Assessing the variability and predictability of adipokines (adiponectin, leptin and resistin and their ratios) in non-obese and obese women with polycystic ovary syndrome


 Objectives

To assess the variability in adiponectin, leptin, and resistin between ovulatory women, and non-obese and obese women with polycystic ovary syndrome (PCOS). The study also explores the ratios of the adipokines and evaluated their predictability for PCOS.

Results

The PCOS group presented with lower adiponectin [13.0(10.49-16.59) vs 18.42(15.72-19.92) µg/ml, p<0.0001], adiponectin: leptin ratio [0.60(0.35-0.88) vs 1.19(0.92-1.37), p<0.0001], and adiponectin: resistin ratio [0.30(0.21-0.43) vs 0.42(0.32-0.62), p<0.0001] but a higher leptin [20.02(14.54-26.80) vs 16.17(14.51-18.36) ng/ml, p<0.0001] and leptin: resistin ratio [0.53(0.37-0.82) vs 0.40(0.27-0.48), p<0.0001] compared to the ovulatory group. The obese PCOS group presented with lower adiponectin [11.04(5.66-13.25) vs 14.18(11.04-18.02) µg/ml, p<0.0001 and 18.42(15.72-19.92) µg/ml, p<0.0001], adiponectin: leptin ratio [0.36(0.27-0.44) vs 0.78(0.61-1.16), p<0.0001 and 1.19(0.92-1.37), p<0.0001], and adiponectin: resistin ratio [0.24(0.17-0.38) vs 0.40(0.23-0.58), p<0.0001 and 0.42(0.32-0.62), p<0.0001] but a higher leptin [26.80(14.28-32.09) vs 17.95(14.86-21.26) ng/ml, p<0.05 and 16.17(14.51-18.36) ng/ml, p<0.0001] and leptin: resistin ratio [0.63(0.46-1.03) vs 0.41(0.30-0.61), p<0.0001 and 0.40(0.27-0.48), p<0.0001] compared to the non-obese PCOS and ovulatory control group, respectively. Adiponectin: leptin ratio presented with the best discriminatory power in predicting PCOS (AUC=0.83) followed by adiponectin alone (AUC=0.79), and leptin: resistin ratio and leptin alone (both AUC=0.69). Resistin alone presented with the poorest discriminatory power (AUC=0.48).


Introduction
Overweight and obesity are pervasive medical conditions which are considered global epidemic and threat to public health [1,2]. A mountain of evidence suggest that obesity is associated with the risk of metabolic diseases such as diabetes mellitus, hypertension, cardiovascular disease, obstructive sleep apnea, various types of cancer and overall mortality [3][4][5]. There have also been reports of associations between obesity and infertility [6,7], particularly among women due to the risk of anovulation [1,8].
Anovulation is a common cause of infertility in women; responsible for not less than 25% fertility problems among couples [9,10] and 25-50% of female infertility, of which polycystic ovary syndrome (PCOS) accounts over 90% of cases [11].
PCOS is a multisystem, endocrinological, reproductive and metabolic disorder characterized by oligo-and/or anovulation, hyperandrogenism, and polycystic ovaries [12]. Obesity-related adverse alterations in adipose tissue that predispose to metabolic dysregulation has been implicated in PCOS pathogenesis. These adverse alterations include derangements in bioactive cytokines and adipokines such as adiponectin, leptin, and resistin [13,14].
Thus far, reports regarding the changes in adiponectin, leptin, and resistin in non-obese and obese women with PCOS remain inconclusive. It is possible that differences in geography and its accompanying diversities in genetic and lifestyle characteristics may be involved in the discrepancies observed by previous studies. Additionally, very limited studies have assessed the usefulness of adipokine ratios in predicting PCOS [17,21].
Sarray et al. [17] recently reported the expediency of adipokine ratio in predicting PCOS.
They however indicated that studies in diverse backgrounds to confirm the utility of adipokine ratios as potential biomarkers of PCOS is warranted. In light of this, and the fact that none of such studies have been conducted to evaluate the validity of the associations in a Ghanaian population, this study aimed at assessing the variability in adiponectin, leptin, and resistin between ovulatory women, non-obese and obese women with PCOS in Ghana. The study also explores the ratios of the adipokines and evaluated their predictability for PCOS.

