|17β-HSD3 deficiency||5α-Reductase deficiency||cAIS|
| Δ4-A: Δ4-androstenedione;|
|Eponymus||17-Ketosteroid reductase deficiency||Pseudovaginal perineoscrotal hypoplasia||Testicular feminization syndrome|
|Inhneritance||Autosomal recessive||Autosomal recessive||X-linked recessive|
|External phenotype||Female or ambiguous (rare)||Ambiguous or female||Female|
|blind-ending vagina||Present (80% of patients)||Present (50% of patients)||Present (100% of patients)|
|prostate||Absent||Absent or hypoplasic||Absent|
|Mullerian structures||Absent||Absent||Absent or rudimental (∼30% of patients)|
|Testes||Extra-abdominal (80-90%)||Extra-abdominal (100%)||Intra-abdominal (70% of patients)|
|Hormone profile||Δ4-A ↑; T ↓, T/Δ4-A ↓, T/DHT N, estrogens N or ↓||Δ4-A N; T N or ↑, DHT ↓, T/DHT ↑, T/Δ4-A N, estrogens N||Δ4-A N, T N or ↑, T/Δ4-A N, T/DHT N or ↓, estrogens ↑|
|Gender assignment at birth||Mainly female||Female or male||Female|
|breast development||Variable||None||Normal female|
|androgen hair||Normal male||Normal male||Absent or scanty|
|Gender role change||Present (30-50% of patients)||Present (∼75% of patients)||Not present|
That last line should tell us something. The difference between 17BHDD and 5ARD is significant, even though the symptoms are so very similar. That should tell us something too.
We have many pieces to the puzzle. There are still many gaps though. Until we can settle the timing of what happens during neurological development in the womb, we have no idea why these differences happen. Given that children in all three categories (with the exception of some only mildly feminised by 5ARD) have the same sociological upbringing, it appears the "nature vs nurture" question is firmly on the side of "nature" here. It's certain though that we don't know the mechanisms.
It's also certain that very many of those who are against basic human rights for the "gender variant" can't, or won't, believe that medical conditions like this can exist. It's against their religion, I suppose.