The Effects of Injected Testosterone Dose and Age on the Conversion of Testosterone to Estradiol and Dihydrotestosterone in Young and Older Men
Background: During testosterone (T) therapy, T is partly converted to 17β-estradiol (E2) and 5α-dihydrotestosterone (DHT). Effects of age, testosterone dose, and body composition on total and free E2 and DHT levels are unknown.
Objective: We evaluated age and dose-related differences in E2 and DHT levels in response to graded doses of testosterone enanthate in young and older men.
Methods: Fifty-one young (aged 19–35 yr) and 52 older (aged 59–75 yr) men completed treatment with monthly injections of a GnRH agonist plus randomly assigned weekly doses of testosterone enanthate (25, 50, 125, 300, or 600 mg) for 5 months.
Results: During testosterone administration, total and free E2 levels increased dose-dependently (dose effect, P < 0.001) in both young and older men. Total and free E2 levels and E2:T ratios during T administration were higher in older than young men, but age-related differences in free E2 and free E2:T ratios were not significant after adjusting for testosterone levels, percentage fat mass, and SHBG. DHT levels and DHT:T ratios were dose-related but did not differ between young and older men. Mechanistic modeling of free hormone data revealed that the conversions of T to E2 and DHT were both consistent with saturable Michaelis-Menten kinetics. The in vivo Km values were estimated to be 1.83 nm for aromatase and 3.35 nm for 5α-reductase, independent of age. The Vmax parameter for E2 was 40% higher in older men than younger men, but Vmax for DHT was not significantly different between age groups.
Conclusions: During im testosterone administration, E2 and DHT levels exhibit saturable increases with dose. The rate of whole body aromatization is higher in older men, partly related to their higher percentage fat mass, SHBG, and testosterone levels.
Testosterone, the major circulating androgen in men, serves both as a hormone and as a prohormone. Testosterone is converted in the body to two active metabolites, 17β-estradiol (E2) and 5α-dihydrotestosterone (DHT), which mediate some actions of testosterone in target tissues. Most of the circulating E2 is derived from peripheral aromatization of testosterone in adipose tissue, and DHT is derived by 5α-reduction of testosterone (1,2). It is generally believed that ratios of E2 to testosterone (E2:T) and DHT to testosterone (DHT:T) do not change during therapy when testosterone is administered parenterally by im injections of testosterone esters. Indeed, DHT and E2 concentrations are low in androgen-deficient men and increase in proportion to the increase in serum testosterone concentrations when these men are treated with testosterone esters (3,4,5,6,7). However, it is not known whether, in the presence of testosterone therapy, E2:T and DHT:T ratios in response are influenced by age and body composition. Serum E2 concentrations are higher in obese men than lean men (8,9,10,11,12), but the effects of body composition, age, and testosterone dose on E2:T ratio during testosterone therapy have not been fully investigated (13).
Using blood samples and data from our previously published testosterone dose-response studies in young and older men (14,15), we determined the effects of age, body composition, and testosterone dose on E2:T and DHT:T ratios. We found that the E2:T ratio, but not the DHT:T ratio, was higher in older men than young men during administration of graded doses of testosterone enanthate (TE) in men whose endogenous testosterone production had been suppressed by administration of a long-acting GnRH agonist. Because older men had higher body mass index (BMI) than young men, we investigated whether age effects on E2:T ratio during testosterone administration could be explained by differences in BMI in young and older men. We also examined free E2 and DHT concentrations during testosterone administration to determine whether conversion of testosterone to E2 and DHT in men conforms to the Michaelis-Menten kinetics and whether these conversions are affected by age.
Subjects and Methods
Blood samples were derived from healthy young men (aged 18–35 yr) and older men (aged 60–75 yr) with normal testosterone levels who were participants in a testosterone dose-response study. The design and main findings of this study have been reported (14,15). The study protocols were approved by the institutional review boards of Charles Drew University and Harbor-University of California Los Angeles Research and Education Institute. All participants provided informed consent.
Exclusion criteria included history of prostate cancer, prostate-specific antigen above 4 ng/ml, American Urological Association (AUA) lower urinary tract symptom score above 7, hematocrit above 48%, severe sleep apnea, diabetes mellitus, congestive heart failure, myocardial infarction in the preceding 6 months, or participation in moderate to intense exercise training. Men who had used androgens, GH, or any other anabolic therapy in the preceding year were excluded also.
After a 4-wk control period, participants received monthly injections of a GnRH agonist (leuprolide depot, 7.5 mg; TAP Pharmaceuticals, North Chicago, IL) to suppress endogenous testosterone production. Concomitantly, the participants were randomized to receive weekly im injections of TE (Delatestryl, 200 mg/ml; Savient Pharmaceuticals, Iselin, NJ) in one of five doses: 25, 50, 125, 300, or 600 mg (14,15). Treatment duration was 20 wk. Hormone levels were measured twice during the control period and every month thereafter during the treatment and recovery periods; blood was drawn before receiving the hormone injections 7 d after the previous TE injection. The Data and Safety Monitoring Board stopped the 600-mg TE dose group due to a number of serious adverse events in older men in this dose group. After this point, randomization was limited to one of four TE dose groups: 25, 50, 125, or 300 mg weekly.
