Influence of Insulin on LH, Testosterone and SHBG in various PCOS Categories based on the Mode of Secretion of LH in relation to FSH Levels

Abstract

Context

Polycystic ovary syndrome (PCOS) or disease (PCOD) is one of the most common causes of female infertility.

Objective

The objective of this study was to find out the influence of insulin on LH, testosterone and SHBG in diffrent PCOS categories.

Experimental design

A total of 800 women who were subjected to infertility treatment at infertility clinics were selected. About 60 healthy females with regular menstrual cycles were considered as control. The data were collected from hospital records using subject’s consent.

Results

Relationship of insulin to LH and testosterone was positive and significant (p<0.05) in the entire PCOS group and in five PCOS subcategories with increased LH rise (i.e. 1.3, 2, 3, 4 & 5 times of LH rise in relation to FSH levels in each group respectively). The correlation pattern showed an increasing trend from lower to increased rise of LH compared to FSH. The relationship between insulin and SHBG was negative and significant (p<0.05) in all PCOS subcategories, except for the group having similar LH and FSH levels and also in another group with FSH levels higher than LH levels. A strong positive correlation was established between insulin and SHBG in normal subjects. The percentage of negative correlation was strong in PCOD subcategories with elevated rises of LH.

Conclusion

This study established the influence of insulin on other marker hormones (LH, testosterone an SHBG) in various PCOS categories in view of their percentage of relationship.

Introduction

Polycystic ovary syndrome (PCOS) or polycystic ovarian disease (PCOD) is the commonest cause of female infertility affecting 5–10% of reproductive aged females. This syndrome showed symptoms of hyperandrogenism, oligomenorrhea, and polycystic ovarian morphology on ultrasound (1, 2). Moreover, it also showed metabolic abnormalities including insulin insensitivity, obesity, hyperinsulinemia and cardiac problems.

All severe insulin-insensitive situations are having connections to PCOS. The resultant hyperinsulinemia of insulin resistance is strongly correlated with the anovulation of PCOS. Subjects having normal ovulation and also diagnosed with PCOS are not much resistant to insulin than others with anovulation and PCOS (3). Hyperinsulinemia regulates hyperandrogenemia causing theca cells to release androgens and triggering increased LH effect on androgen production in the ovaries. The decline in the production of SHBG by the combined action of androgens and insulin enhances free testosterone and causing also clinical androgen further reduced (2). Considering insulin levels in PCOS patients, numerous papers have assessed the role of β-cell in PCOS, yet the outcomes are ambiguous. Researchers established that response of insulin is increased in PCOS cases, probably as a reimbursement for the peripheral insulin resistance (4). In the current study, an attempt has been made to evaluate the influence of insulin on LH, testosterone and SHBG of normal and different PCOS cases categorised on the basis of mode of secretion of LH in relation to FSH levels.

Methodology

This cross-sectional study was carried out in hospitals (specialized for infertility and assisted reproduction) in selected districts of Kerala, India. About 800 infertile women (age ranges from 25 to 32) were chosen based on the viewpoint of widely accepted Rotterdam criteria for PCOS identification. The conditions for PCOS comprise either abnormal periods together with hyperandrogenemia, or hyperandrogenemia with polycystic ovaries. Moreover, this criterion also included the condition of abnormal periods together with polycystic ovaries on ultrasound. The total PCOS subjects were grouped into seven subcategories in view of the pattern of LH rise in relation to FSH levels (i.e. how may times LH level is greater than FSH level). The normal ovulatory women (Category I) comprised 60 subjects (age ranges from 25 to 32). The various subcategories of PCOS cases were PCOS having same LH and FSH levels (Category II), PCOS having FSH higher than LH levels (Category III), PCOS having LH levels 1.3 times higher than FSH levels (Category IV), PCOS having LH levels 2 times higher than FSH levels (Category V), PCOS having LH levels 3 times higher than FSH levels (Category VI), PCOS having LH levels 4 times higher than FSH levels (Category VII) and PCOS having LH levels 5 times higher than FSH levels (Category VIII). Selected normal subjects had regular menses, without polycystic ovaries on ultrasound. All the women who joined for this research work provided permission for taking their data from the hospital authorities. The data regarding endocrine parameters of the PCOS subjects and normal cases were collected. All the aspects of this research work were approved by members of Institutional Review Board and normal precautions were applied for the human subjects’ data. Our work obeys the principles for improving the value and openness of human health related research.

Statistics

The data analysis and interpretation were done using Microsoft Excel and SPSS 16.0 for Windows. The degree of relationship between endocrine parameters (Insulin and LH; insulin and testosterone; insulin and SHBG) in the control and PCOS subcategories (differentiated based on LH rise in relation to FSH levels) was calculated using Pearson correlation and its significance was noted. The significant difference between these estimated correlation coefficients (normal cases vs. each PCOS subcategory) in three instances of relationship was calculated using Fisher’s Z transformation test. The calculated test statistic (z) values depict the influence of insulin on LH, testosterone and SHBG of various PCOS subcategories and normal subjects. The statistical significance of data was expressed at p<0.05.

