Printer Friendly

Combined Oral Contraceptive Effects on Low-Grade Chronic Inflammatory Mediators in Women with Polycystic Ovary Syndrome: A Systematic Review and Meta-Analysis.

1. Introduction

Polycystic ovary syndrome (PCOS), which is common in women of reproductive age, presents with different phenotypes regarding androgen levels, oligo-/anovulation, ovary morphology, dyslipidemia, dysglycemia, and fat body mass. Basically, four phenotypes are clearly defined: hyperandrogenism, oligo- or anovulation plus polycystic ovary morphology (A), hyperandrogenism plus oligo-/anovulation (B), oligo-/anovulation plus polycystic ovary morphology (C), and hyperandrogenism plus polycystic ovary morphology (D). Clinical and/or biochemical hyperandrogenism, abnormal fat distribution, central obesity, and insulin resistance are observed in 50%-70% and metabolic syndrome in about 40% of PCOS patients [1]. The different phenotypes of PCOS may induce endothelial dysfunction and atherosclerosis [2]. Low-grade chronic inflammation is characterized by increased blood levels of several mediators of endothelial dysfunction and inflammatory processes. The principal markers of the inflammatory state include C-reactive protein (CRP), tumor necrosis factor alpha (TNF-[alpha]), homocysteine (Hcy), interleukins 1 and 6 (IL-1, IL-6), follistatin, and total number of white blood cells (WBCs) [3-8]. The initiation of the inflammatory process is complex but it is tightly associated with body mass and android fat distribution [1, 9].

Despite the complex pathophysiology and multiple phenotypes of PCOS, the first-line treatment of all women with PCOS not planning to become pregnant is combined oral contraceptives (COCs). The major risks and benefits of COCs in PCOS patients were recently reviewed [10]. Because COC use itself may induce an inflammatory profile [11], the combination of oral contraceptives and PCOS could worsen or increase the risks for venous thromboembolism and cardiovascular disease (CVD) in these patients. The main objective of this meta-analysis was to examine the potential additional risk of all COC preparations in women with PCOS by investigating each markers of low-grade chronic inflammation.

2. Methods and Methods

2.1. Search Procedure. To answer the question whether combined oral contraceptives modify or not inflammatory markers, a search and review of all studies published between 1990 and 2017 on the impact of COC on inflammatory markers in PCOS patients that were written in English, using PubMed, Scopus, and the Cochrane database were done. Keywords in different combination for the search included oral contraceptive, polycystic ovary syndrome, polycystic ovary disease, polycystic ovarian disease, chronic inflammation, inflammatory mediators, and contraception. Online searches of specialized journals were also used by adding appropriate keywords. In addition, these databases were expanded by manual search of reference lists of relevant studies from the obtained articles.

2.2. Criteria for Study Inclusion. Study inclusion was based on a description of the PCOS diagnostic using the National Institute of Health (NIH) or the Rotterdam criteria and comparison of the inflammatory markers concentrations before and after use of any COC formulation for at least three months. Eligibility was double-checked independently by two authors (SFM, MMWY). Taking into account the included studies, the effects of COCs on inflammatory mediators were evaluated in 838 PCOS women by comparison of the data obtained before and after 3-18 months of COC use.

2.3. Data Extraction and Quality Assessment. Two authors (SFM, MMWY) extracted information regarding publication intervention, study design, outcome, statistical methods, and results. Further, they independently assessed quality and bias using the Newcastle-Ottawa scale for evaluating the quality of randomized and nonrandomized observational studies in meta-analysis [12]. As recommended three factors were considered to score the quality of included studies: (A) selection (representativeness of exposed cohort, selection of no exposed cohort, and ascertainment of exposure), (B) demonstration (at the start of the study the outcome of interest of modified), comparability (assessed on the bias of study design and analysis), and (C) outcome (assessment of outcome, sufficient follow-up, and adequate follow-up). The quality of the studies was rated good (3 stars in a, to 2 stars in b, and 2 to 3 stars in c), fair (2 stars in a; 1 to 2 stars in b, and 2 to 3 stars in c), and poor (0 to 2 stars in a, 0 stars in b, and 0 to 1 stars in c) (Table 1).

2.4. Outcome Measures. As the primary outcomes were changes in the concentrations of CRP, Hcy, follistatin, TNF-[alpha], IL-6, leukocyte total number, and several adhesion molecules with the use of COCs, only studies in which individuals were used as their own control were included.

2.5. Data Analysis. Because studies did not use the same assays to measure the principal marker (CRP), the effect sizes are presented as the standardized mean difference (SMD) and standard error (SE). When adequate these methods were used for analysis of the inflammatory markers. Data were pooled by using a random-effects model. Under this model, it was assumed that the true effect size may vary from study to study, and the summary effect was the estimate of the mean of the distribution of effect sizes [13]. The variance of the effect size across the studies ([Tau.sup.2]) was estimated using the method of moments [14]. Heterogeneity was assessed for each result by using Cochran's Q statistic and the I score to quantify the degree of heterogeneity [15]. When pooling the results of different studies was not possible, the analysis of the effect size in a particular study was described as the raw mean difference reported in the original studies, and these differences were analyzed using ANOVA, two-sided Student's t tests, or Wilcoxon or Mann-Whitney tests as described in each study.

3. Results

3.1. Studies Selection. A total of 1995 articles were initially identified. After screening by one of the authors (SFM) using the reported keywords, 433 studies were maintained. Three reviewers (NSS, MMWY, and MASM) screened abstracts to assess the eligibility of the studies. With this procedure, 1562 articles were excluded (Figure 1). Furthermore, another 54 full articles were excluded because they did not compare the values of the variables of interest obtained before and after the use of COCs or did not provide sufficient data for analysis. Finally, 27 full observational papers providing all necessary data were selected (Table 2).

3.2. Risk of Bias. The study quality was verified using the risk of bias assessment tool Newcastle-Ottawa scale for observational studies and summarized in Table 2.

3.3. Effects of Interventions. The pooled results of the studies are presented in the following sections. In some sections, the results were not pooled and instead are presented as single mean differences. Tables 3 and 4 summarize the effects of interventions using different COC formulations.

3.3.1. C-Reactive Protein. Data regarding the impact of COCs containing 35 [micro]g ethinylestradiol (EE) and 2 mg cyproterone acetate (CPA) on CRP levels of PCOS patients after 3-6 months of use were extracted from eleven articles, which were published in peer-reviewed journals between 2003 and 2016 [5, 16-25]. All studies were randomized or nonrandomized open trials. A total of 328 PCOS patients were enrolled, and their CRP levels were compared before and after the COC use. The SMD was -0.511 (SE = 0.117, p <0.001) (Figure 2(a)). Perhaps because different assays were used to measure CRP (standard or highly sensitive), these studies were heterogeneous ([Tau.sup.2] = 0.099, Q = 34.94, [I.sup.2] = 71.38, p <0.001).

Changes in the serum concentrations of CRP before and after the use of a combined formulation containing 30 [micro]g EE and 3 mg drospirenone (DRSP) for 3-6 months were reported in six studies published between 2011 and 2016 [22, 23, 2528]. A total of 206 PCOS patients were enrolled, and the SMD was -0.455 (SE=0.074, p <0.001). These studies were homogeneous ([Tau.sup.2] = 0.000, Q = 4.32, [I.sup.2] =0.00, p <0.001) (Figure 2(a)). Comparison of the difference in CRP levels before and after the use of a combination containing 30 [micro]gEE and 2 mg CMA was performed in two reports that included 71 PCOS patients [28, 29]. Either a significant increase of 195% [29] or a nonsignificant increase of 12% was reported [28]. The SMD was -0.435 (SE = 0.124, p <0.001). Both studies were highly homogeneous ([Tau.sup.2] = 0.00, Q = 0.01, [I.sup.2] = 0.00, p = 0.918) (Figure 2(b)).

The impact of 150 [micro]g of the progestin desogestrel (DSG) combined with 30 [micro]g EE on CRP levels was verified in two studies published in 2008 and 2011, which included only 60 PCOS patients [30, 31]. The SMD was -0.059 (SE = 0.429, p = 0.891). The studies were highly heterogeneous ([Tau.sup.2] = 0.486, Q = 17.52, [I.sup.2] = 88.58, p <0.001) (Figure 2(b)). A combination of 20 [micro]g EE and 150 [micro]g DSG for 3 months induced a nonsignificant increase in the CRP level of PCOS patients from 1.6 mg/l to 1.70 mg/l (p = 0.720) [32]. A single study [33] reported a nonsignificant increase in CRP levels from 4.9 mg/l to 5.6 mg/l (p >0.05) after the use of a combination of 30 [micro]g EE and 0.15-0.25 mg of the progestin norgestimate (NGM).

3.3.2. Homocysteine. Three studies with PCOS patients using a combination of 35 [micro]g EE and 2 mg CPA reported increases in Hcy levels of 6%-18% [23, 34, 35]. Comparison of the Hcy levels before and after the COC administration performed in these studies, with 59 PCOS patients, showed SMD of 0.219 (SE = 0.377, p = 0.561). The studies were highly heterogeneous ([Tau.sup.2] = 0.356, Q = 12.99, [I.sup.2] = 4.60, p = 0.002) (Figure 3).

