- Top of page
- Materials and methods
In humans, recent studies have correlated anogenital distance (AGD) in adult men to intrinsic testicular function. Although rodent studies suggest that AGD is determined in utero and remains constant in adult life, it is not certain if AGD remains constant across a man's adult life. We sought to determine if adult male AGD varies based on age. A cross-sectional study of men being evaluated at a men's health clinic. Anogenital distance (the distance from the posterior aspect of the scrotum to the anal verge) and penile length (PL) were measured using digital callipers. anova and linear regression were used to determine correlations between AGD, fatherhood status and age. In all, 473 men were included in the analysis with a mean age of 43 ± 13 years. The mean AGD for the group was 39 ± 13 mm. Anogenital distance did not vary between age categories for the entire group, for fathers, and for childless men. Moreover, penile length also remained constant across age categories. On adjusted analyses stratified by fatherhood status, there was no relationship between AGDp and age. The current cross-sectional study demonstrates that anogenital distance, defined as the distance from the posterior scrotum to the anal verge, is similar for men of different ages. As such, AGD may provide a measure for genital development and function throughout adult life. However, confirmation with longitudinal studies is needed.
- Top of page
- Materials and methods
Anogenital distance (AGD), a marker for genital development, has been examined in both animals and humans.(Swan et al., 2005; Scott et al., 2007; Hsieh et al., 2008; Torres-Sanchez et al., 2008) As it is a sexually dimorphic measure, AGD was initially used to sex animals.(Marois, 1968; Greenham & Greenham, 1977; Hsieh et al., 2008) More recently, human studies have also shown that boys have longer perineal lengths than girls.(Salazar-Martinez et al., 2004; Torres-Sanchez et al., 2008; Thankamony et al., 2009; Sathyanarayana et al., 2010) Investigators have also used AGD to show that agents which disrupt androgen signalling in animal models can lead to abnormal genital lengths and even altered testicular function as measured by testosterone and sperm production.(Foster et al., 2001; Scott et al., 2008; Martino-Andrade et al., 2009; Cowin et al., 2010) Moreover, rodent studies suggest a narrow masculinization programming window, suggesting that prenatal influences on androgen signalling have more profound impacts on genital lengths than postnatal insults. (Welsh et al., 2008, 2010; van den Driesche et al., 2011).
In humans, recent studies have correlated AGD in adult men to intrinsic testicular function. (Eisenberg et al., 2011,2012b,c; Mendiola et al., 2011) Assuming that human AGD is determined in utero as it is in animals, such studies suggest that in utero influences may impact genital development and adult testicular function. Thus, AGD may provide an assessment of a man's genital development and intrinsic testicular function. However, it is not certain if AGD remains constant across a man's adult life as other anthropomorphic measures have been shown to change.(Cline et al., 1989) We sought to determine if adult male AGD varies based on age.
Materials and methods
- Top of page
- Materials and methods
The methods of collection and cohort assembly have been previously reported. (Romano-Riquer et al., 2004; Eisenberg et al., 2011) Briefly, after obtaining Institutional Review Board approval from Baylor College of Medicine, patients were recruited from a urology clinic specializing in reproductive and sexual medicine from August 2010 through October 2011. Date of visit, reason for visit, observer, anthropomorphic measurements and relevant laboratory data were recorded. All men provided written consent for participation.
The methods of genital measurement have been previously described. (Eisenberg et al., 2011) In the supine, frog-legged position 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 calliper (Model no. 01407A; Neiko, USA) This was defined as the posterior AGD (AGDp). From the same position, the distance from the anal verge to the superior border of the penis was also measured (AGDa). It is important to note that others have defined the anogenital distance (AGD) from the anus to the anterior base of the penis (AGDa in this study) and the distance from the posterior scrotum to the anus (as was defined in this study as AGDa) as the anoscrotal distance (ASD). (Swan et al., 2005; Hsieh et al., 2008; Sathyanarayana et al., 2010) We originally only measured AGDp, but began measuring AGDa at the midpoint of study accrual.
From 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.
anova was used to compare continuous variables. Pearson correlation coefficients were calculated to determine the association between continuous variables. Linear regression models were used to determine the relationship between genital measures and age. Covariates (i.e. race, fatherhood, BMI, height, weight) included in the models were included if bivariate analysis showed correlation with AGD. Age was analysed as a continuous and categorical variable (<30, 30–40, 40–50, 50–60, 60–70, 70+) with no differences in the overall conclusions. Analyses listed reflect age as a continuous variable. Given the nonparametric distribution of the semen parameters and AGD, linear regression models were also run with log transformed variables with no differences in the overall conclusions. AGD, PL and testis volume were analysed as continuous values for all analyses. To assess for effect modification by age or ethnicity, stratified analyses were performed with no change in the conclusions. Moreover, serum testosterone levels were also included in separate models with no change in overall conclusions. Limited patient numbers prevented meaningful analysis after stratification by patient diagnosis. However, patients evaluated for infertility had lower AGD, although other diagnoses had similar AGDs. All p values were two-sided ones. Analyses were performed using Stata 10 (StataCorp LP, College Station, TX, USA).
