| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Gynecological Endocrinology and Reproductive Medicine (S.E., N.S., A.G., M.v.W., T.S.), Obstetrics and Gynecology (S.E.), and Internal Medicine (A.H.), Division of Endocrinology and Metabolism, and Biostatistics (V.H.), University of Heidelberg, 69115 Heidelberg, Germany
Address all correspondence and requests for reprints to: S. Eisenhardt, M.D., Women’s University Hospital, Department of Gynecological Endocrinology and Reproductive Medicine and Department of Obstetrics and Gynecology, Vossstr. 9, 69115 Heidelberg, Germany. E-mail: stefan.eisenhardt{at}med.uni-heidelberg.de.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: The objective of the study was to evaluate the early potential effects of metformin treatment, their time of onset, and the role of IR on metformin’s efficacy.
Design: This was a prospective randomized, double-blind, placebo-controlled trial.
Setting: The study was conducted at the University of Heidelberg, Heidelberg, Germany.
Patients: The patient population was 45 oligo-/anovulatory PCOS women with typical ovaries.
Interventions: Women were stratified for IR (32 of 13) and then randomly allocated to receive either metformin (n = 22) or placebo (n = 23) and were assessed before and every 4 wk within a treatment period of 12 wk.
Main Outcome Measures: Menstrual disturbance and markers of insulin metabolism were measured.
Results: The main outcome criterion menstrual disturbance was successfully improved in the metformin-treated group, depending on IR (12 of 15 vs. three of 17), whereas women without IR (four of seven vs. four of six) had no significant amelioration of their menstrual irregularities (P < 0.05). Estradiol levels increased continuously only in the treatment group (P < 0.005), indicating an improvement of ovulatory function. Sixty-seven percent of metformin-treated women had at least one ovulation, compared with only 45% in the placebo group, shown by biphasic body temperature curves. Insulin sensitivity improved within 4 wk after beginning of metformin as shown by an increased area under the curve glucose to insulin ratio, compared with baseline (P < 0.005).
Conclusions: IR is a baseline predictor of clinical efficacy in metformin treatment in PCOS women measured by improved menstrual cyclicity and ovulatory function.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A link between disturbed insulin action and PCOS was first highlighted in 1980 (6), and subsequent studies have shown that insulin resistance is an integral feature of PCOS (3, 7, 8, 9, 10). The associated hyperinsulinemia may promote abnormal ovarian androgen secretion and therewith abnormal follicular development, leading to dysfunctional ovarian and menstrual activity (3, 7, 8, 11, 12). As a consequence of insulin resistance, women with PCOS exhibit a greater risk for dyslipidemia (13), coagulation disorders and endothelial dysfunction (14), increased incidence of hypertension (15), and type 2 diabetes mellitus (16, 17) in later life, which are strong and established risk factors for cardiovascular disease (18).
Improved understanding of the pathogenesis and the recognition of the critical role of hyperinsulinemia have provided the rationale for trials examining the therapeutic value of insulin-sensitizing agents like metformin (19). Meanwhile metformin treatment has been described extensively (20, 21, 22, 23, 24, 25), although the possible early onset of metformin-related effects has not yet been studied in a controlled study.