COCs with DRSP were reported to either decrease or increase Hcy levels [23, 27, 36]. The analysis combining the results of 71 PCOS patients demonstrated an SMD of 0.057 (SE = 0.407, p = 0.888), and the studies were highly heterogeneous ([Tau.sup.2] = 0.585, Q = 27.26, [I.sup.2] = 88.89, p = 0.001) (Figure 3). The combination of 30 [micro]g EE and 150 [micro]g DSG decreased or did not change the Hcy levels in PCOS users in two studies [31, 34]. When obese and nonobese PCOS patients were taken into account, the SMD was 0.346 (SE = 0.391, p = 0.376). These studies were also shown to be highly heterogeneous ([Tau.sup.2] = 0.387, Q = 14.10, [I.sup.2] = 85.81, p = 0.001) (Figure 3).

3.3.3. Follistatin. Two studies examined follistatin changes after the use of COCs containing 35 [micro]g EE and 2 mg CPA in 73 PCOS patients [5, 18] and, in both studies, the follistatin levels were significantly increased (p <0.001 for both). The SMD in follistatin levels was -0.653 (SE = 0.246, p = 0.008). The studies only showed a moderate nonsignificant heterogeneity ([Tau.sup.2] = 0.076, Q = 2.43, [I.sup.2] = 58.93, p = 0.119) (Figure 4).

3.3.4. Interleukin-6. Five studies with 84 PCOS patients examined changes in IL-6 with the use of different COC compositions [20, 29, 33, 37, 38] but it was not possible to pool the data, and the effect sizes are reported individually as single mean differences (Table 4). A nonsignificant decrease in IL-6 levels, from 1.67 pg/ml to 1.39 pg/ml (ES = -0.280; p = 0.911), was reported with the use of 35 [micro]g EE and 2 mg CPA [20]. PCOS patients using COCs with DRSP showed a nonsignificant increase in IL-6 levels, from 0.95 pg/ml to 1.09 pg/ml (ES = 0.140; p >0.05), after nine months of use [37]. One study found a nonsignificant decrease in IL-6 levels in users of COCs containing CMA, from 1.9 pg/ml to 1.7 pg/ml (ES = -0.200; p >0.05) [29]. After the use of COCs containing the progestin DSG for 12 months, IL-6 levels showed a nonsignificant decrease from 1.7 pg/ml to 1.4 pg/ml (ES= -0.300; p >0.05). Use of COCs with NGM for three months by PCOS patients resulted in a nonsignificant decrease in IL-6 levels from 2.1 pg/ml to 1.6 pg/ml (ES = -0.500; p = 0.060). Collectively, COCs do not increase the levels of IL-6 in PCOS patients. The results of the studies are limited because the women used different COC compositions and studies had a small number of patients.

3.3.5. Plasminogen Activator Inhibitor-1. The use of a combination of 30 [micro]g EE and 150 [micro]g DSG resulted in a significant decrease in the PAI-1 levels after six months of use from 46.9 ng/ml to 29.50 ng/ml (ES = -17.4; p <0.05) [30]. A pill containing 35 [micro]g EE and 0.18-0.25 [micro]g NGM induced a nonsignificant decrease in PAI-1 levels, from 81.2 ng/ml to 76.5 ng/ml (ES = -4.7; p = 0.520), after three months of use by PCOS patients [33]. Furthermore, a combination of 30 [micro]g EE and 3 mg DRSP also resulted in a nonsignificant decrease in PAI-1 levels from 2.6 UI/ml to 2.4 IU/ml (ES = -0.20; p >0.05) after six months of use [25]. Collectively, COCs decrease PAI-1 levels in PCOS patients, but studies with larger sample size are still needed to definitively establish the clinical significance of this change. In another study, which was not included in this meta-analysis due to missing data, PAI-1 levels were significantly decreased after six months of use of 35 [micro]g EE and 2 mg CPA by PCOS patients (ES = -0.7 IU/ml, p = 0.004) [17].

3.3.6. Leukocyte Total Number. In a randomized open clinical trial, COC preparations with 30 [micro]g EE and 3 mg DRSP or 75 [micro]g GSD (pooled) demonstrated a nonsignificant increase in the total neutrophil count, from 3.9 x 1000/[mm.sup.3] to 4.7 x 1000/[mm.sup.3] (ES = 0.80; p >0.05) and a nonsignificant decrease in lymphocytes from 2.3 x 1000/[mm.sup.3] to 2.2 x 1000 [mm.sup.3] (ES = -0.10; p >0.05) [3] (Table 2). Therefore, COCs might not increase leukocyte number in PCOS users.

3.3.7. Tumor Necrosis Factor-[alpha]. Only a single study reported the impact of COCs on TNF-[alpha] in PCOS subjects. A combination of 30 [micro]g EE and 2 mg CMA showed a nonsignificant increase of 10.6 pg/ml to 12.0 pg/ml (ES = 1.4; p >0.05) in TNF-[alpha] levels in PCOS patients [29] (Table 4).

3.3.8. Other Chronic Inflammatory Markers. The influence of COCs on MCP-1 was evaluated in a single study [38] and this study, using a combination of 30 [micro]g EE and 150 [micro]g DSG, showed a nonsignificant decrease of MCP-1 from 50 pg/ml to 44 pg/ml (ES = -0.60; p >0.05). The impact of a combination of 35 EE [micro]g and 2 mg CPA for three months on several adhesion molecules markers in PCOS women was also described in a single study [39]. With the EE/CPA combination, soluble intracellular adhesion molecule-1 (sICAM-1) concentrations decreased from 370.1 ng/ml to 364.7 ng/ml (ES = -5.4; p = 0.800), and vascular cell adhesion molecule-1 (sVCAm-1) decreased from 776.3 ng/ml to 742.9 ng/ml (ES = -33.4; p = 0.100) with EE/CMA [32]. A combination of 20 [micro]g EE and 150 [micro]g DSG used for three months significantly decreased sVCAM-1 in PCOS patients, from 583 ng/ml to 522 mg/l (ES = -61.0; p = 0.003) [39]. With EE/CPA, se-selectin increased from 29.1 ng/ml to 37.1 ng/ml (ES = 8.0; p = 0.100), and sp-selectin decreased from 229.4 ng/ml to 189.6 ng/ml (ES = -39.8; p = 0.080). Thus, COCs did not negatively influence the levels of adhesion molecules in PCOS users and could even improve their levels in these patients. COCs containing 35 [micro]g EE and 2 mg CPA were shown to significantly increase sCD40L from 1.33 ng/ml to 2.7 ng/ml (ES = 1.37; p = 0.011) after three months of use in PCOS subjects [40]. Table 2 summarizes all the changes in markers of chronic inflammation following the use of different COCs in PCOS subjects.

4. Discussion

Given the frequent prescription of COCs as first-line treatment for PCOS women not seeking to become pregnant and the possible influence of COCs on low-grade chronic inflammatory markers in these patients, this meta-analysis was justified. The present study examined the data from a total of 27 reports with the inclusion of over than eight hundred PCOS women. Despite the design including patients as their own control, most reports suffer from small sample size with the inclusion of less 50 patients. Unfortunately, some data could not be pooled, and the effect sizes were given individually, according to the different COC compositions. The meta-analysis showed that most COCs were associated with worsening CRP concentrations, nonsignificant changes in Hcy levels, increase in follistatin, and conflicting results regarding IL-6 concentrations. It was also found that COCs did not change significantly the leukocyte total number and TNF-[alpha] levels. Otherwise, COCs tended to decrease PAI-1 and adhesion molecule levels in the blood.

The small sample sizes for comparisons of some inflammatory mediators must be considered. The short-term follow-up times, which were frequently limited to three months, must also be taken into account before definitive conclusions. Finally, the heterogeneity between studies indicates that the results must be interpreted with caution. The strengths of the current study include the evaluation of several mediators of chronic inflammation, the use of established criteria for PCOS diagnosis, and the medium-to-high quality requirement for inclusion of studies. To overcome the impact of the heterogeneity due to clinical or methodological issues, the SMD and a random-effects model were used to compensate for heterogeneity among the studies.

CRP was reported to be 2-fold higher in PCOS patients than controls and is considered a major predictor of metabolic dysfunction and low-grade chronic inflammation in PCOS subjects [41]. Moreover, an additional increase in CRP levels with the use of COCs was reported in these patients [7] and this increase in CRP levels is believed to be due to a direct effect from increased hepatocyte synthesis rather than from the IL-6-dependent inflammatory process [5, 7, 24, 42]. The results of the current meta-analysis endorse previous findings and demonstrate that CRP is increased with the use of COCs. Among eleven studies reporting on CRP levels in PCOS users of COCs containing EE/CPA, only one did not report an increase in this protein [42].

PCOS users of COCs containing EE/DRSP have also been reported to have increased CRP levels [23, 25, 27]. The current meta-analysis has confirmed a significant increase in CRP with this formulation. One single comparison showed that the changes in CRP levels were higher with EE/CMA than the increase observed with the EE/DRSP combination [28]. This meta-analysis also demonstrated that pills containing EE/DSG resulted in a nonsignificant increase in CRP in either lean or obese PCOS patients [30, 31].