Serum total testosterone levels were measured by a specific RIA that has been validated previously against liquid chromatography–tandem mass spectrometry (LC–MS/MS) (15) and published extensively (14,16,17). The intra- and interassay coefficients of variation for total testosterone assay were 8.2 and 13.2%, respectively, and sensitivity was 0.6 ng/dl. The RIA and LC–MS/MS methods were compared by analyzing samples prepared in charcoal stripped serum to which known amounts of testosterone had been added; additionally, serum samples from men were analyzed by both methods. These measurements demonstrated a correlation of 0.99 between the RIA and LC–MS/MS measurement. Serum SHBG levels were measured by an immunofluorometric assay that has a sensitivity of 6.25 nmol/liter. Serum E2 levels were measured by a previously described RIA that had a sensitivity of 2.5 pg/ml (ICN Pharmaceuticals, Costa Mesa, CA) (18). The cross-reactivity of testosterone, dehydroepiandrosterone, dehydroepiandrosterone sulfate, androstenedione, and progesterone in the estradiol assay was less than 0.1%. The intra- and interassay coefficients of variation for the estradiol assay were 5 and 12%, respectively. Dihydrotestosterone concentrations were measured after extraction of serum samples by hexane-ethyl acetate mixture followed by celite chromatography, as described (19). For measurement of DHT levels, serum samples were extracted and chromatographed through a celite column. DHT concentrations were measured in the eluate from the celite fractions by RIA (20) and corrected for recovery. The mean recovery was greater than 75%. The sensitivity of the DHT assay was 0.1 nmol/liter. Intra- and interassay coefficients of variation were ±9 and ±15%. Based on previous analysis (21), which showed that total and free testosterone levels were steady-state by treatment d 84, all hormones (testosterone, E2, and DHT) were measured on d 84 and 112 (7 d after the previous TE injection) and averaged. Free testosterone, free E2, and free DHT were calculated from the total concentrations of testosterone, E2, and DHT and the measured concentrations of albumin and SHBG using a novel spreadsheet (22) calculator based on the law of mass action (23).
Body composition assessment
Body composition was assessed at baseline and during wk 20 by dual-energy x-ray absorptiometry using a Hologic 4500 densitometer (Hologic, Inc., Waltham, MA). A body composition phantom was used to calibrate the machine before each measurement.
Graphical and tabular displays were used to assess normality of variables, the prevalence of outliers and influence points, and preliminary evidence of associations. The relationships between age, dose, and total E2 and DHT were evaluated using two-way ANOVA. Ratios of DHT and E2 to testosterone were assessed using the same approach. We conducted a step-wise regression analysis using age group, treatment testosterone levels, treatment SHBG levels, and treatment percentage fat mass to predict the metabolites of testosterone. These analyses were performed using SAS version 9.2 (SAS Institute, Cary, NC).
We employed two models to explore the relationship between testosterone and its metabolites. First, we used an empirical power law to assess the relationship of posttreatment circulating testosterone levels to E2 levels. We used exploratory plots stratified by age group, with scatterplot smoothing using generalized additive models (GAM) (24); these models apply semiparametric locally weighted smoothing to diagnose nonlinearities in associations. These results were used to motivate the fitting of parametric models of E2 on total testosterone and covariates using generalized linear models (GLM), assuming a γ distribution for outcomes and a log link function. The overall fit of the parametric models was measured by the Akaike Information Criterion (25)—a penalized likelihood-based statistic of model deviation from the data—for GLM and coefficient of determination (R2) for linear models. The statistical significance of individual regression coefficients was assessed using Wald tests. To assess the independent contributions of individual covariates to model fit while controlling for the influence of other variables, a forward stepwise fitting procedure was used. Results were considered statistically significant if null hypotheses of no association could be rejected at the 0.05 level. Analyses were performed using R (26) version 2.9.2 (R Foundation, Vienna, Austria).
We also analyzed the data using a mechanistic model in which the four relationships (i.e. total E2 vs. total testosterone, total DHT vs. total testosterone, free E2 vs. free testosterone, and free DHT vs. free testosterone) were modeled using rectangular hyperbolae, i.e. Y = A X/ (B + X). Four equations (with corresponding A and B values) were derived for young men and old men, respectively. Based on pharmacokinetic principles and Michaelis-Menten kinetics (see Appendix A, published on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org), the parameter A corresponds to the ratio of Vmax (the maximal rate of conversion of total or free testosterone to the metabolite) divided by MCR (the metabolic clearance rate of the total or free metabolite); the B parameter corresponds to the Km value for the enzymatic conversion of total or free testosterone to the metabolite. The model parameter estimation (including the se of the estimate, se, and R2 value) was performed using the 2D Michaelis-Menten equation curve fitting software available online at http://zunzun.com/Equation/2/BioScience/Michaelis-Menten/. Parameter estimates are derived using a downhill simplex method, with initial values derived from a genetic algorithm (personal correspondence from zunzun.com). The data points were equally weighted in the present analyses. Statistical comparison of the A and B parameters between young and old men were based on unpaired t tests. Because the results of the empiric power model and mechanistic model were similar, only the results of the mechanistic model are described in detail.
