0021-972X/01/$03.00/0
The Journal of Clinical Endocrinology & Metabolism
Copyright © 2001 by The Endocrine Society
Vol. 86, No. 3
Printed in U.S.A.
Interrelationships among Lipoprotein Levels, Sex
Hormones, Anthropometric Parameters, and Age in
Hypogonadal Men Treated for 1 Year with a PermeationEnhanced Testosterone Transdermal System*
ADRIAN S. DOBS, PAUL S. BACHORIK, STEFAN ARVER, A. WAYNE MEIKLE,
STEVEN W. SANDERS, KIM E. CARAMELLI, AND NORMAN A. MAZER
The Johns Hopkins Medical Center (A.S.D., P.S.B.), Baltimore, Maryland 21287; Departments of
Medicine (A.W.M.) and Pharmaceutics (N.A.M.), University of Utah, Salt Lake City, Utah 84132;
Karolinska Hospital (S.A.), Stockholm, Sweden; and Watson Laboratories, Inc. (S.W.S., K.E.C.,
N.A.M.), Salt Lake City, Utah 84108
ABSTRACT
Serum lipoproteins and cardiovascular risk are affected by endogenous and exogenous sex hormones. As part of a multicenter evaluation of a permeation-enhanced testosterone transdermal system
(TTD), the interrelationships among serum lipoproteins, hormone
levels, anthropometric parameters, and age were investigated in 29
hypogonadal men.
Subjects (aged 21– 65 yr) were first studied during prior treatment
with im testosterone esters (IM-T), then during an 8-week period of
androgen withdrawal resulting in a hypogonadal state (HG), and
finally during a 1-yr treatment period with the TTD. Compared with
treatment with IM-T, the HG period produced increases in high density lipoprotein [HDL; 12.0 ⫾ 1.6% (⫾SEM); P ⬍ 0.001] and total
cholesterol (4.2 ⫾ 1.9%; P ⫽ 0.02) and a decrease in the cholesterol/
HDL ratio (⫺9.7 ⫾ 2.8%; P ⫽ 0.02). Compared with the HG period,
TTD treatment produced decreases in HDL (-7.6 ⫾ 2.5%; P ⫽ 0.002)
and increases in the cholesterol/HDL ratio (9.0 ⫾ 2.5%; P ⫽ 0.01) and
triglycerides (20.7 ⫾ 6.4%; P ⫽ 0.03). Small decreases in total cho-
C
ARDIOVASCULAR DISEASE is the leading cause of
mortality in the United States (1). Its increased prevalence among men has been attributed to lower serum high
density lipoprotein (HDL) cholesterol levels compared with
those in premenopausal women (2– 4). The different concentrations of sex steroids in men and women are thought to be
important factors contributing to the gender difference in
lipoprotein profiles (5).
In women, the levels of estrogenic hormones, whether
present endogenously or given exogenously, have been
shown to correlate with increasing HDL levels, resulting in
Received May 12, 2000. Revision received November 6, 2000. Accepted November 20, 2000.
Address all correspondence and requests for reprints to: Adrian S.
Dobs, M.D., M.H.S., Department of Medicine, The Johns Hopkins University, 1830 East Monument Street, Room 328, Baltimore, Maryland
21205. E-mail: [email protected].
* This work was supported by grants from Watson Laboratories, Inc.
(formerly TheraTech, Inc.); NIH General Clinical Research Center Grants
5-M01-RR-00722, RR-0003524, and M01-RR-00064; USPHS Grants DK45760 and DK-43344; and the Swedish Medical Research Council (MFR
11615 and 400/96). Portions of this manuscript were presented at the
76th Annual Meeting of The Endocrine Society, Anaheim, California,
1994.
lesterol (⫺1.2 ⫾ 1.8%; P ⫽ 0.1) and low density lipoprotein (⫺0.8 ⫾
2.6%; P ⫽ 0.07) were also observed during TTD, but did not reach
statistical significance. Likewise, there were no significant differences between the IM-T and TTD treatments. Serum HDL levels
showed a strong negative correlation with body mass index and other
obesity parameters in all three study periods (r ⬍ ⫺0.45; P ⬍ 0.02).
During treatment with TTD, serum testosterone levels also correlated
negatively with body mass index (r ⫽ ⫺0.621; P ⬍ 0.001). As a
consequence of these relationships, a positive trend was observed
between HDL and testosterone levels during TTD treatment (r ⫽
0.336; P ⫽ 0.07). Interestingly, the changes in lipoprotein levels during TTD treatment indicated a more favorable profile (decrease in
cholesterol and low density lipoprotein levels) with increasing age of
the patients.
In hypogonadal men the effects of transdermal testosterone replacement on serum lipoproteins appear consistent with the physiological effects of testosterone in eugonadal men. (J Clin Endocrinol
Metab 86: 1026 –1033, 2001)
a lower cardiovascular risk profile (6). In addition, estrogens
have been shown to produce vasodilating effects that are
beneficial to cardiovascular function (7).
In men, however, testosterone levels have a more complicated and controversial relationship to HDL levels and
cardiovascular risk. During puberty, rising endogenous testosterone levels are associated with a fall in serum HDL (8).
HDL levels also decrease with administration of exogenous
androgens to healthy young men (9), athletes (10), and hypogonadal patients (11), although in elderly men the effect of
testosterone administration on HDL levels appears to be
much weaker (12–15). Conversely, experimental hypogonadism in men, induced by the administration of a GnRH agonist, results in increased serum HDL levels, implying that
androgen levels in the normal adult male range have a suppressive effect on HDL levels (16). However, several crosssectional studies of adult men (17–20), including patients
with coronary artery disease (21), have shown that higher
testosterone levels are associated with higher HDL concentrations. Numerous factors could account for the apparent
positive correlation between testosterone and HDL in men,
including obesity, fat distribution, diet, age, alcohol intake,
exercise, and smoking (17, 22).
1026
TESTOSTERONE AND LIPOPROTEINS
As part of a multicenter phase III evaluation of the efficacy
and safety of Androderm, a permeation-enhanced testosterone transdermal system (TTD), we investigated the interrelationships among serum lipoproteins, sex hormone levels,
anthropometric parameters, and age in 29 hypogonadal men.