Overall, the results of this review demonstrate that the use of COCs by PCOS women increases CRP independent of the formulation and could be associated with an increased risk of atherogenesis. In addition, these findings suggest that the CRP increase varies with the progestin or the estrogen dose [43]. In summary, COCs maintain or even worsen the CRP levels in PCOS women. However, it is yet unclear if the elevated CRP levels observed after COC therapy in women with PCOS actually represent aggravation of the inflammatory process in target tissues [5]. A positive correlation among CRP concentrations, endothelial dysfunction, and the severity of the atherosclerotic process have been reported in several studies [7, 42].

Independent of other risk factors, hyperhomocysteinemia is a recognized risk marker for atherosclerosis [44] due to its capacity to increase oxidative stress in the vascular endothelium and activate platelets [45]. Hcy has been reported to be increased in PCOS patients [46]. The effects of different COC formulations for 3-12 months on Hcy levels in PCOS patients are inconsistent and it seems to depend on the specific composition. The studies with COCs containing EE and CPA have been reported to either increase [23] or decrease [34, 35] Hcy concentrations in PCOS users. In the current meta-analysis, a nonsignificant reduction in Hcy levels was demonstrated with pills containing CPA. COCs containing the progestin DSG also resulted in a small nonsignificant decrease in Hcy levels [31,34]. Oral contraceptives containing the progestin DRSP did not significantly modify the Hcy levels in PCOS subjects. Regarding Hcy, the current data showed that COCs tend to decrease Hcy levels of PCOS subjects.

Acting as an antagonist of aromatase activity, the glycosylated polypeptide follistatin has been reported to facilitate PCOS via a polymorphism in its gene [47]. Follistatin levels were found to be increased in PCOS women, either obese and nonobese [5, 48]. The increased levels of follistatin in PCOS users of COCs may be attributed to increased secretion from hepatocytes rather than to an acute inflammatory response. However, the impact of different COCs on follistatin concentrations in PCOS patients has rarely been reported. A combination of EE and CPA resulted in a significant increase in follistatin in PCOS patients [5, 18]. In young hyperandrogenemic and hyperinsulinemic patients who were not diagnosed with PCOS, follistatin was also reported to be increased by nearly 4-fold after treatment with COCs containing EE and DRSP [18]. The current metaanalysis confirmed a significant increase in follistatin levels of PCOS users of COCs containing the progestin CPA. COCs containing other progestins have not been examined in PCOS women. Therefore, more studies including clinical variables are needed to evaluate the impact of follistatin changes on the cardiovascular system of PCOS women.

The increased IL-6 levels previously observed in PCOS seem to be related to obesity and not to PCOS itself [41]. A nonsignificant decrease in IL-6 of PCOS women has been observed after using different COC formulations [20, 29, 33, 38]. Only one study reported a small and nonsignificant increase in IL-6 with the use of COCs containing the progestin DRSP [37]. Thus, it can be concluded that COCs do not significantly change IL-6 concentrations in PCOS women.

The increased number of leukocytes, primarily neutrophils, observed in PCOS is likely a result of hyperandrogenism [3, 49]. The relative increase in neutrophils is believed to be amplified by COC use. In the current review, the use of a COC formulation containing EE and DRSP did not result in significant changes in total number of neutrophils and lymphocytes [3]. So, regarding the use of COCs by PCOS patients and leukocyte number, no conclusion can be made at this time, but it is likely that COCs do not modify leukocytes in this condition.

TNF-[alpha] participates in the pathogenesis of insulin resistance and is elevated in both obese and nonobese PCOS women [50]. A COC preparation with 30 [micro]g EE and 2 mg CMA nonsignificantly increased TNF-[alpha] in PCOS users [29]. Although TNF-[alpha] is considered an important marker of low-grade chronic inflammation, more studies are needed regarding the use of COC in PCOS, and, due to the paucity of data, no conclusion can be drawn at this time. Several adhesion molecules may be higher in PCOS patients than in healthy subjects [51] and could be a risk factor for atherosclerosis and chronic inflammation in these patients. The influence of COCs on the concentrations of these molecules in PCOS women was reported in a few studies [32, 38]. Only a combination of 30 [micro]g EE and 0.15 mg DSG resulted in a significant decrease in sVCAM-1 levels in PCOS users [38]. Moreover, COCs containing the progestins CPA and CMA did not significantly change sICAM-1, sVICAm-1, and selectins.

PAI-1 levels have been reported to decrease with the use of COCs containing different progestins. Specifically, this decrease was significant with the progestins CPA and DSG [30]. Though, in the current review, the raw effect size varied from -0.2 [25] to -17.4 UI/ml [30], the data could not be pooled for analysis. The extent to which decrease in PAI-1 may improve endothelial function in PCOS patients is still unclear and may be mediated by diminishing its prothrombotic, antithrombotic, and vascular smooth muscle proliferation effects. The weak negative effect of COCs on insulin resistance does not negate the estrogen-induced beneficial effect on PAI-1 [17].

5. Conclusion

In summary, the current study demonstrated that the studies included in the meta-analysis have fair to good quality, although the follow-up had been limited to 3-18 months. C-reactive protein concentrations increased with most preparations of combined oral contraceptives, mainly with the most frequently used in PCOS patients. The decrease in plasminogen activator inhibitor-1 levels with COCs containing DSG, NGM, and DRSP must be evaluated regarding the clinical relevance and impact on the cardiovascular system. Finally, it was demonstrated that most inflammatory markers are modulated by COCs use in PCOS women but the clinical implication for practice needs more extensive investigation and future studies should consider the inclusion of larger sample size, much longer follow-up periods, and the inclusion of objective clinical parameters in their outcome.

Conflicts of Interest

The authors fully declare that there is no either any financial or other conflicts of interest that could be perceived as prejudicing the impartiality of this study.

https://doi.org/10.1155/2018/9591509

Acknowledgments

The authors are grateful to American Journal of Experts for English revision. This research did not received any specific grant from any funding agency.

References

[1] D. A. Dumesic, A. L. Akopians, V. K. Madrigal et al., "Hyperandrogenism accompanies increased intra-abdominal fat storage in normal weight polycystic ovary syndrome women," The Journal of Clinical Endocrinology & Metabolism, vol. 101, no. 11, pp. 4178-4188, 2016.

[2] L. L. Y. Cardozo, D. G. Romero, and J. F. Reckelhoff, "Cardiometabolic features of polycystic ovary syndrome: Role of androgens," Physiology Journal, vol. 32, no. 5, pp. 357-366, 2017.

[3] L. Ibanez, A. M. Jaramillo, A. Ferrer, and F. de Zegher, "High neutrophil count in girls and women with hyperinsulinaemic hyperandrogenism: Normalization with metformin and flutamide overcomes the aggravation by oral contraception," Human Reproduction, vol. 20, no. 9, pp. 2457-2462, 2005.

[4] F. Orio Jr., S. Palomba, T. Cascella et al., "The increase of leukocytes as a new putative marker of low-grade chronic inflammation and early cardiovascular risk in polycystic ovary syndrome," The Journal of Clinical Endocrinology & Metabolism, vol. 90, no. 1, pp. 2-5, 2005.

[5] M.-J. Chen, W.-S. Yang, H.-F. Chen et al., "Increased follistatin levels after oral contraceptive treatment in obese and non-obese women with polycystic ovary syndrome," Human Reproduction, vol. 25, no. 3, pp. 779-785, 2010.

[6] A. Repaci, A. Gambineri, and R. Pasquali, "The role of low-grade inflammation in the polycystic ovary syndrome," Molecular and Cellular Endocrinology, vol. 335, no. 1, pp. 30-41, 2011.

[7] S. Cauci, M. Di Santolo, J. F. Culhane, G. Stel, F. Gonano, and S. Guaschino, "Effects of third-generation oral contraceptives on high-sensitivity C-reactive protein and homocysteine in young women," Obstetrics & Gynecology, vol. 111, no. 4, pp. 857-864, 2008.

[8] C. Christakou and E. Diamanti-Kandarakis, "Structural, biochemical and non-traditional cardiovascular risk markers in PCOS," Current Pharmaceutical Design, vol 19, no. 32, pp. 5764-5774, 2013.

[9] F. Gonzalez, "Inflammation in polycystic ovary syndrome: underpinning of insulin resistance and ovarian dysfunction," Steroids, vol 77, no. 4, pp. 300-305, 2012.

[10] S. F. De Medeiros, "Risks, benefits size and clinical implications of combined oral contraceptive use in women with polycystic ovary syndrome," Reproductive Biology and Endocrinology, vol. 15, no. 1, pp. 1-17, 2017.

[11] C. J. S0rensen, O. B. Pedersen, M. S. Petersen et al., "Combined oral contraception and obesity are strong predictors of low-grade inflammation in healthy individuals: Results from the Danish Blood Donor Study (DBDS)," PLoS ONE, vol. 9, no. 2, 2014.

[12] G. A. Wells, B. Shea, D. OConnell et al., "The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in meta-analyses," http://www.ohri.ca/programs/ clinicaLepidemiology/oxford.asp, 2018 (accessed 03 July 2018).

[13] M. Borenstein, L. V. Hedges, J. P. T. Higgins, and H. R. Rothstein, Introduction to Meta-Analysis, John Wiley & Sons, Ltd., 2009.

[14] R. Der Simonian and R. Kacker, "Random-effects model for meta-analysis of clinical trials: an update," Contemporary Clinical Trials, vol. 28, no. 2, pp. 105-114, 2007.

