🧬 Aging Wiki

androgenetic-alopecia

Properties1
aliasesAGA, male pattern baldness, female pattern hair loss, FPHL, MPHL, alopecia, hair miniaturization, age-related hair loss

16 min read

Androgenetic Alopecia

Androgenetic alopecia (AGA) is the most common form of progressive hair loss in humans, driven by the intersection of androgen signaling and age-related decline of hair follicle stem cells (HFSCs). It manifests as patterned miniaturization of scalp hair follicles — a gradual conversion from thick terminal hairs to fine, unpigmented vellus-like hairs — following characteristic distributions (Hamilton–Norwood scale in men; Ludwig scale in women). Prevalence rises steeply with age: >50% of men by age 50 and ~80% of white men by age 70; roughly 50% of women show some female pattern hair loss (FPHL) by age 65 1. Two intersecting biological mechanisms drive this: (1) an androgen-dependent pathway (DHT via 5-alpha-reductase) that shortens the anagen growth phase and miniaturizes genetically susceptible follicles, and (2) an age-dependent pathway (HFSC depletion via COL17A1 proteolysis) that erodes the stem cell reservoir independently of androgens 2.

Definition and clinical features

AGA is characterized by:

  • Patterned distribution — frontal recession and vertex thinning in men (Hamilton–Norwood classification; Hamilton types described in Ann N Y Acad Sci 1951 3); diffuse crown thinning with frontal line preservation in women (Ludwig scale).
  • Hair follicle miniaturization — progressive reduction in follicle cross-sectional area, hair shaft diameter, and follicle depth across successive anagen cycles 1.
  • Anagen shortening — the proportion of follicles in anagen declines; telogen fraction rises. Horizontal scalp biopsy in established AGA shows a terminal:vellus ratio <4:1 (normal >8:1) 1.
  • Preserved follicle count (early) — a critical finding: bald scalp retains near-normal KRT15hi stem cell density (n=8, P=0.3 vs haired scalp), but both the CD200hiITGA6hi progenitor population (n=9, P=0.005) and CD34hi cells (n=3, P=0.01) are markedly depleted 4. This implies the primary defect is in stem-cell-to-progenitor conversion, not wholesale follicle or stem cell loss.
  • Eventual follicle loss — in advanced AGA, follicle miniaturization proceeds to fibrous tract replacement; this is partly irreversible 1.

AGA is not merely cosmetic. It is associated with increased cardiovascular risk (contested), metabolic syndrome, and prostate disease in men — though whether these associations reflect shared androgenic drive or confounding is debated. contradictory-evidence

Mechanism 1 — androgen-driven hair miniaturization (DHT pathway)

The androgen-dependent arm involves dihydrotestosterone (DHT), the most potent natural androgen, produced locally in scalp dermal papilla cells (DPCs) by 5-alpha-reductase (SRD5A1/2 isoforms). DHT binds the androgen receptor (AR) in DPCs, triggering transcriptional programs that shorten anagen duration and impair follicle morphogenesis.

Key steps:

  1. Testosterone → DHT conversion — catalyzed by 5α-reductase type II (encoded by SRD5A2, predominantly in frontal scalp DPCs) and type I (SRD5A1, also expressed in scalp). DHT has ~5× higher affinity for AR than testosterone 5.
  2. AR activation in DPCs — DHT-AR complex suppresses Wnt/β-catenin signaling in DPCs, reducing Wnt ligand production needed for epithelial follicle cell proliferation and anagen entry 6. Wnt/β-catenin drives the anagen-promoting cross-talk between DPCs and follicle epithelium.
  3. Anagen shortening — successive cycles produce progressively shorter anagen phases; follicle fails to reach full depth.
  4. Miniaturization — smaller DPC volume → smaller hair matrix → smaller hair shaft. Over multiple cycles, terminal follicles become vellus-like.

Evidence for androgen causation:

  • Men castrated before puberty do not develop AGA 3.
  • Finasteride (5α-reductase inhibitor) substantially reduces AGA progression in men — see Treatment landscape below.
  • Women with hyperandrogenism (PCOS, congenital adrenal hyperplasia) show higher AGA prevalence.
  • Genetic risk loci for AGA are enriched near AR (Xq11-12) and SRD5A2 loci in GWAS.

Note: the androgen-independent component of FPHL is significant — many women with FPHL have normal androgen levels, pointing to alternate drivers including the HFSC aging mechanism.

Mechanism 2 — age-related HFSC depletion via COL17A1 proteolysis

In a landmark 2016 study, Matsumura et al. identified a second, androgen-independent mechanism of hair follicle aging and loss, driven by progressive depletion of hair follicle stem cells 2.

