Correlation Between Glucose Metabolism and Serum Steroid Hormones in Patients With Polycystic Ovary Syndrome

Xuelin Li; Tianyue Zhang; Shengxian Li; Yuying Deng; Lihua Wang; Tao Tao; Shujie Wang; Yanyun Gu; Weiqiong Gu; Jie Hong; Wei Liu; Weiqing Wang; Yifei Zhang


Clin Endocrinol. 2020;92(4):350-357. 

In This Article


This study explored the association between glucose metabolism and serum steroid hormone profiles in a large PCOS cohort in China, in which 371 (40.6%) of the 914 patients with PCOS had prediabetes and T2DM, showing a relatively high prevalence of abnormal glucose metabolism in this young PCOS cohort. In fact, the prevalence of prediabetes and diabetes in Chinese adults under 40 years old was 28.8% and 5.9%, respectively.[18] Considering that premenopausal women have a lower risk of dysglycaemia than men of the same age, this PCOS cohort exhibited a relatively higher prevalence of prediabetes and T2DM than Chinese women of a similar age. We focused on serum steroid hormone profiling in these patients using LC-MS/MS and found that aldosterone, androstenedione, oestrone, pregnenolone and FAI showed a positive correlation with dysglycaemia, while progesterone showed a negative correlation. After adjusting for age, BMI and fasting insulin levels in the logistic regression model, aldosterone, androstenedione and oestrone (in quartiles) remained significantly correlated with the risk of abnormal glucose metabolism, suggesting that in addition to insulin resistance and obesity, there are likely other underlying mechanisms linking abnormal glucose metabolism to serum steroid hormone levels in patients with PCOS.

Patients with PCOS have increased risks of metabolic disorders such as IGT and T2DM compared with those without PCOS.[6] Obesity and insulin resistance, as two important risk factors for T2DM, frequently occur in patients with PCOS,[19] and the prevalence of T2DM in these patients increases with increasing BMI. However, it remains unclear whether the PCOS status has an independent effect on hyperglycaemia. Hyperandrogenism is considered as a cardinal element in the pathogenesis of PCOS and has a close interaction with hyperinsulinaemia.[20] Previous studies have found that hyperandrogenism can induce insulin resistance in subcutaneous tissue by inhibiting the phosphorylation of protein kinase C,[12] upregulating the uncoupling of protein-1 expression in brown adipose tissue,[11] or promoting lipolysis in visceral fat cells,[21] thus influencing glucose metabolism. Our previous study revealed that patients with PCOS had higher plasma glucose levels than their corresponding control groups.[16] In addition, the levels of several serum steroid hormones, such as DHEAS, 17-hydroprogesterone and oestrone, were different in lean and obese patients, indicating that androgen, androgen metabolites and other steroid hormones may play a role in the occurrence of obesity and abnormal glucose metabolism.

Clinical evidence has indicated that aldosterone exerts a direct effect on insulin action and that the removal of excess aldosterone improves plasma glucose levels.[22] A large-sample multi-ethnic study has also verified that excess aldosterone is related to an increased risk of diabetes and that insulin resistance is involved in the pathophysiology of T2DM.[23] Similarly, O'Reilly et al used LC-MS/MS to measure serum androgen levels in 86 British women with PCOS and found a correlation between high androstenedione levels and insulin resistance as well as dysglycaemia.[24] However, large-scale studies assessing the effect of serum steroid hormones on glucose metabolism are lacking, and the underlying mechanisms remain unclear. We performed a relatively large-sample study using this method and found that aldosterone and androstenedione were correlated with glucose metabolism, independent of age, BMI and fasting insulin levels. The trend for aldosterone remained significant also when we excluded subjects with hypertension. These results indicate that there may be additional mechanisms independent of insulin resistance that mediates the relationship between aldosterone or androstenedione and glucose metabolism in PCOS.

Oestrogens participate in the metabolism of different tissues and produce many positive metabolic benefits.[25] However, several large-sample studies in men have indicated that oestrogens such as oestradiol and oestrone have a positive correlation with diabetes risk.[26,27] Currently, the relationship between oestrogens and diabetes in women with PCOS is not fully understood. In our study, we found that after adjusting for confounding factors, oestrone was correlated with the risk of abnormal glucose metabolism in women with PCOS. However, the correlation was only found with quartile 3 of oestrone, whereas no significant correlation was found with quartiles 2 and 4 when using quartile 1 as a reference. Overall, different levels of oestrone may have different effects on glucose metabolism. The odds ratio for quartile 2 of oestrone and abnormal glucose metabolism was less than 1, suggesting that compared with quartile 3 and quartile 4 of oestrone, a relatively lower oestrone level in patients with PCOS may have a protective impact on glucose metabolism (though no statistical significance was found). The protective roles of oestrone are consistent with several clinical applications. For example, oestrogens therapy reduces diabetes risk in postmenopausal women.[28] However, since a relatively higher conversion of androgen into oestrone in the peripheral tissue occurs in PCOS patients with hyperandrogenaemia,[29] a positive correlation may be found between oestrone (in higher quartiles) and hyperglycaemia. Whether it is possible to link oestrone to the prevalence of abnormal glucose metabolism needs further verification.

Although numerous studies have found that testosterone is closely associated with insulin resistance and could further lead to diabetes, no obvious correlations were found between testosterone and FPG, 2hPG or HbA1c in the present study, which was consistent with a Nordic multicentre cross-sectional study.[30] However, the effects of other confounding factors could not be excluded, such as physical activity, smoking, alcohol consumption and family history of diabetes,[31] which may cause differences in the results of association studies.

There are some limitations in our study that need attention. Firstly, we did not collect information about some confounding factors, such as socioeconomic status, lifestyle and family history of T2DM, which may be involved in the development of abnormal glucose metabolism. Secondly, 11-oxygenated androgens, which are closely correlated with markers of metabolic risk, were not measured. Therefore, we should be cautious in interpreting the current findings. Nonetheless, we performed a two-centre study based on a relatively large sample of Chinese patients with PCOS and measured serum steroid hormone profiles using LC-MS/MS, which is featured by high sensitivity and specificity, contributing to the reliability of the study results. However, it should be noted that the measurement of low serum aldosterone concentrations by LC-MS/MS remains a challenge, despite being a relatively more accurate and reliable method when compared with immunoassays.[32–34] Although no clear causality between steroid hormones and abnormal glucose metabolism was proven in our study, increasing evidences indicate that abnormal patterns of steroid hormones, such as hyperandrogenaemia and hyperaldosteronaemia in women with PCOS, contribute to hyperglycaemia and may increase the risk of prediabetes and T2DM.[20,23]

In conclusion, our study indicated that the PCOS population is more susceptible to developing prediabetes and diabetes. By using LC-MS/MS to measure serum steroid hormone profiles, we found that serum aldosterone, androstenedione and oestrone were positively correlated with abnormal glucose metabolism after adjusting for multiple confounding factors. These findings help further elucidate the roles of steroid hormones in abnormal glucose metabolism in PCOS and confirm the need to screen women with PCOS for dysglycaemia. However, further mechanism and intervention studies are needed to confirm the above results.