Longitudinal evaluations of these parameters were first
made during prior treatment with im testosterone esters
(IM-T), then during an 8-week period of androgen withdrawal resulting in a hypogonadal state (HG), and finally
during a 1-yr treatment period with TTD. By studying a
population of hypogonadal men during periods of androgen
replacement and withdrawal, we were able to assess both the
effects of exogenous testosterone administration on serum
lipoproteins as well as the cross-sectional relationship between lipoproteins and sex hormones in the same
individuals.
Other aspects of this multicenter study, including pharmacokinetics, sexual function, prostate evaluations, and
overall clinical outcomes have been published previously
(23–26).
Subjects and Methods
Patient population
Hypogonadal males, 20 – 65 yr of age, who had been receiving androgen replacement therapy (primarily testosterone ester injections) for
at least 3 months were eligible for this study. Hypogonadism was defined as a pretreatment serum testosterone of 8.7 nmol/L or less and was
confirmed during the androgen withdrawal period. Patients were excluded from the study for serum prostate-specific antigen levels greater
than 3.9 ␮g/L, prostate volume greater than 30 cm3 on transrectal ultrasound examination, postvoid residual volume greater than 60 cm3,
unstable or untreated endocrine disorders, poorly controlled diabetes,
psychiatric disturbances, or use of tricyclic antidepressants or any drugs
with antiandrogenic properties. Patients were also excluded for severe
hyperlipidemia, defined as two or more of the following abnormalities:
greater than 95th percentile (for age) of total cholesterol, low density
lipoprotein (LDL), or tryglycerides or less than 5th percentile of HDL,
based on normative data from the Lipid Research Clinics Program (27).
The study protocol was approved by the institutional review boards
of the participating institutions (center 1, University of Utah School of
Medicine, Salt Lake City, UT; center 2, Karolinska Hospital, Stockholm,
Sweden; and center 3, Johns Hopkins University, Baltimore, MD), and
written informed consent was obtained from all patients.
Study design
This was an open label, multicenter study with four consecutive
evaluation periods (26). In the first period (denoted IM-T), patients were
monitored for 21 days after receiving a final testosterone ester injection.
The second period (denoted HG) was an 8-week androgen withdrawal
phase in which patients returned to a hypogonadal state. In the third
period, lasting approximately 1 month, patients underwent a series of
single dose transdermal pharmacokinetic studies using different application sites (23). No lipoprotein or anthropometric measurements were
made during this period. In the last period (denoted TTD), patients were
treated for 12 months with transdermal testosterone. During the TTD
period, inpatient pharmacokinetic studies and physical examinations
were performed at months 3, 6, and 12 (26).
1027
During the 12 month TTD period, all patients began treatment with 2
2.5-mg/day patches nightly (total dose, 5 mg/day). Twenty-seven patients were maintained on this regimen, and 2 were changed to a 3-patch
nightly regimen (7.5 mg/day) based on their initially low testosterone
levels. The TTD systems were applied to the recommended sites (back,
abdomen, thighs, and upper arms) on a rotating basis.
Serum hormone measurements
Morning serum levels of total testosterone (T), bioavailable testosterone (BT), dihydrotestosterone (DHT), estradiol (E2), and sex hormone-binding globulin (SHBG) were measured between 0800 and 1200 h
at 1- to 4-week intervals during the IM-T, HG, and TTD periods. Samples
were analyzed by Endocrine Sciences, Inc. (Calabasas Hills, CA) using
validated hormone assays (23).
Plasma lipoprotein parameters
All plasma samples for lipoprotein measurements were obtained in
the morning after a 12-h fast and were assayed at the Lipid Research Unit
of The Johns Hopkins School of Medicine using validated methods (28,
29), as described below. Samples from centers 1 and 2 were frozen at ⫺70
C and shipped under dry ice to the laboratory. Samples from center 3
were delivered unfrozen to the laboratory. Total cholesterol and triglycerides were measured enzymatically with a Hitachi 704 clinical
chemistry analyzer (Roche Molecular Biochemicals, Indianapolis, IN).
HDL was measured enzymatically in the clear supernatant after precipitation of the apolipoprotein B-containing lipoproteins with heparin
sulfate at 1.3 g/L and manganese chloride at 0.092 mol/L (28). LDL was
calculated from cholesterol, HDL, and triglycerides using the Friedewald equation (29). The cholesterol/HDL ratio was computed as an
index of cardiovascular risk (30).
Lipoprotein samples were obtained during out-patient visits on days
7 and 21 of the IM-T period, at weeks 4 and 8 of the HG period, and at
months 0 (pretreatment), 1, 2, 4, 5, 7, 8, 9, 10, and 11 of the TTD period.
Samples obtained during in-patient visits during the TTD period
(months 3, 6, and 12) were excluded from statistical comparisons due to
differences in posture and activity at the time of sampling, which are
known to affect lipoprotein concentrations (31).
Anthropometric measurements
Anthropometric measurements of body weight, height, waist circumference, and hip circumference were made with the patient dressed
in underwear and examining gown at the following time points: screening visit of the IM-T period, week 8 of the HG period, and months 3, 6,
and 12 of the TTD period. Waist circumference was measured at the level
of the umbilicus; hip circumference was measured at the level of the
greater trochanter. The waist to hip ratio was computed as an index of
upper body adiposity. Body mass index (BMI) was calculated as weight
(in kilograms) divided by height (in meters) squared.
Statistical analysis
Normality was assessed by Kolmogorov-Smirnov tests, and appropriate transformations were used when needed. Data that were measured more than once within a study period were reduced to period
averages, defined for each parameter on a pharmacokinetic or statistical
basis (no significant trend over time), as follows.
Androgen dosing regimens
Hormone data. For the IM-T period, the period averages for T, BT, DHT,
E2, and SHBG were defined as the AUC from days 0 –21 divided by 21,
i.e. the time-averaged values. For the HG period, the period averages
were defined as the mean of the week 4 and week 8 values. For the TTD
period, the period averages were the mean of the 10 morning hormone
levels measured from months 3–12.
During the first period, 27 men were treated with IM-T formulations
(26 with testosterone enanthate and 1 with testosterone propionate). The
average IM-T dose was 229 mg (range, 150 –300 mg), and the average
dosing interval before the last injection was approximately 26 days
(range, 10 – 44 days). Two additional patients used TTD (2 2.5-mg/day
patches nightly) based on participation in an earlier open label trial.