- Top of page
- Materials and methods
In all, 473 men were included in the analysis with a mean age of 43 ± 13.0. Men were seen in clinic for infertility, hypogonadism, erectile dysfunction, vasectomy evaluation or reversal, or general urologic complaints (e.g. voiding dysfunction, urolithiasis, prostate cancer screening). A majority of the men were white and overweight or obese (Table 1). There were approximately equal numbers of fathers and childless men.
Table 1. Demographic, reproductive and anthropomorphic characteristics of the cohort
|Characteristic|| || n ||% or Mean (SD)|
|Office visit||General urology||38||8.1|
|Height (in)|| ||453||70.6 (2.88)|
|Weight (lbs)|| ||452||205.4 (39.9)|
Anogenital distance did not vary between age categories (Fig. 1). The relationship remained for both fathers and childless men (Table 2). Moreover, penile length also remained constant across age categories. Consistent with earlier work, AGD was significantly shorter in childless men compared to fathers (36.4 mm vs. 41.9 mm, p < 0.01) whereas PL was similar between groups (113.3 mm vs. 115.6 mm, p = 0.29). AGDa was not different between childless men and fathers (118.2 mm vs. 119.9 mm, p = 0.41).
Figure 1. Boxplot showing the interquartile range (IQR) of the AGDp stratified by age deciles. Median value is denoted with horizontal bar. Whiskers designate 1.5 × IQR.
Download figure to PowerPoint
Table 2. Distribution of genital measures stratified by age deciles. Comparisons are made by anova. AGDp, AGDa, PL by age categories
| || ||AGDp||AGDa||PL|
|Subjects||Age|| n ||Mean (SD)|| n ||Mean (SD)|| n ||Mean (SD)|
|All||<30||54||38.4 (12.8)||25||117.2 (14.4)||54||115.4 (23.9)|
|30–40||194||38.2 (12.8)||97||116.9 (16.5)||194||112.1 (23.5)|
|40–50||97||41.0 (15.0)||63||122.8 (18.1)||95||114.5 (24.8)|
|50–60||62||40.1 (14.1)||43||119.1 (19.6)||62||117.3 (22.3)|
|60–70||46||37.1 (11.4)||40||120.6 (14.7)||46||119.5 (20.0)|
|70+||20||41.2 (14.7)||19||117.1 (15.5)||19||112.5 (23.5)|
|p for difference||0.45|| ||0.34|| ||0.39|
|Fathers||<30||3||42 (5.3)||0|| ||3||115.3 (21.5)|
|30–40||58||41.4 (12)||26||117.4 (13.5)||58||114.6 (25.7)|
|40–50||61||44.7 (14.6)||40||121.7 (18.6)||61||113.2 (26.1)|
|50–60||49||42.3 (14.7)||33||121.3 (20.4)||49||118.9 (20.4)|
|60–70||38||38.5 (11.4)||32||120 (15.8)||38||118.9 (21.4)|
|70+||18||40.1 (15.1)||17||116.3 (15)||17||110.5 (24)|
|p for difference||0.36|| ||0.75|| ||0.68|
|Childless||<30||51||38.2 (13.1)||25||117.2 (14.4)||51||115.4 (24.2)|
|30–40||136||36.8 (13)||71||116.6 (17.6)||136||111.1 (22.6)|
|40–50||36||34.7 (13.9)||23||124.8 (17.5)||34||116.7 (22.4)|
|50–60||13||31.7 (7.6)||10||111.9 (15.5)||13||111.5 (28.6)|
|60–70||8||30.6 (9)||8||123 (9.7)||8||122.2 (11.5)|
|70+||2||51 (7.1)||2||123.5 (24.7)||2||129.5 (3.5)|
|p for difference||0.17|| ||0.26|| ||0.45|
AGDp showed no linear relationship with height (r = 0.0003, p = 0.99), weight (r = 0.04, p = 0.42), nor BMI (r = 0.04, p = 0.41). In contrast, AGDa showed a linear relationship with weight (r = 0.24, p < 0.01), BMI (r = 0.22, p < 0.01), but not height (r = 0.09, p = 0.15). Penile length was correlated with height (r = 0.18, p < 0.01), weight (r = 0.13, p < 0.01), and BMI (r = 0.21, p < 0.01).