The relationship between insulin and LH was not significant in the normal subjects (Category I) and in two PCOS subcategories, namely PCOS having the same LH and FSH levels (Category II) and PCOS having FSH levels higher than LH levels (Category III). Conversely a significant (p<0.05) positive correlation was observed between insulin and LH in PCOS having LH levels 1.3 times higher than FSH levels (Category IV), PCOS having LH levels 2 times higher than FSH levels (Category V), PCOS having LH levels 3 times higher than FSH levels (Category VI), PCOS having LH levels 4 times higher than FSH levels (Category VII), PCOS having LH levels 5 times higher than FSH levels (Category VIII) and in the whole PCOS group. Moreover, the degree of positive relationship between insulin and LH showed a gradation pattern from category IV to category VIII (0.547, 0.594, 0.602, 0.626 and 0.695) of these different PCOS categories with corresponding rise in LH levels in relation to FSH levels (Table 1). Similarly, the relationship of insulin to testosterone and SHBG was not significant in the normal subjects (Category I) and in the PCOS subcategories including PCOS having same LH and FSH levels (Category II) and in PCOS having FSH levels higher than LH levels (Category III) and was significant in the remaining PCOS subcategories and in the whole PCOS group (Table1). Also, the correlation between insulin and testosterone was positive and showed a similar increasing pattern (0.664, 0.767, 0.805, 0.808 and 0.830) from category IV to category VIII as in the case of the correlation between insulin and LH. However, an inverse relationship was observed between insulin and SHBG in various PCOS subcategories except for PCOS having FSH levels higher than LH levels (Category III). The degree of inverse relationship also exhibited a progression (-0.658,-0.747, and -0.830) from category VI to category VIII (Table 1). The confidence interval for each correlation in each group was also calculated.

Table 1.

Relationship of insulin to LH, testosterone and SHBG in the control and different PCOS categories

Groups	Relationship between insulin and LH [Correlation coefficient (r)]	95% C.I. for ‘r’ between insulin and LH		Relationship between insulin and testosterone [Correlation coefficient (r)]	95% C.I. for ‘r’ between insulin and testosterone		Relationship between insulin and SHBG [Correlation coefficient (r)]	95% C.I. for ‘r’ between insulin and SHBG
		Lower C.I.	Upper C.I.		Lower C.I.	Upper C.I.		Lower C.I.	Upper C.I.
Normal subjects (Category I)	.059 (n=50)	-0.223	0.331	-.146 (n=38)	-0.444	0.182	.827*(n=18)	0.586	0.933
PCOS having same LH and FSH levels (Category II)	-.066 (n=14)	-0.576	0.481	-.218 (n=12)	-0.703	0.406	-.437 (n=10)	-0.836	0.265
PCOS having FSH levels higher than LH levels (Category III)	-.043 (n=53)	-0.309	0.229	-.337 (n=13)	-0.748	0.262	.183 (n=13)	-0.409	0.666
PCOS having LH levels 1.3 times higher than FSH levels (Category IV)	.547* (n=50)	0.316	0.716	.664*(n=10)	0.059	0.912	-.553*(n=14)	-0.837	-0.031
PCOS having LH levels 2 times higher than FSH levels (Category V)	.594* (n=64)	0.407	0.732	.767*(n=20)	0.491	0.903	-.529*(n=20)	-0.787	-0.112
PCOS having LH levels 3 times higher than FSH levels (Category VI)	.602* (n=24)	0.262	0.808	.805*(n=15)	0.498	0.932	-.658*(n=13)	-0.887	-0.167
PCOS having LH levels 4 times higher than FSH levels (Category VII)	.626* (n=14)	0.142	0.868	.808*(n=12)	0.436	0.944	-.747*(n=8)	-0.951	-0.089
PCOS having LH levels 5 times higher than FSH levels (Category VIII)	.695*(n=11)	0.163	0.913	.830*(n=11)	0.458	0.954	-.830*(n=8)	-0.968	-0.301
Whole PCOS	.481*(n=230)	0.374	0.574	.668* (n=93)	0.537	0.767	-.444*(n=86)	-0.599	-0.256

*significant (p<0.05).

The influence of insulin on LH and testosterone was significantly (p<0.05) higher in PCOS subcategories and whole PCOS group (Table 2) when compared to the control. However, the effect of insulin on SHBG was significant (p<0.05) in all PCO subcategories except for category II comprising PCOS having the same LH and FSH levels, as in the case of normal subjects. The effect was also significant (p<0.05) in the whole PCOS group compared to control.