Therefore, the trial was planned to characterize early potential effects of metformin and the exact point of time of metformin’s efficacy in PCOS. Moreover, the study was designed to establish basic knowledge concerning subsequent therapeutic concepts in patients with PCOS with or without insulin resistance.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Study population
Women with PCOS (n = 45), aged 21–36 yr, whose chief complaints were menstrual disturbances and infertility and/or clinical signs of hyperandrogenism (e.g. hirsutism and acne) were recruited between 2002 and 2004 from the Department of Gynecological Endocrinology and Reproductive Medicine of the Women’s University Hospital Heidelberg (Heidelberg, Germany). The diagnosis of PCOS was based on at least two of the three following abnormalities: according to our including criteria all patients were expected to have disturbed ovulatory function with chronic oligomenorrhea (cycle length > 35 d; less than nine cycles per year) or amenorrhea (cycle length > 12 wk) and typical appearance of polycystic ovaries by ultrasound according to the criteria of the Rotterdam consensus meeting 2003 (2), which all patients fulfilled. Facultatively patients could have clinical and/or biochemical signs of hyperandrogenism [serum total testosterone (T) concentration 
60 ng/dl or greater (2.1 nmol/liter) or serum androstenedione (A) concentration greater than 2.9 ng/ml (>10.1 nmol/liter)]. The body mass index (BMI) was not considered as inclusion or exclusion criterion (BMI between 18 and 49 kg/m2, median 31 kg/m2). A fasting glucose to insulin ratio (FGIR) less than 4.5 (26) indicating insulin resistance was observed in 32 of the 45 patients. The presence of the following disorders was excluded by specific laboratory tests: impaired glucose tolerance test (fasting glucose > 5.6 mmol/liter and/or 2-h glucose > 7.8 mmol/liter) or any form of diabetes mellitus, hyperprolactinemia, thyroid disorders, late-onset congenital adrenal hyperplasia (exclusion of 21-hydroxylase deficiency by molecular genetic analysis), and Cushing’s syndrome (normal basal free serum cortisone and 2-mg dexamethasone suppression test).
None of the women had taken any medications likely to influence hormonal profiles or antiobesity compounds during 6 months before inclusion in the study. We confirmed absence of heart, liver, or kidney diseases (predisposing lactic acidosis) and unsuspected pregnancy in all participants before inclusion in the study. Some patients prematurely dropped out of the study in case of diagnosis of pregnancy or because of loss to follow-up.
Study design
Treatments. Patients were randomly allocated to one of two groups: group 1 (n = 22) received metformin, group 2 (n = 23) received placebo. Metformin and placebo were provided by Lipha S.A. [Pharmacie Centrale (Clinical Trial Supply Group), Meyzieu, France]. Metformin was administered at a dosage of 3 x 500 mg daily, except for the first week of treatment when 500 mg were given only twice a day to reduce the incidence and severity of gastrointestinal side effects.
Randomization was done in a prospective, placebo-controlled, double-blind fashion stratified for insulin resistance. Patients received either metformin or placebo according to computer-generated code with a randomization in blocks of six. A copy of the code was stored in a sealed envelope by a third party who did not participate in the study for emergency situations. The randomization code was not broken until the last patient completed all observations.
Patients were advised to use barrier contraception if fertility was not desired and were carefully instructed to stop taking the drug immediately on confirmation of pregnancy.
Assessment program.
All patients underwent clinical, metabolic, and hormonal evaluations at baseline and in regular intervals of 4 wk during the whole treatment period of 12 wk after randomization. In the baseline study, vaginal bleeding was induced by progesterone withdrawal or spontaneous bleeding. Clinical assessment included menstrual cycle frequency, basal body temperature curve, height, weight, BMI, and hirsutism. The following studies were performed on d 2–5: After a 12-h overnight fast, a nonheparinized venous blood sample was obtained between 0800 and 0900 h to measure the circulating concentrations of prolactin, LH, FSH, estradiol, total T, SHBG, progesterone, TSH, total T3, free T4, dehydroepiandrosterone sulfate (DHEA-S), A, 17
-hydroxyprogesterone (17-OHP), cortisol, fasting glucose and insulin, total cholesterol (TC), triglycerides, high-density lipoprotein (HDL) cholesterol, and low-density lipoprotein (LDL) cholesterol. As a monitor of general drug safety, complete blood count, hepatic function tests [including alanine aminotransferase (ALT), aspartate aminotransferase (AST), and alkaline phosphatase], and renal chemistry were ascertained to recognize and prevent possible metformin-induced complications.
After obtaining the basal blood sample, a 2-h oral glucose tolerance test (OGTT) was immediately performed with an oral glucose load of 75 g, and nonheparinized blood samples were obtained after 30, 60, 90, and 120 min to measure serum glucose and insulin concentrations.
ACTH-stimulation test: blood samples were drawn at 0 and 60 min for cortisol, dehydroepiandrosterone (DHEA), A, hydroxyprogesterone, deoxycortisol, and hydroxypregnenolone after iv administration of 0.25 mg ACTH (Synacthen, Novartis, Nürnberg, Germany).