[15] J. P. T. Higgins, S. G. Thompson, J. J. Deeks, and D. G. Altman, "Measuring inconsistency in meta-analyses," British Medical Journal, vol. 327, no. 7414, pp. 557-560, 2003.

[16] L. Morin-Papunen, K. Rautio, A. Ruokonen, P. Hedberg, M. Puukka, and J. S. Tapanainen, "Metformin Reduces Serum C-Reactive Protein Levels in Women with Polycystic Ovary Syndrome," The Journal of Clinical Endocrinology & Metabolism, vol. 88, no. 10, pp. 4649-4654, 2003.

[17] H. J. Teede, C. Meyer, S. K. Hutchison, S. Zoungas, B. P. McGrath, and L. J. Moran, "Endothelial function and insulin resistance in polycystic ovary syndrome: the effects of medical therapy," Fertility and Sterility, vol. 93, no. 1, pp. 184-191, 2010.

[18] L. Ibanez, M. Diaz, G. Sebastiani et al., "Treatment of androgen excess in adolescent girls: Ethinylestradiol-cyproteroneacetate versus low-dose pioglitazone-flutamide-metformin," The Journal of Clinical Endocrinology & Metabolism, vol. 96, no. 11, pp. 3361-3366, 2011.

[19] F. Gode, C. Karagoz, C. Posaci et al., "Alteration of cardiovascular risk parameters in women with polycystic ovary syndrome who were prescribed to ethinyl estradiol-cyproterone acetate," Archives of Gynecology and Obstetrics, vol. 284, no. 4, pp. 923-929, 2011.

[20] M. Diaz, M. R. Chacan, A. Lapez-Bermejo et al., "Ethinyl estradiol-cyproterone acetate versus low-dose pioglitazoneflutamide-metformin for adolescent girls with androgen excess: Divergent effects on CD163, TWEAK receptor, ANGPTL4, and LEPTIN Expression in subcutaneous adipose tissue," The Journal of Clinical Endocrinology & Metabolism, vol. 97, no. 10, pp. 3630-3638, 2012.

[21] L. Ibanez, M. Diaz, G. Sebastiani, M. V. Marcos, A. Lapez-Bermejo, and F. De Zegher, "Oral contraception vs insulin sensitization for 18 months in nonobese adolescents with androgen excess: Posttreatment differences in C-reactive protein, intima-media thickness, visceral adiposity, insulin sensitivity, and menstrual regularity," The Journal of Clinical Endocrinology & Metabolism, vol. 98, no. 5, pp. E902-E907, 2013.

[22] C. Christakou, A. Kollias, C. Piperi, I. Katsikis, D. Panidis, and E. Diamanti-Kandarakis, "The benefit-to-risk ratio of common treatments in PCOS: effect of oral contraceptives versus metformin on atherogenic markers," Hormones, vol. 13, no. 4, pp. 488-497, 2014.

[23] K. Kahraman, Y. E. Sukur, C. S. Atabekoglu et al., "Comparison of two oral contraceptive forms containing cyproterone acetate and drospirenone in the treatment of patients with polycystic ovary syndrome: A randomized clinical trial," Archives of Gynecology and Obstetrics, vol. 290, no. 2, pp. 321-328, 2014.

[24] J. A. Dardzinska, D. Rachon, M. Kuligowska-Jakubowska et al., "Effects of metformin or an oral contraceptive containing cyproterone acetate on serum C-reactive protein, interleukin-6 and soluble vascular cell adhesion molecule-1 concentrations in women with polycystic ovary syndrome," Experimental and Clinical Endocrinology & Diabetes, vol. 122, no. 2, pp. 118-125, 2014.

[25] F. Orio, G. Muscogiuri, F. Giallauria et al., "Oral contraceptives versus physical exercise on cardiovascular and metabolic risk factors in women with polycystic ovary syndrome: a randomized controlled trial," Clinical Endocrinology, vol. 85, no. 5, pp. 764-771, 2016.

[26] H. Tfayli, J. W. Ulnach, S. Lee, K. Sutton-Tyrrell, and S. Arslanian, "Drospirenone/ethinyl estradiol versus rosiglitazone treatment in overweight adolescents with polycystic ovary syndrome: Comparison of metabolic, hormonal, and cardiovascular risk factors," The Journal of Clinical Endocrinology & Metabolism, vol. 96, no. 5, pp. 1311-1319, 2011.

[27] A. Harmanci, N. Cinar, M. Bayraktar, and B. O. Yildiz, "Oral contraceptive plus antiandrogen therapy and cardiometabolic risk in polycystic ovary syndrome," Clinical Endocrinology, vol. 78, no. 1, pp. 120-125, 2013.

[28] R. Yildizhan, A. I. Gokce, B. Yildizhan, and N. Cim, "Comparison of the effects of chlormadinone acetate versus drospirenone containing oral contraceptives on metabolic and hormonal parameters in women with PCOS for a period of two-year follow-up," Gynecological Endocrinology, vol. 31, no. 5, pp. 396-400, 2015.

[29] C. S. Vieira, W. P. Martins, J. B. F. Fernandes et al., "The effects of 2 mg chlormadinone acetate/30 mcg ethinylestradiol, alone or combined with spironolactone, on cardiovascular risk markers in women with polycystic ovary syndrome," Contraception, vol. 86, no. 3, pp. 268-275, 2012.

[30] K. Hoeger, K. Davidson, L. Kochman, T. Cherry, L. Kopin, and D. S. Guzick, "The impact of metformin, oral contraceptives, and lifestyle modification on polycystic ovary syndrome in obese adolescent women in two randomized, placebo-controlled clinical trials," The Journal of Clinical Endocrinology & Metabolism, vol. 93, no. 11, pp. 4299-4306, 2008.

[31] S. Kilic, N. Yilmaz, E. Zulfikaroglu, G. Erdogan, M. Aydin, and S. Batioglu, "Inflammatory-metabolic parameters in obese and nonobese normoandrogenemic polycystic ovary syndrome during metformin and oral contraceptive treatment," Gynecological Endocrinology, vol. 27, no. 9, pp. 622-629, 2011.

[32] B. Banaszewska, L. Pawelczyk, R. Z. Spaczynski, J. Dziura, and A. J. Duleba, "Effects of simvastatin and oral contraceptive agent on polycystic ovary syndrome: Prospective, randomized, crossover trial," The Journal of Clinical Endocrinology & Metabolism, vol. 92, no. 2, pp. 456-461, 2007.

[33] P. A. Essah, J. A. Arrowood, K. I. Cheang, S. S. Adawadkar, D. W. Stovall, and J. E. Nestler, "Effect of combined metformin and oral contraceptive therapy on metabolic factors and endothelial function in overweight and obese women with polycystic ovary syndrome," Fertility and Sterility, vol. 96, no. 2, pp. 501-504, 2011.

[34] A. Cagnacci, A. Tirelli, A. Renzi, A. M. Paoletti, and A. Volpe, "Effects of two different oral contraceptives on homocysteine metabolism in women with polycystic ovary syndrome," Contraception, vol. 73, no. 4, pp. 348-351, 2006.

[35] O. B. Gul, A. Somunkiran, O. Yucel, F. Demirci, and I. Ozdemir, "The effect of ethinyl estradiol-cyproterone acetate treatment on homocysteine levels in women with polycystic ovary syndrome," Archives of Gynecology and Obstetrics, vol. 277, no. 1, pp. 25-30, 2008.

[36] F. Mancini, A. Cianciosi, N. Persico, F. Facchinetti, P. Busacchi, and C. Battaglia, "Drospirenone and cardiovascular risk in lean and obese polycystic ovary syndrome patients: a pilot study," American Journal of Obstetrics & Gynecology, vol. 202, no. 2, pp. 169.e1-169.e8, 2010.

[37] L. Ibanez and F. De Zegher, "Ethinylestradiol-drospirenone, flutamide-metformin, or both for adolescents and women with hyperinsulinemic hyperandrogenism: Opposite effects on adipocytokines and body adiposity," The Journal of Clinical Endocrinology & Metabolism, vol. 89, no. 4, pp. 1592-1597, 2004.

[38] D. Glintborg, H. Mumm, M. L. Altinok, B. Richelsen, J. M. Bruun, and M. Andersen, "Adiponectin, interleukin-6, monocyte chemoattractant protein-1, and regional fat mass during 12-month randomized treatment with metformin and/or oral contraceptives in polycystic ovary syndrome," Journal of Endocrinological Investigation, vol. 37, no. 8, pp. 757-764, 2014.

[39] O. Bilgir, L. Kebapcilar, C. Taner et al., "The effect of ethinylestradiol (EE)/cyproterone acetate (CA) and EE/CA plus metformin treatment on adhesion molecules in cases with polycystic ovary syndrome (PCOS)," Internal Medicine, vol. 48, no. 14, pp. 1193-1199, 2009.

[40] L. Kebapcilar, C. E. Taner, A. G. Kebapcilar, A. Alacacioglu, and I. Sari, "Comparison of four different treatment regimens on coagulation parameters, hormonal and metabolic changes in women with polycystic ovary syndrome," Archives of Gynecology and Obstetrics, vol. 281, no. 1, pp. 35-42, 2010.