The COL17A1 mechanism:

  1. COL17A1 (Collagen XVII / BP180) is a hemidesmosomal transmembrane collagen expressed in HFSCs that anchors them to the basement membrane and maintains their identity. It functions as an epigenetic and structural “youth” factor for the HFSC niche.
  2. Replication-associated DNA damage in HFSCs triggers activation of p53 and downstream proteases (including MMP and ADAM family metalloproteinases), leading to ectodomain shedding (proteolysis) of COL17A1 from the cell surface.
  3. Loss of COL17A1 destabilizes HFSC adhesion to the basement membrane. HFSCs undergo transepidermal elimination — they are pushed upward and shed through the epidermis, physically depleting the HFSC pool.
  4. Follicle size reduction — as the HFSC pool shrinks, follicle regenerative capacity declines, producing smaller follicles and eventually follicle loss. This mirrors aged human scalp histology.
  5. COL17A1 protein levels decline with age — in human scalp biopsies, COL17A1 expression in HFSCs shows age-dependent decline; this correlates with follicle thinning and is independent of androgen status 2.

Extrapolation from mouse model:

DimensionStatus
Pathway conserved in humans?yes — COL17A1 expressed in human HFSCs; human scalp biopsy data confirms age-related decline
Phenotype conserved in humans?yes — age-related hair thinning in women (FPHL without hyperandrogenism) likely reflects this mechanism
Replicated in humans?partial — human biopsy correlative data in Matsumura 2016; no interventional human replication yet

The COL17A1 mechanism explains why hair loss continues in women without hyperandrogenism, and why both sexes experience age-progressive thinning independently of androgen levels. It also explains why AGA worsens with age even in men with stable androgen levels. needs-replication — intervention studies targeting COL17A1 or its protease cascade have not yet been completed in humans.

Hair cycle disruption — anagen shortening

The hair cycle in normal scalp follicles progresses: anagen (growth, 2–6 years) → catagen (regression, ~2 weeks) → telogen (rest, ~3 months) → exogen (shedding). In AGA:

  • Anagen duration is progressively shortened across cycles (from years to months or weeks).
  • Telogen fraction is increased; biopsies in AGA show telogen:anagen ratio inversion.
  • In advanced AGA, kenogen (empty follicle, no hair shaft) intervals appear between telogen and the next anagen.
  • A 4:1 terminal:vellus follicle ratio is pathognomonic for AGA on horizontal scalp biopsy 1.

Anagen shortening is the proximate output of both DHT signaling (via DPC suppression of Wnt/β-catenin) and HFSC depletion (fewer HFSCs available to drive the anagen entry signal). These two mechanisms therefore converge on the same hair cycle readout.

Stem cell exhaustion — HFSC compartment in AGA

The bulge region of the outer root sheath houses the canonical HFSC population (KRT15hi/ITGA6+ in humans per Garza 2011; CD34+/K15+ in mice). Note: CD34 and CD200 mark progenitor populations derived from the bulge, not the stem cells themselves — a distinction critical to interpreting AGA pathology. Key findings in AGA scalp:

  • Garza 2011 (JCI): Bald scalp in men with AGA retains KRT15hi stem cells at near-normal density (n=8, P=0.3), but the CD200hiITGA6hi progenitor population is markedly depleted (2.3% ± 0.7% vs 0.28% ± 0.1% of all epithelial cells; n=9, P=0.005) and CD34hi cells are depleted roughly 10-fold (1.9% ± 1% vs 10.5% ± 0.3%; n=3, P=0.01) 4. Study subjects: men aged 40–65 years, n=54 total donors; 1 subject using minoxidil (excluded from key analyses). This implies a defect in KRT15hi stem-cell-to-progenitor conversion, rather than loss of the stem cell pool itself.
  • Matsumura 2016 (Science): In aged mouse and human skin, COL17A1+ HFSCs are progressively eliminated; total HFSC number declines with age 2. This is mechanistically distinct from the Garza finding (which was in AGA-affected men of various ages) and may represent the pure aging-only trajectory.
  • Reconciliation: In early/mid AGA, HFSCs may still be present in the bulge but fail to generate competent progenitors (Garza); in advanced or aged AGA, the upstream HFSC pool itself is also depleted (Matsumura). needs-replication — the two populations may represent different stages of the same disease trajectory; longitudinal HFSC tracking studies are lacking.

The niche context matters: niche signals including Wnt, BMP, FGF, and cell-cell adhesion via COL17A1 are all required for HFSC maintenance and activation. Androgen-driven suppression of DPC Wnt production impairs the niche signal from below; COL17A1 loss impairs the HFSC’s structural integration into the niche from within.