GnRH-stimulation test after suppression with dexamethasone. Blood samples were collected at 0, 30, and 60 min and 24 h for A, hydroxyprogesterone, T, cortisol, and hydroxypregnenolone as well as LH and FSH at 0 and 60 min after a single sc administration of 0.5 mg of native GnRH (Decapeptyl, Ferring, Kiel, Germany). The evening before, dexamethasone was administered at a dosage of 2 mg orally.
ACTH- and GnRH-stimulation tests were repeated at the 12-wk assessment point.
Every 4 wk during the 3-month study period, the baseline clinical, metabolic (including the OGTT), and hormonal evaluations were performed (except for thyroid function tests, ACTH function, and dexamethasone function test, which were reassessed only at the end of the study period) or on d 2–5 of a spontaneous menstrual bleeding.
Methods
Improvement of cycle disorders was defined as a change among clinically classified cycle groups (amenorrhea/oligomenorrhea/eumenorrhea) or a reduction in cycle length of at least 4 wk or pregnancy.
BMI was calculated using the equation: weight (kilograms)/height (meters)2. Hirsutism was clinically evaluated using the Ferriman-Gallwey (F-G) score (assessment of 11 body areas); a score greater than 8 was defined as hirsutism (27).
Using the serum glucose and insulin concentrations during fasting and the 2-h OGTT, we calculated the following parameters:
A. FGIR (3, 26) = fasting serum glucose concentration (milligrams per deciliter)/fasting serum insulin concentration (microinternational units per milliliter).
B. Homeostasis model assessment for insulin resistance (HOMA-IR) (28, 29, 30) = fasting serum insulin (microinternational units per milliliter) x fasting serum glucose (millimoles per liter)/22.5.
C. Quantitative insulin sensitivity check index (QUICKI) (29, 30, 31) = 1/[log (I0) + log (G0)], where I0 = fasting serum insulin (microinternational units per milliliter) concentration and G0 = fasting serum glucose (milligrams per deciliter) concentration.
D. ß-Cell function index (30, 32) = (20) x [fasting serum insulin (microinternational units per milliliter) concentration/fasting serum glucose concentration (millimoles per liter)] – 3.5.
E. Area under the insulin curve (30, 33) using a trapezoidal method (AUC-insulin = insulin 0' x 0.25 + insulin 30' x 0.5 + insulin 60' x 0.75 + insulin 120' x 0.5) and the same for the area under the glucose curve (AUC-glucose).
Assay methods
Concentrations of estradiol, FSH, LH, TSH, HCG, PRL, and T were analyzed using commercially available immunoradiometric kits [electrochemiluminescence immunoassays were purchased from Boehringer (Mannheim, Germany)] and analyzed on the Elecsys immunoassay analyzer (Roche Molecular Biochemicals, Mannheim, Germany). Electrochemiluminescence immunoassays for SHBG, DHEA-S, and progesterone were purchased from Roche Diagnostics GmbH (Mannheim, Germany). ELISA for A, 17-OHP, and DHEA were purchased from DRG Instruments GmbH (Marburg, Germany). These previously mentioned assays were performed in the endocrine laboratory of the Department of Gynecological Endocrinology and Reproductive Medicine.
The serum levels of the following hormones were analyzed in the steroid laboratory of the Department of Pharmacology: 17-OHP, DHEA, DHEA-S, A, T, dihydrotestosterone, and 21-deoxycortisol. These steroids were measured after extraction and chromatographic purification by RIA using specific antibodies. Intra- and interassay coefficients of variation were less than 8.1 and 9.8%, respectively.
Plasma glucose levels were assayed by the glucose hexokinase method using an automatic autoanalyzer (ADVIA 2400, Bayer Diagnostics, Giessen, Germany). The intraassay coefficient of variation was 1.8% or less. Insulin was measured by RIA.
Serum TC, triglycerides, HDL, LDL, ALT, AST, and alkaline phosphatase were measured using an automatized autoanalyzer (ADVIA 2400, Bayer Diagnostics). The intraassay coefficients of variation were 2.3% or less.