The characteristics of young and older men have been described (17,18). Fifty-two of 60 randomized older men and 54 of 61 randomized young men completed the study. For the current analysis, data were available on 51 young men who completed the study (11 in the 25-mg, 8 in the 50-mg, 11 in the 125-mg, 10 in the 300-mg, and 11 in the 600-mg group) and 52 older men who completed the study (12 in the 25-mg, 12 in the 50-mg, 11 in the 125-mg, 10 in the 300-mg, and 7 in the 600-mg group).
Baseline characteristics of the 51 young and 52 older men included in these analyses are shown in Table 11.. Young men had higher levels of total testosterone and free testosterone but lower levels of total E2 than older men. However, free E2 did not differ between the two groups. Total E2:total T and free E2:free T ratios were higher in older men than young men. Total and free DHT levels were higher in young men than older men. Total DHT:total T and free DHT:free T ratios were higher in young men than older men. Lean body mass did not differ between the two age groups, but older men had greater mean BMI, fat mass, and percentage fat mass than young men.
|Variable||Young men||Older men||P|
|Age (yr)||26.3 ± 4.5||65.6 ± 4.3|
|Total T (ng/dl)||581.7 ± 168.6||335.4 ± 93.8||<0.0001|
|Free T (pg/ml)||138.8 ± 49.5||53.7 ± 17.8||<0.0001|
|Total E2 (pg/ml)||20.8 ± 7.3||26.0 ± 5.3||<0.0001|
|Free E2 (pg/ml)||0.5 ± 0.2||0.5 ± 0.2||0.5|
|Total DHT (ng/dl)||54.0 ± 16.8||28.1 ± 14.6||<0.0001|
|Free DHT (pg/ml)||6.0 ± 2.42||1.9 ± 0.9||<0.0001|
|Total E2:total T ratio (%)||0.4 ± 0.1||0.8 ± 0.2||<0.0001|
|Total DHT:total T ratio (%)||9.5 ± 2.1||8.3 ± 3.5||0.05|
|Free E2:free T ratio (%)||0.4 ± 0.1||1.0 ± 0.2||<0.0001|
|Free DHT:free T ratio (%)||4.3 ± 1.0||3.5 ± 1.4||0.0017|
|SHBG (nm/liter)||31.4 ± 14.1||52.8 ± 23.8||<0.0001|
|BMI (kg/m2)||24.19 ± 3.0||26.8 ± 3.6||0.0002|
|Lean body mass (kg) (%)||57.7 ± 7.3||57.8 ± 6.4||0.94|
|Percentage fat mass (%)||18.6 ± 5.4||26.4 ± 5.5||<0.0001|
|Fat mass (kg)||14.4 ± 5.7||22.3 ± 7.4||<0.0001|
|Height (cm)||176.5 ± 5.7||176.2 ± 5.7||0.82|
Data are expressed as mean ± sd. To convert total testosterone to nanomoles per liter, multiply by 0.03467; to convert free testosterone to picomoles per liter, multiply by 3.467. To convert E2 concentrations to picomoles per liter, multiply by 3.671. To convert total DHT to nanomoles per liter, multiply by 0.0344; to convert free DHT to picomoles per liter, multiply by 3.44.
Exploratory analyses by age group
Untransformed total testosterone was positively associated with total E2 (r = 0.47; P < 0.01) and free E2 (r = 0.36; P < 0.01) in young but not older men. Total testosterone was also correlated with total DHT (r = 0.75; P < 0.01) and free DHT (r = 0.53; P < 0.01) in young men but only with total DHT in older men (r = 0.51, P < 0.01). Baseline total E2, free E2, total E2:T ratio, and free E2:T ratio were not significantly associated with body composition measures such as body weight, BMI, lean body mass, fat mass, or percentage fat mass. Free E2 was significantly inversely correlated with SHBG in young (r = −0.47; P < 0.001) and older (r = −0.55; P < 0.001) men.
Response to treatment (Fig. 11)
Results of two-way ANOVA
As reported previously (15), both total testosterone and free testosterone concentrations increased with TE dose and were significantly higher in older men than young men in the 125-, 300-, and 600-mg dose groups (P < 0.05). In the current analysis, total E2 increased in a dose-dependent manner in both young and older men (dose effect, P < 0.0001). Additionally, during testosterone administration, older men had higher total E2 levels than young men (age effect, P < 0.0001) with significant differences in the 50-, 125-, 300-, and 600-mg dose groups (P < 0.05). Similarly, free E2 levels increased dose dependently (dose effect, P = 0.001) and were higher in older men (age effect, P < 0.0001) than young men with significant differences in the 125-, 300-, and 600-mg dose groups (P < 0.05).
Total E2:total T and free E2:free T ratios decreased dose dependently in both young and older men (dose effect, P < 0.0001 for each). Older men had higher total E2:total T and higher free E2:free T ratios than young men (age effects, P = 0.0003 and <0.001, respectively).