Melanocyte SC parallel: Nishimura et al. (2002) demonstrated that the hair follicle bulge (lower permanent portion) is the niche for melanocyte stem cells (McSCs) — these Dct+ cells are slow-cycling, self-maintaining, and supply melanocyte progeny to the hair matrix at each anagen 7. The 2002 paper established niche localization using Dct-lacZ transgenic mice and ACK2-mediated Kit blockade. (Note: subsequent work by Nishimura et al. 2010, Cell Stem Cell, identified TGF-β signaling via TGFβRII as a key maintenance signal for McSCs in the niche — that mechanism is not from the 2002 paper.) McSC depletion from the bulge underlies age-related hair graying, and follows a similar geometry: repeated hair cycles activate McSCs, eventually exhausting the pool — producing greying at a rate that parallels but is distinct from the HFSC depletion producing thinning.

Inflammation contribution

Perifollicular inflammation has been recognized as a component of AGA histology since early biopsy studies, though its causal role is debated:

  • Histological findings: Lymphocytic perifollicular infiltrates, primarily around the upper follicle isthmus and infundibulum, are present in ~40–75% of AGA biopsies 8. Perifollicular fibrosis (lamellar fibrosis around miniaturized follicles) is associated with advanced disease.
  • Inflammatory mediators: Elevated IL-1α, IL-6, and PGD2 (prostaglandin D2) in affected scalp; PGD2 via the CRTH2/DP2 receptor suppresses hair growth in in-vitro models 6. needs-replication — causal role of PGD2 in human AGA requires interventional confirmation.
  • SASP-adjacent signaling: Senescent DPCs accumulate with age and in AGA and secrete a pro-inflammatory secretome consistent with SASP; this further impairs follicle cycling in a paracrine fashion. See cellular-senescence.
  • Practical implication: The inflammatory component may explain partial responsiveness of some AGA cases to anti-inflammatory adjuncts (e.g., topical cetirizine, ketoconazole shampoo), but this remains secondary to anti-androgen or HFSC-restoration approaches.

Whether inflammation is a cause or consequence of follicle miniaturization is not resolved. The current view favors it as a secondary amplifier rather than a primary initiating event in classical AGA. contradictory-evidence

Treatment landscape

Treatments divide into approved pharmacological, physical, and emerging biological approaches.

Finasteride (5α-reductase inhibitor, type II)

Oral finasteride 1 mg/day (FDA-approved for men) inhibits SRD5A2, reducing scalp DHT by ~64% and serum DHT by ~70%. Clinical trial data (2-year RCTs and extension studies):

  • Halts AGA progression in ~85% of men and produces measurable hair count increases in ~65% at 1 year 5.
  • Reversal of miniaturization demonstrated histologically (increased terminal:vellus ratio) 9.
  • Efficacy maintained at 5 years in extension studies; discontinuation leads to regression within 12 months.
  • Side effects: Sexual dysfunction (decreased libido, erectile dysfunction, ejaculatory disorder) in ~2–4% of men; post-finasteride syndrome (persistent side effects after discontinuation) is described but prevalence uncertain. Finasteride is contraindicated in pregnancy (teratogenic in male fetuses).
  • Women: Not FDA-approved for premenopausal women; some off-label use in postmenopausal women. long-term-unknown for women.

Minoxidil (topical/oral potassium channel opener)

Topical minoxidil 2% (women) and 5% (men) and low-dose oral minoxidil (0.25–2.5 mg/day) are FDA-approved or commonly used. Mechanism involves potassium channel (K_ATP) opening in vascular smooth muscle and possibly direct effects on follicle cycling; precise mechanism of hair growth promotion remains uncertain. no-mechanism — the exact molecular pathway by which minoxidil prolongs anagen is not fully characterized.

  • Maintains hair count and density; approximately 30–40% of men and women show objective improvement.
  • Requires ongoing use; cessation leads to return of hair loss within months.

Dutasteride (dual 5α-reductase inhibitor, types I + II)

Inhibits both SRD5A1 and SRD5A2; reduces scalp DHT more completely than finasteride (~97% vs ~64%). FDA-approved in South Korea and Japan for AGA; off-label use in the US/Europe. Meta-analyses suggest superior hair count improvement vs finasteride 10. Similar side-effect profile to finasteride; longer half-life (~5 weeks) means side effects persist after discontinuation. long-term-unknown — long-term (>4 year) safety data in young men are limited.

