Introduction

The polycystic ovary syndrome (PCOS) is a predominantly hyperandrogenic and very prevalent disorder (1,2). When considered as a whole and compared to nonhyperandrogenic women, PCOS patients have insulin resistance irrespective of their grade of obesity (3) and associate frequently abdominal adiposity and/or obesity (4,5).

During the past years substantial research efforts have addressed the role of the adipokines adiponectin and leptin in the pathogenesis of the metabolic and cardiovascular complications of abdominal adiposity and obesity (6). The physiological roles of leptin include the regulation of adipose tissue homeostasis, mostly by modulating appetite and food intake (7), and also modulates reproductive function serving as a marker of the adipose tissue energy depots (8). Human obesity is characterized by resistance to the actions of leptin in several target tissues and the development of compensatory hyperleptinemia (6). In women with PCOS serum leptin levels are largely dependent on the association with obesity, and the abnormalities found in these women do not appear to be specific of PCOS (9).

On the contrary, adiponectin levels are specifically decreased in PCOS patients and this abnormality is independent from the grade of obesity, being present even in lean women with the syndrome (10,11). Adiponectin exhibits anti-inflammatory and insulin-sensitizing effects and its serum levels are decreased in abdominal adiposity, in obesity and in disorders of glucose tolerance (12).

We have proposed recently that PCOS patients suffer from a vicious circle consisting of androgen excess favoring the abdominal deposition of fat in affected women, and the dysfunction of visceral adipose tissue, by inducing hypoadiponectinemia, insulin resistance, and compensatory hyperinsulinism, further facilitating androgen secretion by their ovaries and adrenal glands (4).

However, the association of PCOS with hypoadiponectinemia and the inverse correlation between serum adiponectin and free testosterone levels reported earlier (10,11) do not provide any scientifically sound conclusion about the possible causal role of androgen excess on hypoadiponectinemia, this demonstration requiring interventions directed toward amelioration of androgen excess.

We here compared the effects on serum adiponectin and leptin levels of two pharmacological strategies based on the amelioration of hyperandrogenism on the one hand (using the antiandrogenic oral contraceptive pill Diane³⁵ Diario) and on the improvement of insulin sensitivity on the other (using the insulin sensitizer metformin). Our recently published data demonstrate that the antiandrogenic oral contraceptive pill clearly outperforms metformin in ameliorating androgen excess without influencing insulin sensitivity, whereas metformin improves insulin sensitivity markedly but have minor effects on androgen excess (13).

As will be seen, only the antiandrogenic oral contraceptive pill, and not metformin, improved hypoadiponectinema by increasing serum adiponectin levels and the adiponectin/leptin ratio suggesting that androgen excess might have a direct inhibitory effect on adiponectin secretion in premenopausal women.

Methods And Procedures

Patients

This study derives from a more ample randomized-controlled clinical trial addressing the effects of treatment with an antiandrogenic oral contraceptive compared with the insulin sensitizer metformin on classic and nonclassic cardiovascular risk factors in PCOS patients (ClinicalTrials.gov NLM Identifier NCT00428311); several results have been reported previously (13,14).

Thirty-four consecutive PCOS patients (age 24.3 6.2 years, range 15–42 years; BMI 29.9 6.3 kg/m², range 19.2–47.5 kg/m²) were recruited. The diagnosis of PCOS was based on the presence of clinical and/or biochemical hyperandrogenism, oligo-ovulation, and exclusion of secondary etiologies (1); the methods used to evaluate these criteria have been published already (13).

None of the patients had either a personal history of hypertension, alcohol abuse, disorders of glucose tolerance, cardiovascular events, sleep apnea, or had received treatment with oral contraceptives, antiandrogens, insulin sensitizers or drugs that might interfere with adiponectin or leptin levels for the previous 6 months. Written informed consent was obtained from all the participants, and the study was approved by the local Ethics Committee.

Study design

After giving informed consent, patients were randomized to receive an antiandrogenic oral contraceptive containing 35 g of ethinyl-estradiol plus 2 mg of cyproterone acetate (Diane³⁵ Diario, Schering España, Madrid, Spain) or 850 mg of metformin (Dianben, Merck Farma y Química, Mollet del Vallés, Spain) daily for 24 weeks. Simple randomization was conducted using blocks of 10 sealed opaque envelopes allocating five patients to receive Diane³⁵ Diario and 5 patients to receive metformin. Patients were instructed to maintain a diet containing 25–30 kcal per kg of body weight per day and moderate physical activity throughout the trial. Patients were submitted to a complete evaluation at baseline and after 12 and 24 weeks of treatment for this and other studies that included anthropometric and laboratory measurements, a 75-g oral glucose tolerance test (OGTT) with measurement of serum insulin and plasma glucose every 30 min for 2 h, and several tests of cardiovascular performance. No masking method was used after randomization.

