| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Clinical Research Centers |
Department Obstetrics and Gynecology (M.P.D., M.C.D.), Division of Reproductive Endocrinology and Infertility, Hutzel Hospital/Wayne State University School of Medicine, Detroit, Michigan 48201; Center for Reproductive Medicine (D.G.), Wichita, Kansas 67214; Department of Internal Medicine (R.S.S.), Yale New Haven Hospital, New Haven, Connecticut 06520; Diabetes Division, Department of Medicine R.A.D.), University of Texas Health Science Center, San Antonio, Texas 78284-7870 ,
Address correspondence and requests for reprints to: Michael P. Diamond, MD, Professor of Obstetrics and Gynecology, Hutzel Hospital/Wayne State University, 4707 St Antoine Boulevard, Detroit, Michigan 48201.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsReferences |
|---|
Short-term methyltestosterone administration had no significant effects on the fasting levels of glucose, insulin, c-peptide, glucagon, or glucose turnover. During the hyperglycemic clamp studies, the mean glucose level during the final hour was 203 ± 2 and 201 ± 1 mg/dL in the methyltestosterone and control studies, respectively. The insulin response to this hyperglycemic challenge was slightly but not significantly greater during methyltestosterone treatment (first phase 59 ± 8 vs. 50 ± 8 µU/mL in controls; second phase 74 ± 9 vs. 67 ± 9 µU/mL in controls; total insulin response 133 ± 16 vs. 117 ± 15 µU/mL in controls). In spite of this, glucose uptake was reduced from the control study value of 10.96 ± 1.11 to 7.3 ± 0.70 mg/kg/min by methyltestosterone (P < 0.05). The ratio of glucose uptake per unit of insulin was also significantly reduced from a control study value of 14.3 ± 1.4 to 9.4 ± 1.3 mg/kg/min per µU/mL x 100 during methyltestosterone administration. In the euglycemic hyperinsulinemic clamp studies, insulin was infused at rates of 0.25 and 1.0 mU/kg/min to achieve insulin levels of approximately 25 and 68 µU/mL, respectively. During low-dose insulin infusion, rates of endogenous hepatic glucose production were equivalently suppressed from basal values of 2.37 ± 0.29 and 2.40 ± 0.27 mg/kg/min to 0.88 ± 0.25 and 0.77 ± 0.26 mg/kg/min in the methyltestesterone and control studies respectively. Whole body glucose uptake during low-dose insulin infusion was minimally affected. During the high-dose insulin infusion, endogenous hepatic glucose production was nearly totally suppressed in both groups. However, whole body glucose uptake was reduced from the control value of 6.11 ± 0.49 mg/kg/min to 4.93 ± 0.44 mg/kg/min during methyltestosterone administration (P < 0.05).
Our data demonstrate that androgen excess leads to the development of insulin resistance during both hyperglycemic and euglycemic hyperinsulinemia. These findings provide direct evidence for a relationship between hyperandrogenemia and insulin resistance, and its associated risk factors for cardiovascular disease.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsReferences |
|---|
A potential role of insulin in stimulating ovarian androgen production is supported by several lines of evidence. In vitro studies demonstrate that ovarian tissue contains insulin receptors (2, 3, 4), and insulin directly stimulates ovarian and adrenal steroidogenesis (5, 6, 7). Additionally, in vivo, an association between basal hyperinsulinemia and elevated basal testosterone levels has been consistently demonstrated in both obese and nonobese women with polycystic ovarian syndrome (PCO) (8, 9, 10). However, in regularly cycling, euandrogenic women without PCO, a significant correlation between basal insulin and androgen levels has yet to be documented (11).
A direct inhibitory effect of androgens on insulin sensitivity remains controversial. Both obese and nonobese PCO women have reduced insulin sensitivity compared with age- and weight-matched non-PCO control subjects (12). Additionally, elevated plasma insulin levels (which may be the result of increased pancreatic insulin secretion) (13) imply the presence of underlying insulin resistance and have been identified in hyperandrogenic women (8, 10, 14) and after DHEA administration in control subjects. (15). Reduction of androgen levels in PCO subjects may not always be associated with improvement in insulin action (16, 17), however; although this might be explained by the fact that the degree of androgen elevation in these women was minimal.