Role of the pharmaceutical company providing the drugs under study
The pharmaceutical company providing the drugs was not involved in the study design, data collection, data analysis, data interpretation, or writing the report. No funding of any kind was received to perform the study or by any of the participants in the study.
Statistical analysis
All data are presented as median and range (1 to 3 quartile) unless otherwise noted. Descriptive statistics were used for continuous data at each visit. The data were evaluated in an intent-to-treat analysis. The success of the treatment was evaluated by logistic regression. Treatment, insulin resistance, their interaction, and the age of the patients were chosen as covariates. P < 0.05 was considered to indicate statistical significance. Changes in parameters over time were assessed using repeated-measures ANOVA, and differences between groups were evaluated using a nonparametric test (Mann-Whitney). Within-group differences were analyzed by the paired Student’s t test/sign test. Data analysis was performed using the Statistical Analysis System SAS (SAS Institute Inc., Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We were able to analyze complete data from 38 of the 45 women. Three of the patients became pregnant during the first (one metformin/one placebo group) and second (one metformin group) months of treatment. The difference in the dropout rates (excluding pregnancies) between the placebo (n = 23) and treatment group (n = 22) was remarkable (placebo three/metformin one). All women lost to follow-up withdrew from the study for reasons unrelated to the use of metformin (socioeconomic reasons). Treatment compliance was good and most of the patients did not report any drug-related adverse effects. No abnormalities were observed in complete blood counts or renal or liver function tests after treatment with metformin. The main outcome variable improvement of cycle length was analyzed in an intent-to-treat fashion so that all 45 patients could be evaluated in a logistic regression model. The final number of patients was 19 in the metformin group and 19 in the placebo group.
No significant differences in clinical baseline characteristics between both groups could be observed (Table 1
). During 3 months of treatment with metformin, no significant changes in body weight or BMI could be observed. All subjects had polycystic-appearing ovaries (PAO) on transvaginal ultrasound according to the criteria mentioned above. Ultrasound scans were performed every 4 wk with no significant decrease in PAO or ovarian volume. The F-G score remained unchanged in both groups after the 3-month treatment period.
|
Thirty-two women were found to have insulin resistance (15 patients of the metformin group/17 patients of the placebo group), whereas 13 patients did not show IR (seven/six).
Stratified by insulin resistance (FGIR < 4.5), the main outcome criterion menstrual disturbance was successfully improved in the metformin-treated group, depending on the presence of insulin resistance (12/15 vs. three/17) whereas women without insulin resistance (four/seven vs. four/six) had no significant amelioration of their menstrual irregularities. In a logistic regression analysis, this result showed statistical significance (P < 0.05).
Before treatment, no woman had regular menses. Eight/13 (metformin/placebo) women were oligomenorrheic and 14/10 were amenorrheic. After 3 months of treatment, eight/two women got their cycles totally restored to eumenorrhea (cycle length 28 ± 7 d), eight/11 women were oligomenorrheic, and four/nine women still had an amenorrhea. Most of the subjects converted only one step (e.g. amenorrhea to oligomenorrhea); a two-step conversion was rare.
Tables 2 and 3 present the subjects’ metabolic parameters. None of the women showed any degree of glucose intolerance at baseline because this was an exclusion criterion. At baseline, metabolic parameters were comparable in both groups. All measures of insulin-related parameters, like FGIR, HOMA-IR, QUICKI, and ß-cell function, were similar in both groups. There was also no difference in levels of fasting glucose, fasting insulin, and for AUC-glucose. AUC-insulin showed a trend toward reduction in the metformin group but was not found to be statistically significant.
|
|
). Other markers of insulin resistance (HOMA-IR) and insulin sensitivity (QUICKI) did not show changes with statistical significance.
