| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Nemours Children’s Clinic (N.M., P.B., S.W.), Jacksonville, Florida 32207; Nemours Children’s Clinic (H.Y.H.), Pensacola, Florida 32504; and All Children’s Hospital (D.S.), St. Petersburg, Florida 33701
Address all correspondence and requests for reprints to: Nelly Mauras, Nemours Children’s Clinic, 807 Childrens Way, Jacksonville, Florida 32207. E-mail: nmauras{at}nemours.org
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: Our objective was to study the metabolic effects of oral vs. TD estrogen in GH-treated girls with Turner syndrome.
Design and Methods: Eleven girls with Turner syndrome, mean age 13.4 ± 0.5 (SE) yr, on GH for at least 6 months were recruited. Studies included [13C]leucine and d5-glycerol infusions, indirect calorimetry, dual-emission x-ray absorptiometry, and hormone and substrate measurements. They received, in random order, 17ß-estradiol orally (0.5, 1, and 2 mg for 2 wk each) and TD (0.025, 0.0375, and 0.05 mg for 2 wk each), and studies were repeated after each 6-wk course with 4 wk washout in between.
Results: Rates of whole-body protein turnover, oxidation and synthesis, lipolysis, lipid and carbohydrate oxidation, and resting energy expenditure were unaffected by either form of estrogen; nor were lipids, insulin, and fibrinogen concentrations affected. Plasma IGF-I concentrations did not change clinically significantly with either form of estrogen, despite higher estrogen concentrations after oral estrogen. Estradiol concentrations did not correlate with any variables measured.
Conclusions: In GH-treated girls with Turner syndrome, neither oral nor TD estrogen adversely affected rates of protein turnover, lipolysis, and lipid oxidation rates or plasma lipids, fibrinogen, or fasting insulin concentrations. There was no clinically significant change in IGF-I concentrations after either form of estrogen. In aggregate, these data suggest that the route of delivery of estrogen does not adversely affect these metabolic effects of GH in young girls with Turner syndrome.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
These findings, however, have not been universally confirmed, and oral estrogen is not necessarily deleterious. In a study comparing two different oral preparations of estrogen vs. TD estrogen, investigators concluded that oral estradiol but not TD prevents the total-body fat increase associated with menopause and that it also prevents the shift from peripheral to central fat distribution observed in that age group (9). The latter pattern of fat distribution is classically associated with increased risk for cardiovascular disease. In adult women with Turner syndrome treated with either oral or TD estrogen and norethisterone for 6 months, similar relative declines in total IGF-I concentrations were observed, yet no differences in the bioavailable free IGF-I concentrations were noted using either route of estrogen treatment (10). In addition, despite higher concentrations of different hepatic-derived binding proteins, young girls with Turner syndrome had similar concentrations of IGF-I, IGF-binding protein 3 (IGFBP-3), and urinary GH after administration of low-dose oral and TD estrogen for 1 month each (11).
We designed these studies to investigate the differential effects of estrogen delivered via oral vs. TD routes in GH-treated hypogonadal girls with Turner syndrome, specifically, its effects on protein and fat metabolism as well as on growth factor concentrations. We hypothesized that in hypogonadal girls with Turner syndrome, oral estrogen would cause a decrease in rates of lipolysis and lipid oxidation, resulting in a less desirable metabolic profile and lower IGF-I concentrations, similar to the reported effects in postmenopausal women.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Study subjects
Eleven girls with Turner syndrome (45X and related karyotypes), between the ages of 11 and 15 yr, were recruited after informed written consent from their parents and the child’s assent when appropriate. Study subjects were referred to the Nemours Children’s Clinic in Jacksonville from several pediatric endocrine clinics and through web-based advertising of the protocol through the Turner Syndrome Society web page. By design, all were on stable daily GH doses for at least 6 months before the studies as well as on adequate thyroid replacement therapy when needed. All were hypogonadal as indicated on physical exam and by elevated FSH concentrations. Subjects with significant overweight (body mass index > 30 kg/m2) or history of systemic illness were excluded from participation.