Lipoprotein data. Triglyceride data were found to be nonnormal and were
log-transformed. For the IM-T period, the period averages for all lipoprotein parameters were the mean of the day 7 and day 21 values. For
the HG period, the period averages were the mean of the week 4 and
week 8 values. For the TTD period, the period averages were the mean
of the month 4, 5, 7, 8, 9, 10, and 11 values.
1028
JCE & M • 2001
Vol. 86 • No. 3
DOBS ET AL.
Anthropometric data. For the IM-T and HG periods, the period averages
corresponded to the single measurements taken during those periods.
For the TTD period, the period average was the mean of the month 3,
6, and 12 values.
Descriptive statistics were computed using Excel. Inferential comparisons of study period averages among the IM-T, HG, and TTD periods were analyzed using repeated measures ANOVA, with clinical
center (1, 2, or 3) as a grouping variable. Pairwise comparisons were
based on contrasts from the ANOVA, with no adjustments for multiple
comparisons. Comparisons between the IM-T and the TTD periods
excluded the two subjects who received TTD in the IM-T period. Within
each treatment period, Pearson correlation coefficients were computed
among the hormone levels, primary lipoprotein parameters, anthropometric parameters, and age using the individual subject period averages.
For the TTD period, univariate and multivariate linear regression analyses were performed between HDL, T, and BMI. In addition, Pearson
correlation coefficients were computed for the changes in each primary
lipoprotein parameter (TTD period minus HG period) vs. age, hormone
levels, and anthropometric parameters in the TTD period.
Results
Patient disposition and characteristics
A total of 37 healthy hypogonadal men were enrolled in
the study. Three patients withdrew during the HG period,
TABLE 1. Demographic data and hypogonadal diagnoses for the
29 patients who completed the study
Demographic data
Age (yr)
Wt (kg)
Body mass index (kg/m2)
Waist to hip ratio
Race [no. (%)]
Caucasian
African-American
Asian
Hypogonadal diagnoses (no.)
Primary
Klinefelter’s syndrome
Infection
Total
Secondary
Pituitary tumor resection
Kallmann’s syndrome
Nonsecretory pituitary tumor
Sarcoidosis
Idiopathic hypogonadotropic
hypogonadism
Craniopharyngioma
Inflammatory process
Total
Mean
38.6
87.5
27.1
0.94
SEM
Range
2.5
2.9
1.0
0.01
21– 65
58.0–127.3
18.6–39.3
0.79–1.11
26 (89.7)
2 (6.9)
1 (3.4)
7
2
9
including 2 who did not meet the subnormal testosterone
level requirement for the study. Of the 34 patients who entered the 1-yr TTD treatment period, 3 patients withdrew due
to skin-related adverse events, 1 for noncompliance, and 1 for
personal reasons. Of the 29 patients (mean age, 38.6 yr) who
completed 12 months of TTD treatment, 9 patients had primary hypogonadism, and 20 had secondary hypogonadism.
A summary of the demographic parameters and hypogonadal diagnoses of the patients is given in Table 1.
Hormone levels
Average morning hormone levels of T, BT, DHT, and E2
were within the normal ranges during the IM-T and TTD
periods and were subnormal during the HG period (Table 2).
Mean SHBG levels increased during the HG period. There
were no significant differences in average morning hormone
levels between the IM-T and TTD periods.
Lipoprotein parameters
Compared with treatment with IM-T, the HG period produced increases in HDL [12.0 ⫾ 1.6% (⫾sem); P ⬍ 0.001] and
total cholesterol (4.2 ⫾ 1.9%; P ⫽ 0.02) and a decrease in the
cholesterol/HDL ratio (⫺9.7 ⫾ 2.8%; P ⫽ 0.02; Table 3).
Triglyceride levels did not change significantly. Compared
with the HG period, TTD treatment produced decreases in
HDL (⫺7.6 ⫾ 2.5%; P ⫽ 0.002) and increases in the cholesterol/HDL ratio (9.0 ⫾ 2.5%; P ⫽ 0.01) and triglycerides
(20.7 ⫾ 6.4%; P ⫽ 0.03; Table 3). Small decreases in total
cholesterol (⫺1.2 ⫾ 1.8%; P ⫽ 0.1) and LDL (⫺0.8 ⫾ 2.6%; P ⫽
0.07) were also observed during TTD, but did not reach
statistical significance. There were no significant differences
in any lipoprotein parameter between the IM-T and TTD
treatments.
Anthropometric parameters
6
5
5
1
1
Compared with IM-T treatment, there were no significant
changes in any parameter during the HG period (Table 4).
However, during the 1-yr TTD treatment, weight and BMI
exhibited small, but significant, increases compared with
values during IM-T treatment and the HG period (Table 4).
In contrast, waist circumference and waist to hip ratio did not
change significantly during TTD treatment.
1
1
20
TABLE 2. Average morning serum hormone levels in hypogonadal men (n ⫽ 29) during im testosterone ester treatment (IM-T), during a
hypogonadal state (HG), and during 1 yr of treatment with a testosterone transdermal system (TTD)
Hormonea
Testosterone (nmol/L)
BT (nmol/L)
DHT (nmol/L)
E2 (pmol/L)
SHBG (nmol/L)
Normal rangeb
IM-T
(n ⫽ 27)
HG (n ⫽ 29)
TTD (n ⫽ 29)
10.6–35.7
3.2–14.6
1.0–2.9
33–132
7.3–79.1
20.4 ⫾ 9.1
11.1 ⫾ 5.6
1.5 ⫾ 0.8
106 ⫾ 40
31.2 ⫾ 22.5
4.2 ⫾ 3.5
1.7 ⫾ 1.7c
0.4 ⫾ 0.3c
33 ⫾ 26c
37.1 ⫾ 25.7c
20.8 ⫾ 6.9d
11.1 ⫾ 3.5d
1.6 ⫾ 0.6d
99 ⫾ 33d
29.1 ⫾ 20.8d
c
Values are the mean ⫾ SD. See statistical analysis methods for definitions of average morning levels corresponding to each treatment period.
Comparisons between IM-T and TTD periods based on n ⫽ 27, all P ⫽ NS.
a
To convert testosterone and BT to nanograms per dL, multiply by 28.84; to convert DHT to nanograms per dL multiply by 29.04; to convert
E2 to nanograms per dL, multiply by 0.02724.
b
Ninety-five percent confidence intervals for morning hormone levels in normal men between the ages of 20 – 65 yr.
c
P ⬍ 0.001 compared to IM-T.
d
P ⬍ 0.001 compared to HG.
