The follicular levels of adipokines and their ratio as the prognostic markers of in vitro fertilization outcomes

Abstract

Introduction

Adipokines are involved in the regulation of the female reproductive system. The purpose was to study the possibility of using adipokines levels in the follicular fluid to predict IVF efficiency.

Materials and methods

Four groups of women were studied: pregnant during IVF, with normal (PN, n = 9) and increased (BMI > 25 kg/m²) body weight (BW) (PI, n = 7), and nonpregnant during IVF, with normal (nPN, n = 16) and increased BW (nPI, n = 21).

Results

In PN group, leptin level was higher than in nPN group (p < .05). In the PI and nPI groups, it did not differ, but was higher than in women with normal BW. In PN group, ghrelin level was lower than in nPN group (p < .05), while in the PI and nPI groups it was comparable. The leptin/ghrelin ratio in PN group was higher than in nPN group (18.10 ± 3.38 vs. 3.93 ± 0.60, p < .05), but lower than in the PI (31.70 ± 15.38) and nPI (24.30 ± 3.45) groups. The leptin/adiponectin ratio in PN group was also higher than in nPN group (6.97 ± 0.64 vs. 2.95 ± 0.39, p < .05), but lower than in the PI (13.60 ± 1.59) and nPI (10.86 ± 0.87) groups. Adiponectin levels differed only between the nPN and nPI groups. In women with normal BW, odds ratio showed that the leptin/ghrelin ratio has the greatest prognostic value for predicting the success of IVF outcomes (OR: 29.53; CI: 1.53–570.83, p =.025) among other indicators. In women with increased BW, none of the indicators had predictive value.

Conclusion

The follicular leptin/ghrelin ratio is a suitable indicator for predicting IVF outcomes in women with normal BW.

Keywords:

• In vitro fertilization
• leptin
• adiponectin
• ghrelin
• follicular fluid
• pregnancy

Introduction

In many countries, the proportion of infertile couples reaches 30%, as a result of which infertility is one of the most urgent problems of modern medicine [1]. The most commonly used approach to increase fertility in such couples is the use of the in vitro fertilization (IVF). However, at present, application of IVF technology requires improvement, since only 35% of the cycles lead to the birth of a healthy living child [2].

Along with gonadotropins, the most important regulators of reproductive functions in women are adipokines produced by the adipose tissue, such as leptin, adiponectin, and ghrelin [3,4]. Adipokines control the functions of all components of the hypothalamic–pituitary–gonadal (HPG) axis. Penetrating the blood-brain barrier, leptin stimulates the production of gonadotropin-releasing hormone (GnRH) by hypothalamic neurons, which leads to an increase in gonadotropins secretion and an activation of the HPG axis [5]. Adiponectin and ghrelin have the opposite effect on GnRH-expressing neurons [6]. Leptin and adiponectin are also able to regulate functional activity of the ovaries, and in some cases, they stimulate folliculogenesis and ovarian steroidogenesis, and in others, they suppress them, and this depends on metabolic and hormonal status, as well as on metabolic disorders, including obesity and diabetes mellitus [7–13].

In obesity, there is an increase in the plasma levels of leptin and a decrease in the levels of its functional antagonists, adiponectin and ghrelin [14]. There is evidence that the changes in adipokines levels in the blood and follicular fluid (FF) during obesity can cause infertility and lead to a decrease in IVF effectiveness [15,16], along with that correction of the leptin levels in leptin-deficit obese ob/ob mouse restores their fertility while the calories restriction has no effect [17]. It is important that adipokines levels in FF are determined not only by the efficiency of adipokines transport from the blood stream through the blood–ovarian barrier, but also by possibility of de novo synthesis of some adipokines in the ovaries [18,19]. We believe that not only adipokines levels in FF, but also their ratio plays an important role in the functioning of the reproductive system in women with normal and increased body weight (BW) and in the success and effectiveness of IVF. To verify this assumption, we studied the relationship between the success of IVF and the levels and the ratios of leptin, adiponectin and ghrelin in FF of women with normal and increased (BMI > 25 kg/m²) BW.