|
Hormonal characteristics
Tables 4 and 5 present hormonal parameters. The baseline hormone concentrations were similar in both groups. Improvements in insulin metabolism were interestingly not accompanied by statistically significant improvement of clinical or biochemical signs of hyperandrogenism. Twenty-one of the 45 subjects (47%) had hyperandrogenism (increased F-G score and/or increased serum T or A). Circulating androgen concentrations and median SHBG levels remained unchanged throughout the whole study period. In contrast, basal LH slightly increased in the first 2 months of the study in both groups and was still seen high at the end of the trial only in the placebo group. Basal FSH and stimulated LH and FSH levels were unaffected by metformin. Estradiol levels continuously increased every 4 wk with evidence in the treatment group only (P < 0.005). Progesterone levels were measured only in the 4-wk intervals of hormonal assessment, which was clearly not a sufficient method to prove ovulation rates. Instead all patients ran a menstrual calendar and a basal body temperature curve to find out and to demonstrate ovulation rates.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Effects on cycle disorders and insulin metabolism
In insulin-resistant women, insulin sensitivity was improved as early as within 4 wk of treatment. Consistent with this finding, the AUC-glucose to AUC-insulin ratio levels rose. Interestingly, although improvements of insulin sensitivity in this group were moderate, effects of metformin on menstrual cyclicity were striking. The main outcome criterion menstrual disturbance was successfully improved only in the metformin group without effect under placebo and restricted to insulin-resistant women, whereas women without insulin resistance failed to benefit significantly. In the present study, 80% of the women with insulin resistance receiving metformin, mostly suffering from severe oligomenorrhea at baseline, reported improvement of their menstrual cycles, in contrast to only 18% in the placebo group.
In accordance to short-term findings of Nestler et al. (20), biphasic basal body temperature curve assessment in women who experienced regular menses after metformin showed that most cycles became ovulatory. Spontaneous ovulation can occur rapidly, and normal menstrual rhythm can be achieved within 3 months of metformin treatment (20, 23). In accordance to our results, several studies show that the ovulation rate increased with no changes in weight, suggesting that the effect is independent of weight loss (20, 21, 34, 35, 36). Fleming et al. (25) have recently shown the mean time until the first ovulation to be significantly shorter in the metformin-treated group (24 d) than the placebo-treated group (42 d). This suggests a rather rapid effect of treatment upon ovarian and ovulatory function, which is further supported by the finding that the significant increase in estradiol concentrations was found already in the first week of treatment when the metformin dosage was restricted to only 850 mg/d. Interestingly no changes in the circulating inhibin-B or -T concentrations could be found suggesting that despite of improved follicular maturation there appeared to be no changes in the remainder of ovarian metabolism (total immature granulosa cell activity and stromal androgen biosynthesis). These data are extended by our findings. We were able to show that estradiol levels increased continuously every 4 wk, with statistical evidence only in the metformin group, indicating an improvement of ovarian function, whereas no significant changes in the T levels could be observed.
Along with the significant increase in estradiol concentrations, 67% of women had at least one ovulation within the treatment period of 12 wk receiving metformin, compared with only 45% in the placebo group, demonstrated by biphasic body temperature curves. In addition, only patients treated with metformin had three ovulatory cycles during the whole study period, indicating a remarkable improvement of ovulatory function. These differences however did not show statistical significance.
Hormonal and metabolic effects
Whereas reproductive and cyclic abnormalities improved strikingly after treatment, the positive effects of treatment were independent of changes in body weight or BMI and were not correlated to hormonal improvements. These findings are consistent with the data of Moghetti et al. (21) and Jakubowicz and Nestler (10). We did not observe effects on serum LH and androgens levels, as mentioned above. Furthermore, SHBG levels did not raise and hirsutism did not show any attenuation in the short treatment period. These findings are consistent with data from Harborne et al. (37). However, our data are not consistent with the results obtained in other short-term studies by the Nestler and Jakubowicz group (10) and in five of seven randomized studies (38), in which insulin metabolism, SHBG, and androgens improved. This discrepancy can be explained by the fact that severe hyperandrogenism was not considered as a necessary inclusion criterion in our study. There are two studies to mention (39, 40) that show a direct effect of metformin in reducing androgen production in theca cells.
The failure of metformin to influence circulating SHBG concentrations beyond placebo or control is another surprising observation that has been recorded previously (41).