Assays

For this study, adiponectin and leptin levels were measured by radioimmunoassay (Human Adiponectin RIA Kit and Human Leptin RIA Kit, Linco Research, St Charles, MO), with a sensitivity of 1 ng/ml and 0.5 ng/ml for adiponectin and leptin, respectively, and intra- and inter-assay coefficients of variation below 10% for both assays. The technical characteristics of the assays employed for plasma glucose and serum hormone measurements have been described elsewhere (15,16,17). The composite insulin sensitivity index (ISI) was calculated from the circulating glucose and insulin concentrations during the OGTT according to Matsuda and DeFronzo (18), and the area under the curve (AUC) for glucose and insulin from 75-g OGTT according to the mathematical method described by Tai (19).

Statistical analysis

Data are shown as means s.d. or as raw numbers and percentages, as appropriate. Sample size analysis was not conducted specifically for the changes on serum adipokine levels during treatment, but was obtained for the whole randomized controlled clinical trial addressing the effects of treatment with an antiandrogenic oral contraceptive compared with the insulin sensitizer metformin on classic and nonclassic cardiovascular risk factors. A priori sample size analysis was performed using the online calculator provided by the Massachusetts General Hospital Mallinckrodt General Clinical Research Center (http://hedwig.mgh.harvard.edu/
sample_size/size.html). Because the possible advantages of metformin over Diane³⁵ Diario rely on the presumed opposite effects of both drugs on insulin resistance, we used the difference in surrogate markers of insulin resistance as the primary outcome for sample size analysis. An experimental design including a total of 22 patients would have power above 0.80 to detect differences in fasting insulin levels between treatments such as those previously reported using a similar experimental design (20,21).

For continuous variables, normality was assessed using the Kolmogorov–Smirnov test and logarithmic transformation was applied as needed to ensure a normal distribution. At baseline, the comparison between arms of treatment was analyzed by unpaired t-test. We used a two-way ANOVA to evaluate the differences in baseline adiponectin, leptin, anthropometric and hormonal variables depending on the presence of obesity (nonobese: BMI <30.0 kg/m²; obese: BMI 30.0 kg/m²) depending on the arm of treatment to which the women were allocated, and the possible interaction between these independent variables. For discontinuous variables, the ²-test or Fisher's exact test were applied, as appropriate.

Treatment effects on serum adiponectin and leptin levels, and on the adiponectin/leptin ratio was analyzed by a repeated-measures general lineal model. The arm of treatment, and the presence or absence of obesity, were introduced as between-subjects effects, and the visit of evaluation (baseline, 12 and 24 weeks) was introduced as the within-subjects effect. The interaction between the arm of treatment and the visit of evaluation was used to estimate the differences in the response to each treatment over time. Also the interaction of both effects with the presence or absence of obesity served to evaluate if the differences in the response to each treatment over time occurred when considering all the patients as a whole, or were restricted to either the nonobese or the obese subgroups. Pearson's correlation analysis was used to correlate the changes (expressed as percentage of baseline values) observed during treatment in serum adiponectin levels and in the adiponectin/leptin ratio with the changes observed in serum androgen levels.

Seven patients discontinued metformin for different reasons (13), as follows: before the 12 weeks visit, two patients discontinued metformin because of gastrointestinal side-effects, one patient got pregnant, and two patients discontinued the study because of protocol violations; between the 12 and 24 weeks visits, one patient changed her address and was lost to follow-up, and another patient was discontinued because of a protocol violation.

Considering the large drop-out rate with metformin, the results obtained when considering only the patients completing the three visits of the protocol were also confirmed by intention-to-treat analysis assuming no changes in dependent variables at the missing visits with respect to the previous evaluation for patients discontinuing metformin. P < 0.05 was considered statistically significant. Analyses were performed using SPSS 10 for Macintosh (SPSS, Chicago, IL).