Materials and methods

The study was approved by local Ethical Committee of D.O. Ott Research Institute of Obstetrics, Gynecology and Reproduction (protocol No 77; May 12, 2016). Fifty-three women participated in the study with following inclusion criteria: (1) age of 18–45 years and (2) infertility due to tubal or male factor. Exclusion criteria were as follows: (1) clinical evidences of polycystic ovarian syndrome, (2) carbohydrate metabolism disorders, (3) blood anti-Müllerian hormone concentration <1.0 ng/mL, (4) contraindications to IVF. According to the body mass index (BMI), women were divided on two groups—with normal and increased BW. The BMI was calculated as weight(kg)/[height(m)]². Normal BW was taken as BMI 18.5–24.9 kg/m², while increased BW as BMI > 25 kg/m². Depending upon pregnancy diagnosis, women in each group were subsequently divided on two subgroups: pregnant and nonpregnant after IVF. All the studied groups were uniform in age. Along with this, the PN (pregnant, normal BW) and nPN (nonpregnant, normal BW) groups and the PI (pregnant, increased BW) and nPI (nonpregnant, increased BW) groups were pairwise uniform in BMI.

IVF was performed using GnRH-antagonist protocol. For controlled ovarian stimulation was used either recombinant follicle-stimulating hormones, such as Gonal F^® (“Merck Serono S.p.A.”, Italy), Puregon^® (‘Organon’, the Netherlands), Follitrop^® (‘LG Chem Ltd.’, Republic of Korea), Pergoveris^® (‘Merck Serono S.A.’, Switzerland), or human menopausal gonadotropins, such as Menopur^® (‘Ferring GmbH’, Germany) and Humog^® (‘Bharat Serums & Vaccines Ltd.’, India). Ovulation induction was performed using recombinant human chorionic gonadotropin (hCG) Pregnyl^® (‘Organon’, the Netherlands) or Ovitrelle^® (‘Merck Serono S.p.A.’, Italy), when minimum of two follicles reached 18 mm. Transvaginal puncture with egg retrieval was made after 34–36 h after hCG injection. The FF from the first punctated follicle was collected for ELISA analysis. After 4–5 days no more than two embryos were transferred, and after 14–15 days post embryo transfer venous blood was collected for β-hCG determination. The β-hCG-positive women were subsequently studied using ultrasonography, and pregnancy was diagnosed on the presence of gestational sac. The leptin, adiponectin and ghrelin levels in FF were measured using ELISA kits, such as the ‘DRG Leptin ELISA’ (‘DRG’, USA), ‘AssayMax Human Adiponectin ELISA’ (‘AssayPro’, USA), ‘Human Ghrelin ELISA’ (‘SCETI’, Japan).

The threshold values for adipokines concentrations and their ratios were determined below which the onset of pregnancy has a low probability. As threshold values, the values with the best sensitivity and specificity were chosen, which were calculated as follows. The sensitivity was calculated by the equation Sn = TP/(TP + FN), where Sn is the sensitivity, TP is the number of truly positive results, FN is the number of false negative results. The specificity was calculated by the equation Sp = TN/(TN + FP), where Sp is specificity, TN is the number of truly negative results, FP is the number of false positive results.

Statistical analysis was performed using SPSS Statistics. Normality was assessed using Kolmogorov–Smirnov test (in all groups p > .05). Investigated groups were compared using analysis of variance (ANOVA), significant differences were taken as p < .05.

Results

Of the 53 women who participated in the study, after IVF clinical pregnancy occurred in 16 patients (30%). The following four groups were formed: women with normal BW who became pregnant (PN, n = 9) and with no pregnancy (nPN, n = 16), and women with increased BW who became pregnant (PI, n = 7) and with no pregnancy (nPI, n = 21).

The leptin level in FF in the PN group was higher than in the nPN group (p < .001) (Table 1). In women with increased BW, both pregnant and without pregnancy, leptin level was higher than in the PN group. At the same time, there were no significant differences between the PI and nPI groups (Table 1). Adiponectin levels differed only between the nPN and nPI groups, while no differences were found between pregnant and nonpregnant women (Table 1). The ghrelin level in the PN group was reduced in comparison with the nPN group. There were no differences in ghrelin level between the groups of women with increased BW (Table 1).

Table 1. The age, body mass index, adipokines levels in the follicular fluid and their ratios in women with the normal and increased body weight (BW) who became pregnant as a result of IVF or with no pregnancy.