Study design
Each girl was admitted three separate times to the Wolfson Children’s Hospital Clinical Research Center. For 3 d before admission, the patients consumed a weight maintenance diet based on the patients’ dietary history. The afternoon of admission, a 3-min step test was performed to determine maximal heart rate (HR). Because 70% increase in HR correlates with about 50% maximal oxygen consumption (VO2max), we used that increase above resting HR as the maximal HR tolerated during the exercise study the next day (12). Body composition was assessed using the sum of skinfold thickness and dual-emission x-ray absorptiometry (Hologic QDR 4500A; Hologic, Waltham, MA). Patients were allowed nothing by mouth except for water after 2100 h. The next morning (D1) at 0600 h, two forearm veins were cannulated, one kept heated for arterialized frequent blood sampling (13) and the other for the administration of stable isotopes. A primed, 240-min infusion of L-[1-13C]leucine was started (
7.6 µmol/kg; 0.13 µmol/kg·min), and at 180 min, an infusion of [1,1,2,3,3-2H5]glycerol (d5-glycerol) was added (1.3 µmol/kg; 0.10 µmol/kg·min) and continued through 300 min (120-min infusion). Frequent blood and breath samples were obtained for isotopic enrichment measurements as well as hormones, growth factors, lipids, and substrates. Indirect calorimetry was performed three times using a CPX-MAX calorimeter to measure substrate oxidation rates using a mouthpiece. At 240 min, after completion of the [1-13C]leucine infusions, subjects were exercised at half-maximal capacity for 60 min using bicycle ergometry to stimulate lipolysis during the d5-glycerol infusion.
After the baseline study, the girls were started on incremental daily doses of oral estrogen (17ß-estradiol, Estrace; Bristol-Myers Squibb, New York, NY) at 0.5, 1, and 2 mg, each dose given for 2 wk and the study repeated again after 6 wk (D2). This was followed by a 4-wk washout. Blood and body composition analyses were done and were followed by initiation of TD estrogen (17ß-estradiol, Vivelle TD system; Novartis Pharmaceuticals, East Hanover, NJ) at 0.025, 0.0375, and 0.05 mg applied twice weekly at 2-wk increments for another 6 wk, and the study was repeated a third time (D3) identical to baseline. The treatment order (oral/TD vs. TD/oral) was randomized (Fig. 1
). Subjects were kept on daily GH throughout the duration of the study at about 0.4 mg/kg·wk given sc at bedtime.
|
-[13C]Ketoisocaproic acid, the -ketoacid of leucine, was measured by electron impact gas chromatography/mass spectrometry (5890/5970 Hewlett-Packard GCMS; Hewlett-Packard, Palo Alto, CA) using N-methyl-N-t-butyldimethysilyl trifluoroacetamide as derivative and 13CO2 enrichments using a dual-inlet isotope ratio mass spectrometer (VG Optima SIRA 3GC; Waters Corp., Milford, MA) in our lab as described (14, 15). The d5-glycerol analysis was performed using Tris-heptafluorobutyrl ester derivatives and negative ion chemical ionization gas chromatography/mass spectrometry (6890/5973 Hewlett-Packard), as described previously (16, 17). Free fatty acids were measured by a colorimetric assay. IGF-I, IGFBP-3, and insulin concentrations were measured by standard immunometric assays, and plasma lipids were measured on the Cobas Mira using cholesterol reagent from Roche Diagnostics (Somerville, NJ). Fibrinogen concentrations were measured by nephelometry. The urea nitrogen concentrations in urine were measured by Paramax urea nitrogen reagent (Baster; DuPont Clinical Systems Division, Wilmington, DE), and plasma glucose concentrations were measured by a glucose analyzer (Beckman Instruments, Palo Alto, CA). Plasma estradiol concentrations were measured in our lab by RIA using a third-generation estradiol assay with a sensitivity of 0.8 pg/ml (Diagnostic Systems Laboratories, Webster, TX). Estrone concentrations were also measured by RIA. There was an approximately 7% cross-reactivity with estrone in the estradiol assay.
Materials, hormones, and drugs
L-[1-13C]Leucine (99% enriched), d5-glycerol (99% enriched) were purchased from Cambridge Isotope Laboratories (Billerica, MA), dissolved in sterile saline, and filtered through a 200-µm Millipore (Bedford, MA) filter in sterile vials. An aliquot from each lot was cultured for pathogens and tested with the limulus lysate assay before human use.
Calculations
For leucine kinetics, the plasma enrichment of -ketoisocaproic acid was used as an index of the intracellular enrichment of leucine (reciprocal pool model), and estimates of whole-body protein metabolism were made using steady-state equations as described previously (18, 19). Leucine’s rate of appearance (Ra) was used as an index of proteolysis and the nonoxidative leucine disposal as an index of whole body protein synthesis. Glycerol’s Ra was calculated as Ra = I x (Ei/Ep – 1), where I is the infusion rate of the tracer, Ei and Ep the enrichment of the infusate and glycerol tracer in plasma at steady state (16). Glycerol turnover was calculated both from the enrichments the hour before and the hour during exercise. Indirect calorimetry data were used to estimate rates of glucose, protein, and lipid oxidation using gas exchange equations (20).