Total and free DHT levels increased dose dependently in both age groups (dose effect, P < 0.0001 for both total and free DHT); however, total and free DHT levels during treatment did not differ significantly between the two groups. Similarly, total DHT:total T and free DHT:free T ratios decreased dose dependently in young and older men (dose effect, P = 0.0004 and P = 0.009, respectively), but without a statistically significant difference between the two age groups.
Results of step-wise regression
Because E2 levels might be affected by testosterone levels, percentage fat mass, and SHBG levels that differed in young and older men, we compared E2 levels after adjusting for percentage fat mass and SHBG, both individually and in combination, in addition to treatment testosterone levels. Differences between old and younger men in total E2 [46.6 vs. 28.9 pg/ml; P (age) < 0.0001] and free E2 [1.4 vs. 0.9 pg/ml; P (age) < 0.001] as well as total E2:T ratios [0.8 vs. 0.5; P (age) = 0.02] and free E2:T ratios [1.0 vs. 0.6; P (age) = 0.03] persisted after adjusting for SHBG alone. After controlling for percentage fat mass alone, older men had significantly higher total E2 levels [43.1 vs. 32.7 pg/ml; P (age) < 0.001], and higher total E2:T ratios [0.8 vs. 0.6; P (age) = 0.04] and free E2:T ratios [1.0 vs. 0.6; P (age) = 0.01] than young men; however, free E2 levels were not significantly different between the two age groups [1.2 vs. 1.0 pg/ml; P (age) = 0.24]. Finally, in the model containing treatment testosterone levels, percentage fat mass, and SHBG levels, older men continued to have significantly higher total E2 [45.8 vs. 30 pg/ml; P (age) = 0.001] and total E2:T ratios (0.8 vs. 0.5; P = 0.04) but free E2 levels [1.3 vs. 1.1 pg/ml; P (age) = 0.24] and free E2:T ratios [1.0 vs. 0.6; P (age) = 0.6] did not differ significantly between the young and older men.
Modeling the relationships between metabolite and testosterone levels
Plotting of E2 vs. total testosterone during treatment using GAM indicated a curvilinear relationship (concave downward), so that cross-sectional increases in total testosterone were associated with increases in E2 of proportionately lesser magnitude. Plots of transformed values (data not shown) indicated a linear association between the logarithm (log) of E2 and the log of total testosterone. These observations motivated fitting a GLM specifying that the log of average E2 is a linear function of the log of total testosterone, with E2 values assumed to follow a γ distribution. This model is appropriate for data where the variance increases with the mean (27), as it did for E2 values, and accounts for the curvilinear relation of E2 to total testosterone. Total testosterone, age group, and percentage fat mass exhibited statistically significant associations with E2, controlling for the other factors. Conditional on percentage fat mass and age group, a cross-sectional doubling of total testosterone was associated with an approximate 40% increase in mean E2 concentrations (Table 22).). For a given total testosterone and percentage fat, older subjects had approximately 37% greater E2 concentrations than their younger counterparts. Comparison of this model with candidate submodels via Akaike Information Criterion indicates preference for the full model.
|Total testosterone (100% increase)b||1.40||1.35, 1.46|
|Age group (older vs. young men)||1.37||1.19, 1.57|
|Fat mass (10%)c||1.02||0.97, 1.06|
CI, Confidence interval.
Using the combined baseline and on-treatment data for all treatment groups, we graphically explored and mathematically modeled the four relationships (total E2 vs. total testosterone, total DHT vs. total testosterone, free E2 vs. free testosterone, and free DHT vs. free T) using separate analyses for young and older men. Figures 22,, A–D, and 3, A–D, show the data and the fitted curves, corresponding to the total and free hormone concentrations, respectively. In all cases, the combined data exhibited curvilinear relationships that were well described by a rectangular hyperbolae, Y = A X/ (B + X), consistent with a saturable conversion of testosterone to metabolite governed by Michaelis-Menten kinetics (Appendix A). Table 33 lists the parameter values estimated from the curve fitting.
|Young men (n = 50, 50)|
|Older men (n = 50, 50)|
|DHT vs. TT|
|Young men (n = 48, 47)|
|Older men (n = 48, 47)|
|FE vs. FT|
|Young men (n = 49, 50)|
|Older men (n = 51, 50)|
|FDHT vs. FT|
|Young men (n = 48, 47)|
|Older men (n = 48, 47)|
Data are expressed as nanograms per deciliter, unless otherwise specified. ″n″ values correspond to baseline and on-treatment data points, respectively. P values correspond to the comparison of young and older men. We analyzed the data using a mechanistic model in which the four relationships (i.e. total E2 vs. total T, free E2 vs. free T, total DHT vs. total T, and free DHT vs. free T) were modeled using rectangular hyperbolae, i.e. Y = A X/ (B + X). Four equations (with corresponding A and B values) were derived for young men and old men, respectively. Based on pharmacokinetic principles and Michaelis-Menten kinetics (see Appendix A), the parameter A corresponds to the ratio of Vmax (maximal rate of conversion of total or free T to the metabolite) divided by MCR (the metabolic clearance rate of the total or free metabolite); the B parameter corresponds to the Km value for the enzymatic conversion of total or free T to the metabolite. The model parameter estimation (including the se of the estimate, and R2 value) was performed using the 2D Michaelis-Menten equation curve fitting software available online at http://zunzun.com/Equation/2/BioScience/Michaelis-Menten/. TT, Total testosterone; FE, free E2; FT, free testosterone; FDHT, free DHT.