In this report, we have attempted to better define the effect of androgens on glucose homeostasis in euandrogenic, non-PCO women. Specifically, we have utilized the hyperglycemic and euglycemic, hyperinsulinemic clamp techniques to examine the effect of short-term administration of the synthetic androgen methyltestosterone on insulin secretion and on hepatic and peripheral insulin sensitivity.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsReferences |
|---|
All women had a normal oral glucose tolerance test, were not taking oral contraceptives or other drugs known to influence carbohydrate metabolism, and had had no recent change in body weight. They were instructed to maintain their normal activity and to consume a normal food intake, containing at least 200 gm carbohydrate per day for three days before the study. All subjects were informed of the nature, purpose, and risk of the study before giving their informed, voluntary, written consent for their participation. The experimental protocol was approved by the Human Investigation Committee of Yale University School of Medicine.
Each subject received a hyperglycemic or euglycemic, hyperinsulinemic clamp study before and after 10–12 days of methyltestosterone administration (5 mg tid, The Brown Pharmaceutical Co., Costa Mesa, CA). The studies were performed with the subjects in the supine position, beginning at 0800 h, after an overnight fast of 10 h. Before the study, a small teflon catheter was inserted into an antecubital vein for infusion of insulin and/or glucose. A second cannula was inserted retrogradely into a wrist vein for blood sampling and kept patent by isotonic saline. The hand then was placed in a box that was heated to approximately 65 C to obtain "arterialized" venous blood samples through the wrist cannula (18).
Hyperglycemic clamp protocol
During the hyperglycemic clamp study the plasma glucose concentration was acutely raised by 125 mg/dL above baseline and maintained at this level for two h, as described by DeFronzo et al. (19). A priming dose of glucose was administered over the first 15 min. Subsequently, the plasma glucose concentration was measured every 5 min, and a 20% glucose solution was adjusted, based on a negative feedback principle, to maintain the plasma glucose constant at the desired hyperglycemic plateau (19). Plasma samples for glucose and insulin were drawn at 5- to 10-min intervals during the basal period, at 2-min intervals during the initial 10 min of the hyperglycemic clamp, and every 5–10 min thereafter.
Euglycemic, hyperinsulinemic clamp protocol
A two-step euglycemic, hyperinsulinemic clamp study was performed. Each insulin infusion step lasted for 100 min. During the first step a primed continuous insulin infusion was administered at the rate of 0.25 mU/kg/min. At 100 min another primary dose of insulin was given, and a continuous insulin infusion was increased to a rate of 1.0 mU/kg/min. Throughout, plasma glucose concentration was monitored every five min and maintained at the fasting level by the periodic adjustment of a variable infusion of 20% dextrose, based on a negative feedback principle (2). Plasma samples for insulin and glucagon were drawn four times during the control period and every 5–20 min thereafter.
In all of the euglycemic hyperinsulinemic clamp studies the subjects were given a primed (25 µCi), continuous (0.25 µCi/min) infusion of tritiated-3-glucose (New England Nuclear Boston, MA) to assess whole body glucose metabolism and hepatic glucose production. The 3H-glucose infusion was initiated 120 min before the euglycemic hyperinsulinemic clamp, and continued throughout the study. Samples for plasma tritiated glucose specific activity determinations were taken four times during the control period and every 5–15 min during insulin/glucose infusion.
In both the hyperglycemic and euglycemic insulin clamp studies, samples for the determination of estradiol (E2), progesterone, testosterone, and free testosterone were obtained three times during the basal period; prolactin, luteinizing hormone (LH), follicle-stimulating hormone (FSH), delydroepiandrosterone sulfate (DHEAS) and androstenedione, cortisol, growth hormone, epinephrine, and norepinephrine were determined twice during the basal period.
Analytic procedures
All arterialized venous blood samples were immediately centrifuged and frozen at -70 C until analyzed. Glucose levels were determined in duplicate at the patient’s bedside on a Glucose Analyzer II (Beckman Coulter, Inc. Instruments, Fullerton, CA). The plasma insulin concentration was determined with a double-antibody radioimmunoassay (Cambridge Medical Diagnostics, Billerica, MA). Plasma glucagon (ICN Biomedicals, Inc., Carcon, CA), estradiol (Ciba-Corning, East Walpole, MA) and progesterone (Ciba-Corning) concentrations were measured by radioimmunoassay. Plasma glucose radioactivity was measured using the Somogyi procedure for precipitation of plasma proteins as previously described (20). Inter- and intraassay coefficients of variation were, respectively, as follows: insulin, 6% and 5%; glucose, 1% and 1%; E2, 8% and 4%; P, 6% and 5%; T, 9% and 9%; epinephrine, 4% and 7%; norepinephrine, 11% and 11%; glucagon, 7% and 4%; GH, 8% and 6%; cortisol, 7% and 4%; and PRL, 6% and 3% (21). For DHEAS, prolactin, LH, and FSH, interassay and intraassay variability was less than 10% (22).