TESTOSTERONE AND LIPOPROTEINS
Interrelationships among anthropometric parameters,
lipoproteins, hormones and age
During the TTD treatment period, highly significant correlations (P ⬍ 0.001) were found within the individual
groups of anthropometric, lipoprotein, and hormone parameters (Table 5). Significant interrelationships among the parameter groups were also noted (Table 5 and Fig. 1). HDL
levels exhibited strong negative correlations with obesity
parameters, e.g. weight, BMI, and waist size, as illustrated for
BMI in Fig. 1A. Similar correlations were also found during
the IM-T and HG periods (r ⬍ ⫺0.45; P ⬍ 0.02; data not
shown). T, BT, and DHT levels also exhibited negative correlations with the obesity parameters, as illustrated by the
relationship between T and BMI in Fig. 1B. In contrast the
E2/T ratio showed highly significant positive correlations
with obesity parameters (r ⬎ 0.62; P ⬍ 0.001; Table 5), exhibiting a 5-fold increase over the BMI range from 20 – 40 (Fig.
1C). As a consequence of the negative correlations between
HDL and BMI (Fig. 1A) and between T and BMI (Fig. 1B), a
positive trend between HDL and T (r ⫽ 0.336; P ⫽ 0.07) was
observed (Fig. 2). The lack of an independent correlation
between HDL and T was shown by a multivariate regression
analysis of HDL vs. BMI and T, which yielded a regression
slope for T that was close to zero (P ⫽ 0.9). With respect to
age, the only significant correlations were obtained with the
waist to hip ratio (r ⫽ 0.588; P ⬍ 0.001) and SHBG level (r ⫽
0.459; P ⬍ 0.05; Table 5).
The changes in lipoprotein parameters induced by TTD
treatment (i.e. the differences between the TTD and HG periods) were further investigated in relation to the anthropoTABLE 3. Lipoprotein parameters during im testosterone ester
treatment (IM-T), during a hypogonadal state (HG), and during 1
yr of treatment with a testosterone transdermal system (TTD)
Parametera
IM-T
HG
TTD
Cholesterol (mmol/L)
HDL (mmol/L)
LDL (mmol/L)
Triglycerides (mmol/L)e
Chol/HDL
5.14 ⫾ 1.05
1.13 ⫾ 0.31
3.25 ⫾ 0.99
1.45 ⫾ 1.00
4.98 ⫾ 1.79
5.42 ⫾ 1.17b
1.29 ⫾ 0.35c
3.40 ⫾ 1.09
1.38 ⫾ 0.84
4.64 ⫾ 1.86b
5.32 ⫾ 1.09
1.17 ⫾ 0.29d
3.31 ⫾ 0.95
1.60 ⫾ 1.15f
4.94 ⫾ 1.74f
Values are the mean ⫾ SD (n ⫽ 29). Comparisons between IM-T and
TTD are based on n ⫽ 27.
a
To convert cholesterol, HDL, and LDL to milligrams per dL,
multiply by 38.67; to convert triglycerides to milligrams per dL, multiply by 88.57.
b
P ⬍ 0.05 compared to IM-T.
c
P ⬍ 0.001 compared to IM-T.
d
P ⬍ 0.01 compared to HG.
e
Geometric mean given for this parameter.
f
P ⬍ 0.05 compared to HG.
1029
metric parameters, hormone levels, and patient age. In contrast to the interrelationships displayed in Table 5, the
changes in HDL were independent of the anthropometric
parameters or hormone levels. Thus, the small reduction in
HDL levels seen during TTD treatment was not significantly
different between men whose BMI was less than the median
value of 27 (⫺0.10 ⫾ 0.06 mmol/L) and those whose BMI was
greater than 27 (⫺0.14 ⫾ 0.03 mmol/L; P ⫽ 0.5). Interestingly,
the only predictive variable for the changes in lipoprotein
levels was the patient’s age, which had a negative correlation
to the changes in total cholesterol and LDL levels (r ⫽ ⫺0.451
and ⫺0.446, respectively; P ⬍ 0.05), but no correlation with
the change in HDL levels (r ⫽ ⫺0.045; P ⫽ 0.8; Fig. 3, A
and B).
Discussion
The magnitude and direction of the lipoprotein changes
observed in our study are consistent with other studies in
hypogonadal men receiving im or transscrotal androgen replacement (11, 32) and with studies in eugonadal men undergoing androgen supplementation (9, 33, 34) or deprivation (16, 35). They differ substantially, however, from the
deleterious lipoprotein effects seen in anabolic steroid abusers (36 –38).
Studies in male weight lifters using high dosages of androgens and anabolic steroids showed large decreases in
HDL levels of approximately ⫺0.78 mmol/L (36, 37). Similar
decreases in the HDL2 subfraction were observed with use of
oral nonaromatizable androgens and were associated with
an increase in hepatic triglyceride lipase activity, which increases HDL catabolism (38). The striking reductions in HDL
produced by orally administered androgens may be related
to hepatic first pass effects of the steroids (33), their inability
to form estrogenic metabolites (34), and/or a direct hepatotoxic effect in some cases (39). The much smaller reduction
in HDL levels in our study during both IM-T and TTD therapy, approximately ⫺0.13 mmol/L, is presumably due to the
lower doses of testosterone used, the aromatization to estradiol, and the avoidance of hepatic first pass metabolism.
In eugonadal men, administration of testosterone enanthate (ⱖ200 mg/week) caused HDL levels to decrease by
⫺0.10 to ⫺0.23 mmol/L (9, 33, 34). Conversely, experimental
hypogonadism, induced by GnRH agonists or antagonists in
eugonadal men, reversibly increased HDL levels by approximately ⫹0.26 mmol/L (16, 35). In Bagatell’s study (16), the
increase in HDL was prevented by the coadministration of
testosterone enanthate at a dose of 100 mg/week, which
maintained testosterone levels in the physiological range.