The total and free E2 vs. testosterone curves in older men (Figs. 2B2B and 3B3B)) had A values (Table 33)) that were approximately 30–40% higher than in young men (Figs. 2A2A and 3A3A),), consistent with a higher maximal rate of whole body aromatization in older men. In contrast, the B values were not significantly different between age groups (Table 33).). Based on the B values from the free E2 vs. free testosterone models (Appendix A), the estimated Michaelis-Menten Km, E2 values for whole body aromatization were comparable in the two age groups [2.12 ± 0.34 nm in young men and 1.56 ± 0.18 nm in older men; P value, nonsignificant (NS)]. The mean value of 1.83 nm characterizes the combined result.
The total and free DHT vs. testosterone curves (Figs. 22,, C and D, and 33,, C and D) exhibited more complex behavior between age groups but were nevertheless consistent with the general predictions given in Appendix A. Based on the total hormone concentrations, the parameters A and B were both significantly smaller for young men than older men. In contrast, the parameters A and B derived from the free hormone concentrations were not significantly different between the age groups (Table 33).). The former observation is consistent with the lower SHBG levels in young men compared with the older men (Table 11),), which would be expected to reduce the A and B values. The latter observation indicates that the maximal whole body production rate of DHT from testosterone is not affected by age and that the Km, DHT value of the 5α-reductase pathway is also the same in young and older men. Based on the B values from free DHT vs. free testosterone models (Table 33),), the estimated Km, DHT values were 2.71 ± 0.53 nm in young men and 3.98 ± 0.82 nm in older men (P value, NS). The mean value of 3.35 nm characterizes the combined result.
We found that during administration of graded doses of TE, the older men had higher total and free E2 concentrations and higher total and free E2:T ratios than young men after adjusting for serum testosterone levels. In contrast to E2 levels, serum DHT levels and DHT:T ratios did not differ between young and older men. Regression modeling revealed that percentage fat mass independently and partly explained the difference in free E2 levels between young and older men, and jointly with SHBG explained most of the differences in free E2 levels as well as free E2:T ratios in the two groups. There also was an apparent effect of testosterone dose on metabolite to testosterone ratios; higher doses of testosterone were associated with seemingly lower metabolite to testosterone ratios in both young and older men. Mechanistic modeling indicated that the whole body conversion of testosterone to E2 and DHT was consistent with saturable Michaelis-Menten kinetics.
In men in whom endogenous testicular steroidogenesis has been suppressed by administration of GnRH agonist, the steady-state circulating E2 concentrations during testosterone administration should reflect the ratio of E2 production from peripheral aromatization of testosterone and plasma E2 clearance (as discussed in Appendix A). Although differences in SHBG levels may contribute to altered clearance rates, the observation that the free E2 levels are also higher in older men is consistent with an increase in whole body aromatization.
Modeling the free E2 vs. free testosterone and free DHT vs. free testosterone relationships showed that our data are consistent with Michaelis-Menten enzyme kinetics in both cases. From this analysis, we estimated the Km, E2 value for aromatization to be approximately 1.83 nm; whereas for the 5α-reductase pathway the Km, DHT value was approximately 3.35 nm. The maximum free testosterone levels in our study only reached two to three times these values, so that full saturation had not been attained. Within the range of testosterone concentrations observed in this experiment, our GAM and GLM analyses did not find aromatization of testosterone to be dose-limited. However, the data conformed to Michaelis-Menten kinetic model, and we cannot exclude the possibility that the aromatization of testosterone may be saturable at testosterone concentrations that are higher than those observed in our study.
In vitro estimates of Km, E2 in human cell lines containing aromatase range from 6.7 to 68 nm using androstenedione as a substrate (28,29,30,31). There are limited data using testosterone as a substrate [Km of 41 nm (31)], and it is conceivable that the greater hydrophobicity of testosterone, as reflected in albumin binding (32) would be associated with a lower Km, E2 than for androstenedione. This might explain the apparent discrepancy between our in vivo estimate of Km, E2 (1.83 nm) and the in vitro values for androstenedione.
Two isoforms of 5α-reductase have been found in human tissues with widely divergent values of Km, DHT estimated in vitro. The type I isoenzyme has a Km, DHT value on the order of 2000 nm, whereas the type 2 isoenzyme has a Km, DHT value on the order of 12 nm (33,34). The type 2 isoenzyme is thought to be the more prevalent form in humans (35) and is more consistent with our in vivo estimate of Km, DHT (3.35 nm). In rat tissue, the type 1 and type 2 isoenzymes have Km, DHT values of 2300 and 0.8 nm, respectively (36); the type 2 value again is consistent with our in vivo estimate.
