What is androgenetic alopecia?
Androgenetic alopecia (AGA), pattern hair loss, is the most common form of hair loss in humans.
It affects about 50% of men over age 50 and 25-40% of women over a lifetime.
In men it is often called "male-pattern baldness," with thinning at the hairline and vertex.
In women it typically presents as diffuse thinning across the crown.
The driving mechanism is dihydrotestosterone (DHT), the active metabolite of testosterone produced
by 5α-reductase. DHT binds the androgen receptor (AR) in hair follicles, triggering
follicle miniaturization, follicles shrink in successive growth cycles until hairs become vellus and
eventually stop growing altogether.
DHT → AGA molecular pathway
1 Testosterone is produced in the testes / ovaries and adrenal glands.
2 5α-reductase type 2 (SRD5A2) converts testosterone to DHT inside hair follicles.
3 DHT binds the androgen receptor (AR) in the cytoplasm of dermal papilla cells.
4 The DHT-AR complex enters the nucleus and activates target genes.
5 Gene-expression changes shorten the anagen (growth) phase and lengthen the telogen (rest) phase.
6 Progressive follicle miniaturization produces vellus hairs and ultimately visible baldness.
AGA is polygenic, influenced by many genes, with estimated heritability of
80-95%. Environmental factors such as stress, nutrition, and hormonal status also significantly
modify how quickly the condition progresses.
The AR gene and CAG / GGN repeats
AR gene at a glance
| Gene name | AR (Androgen Receptor) |
| NCBI Gene ID | 367 |
| Chromosomal location | Xq12 (X chromosome, long arm, band 12) |
| RefSeq mRNA | NM_000044.6 |
| RefSeq protein | NP_000035.2 |
| mRNA length | ~4.3 kb (8 exons) |
| Protein size | 919 amino acids |
Why X-linked inheritance matters
Because AR sits on the X chromosome, its inheritance is X-linked. Men carry a single X (from their mother)
and a Y (from their father), so they inherit AR entirely from their mother. The mother in turn inherited her
X chromosome from her father, the maternal grandfather. That is why a bald maternal grandfather
is the single strongest family-history predictor of male-pattern baldness in his grandsons.
The CAG repeat tract (polyglutamine)
Exon 1 of AR contains a tandem (CAG)n repeat encoding a polyglutamine tract in the
N-terminal domain of the receptor. Repeat length varies between individuals and is a key determinant of how
sensitively the receptor responds to DHT.
| CAG repeats | Risk band | Mechanism | Reference |
|---|
| < 18 | Very high | Strongly DHT-sensitive receptor; sharply elevated androgen-responsive transcription | Choong et al. 1996 |
| 18 - 21 | High | High AR sensitivity; strong androgenic effect at the follicle | Hillmer et al. 2005 |
| 22 - 24 | Moderate | Moderate AR sensitivity; moderate risk of progression | Hillmer et al. 2005 |
| 25 - 29 | Low | Relatively low AR sensitivity; follicles more resistant to DHT | Ellis et al. 2001 |
| ≥ 30 | Protective | Long CAG tracts reduce AR transcriptional activity | Giovannucci et al. 1997 |
The GGN repeat tract (polyglycine)
Exon 1 also carries a (GGN)n repeat encoding a polyglycine tract. GGN interacts with CAG
in regulating AR activity. Short GGN (< 18) is associated with higher risk, while long GGN (≥ 24) is protective.
Its effect is weaker than CAG (weight 15% vs. 40% in Folliscope's PRS formula).
The nine-SNP AGA panel
Beyond AR, large genome-wide association studies (GWAS) have mapped many loci that contribute to AGA risk.
Folliscope implements nine of the most well-supported SNPs from peer-reviewed literature:
| rs ID | Gene | Chr | Risk allele | OR | PRS weight | Function |
|---|
| rs6152 | AR X-linked | X | G | 2.50× | 0.90 | AR exon-1 variant, boosts receptor sensitivity to DHT |
| rs1385699 | EDA2R X-linked | X | C | 2.20× | 0.85 | Ectodysplasin A2 receptor, regulates follicle signaling via the EDA pathway |
| rs12558842 | AR X-linked | X | G | 1.80× | 0.70 | AR regulatory region, affects AR expression in follicles |
| rs2497938 | AR X-linked | X | C | 1.75× | 0.65 | AR intronic variant, transcriptional modulator |
| rs7349332 | WNT10A Autosomal | 2 | T | 1.45× | 0.50 | WNT signaling, regulates follicle cycling via Wnt/β-catenin |
| rs9479482 | HDAC9 Autosomal | 7 | C | 1.35× | 0.45 | Histone deacetylase 9, epigenetic regulation of follicle genes |
| rs1160312 | PAX1/FOXA2 Autosomal | 20 | A | 1.60× | 0.60 | PAX1 transcription factor, regulates follicle differentiation |
| rs929626 | EBF1 Autosomal | 5 | C | 1.30× | 0.35 | Early B-cell factor 1, regulates hair-growth cycling |
| rs523349 | SRD5A2 Autosomal | 2 | G | 1.40× | 0.55 | 5α-reductase type 2, converts testosterone to DHT in follicles |
Polygenic risk score (PRS) formula
Hybrid mode (genetic + clinical data)
Hybrid formula
HybridScore = 0.45 × GeneticScore + 0.30 × ClinicalScore + 0.15 × FamilyScore + 0.10 × LifestyleScore
GeneticScore = 0.40 × CAGScore + 0.15 × GGNScore + 0.45 × SNPScore
ClinicalScore = 0.35 × NorwoodLudwig + 0.20 × PatternArea + 0.15 × HairPull + 0.10 × LossVolume + 0.10 × Miniaturization + 0.10 × Duration
FamilyScore = 0.35 × MaternalGrandfather + 0.25 × Father + 0.15 × PaternalGrandfather + 0.10 × Brothers + 0.08 × MotherThinning + 0.07 × Generations
LifestyleScore = 0.25 × Comorbidities + 0.25 × Stress + 0.20 × Smoking + 0.15 × Diet + 0.15 × Sleep
All component scores range 0-100. An age modifier is applied to the final score (early onset = ×1.15).