This finding suggests that physiological levels of testosterone
TABLE 4. Anthropometric parameters in hypogonadal men (n ⫽ 29) during im testosterone ester treatment (IM-T), during a
hypogonadal state (HG), and during 1 yr of treatment with a testosterone transdermal system (TTD)
Parameter
IM-T
HG
TTD
Wt (kg)
BMI (kg/m2)
Waist circumference (cm)
Waist to hip ratio
87.5 ⫾ 15.4
27.1 ⫾ 5.5
99.2 ⫾ 11.1
0.945 ⫾ 0.072
87.9 ⫾ 15.7
27.3 ⫾ 5.5
99.0 ⫾ 11.1
0.954 ⫾ 0.085
89.2 ⫾ 15.9a,b
27.8 ⫾ 5.5a,b
98.1 ⫾ 10.2
0.940 ⫾ 0.055
Values are the mean ⫾ SD. Comparisons between IM-T and TTD periods are based on n ⫽ 27.
a
P ⬍ 0.01 compared to IM-T.
b
P ⬍ 0.05 compared to HG.
0.459b
0.137
0.193
⫺0.226
0.261
0.146
0.196
0.052
⫺0.196
0.073
⫺0.212
Age
0.588a
⫺0.250
⫺0.178
⫺0.205
⫺0.022
⫺0.128
0.176
⫺0.354
⫺0.303
⫺0.309
⫺0.143
⫺0.080
0.063
0.336
0.223
0.212
⫺0.379b
0.104
⫺0.557c
⫺0.354
⫺0.297
⫺0.327
⫺0.243
⫺0.110
⫺0.003
⫺0.705
⫺0.480c
⫺0.632a
0.231
⫺0.375b
0.764a
T
BT
DHT
E2
SHBG
E2/T
⫺0.045
⫺0.145
0.014
0.276
0.065
0.105
⫺0.554
⫺0.530c
⫺0.511c
0.248
⫺0.169
0.623a
⫺0.612
⫺0.482c
⫺0.594a
0.238
⫺0.272
0.789a
a
c
b
a
P ⬍ 0.001.
P ⬍ 0.05.
P ⬍ 0.01.
0.034
a
0.190
c
0.300
⫺0.475c
0.352
0.251
0.462b
0.106
⫺0.520c
0.165
0.182
0.390b
0.108
⫺0.432b
0.215
0.047
0.319
Chol
HDL
LDL
TG
Chol/HDL
0.801a
0.182
0.598a
BMI
Wgt
BMI
Waist
WHR
0.804a
0.801a
0.101
Waist
WHR
0.331
⫺0.202
0.291
0.313
0.323
Chol
⫺0.176
0.939a
0.490c
0.657a
HDL
⫺0.240
⫺0.536c
⫺0.811a
LDL
0.250
0.636a
TG
0.732a
⫺0.385b
⫺0.291
⫺0.268
0.130
⫺0.124
0.369b
Chol/HDL
T
0.534c
0.844a
0.044
0.586a
⫺0.716a
BT
0.489c
0.426b
⫺0.315
⫺0.265
DHT
0.200
0.492c
⫺0.524c
E2
⫺0.381b
0.556c
SHBG
⫺0.586a
E2/T
⫺0.090
DOBS ET AL.
TABLE 5. Correlation matrix among anthropometric parameters, lipoproteins, morning hormone levels, and age in hypogonadal men (n ⫽ 29) during 1 yr of TTD treatment
1030
JCE & M • 2001
Vol. 86 • No. 3
suppress HDL levels in normal eugonadal men (16) to the
same degree seen in hypogonadal patients receiving TTD
treatment.
In contrast to exogenous testosterone, the relationship between endogenous testosterone and lipoprotein levels is
more controversial. A positive correlation between HDL cholesterol and testosterone levels has been demonstrated in
several cross-sectional studies (17–20, 40). Other crosssectional studies have found a negative correlation (5) or no
correlation (22, 41, 42). The cross-sectional data analysis of
our study revealed a positive trend between HDL and testosterone levels during TTD treatment. This relationship appears to be due to obesity, which is negatively correlated to
both HDL and testosterone.
The interrelationships among body size, body composition, testosterone, and lipoproteins are complex. Crosssectional studies in men with varying degrees of obesity have
shown that total testosterone levels and, to a lesser extent,
bioavailable and/or free testosterone levels decrease with
increasing BMI values (43– 48). These changes are associated
with decreases in SHBG (43, 44, 47) and IGF-I levels (48),
increases in insulin and estradiol levels (47, 48), and low/
normal levels of LH (48), which are indicative of a state of
insulin resistance, increased aromatization of testosterone,
and hypogonadotropic hypogonadism. As endogenous testosterone levels were low in our hypogonadal patients, and
testosterone absorption from the TTD was independent of
weight or BMI (26), the observed decrease in testosterone
levels with increasing BMI must be a consequence of increasing testosterone clearance. The latter could result from
an increase in the volume of distribution, a decrease in halflife, and/or an increase in metabolic conversion (49). Lower
SHBG levels, which showed a negative trend with BMI,
would also be expected to increase testosterone clearance
rates (49). The striking increase in E2/T ratios with BMI
observed in our patients is consistent with data in women
that show a positive correlation between obesity and the
rates of testosterone aromatization (50). In obese men with
intact hypothalamic-pituitary-gonadal axes, the increase in
E2/T ratio and the negative feedback effects of circulating E2
levels on LH secretion (51, 52) could be important factors in
the development of hypogonadotropic hypogonadism (48).
As a potent inhibitor of abdominal (sc) lipoprotein lipase
(53) and a stimulator of muscle protein synthesis (54), exogenous testosterone has been shown to decrease abdominal
fat mass and increase muscle mass (55, 56). In the present
study the 1.3-kg increase in weight during the TTD period
was associated with small (but not statistically significant)
decreases in waist size and waist to hip ratio. This suggests
that the observed weight gain may have resulted from an
increase in lean mass rather than fat mass, as seen in other
testosterone replacement studies in hypogonadal men in
which body composition was measured (57, 58).
The negative correlation between HDL levels and BMI
seen in the present study has been observed in numerous
studies of eugonadal men (59 – 61) as well as in patients with
coronary artery disease (62). Insulin resistance and increased
hepatic triglyceride lipase activity have been postulated to
explain the metabolic defect in obesity that causes low HDL
levels (60 – 62).
TESTOSTERONE AND LIPOPROTEINS
1031
FIG. 1. Correlations between HDL levels vs. BMI (A), morning testosterone levels vs. BMI (B), and estradiol to testosterone ratios (E2/T) vs.