Low-level laser therapy (LLLT)

FDA-cleared (510k) devices (laser combs, caps). Proposed mechanism: photobiomodulation stimulating mitochondrial activity in follicle cells. Modest effect size; low-grade evidence (heterogeneous trials) [^unsourced-lllt]. no-mechanism unsourced

Topical antiandrogens

  • Topical finasteride — formulations delivering high local scalp DHT suppression with minimal systemic exposure; approved in some EU countries; Phase 3 data showing comparable hair count to oral finasteride with lower systemic side effects.
  • Clascoterone (Winlevi) — FDA-approved for acne; topical AR antagonist; Phase 2 trial in AGA (NCT03143803) showed hair count improvement vs placebo in women with FPHL. long-term-unknown.

Hair transplant surgery

Follicular unit transplantation (FUT) and follicular unit extraction (FUE) redistribute DHT-resistant occipital follicles to bald areas. Occipital follicles are androgen-insensitive (lower AR expression, lower 5α-reductase), a property that is intrinsic to the follicle and maintained after transplantation. Gold standard for restoration of severe AGA; does not halt ongoing progressive loss in non-transplanted areas.

Emerging: HFSC-targeted and cell therapies

  • PRP (platelet-rich plasma): Multiple small RCTs show modest hair density improvement; mechanism unclear; evidence quality is low (small n, heterogeneous preparation protocols). needs-replication
  • COL17A1 stabilization / protease inhibition: Experimental; no clinical trial data yet. Matsumura 2016 suggests that interventions preventing COL17A1 ectodomain shedding could slow age-related HFSC depletion 2. needs-human-replication
  • Wnt agonists / DKK1 inhibition: DKK1 (Dickkopf 1) is elevated in AGA DPCs and suppresses Wnt signaling required for anagen; topical DKK1 inhibitors are in early development. long-term-unknown

Hallmark mapping

HallmarkConnection
stem-cell-exhaustionHFSC pool depleted by COL17A1 proteolysis and/or progenitor failure; McSC depletion produces concurrent greying
cellular-senescenceSenescent DPCs secrete SASP factors impairing follicle cycling; p16+/p21+ cells accumulate in aged follicles
chronic-inflammationPerifollicular lymphocytic infiltrates, PGD2, IL-1α amplify miniaturization
genomic-instabilityDNA damage in HFSCs triggers p53 → COL17A1 proteolysis cascade (Matsumura 2016)
deregulated-nutrient-sensingmTOR activity in HFSCs modulates stem cell quiescence vs activation; under-explored in AGA specifically

The androgen-driven arm (DHT/AR/Wnt) is not a canonical aging hallmark but an endocrine-genetic susceptibility that is unmasked and amplified by the aging-hallmark mechanisms above.

Limitations and gaps

  • HFSC longitudinal tracking — No prospective human study has tracked HFSC pool size and COL17A1 expression across decades. Cross-sectional biopsy data exists; causal trajectory in humans is inferred from mouse models. needs-human-replication
  • Female pattern hair loss mechanism — FPHL is heterogeneous; androgen-driven, androgen-independent (HFSC aging), and possibly auto-immune subtypes likely coexist. Distinguishing these clinically impacts treatment selection. no-mechanism
  • Inflammation causality — Whether perifollicular inflammation initiates, sustains, or merely co-occurs with follicle miniaturization is unresolved. RCT data testing anti-inflammatory agents as monotherapy in AGA are lacking. contradictory-evidence
  • Combination therapy evidence — Finasteride + minoxidil is widely used in combination but RCT data specifically on combination vs monotherapy are sparse; most evidence is observational or small n. needs-replication
  • Systemic associations — Proposed associations between AGA and cardiovascular disease, metabolic syndrome, and prostate disease are largely observational and confounded by shared androgen drive; mechanistic links are not established. contradictory-evidence
  • Long-term dutasteride safety in young men — Use of dutasteride (off-label) in men in their 20s–30s for cosmetic hair preservation has minimal long-term safety data. long-term-unknown
  • ICD-11 code — ICD-11 code for AGA not confirmed; the WHO ICD-11 browse API requires bearer-token authentication and was not accessible during verification. Left null pending manual lookup. unsourced

See also

Footnotes

Footnotes

  1. doi:10.1016/0190-9622(93)70106-4 · Whiting DA · J Am Acad Dermatol 1993;28(5 Pt 1):755–763 · observational (horizontal scalp biopsy series) · model: human AGA · “Diagnostic and predictive value of horizontal sections of scalp biopsy specimens in male pattern androgenetic alopecia” · closed-access (not_oa per a local paper archive) — terminal:vellus ratio cutpoints (<4:1 pathognomonic; >8:1 normal), prevalence figures, and progressive miniaturization claims sourced from this paper cannot be verified against the full text no-fulltext-access ↩ ↩2 ↩3 ↩4 ↩5