Clinical-only mode (no DNA data)
Clinical-only formula
ClinicalOnlyScore = 0.55 × ClinicalScore + 0.30 × FamilyScore + 0.15 × LifestyleScore
Risk categories
| Score range | Category | Color | Interpretation |
|---|
| 0 - 19 | Minimal | Green | No significant susceptibility signal |
| 20 - 39 | Low | Dark green | Low susceptibility, periodic monitoring |
| 40 - 59 | Moderate | Orange | Early warning, consider a clinical consult |
| 60 - 79 | High | Red-orange | Significant risk, early intervention recommended |
| 80 - 100 | Very high | Red | Very high risk, prompt consultation advised |
Limitations
Folliscope is an educational computational-biology project, not a clinical diagnostic tool.
Its results do not replace professional evaluation by a licensed dermatologist.
- The nine-SNP panel is representative, not comprehensive, over 200 loci have been linked to AGA in large GWAS studies.
- CAG / GGN thresholds come from specific population studies (predominantly Caucasian) and may differ in Asian or African populations.
- Complex gene-environment interactions (epistasis) are not modeled.
- Uploaded FASTA sequences are synthetic / individual, not clinically validated data.
- The PRS formula is simplified for educational use. Real clinical PRS models use millions of variants with carefully tuned coefficients.
- Other forms of alopecia (alopecia areata, telogen effluvium, scarring alopecias) are out of scope.
- The clinical questionnaire is self-reported and subject to reporting bias.
- Not validated on a real patient cohort, clinical validation would be required for diagnostic use.
References
- Liu Y, Tosti A, Wang ECE, Heilmann-Heimbach S, Aguh C, Jimenez F, et al. (2025). Androgenetic alopecia. Nat Rev Dis Primers, 11(1), 73.
- Gupta AK, Dennis DJ, Economopoulos V, Piguet V. (2026). The genetic landscape of androgenetic alopecia: current knowledge and future perspectives. Biology, 15(2), 192.
- Hillmer AM, Hanneken S, Ritzmann S, et al. (2005). Genetic variation in the human androgen receptor gene is the major determinant of common early-onset androgenetic alopecia. Am J Hum Genet, 77(1), 140-148.
- Heilmann-Heimbach S, Herold C, Hochfeld LM, et al. (2017). Meta-analysis identifies novel risk loci and yields systematic insights into the biology of male-pattern baldness. Nat Commun, 8, 14694.
- Choong CS, Kemppainen JA, Zhou ZX, Wilson EM. (1996). Reduced androgen receptor gene expression with first exon CAG repeat expansion. Mol Endocrinol, 10(12), 1527-1535.
- Ellis JA, Stebbing M, Harrap SB. (2001). Polymorphism of the androgen receptor gene is associated with male pattern baldness. J Invest Dermatol, 116(3), 452-455.
- Prodi DA, Pirastu N, Maninchedda G, et al. (2008). EDA2R is associated with androgenetic alopecia. J Invest Dermatol, 128(9), 2268-2270.
- Giovannucci E, Stampfer MJ, Krithivas K, et al. (1997). The CAG repeat within the androgen receptor gene and its relationship to prostate cancer. Proc Natl Acad Sci USA, 94(7), 3320-3323.
- Yip L, Zaloumis S, Irwin D, et al. (2009). Gene-wide association study between the aromatase gene (CYP19A1) and female pattern hair loss. Br J Dermatol, 161(2), 289-294.
- Norwood OT. (1975). Male pattern baldness: classification and incidence. South Med J, 68(11), 1359-1365.
- Ludwig E. (1977). Classification of the types of androgenetic alopecia (common baldness) occurring in the female sex. Br J Dermatol, 97(3), 247-254.
- Cheng Y, Lv LJ, Cui Y, Han XM, Zhang Y, Hu CX. (2024). Psychological stress impacts neurotrophic factor levels in patients with androgenetic alopecia and correlates with disease progression. World J Psychiatry, 14(10), 1437-1447.
- Kavadya Y, Mysore V. (2022). Role of smoking in androgenetic alopecia: a systematic review. Int J Trichology, 14(2), 41-48.