BMI (C) in 29 hypogonadal men during 1 yr of transdermal testosterone treatment. Dashed lines derived from linear regression. r and P values
are denoted in the upper right corners of each panel. To convert testosterone levels to nanograms per dL, multiply by 28.84. To convert HDL
levels to milligrams per dL, multiply by 38.67.
FIG. 2. Correlation between HDL levels and morning testosterone
levels in 29 hypogonadal men during 1 yr of transdermal testosterone
treatment. The dashed line is derived from linear regression. The r
and P values are denoted in the upper right corner. To convert testosterone levels to nanograms per dL, multiply by 28.84. To convert
HDL levels to milligrams per dL, multiply by 38.67.
In contrast to the effects of testosterone replacement on
HDL levels and cholesterol/HDL ratios, recent studies suggest that testosterone may have cardioprotective effects in
men related to vasodilation (63) and other antiischemic
mechanisms (64, 65), and that testosterone may be potentially
beneficial as a cardiovascular drug (66, 67). Although some
contradictions appear in the literature (68), it is likely that the
overall influence of testosterone replacement on cardiovas-
FIG. 3. Correlations between changes in LDL levels vs. age (A) and
changes in HDL levels vs. age (B) in 29 hypogonadal men during 1 yr
of transdermal testosterone treatment. The dashed line is derived
from linear regression. The r and P values are denoted in the upper
right corners of each panel. To convert LDL and HDL levels to milligrams per dL, multiply by 38.67.
cular risk in men will ultimately involve effects on lipoproteins as well as other factors, similar to current views regarding estrogen replacement and cardiovascular risk in
women (7).
1032
DOBS ET AL.
Lastly, our findings regarding the influence of age on
lipoprotein changes during testosterone replacement are
consistent with a number of recent studies conducted in older
men that have shown decreases in LDL or total cholesterol
during testosterone therapy (12–15, 69, 70) and little or no
change in HDL levels (12–15). The mechanism by which age
moderates the effects of exogenous testosterone on lipoprotein levels remains to be further understood.
In conclusion, the longitudinal component of our study
showed a small reduction in HDL levels in hypogonadal men
treated with im or transdermal testosterone preparations,
whereas the cross-sectional data analysis showed a positive
trend between HDL and testosterone levels during transdermal treatment. These seemingly contradictory findings
are reconciled by the anthropometric data from our patients,
which showed that HDL and testosterone levels were both
negatively correlated to BMI and that the positive crosssectional trend between them was a consequence of these
relationships. In regard to our longitudinal observations, the
effects of transdermal testosterone replacement on serum
lipoprotein levels in hypogonadal men appear to be consistent with the physiological effects of testosterone in healthy
eugonadal men.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Acknowledgments
N.A.M. would like to acknowledge helpful discussions on lipoprotein
measurement, cardiovascular risk, and anthropometrics with Drs. Paul
N. Hopkins and Maria E. Ramirez, and the statistical expertise of Drs.
John Burkhart, L. Rajaram, and Heather Thomas.
References
1. Hoyert DL, Kochanek KD, Murphy SL. 1999 Deaths: final data for 1997. Natl
Vital Stat Rep. 47:1–104.
2. Castelli WP, Garrison RJ, Wilson PW, et al. 1986 Incidence of coronary heart
disease and lipoprotein cholesterol levels–The Framingham Study. JAMA.
256:2835–2838.
3. Goldbourt U, Holtzman E, Neufeld HN. 1985 Total and high density lipoprotein cholesterol in the serum and risk of mortality: evidence of a threshold
effect. Br Med J. 290:1239 –1243.
4. Gordon DJ, Rifkind BM. 1989 High-density lipoprotein: the clinical implications of recent studies. N Engl J Med. 321:11311–1316.
5. Semmons J, Rouse I, Beilin LJ, et al. 1983 Relationship of plasma HDLcholesterol to testosterone, estradiol and sex-hormone-binding globulin levels
in men and women. Metabolism. 32:428 – 432.
6. Gorodeski GI. 1994 Impact of the menopause on the epidemiology and risk
factors of coronary artery heart disease in women. Exp Gerontol. 29:357–375.
7. Adams MR, Washburn SA, Wagner JD, Williams JK, Clarkson TB. 1994
Arterial changes: estrogen deficiency and effects of hormone replacement. In:
Lobo RA, ed. Treatment of the postmenopausal woman: basic and clinical
aspects, chapt 22. New York: Raven Press; 243–250.
8. Kirkland RT, Keenan BS, Probstfield JL, et al. 1987 Decrease in plasma
high-density lipoprotein cholesterol levels at puberty in boys with delayed
adolescence. JAMA. 257:502–507.
9. Meriggiola MC, Marcovina S, Paulsen CA, Bremenr WJ. 1995 Testosterone
enanthate at a dose of 200 mg/week decreases HDL-cholesterol levels in
healthy men. Int J Androl. 18:237–242.
10. Applebaum-Bowden D, Haffner SM, Hazzard WR. 1987 The dysproteinemia
of anabolic steroid therapy: increase in hepatic triglyceride lipase precedes the
decrease in high density lipoprotein2 cholesterol. Metabolism. 36:949 –952.
11. Sorva R, Kuusi T, Taskinen NM, et al. 1988 Testosterone substitution increases the activity of lipoprotein lipase and hepatic lipase in hypogonadal
males. Atherosclerosis. 69:191–197.
12. Tenover JS. 1992 Effects of testosterone supplementation in the aging male.
J Clin Endocrinol Metab. 75:1092–1098.
13. Zgliczynski S, Ossowski M, Slowinska-Srzednicka J, et al. 1996 Effect of
testosterone replacement therapy on lipids and lipoproteins in hypogonadal
and elderly men. Atherosclerosis. 121:35– 43.
14. Münzer T, Harman SM, Christmas C, et al. 2000 Effects of administration of
testosterone and/or GH in healthy aged men. Aging Male. 3:3.
15. Kenny AM, Prestwood KM, Marcello KD, Fall PM, Raisz LG. 2000 The effects
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
JCE & M • 2001
Vol. 86 • No. 3
of transdermal testosterone on bone metabolism, body composition, lipids and
health-related quality of life in older men. Aging Male. 3:3.