Study Design/Setting
The study was a case-control study. Consecutive consenting women clinically diagnosed of PCOS visiting the Obstetrics and Gynaecology units of Trust Care, Ruma and Asbury were included in the study as cases. PCOS diagnosis was based on the 2003 Rotterdam criteria, in which PCOS diagnosis is confirmed when two of three conditions are met: oligo-and/or anovulation, hyperandrogenism (clinical and/or biochemical), and polycystic ovaries on ultrasound examination [12]. Fertile (eumenorrheic) women visiting the hospital for routine check-up were included as controls. Relevant clinical data of each participant was extracted from the hospital's archive. Women with Cushing syndrome, hyperprolactinemia, androgen-producing tumors, non-classic adrenal hyperplasia, active thyroid disease, and diabetes were excluded.
Blood sampling, processing and analysis Five milliliters of venous blood was obtained from each participant and dispensed into gel separator tubes. The tubes were centrifuged at 1500 x g for 10 minutes at 4 °C to obtain the serum which were stored at -20 °C until analysis. Serum levels of adiponectin, leptin, and resistin were measured based on solid-phase sandwich Enzyme Linked Immunosorbent Assay (ELISA) technique (standardized with an intra-and inter-assay %CVs <10%) (Green Stone Swiss Co Limited, China) according to the manufacturer's instructions.

Statistical analysis
Statistical analysis and graphical presentation was performed using the R Language for Statistical Computing version 3.6.0 (R Core Team, Vienna, Austria) [25]. Chi squared test was used to assess significance of association between the participant characteristics and fertility status. Distribution of adipokines were presented with density plots. Hierarchical clustering by Spearman's correlation was used to assess relationship between adipokines (and their ratios) with obesity indices. Independent t-test and one-way ANOVA with Tukey test or Mann-Whitney U t-test and Kruskal-Wallis test with Dunn's test were used to test for significance of difference between groups where applicable. The receiver operating characteristics (ROC) curve analysis was used to evaluate the performance of the adipokines (and their ratios) in predicting PCOS. Confidence was set at 95% and a value < 0.05 was considered statistically significant.   There was no statistically significant correlation between leptin and obesity indices.

Discussion
This study reports lower levels of adiponectin, adiponectin: leptin ratio, and adiponectin: shown to decrease in obesity and increase with weight loss [20]. It is considered a 'beneficial' adipokine in reproduction [31]. The levels of leptin have been reported to constitutively secreted by adipocytes in proportion to the adipose mass [32]. In obese patients, the levels of leptin are even more elevated due to leptin resistance [33].
Additionally, increased expression of the resistin gene has been observed in human preadipocytes, which decreased during adipocyte differentiation. The relationship between resistin and obesity is however convoluted. Whereas some report a direct association [34,35], others show no significant association with obesity [36-38]. Consistent with these previous reports, we also found that, with the exception of resistin, all other adipokines including their ratios were strongly and more correlated with various obesity indices among women with PCOS compared to the controls. Specifically, adiponectin, adiponectin: leptin ratio, and adiponectin: resistin ratio showed a positive association whereas leptin, and leptin: resistin ratio correlated positively with the obesity indices. This finding is also coherent with studies by Sarray et al. [17] and Golbahar et al. [21].
In order to assess the predictive capabilities of the adipokines and their ratios, we employed the ROC curve analysis with reference to PCOS. We found adiponectin: leptin ratio to have the best discriminatory power in predicting PCOS with an AUC of 0.83.
Golbahar et al. also found adiponectin: leptin ratio to have a similarly high discriminatory power with comparable AUC of 0.86 among Bahraini women with PCOS [21]. A much higher discriminatory power (AUC of 0.94) for adiponectin: leptin ratio in predicting PCOS has been reported by Sarray et al. also among women with PCOS in Bahrain [17]. The discrepancies in the predictive power may be attributed to differences in characteristics of the study population, sample size, and methods for biochemical analysis.

Conclusion
This study shows significantly altered serum levels adiponectin and leptin but not resistin in Ghanaian women with PCOS compared to healthy subjects. Obese PCOS patients have the most altered levels of adipokines compared to non-obese PCOS and healthy subjects.
Adiponectin: leptin ratio is the best predictor of PCOS compared to individual adipokines.

Limitations
The major limitation of this present study is the relatively small sample size. We recommend the use of larger sample size in future studies.

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.
Additional file 1.pdf