Higher levels of total and free E2 in older men could potentially be due to age-related differences in body composition, circulating testosterone levels, and SHBG levels. Because adipose tissue is an important site of testosterone to E2 conversion in adult men, older men might be expected to have higher E2 levels because of higher fat mass. The proinflammatory state associated with adiposity and aging that is believed to promote aromatase activity via IL-6 might be one contributory factor (37). Although the age-related differences in total E2 levels and total E2:T ratios during testosterone therapy persisted even after adjusting for the higher percentage fat mass and SHBG levels in older men, the differences were attenuated after adjustment for these covariates. Furthermore, free E2 levels and free E2:T ratios were not significantly different between the two groups, after adjusting for fat mass, testosterone levels, and SHBG levels. Thus, age-related differences in E2 levels and E2:T ratios are at least partly related to differences in body composition between the young and older men and partly to differences in testosterone and SHBG levels.
Anecdotally, high doses of testosterone and other androgens have been reported to be associated with gynecomastia. Our data indicate that the occurrence of gynecomastia with high doses of testosterone cannot be explained on the basis of high estradiol to testosterone ratios; however, the high levels of E2 attained at high doses may be sufficient to account for it.
The clinical consequences of higher E2 levels and higher E2:T ratios in older men remain poorly understood. In longitudinal studies in older men, higher E2 levels have been associated with adverse outcomes such as stroke (38) and cognitive decline (39). In cross-sectional studies, higher E2 levels have been associated with the increased risk of metabolic syndrome (40) and type 2 diabetes mellitus. Very low E2 levels have been associated with bone loss and fracture risk (41,42,43), insulin resistance and premature atherosclerosis (44), and higher risk of mortality in elderly men (45). Hence, too much as well as too little estrogen may be detrimental to the health in men. Further research is needed to elucidate the role of estrogen in men’s health, including defining thresholds beyond which E2 levels might be detrimental.
This work was supported by National Institute on Aging Grants 1 UO1AG14369 and 1RO1AG31206, and by Boston Claude D. Pepper Older Americans Independence Center Grant 5P30 AG31679.
Disclosure Summary: The authors have nothing to disclose.
First Published Online June 9, 2010
Abbreviations: BMI, Body mass index; DHT, 5α-dihydrotestosterone; E2, 17β-estradiol; GAM, generalized additive models; GLM, generalized linear models; LC–MS/MS, liquid chromatography–tandem mass spectrometry; log, logarithm; NS, nonsignificant; TE, testosterone enanthate.
- Simpson ER 2003 Sources of estrogen and their importance. J Steroid Biochem Mol Biol 86:225–230 [PubMed] [Google Scholar]
- Longcope C, Kato T, Horton R 1969 Conversion of blood androgens to estrogens in normal adult men and women. J Clin Invest 48:2191–2201 [PMC free article] [PubMed] [Google Scholar]
- Meikle AW, Mazer NA, Moellmer JF, Stringham JD, Tolman KG, Sanders SW, Odell WD 1992 Enhanced transdermal delivery of testosterone across nonscrotal skin produces physiological concentrations of testosterone and its metabolites in hypogonadal men. J Clin Endocrinol Metab 74:623–628 [PubMed] [Google Scholar]
- Schubert M, Minnemann T, Hübler D, Rouskova D, Christoph A, Oettel M, Ernst M, Mellinger U, Krone W, Jockenhövel F 2004 Intramuscular testosterone undecanoate: pharmacokinetic aspects of a novel testosterone formulation during long-term treatment of men with hypogonadism. J Clin Endocrinol Metab 89:5429–5434 [PubMed] [Google Scholar]
- von Eckardstein S, Nieschlag E 2002 Treatment of male hypogonadism with testosterone undecanoate injected at extended intervals of 12 weeks: a phase II study. J Androl 23:419–425 [PubMed] [Google Scholar]
- Behre HM, von Eckardstein S, Kliesch S, Nieschlag E 1999 Long-term substitution therapy of hypogonadal men with transscrotal testosterone over 7–10 years. Clin Endocrinol (Oxf) 50:629–635 [PubMed] [Google Scholar]
- Swerdloff RS, Wang C, Cunningham G, Dobs A, Iranmanesh A, Matsumoto AM, Snyder PJ, Weber T, Longstreth J, Berman N 2000 Long-term pharmacokinetics of transdermal testosterone gel in hypogonadal men. J Clin Endocrinol Metab 85:4500–4510 [PubMed] [Google Scholar]