  2. doi:10.1126/science.aad4395 · Matsumura H et al. · Science 2016;351:aad4395 · in-vivo (mouse + human biopsy) · n=not extracted (multiple cohorts) · model: mouse aging + human scalp biopsies · “Hair follicle aging is driven by transepidermal elimination of stem cells via COL17A1 proteolysis” · closed-access (not_oa per a local paper archive) — quantitative claims on this wiki page (COL17A1 proteolysis mechanism, transepidermal elimination, age-dependent COL17A1 decline) cannot be verified against the full text no-fulltext-access ↩ ↩2 ↩3 ↩4 ↩5

  3. Hamilton JB · “Male hormone stimulation is prerequisite and an incitant in common baldness” · Am J Anat 1942;71(3):451–480 · doi:10.1002/aja.1000710306 · observational (castration + androgen administration studies) · model: human + primate · castration prevents AGA; testosterone administration in castrated men restores pattern. (Note: a second Hamilton paper, “Patterned loss of hair in man: types and incidence,” appeared in Ann N Y Acad Sci 1951;53:708–728, doi:10.1111/j.1749-6632.1951.tb31971.x — this is the classification paper. The wiki previously cited a non-existent “Am J Anat 1951;86(3):399–476” entry; the Am J Anat paper is 1942 and the 1951 reference belongs to Ann N Y Acad Sci.) ↩ ↩2

  4. doi:10.1172/jci44478 · Garza LA et al. · J Clin Invest 2011;121(2):613–622 · observational (human scalp biopsy, flow cytometry + hair reconstitution) · n=54 total male donors aged 40–65 (paired bald + haired scalp); KRT15hi comparison n=8, CD200hiITGA6hi comparison n=9, CD34hi comparison n=3 · model: human AGA · KRT15hi stem cells preserved (P=0.3); CD200hiITGA6hi progenitors depleted (P=0.005); CD34hi cells depleted ~10-fold (P=0.01) · full PDF verified 2026-05-05 ↩ ↩2

  5. doi:10.1046/j.1523-1747.2003.12167.x · Shapiro J, Kaufman KD · J Investig Dermatol Symp Proc 2003;8(1):20–23 · review · model: human AGA · “Use of Finasteride in the Treatment of Men With Androgenetic Alopecia” · 52 citations · bronze OA (pending download) ↩ ↩2

  6. doi:10.1007/s00403-018-1826-8 · Premanand A, Reena Rajkumari B · Arch Dermatol Res 2018;310(8):621–632 · review · model: human AGA molecular mechanisms · “Androgen modulation of Wnt/β-catenin signaling in androgenetic alopecia” · closed-access (not_oa per a local paper archive) — DHT-AR suppression of Wnt/β-catenin and PGD2/IL-1α inflammation claims attributed to this review cannot be verified against the full text no-fulltext-access ↩ ↩2

  7. doi:10.1038/416854a · Nishimura EK et al. · Nature 2002;416:854–859 · in-vivo (Dct-lacZ transgenic mouse; ACK2 Kit-blockade; vibrissal reconstitution) · model: mouse hair follicle bulge · “Dominant role of the niche in melanocyte stem-cell fate determination” · Established that Dct+ McSCs localize exclusively to the lower permanent portion/bulge, are slow-cycling and self-maintaining, and regenerate melanocytes at each anagen. Does NOT address TGF-β or SCF signaling (those findings are from Nishimura et al. 2010, Cell Stem Cell, doi:10.1016/j.stem.2009.12.010) · local PDF verified 2026-05-05 ↩

  8. doi:10.1016/j.jaad.2021.09.040 · Plante J et al. · J Am Acad Dermatol 2022;86(2):437–438 · observational (histology) · model: human AGA biopsy series · “Perifollicular inflammation and follicular spongiosis in androgenetic alopecia” · 30 citations · bronze OA (pending download) ↩

  9. doi:10.1038/sj.jidsp.5640230 · Whiting DA · J Investig Dermatol Symp Proc 1999;4(3):308–311 · in-vivo (human biopsy, finasteride-treated) · model: human AGA · “Measuring Reversal of Hair Miniaturization in Androgenetic Alopecia by Follicular Counts” · bronze OA but download failed (no accessible URL after filtering) — finasteride reversal of miniaturization claim cannot be verified against full text no-fulltext-access ↩

  10. doi:10.2174/1574884712666170310111125 · review · model: human AGA trials · “Dutasteride in Androgenetic Alopecia: An Update” · Curr Clin Pharmacol 2017 · 75 citations · closed-access, not downloaded no-fulltext-access ↩

Graph View

Backlinks