Results

No differences were observed in the baseline characteristics of the PCOS patients allocated to either treatment (13), including serum adiponectin concentrations (11.3 4.8 vs. 12.5 6.2 g/ml for the women allocated to Diane³⁵ Diario or to metformin, P = 0.622), serum leptin levels (25.6 8.1 vs. 29.2 14.6 ng/ml for the women allocated to Diane³⁵ Diario or to metformin, P = 0.370), and the adiponectin/leptin ratio (0.50 0.31 vs. 0.69 0.84 g/ng for the women allocated to Diane³⁵ Diario or to metformin, P = 0.421) (Table 1).

Table 1 - Baseline clinical and biochemical characteristics of the PCOS patients randomized to receive Diane³⁵ Diario or metformin as function of obesity.

Full table (81K)

At baseline, obese women (BMI 30.0 kg/m²) presented with higher abdominal circumference, waist-to-hip ratio, fat mass expressed as percentage of total body weight, free testosterone, AUC glucose and AUC insulin, and serum leptin levels compared with nonobese PCOS patients, irrespective of the arm of treatment to which these women were allocated (Table 1). Obese PCOS patients also showed a lower ISI and adiponectin/leptin ratio compared with their nonobese counterparts (Table 1).

As previously reported (13), Diane³⁵ Diario was superior to metfomin for the control of hyperandrogenism and menstrual disturbances and increased plasma Apo A-I and HDL-phospholipids levels, whereas the ISI increased with metformin but did not change with Diane³⁵ Diario. Of note, this occurred in the absence of any statistically significant change in BMI, waist circumference and waist-to-hip ratio and the percentage of body fat with any of the drugs (13).

Effects of the treatments in obese and nonobese PCOS patients

When introducing the presence of obesity as between subject effect in the general linear model, we could observe a significant decrease of waist circumference in the whole group patients (Wilks' = 0.588, F = 7.707, P = 0.003, Figure 1; intention-to-treat analysis: Wilks' = 0.712, F = 5.859, P = 0.007), that occurred irrespective of the treatment applied (Wilks' = 0.964, F = 0.410, P = 0.669, Figure 1; intention-to-treat analysis: Wilks' = 0.998, F = 0.025, P = 0.975), yet this decrease in waist circumference was actually related to the change observed only in obese patients (Wilks' = 0.546, F = 9.135, P = 0.001, Figure 1; intention-to-treat analysis: Wilks' = 0.662, F = 7.397, P = 0.003). The results regarding other anthropometrical, metabolic and hormonal parameters previously reported (13), including the lack of changes observed in body weight and the percentage of body fat, were unchanged when also considering the interaction between the arm of treatment and obesity.

Figure 1.

Changes in waist circumference during treatment of PCOS women with Diane³⁵ Diario or metformin as a function of obesity. Black circles represent the data from the patients allocated to Diane³⁵ Diario whereas white circles represent the data from the patients allocated to metformin. The small figures under the x-axis indicate the number of patients at each visit. Data are means s.e.m., and were submitted to a repeated-measures general lineal model introducing the arm of treatment and the presence or absence of obesity as between-subjects effects, and the visit of evaluation (baseline, 12 and 24 weeks) as the within-subjects effect. *P < 0.05 for the interaction between the visit of evaluation and obesity, meaning that the changes in waist circumference during treatment occurred only in obese patients, in both the analysis of patients who completed the study and in the intention-to-treat analysis.

Full figure and legend (15K)

Effects of treatment on serum adiponectin and leptin levels and on adiponectin/leptin ratio as a function of the presence of obesity

Serum adiponectin levels increased during the 24 weeks of the study in the group of PCOS patients treated with Diane³⁵ Diario as demonstrated by the statistically significant interaction of the effect of the visit of evaluation with the arm of treatment (Wilks' = 0.705, F = 4.603, P = 0.021; Figure 2; intention-to-treat analysis: Wilks' = 0.866, F = 2.249, P = 0.124), yet this change was actually caused by the increase observed in the obese women treated with Diane³⁵ Diario as demonstrated by the statistically significant interaction of the effect of the visit of evaluation with the arm of treatment and grade of obesity (Wilks' = 0.755, F = 3.570, P = 0.045; Figure 2; intention-to-treat analysis: Wilks' = 0.854, F = 2.482, P = 0.101).

Figure 2.

