recruited from August 2010 through November 2010

from a urology clinic specializing in reproductive medi-

cine. Men with a history of infertility, sexual dysfunction,

hypogonadism, fecundity anxiety, or vasectomy and age

18 or older were eligible. Men with a history of orchiec-

tomy, testicular torsion, prior malignancy, prior testoster-

one use, or prior chemotherapy exposure were excluded.

Within the cohort of men who had anogenital distance

measured, we searched for men who were also azoosper-

mic deFned as an absence of sperm in the ejaculate. The

distinction between OA and NOA was made based on

history (e.g. previous vasectomy), physical examination

(e.g. testicular size), laboratory, and surgical Fndings (e.g.

testicular biopsy) (Schoor

et al.

, 2002). In all, 98 men

were azoospermic. Mean age was 36.1 ± 8.0. Of the

cohort, 58.6% were white, 13.8% were Hispanic, and

13.8% were African American. All men provided written

consent for participation.

Genital measurements

The methods of genital measurement have been previ-

ously described. Brie²y, in the supine, frog-legged posi-

tion with the legs abducted allowing the soles of the feet

to meet, the distance from the posterior aspect of the

scrotum to the anal verge was measured using a digital

caliper (Neiko USA, Model No. 01407A). It is important

to note that others have deFned the anogenital distance

(AGD) from the anus to the anterior base of the penis

and the distance from the posterior scrotum to the anus

(as was measured in this study) as the anoscrotal distance

(ASD) (Swan

et al.

, 2005; Hsieh

et al.

, 2008; Sathyanara-

yana

et al.

, 2010). The inter-rater reliability of our mea-

surements were 0.91 for anogenital measurements at our

institution. Given the age of the patients measured, the

posterior scrotum was measured as the anterior border as

it was felt to be a more comfortable, reliable, and repro-

ducible measure.

³rom the same position, the stretched penile length

(PL) was measured from the base of the dorsal surface of

the penis to the tip of the glans. Testicular volume was

estimated at physical examination by one investigator

(LIL) in a room at approximately 25–27

C. Total testicu-

lar volume represents the sum of the right and left testes.

Hormone analysis

All

hormone

assays

were

processed

single,

experienced laboratory (Laboratory for Male Reproduc-

tive Research and Testing, Baylor College of Medi-

cine,

Houston,

TX).

Testosterone

(reference

range:

6.9–34.7 nmol

⁄

L), LH (reference range: 6–19 mIU

⁄

mL),

and ³SH (reference range: 4–10 mIU

⁄

mL) values were

assessed using an automated, one-step competitive bind-

ing assay with the Beckman Coulter Access II Immuno-

assay system (Beckman Coulter, Inc., Brea, CA). The

assays were recalibrated everyday with controls that

spanned the normal range for all hormones.

Statistical analysis

Comparisons were made using

anova

for most continu-

ous variables and Chi-squared for categorical variables.

Given the nonparametric distribution of the genital mea-

sures (i.e. AGD and PL), the nonparametric Mann–

Whitney

-test was used for comparisons. To assess the

performance of AGD length to predict the etiology of

azoospermia, the data were stratiFed on the basis of AGD

length. Multivariable logistic regression was used to deter-

mine the relationship between AGD length and etiology

of azoospermia. Bivariable regression coefFcient between

genital measures, anthropomorphic variables (height,

weight, BMI), and the etiology of azoospermia were

determined, and relationships with a

value < 0.2 were

included in the multivariable models. Receiver operating

characteristic (ROC) curves were generated using maxi-

mum likelihood estimation to Ft a binomial ROC curve

to either continuously distributed data or ordinal category

data. All

values were two-sided. Analyses were per-

formed using Stata 10 (StataCorp LP, College Station,

TX, USA).

Results

In all, 98 men were azoospermic and available for analysis

– 69 men had obstructive azoospermia (OA) and 29 men

had nonobstructive azoospermia (NOA). Of the OA men,

63 had previously undergone vasectomy, 4 had CBAVD

(congenital bilateral absence of the vas deferens), and 2

had idiopathic epididymal obstruction. Mean age was

44.2 for OA men and 32.8 for NOA men. Other demo-

graphic and anthropomorphic characteristics of the OA

and NOA men were similar (Table 1). Of the NOA men,

20 had karyotypes, of which 17 were normal, and 2

showed XXY and 1 showed a Y chromosome inversion.

Seventeen NOA men had a Y chromosome microdeletion

assessment with 3 men showing abnormalities.

Men with OA had signiFcantly longer AGD than those

with NOA (mean: 41.9 vs. 36.3 mm; median 40 vs.

31.2 mm;

= 0.01). Men with OA also had signiFcantly

larger total testis size than did those with NOA (mean:

41.7 vs. 26.4 mL, median: 44 vs. 20 ml;

< 0.01). In con-

trast, there were no signiFcant differences between penile

length in men with OA and those with NOA (mean:

125.4 vs. 123.0 mm; median: 127 vs. 121 mm;

= 0.2,

Table 2).

M. L. Eisenberg

et al.

The relationship between anogenital distance and azoospermia in adult men

2012 The Authors

International Journal of Andrology

, 2012,

, 726–730

International Journal of Andrology

2012 European Academy of Andrology

727