Indicator	Women with normal BW		Women with increased BW
	nPN (n = 16)	PN (n = 9)	nPI (n = 21)	PI (n = 7)
Age, years	31.8 ± 0.8	32.1 ± 1.6	34.3 ± 1.1	31.6 ± 1.3
Body mass index, kg/m^2	21.3 ± 0.3	23.0 ± 0.4	29.9 ± 0.8	30.6 ± 0.7
Leptin, ng/mL	10.54 ± 1.16^a,b	21.61 ± 2.55^a,с	32.16 ± 2.47^b	39.62 ± 4.35^с
Ghrelin, ng/mL	3.80 ± 0.78^a,b	1.39 ± 0.19^a	1.93 ± 0.34^b	2.24 ± 0.39
Adiponectin,µg/mL	3.72 ± 0.13^b	3.20 ± 0.30	2.98 ± 0.08^b	2.95 ± 0.17
Leptin/Adiponectin	2.95 ± 0.39^a,b	6.97 ± 0.64^a,с	10.86 ± 0.87^b	13.60 ± 1.59^с
Leptin/Ghrelin	3.93 ± 0.60^a,b	18.10 ± 3.38^a	24.30 ± 3.45^b	31.70 ± 15.38
Adiponectin/ Ghrelin	1.61 ± 0.29	2.64 ± 0.44	2.37 ± 0.31	1.94 ± 0.64

Note. ^athe differences between the nPN and PN groups are statistically significant at p < .05; ^bthe differences between the nPN and nPI groups are statistically significant at p < .05; ^cthe differences between the PN and PI groups are statistically significant at p < .05.

The leptin/adiponectin ratio was increased in both groups of women with increased BW, while in women with normal BW, it was increased only in pregnant women (p < .001) (Table 1). In women with increased BW, the leptin/ghrelin ratio was higher than in women with normal BW. In the PN group, it was higher than in the nPN group. A similar pattern was observed in women with increased BW, but difference between the PI and nPI groups was not significant (Table 1). The adiponectin/ghrelin ratio in all the studied groups did not differ (Table 1).

According to calculated threshold values for adipokines concentrations and their ratios, the most sensitive-specific indicator for prediction of IVF effectiveness in women with normal BW was the leptin/ghrelin ratio. This is indicated by the high value of the odds ratio for the leptin/ghrelin ratio (Table 2). Based on calculated odds ratio, in women with normal BW the follicular leptin level and the leptin/adiponectin ratio can be used to estimate the IVF outcomes, but their prognostic significance is significantly lower (Table 2). At the same time, these indicators for women with increased BW had low sensitivity and specificity. Their odds ratios were not statistical significant (Table 2), so they cannot be used as indicators for IVF effectiveness.

Table 2. The prognostic significance of the leptin level and the ratios of leptin/ghrelin and leptin/adiponectin in the follicular fluid.

	Parameter	The value at which pregnancy becomes unlikely	Sensitivity	Specificity	Odds ratio	CI	p
Normal BW	Leptin, ng/mL	<15	89 %	88 %	7.11	1.23–40.99	.028
	Leptin/Ghrelin	<9	89 %	100 %	29.53	1.53–570.83	.025
	Leptin/Adiponectin	<5	89 %	88 %	7.11	1.23–40.99	.028
Increased BW	Leptin, ng/mL	<35	71 %	57 %	1.67	0.42–6.68	.47
	Leptin/Ghrelin	<20	43 %	57 %	1	0.21–4.77	1
	Leptin/Adiponectin	<12	71 %	71 %	2.5	0.58–10.80	.22

Discussion

Unsuccessful IVFs are usually associated with the following factors, such as poor quality embryos, poor implantation and also the early development of an implanted embryo leading to miscarriage during early pregnancy. In addition to genetic abnormalities causing poor quality embryos [20], the above-mentioned factors are largely dependent on steroidogenic function of the ovaries and hormonal regulation of the follicles. These processes are controlled by gonadotropins, insulin-like growth factor-1 and adipokines, primarily by leptin. It was shown that leptin at the physiological concentration stimulates folliculogenesis, enhances progesterone production, activates aromatase, and increases estradiol production [7–9,21,22]. Leptin deficiency in female mice knocked out by the ob gene encoding leptin causes infertility [23]. To regulate folliculogenesis, leptin level in FF is most important, since only follicular leptin is able to directly affect physiological processes in follicular cells and regulate steroidogenesis in them. We showed that in women with normal BW who became pregnant as a result of IVF, leptin level in FF was significantly higher than in nonpregnant women. This indicates a positive effect of follicular leptin on both ovarian function and IVF effectiveness in women with normal BW.

A prolonged increase in leptin levels and systemic hyperleptinemia in obesity leads to a decrease in basal and insulin-like growth factor-1-induced production of estradiol and progesterone, to an inhibition of gonadotropin-stimulated steroidogenesis and ovulation [10,11,24]. This may be due to the fact that an increased level of follicular leptin causes a decrease in sensitivity of follicular cells to leptin and a weakening of the follicular leptin signaling, similar to how it occurs in the case of a leptin deficiency. At the same time, we showed for the first time that follicular leptin level was similarly increased both in women with increased BW who became pregnant as a result of IVF, and in those who did not become pregnant. Therefore, in women with obesity-associated hyperleptinemia, an assessment of leptin level in FF is not suitable for predicting IVF outcome.