Bagatell CJ, Knopp RH, Vale WW, Rivier JE, Bremner WJ. 1992 Physiologic
testosterone levels in normal men suppress high-density lipoprotein cholesterol levels. Ann Intem Med. 116:967–973.
Barrett-Connor EL. 1995 Testosterone and risk factors for cardiovascular disease in men. Diabetes Metab. 21:151–161.
Hamalainen E, Adlercreutz K Ehnholm C, Puska P. 1986 Relationships of
serum lipoproteins and apolipoproteins to sex hormones and to the binding
capacity of sex hormone binding globulin in healthy Finnish men. Metabolism.
35:535–541.
Nordoy A, Aakvaag A, Thelle D. 1979 Sex hormones and high densitv lipoproteins in healthy mates. Atherosclerosis. 34:431– 436.
Heller RF, Wheeler MJ, Micallef J, Miller NE, Lewis B. 1983 Relationship of
high density lipoprotein cholesterol with total and free testosterone and sex
hormone binding globulin. Acta Endocrinol (Copenh). 104:253–256.
Phillips GB, Pinkernell BH, Jing T-Y. 1994 The association of hypotestosteronemia with coronary artery disease in men. Arterioscler Thromb.
14:701–706.
Stefanick ML, William PT,, Krauss RM, Terry RB, Vranizan KM, Wood PD.
1987 Relationships of plasma estradiol, testosterone, and sex hormone-binding
globulin with lipoproteins, apolipoproteins, and high density lipoprotein subfractions in men. J Clin Endocrinol Metab. 64:723–729.
Meikle AW, Arver S, Dobs AS, Sanders SW, Rajaram L, Mazer NA. 1996
Pharmacokinetics and metabolism of a permeation-enhanced testosterone
transdermal system in hypogonadal men: influence of application site–a clinical research center study. J Clin Endocrinol Metab. 81:1832–1840.
Arver S, Dobs AS, Meikle AW, Allen RP, Sanders SW, Mazer NA. 1996
Improvement in sexual function in testosterone deficient men treated for 1 year
with a permeation enhanced testosterone transdermal system. J Urol.
155:1604 –1608.
Meikle AW, Arver S, Dobs AS, et al. 1997 Prostate size in hypogonadal men
treated with a nonscrotal permeation-enhanced testosterone transdermal system. Urology. 49:191–196.
Arver S, Dobs AS, Meikle AW, Caramelli KE, Rajaram L, Sanders SW, Mazer
NA. 1997 Long-term efficacy and safety of a permeation-enhanced testosterone
transdermal system in hypogonadal men. Clin Endocrinol (Oxf). 47:727–737.
Lipid Research Clinics Program. 1980 The lipid research clinics population
studies data book, vol 1. NIH publication 80 –1527. Bethesda: National Institutes of Health.
Bachorik PS, Kwiterovich PO. 1991 Measurement of plasma cholesterol, lowdensity lipoprotein cholesterol, and high-density lipoprotein cholesterol. In:
Techniques in diagnostic human biochemical genetics: a laboratory manual,
chapt 26. New York: Wiley-Liss; 425– 459.
Friedewald W, Levy RI, Frederickson DS. 1972 Estimation of the concentration of low-density lipoprotein cholesterol in plasma without the use of the
preparative centrifuge. Clin Chem. 18:499 –502.
Stampfer MJ, Sacks FM, Salvini S, Willett WC, Hennekens CH. 1991 A
prospective study of cholesterol apolipoproteins, and the risk of myocardial
infarction. N Engl J Med. 325:373–381.
Miller M, Bachorik PS, Cloey TA. 1992 Normal variation of plasma lipoproteins: postural effects on plasma concentrations of lipids, lipoproteins, and
apolipoproteins. Clin Chem. 38:569 –574.
Place VA, Atkinson L, Prather DA, Trunnell N, Yates FE. 1990 Transdermal
testosterone replacement through gential skin. In: Nieschlag E, Behre HM, eds.
Testosterone: action, deficiency and substitution. Berlin: Springer Verlag;
165–181.
Thompson PD, Cullinane EM, Sady S, et al. 1989 Contrasting effects of
testosterone and stanozolol on serum lipoprotein levels. JAMA. 261:1165–1168.
Friedl KE, Hannan CJ, Jones RE, Plymate SR. 1990 High-density lipoprotein
cholesterol is not decreased if an aromatizable androgen is administered.
Metabolism. 39:69 –74.
Goldberg RB, Rabin D, Alexander AN, Doelle GC, Getz GS. 1985 Suppression of plasma testosterone leads to an increase in serum total and high density
lipoprotein cholesterol and apoproteins a-i and b. J Clin Endocrinol Metab.
60:203–207.
Webb OL, Larkarzewski PM, Glueck CJ. 1984 Severe depression of highdensity lipoprotein cholesterol levels in weight lifters and body builders by
self-administered exogenous testosterone and anabolic-androgenic steroids.
Metabolism. 33:971–975.
Alen M, Rahkila P, Marniemi J. 1985 Serum lipids in power athletes selfadministering testosterone and anabolic steroids. Int J Sports Med. 6:139 –144.
Kantor MA, Bianchini A, Bernier D, Sady SP, Thompson PD. 1985 Androgens reduce HDL2-cholesterol and increase hepatic triglyceride lipase activity.
Med Sci Sports Exerc. 17:462– 465.
Furman RH, Howard RP, Norcia LN, Keaty EC. 1958 The influence of androgens, estrogens and related steroids on serum lipids and lipoproteins/
observations in hypogonadal and normal human subjects. Am J Med. 24:80 –97.
Gutai L, LaPorte R, Kuller L, Dai W, Falvo-Gerard L, Caggiula A. 1981 Plasma
testosterone, high density lipoproteins in healthy males. Atherosclerosis.
34:431– 436.
TESTOSTERONE AND LIPOPROTEINS
41. Kiel DP, Baron JA, Plymate SR, Chute CG. 1989 Sex hormones and lipoproteins in men. Am J Med. 87:35–39.
42. Mendoza SG, Zerpa A, Carrasco K, et al. 1983 Estradiol, testosterone, apolipoprotein, lipoprotein cholesterol and lipolytic enzymes in men without
premature myocardial infarction and angiographically assesses coronary occlusion. Artery. 12:1–23.