Statistical analysis
Repeated-measures analysis of covariance with appropriate post hoc comparison procedures was used to test for differences between baseline, oral estrogen administration, and TD estrogen administration for major study variables, including hormones, growth factors, and kinetic data. Order of treatment was used as a between-subjects factor in all analyses. Baseline estradiol concentrations were used as a covariate because of moderate correlations found between baseline estradiol and other study parameters. In addition, age was added as a second covariate in the analysis of body composition due to the influence of this factor on growth. Paired t tests were used to test for significant differences between pre- and postexercise glycerol turnover. Significance was established at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
.
|
57 µg/kg·d), and this was continued throughout the study. Fat and protein kinetics
Glycerol concentrations before and after exercise, respectively, were as follows (µmol/liter): baseline, 81 ± 10 and 218 ± 30; oral, 97 ± 13 and 190 ± 30; TD, 82 ± 8 and 152 ± 21 (all comparisons P < 0.01). Comparison of glycerol concentrations among treatment groups showed a significant difference after exercise between baseline and TD treatment with lower glycerol concentrations TD (P = 0.01). However, rates of whole-body lipolysis, as measured by glycerol turnover, were not different between treatment arms. Rates of lipolysis increased significantly during exercise as expected, but there were no differences in these rates before or after estrogen regardless of the route of administration when expressed as micromoles per kilogram per minute or micromoles per kilogram fat mass (FM) per minute [before and after exercise, respectively (µmol/kg FM·min): baseline, 12.7 ± 1.8 and 23.5 ± 2.4 (P = 0.0003); oral, 17.2 ± 2.1 and 26.8 ± 4.3 (P = 0.01); TD, 16.4 ± 3.0 and 22.2 ± 3.4 (P = 0.006)]; P = not significant (NS) among the groups before or after exercise (Fig. 2). Rates of whole-body proteolysis, oxidation, and synthesis were not altered by estrogen, regardless of the route of administration (Table 2).
|
|
Lipid oxidation rates were also unaffected by estrogen treatment, as were the rates of carbohydrate oxidation and resting energy expenditure (Table 3). There was also no change in plasma lipids (high-density lipoprotein, low-density lipoprotein, and total cholesterol), fibrinogen or fasting insulin concentrations during either form of treatment (Table 4).
|
|
Because body composition changes can be naturally affected by the length of the studies, we performed the analysis based solely on chronological treatment order. There were no changes in adiposity (percent FM) over the course of estrogen therapy regardless of route of administration (Table 5).
|
Plasma IGF-I concentrations decreased modestly after oral estrogen but were not significantly affected by TD delivery (ng/ml): baseline, 916 ± 43; oral, 790 ± 57 (P = 0.01 vs. baseline, paired t test); TD, 808 ± 70 (P = 0.18 vs. baseline) (Fig. 3). [To convert IGF-I to SI units (nmol/liter) multiply by 0.131; to convert IGFBP-3 to nmol/liter, multiply by 0.035.] However, ANOVA of the IGF-I concentrations after the different forms of delivery was not significant (P = 0.6). IGFBP-3 concentrations were unchanged by estrogen treatment (mg/liter): baseline, 49 ± 2; oral, 43 ± 2; TD, 44 ± 1 (P = NS) (Table 4
).
|
This lack of differences in metabolic effects on protein and fat metabolism after the oral route was remarkable because estradiol concentrations were three times higher after the oral route (pg/ml): at baseline, 22 ± 3; after oral, 280 ± 43; washout, 31 ± 3; TD, 79 ± 5 (to convert estradiol to pmol/liter, multiply by 3.671). We performed analysis of covariance and ran repeated-measures ANOVA using the estradiol concentration as a covariate. The estradiol concentration did not show any effect on any of the dependent variables measured here and reported above. Estrone concentrations were also higher after the oral route (pg/ml): baseline, 16.8 ± 4.9; oral, 164 ± 65; washout, 13.4 ± 3.9; TD, 18.7 ± 5.5.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
These studies show lack of an estrogen effect, regardless of route, on overall protein anabolism during induction of puberty in GH-treated girls, as it was shown using low-dose estrogen given alone in similar girls reported previously (21). Although low doses of estrogen can promote linear growth and hence protein anabolism, the whole-body protein turnover rates measured by stable isotopic infusions overall reflect nitrogen turnover in different protein pools, the largest of which is skeletal muscle. We have measured whole-body kinetics in girls with Turner’s using very low doses of ethinyl estradiol (100 ng/kg·d, or 3–4 µg/d), the same doses reported to be growth promoting (22, 23, 24), and observed no protein-anabolic effect using those doses either. However, using the same isotope infusion methods, comparable size cohort, and comparable lengths of treatment, we observed significant protein-anabolic effects using testosterone (25), GH (26), and IGF-I (27), suggesting that estrogen, regardless of dose and route, does not affect the predominant protein pool we can measure, skeletal muscle. They also support data by Gravholt et al. (28) in a study of girls with Turner syndrome in which growth factors, insulin, glucose, lipolytic and gluconeogenic precursors, glucose tolerance, and assessment of body composition and bone mineral content were measured. They showed that the coadministration of GH and oral estrogen (17ß-estradiol) did not compromise any of the metabolic effects of GH measured (28).
