birth to sons with shorter perineal length, linking
environmental exposure and human genital devel-
opment.
8,9
Hsieh et al demonstrated shorter ano-
genital distances in boys with genital anomalies (ie
hypospadias and cryptorchidism), further establish-
ing a link between normal genital development and
perineal length in humans.
6
A relationship between genital measures and tes-
ticular function was recently shown by 2 groups who
related sperm production to anogenital distance in
adult men. Mendiola et al examined healthy volun-
teers, and found a positive relationship between se-
men parameters and anogenital distance.
19
Our
group simultaneously found differences in perineal
lengths between infertile and fertile men.
20
Several rodent studies have established a critical
gestational masculinization programming window
in which endocrine disruptors can permanently al-
ter genital development, growth and function.
21,22
The current report establishes a similar relationship
between genital development and hormone levels in
humans. Indeed, Andersson et al demonstrated that
infertile men have lower serum testosterone levels
than fertile men, establishing that infertile men
seem to have global testicular impairment.
23
Welsh et al showed that if the genital lengths of a
male rat were altered with flutamide in utero, they
could not be rescued by postnatal testosterone treat-
ment, suggesting that adult genital length may be
firmly established in utero.
22
As the testicular dys-
genesis syndrome hypothesizes an environmental
cause for declining testicular function, increasing
testis cancer and increasing genital anomalies, the
current study links subtle defects in genital devel-
opment to impaired adult function, suggesting that
gestational exposure may have a critical role in tes-
ticular function, supporting a fetal origin of the tes-
ticular dysgenesis syndrome.
It is important to note that of the 4 hormones as-
sessed in the current report (ie testosterone, estrogen,
LH and FSH) only testosterone is synthesized in the
testis. Thus, it is perhaps not surprising that a rela-
tionship with genital distance was only found for tes-
tosterone. While our laboratory does not routinely test
for other testis derived hormones, it would be interest-
ing to determine if antimüllerian hormone and inhibin
levels were also correlated to genital lengths.
Certain limitations warrant mention. In a refer-
ral center for male infertility it was not always pos-
sible to blind observers to the men’s diagnoses or
fatherhood status which theoretically can lead to
observer bias. However, hormone levels were not
known at the time of genital measurement. Thus,
investigators were blind to hormone measurements.
Table 2.
Multivariable linear regression model of the relationship between genital measurements and hormone parameters
±
(95% CI) Unadjusted
p Value
±
(95% CI) Adjusted*
p Value
Testosterone:
AGD
1.98
(0.18, 3.78)
0.03
2.01
(0.18, 3.84)
0.03
Penile length
1.11
(0.08, 2.14)
0.03
0.95
(
±
0.13, 2.02)†
0.08
Testicular vol
2.70
(
±
0.26, 5.65)
0.07
2.87
(
±
0.22, 5.95)
0.07
FSH:
AGD
±
0.07
(
±
0.20, 0.06)
0.30
±
0.06
(
±
0.19, 0.07)
0.34
Penile length
0.00
(
±
0.07, 0.07)
0.98
0.00
(
±
0.07, 0.07)
0.97
Testicular vol
±
0.55 (
±
0.73,
±
0.37)
²
0.01
±
0.55 (
±
0.73,
±
0.36)
²
0.01
LH:
AGD
±
0.03
(
±
0.07, 0.02)
0.21
±
0.03
(
±
0.07, 0.02)
0.26
Penile length
0.01
(
±
0.02, 0.03)
0.66
0.00
(
±
0.02, 0.03)
0.69
Testicular vol
±
0.16 (
±
0.23,
±
0.10)
²
0.01
±
0.15 (
±
0.22,
±
0.09)
²
0.01
Estradiol:
AGD
±
0.02
(
±
0.07, 0.03)
0.49
±
0.03
(
±
0.08, 0.02)
0.28
Penile length
±
0.01
(
±
0.03, 0.02)
0.71
0.00
(
±
0.03, 0.02)
0.80
Testicular vol
±
0.02
(
±
0.09, 0.05)
0.62
±
0.03
(
±
0.10, 0.04)
0.36
Testosterone/LH:
AGD
0.98
(0.12, 1.83)
0.03
0.98
(0.12, 1.85)
0.03
Penile length
0.22
(
±
0.25, 0.68)
0.35
0.23
(
±
0.27, 0.74)†
0.36
Testicular vol
2.47
(1.18, 3.76)
²
0.01
2.59
(1.27, 3.92)
²
0.01
* Adjusted for age unless otherwise noted.
† Adjusted for age and body mass index.
Table 3.
Comparison of AGD in men stratified by
hypogonadal testosterone using the Wilcoxon rank sum test
Testosterone
cutoff (ng/dl)
Less than Cutoff
Cutoff or Greater
p Value
No.
Mean AGD (SD)
No.
Mean AGD (SD)
200
21
30.7 (2.1)
95
35.0 (1.4)
0.18
220
27
30.4 (1.8)
89
35.4 (1.5)
0.09
240
42
30.5 (1.5)
74
36.4 (1.7)
0.02
260
45
30.1 (1.4)
71
36.9 (1.7)
0.01
280
50
30.2 (1.4)
66
37.3 (1.8)
²
0.01
300
62
31.6 (1.5)
54
37.3 (2.0)
0.02
ANOGENITAL DISTANCE AND REPRODUCTIVE HORMONES
597