Calculations
During the hyperglycemic clamp study, the weighted mean of plasma insulin concentration from 0–10, 10–120, and 0–120 min was determined to quantitate the first phase, second phase, and total plasma insulin responses. The basal plasma insulin concentration represents the mean of at least three values obtained at 5- to 10-min intervals during the baseline period.
Under steady state plasma glucose concentrations, as occurred during the last hour of the hyperglycemic clamp, hepatic glucose production was completely suppressed (23), and the amount of glucose infused was expected to equal the amount of glucose taken up by all the cells of the body after a small correction for urinary glucose losses, which was less than 0.2 mg/kg/min in all subjects (19). During the last hour of the hyperglycemic clamp, the amount of glucose metabolized (M) divided by the plasma insulin concentration (I) provided an estimate of whole body sensitivity to insulin under hyperinsulinemic conditions (19).
During the euglycemic, hyperinsulinemic clamp studies, the glucose infusion rate, corrected for minor changes in the plasma glucose level, was averaged over 20-min intervals as previously described (24). The residual rate of hepatic glucose production during the last 40 min of the insulin clamp, when added to the mean glucose infusion rate during the same time period, provided a measure of whole body insulin-mediated glucose metabolism. In the basal state the rate of hepatic glucose production was determined by dividing the mean 3-3H-glucose infusion rate (DPM/min) by the steady state plasma tritiated glucose specific activity (DPM/mg). During insulin infusion nonsteady state conditions for plasma tritiated glucose specific activity prevailed, and the rate of glucose appearance was calculated using a two-compartment model, as previously described (25). Endogenous hepatic glucose production was determined from the difference between the tracer-determined rate of glucose production and the rate of exogenous glucose infusion rates.
Data analysis
All results are expressed as the mean ± SEM. Statistical differences between hyperglycemic and euglycemic, hyperinsulinemic clamp studies, before and after methyltestosterone treatment, were performed by analysis of variance for repeated measures or paired t test. Comparisons between the control period and the hyperglycemic or euglycemic, hyperinsulinemic steps in each study were also calculated by analysis of variance or paired t test. Significance was defined as P < .05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsReferences |
|---|
The hormonal milieu of the patients at the time of control and
methyltestosterone studies is summarized in Table 1
. Methyltestosterone
administration did not cause any significant changes in the levels of
steroid hormones, protein hormones, or catecholamines.
|
Methyltestosterone administration for 10–12 days had
no significant effect on fasting levels of glucose (82 ± 2
vs. 82 ± 1 mg/dL), insulin (11 ± 1
vs. 10 ± 1 µU/mL), C-peptide (434 ± 35
vs. 463 ± 63 pmol/L), or glucagon (136 ± 21
vs. 136 ± 29 pg/mL). During the hyperglycemic clamp
the plasma glucose concentration was acutely raised and maintained at
203 ± 2 and 201 ± 1 mg/dL (60–120 minute time period)
during the methyltestosterone and the control studies,
respectively (Fig. 1A). The coefficient
of variation in plasma glucose concentration was less than 5% in all
studies. During the hyperglycemic clamp, there was a rapid elevation in
the plasma insulin concentration (Figure 1b). After
methyltestosterone treatment, neither first phase insulin
secretion (59 ± 8 vs. 50 ± 8 µU/mL), second
phase insulin secretion (74 ± 9 vs. 67 ± 9
µU/mL), nor total plasma insulin response (133 ± 16
vs. 117 ± 15 µU/mL), was significantly different in
the methyltesterone and control studies, respectively. First and second
phase C-peptide levels were also not significantly changed after
methyltestosterone treatment (first phase 1166 ± 126
vs. 1101 ± 121 pmol/L methyltestosterone
vs. control; second phase 2090 ± 211 vs.