Changes in serum adiponectin and leptin levels, and in the adiponectin/leptin ratio (g/ng), during treatment of PCOS women with Diane³⁵ Diario or metformin as a function of obesity. Black circles represent the data from the patients allocated to Diane³⁵ Diario whereas white circles represent the data from the patients allocated to metformin. The small figures under the x-axis indicate the number of patients at each visit. Data are means s.e.m., and were submitted to a repeated-measures general lineal model introducing the arm of treatment and the presence or absence of obesity as between-subjects effects, and the visit of evaluation (baseline, 12 and 24 weeks) as the within-subjects effect. *P < 0.05 for the interaction of the visit of evaluation with the arm of treatment, meaning the differences in the response to both drugs was present when considering obese and nonobese PCOS patients as a whole, only in patients who completed the study. ^†P < 0.05 for the interaction between the effects of the visit of evaluation, the arm of treatment and of obesity (meaning that the differences in the response to both drugs was present only in the obese PCOS patients) when analyzing only those patients who completed the study. ^‡P < 0.05 for the interaction between the effects of the visit of evaluation, the arm of treatment and of obesity (meaning that the differences in the response to both drugs was present only in the obese PCOS patients) in both the analysis of the patients who completed the study and in intention-to-treat analysis.

Full figure and legend (24K)

Nevertheless, leptin concentrations did not show any statistically significant change along the study in the whole group of patients (Wilks' = 0.961, F = 0.445, P = 0.647; Figure 2; intention-to-treat analysis: Wilks' = 0.985, F = 0.216, P = 0.807), even when considering also the interaction with obesity (Wilks' = 0.862, F = 1.755, P = 0.196; Figure 2; intention-to-treat analysis: Wilks' = 0.896, F = 1.678, P = 0.204). Similarly, no statistically significant influence of the arm of treatment on the evolution of serum leptin levels (Wilks' = 0.945, F = 0.645, P = 0.534; Figure 2; intention-to-treat analysis: Wilks' = 0.973, F = 0.396, P = 0.677), even when considering also the effect of obesity (Wilks' = 0.886, F = 1.421, P = 0.263; Figure 2; intention-to-treat analysis: Wilks' = 0.867, F = 2.227, P = 0.126).

Finally, the adiponectin/leptin ratio improved only in the group of obese PCOS patients who were allocated to receive Diane³⁵ Diario, as demonstrated by the statistically significant interaction between the effects of visit of evaluation, arm of treatment and obesity (Wilks' = 0.707, F = 4.565, P = 0.022; Figure 2; intention-to-treat analysis: Wilks' = 0.745, F = 4.970, P = 0.014).

Correlation between the changes observed during treatment in serum adiponectin levels and in the adiponectin/leptin ratio and the changes observed in serum androgen levels

The changes observed during treatment in serum adiponectin levels did not correlate with serum total or free testosterone concentrations or with serum androstendione levels, either when considering all the patients as a whole or when studying the women treated with Diane³⁵ Diario or metfomin separately (Figure 3). On the contrary, the changes in the adiponectin/leptin ratio showed a strong inverse correlation with the changes in serum androstendione levels both when considering all the patients as a whole or when studying the women treated with Diane³⁵ Diario separately (Figure 3). Finally, the changes in the adiponectin/leptin ratio did not correlate with the changes in serum total or free testosterone concentrations (Figure 3).

Figure 3.

Correlation analysis of the changes observed during treatment in serum adiponectin levels and in the adiponectin/leptin ratio with the changes observed in serum androgen levels. Black circles represent the data from the patients allocated to Diane³⁵ Diario whereas white circles represent the data from the patients allocated to metformin. The changes are as percentage of baseline values and the analysis was conducted either considering all patients as a whole, or considering the women treated with Diane³⁵ Diario or metfomin separately. The thick line corresponds to the correlation analysis considering all the patients as a whole; the thin dashed line corresponds to the analysis of the patients treated with Diane³⁵ Diario, and the thin dotted line corresponds to the analysis of the patients treated with metfomin.