Because insulin values are not standardized assessable, they need individual interpretation in synopsis with the clinical appearance of the patient to diagnose peripheral insulin resistance (42). In our study population, this procedure led to the same cut-off value for the FGIR (<4.5), as previously reported in literature (26).
Predictive factors
Besides describing early metformin effects, we were interested in defining subgroups, which benefit most from metformin treatment. Success of metformin treatment depended on the degree of insulin resistance. Theoretically, metformin should offer best results on ovulation rates in infertile and severely insulin-resistant women (37).
Moghetti et al. (21) defined baseline-independent predictors of clinical response to metformin in a logistic regression analysis including higher plasma insulin, lower serum A, and less severe menstrual abnormalities. These data further strengthened the hypothesis that insulin sensitizers may be effective only in the insulin-resistant subset of this heterogeneous syndrome.
In contrast, other studies did not confirm these predictive factors (25). Jakubowicz and Nestler (10) have shown the importance of glucose load and androgens at baseline. In contrast, waist to hip ratio was reduced in the active-treatment group alone. Because waist to hip ratio or waist circumference are established markers for visceral obesity, these data suggest that even without body weight loss, metformin treatment could reduce the metabolically most active form of adipose tissue that is most important for insulin resistance and cardiovascular risk.
In summary, the results of this prospective, randomized, double-blind, placebo-controlled study open the prospect toward a realistic treatment with metformin for a large number of women with PCOS. Insulin resistance or related metabolic pathways seem to be the important determinants of metformin’s efficacy in women with PCOS suffering from menstrual irregularities.
Therefore, we conclude that the presented data show with statistical evidence that insulin resistance is a baseline predictor of clinical efficacy in treatment with metformin in PCOS women. Only in this specific subgroup of women with PCOS suffering from insulin resistance, menstrual disturbance, ovulation rates, and consecutively infertility can be improved significantly.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: A, Androstenedione; ALT, alanine aminotransferase; AST, aspartate aminotransferase; AUC-glucose, area under the glucose curve; AUC-insulin, area under the insulin curve; BMI, body mass index; DHEA, dehydroepiandrosterone; DHEA-S, DHEA sulfate; F-G, Ferriman-Gallwey; FGIR, fasting glucose to insulin ratio; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment for insulin resistance; LDL, low-density lipoprotein; OGTT, oral glucose tolerance test; 17-OHP, 17-hydroxyprogesterone; PAO, polycystic-appearing ovaries; PCOS, polycystic ovary syndrome; QUICKI, quantitative insulin sensitivity check index; T, testosterone; TC, total cholesterol.
All authors of the manuscript have nothing to declare.
Received September 6, 2005.
Accepted December 7, 2005.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. A. Creanga, H. M. Bradley, C. McCormick, and C. T. Witkop Use of Metformin in Polycystic Ovary Syndrome: A Meta-Analysis Obstet. Gynecol., April 1, 2008; 111(4): 959 - 968. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Legro, H. X. Barnhart, W. D. Schlaff, B. R. Carr, M. P. Diamond, S. A. Carson, M. P. Steinkampf, C. Coutifaris, P. G. McGovern, N. A. Cataldo, et al. Ovulatory Response to Treatment of Polycystic Ovary Syndrome Is Associated with a Polymorphism in the STK11 Gene J. Clin. Endocrinol. Metab., March 1, 2008; 93(3): 792 - 800. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Moll, F. van der Veen, and M. van Wely The role of metformin in polycystic ovary syndrome: a systematic review Hum. Reprod. Update, November 1, 2007; 13(6): 527 - 537. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Tian, H. Shen, Q. Lu, R. J. Norman, and J. Wang Insulin Resistance Increases the Risk of Spontaneous Abortion after Assisted Reproduction Technology Treatment J. Clin. Endocrinol. Metab., April 1, 2007; 92(4): 1430 - 1433. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Rosenfield Identifying Children at Risk for Polycystic Ovary Syndrome J. Clin. Endocrinol. Metab., March 1, 2007; 92(3): 787 - 796. [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 |