We also measured the rate of lipolysis using stable isotope methods as well as rates of lipid oxidation. As expected, exercise markedly increased the rates of lipolysis; however, there were no differences in lipolysis rates with either oral or TD estrogen administration and no changes in the oxidation of substrates or lipid and insulin concentrations. These studies of estrogen and GH in girls contrast with the synergistic effect of testosterone observed when given in combination with GH in young GH-deficient boys reported by us previously (29), highlighting the more potent effect of androgens in measures affecting protein anabolism and lean mass accrual.
The results here differ from the described decrease in lipid oxidation and whole-body protein synthesis rates reported in GH-deficient women treated with both medications reported and summarized previously (7). Cook et al. (30) also reported that when using oral estrogen, the dose of GH used in GH-deficient states needs to be increased. A first-pass effect of oral estrogen by the liver has been postulated as a principal mechanism for the reported effects of oral estrogen inhibiting IGF-I production (7) as well as up-regulation of suppressors of cytokine signaling (SOCS)-2 and 3, which act as negative regulators of GH in the liver by inhibiting GH signaling (31). These young girls with Turner syndrome had no clinically significant changes in plasma IGF-I concentrations with oral estrogen with significant overlap in the responses (Fig. 3). This is similar to data in a small study comparing oral ethinyl estradiol vs. TD estradiol in which despite lower concentrations of different hepatic-derived binding proteins after oral estrogen, plasma IGF-I concentrations were found to be similar in both groups, with no GH used (11). Despite the fact that the estradiol concentrations achieved here were higher than the TD route, the differences in IGF-I concentrations after the oral route (790 ± 57 ng/ml) and after TD (808 ± 70 ng/ml), although statistically lower in the oral route compared with baseline, are likely inconsequential clinically, considering the wide range of normal IGF-I concentrations in children and the variability of the response. In adult women with Turner syndrome, investigators observed similar mild decreases in plasma IGF-I concentrations but no changes in free IGF-I depending on the route of estrogen therapy and a normalization of IGFBP-3 proteolysis rates with either form of estrogen (10, 32).
The dose of GH used in these girls on average was about 57 µg/kg·wk, which is much higher than those used in all the reported adult studies. It is possible that the lack of substantial effect of oral vs. TD estrogen on whole-body protein rates, lipolysis, and lipid oxidation and modest effect described on IGF-I concentrations is due in part to the potent protein-anabolic and considerable lipolytic effects of GH, which, after at least 6 months of prior daily administration, have now stabilized. These GH effects likely override the estrogen effects in the measured parameters.
Measurement of fibrinogen concentrations has become an important marker for cardiovascular disease. Studies in adult women on hormone replacement therapy have shown either no change or a decrease in fibrinogen concentrations after estrogen (33, 34, 35), the latter mostly after oral, not TD use (35). Interestingly, in the cohort studied here, fibrinogen concentrations were normal at baseline, similar to adult women with Turner syndrome reported previously (36), and were not affected by either form of estrogen. The results are also congruent with findings in postmenopausal women treated short-term with either oral or transvaginal estrogen, where no differences in hemostatic factors were observed depending on route of treatment (37). Collectively, the data suggest that neither oral nor TD estrogen negatively affect measures of fibrinogen, a measure of cardiovascular risk.