Full figure and legend (31K)

Discussion

The association of PCOS with hypoadiponectinemia and the inverse correlation between serum adiponectin and free testosterone levels reported earlier suggested a possible causal role of the androgen excess on the hypoadiponectinemia characteristic of these women (10,11). Our present results, by demonstrating that the antiandrogenic oral contraceptive Diane³⁵ Diario increases serum adiponectin levels and the adiponectin/leptin ratio in obese PCOS patients in parallel with a marked decrease in serum androgen levels (13), might suggest that androgen excess directly influences adipocyte function by modulating adiponectin secretion. This is also supported by the finding of a strong and statistically significant correlation between the improvement in the adiponectin/leptin ratio and the decrease in serum androstendione concentrations in the women treated with Diane³⁵ Diario. In conceptual agreement, androgen receptors are present in preadipocytes and adipocytes, particularly in visceral adipose tissue, possibly influencing the distribution of body fat (22) and administration of testosterone decreases the secretion of adiponectin both in vitro (23) and in vivo (24). However, because the increase in circulating adiponectin levels did not correlate with the decrease in serum androgen levels, and also because the improvement in the adiponectin/leptin ratio only correlated with the decrease in serum androstendione concentrations and not with the decrease in serum total or free testosterone levels, it is also possible that oral contraceptives increase serum adiponectin levels by mechanisms not directly related to the amelioration of androgen excess.

Furthermore, the increase in the serum levels of the insulin sensitizing molecule adiponectin might be related to the apparent paradox of the different effects on insulin resistance and glucose tolerance of oral contraceptives in women from the general population and in hyperandrogenic women: whereas oral contraceptives may worsen insulin resistance and glucose tolerance in the former (25,26), they exert neutral or even beneficial effects in the latter (21,27), including the PCOS women studied here (13). We hypothesize that the increase in adiponectin secretion occurred in response to the amelioration of androgen excess in our obese PCOS patients treated with Diane³⁵ Diario overcame the possible deleterious effects of the drug on insulin resistance and glucose tolerance.

Also, the amelioration of androgen excess that occurred with both Diane³⁵ Diario and metformin in our series (13)—yet much more marked with the oral contraceptive—might have been related to the decrease in waist circumference in the obese subset of PCOS patients studied here, suggesting a certain redistribution of body fat to peripheral depots because it occurred in absence of significant changes in BMI and in the percentage of body fat (13). A similar change in the distribution of body fat has been described with the use of the pure antiandrogen flutamide in PCOS patients (28).

As opposed to adiponectin, serum leptin levels did not change during our clinical trial either with Diane³⁵ Diario or with metformin. Leptin acts as a marker of overall fat depots both in PCOS patients and in the general population (9,29). In agreement with our present results, serum leptin levels did not change in PCOS patients during treatment with metformin, contraceptive pills or even gonadotropin-releasing hormone analogues in the absence of changes in BMI (30,31,32). However, when considering the effects of metformin in obese and nonobese PCOS patients separately, it appears that leptin tends to decrease in obese subjects, and to increase in the lean patients, yet these changes do not result in statistically significant differences.

The adiponectin/leptin ratio is a better predictor of insulin resistance, inflammation and atherosclerosis than the separate measure of their individual serum concentrations (33) and, together with the increase observed in serum adiponectin levels, the increase in this ratio highlights the beneficial impact of treatment with Diane³⁵ Diario on the dysfunction of visceral adipose tissue in obese PCOS women.

Finally, the insulin sensitizer metformin did not influence adiponectin and leptin secretion in our patients despite a clear improvement in insulin sensitivity (13), indicating that the insulin-sensitizing effects of metformin in PCOS patients are not mediated by changes in adipokine secretion, as expected from the mostly hepatic, muscular, and intestinal effects of this drug. The absence of changes in serum adiponectin levels during metformin treatment has been reported previously in a small series of adult PCOS patients (34).

In summary, the amelioration of androgen excess by using an antiandrogenic oral contraceptive pill improves adipose tissue dysfunction in obese PCOS patients, as reflected by a clear increase in serum adiponectin levels and in the adiponectin/leptin ratio. Considering the insulin sensitizing effects of adiponectin, the increase in the circulating levels of this adipokine (which is not observed with metformin, and is especially important in obese patients) might contribute to explain why oral contraceptives do not deteriorate insulin resistance and glucose tolerance in PCOS patients.

Disclosure

The authors declared no conflict of interest.