43. Glass AR, Swerdloff RS, Bray GA, Dahms WT, Atkinson RL. 1977 Low
serum testosterone and sex-hormone-binding-globulin in massively obese
men. J Clin Endocrinol Metab. 45:1211–1219.
44. Amatruda JM, Harman SM, Pourmotabbed G, Lockwood DH. 1978 Depressed plasma testosterone and fractional binding of testosterone in obese
males. J Clin Endocrinol Metab. 47:268 –271.
45. Zumoff B, Strain GW, Miller LK, et al. 1990 Plasma free and non-sexhormone-binding-globulin-bound testosterone are decreased in obese men in
proportion to their degree of obesity. J Clin Endocrinol Metab. 71:929 –931.
46. Vermeulen A, Kaufman JM, Deslypere JP, Thomas G. 1993 Attenuated luteinizing hormone (LH) pulse amplitude but normal LH pulse frequency, and
its relation to plasma androgens in hypogonadism of obese men. J Clin Endocrinol Metab. 76:1140 –1146.
47. Vermeulen A, Kaufman JM, Giagulli VA. 1996 Influence of some biological
indexes on sex hormone-binding globulin and androgen levels in aging or
obese males. J Clin Endocrinol Metab. 81:1821–1826.
48. Giagulli VA, Kaufman JM, Vermeulen A. 1994 Pathogenesis of the decreased
androgen levels in obese men. J Clin Endocrinol Metab. 79:997–1000.
49. Wilkinson GR, Shand DG. 1975 Commentary: a physiological approach to
hepatic drug clearance. Clin Pharmacol Ther. 18:377–390.
50. Longcope C, Baker R, Johnston CC. 1986 Androgen and estrogen metabolism:
relationship to obesity. Metabolism. 35:235–237.
51. Finkelstein JS, O’Dea LSL, Whitcomb RW, Crowley WF. 1991 Sex steroid
control of gonadotropin secretion in the human male. I. Effects of testosterone
administration in normal and gonadotropin-releasing hormone-deficient men.
J Clin Endocrinol Metab. 73:609 – 620.
52. Finkelstein JS, Whitcomb RW, O’Dea LSL, Longcope C, Schoenfeld DA,
Crowley WF. 1991 Sex steriod control of gonadotropin secretion in the human
male. II. Effects of estradiol administration in normal and gonadotropinreleasing hormone-deficient men. J Clin Endocrinol Metab. 73:621– 628.
53. Ramirez ME, McMurry MP, Wiebke GA, et al. 1997 Evidence for sex steroid
inhibition of lipoprotein lipase in men: comparison of abdominal and femoral
adipose tissue. Metabolism. 46:179 –185.
54. Urban RJ, Bodenburg YH, Gilkison C, et al. 1995 Testosterone administration
to elderly men increases skeletal muscle strength and protein synthesis. Am J
Physiol. 269:E820 –E826.
55. Mårin P, Holmäng S, Jönsson L, et al. 1992 The effects of testosterone treatment on body composition and metabolism in middle aged obese men. Int J
Obesity. 16:991–997.
1033
56. Bhasin S, Storer TW, Berman N, et al. 1996 The effects of supraphysiologic
doses of testosterone on muscle size and strength in men. N Engl J Med.
335:1–7.
57. Katznelson L, Finkelstein JS, Schoenfeld DA, Rosenthal DI, Anderson EJ,
Klibanski A. 1996 Increase in bone density and lean body mass during testosterone administration in men with acquired hypogonadism. J Clin Endocrinol Metab. 81:4358 – 4365.
58. Bhasin S, Storer TW, Asbel-Sethi N, et al. 1998 Effects of testosterone replacement with a non-genital, transdermal system, Androderm, in human
immunodeficiency virus-infected men with low testosterone levels. J Clin
Endocrinol Metab. 83:3155–3162.
59. Heiss G, Johnson NJ, Reiland S, Davis CE, Tyroler HA. 1980 The epidemiology of plasma high-density lipoprotein cholesterol levels. The Lipid Research
Clinics Program Prevalence Study Summary. Circulation. 62:IV116 –IV136.
60. Haffner SM, Stern MP, Hazuda HP, Pugh J, Patterson JK. 1987 Do upperbody and centralized adiposity measure different aspects of regional body-fat
distribution? Relationship to non-insulin-dependent diabetes mellitus, lipids,
and lipoproteins. Diabetes. 36:43–51.
61. Freedman DS, Jacobsen SJ, Barboriak JJ, et al. 1990 Body fat distribution and
male/female differences in lipid and lipoprotein. Circulation. 81:1498 –1506.
62. Lichtenstein MJ, Yarnell JW, Elmwood PC, et al. 1987 Sex hormones, insulin,
lipids, and prevalent ischemic heart disease. Am J Epidemiol. 126:647– 657.
63. Webb CM, McNeill JG, Hayward CS, de Zeigler D, Collins P. 1999 Effects
of testosterone on coronary vasomotor regulation in men with coronary heart
disease. Circulation. 100:1690 –1696.
64. Webb CM, Adamson DL, de Zeigler D, Collins P. 1999 Effect of acute testosterone on myocardial ischemia in men with coronary artery disease. Am J
Cardiol. 83:437– 439.
65. Rosano GM, Leonardo F, Pagnotta P, et al. 1999 Acute anti-ischemic effect of
testosterone in men with coronary artery disease. Circulation. 99:1666 –1670.
66. Alexandersen P, Haarbo J, Christiansen C. 1996 The relationship of natural
androgens to coronary heart disease in males: a review. Atherosclerosis.
125:1–13.
67. Shapiro J, Christiana J, Frishman WH. 1999 Testosterone and other anabolic
steroids as cardiovascular drugs. Am J Ther. 6:167–174.
68. White CM, Ferraro-Borgida MJ, Moyna NM, et al. 1999 The effect of pharmacokinetically guided acute intravenous testosterone administration on electrocardiographic and blood pressure variables. J Clin Pharmacol. 39:1038 –1043.
69. Morley JE, Perry HM, Kaiser FE, et al. 1993 Effects of testosterone replacement
therapy in males: a preliminary study. J Am Geriatr Soc. 41:149 –152.
70. Sih R, Morley JE, Kaiser FE, Perry HM, Patrick P, Ross C. 1997 Testosterone
replacement in older hypogonadal men: a 12-month randomized controlled
trial. J Clin Endocrinol Metab. 82:1661–1667.