[41] H. F. Escobar-Morreale, M. Luque-Ramirez, and F. Gonzalez, "Circulating inflammatory markers in polycystic ovary syndrome: a systematic review and meta-analysis," Fertility and Sterility, vol. 95, no. 3, pp. 1048.e2-1058.e2, 2011.

[42] M. van Rooijen, L. O. Hansson, J. Frostegard, A. Silveira, A. Hamsten, and K. Bremme, "Treatment with combined oral contraceptives induces a rise in serum C-reactive protein in the absence of a general inflammatory response," Journal of Thrombosis and Haemostasis, vol. 4, no. 1, pp. 77-82, 2006.

[43] G. M. Prelevia, M. I. Wurzburger, D. Trpkovic, and L. Balint-Periac, "Effects of a Iow-dose estrogen-antiandrogen combination (Diane-35) on lipid and carbohydrate metabolism in patients with polycystic ovary syndrome," Gynecological Endocrinology, vol. 4, no. 3, pp. 157-168, 2009.

[44] M. R. Nehler, L. M. Taylor Jr., and J. M. Porter, "Homocysteinemia as a risk factor for atherosclerosis: A review," Cardiovascular Pathology, vol. 6, no. 1, pp. 1-9, 1997.

[45] A. Coppola, G. Davi, V. De Stefano, F. P. Mancini, A. M. Cerbone, and G. Di Minno, "Homocysteine, coagulation, platelet function, and thrombosis," Seminars in Thrombosis and Hemostasis, vol. 26, no. 3, pp. 243-254, 2000.

[46] Y. Meng, X. Chen, Z. Peng, X. Liu, Y. Sun, and S. Dai, "Association between high serum homocysteine levels and biochemical characteristics in women with polycystic ovarian syndrome: A systematic review and meta-analysis," PLoS ONE, vol. 11, no. 6, 2016.

[47] M. R. Jones, S. G. Wilson, B. H. Mullin, R. Mead, G. F. Watts, and B. G. A. Stuckey, "Polymorphism of the follistatin gene in polycystic ovary syndrome," Molecular Human Reproduction, vol. 13, no. 4, pp. 237-241, 2007.

[48] H. Teede, S. Ng, M. Hedger, and L. Moran, "Follistatin and activins in polycystic ovary syndrome: Relationship to metabolic and hormonal markers," Metabolism--Clinical and Experimental, vol. 62, no. 10, pp. 1394-1400, 2013.

[49] Z. H. Huang, B. Manickam, and V. Ryvkin, "PCOS is associated with increased CD11c expression and crown-like structures in adipose tissue and increased central abdominal fat depots independent of obesity," The Journal of Clinical Endocrinology & Metabolism, vol. 98, no. 1, pp. E17-E24, 2013.

[50] F. Gonzalez, K. Thusu, E. Abdel-Rahman, A. Prabhala, M. Tomani, and P. Dandona, "Elevated serum levels of tumor necrosis factor alpha in normal-weight women with polycystic ovary syndrome," Metabolism--Clinical and Experimental, vol. 48, no. 4, pp. 437-441, 1999.

[51] E. Diamanti-Kandarakis, K. Alexandraki, C. Piperi et al., "Inflammatory and endothelial markers in women with polycystic ovary syndrome," European Journal of Clinical Investigation, vol. 36, no. 10, pp. 691-697, 2006.

Sebastiao Freitas de Medeiros [iD], (1,2) Matheus Antonio Souto de Medeiros, (2) Nayara de Souza Santos, (2) Bruna Barcelo Barbosa, (2) and Marcia Marly Winck Yamamoto (2)

(1) Department of Gynecology and Obstetrics, Medical School, Federal University of Mato Grosso, Cuiaba, MT, Brazil

(2) Tropical Institute of Reproductive Medicine, Cuiaba, MT, Brazil

Correspondence should be addressed to Sebastiao Freitas de Medeiros; de.medeiros@terra.com.br

Received 7 August 2018; Accepted 4 October 2018; Published 25 November 2018

Academic Editor: Han J. Moshage

Caption: Figure 1: Flow chart.
Table 1: Characteristics of clinical studies comparing before and
after changes in markers of chronic inflammation in PCOS patients
with the use of different combined oral contraceptives.

Study                           Country

Morin-Papunen et al, 2003       Finland
Ibanez and de Zegher, 2004   Belgium/ Spain
Ibanez et al, 2005           Spain/ Belgium
Cagnacciet al, 2006              Italy
Cagnacciet al, 2006              Italy
Banaszewska et al, 2007           USA
Gul et al, 2008                  Turkey
Hoeger et al, 2008                USA
Bilgir et al, 2009               Turkey
Kebapcilar et al, 2010           Turkey
Chen et al, 2010                 Taiwan
Mancini et al, 2010              Italy
Teede et al, 2010              Australia
Tfayli et al, 2011                USA
Essah et al, 2011                 USA
Kilic et al, 2011                Turkey
Ibanez et al, 2011           Belgium/ Spain
Code et al, 2011                 Turkey
Vieira et al, 2012               Brazil
Diaz et al, 2012             Spain/ Belgium
Harmanci et al, 2013             Turkey
Ibanez et al, 2013           Spain/ Belgium
Christakou et al, 2014           Greece
Christakou et al, 2014           Greece
Glintborg et al, 2014           Denmark
Kahraman et al, 2014             Turkey
Kahraman et al, 2014             Turkey
Dardzinska et al, 2014          Finland
Yildzhan et al, 2015             Turkey
Yildzhan et al, 2015             Turkey
Orio et al, 2016                 Italy

Study                                 Study design               No
                                                              patients

Morin-Papunen et al, 2003        Randomized open trial           10
Ibanez and de Zegher, 2004       Randomized open trial           16
Ibanez et al, 2005               Randomized open trial           21
Cagnacciet al, 2006              Randomized open trial           10
Cagnacciet al, 2006              Randomized open trial           10
Banaszewska et al, 2007           Crossover open trial           24
Gul et al, 2008                Non-randomized open trial         30
Hoeger et al, 2008               Randomized open trial           11
Bilgir et al, 2009               Randomized open trial           20
Kebapcilar et al, 2010           Randomized open trial           28
Chen et al, 2010               Non-randomized open trial         56
Mancini et al, 2010            Non-randomized open trial         28
Teede et al, 2010                Randomized open trial           26
Tfayli et al, 2011             Non-randomized open trial         20
Essah et al, 2011                Randomized open trial           10
Kilic et al, 2011                Randomized open trial           49
Ibanez et al, 2011               Randomized open trial           17
Code et al, 2011               Non-randomized open trial         40
Vieira et al, 2012               Randomized open trial           21
Diaz et al, 2012                 Randomized open trial           14
Harmanci et al, 2013           Non-randomized open trial         23
Ibanez et al, 2013               Randomized open trial           17
Christakou et al, 2014         Non-randomized open trial         40
Christakou et al, 2014         Non-randomized open trial         40
Glintborg et al, 2014            Randomized open trial           23
Kahraman et al, 2014             Randomized open trial           19
Kahraman et al, 2014             Randomized open trial           20
Dardzinska et al, 2014            Crossover open trial           42
Yildzhan et al, 2015             Randomized open trial           56
Yildzhan et al, 2015             Randomized open trial           50
Orio et al, 2016             Randomized double-blind trial       47

Study                                       COC
                                       type and dose

Morin-Papunen et al, 2003        EE 35 [micro]g/ CPA 2 mg
Ibanez and de Zegher, 2004       EE 30 [micro]g/ DRSP 3 mg
Ibanez et al, 2005               EE 30 [micro]g/ DRSP 3 mg
Cagnacciet al, 2006              EE 35 [micro]g/ CPA 2 mg
Cagnacciet al, 2006             EE 30 [micro]g/ DSG 0.15 mg
Banaszewska et al, 2007         EE 20 [micro]g/ DSG 0.15 mg
Gul et al, 2008                  EE 35 [micro]g/ CPA 2 mg
Hoeger et al, 2008              EE 30 [micro]g/ DSG 0.15 mg
Bilgir et al, 2009               EE 35 [micro]g/ CPA 2 mg
Kebapcilar et al, 2010           EE 35 [micro]g/ CPA 2 mg
Chen et al, 2010                 EE 35 [micro]g/ CPA 2 mg
Mancini et al, 2010                 EE 30/ig/ DRSP 3 mg
Teede et al, 2010                EE 35 [micro]g/ CPA 2 mg
Tfayli et al, 2011               EE 30 [micro]g/ DRSP 3 mg
Essah et al, 2011            EE 35 [micro]g/ NGM 0.18-0.25 mg
Kilic et al, 2011               EE 30 [micro]g/ DSG 0.15 mg
Ibanez et al, 2011               EE 35 [micro]g/ CPA 2 mg
Code et al, 2011                 EE 35 [micro]g/ CPA 2 mg
Vieira et al, 2012               EE 30 [micro]g/ CMA 2 mg
Diaz et al, 2012                 EE 35 [micro]g/ CPA 2 mg
Harmanci et al, 2013             EE 30 [micro]g/ DRSP 3 mg
Ibanez et al, 2013               EE 35 [micro]g/ CPA 2 mg
Christakou et al, 2014           EE 35 [micro]g/ CPA 2 mg
Christakou et al, 2014           EE 30 [micro]g/ DRSP 3 mg
Glintborg et al, 2014           EE 30 [micro]g/ DSG 0.15 mg
Kahraman et al, 2014             EE 35 [micro]g/ CPA 2 mg
Kahraman et al, 2014             EE 30 [micro]g/ DRSP 3 mg
Dardzinska et al, 2014           EE 35 [micro]g/ CPA 2 mg
Yildzhan et al, 2015             EE 30 [micro]g/ DRSP 3 mg
Yildzhan et al, 2015             EE 30 [micro]g/ CMA 2 mg
Orio et al, 2016                 EE 35 [micro]g/ CPA 2 mg