Table 2.

Comparison of significance of difference between control and various PCOS categories with respect to correlation coefficients (regarding insulin to LH, testosterone and SHBG) using Fisher’s Z transformation test

Groups	‘r’ between insulin and LH [Test statistic (z)]	‘r’ between insulin and testosterone [Test statistic (z)]	‘r’ between insulin and SHBG [Test statistic (z)]
r of normal subjects (Category I) vs. r of PCOS having same LH and FSH levels (Category II)	0.37	0.2	1.55
r of normal subjects (Category I) vs. PCOS having FSH levels higher than LH levels (Category III)	0.5	0.54	2.43*
r of normal subjects (Category I) vs. r of PCOS having LH levels 1.3 times higher than FSH levels (Category IV)	-2.69*	-2.29*	4.54*
r of normal subjects (Category I) vs. r of PCOS having LH levels 2 times higher than FSH levels (Category V)	-3.22*	-3.92*	4.99*
r of normal subjects (Category I) vs. r of PCOS having LH levels 3 times higher than FSH levels (Category VI)	-2.43*	-3.77*	4.79*
r of normal subjects (Category I) vs. r of PCOS having LH levels 4 times higher than FSH levels (Category VII)	-2.02*	-3.39*	4.15*
r of normal subjects (Category I) vs. r of PCOS having LH levels 5 times higher than FSH levels (Category VIII)	-2.09*	-3.41*	4.58*
r of normal subjects vs. r of whole PCOS	-2.9*	-4.78*	5.9*

*significant (p<0.05).

Discussion

This study was carried out to evaluate whether the changes observed in LH, testosterone and SHBG are dependent on insulin or not in various PCOS subcategories, differentiated based on the pattern of LH rise in relation to FSH levels, which were not reported in previous works. We hypothesized that elevated insulin levels exert their effect on other hormones to develop hormonal imbalances in PCOS. In our study, insulin influences LH and testosterone levels which are markedly higher in these PCOS subcategories with varied multiples of LH rise with regard to FSH levels. The levels of LH together with testosterone increase with an increase in the amount of insulin in these PCOS cases. However, these variations were not detected in normal subjects. Moreover, the degree of positive relationship of insulin to LH and testosterone showed a gradation from category IV to category VIII of these PCO subgroups, i.e. this study showed that the higher the LH levels, the higher would be the degree of relationship. Also the degree of relationship was feeble in PCOS subcategories, namely PCOS having same LH and FSH levels (Category II) and in PCOS having FSH levels higher than LH levels (Category III). PCOS subjects, having lower rise of LH in relation to FSH levels, revealed poor association between insulin and LH levels and also between insulin and testosterone levels. The amount of relationship is associated with the increase in the level of LH rise in relation to FSH levels. The PCOS subgroups were categorized on that basis of increasing pattern of LH levels with respect to FSH levels. When considering PCOS as a heterogeneous group, the degree of relationship was not much increased in comparison to PCOS subcategories. Considering PCOS as a whole group, the rising pattern of LH is not very evident as in PCOS subcategories, where subjects having identical and elevated LH levels are assembled under separate groups in accordance with the LH level rise.

Insulin and its related growth factors could enhance the activities of gonadotropin hormones, i.e. follicle stimulating hormone (FSH) and luteinizing hormone (LH) in triggering luteal cell formation and stimulating steroid synthesis. Insulin has a role in the augmented level and regularity in the GnRH and LH release observed in PCOS. Moreover, in vitro studies have validated the elevation in these hormonal levels as a consequence of insulin infusion techniques (5). Accordingly, elevated GnRH synthesis and secretion bring about later increase in levels of LH and also the enhanced production of testosterone (6). Likewise, insulin enhances testosterone levels of PCOS subjects in this study. Thus, an increase in the level of testosterone is associated with elevated insulin levels and consequent increase in LH levels, established in this study. In this case also, the degree of relationship between insulin and testosterone was higher in groups consisting of PCOS with elevated LH rise in relation to FSH levels, than in the whole PCOD group. In spite of the classic insulin resistance of PCOS, earlier research works had observed the increased sensitivity of theca cells to insulin and caused increased androgen synthesis as an outcome of it. Moreover, they established that insulin only is unable to promote androgen synthesis (7). Researchers reported that cell culture experiments carried out in theca cells revealed elevation in microsomal enzyme expression after insulin and LH stimulation (8), while other researchers were not successful in substantiating these effects (9).