1882 ± 162 pmol/L).
|
, whole body glucose
uptake was significantly reduced after methyltestosterone
treatment (7.30 ± 0.70 vs. 10.96 ± 1.11
mg/kg/min in control, P < 0.05). Similarly, the ratio
of glucose uptake to the circulating insulin concentration was reduced
in every subject after methyltestosterone administration
(9.4 ± 1.3 vs. 14.3 ± 1.4 mg/kg/min per
µU/mL x 100, P < 0.02).
|
In the control and methyltestosterone studies, the
basal plasma glucose concentrations were 83 ± 1 and 78 ± 1
mg/dL, respectively. The fasting plasma insulin and C-peptide
concentrations were similar in the two groups (11 ± 1
vs. 14 ± 2 µU/mL, and 427 ± 68 vs.
408 ± 36 pmol/mL). During the low-dose insulin infusion, the
steady plasma insulin concentrations were not significantly different;
22 ± 2 and 28 ± 3 µU/mL in the control and
methyltestosterone treated groups, respectively (Fig. 3). During the higher
dose insulin infusion, mean plasma insulin concentrations were also
comparable in the control and methyltestosterone studies,
i.e. 65 ± 5 vs. 69 ± 5 µU/mL,
respectively. The mean plasma glucose concentration during the low-dose
(85 ± 1 vs. 80 ± 2 mg/dL, control vs.
methyltestosterone) and high-dose insulin infusion
(84 ± 1 vs. 81 ± 2 mg/dL) were not significantly
different and varied by less than 5% in all studies.
|
Discussion
The present results demonstrate that-short term androgen administration to nonobese, normally menstruating women caused a significant reduction in whole body insulin sensitivity. Hepatic sensitivity to insulin was not altered by methyltestosterone administration. The insulin resistance was identified under both hyperglycemic and euglycemic hyperinsulinemic conditions, but tended to be more pronounced under hyperglycemic conditions. These findings are consistent with the suggestion that hyperandrogenism may contribute to insulin resistance in women with PCO. However, they do not preclude the possibility of the converse, i.e. hyperinsulinemia might cause hyperandrogenism. The normal response to the development of insulin resistance is a compensatory increase in insulin secretion. The failure to observe a compensation increase in either first or second phase insulin secretion during the hyperglycemic clamp suggests that methyltestosterone administration may also inhibit insulin secretion.
Our observation that androgens induce insulin insensitivity in man are consistent with observations in animals. In oophorectomized female rats, pharmacologic doses of testosterone for 12 weeks caused an 84% reduction in insulin action as assessed by the euglycemic clamp technique (26). The defect in insulin action was explained by diminished 2-deoxyglucose uptake and by glycogen synthesis in type II muscle fibers. In dogs, oral administration of the potent androgenic progestin, levonorgestrel, enhanced the basal rate of gluconeogenesis and impaired the ability of insulin to inhibit hepatic glucose production (27). Unlike the results in dogs, we failed to observe the development of hepatic insulin resistance in the present study. This could be explained by the different androgen (methyltestosterone) used in our studies or by species differences.
Androgen-induced insulin resistance also has been observed in women. In paired hyperglycemic clamp studies conducted before and during Norplant administration (containing levonorgestrel capsules), the rate of insulin-mediated glucose uptake was significantly suppressed during androgen therapy (28, 29). These results are also consistent with reports that have demonstrated a reduction in insulin resistance after antiandrogen therapy (30) or oophorectomy (31). Furthermore, our results are consistent with the report that women receiving intramuscular testosterone esters for four months manifested a reduction in insulin-stimulated glucose utilization at physiological insulin levels, without altering insulin’s ability to suppress hepatic glucose production (32).
Our data are consistant with those of Dunaif et al. (33). Using the euglycemic, hyperinsulinemic clamp technique these investigators demonstrated that both lean and obese women with PCO were insulin resistant compared with weight-matched lean and obese controls with normal androgen levels (33). However, these observations (33) do not demonstrate whether the insulin resistance was the result of elevated androgen levels or of some other characteristics of the PCO syndrome. In our experiments, healthy normally menstruating women were studied, thereby allowing us to more directly demonstrate that hyperandrogenemia can lead to the development of insulin resistance. Elkind-Hirsch et al. (34) treated 12 PCO women with elevated testosterone concentrations with leuprolide acetate. In association with a reduction in their circulating testosterone levels, these women had an improvement in insulin sensitivity as assessed by the minimal model technique.