Ghrelin and adiponectin, being functional antagonists of leptin, are also involved in the regulation of women reproduction [6,25]. An increase in ghrelin level leads to decreased leptin sensitivity, which is based on activation of the negative regulators of leptin signaling, including the suppressor-3 of cytokine signaling [26]. Accordingly, a decrease in ghrelin level and an increase in the leptin/ghrelin ratio in FF of women without systemic hyperleptinemia should lead to an improvement in folliculogenesis and fertility, which is confirmed by our data. In FF of pregnant women with normal BW, we showed the decreased ghrelin level and increased leptin/ghrelin ratio as compared with nonpregnant women (Table 1). In the PI and nPI groups, the ghrelin levels and the leptin/ghrelin ratio were comparable, and did not differ from that in pregnant women with normal BW (Table 1). Therefore, in women with normal BW, but not in women with increased BW, the leptin/ghrelin ratio, along with follicular leptin level, is an important prognostic indicator of the success of an IVF outcome. This is illustrated by threshold values for leptin (<15 ng/mL) and leptin/ghrelin ratios (<9), associated with failed IVF attempts (Table 2). The leptin/ghrelin ratio is more sensitive and specific as compared to that of leptin level, which is illustrated by a higher value of the odds ratio for leptin/ghrelin ratio (Table 2).

Adiponectin suppresses steroidogenesis in the ovaries and inhibits secretion of GnRH by hypothalamic neurons [6]. The regulatory effects of adiponectin on ovarian functions depend on metabolic and hormonal status, concentration and pattern of adiponectin isoforms and the ratio of types 1 and 2 adiponectin receptors [4]. We found that nonpregnant women with normal BW had higher follicular adiponectin level than nonpregnant women with increased BW, but this indicates only limited relationship between adiponectin concentration in FF and IVF outcomes. At the same time, an increase in the leptin/adiponectin ratio was observed in pregnant women with normal BW as compared with nonpregnant women, which illustrated by increased odds ratio for the leptin/adiponectin ratio in FF. This indicates the possibility of using the leptin/adiponectin ratio for predicting IVF effectiveness. At the same time, we did not find differences in the leptin/adiponectin ratio between pregnant and nonpregnant women with increased BW, which does not allow us to use this indicator to evaluate IVF outcomes in these groups of patients.

Conclusion

Based on the results of a comprehensive study of the levels and ratios of leptin, ghrelin and adiponectin, we proposed to use the leptin/ghrelin ratio in FF and, although a lower significance, the follicular leptin level and the leptin/adiponectin ratio to predict IVF outcomes in women with normal BW. However, these indicators cannot be used to predict IVF effectiveness in women with increased BW who have systemic hyperleptinemia and, therefore, have high leptin levels in FF and increased leptin/ghrelin and leptin/adiponectin ratios. Our findings indicate that an improvement of the follicular leptin signaling and normalization of the balance between leptin and its functional antagonists, ghrelin and adiponectin, may be one of the effective ways to increase the effectiveness of IVF. In the case of a decrease in leptin levels (nPN group), leptin replacement therapy can be considered as a promising approach for normalizing the adipokine status in the ovaries [27]. In women with overweight, hyperleptinemia and leptin resistance (nPI group), two different approaches can be used. The first is to reduce body weight and improve metabolic status, which reduces hyperleptinemia, normalizes leptin transport into the ovaries and improves the leptin signaling in follicular cells [28]. The second approach is to use the activators of leptin signaling, for example, the inhibitors of protein phosphotyrosine phosphatase 1B [29]. However, drugs capable of activating the leptin signaling are currently still under investigation and are not widely used in clinical practice. In the future, studies of follicular adipokines levels and their ratios in women, differing not only in the BW, but also in the metabolic and hormonal parameters, as well as those with metabolic disorders associated with changes in adipokine production and their balance, are needed.

Geolocation information

Saint-Petersburg, Russia; Word count — 2095

Disclosure statement

The authors report no conflict of interest

Additional information

Funding

This work was supported by the Fundamental research program of Russian Academy of Sciences under Grant АААА-А19-119030490047-8 and partially supported by the Russian Foundation of Basic Investigations and DST [№ 18-515-45004 IND-a; № INT/RUS/RFBR/P-310].