Researchers (10) reported a steady increase in the mean values of insulin (29.58μIU/mL, 34.89 μIU/mL, 39.38μIU/mL, 41.11μIU/mL and 45.90μIU/mL) and testosterone (1.38 nmol/L, 1.98 nmol/L, 2.03 nmol/L, 2.06 nmol/L and 2.44 nmol/L) respectively in PCOS subgroups and this increasing trend was in accordance with the corresponding increase in the range of LH :FSH ratios (1.1- 1.5, 1.6-2.5, 2.6-3.5, 3.6-4.5, 4.6 -5.5) in these groups. This study has prompted our conclusion that insulin definitely played a putative role in generating gonadotropin dysfunction and steroid hormone imbalance in PCOS subjects. Insulin appears to play in collaboration with LH to raise intracellular concentration of cAMP, which triggers steroidogenic acute regulatory (StAR) protein, stimulating steroidogenic action. Though this influence may be straight via the phosphatidylinositol 3-kinases (PI3K) pathway, the demand of cyclic AMP in this process implies deviation from the normal insulin flow; however the discrepancies in actions at molecular level are unidentified (11). Moreover, insulin could also increase the synthesis of steroid hormones by aromatase up regulation in granulosa cells, which may function as substrates for theca cells so as to convert them into androgens (12).

Consistent with earlier findings, our study also observed a negative correlation between insulin and SHBG levels (13). Also the degree of negative relationship was very higher in PCOS subcategories consisting of subjects with the same and elevated LH levels in relation to FSH levels, than the entire PCOS category with various LH levels with respect to FSH levels. However, a moderate inverse relationship (-0.437) was observed in PCOS subjects having the same amount of FSH and LH levels, and a very lower positive relationship was established in PCOS subcategory with FSH levels higher than LH levels. A strong positive relationship between insulin and SHBG was detected in normal subjects.

The decrease in the level of SHBG can be attributed to increased insulin levels and the resultant elevation in the levels of LH and testosterone. Insulin seems to suppress the levels of sex hormone binding globulin (SHBG) in PCOS cases. Lower SHBG concentration was observed in PCOS with elevated insulin, in comparison to the normal subjects. Increased insulin levels were correlated with reduced SHBG levels, directing into augmented occurrence of androgens (14). However, research works did not reveal any direct inhibition of SHBG by insulin or growth factors of insulin. Moreover, they act indirectly and diminished the synthesis of protein in liver cells. Thus, increased levels of monosaccharides like glucose and fructose caused a decline in human SHBG production via a down regulation of activity of hepatocyte protein. However, considering insulin resistance, it is associated with elevated concentrations of these monosaccharides as a result of impaired insulin signalling and without insulin action by itself (15,16). Instead, insulin inhibited the synthesis of its growth factor binding protein in the liver and ovaries and thereby allowing increased availability of its growth factor. In the liver, the growth factor enhances the actions of insulin and thus stimulates the suppression of SHBG. In the ovaries, growth factors intensify the PCOS conditions and hormonal imbalances (17).

Most of the research work carried out in this field in recent times also highlighted the prime importance of insulin in the development of PCOS and our study also reinforced the role of insulin. Regarding the importance of insulin resistance and the resultant hyperinsulinemia, various dieting routines could decrease growth factors and receptors specific to insulin and thus have positive results in improving hormonal variations and irregularities in ovulation (18). Research works based on insulin sensitizer drugs like D-Dhiro-inositol (DCI) and myo-inositol [MI] received more attention due to their effectiveness in improving ovarian function and metabolism in subjects with PCOS. In view of accumulating evidence based on insulin sensitizer drugs, the present study again confirmed the influence of elevated insulin and loss of insulin sensitivity in stimulating ovulation problems and thereby the development of PCOS (19-21).

The present study concluded that insulin may influence LH, testosterone and SHBG levels in PCOS subjects of this study. Also the changes observed in LH, testosterone and SHBG levels were associated with changes in insulin levels. The influence of insulin on these hormones in terms of comparison of degree of relationship was very obvious and the amount of association was higher in PCOS subcategories having an identical and increased rise of LH levels compared to the heterogeneous whole PCOS group with varied LH levels. Moreover, this can be inferred that the altered rise of LH levels in relation to FSH levels could be considered as an investigative factor to find out PCOS in a preliminary stage, since in the control subjects, LH and FSH levels are more or less equal and alterations were not found. Categorization of PCOS on the basis of pattern of LH rise, not carried out in previous works, and the determination of a relationship between insulin and other endocrine parameters in these subcategories facilitated an enhanced approach for evaluating the effect of insulin on certain hormones in generating PCOS in women. In earlier studies, researchers were considering PCOS as a whole group and it is very difficult to predict how much the amount of LH in relation to FSH is surpassed in PCOS cases. This work has certain limitations, as the interpretations were made only on the basis of the blood test report of endocrine profile of the selected subjects. Further in vitro studies based on cellular and molecular level mechanisms are required to explain the underlying principle of hormonal interaction in PCOS.

Conflict of interest

The authors declare that they have no conflict of interest.