Given the uneven distribution of fatherhood status across the extremes of age categories (6% of men <30 were fathers vs. 85% of men >60), the analyses were stratified on fatherhood status or limited to men between 30 and 60 years of age so that all assumptions of the regression models were met. On adjusted analyses stratified by fatherhood status, there was no relationship between AGDp and age. The lack of relationship remained after adjusting for fatherhood status and race among men 30–60 years of age (Table 3). No significant changes in the conclusions resulted from categorizing age. AGDa and PL also did not vary based on age.
Table 3. Mulivariable linear regression models exploring association between age and genital measures. Linear regression coefficient of age (β, 95% CI) is listed. Model a is adjusted for race and fatherhood. Model b is adjusted for race. Model c is adjusted for BMI, race, and fatherhood. Model d is adjusted for BMI and race
|Genital measure||Subjects||β (95% CI)Model|| p |
|AGDp||All||−0.11 (−0.30, 0.09)a||0.28|
|Childless||−0.14 (−0.31, 0.03)b||0.12|
|Fathers||−0.11 (−0.25, 0.02)b||0.10|
|AGDa||All||0.07 (−0.26, 0.40)c||0.27|
|Childless||−0.04 (−0.26, 0.19)d||0.75|
|Fathers||0.14 (−0.11, 0.40)d||0.68|
|PL||All||0.02 (−0.32, 0.35)d||0.92|
|Childless||0.13 (−0.15, 0.41)d||0.37|
|Fathers||−0.03 (−0.28, 0.21)d||0.79|
- Top of page
- Materials and methods
The current study demonstrates that anogenital distance, defined as the distance from the posterior scrotum to the anal verge is similar for men of different ages and suggests that AGD stays constant across a man's adult life. In addition, although other genital measures can vary based on body size, once linear growth is completed, the anogenital distance appears constant. AGDp was associated with fatherhood status and sperm count (data not shown) whereas AGDa was not, suggesting that AGDp may be a more reliable assessment of genital development.
An assessment of intrinsic fertility or testicular function often relies on a genital examination and a laboratory evaluation including a semen analysis, testosterone and FSH (Lipshultz et al., 2009). However, each currently used metric can be manipulated by exposure to gonadotoxic agents (e.g. DBCP, anabolic steroids). (Egnatz et al., 1980; Bonetti et al., 2008) Semen analysis can also be affected by abstinence period or recent exposure to elevated scrotal temperatures. Anthropomorphic measures may be more reliable as they should not be affected by exposures occurring after puberty. The current report suggests that AGD remains constant over a man's adult life and is not impacted by body habitus or age. As such, AGD may represent a constant assessment of a man's presumptive early life intrinsic testicular function.
During sexual development the immature genital precursors migrate ventrally via an androgen mediated pathway.(Larson, 1997) Human studies in infants have established that boys have longer perineal lengths than girls.(Salazar-Martinez et al., 2004; Torres-Sanchez et al., 2008; Thankamony et al., 2009; Sathyanarayana et al., 2010) Investigators have used the anogenital distance as a marker for normal genital development. Hsieh et al. demonstrated shorter anogenital distances in boys with genital anomalies (i.e. hypospadias and cryptorchidism), establishing a link between normal genital development and perineal length in humans.(Hsieh et al., 2008) Other studies in human have also established the relationship between adult AGD and testicular function as assessed by fertility, sperm production and testosterone production. (Eisenberg et al., 2012a; Eisenberg et al., 2011; Mendiola et al., 2011).
Rodent studies have also established a critical masculinization programming window where endocrine disruptors can permanently alter genital development, growth, and function.(Macleod et al., 2010; Welsh et al., 2010) Human studies have also established that exposure to endocrine disrupting agents can shorten male infant perineal lengths. (Swan et al., 2005) (Torres-Sanchez et al., 2008) Thus, current studies suggest that gestational exposures may play a critical role in male fertility and adult testicular function. The current report suggests that AGD remains a constant assessment of an adult man's intrinsic testicular function. However, any changes in AGD that occur during the transition from childhood to adulthood remain uncertain.
It is important to note that although two separate anogenital distance measures were made, only AGDp appeared associated with fatherhood. This is similar to the findings of Mendiola and colleagues from measurements collected on volunteers from Rochester, New York.(Mendiola et al., 2011) In our hands, AGDa was less comfortable for participants and also required simultaneous calibration between two points separated by over 100 cm which made it difficult to measure.
Certain limitations warrant mention. Rather than following individual men longitudinally, we measured multiple men of different ages. Although we did not observe any significant differences in AGD based on age, it is possible that if we followed individual men longitudinally over time, changes would become apparent. As a referral centre for men's health, it was not always possible to blind observers to the men's diagnoses or fatherhood status, which theoretically can lead to observer bias. In addition, only men referred to and evaluated in our clinic were eligible for enrolment; therefore, it is possible that our patient population does not represent all men. Nevertheless, our study represents the largest adult study of AGD and suggests that AGD is similar for men of different ages. Confirmation with longitudinal studies would help support the constancy of AGD over a man's adult life.