- Muller M, den Tonkelaar I, Thijssen JH, Grobbee DE, van der Schouw YT 2003 Endogenous sex hormones in men aged 40–80 years. Eur J Endocrinol 149:583–589 [PubMed] [Google Scholar]
- Orwoll E, Lambert LC, Marshall LM, Phipps K, Blank J, Barrett-Connor E, Cauley J, Ensrud K, Cummings S, for the Osteoporotic Fractures in Men Study G 2006 Testosterone and estradiol among older men. J Clin Endocrinol Metab 91:1336–1344 [PubMed] 
- Bjørnerem A, Straume B, Midtby M, Fønnebø V, Sundsfjord J, Svartberg J, Acharya G, Oian P, Berntsen GK 2004 Endogenous sex hormones in relation to age, sex, lifestyle factors, and chronic diseases in a general population: the Tromso study. J Clin Endocrinol Metab 89:6039–6047 [PubMed] [Google Scholar]
- Schneider G, Kirschner MA, Berkowitz R, Ertel NH 1979 Increased estrogen production in obese men. J Clin Endocrinol Metab 48:633–638 [PubMed] [Google Scholar]
- Vermeulen A, Kaufman JM, Goemaere S, van Pottelberg I 2002 Estradiol in elderly men. Aging Male 5:98–102 [PubMed] 
- Dobs AS, Bachorik PS, Arver S, Meikle AW, Sanders SW, Caramelli KE, Mazer NA 2001 Interrelationships among lipoprotein levels, sex hormones, anthropometric parameters, and age in hypogonadal men treated for 1 year with a permeation-enhanced testosterone transdermal system. J Clin Endocrinol Metab 86:1026–1033 [PubMed] [Google Scholar]
- Bhasin S, Woodhouse L, Casaburi R, Singh AB, Bhasin D, Berman N, Chen X, Yarasheski KE, Magliano L, Dzekov C, Dzekov J, Bross R, Phillips J, Sinha-Hikim I, Shen R, Storer TW 2001 Testosterone dose-response relationships in healthy young men. Am J Physiol Endocrinol Metab 281:E1172–E1181 [PubMed] [Google Scholar]
- Bhasin S, Woodhouse L, Casaburi R, Singh AB, Mac RP, Lee M, Yarasheski KE, Sinha-Hikim I, Dzekov C, Dzekov J, Magliano L, Storer TW 2005 Older men are as responsive as young men to the anabolic effects of graded doses of testosterone on the skeletal muscle. J Clin Endocrinol Metab 90:678–688 [PubMed] [Google Scholar]
- Bhasin S, Storer TW, Berman N, Callegari C, Clevenger B, Phillips J, Bunnell TJ, Tricker R, Shirazi A, Casaburi R 1996 The effects of supraphysiologic doses of testosterone on muscle size and strength in normal men. N Engl J Med 335:1–7 [PubMed] [Google Scholar]
- Bhasin S, Storer TW, Javanbakht M, Berman N, Yarasheski KE, Phillips J, Dike M, Sinha-Hikim I, Shen R, Hays RD, Beall G 2000 Testosterone replacement and resistance exercise in HIV-infected men with weight loss and low testosterone levels. JAMA 283:763–770 [PMC free article] [PubMed] [Google Scholar]
- Singh AB, Hsia S, Alaupovic P, Sinha-Hikim I, Woodhouse L, Buchanan TA, Shen R, Bross R, Berman N, Bhasin S 2002 The effects of varying doses of T on insulin sensitivity, plasma lipids, apolipoproteins, and C-reactive protein in healthy young men. J Clin Endocrinol Metab 87:136–143 [PubMed] [Google Scholar]
- Bhasin S, Swerdloff RS, Steiner B, Peterson MA, Meridores T, Galmirini M, Pandian MR, Goldberg R, Berman N 1992 A biodegradable testosterone microcapsule formulation provides uniform eugonadal levels of testosterone for 10–11 weeks in hypogonadal men. J Clin Endocrinol Metab 74:75–83 [PubMed] [Google Scholar]
- Rao PN, Khan AH, Moore Jr PH 1977 Synthesis of new steroid haptens for radioimmunoassay. Part III. 15β-Carboxyethylmercaptosteroid-bovine serum albumin conjugates. Specific antisera for radioimmunoassay of 5α-dihydrotestosterone, 5α-androstane-3β, 17β-diol and 5α-androstane-3α, 17β-diol. Steroids 29:171–184 [PubMed] [Google Scholar]
- Coviello AD, Lakshman K, Mazer NA, Bhasin S 2006 Differences in the apparent metabolic clearance rate of testosterone in young and older men with gonadotropin suppression receiving graded doses of testosterone. J Clin Endocrinol Metab 91:4669–4675 [PubMed] [Google Scholar]
- Mazer NA 2009 A novel spreadsheet method for calculating the free serum concentrations of testosterone, dihydrotestosterone, estradiol, estrone and cortisol: with illustrative examples from male and female populations. Steroids 74:512–519 [PubMed] 
- Vermeulen A, Verdonck L, Kaufman JM 1999 A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 84:3666–3672 [PubMed] [Google Scholar]
- Wood S 2006 Generalized additive models: an introduction with R. 1st ed. Boca Raton, FL: Chapman, Hall/CRC [Google Scholar]
- Akaike H 1974 A new look at the statistical model identification. IEEE Trans Automat Control 19:716–723 [Google Scholar]