Does the route of delivery matter in feminizing girls with Turner syndrome? Based on our data and those of others, probably not if concomitantly treated with GH. Our results nonetheless have several possible limitations to consider. One is the relatively short-term nature of these experiments. Although it is possible that longer-term treatment could have unraveled differences between the two routes of estrogen delivery, the changes in lipid oxidation and lipid turnover reported previously in GH-deficient postmenopausal women were observed after 8-wk experiments using each route (8), not dissimilar from the 6-wk experiments for each oral vs. TD route reported here. Based on the significant changes in protein and lipid kinetics using stable isotope methods observed after only 4 wk of testosterone and/or GH in children (25, 29) and after 8 wk of GH or IGF-I in young adults (26, 27), we think that 6 wk is likely enough time to have observed any changes in these same parameters with estrogen therapy.
Could the doses of oral estrogen used have masked the differences in effects vs. the TD route? Most studies reporting the differential effects of estrogen delivery on different metabolic parameters used different types of estrogen preparations such as conjugated estrogens orally and 17ß-estradiol TD. In the last 2 wk of each treatment arm, we chose to administer estradiol doses based on available pharmacokinetic data to achieve plasma estradiol concentrations in the early follicular phase of the menstrual cycle. However, we achieved much higher estradiol concentrations after oral estrogen than after the TD route, similar to those achieved in another group of girls with Turner syndrome treated with GH and 17ß-estradiol orally for 2 months (28). Despite these higher concentrations after the oral route, the expected results of lower protein synthesis rates and lower lipid oxidation rates after oral estrogen were not observed in these experiments, making the lack of adverse effect of oral estrogen in the measured parameters even more compelling. Although the doses chosen raised the concentrations of estradiol to pubertal levels, these are the type of doses these girls will need long term for feminization, a fact that makes these results highly relevant.
Lastly, could diurnal variation in metabolic effects be partly responsible for the lack of differential effects of estrogen observed? The critical factor to compare the data gathered here with those in postmenopausal women more than the time of the day is the relation to meals. These studies were carried out in the morning in the postabsorptive state using isotopic infusions of leucine, similar to studies by Wolthers et al. (8) in postmenopausal women. Hence, it is unlikely that diurnal changes in protein balance contributed to these different results. However, the reported decrease in lipid oxidation with oral estrogen after menopause was not observed in the postabsorptive state but only 30–60 min after a mixed liquid meal (3). We did not measure postprandial substrate oxidation in these studies; hence, it is possible such differences in estrogen effects could have been missed. However, the measurement of postabsorptive oxidation rates is complex and dependent in part on the relative adiposity and insulin sensitivity of the individuals (38). On the other hand, we measured lipolysis rates using d5-glycerol infusions before and after exercise and did not detect any differences in the lipolytic response to exercise depending in the estrogen route. These are the first studies, to our knowledge, measuring rates of whole-body kinetics using different routes of estrogen. These patients were investigated while getting GH; hence, the effects of different routes of estrogen without concomitant GH in this population would require a different study. To date, there has been no prospective, randomized trial using oral vs. TD estradiol in girls with Turner syndrome comparing their long-term linear growth. This too requires additional studies.
In conclusion, in GH-treated girls with Turner syndrome, neither oral nor TD estrogen adversely affected measures of whole-body anabolism, including protein turnover, lipolysis, and lipid oxidation rates. We also observed no significant differences in plasma lipids, fibrinogen, and fasting insulin concentrations. Although significant statistically, there was no clinically significant change in IGF-I concentrations after either form of estrogen. Whether longer-term exposure to either form of estrogen could translate into differences in linear growth requires further study. Taken in aggregate, these data suggest that the route of delivery of estrogen does not adversely affect these metabolic effects of GH in young girls with Turner syndrome.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Summary: D.S., H.Y.H., P.B., and S.W. have nothing to declare. N.M. consults for Genentech and AstraZeneca and has received grant support from Eli Lilly, Genentech, and Serono and received lecture fees from Serono, Genentech, and Pfizer.
First Published Online August 21, 2007
Abbreviations: d5-glycerol, [1,1,2,3,3-2H5]Glycerol; FM, fat mass; HR, heart rate; IGFBP-3, IGF-binding protein 3; NS, not significant.
Received March 26, 2007.
Accepted August 14, 2007.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  Z. M. Nabhan, L. A. DiMeglio, R. Qi, S. M. Perkins, and E. A. Eugster Conjugated Oral versus Transdermal Estrogen Replacement in Girls with Turner Syndrome: A Pilot Comparative Study J. Clin. Endocrinol. Metab., June 1, 2009; 94(6): 2009 - 2014. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. E. Hjerrild, K. H. Mortensen, and C. H. Gravholt Turner syndrome and clinical treatment Br. Med. Bull., June 1, 2008; 86(1): 77 - 93. [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 |