Study                        Duration          PCOS
                              months    diagnostic criteria

Morin-Papunen et al, 2003       06           NIH (a)
Ibanez and de Zegher, 2004      09        Rotterdam (a)
Ibanez et al, 2005              03         Rotterdam (a)
Cagnacciet al, 2006             06           NIH (a)
Cagnacciet al, 2006             06           NIH (a)
Banaszewska et al, 2007         03           Rotterdam
Gul et al, 2008                 03           Rotterdam
Hoeger et al, 2008              06           NIH (a)
Bilgir et al, 2009              03           Rotterdam
Kebapcilar et al, 2010          03           Rotterdam
Chen et al, 2010                03           Rotterdam
Mancini et al, 2010             06           Rotterdam
Teede et al, 2010               06              NIH
Tfayli et al, 2011              06              NIH
Essah et al, 2011               03        Rotterdam (a)
Kilic et al, 2011               06           Rotterdam
Ibanez et al, 2011              06           NIH (a)
Code et al, 2011                06           Rotterdam
Vieira et al, 2012              12           Rotterdam
Diaz et al, 2012                12           NIH (a)
Harmanci et al, 2013            06           Rotterdam
Ibanez et al, 2013              18           NIH (a)
Christakou et al, 2014          06              NIH
Christakou et al, 2014          06              NIH
Glintborg et al, 2014           12           Rotterdam
Kahraman et al, 2014            12              AES
Kahraman et al, 2014            12           Rotterdam
Dardzinska et al, 2014          04           Rotterdam
Yildzhan et al, 2015            06           Rotterdam
Yildzhan et al, 2015            06           Rotterdam
Orio et al, 2016                06         Rotterdam (a)

(a): not clearly stated.

Table 2: Risk of bias assessment of the included studies (a).

                                              Selection

Study, Year                  Representativeness     Selection of
                             of exposed cohort    nonexposed cohort

Morin-Papunen et al, 2003             *                   *
Ibanez and de Zegher, 2004            *                   *
Ibanez et al, 2005                    *                   *
Cagnacciet al, 2006                   *                   *
Banaszewska et al, 2007               *                   *
Gul et al, 2008                       *                   *
Hoeger et al, 2008                    *                   *
Bilgir et al, 2009                    *                   *
Kebapcilar et al, 2010                *                   *
Chen et al, 2010                      *                   *
Mancini et al, 2010                   *                   *
Teede et al, 2010                     *                   *
Tfayli et al, 2011                    *                   *
Essah et al, 2011                     *                   *
Kilic et al, 2011                     *                   *
Ibanez et al, 2011                    *                   *
Code et al, 2011                      *                   *
Vieira et al, 2012                    *                   *
Diaz et al, 2012                      *                   *
Harmanci et al, 2013                  *                   *
Ibanez et al, 2013                    *                   *
Christakou et al, 2014                *                   *
Glintborg et al, 2014                 *                   *
Kahraman et al, 2014                  *                   *
Dardzinska et al, 2014                *                   *
Yildzhan et al, 2015                  *                   *
Orio et al, 2016                      *                   *

                               Selection            Outcome

Study, Year                  Ascertainment of   Outcome not present
                                 exposure           at baseline

Morin-Papunen et al, 2003            *                  --
Ibanez and de Zegher, 2004           *                  --
Ibanez et al, 2005                   *                  --
Cagnacciet al, 2006                  *                  --
Banaszewska et al, 2007              *                  --
Gul et al, 2008                      *                  --
Hoeger et al, 2008                   *                  --
Bilgir et al, 2009                   *                  --
Kebapcilar et al, 2010               *                  --
Chen et al, 2010                     *                  --
Mancini et al, 2010                  *                  --
Teede et al, 2010                    *                  --
Tfayli et al, 2011                   *                  --
Essah et al, 2011                    *                  --
Kilic et al, 2011                    *                  --
Ibanez et al, 2011                   *                  --
Code et al, 2011                     *                  --
Vieira et al, 2012                   *                  --
Diaz et al, 2012                     *                  --
Harmanci et al, 2013                 *                  --
Ibanez et al, 2013                   *                  --
Christakou et al, 2014               *                  --
Glintborg et al, 2014                *                  --
Kahraman et al, 2014                 *                  --
Dardzinska et al, 2014               *                  --
Yildzhan et al, 2015                 *                  --
Orio et al, 2016                     *                  --

                                             Outcome

Study, Year                     Comparability of      Assessment of
                                cohorts (matched         outcome
                                      for)

Morin-Papunen et al, 2003         BMI, age **               *
Ibanez and de Zegher, 2004           BMI *                  *
Ibanez et al, 2005                 BMI, age"                *
Cagnacciet al, 2006                BMI, age"                *
Banaszewska et al, 2007              Age *                  *
Gul et al, 2008                   BMI, age **               *
Hoeger et al, 2008                BMI, age **               *
Bilgir et al, 2009                BMI, age **               *
Kebapcilar et al, 2010            BMI, age **               *
Chen et al, 2010              Weight, BMI, age ***          *
Mancini et al, 2010                  BMI *                  *
Teede et al, 2010                 BMI, age **               *
Tfayli et al, 2011           BMI, age, fat mass ***         *
Essah et al, 2011                    BMI *                  *
Kilic et al, 2011                    BMI *                  *
Ibanez et al, 2011                BMI, age **               *
Code et al, 2011                BMI, WC, age ***            *
Vieira et al, 2012                BMI, age **               *
Diaz et al, 2012                  BMI, age **               *
Harmanci et al, 2013           BMI, WHR, age ***            *
Ibanez et al, 2013                BMI, age **               *
Christakou et al, 2014            BMI, age **               *
Glintborg et al, 2014           BMI, fat mass **            *
Kahraman et al, 2014           BMI, WHR, age ***            *
Dardzinska et al, 2014            BMI, age **               *
Yildzhan et al, 2015           BMI, WHR, age ***            *
Orio et al, 2016                BP, BMI, age ***            *

                                              Outcome

Study, Year                  Sufficient follow-   Adequate    Total
                                up duration       follow-up   score

Morin-Papunen et al, 2003             *               *         8
Ibanez and de Zegher, 2004            *               *         7
Ibanez et al, 2005                   --               *         7
Cagnacciet al, 2006                   *               *         8
Banaszewska et al, 2007              --               *         6
Gul et al, 2008                      --               *         7
Hoeger et al, 2008                    *               *         8
Bilgir et al, 2009                   --               *         7
Kebapcilar et al, 2010               --               *         7
Chen et al, 2010                     --               *         8
Mancini et al, 2010                   *               *         7
Teede et al, 2010                     *               *         8
Tfayli et al, 2011                    *               *         9
Essah et al, 2011                    --               *         6
Kilic et al, 2011                     *               *         7
Ibanez et al, 2011                    *               *         8
Code et al, 2011                      *               *         9
Vieira et al, 2012                    *               *         8
Diaz et al, 2012                      *               *         8
Harmanci et al, 2013                  *               *         9
Ibanez et al, 2013                    *               *         8
Christakou et al, 2014                *               *         8
Glintborg et al, 2014                 *               *         8
Kahraman et al, 2014                  *               *         9
Dardzinska et al, 2014                *               *         7
Yildzhan et al, 2015                  *               *         9
Orio et al, 2016                      *               *         9

(a): Newcastle-Ottawa scale, BMI: body mass index, WC: waist
circumference, WHR: waist-hip ratio, and BP: blood pressure.

Table 3: Summary of the influence of different oral contraceptive
combination on C-reactive protein in polycystic ovary syndrome *.