References

REFERENCES

1. Azziz R, Carmina E, Dewailly D et al. Position statement: criteria for defining polycystic ovary syndrome as a predominantly hyperandrogenic syndrome: an Androgen Excess Society guideline. J Clin Endocrinol Metab 2006;91:4237–4245. | Article | PubMed | ISI | ChemPort |
2. Asuncion M, Calvo RM, San Millan JL et al. A prospective study of the prevalence of the polycystic ovary syndrome in unselected Caucasian women from Spain. J Clin Endocrinol Metab 2000;85:2434–2438. | Article | PubMed | ISI | ChemPort |
3. Ehrmann DA. Polycystic ovary syndrome. N Engl J Med 2005;352:1223–1236. | Article | PubMed | ISI | ChemPort |
4. Escobar-Morreale HF, San Millán JL. Abdominal adiposity and the polycystic ovary syndrome. Trends Endocrinol Metab 2007;18:266–272. | Article | PubMed | ChemPort |
5. Alvarez-Blasco F, Botella-Carretero JI, San Millan JL, Escobar-Morreale HF. Prevalence and characteristics of the polycystic ovary syndrome in overweight and obese women. Arch Intern Med 2006;23:2081–2086. | Article |
6. Trujillo ME, Scherer PE. Adipose tissue-derived factors: impact on health and disease. Endocr Rev 2006;27:762–778. | PubMed | ChemPort |
7. Casanueva FF, Dieguez C. Neuroendocrine regulation and actions of leptin. Front Neuroendocrinol 1999;20:317–363. | Article | PubMed | ChemPort |
8. Moschos S, Chan JL, Mantzoros CS. Leptin and reproduction: a review. Fertil Steril 2002;77:433–444. | Article | PubMed | ISI |
9. Glintborg D, Andersen M, Hagen C et al. Evaluation of metabolic risk markers in polycystic ovary syndrome (PCOS). Adiponectin, ghrelin, leptin and body composition in hirsute PCOS patients and controls. Eur J Endocrinol 2006;155:337–345. | Article | PubMed | ChemPort |
10. Carmina E, Orio F, Palomba S et al. Evidence for altered adipocyte function in polycystic ovary syndrome. Eur J Endocrinol 2005;152:389–394. | Article | PubMed | ISI | ChemPort |
11. Escobar-Morreale HF, Villuendas G, Botella-Carretero JI et al. Adiponectin and resistin in PCOS: a clinical, biochemical and molecular genetic study. Hum Reprod 2006;21:2257–2265. | Article | PubMed | ChemPort |
12. Schulze MB, Shai I, Rimm EB et al. Adiponectin and future coronary heart disease among men with type 2 diabetes. Diabetes 2005;54:534–539. | Article | PubMed | ChemPort |
13. Luque-Ramírez M, Alvarez-Blasco F, Botella-Carretero JI et al. Comparison of ethinyl-estradiol plus cyproterone acetate versus metformin effects on classic metabolic cardiovascular risk factors in women with the polycystic ovary syndrome. J Clin Endocrinol Metab 2007;92:2453–2461. | Article | PubMed | ChemPort |
14. Luque-Ramírez M, Alvarez-Blasco F, Botella-Carretero JI et al. The increased body iron stores of obese women with polycystic ovary syndrome are a consequence of insulin resistance and hyperinsulinism, and do not result from reduced menstrual losses. Diabetes Care 2007;30:2309–2313. | Article | PubMed | ChemPort |
15. Escobar-Morreale HF, Serrano-Gotarredona J, Garcia-Robles R, Sancho J, Varela C. Mild adrenal and ovarian steroidogenic abnormalities in hirsute women without hyperandrogenemia: does idiopathic hirsutism exist? Metabolism 1997;46:902–907. | Article | PubMed | ChemPort |
16. Escobar-Morreale HF, Serrano-Gotarredona J, Garcia-Robles R, Sancho JM, Varela C. Lack of an ovarian function influence on the increased adrenal androgen secretion present in women with functional ovarian hyperandrogenism. Fertil Steril 1997;67:654–662. | Article | PubMed | ChemPort |
17. Escobar-Morreale HF, Lasunción MA, Sancho J. Treatment of hirsutism with ethinyl estradiol–desogestrel contraceptive pills has beneficial effects on the lipid profile and improves insulin sensitivity. Fertil Steril 2000;74:816–819. | Article | PubMed | ChemPort |
18. Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 1999;22:1462–1470. | Article | PubMed | ISI | ChemPort |
19. Tai MM. A mathematical model for the determination of total area under glucose tolerance and other metabolic curves. Diabetes Care 1994;17:152–154. | Article | PubMed | ISI | ChemPort |
