health

Footnotes

doi:10.1016/j.bone.2026.117890 · PMID 41962789 · Liu CJ et al. · in-vivo · model: senile osteoporosis mouse (C57BL/6J, age not specified in summary) · Bone 2026 Aug · AGE accumulation in osteocyte canalicular network associates with deterioration of canalicular physical properties; impaired fluid flow → reduced mechanosensing · DOI confirmed via PubMed esummary 2026-05-23 (corrected from 117400 → 117890); not in local archive ↩
doi:10.1007/s11914-020-00619-x · PMID 32794138 · PMC7541777 · Farr JN, Kaur J, Doolittle ML, Khosla S · review · Current Osteoporosis Reports 2020 Oct · narrative review of osteocyte cellular senescence; documents p16/p21 accumulation in aged bone and downstream SASP-driven resorption promotion · PMC confirmed 2026-05-23; DOI lookup failed (404 at PMC URL); PMC7541777 available as verification fallback ↩
doi:10.1007/s00223-023-01100-4 · PMID 37256358 · PMC10230496 · Fang CL, Liu B, Wan M · review · Calcified Tissue International 2023 · formalizes “Bone-SASP” concept; osteocyte SASP acts locally on bone remodeling and systemically on marrow niche; includes glucocorticoid-induced bone damage and osteoarthritis mechanisms; links bone aging SASP to broader inflammaging · local PDF downloaded and verified at concept level 2026-05-23 ↩
doi:10.1172/JCI162519 · PMID 36809340 · Farr JN, Saul D, Doolittle ML et al., Khosla S · in-vivo · model: aged C57BL/6 mice (20 months); local model: DMP1-Cre⁺/⁻ × p16-LOX-ATTAC (osteocyte-specific); systemic model: p16-INK-ATTAC · Journal of Clinical Investigation 2023;133(8):e162519 · AP20187 10 mg/kg twice-weekly i.p. for 4 months; local-model n=15 females + 10 males per group; systemic-model comparison from prior publication; transplant arm: ~10⁶ Sn/non-Sn fibroblasts i.p. into 4-month-old male C57BL/6 WT mice (n=11/group) · local senolysis improved lumbar spine trabecular BV/TV (via increased bone formation) but had no effect on femur trabecular/cortical bone, resorption, or marrow adiposity; systemic senolysis improved both spine and femur parameters AND reduced resorption and marrow adipocyte numbers; Rankl reduction was systemic-only; transplanted Sn cells induced TAF⁺ osteocyte senescence in young host mice at 60 days post-transplant, confirming non-cell-autonomous SASP propagation · local PDF verified end-to-end 2026-05-23 ↩ ↩²
doi:10.7554/eLife.81480 · PMID 36305580 · Ding P, Gao C, Gao Y et al. · in-vivo · model: DMP1cre × Rosa26-DTA heterozygotes (DTAhet); partial osteocyte ablation (complete ablation is perinatally lethal) · eLife 2022;11:e81480 · partial osteocyte depletion caused severe osteoporosis, sarcopenia, degenerative kyphosis (apparent by 13 weeks), and shortened lifespan (~20–40 weeks vs >2 years WT); scRNA-seq confirmed SASP induction in mesenchymal (osteoblast-lineage) and hematopoietic (myeloid-lineage) cells; myeloid-biased hematopoiesis with expanded neutrophils/monocytes and impaired B-cell lymphopoiesis; causal demonstration that osteocyte dysfunction drives tissue-wide aging phenotypes via SASP · local PDF downloaded and verified 2026-05-23 ↩
doi:10.1016/j.arteri.2019.03.008 · PMID 31221532 · García-Gómez MC, Vilahur G · review · Clínica e Investigación en Arteriosclerosis 2020 Jan-Feb · narrative review of shared osteoporosis/vascular calcification mechanisms; summarizes epidemiological and mechanistic overlap including RUNX2, MGP, estrogen withdrawal, and inflammation · DOI corrected 2026-05-23 (was 2020.01.002; correct suffix is 2019.03.008 per PubMed record for PMID 31221532); closed-access (not_oa); verified at abstract/metadata level only ↩
doi:10.7150/ijms.130435 · PMID 42158825 · Chen KH et al. · systematic-review and meta-analysis · International Journal of Medical Sciences 2026 · nutritional supplementation combined with exercise for musculoskeletal health in women across reproductive stages; includes postmenopausal bone density outcomes · DOI confirmed via PubMed esummary 2026-05-23; cited for intervention-evidence framing only; primary fracture-endpoint RCT data for exercise requires dedicated bone-exercise meta-analysis citation needs-replication ↩
doi:10.1038/s41591-024-03096-2 · PMID 38956196 · PMC11705617 · Farr JN, Atkinson EJ, Achenbach SJ et al., Khosla S · phase-2 rct · n=60 postmenopausal women · Nature Medicine 2024;30(9):2605-2612 · intermittent dasatinib + quercetin (D+Q) vs placebo; primary endpoint CTx bone resorption at 20 weeks: no significant difference (median D+Q −4.1% vs control −7.7%; p=0.611); P1NP bone formation transiently increased relative to control at 2 weeks (+16%, p=0.020) and 4 weeks (+16%, p=0.024) but not at 20 weeks; exploratory high-senescent-burden tertile (T-cell p16 mRNA): D+Q increased P1NP (+34%, p=0.035) and reduced CTx (−11%, p=0.049) at 2 weeks post-dosing; radius BMD increased +2.7% (p=0.004) at 20 weeks (the only BMD signal at the 20-week timepoint); no serious adverse events · TIMING CORRECTION 2026-05-23 via osteoporosis-phenotype verifier pass: original footnote stated all three tertile metrics at 20 weeks; correct PDF reading is P1NP/CTx at 2 weeks + BMD at 20 weeks · SUPERSESSION CANDIDATE relative to preclinical-only framing: first human RCT of senolytics for bone; primary endpoint negative in overall group; exploratory biomarker-stratified benefit suggests senescent cell burden may need to be pre-selected; prospective stratified trial needed ↩

🧬 Aging Wiki

Explorer

bone