The collective observations reviewed above strongly implicate androgen excess in the development of insulin resistance under both hyperglycemic and euglycemic, hyperinsulinemic conditions. These observations may have important implications. Epidemiologic data, as well as in vivo and in vitro animal and human studies, have implicated hyperinsulinism and insulin resistance in the development of a metabolic-cardiovascular syndrome characterized by obesity, type II diabetes mellitus, hypertension, hypercoagulability, hyperlipidemia, and atherosclerotic cardiovascular disease (35). Thus, chronic androgen-induced impairment in insulin sensitivity, whether induced by endogenous hypersecretion or exogenous administration, may have undesirable long-term health consequences. These associations suggest that management of PCO women with hyperandrogenism should include attempts to reduce the elevated androgen levels. Similarly, when exogenous sex steroid hormone replacement is indicated, only preparations that contain physiologic amounts of androgens should be administered. Lastly, a number of studies have implicated elevated androgen and sex binding globulins in the development of NIDDM, hyperlipidemia, and central obesity. The present findings raise the intriguing possibility that relative hyperandrogenism may contribute to the insulin resistance and hyperinsulinemia in these common metabolic disorders.
|
|
Footnotes |
|---|
Received September 16, 1997.
Revised July 17, 1998.
Accepted August 14, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsReferences |
|---|
of Contents].
This article has been cited by other articles:
 |
|
 |
|
  H. Rashid, S. Ormerod, and E. Day Anabolic androgenic steroids: what the psychiatrist needs to know Advan. Psychiatr. Treat., May 1, 2007; 13(3): 203 - 211. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. Golden, A. S. Dobs, D. Vaidya, M. Szklo, S. Gapstur, P. Kopp, K. Liu, and P. Ouyang Endogenous Sex Hormones and Glucose Tolerance Status in Postmenopausal Women J. Clin. Endocrinol. Metab., April 1, 2007; 92(4): 1289 - 1295. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Lidfeldt, G. Samsioe, and C.-D. Agardh Obese women and the relation between cardiovascular risk profile and hormone therapy, glucose tolerance, and psychosocial conditions. Diabetes Care, November 1, 2006; 29(11): 2477 - 2482. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Zanolin, F. Tosi, G. Zoppini, R. Castello, G. Spiazzi, R. Dorizzi, M. Muggeo, and P. Moghetti Clustering of Cardiovascular Risk Factors Associated With the Insulin Resistance Syndrome: Assessment by principal component analysis in young hyperandrogenic women Diabetes Care, February 1, 2006; 29(2): 372 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. M C. S M Leao, M. P. C Duarte, D. M. B Silva, P. R. V Bahia, C. M. Coeli, and M. L. F. de Farias Influence of methyltestosterone postmenopausal therapy on plasma lipids, inflammatory factors, glucose metabolism and visceral fat: a randomized study Eur. J. Endocrinol., January 1, 2006; 154(1): 131 - 139. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. Golden, J. Ding, M. Szklo, M. I. Schmidt, B. B. Duncan, and A. Dobs Glucose and Insulin Components of the Metabolic Syndrome Are Associated with Hyperandrogenism in Postmenopausal Women: The Atherosclerosis Risk in Communities Study Am. J. Epidemiol., September 15, 2004; 160(6): 540 - 548. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. C. W. Wu and A. von Eckardstein Androgens and Coronary Artery Disease Endocr. Rev., April 1, 2003; 24(2): 183 - 217. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. S. Zhou, J. A. Rowley, F. Demirovic, M. P. Diamond, and M. S. Badr Effect of testosterone on the apneic threshold in women during NREM sleep J Appl Physiol, January 1, 2003; 94(1): 101 - 107. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. R. Sattler, E. T. Schroeder, M. P. Dube, S. V. Jaque, C. Martinez, P. J. Blanche, S. Azen, and R. M. Krauss Metabolic effects of nandrolone decanoate and resistance training in men with HIV Am J Physiol Endocrinol Metab, December 1, 2002; 283(6): E1214 - E1222. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Poretsky, N. A. Cataldo, Z. Rosenwaks, and L. C. Giudice The Insulin-Related Ovarian Regulatory System in Health and Disease Endocr. Rev., August 1, 1999; 20(4): 535 - 582. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
)
methyl-testosterone.
) and control studies (
).