20. Morin-Papunen LC, Vauhkonen I, Koivunen RM et al. Endocrine and metabolic effects of metformin versus ethinyl estradiol-cyproterone acetate in obese women with polycystic ovary syndrome: a randomized study. J Clin Endocrinol Metab 2000;85:3161–3168. | Article | PubMed | ChemPort |
21. Morin-Papunen L, Vauhkonen I, Koivunen R et al. Metformin versus ethinyl estradiol-cyproterone acetate in the treatment of nonobese women with polycystic ovary syndrome: a randomized study. J Clin Endocrinol Metab 2003;88:148–156. | Article | PubMed | ChemPort |
22. Dieudonne MN, Pecquery R, Boumediene A, Lenevu MC, Giudicelli Y. Androgen receptors in human preadipocytes and adipocytes: regional specificities and regulation by sex steroids. Am J Physiol 1998;274:C1645–C1652. | PubMed | ISI | ChemPort |
23. Nishizawa H, Shimomura I, Kishida K et al. Androgens decrease plasma adiponectin, an insulin-sensitizing adipocyte-derived protein. Diabetes 2002;51:2734–2741. | Article | PubMed | ChemPort |
24. Lanfranco F, Zitzmann M, Simoni M, Nieschlag E. Serum adiponectin levels in hypogonadal males: influence of testosterone replacement therapy. Clin Endocrinol (Oxf) 2004;60:500–507. | Article | PubMed | ChemPort |
25. Chasan-Taber L, Stampfer MJ. Epidemiology of oral contraceptives and cardiovascular disease. Ann Intern Med 1998;128:467–477. | PubMed | ChemPort |
26. Baillargeon J-P, McCliah DK, Essah PA, Nestler JE. Association between the current use of low-dose oral contraceptives and cardiovascular arterial disease: a meta-analysis. J Clin Endocrinol Metab 2005;90:3863–3870. | Article | PubMed | ChemPort |
27. Harborne L, Fleming R, Lyall H, Sattar N, Norman J. Metformin or antiandrogen in the treatment of hirsutism in polycystic ovary syndrome. J Clin Endocrinol Metab 2003;88:4116–4123. | Article | PubMed | ChemPort |
28. Gambineri A, Patton L, Vaccina A et al. Treatment with flutamide, metformin and their combination added to hypocaloric diet in overweight-obese women with polycystic ovary syndrome: a randomized, 12-month, placebo-controlled study. J Clin Endocrinol Metab 2006;91:3970–3980. | Article | PubMed | ChemPort |
29. Considine RV, Sinha MK, Heiman ML et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N Engl J Med 1996;334:292–295. | Article | PubMed | ISI | ChemPort |
30. Morin-Papunen LC, Koivunen RM, Tomás C, Ruokonen A, Martikainen HK. Decreased serum leptin concentrations during metformin therapy in obese women with polycystic ovary syndrome. J Clin Endocrinol Metab 1998;83:2566–2568. | Article | PubMed | ISI | ChemPort |
31. Romualdi D, Campagna G, Selvaggi LJ et al. Metformin treatment does not affect total leptin levels and free leptin index in obese patients with polycystic ovary syndrome. Fertil Steril 2008;89:1273–1276. | Article | PubMed | ChemPort |
32. Krotkiewski M, Landin K, Dahlgren E, Janson PO, Holm G. Effect of two modes of antiandrogen treatment on insulin sensitivity and serum leptin in women with PCOS. Gynecol Obstet Invest 2003;55:88–95. | Article | PubMed | ChemPort |
33. Satoh N, Naruse M, Usui T et al. Leptin-to-adiponectin ratio as a potential atherogenic index in obese type 2 diabetic patients. Diabetes Care 2004;27:2488–2490. | Article | PubMed | ChemPort |
34. Spranger J, Mohlig M, Wegewitz U et al. Adiponectin is independently associated with insulin sensitivity in women with polycystic ovary syndrome. Clin Endocrinol (Oxf) 2004;61:738–746. | Article | PubMed | ChemPort |

Acknowledgments

We thank Ms. Genoveva González for excellent technical help. This study was supported by the Spanish Ministry of Health and Consumer Affairs, Instituto de Investigación Carlos III, grants FIS PI050341, PI080944 and REDIMET RD06/0015/0007, and by economic aid from Hospital Ramón y Cajal. Centro de Investigación Biomédica en Red sobre Diabetes y Enfermedades Metabólicas Asociadas CIBERDEM is an inititative of Instituto de Investigación Carlos III.