- Development Core Team 2008 R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing [Google Scholar]
- McCullagh P, Nelder JA 1989 Generalized linear models. 2nd ed. Boca Raton, FL: Chapman, Hall/CRC [Google Scholar]
- Berkovitz GD, Brown TR, Fujimoto M 1987 Aromatase activity in human skin fibroblasts grown in cell culture. Steroids 50:281–295 [PubMed] [Google Scholar]
- Ma CX, Adjei AA, Salavaggione OE, Coronel J, Pelleymounter L, Wang L, Eckloff BW, Schaid D, Wieben ED, Adjei AA, Weinshilboum RM 2005 Human aromatase: gene resequencing and functional genomics. Cancer Res 65:11071–11082 [PubMed] [Google Scholar]
- Vinggaard AM, Hnida C, Breinholt V, Larsen JC 2000 Screening of selected pesticides for inhibition of CYP19 aromatase activity in vitro. Toxicol in Vitro 14:227–234 [PubMed] [Google Scholar]
- Yoshida N, Osawa Y 1991 Purification of human placental aromatase cytochrome P-450 with monoclonal antibody and its characterization. Biochemistry 30:3003–3010 [PubMed] [Google Scholar]
- Amundsen LK, Sirén H 2007 Determination of association constants between steroid compounds and albumins by partial-filling ACE. Electrophoresis 28:3737–3744 [PubMed] [Google Scholar]
- Span PN, Sweep CG, Benraad TJ, Smals AG 1996 3α-Hydroxysteroid oxidoreductase activities in dihydrotestosterone degradation and back-formation in rat prostate and epididymis. J Steroid Biochem Mol Biol 58:319–324 [PubMed] [Google Scholar]
- Tunn S, Hochstrate H, Grunwald I, Flüchter SH, Krieg M 1988 Effect of aging on kinetic parameters of 5α-reductase in epithelium and stroma of normal and hyperplastic human prostate. J Clin Endocrinol Metab 67:979–985 [PubMed] [Google Scholar]
- Keam SJ, Scott LJ 2008 Dutasteride: a review of its use in the management of prostate disorders. Drugs 68:463–485 [PubMed] [Google Scholar]
- Stuart JD, Lee FW, Simpson Noel D, Kadwell SH, Overton LK, Hoffman CR, Kost TA, Tippin TK, Yeager RL, Batchelor KW, Bramson HN 2001 Pharmacokinetic parameters and mechanisms of inhibition of rat type 1 and 2 steroid 5α-reductases: determinants for different in vivo activities of GI198745 and finasteride in the rat. Biochem Pharmacol 62:933–942 [PubMed] [Google Scholar]
- Maggio M, Ceda GP, Lauretani F, Bandinelli S, Metter EJ, Artoni A, Gatti E, Ruggiero C, Guralnik JM, Valenti G, Ling SM, Basaria S, Ferrucci L 2009 Estradiol and inflammatory markers in older men. J Clin Endocrinol Metab 94:518–522 [PMC free article] [PubMed] 
- Abbott RD, Launer LJ, Rodriguez BL, Ross GW, Wilson PW, Masaki KH, Strozyk D, Curb JD, Yano K, Popper JS, Petrovitch H 2007 Serum estradiol and risk of stroke in elderly men. Neurology 68:563–568 [PubMed] 
- Muller M, van den Beld AW, Grobbee DE, de Jong FH, Lamberts SWJ 2009 Sex hormones and cognitive decline in elderly men. Psychoneuroendocrinology 34:27–31 [PubMed] [Google Scholar]
- Maggio M, Lauretani F, Ceda GP, Bandinelli S, Basaria S, Paolisso G, Giumelli C, Luci M, Najjar SS, Metter EJ, Valenti G, Guralnik J, Ferrucci L 2010 Estradiol and metabolic syndrome in older Italian men: the InCHIANTI study. J Androl 31:155–162 [PMC free article] [PubMed] 
- Mellström D, Johnell O, Ljunggren O, Eriksson AL, Lorentzon M, Mallmin H, Holmberg A, Redlund-Johnell I, Orwoll E, Ohlsson C 2006 Free testosterone is an independent predictor of BMD and prevalent fractures in elderly men: MrOS Sweden. J Bone Miner Res 21:529–535 [PubMed] [Google Scholar]
- Mellström D, Vandenput L, Mallmin H, Holmberg AH, Lorentzon M, Odén A, Johansson H, Orwoll ES, Labrie F, Karlsson MK, Ljunggren O, Ohlsson C 2008 Older men with low serum estradiol and high serum SHBG have an increased risk of fractures. J Bone Miner Res 23:1552–1560 [PubMed] 
- Khosla S, Melton 3rd LJ, Atkinson EJ, O’Fallon WM 2001 Relationship of serum sex steroid levels to longitudinal changes in bone density in young versus elderly men. J Clin Endocrinol Metab 86:3555–3561 [PubMed] [Google Scholar]
- Maffei L, Murata Y, Rochira V, Tubert G, Aranda C, Vazquez M, Clyne CD, Davis S, Simpson ER, Carani C 2004 Dysmetabolic syndrome in a man with a novel mutation of the aromatase gene: effects of testosterone, alendronate, and estradiol treatment. J Clin Endocrinol Metab 89:61–70 [PubMed] 
- Tivesten A, Vandenput L, Labrie F, Karlsson MK, Ljunggren O, Mellström D, Ohlsson C 2009 Low serum testosterone and estradiol predict mortality in elderly men. J Clin Endocrinol Metab 94:2482–2488 [PubMed]