Inflammatory marker

Study                                    Formulation

CRP (mg/1)
  Morin-Papunen et al, 2003       EE 35 [micro]g/ CPA 2 mg
  Chen et al, 2010                EE 35 [micro]g/ CPA 2 mg
  Teede et al, 2010               EE 35 [micro]g/ CPA 2 mg
  Ibanez et al, 2011              EE 35 [micro]g/ CPA 2 mg
  Gode et al, 2011                EE 35 [micro]g/ CPA 2 mg
  Diaz et al, 2012                EE 35 [micro]g/ CPA 2 mg
  Ibanez et al, 2013              EE 35 [micro]g/ CPA 2 mg
  Christakou et al, 2014          EE 35 [micro]g/ CPA 2 mg
  Kahraman et al, 2014            EE 35 [micro]g/ CPA 2 mg
  Dardzinska et al, 214           EE 35 [micro]g/ CPA 2 mg
  Orio et al, 2016                EE 35 [micro]g/ CPA 2 mg
  Tfayli et al, 2011              EE 30 [micro]g/ DRSP 3 mg
  Harmanci et al, 2013            EE 30 [micro]g/ DRSP 3 mg
  Kahraman et al, 2014            EE 30 [micro]g/ DRSP 3 mg
  Christakou et al, 2014          EE 30 [micro]g/ DRSP 3 mg
  Yildzhan et al, 2015            EE 30 [micro]g/ DRSP 3 mg
  Orio et al, 2016                EE 30 [micro]g/ DRSP 3 mg
  Vieira et al, 2012              EE 30 [micro]g/ CMA 2 mg
  Yildzhan et al, 2015            EE 30 [micro]g/ CMA 2 mg
  Hoeger et al, 2008             EE 30 [micro]g/ DSG 0.15 mg
  Kilic et al, 2011              EE 30 [micro]g/ DSG 0.15 mg
  Banaszewska et al, 2007        EE 20 [micro]g/ DSG 0.15 mg
  Essah et al, 2011           EE 30 [micro]g/ DSG 0.15-0.25 mg

Study                         Before   After   ES **   P ***

CRP (mg/1)
  Morin-Papunen et al, 2003    2.91    4.58    +1.67   <0.001
  Chen et al, 2010             0.7      1.2    +0.5    0.002
  Teede et al, 2010            3.5      4.7    +1.2    0.001
  Ibanez et al, 2011           0.9      1.7    +0.8    <0.05
  Gode et al, 2011             3.31    3.87    +0.56   >0.05
  Diaz et al, 2012             0.9      2.6    +1.7    <0.05
  Ibanez et al, 2013           0.9      3.2    +2.3    <0.01
  Christakou et al, 2014       1.36    2.63    +1.27   <0.001
  Kahraman et al, 2014         1.21    3.31    +2.1    <0.05
  Dardzinska et al, 214        0.77    1.70    +0.93   <0.001
  Orio et al, 2016             1.80    1.90    +0.1    >0.05
  Tfayli et al, 2011           1.7      3.8    +2.1    <0.001
  Harmanci et al, 2013         0.50     1.5     +1     <0.05
  Kahraman et al, 2014         0.93    1.22    +0.29   0.040
  Christakou et al, 2014       1.09    1.93    +0.84   <0.001
  Yildzhan et al, 2015         3.77    4.32    +0.55   0.005
  Orio et al, 2016             1.8      1.9    +0.1    <0.05
  Vieira et al, 2012           2.1      6.0    +3.9    <0.05
  Yildzhan et al, 2015         4.26    0.72    -3.54   0.004
  Hoeger et al, 2008           6.8      9.5    +2.7    <0.05
  Kilic et al, 2011            1.67    3.23    +1.56   <0.001
  Banaszewska et al, 2007      1.61    1.70    +0.09    0.74
  Essah et al, 2011            4.9      5.6    +0.7    0.280

* All abbreviations are given in the text.

** ES: effect-size.

*** p: values taken from each study.

Table 4: Influence of different oral contraceptive combination on
markers of chronic inflammation in polycystic ovary syndrome*.

Inflammatory marker
Study                                        Formulation

IL-6 (pg/ml)
  Diaz et al, 2012                    EE 35 [micro]g/ CPA 2 mg
  Ibanez and de Zegher, 2004          EE 30 [micro]g/ DRSP 3 mg
  Vieira et al, 2012                  EE 30 [micro]g/ CMA 2 mg
  Glintborg et al, 2014              EE 30 [micro]g/ DSG 0.15 mg
  Essah et al, 2011               EE 30 [micro]g/ NGM 0.15-0.25 mg
PAI-1 (mg/ml)/(IU/ml)
  Hoeger et al, 2008                 EE 30 [micro]g/ DSG 0.15 mg
  Essah et al, 2011                EE 30 [micro]g/ NGM 0.18-25 mg
  Orio et al, 2016                    EE 30 [micro]g/ DRSP 3 mg
Neutrophils (nx1000/[mm.sup.3])
  Ibanez et al, 2005                  EE 30 [micro]g/ DRSP 3 mg/
                                            GSD 75 [micro]g
Lymphocytes (nx1000/[mm.sup.3])
  Ibanez et al, 2005                 EE 30 [micro]g/ DRSP 3 mg/
                                            GSD 75 [micro]g
TNF-[alpha] (pg/ml)
  Vieira et al, 2012                  EE 30 [micro]g/ CMA 2 mg
MCP-1 (pg/ml)
  Glintborg et al, 2014              EE 30 [micro]g/ DSG 0.15 mg
sICAM-1(ng/ml)
  Bilgir et al, 2009                  EE 35 [micro]g/ CPA 2 mg
sVCAM-1(ng/ml)
  Bilgir et al, 2009                 EE 30 [micro]g/ DSG 0.15 mg
  Banaszewska et al, 2007             EE 30 [micro]g/ CMA 2 mg
Se-selectin (ng/ml)
  Bilgir et al, 2009                  EE 35 [micro]g/ CPA 2 mg
Sp-selectin (ng/ml)
  Bilgir et al, 2009                  EE 35 [micro]g/ CPA 2 mg
sCD40L (ng/ml)
  Kebapcilar et al, 2010               EE 35 [micro]/ CPA 2 mg

Inflammatory marker
Study                           Before   After   ES **   p **

IL-6 (pg/ml)
  Diaz et al, 2012               1.67    1.39    -0.28   0.911
  Ibanez and de Zegher, 2004     0.95    1.09    +0.14   >0.05
  Vieira et al, 2012             1.90    1.70    -0.20   >0.05
  Glintborg et al, 2014          1.70    1.40    -0.30   >0.05
  Essah et al, 2011              2.10    1.60    -0.50   0.060
PAI-1 (mg/ml)/(IU/ml)
  Hoeger et al, 2008             46.9    29.5    -17.4   <0.05
  Essah et al, 2011              81.2    76.5    -04.7   0.520
  Orio et al, 2016               02.6    02.4    -0.2    >0.05
Neutrophils (nx1000/[mm.sup.3])
  Ibanez et al, 2005             03.9    04.7    +0.80   >0.05
Lymphocytes (nx1000/[mm.sup.3])
  Ibanez et al, 2005             02.3    02.2    -0.1    >0.05
TNF-[alpha] (pg/ml)
  Vieira et al, 2012             10.6    12.0    +1.4    >0.05
MCP-1 (pg/ml)
  Glintborg et al, 2014          50.0    44.0    -06.0   >0.05
sICAM-1(ng/ml)
  Bilgir et al, 2009            370.1    364.7   -5.4    0.800
sVCAM-1(ng/ml)
  Bilgir et al, 2009            583.0    522.0   -61.0   0.003
  Banaszewska et al, 2007       776.3    742.9   -33.4   0.100
Se-selectin (ng/ml)
  Bilgir et al, 2009             29.1    37.1    +8.0    0.100
Sp-selectin (ng/ml)
  Bilgir et al, 2009            229.4    189.6   -39.8   0.080
sCD40L (ng/ml)
  Kebapcilar et al, 2010         1.33    2.70    +1.37   0.011

* All abbreviations are given in the text.

** ES: effect-size.

*** p: values taken from each study.

Figure 2: (a) Forest plot of the standardized difference means of
C-reactive protein taken before and after the use of combined oral
contraceptive in PCOS patients. (b) Forest plot of the standardized
difference means of C-reactive protein taken before and after the
use of combined oral contraceptive in PCOS patients.

(a)

                                     Statistics for each study

Study Name                   Std diff in   Standard      Variance
DRSP                            means       error

Morin-Papunen et al., 2003     -1.512       0.463         0.214
Chen et al., 2010              -0.434       0.140         0.020
Teede et al., 2010             -0.731       0.221         0.049
Ibanez et al., 2011            -0.542       0.260         0.067
Gode et al., 2011               0.306       0.162         0.026
Diaz et al., 2012              -0.610       0.291         0.085
Ibanez et al., 2013            -0.947       0.294         0.087
Christakou et al., 2014        -0.428       0.165         0.027
Kahraman et al., 2014          -0.757       0.260         0.068
Dardzinzka et al., 2014        -0.547       0.165         0.027
Orio et al., 2016              -0.287       0.149         0.022
Rondom-effects model           -0.511       0.117         0.014

                                Statistics for each study

Study Name                    Lower limit     Upper    Z-value
DRSP                                          limit

Morin-Papunen et al., 2003       -2.419       -0.605   -3.266
Chen et al., 2010                -0.708       -0.160   -3.103
Teede et al., 2010               -1.163       -0.298   -3.310
Ibanez et al., 2011              -1.051       -0.033   -2.087
Gode et al., 2011                -0.011       0.623     1.893
Diaz et al., 2012                -1.180       -0.039   -2.095
Ibanez et al., 2013              -1.551       -0.397   -3.307
Christakou et al., 2014          -0.752       -0.104   -2.592
Kahraman et al., 2014            -1.267       -0.247   -2.908
Dardzinzka et al., 2014          -0.871       -0.223   -3.306
Orio et al., 2016                -0.579       0.005    -1.930
Rondom-effects model             -0.740       -0.282   -4.367

                              Statistics for each study

Study Name                   p-value   Total     Relative
DRSP                                              weight

Morin-Papunen et al., 2003    0.001     10         4.36
Chen et al., 2010             0.002     56        11.50
Teede et al., 2010            0.001     26         9.24
Ibanez et al., 2011           0.037     17         8.20
Gode et al., 2011             0.058     40        10.89
Diaz et al., 2012             0.036     14         7.43
Ibanez et al., 2013           0.001     17         7.35
Christakou et al., 2014       0.010     40        10.80
Kahraman et al., 2014         0.004     19         8.19
Dardzinzka et al., 2014       0.001     42        10.79
Orio et al., 2016             0.054     47        11.25
Rondom-effects model          0.000

Heterogeneity: [Tau].sup.2] = 0.099. Q = 34.94.
[I.sup.2] = 71.38%. p < 0.001

                               Statistics for each study

Study Name                Std diff in   Standard      Variance
DRSP                        means       error

Tfalyli et al., 2011       -0.868       0.262         0.069
Harmanci et al., 2013      -0.455       0.219         0.048
Kahraman et al., 2014      -0.493       0.237         0.056
Christakou et al., 2014    -0.563       0.170         0.029
Yildzhan et al., 2015      -0.401       0.139         0.019
Orio et al., 2016          -0.287       0.149         0.022
Rondom-effects model       -0.455       0.074         0.005

                               Statistics for each study

Study Name                 Lower limit     Upper    Z-value   p-value
DRSP                                       limit

Tfalyli et al., 2011          -1.383       -0.354   -3.309     0.001
Harmanci et al., 2013         -0.884       -0.026   -2.078     0.038
Kahraman et al., 2014         -0.957       -0.029   -2.082     0.037
Christakou et al., 2014       -0.896       -0.229   -3.306     0.001
Yildzhan et al., 2015         -0.674       -0.129   -2.890     0.004
Orio et al., 2016             -0.579       0.005    -1.930     0.054
Rondom-effects model          -0.600       -0.311   -6.187     0.000

                       Statistics for each study

Study Name                Total   Relative
DRSP                               weight

Tfalyli et al., 2011       20       7.86
Harmanci et al., 2013      23      11.28
Kahraman et al., 2014      20       9.66
Christakou et al., 2014    40      18.70
Yildzhan et al., 2015      56      24.44
Orio et al., 2016          47      28.06
Rondom-effects model

Heterogeneity: [Tau].sup.2] = 0.000. Q = 4.32.
[I.sup.2] = 0.00%. p < 0.000

(b)

                           Statistics for each study

Study Name             Std diff in   Standard      Variance
DSG                      means       error

Hoeger et al., 2008     -0.672        0.334         0.111
Kilic et al., 2011      -0.682        0.206         0.043
Rondom-effects model    -0.679        0.175         0.031

                              Statistics for each study

Study Name              Lower limit     Upper    Z-value   p-value
DSG                                     limit

Hoeger et al., 2008        -1.326       -0.018   -2.013     0.044
Kilic et al., 2011         -1.086       -0.278   -3.309     0.001
Rondom-effects model       -1.023       -0.336   -3.873     0.000

                       Statistics for each study

Study Name             Total   Relative
DSG                             weight

Hoeger et al., 2008     11      27.61
Kilic et al., 2011      29      72.39
Rondom-effects model

Heterogeneity: [Tau.sup.2] = 0.000. Q = 0.00.
[I.sup.2] = 0.00 %. P = 0.979

                              Statistics for each study

Study Name               Std diff in   Standard      Variance
CMA                        means       error

Vieira et al., 2012       -0.455       0.229         0.053
Vildzhan et al., 2015     -0.427       0.148         0.022
Rondom-effects model      -0.435       0.124         0.015

                              Statistics for each study

Study Name                Lower limit     Upper    Z-value   p-value
CMA                                       limit

Vieira et al., 2012          -0.905       -0.006   -1.986     0.047
Vildzhan et al., 2015        -0.717       -0.138   -2.892     0.004
Rondom-effects model         -0.679       -0.192   -3.506     0.000

                     Statistics for each study

Study Name               Total    Relative
CMA                                weight

Vieira et al., 2012       21       29.34
Vildzhan et al., 2015     50       70.66
Rondom-effects model

Heterogeneity: [Tau].sup.2] = 0.000. Q = 0.00.
[I.sup.2] = 0.00 %. p = 0.918

Figure 3: Forest plot of the standardized difference means of
homocysteine between before and after the use of combined oral
contraceptive in PCOS patients.

                             Statistics for each study

Study Name               Std diff in   Standard      Variance
CPA                         means       error

Cagnacci et al., 2006       0.715       0.354         0.128
Kahraman et al., 2014      -0.508       0.244         0.059
Gul et al., 2008            0.503       0.194         0.038
Rondom-effects model        0.219       0.377         0.142

                               Statistics for each study

Study Name                Lower limit     Upper    Z-value   p-value
CPA                                       limit

Cagnacci et al., 2006        0.021        1.410     2.019     0.044
Kahraman et al., 2014        -0.988       -0.030   -2.083     0.037
Gul et al., 2008             0.123        0.883     2.597     0.009
Rondom-effects model         -0.520       0.959     0.582     0.561

                         Statistics for each study

Study Name               Total    Relative
CPA                               weight

Cagnacci et al., 2006     10      29.56
Kahraman et al., 2014     19      34.27
Gul et al., 2008          30      36.17
Rondom-effects model

Heterogeneity: [Tau.sup.2] = 0.356. Q = 12.99.
[I.sup.2] = 84.60%. P=0.002

                             Statistics for each study

Study Name               Std diff in   Standard      Variance
CPA                         means       error

Harmanci et al., 2013      -0.588       0.226         0.051
Kahraman et al., 2014       0.493       0.237         0.056
Mancini et al., 2010        0.822       0.289         0.084
Mancini et al., 2010       -1.009       0.355         0.126
Rondom-effects model       -0.057       0.407         0.166

                                 Statistics for each study

Study Name                Lower limit     Upper    Z-value   p-value
CPA                                       limit

Harmanci et al., 2013        -1.030       -0.145   -2.603     0.009
Kahraman et al., 2014        0.029        0.957     2.082     0.037
Mancini et al., 2010         0.255        1.388     2.841     0.004
Mancini et al., 2010         -1.705       -0314    -2.848     0.004
Rondom-effects model         -0.855       0.740    -0.141     0.888

                         Statistics for each study

Study Name               Total    Relative
CPA                                weight

Harmanci et al., 2013     23       26.06
Kahraman et al., 2014     20       25.85
Mancini et al., 2010      16       24.78
Mancini et al., 2010      12
Rondom-effects model

Heterogeneity: [Tau.sup.2] = 0.585. Q = 27.26.
[I.sup.2] = 88.99%. P = 0.001

                            Statistics for each study

Study Name               Std diff in   Standard   Variance
CPA                      means         error

Cagnacci et al., 2006       0.680       0.351         0.123
Kilic et al., 2011          0.749       0.226         0.051
Kilic et al., 2011         -0.333       0.210         0.044
Rondom-effects model        0.346       0.391         0.152

                               Statistics for each study

Study Name               Lower limit      Upper    Z-value   p-value
CPA                                       limit

Cagnacci et al., 2006        -0.008       1.368     1.938     0.053
Kilic et al., 2011           0.305        1.193     3.310     0.001
Kilic et al., 2011           -0.744       0.078    -1.590     0.112
Rondom-effects model         -0.420       1.111     0.885     0.376

                         Statistics for each study

Study Name               Total    Relative
CPA                                weight

Cagnacci et al., 2006     10       29.87
Kilic et al., 2011        25       34.27
Kilic et al., 2011        24       35.35
Rondom-effects model

Heterogeneity: [Tau.sup.2] = 0 J87. Q = 14.10.
[I.sup.2] -12.81%. p = 0.001

Figure 4: Forest plot of the standardized difference means of
follistatin before and after the use of combined oral contraceptive
in PCOS patients.

                             Statistics for each study

Study Name             Std diff in   Standard      Variance
CPA                      means       error

Chen et al., 2010        0.465       0.141         0.020
Ibanez et al., 2011     -0.974       0.294         0.087
Rondom-effects model    -0.653       0.246         0.061

                                  Statistics for each study

Study Name              Lower limit     Upper    Z-value   p-value
CPA                                     limit

Chen et al., 2010          -0.740       0.189    -3.303     0.001
Ibanez et al., 2011        -1.551       -0.397   -3.307     0.001
Rondom-effects model       -1.136       -0.171   -2.656     0.008

                       Statistics for each study

Study Name             Total      Relative
CPA                                weight

Chen et al., 2010       56         62.91
Ibanez et al., 2011     17         37.09
Rondom-effects model

Heterogeneity: [Tau.sup.2] = 0.076. Q = 2.43.
[I.sup.2] = 58.93%. P = 0.119
COPYRIGHT 2018 COPYRIGHT 2010 SAGE-Hindawi Access to Research
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:de Medeiros, Sebastiao Freitas; de Medeiros, Matheus Antonio Souto; Santos, Nayara de Souza; Barbosa
Publication:International Journal of Inflammation
Date:Jan 1, 2018
Words:10763
Previous Article:S-Equol, a Major Isoflavone from Soybean, Inhibits Nitric Oxide Production in Lipopolysaccharide-Stimulated Rat Astrocytes Partially via the...
Next Article:The Effect of Submucosal Injection of Plasma-Rich Platelets on Blood Inflammatory Markers for Patients with Bimaxillary Protrusion Undergoing...
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |