Key words
obesity - body mass index - waist circumference - ovarian reserve
Schlüsselwörter
Übergewicht - Body-Mass-Index - Taillenumfang - ovarielle Reserve
Introduction
As simple definitions, obesity is an excess storage of triglycerides in adipose tissues
and overweight is having a body weight, including muscles, fat, and body water, in
excess of a standard or ideal body weight [1]. According to the World Health Organization (WHO), the term overweight has been
defined as having a body mass index (BMI) equal to or greater than 25 kg/m2, while obesity is having a BMI equal to or greater than 30 kg/m2
[2]. The BMI is a practical and globally-used measure of obesity. However, it does not
consider the personʼs degree of abdominal obesity [3]. On the other hand, waist circumference (WC) is an easy and reliable measure of
visceral adipose tissue and can be used as a simple index of cardiovascular risk [4]. The WHO has reported sex-specific cutoff values that define overweight and obesity
[3]. With regard to abdominal obesity, the WC cutoff value specific to the population
of Turkish women is ≥ 80 cm estimated as 80 cm [5].
Obesity has emerged as a global epidemic. The International Obesity Task Force reported
that 1.1 billion adults are overweight, and 312 million of these are women who are
considered obese [6]. The WHO has reported that 60% of women are overweight in the United States and
most European countries, and 30% of these women are considered obese [2], [6]. The prevalence of obesity has reached more than one third of the adult population
in Turkey, similar to most of the worldwide populations [7], [8].
Obesity is associated with numerous personal characteristics including social, psychological,
and demographical issues, and health problems. Moreover, being obese has a detrimental
impact on most organ systems, including the reproductive system. Although the mechanisms
are complex and multifactorial, obese women are more prone to anovulation and having
abnormal uterine bleeding, endometrial hyperplasia, gynecological cancers, infertility,
miscarriage, and antenatal and neonatal complications [9], [10], [11]. Obese women also exhibit impairments in ovarian, follicular, and oocyte development,
fertilization, and embryo development and implantation [12], [13].
The quality and quantity of the oocyte reservoir has been described as ovarian reserve.
This refers to an indirect index of a femaleʼs reproductive capacity [14]. Like all screening tests, ovarian reserve tests are aimed at identifying individuals
at risk for diminished ovarian reserve. Over the past two decades, several methods
have been described to predict the ovarian reserve. Anti-Mullerian hormone (AMH) is
a member of the transforming growth factor β superfamily of glycoproteins and is produced
by the granulosa cells of the preantral and small antral follicles [15]. AMH reflects the quantity of growing follicles and the ovarian follicular pool
[16]. Although basal follicle-stimulating hormone (FSH) levels are the most commonly
used ovarian reserve test, antral follicle counts and AMH levels are promising predictors
with significant potential advantages [17]. For now, AMH is considered as a reliable indicator of declining ovarian function
since AMH exhibits minimal inter- and intra-cycle variations [18].
There are numerous publications stating that the obesity leads to infertity through
various pathways, including impaired follicular development, quantitative and qualitative
development of the oocyte [12], [19]. In the existing literature, conflicting results from studies examining the relationship
between obesity and ovarian reserve have been reported. While some studies have demonstrated
a significant reverse correlation [20], [21], others have demonstrated no relationship between ovarian reserve markers and obesity
[22], [23], [24].
Therefore, we aimed to investigate whether obesity adversely affects serum concentrations
of ovarian reserve markers in women with different ovarian reserve status in the present
study.
Material and Methods
Study population
A total of 485 infertile women who had attended the Reproductive Endocrinology Department
at Hitit University Hospital between February and December 2015 were included in this
prospective cross-sectional study. The study was performed in accordance with the
Declaration of Helsinki 2013 Brazil version and was approved by the local ethics board
of Ankara Numune Education and Research Hospital (20796219–724.087). All participants
gave their written informed consent prior to the beginning of the study. The comprehensive
personal and family histories were obtained from all participants. The participants
who had a history of pelvic surgery, endometriosis, tubo-ovarian masses, cardiovascular,
hepatic, renal and respiratory diseases, diabetes mellitus, malignancies, chemotherapy,
or radiotherapy were excluded from the study. The current use of hormones or drugs
that may have affected ovarian functions, smoking, pregnancy and lactation, an FSH
level > 15 IU/L, hyperprolactinemia, and an age ≥ 45 years were also considered exclusion
criteria. In total, 402 participants who met the inclusion criteria were included
into the study.
Study design
A physical examination and measurements for weight, height, and WC were performed
on all participants. BMI was calculated by dividing the weight in kilograms by the
height in square meters. WC values were measured at the level of the iliac processes
and umbilicus with the same scale to evaluate abdominal obesity. The transvaginal
ultrasound examinations were also performed during the early follicular phase of the
menstrual cycle (day 2 to 4) by means of a transvaginal 7.5 MHz probe (Toshiba Xario
100, Toshiba Medical Systems Co., Nasu, Japan) by the same clinician. The total antral
follicle count (AFC) was the sum of the small follicles (2 – 10 mm in diameter) in
each ovary.
All participants were assigned to three groups according to BMI (< 25 kg/m2: normal group, n = 198; 25.0 – 29.9 kg/m2: overweight group, n = 126; and ≥ 30 kg/m2: obese group, n = 78). The women were also divided into two groups according to WC
(< 80 cm: normal group, n = 103; and ≥ 80 cm: obese group, n = 299). Additionally,
all participants were categorized into three types of ovarian reserve patterns. These
patterns included the normal ovarian group (n = 146), which included women with an
adequate reserve pattern (women with ≥ 7 baseline AFC), the high ovarian group (n = 112)
including women with a high reserve pattern according to the Rotterdam Consensus on
Diagnostic Criteria for polycystic ovarian syndrome (PCOS) [25], and the poor ovarian group (n = 144) that included women with a diminished reserve
pattern (women with < 7 baseline AFC), based on a committee opinion concerning ovarian
reserve by the American College of Obstetricians and Gynecologists [26]. The comparisons of all data in all ovarian reserve groups were then drawn.
Data collection and assays
The blood samples were obtained from the antecubital vein after an overnight fast
between 8:00 a. m. and 10:00 a. m. in the early follicular phase on day 2 or 3. The
blood samples were collected into 5 mL serum separator tubes (BD Vacutainer, Becton
Dickinson, New Jersey, USA). The samples were allowed to clot completely at room temperature
and were then centrifuged at 1500 g for 4 min within 30 min. Serums were analyzed
daily for estradiol (E2) and FSH with an electrochemiluminescence immunoassay (ECLIA)
method using an autoanalyzer (Cobas 6000, E 601 Roche Diagnostics, GmbH, Mannheim,
Germany). To obtain minimal fluctuations in the samples, the serum for AMH measurements
were frozen at − 20 °C within 2 hours for a maximum of 7 days and then analyzed. All
analyses of AMH samples were also performed on a weekly basis by the ECLIA method
using an autoanalyzer (Cobas 6000, E 601 Roche Diagnostics, GmbH, Mannheim, Germany).
Statistical analyses
All analyses were performed using SPSS (Statistical Packages for The Social Sciences)
software version 21 (SPSS Inc., Chicago, IL, USA). For the categorical variables,
chi-square tests were used. The continuous variables were first evaluated for normality
of statistical distribution by Shapiro-Wilk tests. As the continuous variables were
not normally distributed, non-parametric methods (Mann-Whitney U tests) were used
to perform statistical analyses. The descriptive statistics were expressed as a median
(minimum-maximum) and number (percentage %). The Spearman correlation tests were used
to determine the correlations of continuous variables. A p value of < 0.05 and a confidence
interval of 95% were considered as statistically significant.
Results
Baseline demographic and biochemical characteristics
The demographic and biochemical characteristics of the study population regarding
the aspects of ovarian reserve groups are presented in [Table 1]. The comparisons of the mean ages in all ovarian reserve groups demonstrated a statistically
significant difference (p < 0.001). The median age was 29 years (18 – 34) for the
normal reserve group, 27.5 years (18 – 34) for the high reserve group, and 36 years
(20 – 45) for the poor reserve group (p < 0.001). There were no significant differences
in BMI and WC between the three ovarian reserve groups (p = 0.813 and p = 0.065, respectively).
As expected, significant differences between all ovarian reserve groups with regard
to AMH, E2, and FSH levels were detected. The median AMH values of the normal, high,
and poor ovarian groups were 2.3 (0.1 – 11.3) ng/dL, 5.4 (1.6 – 20.0) ng/dL, and 1.2
(0.1 – 10.7) ng/dL respectively, (p < 0.001). Thus, the women with poor ovarian reserve
had lower AMH concentrations. The median E2 values of the normal, high, and poor ovarian
groups were 44.0 (3.0 – 263.0) pg/mL, 37.0 (5.0 – 167.1) pg/mL, and 46.0 (5.0 – 347)
pg/mL (p < 0.001), respectively. The normal, high, and poor ovarian groups also had
statistically different median FSH levels with values of 7.0 (0.1 – 13.8) IU/L, 6.0
(3.3 – 9.9) IU/L, and 7.6 (0.2 – 15.5) IU/L respectively, (p < 0.001).
Table 1 The demographic and biochemical charecteristics of the study population.
|
|
Normal ovarian reserve group (n = 146)
|
High ovarian reserve group (n = 112)
|
Poor ovarian reserve group (n = 144)
|
p
|
|
Values are shown as median (minimum-maximum). BMI = Body mass index; WC = Waist circumference;
AMH = Anti-müllerian hormone; E2 = Estradiol; FSH = Follicular stimulating hormone.
* p-values indicate statistically significant (p < 0.05).
|
|
Age (years)
|
29 (18 – 34)
|
27.5 (18 – 34)
|
36 (20 – 45)
|
0.000*
|
|
BMI (kg/m2)
|
|
|
77 (38.9%)
|
55 (27.8%)
|
66 (33.3%)
|
0.813
|
|
|
44 (34.9%)
|
35 (27.8%)
|
47 (37.3%)
|
|
|
25 (32.1%)
|
22 (28.2%)
|
31 (39.7%)
|
|
WC (cm)
|
|
|
37 (43.1%)
|
29 (36.1%)
|
37 (20.8%)
|
0.065
|
|
|
109 (31.6%)
|
83 (30.6) %
|
107 (37.8%)
|
|
AMH (ng/dL)
|
2.3 (0.1 – 11.3)
|
5.4 (1.6 – 20.0)
|
1.2 (0.1 – 10.7)
|
0.000*
|
|
E2 (pg/mL)
|
44.0 (3.0 – 263.0)
|
37.0 (5 – 167.1)
|
46.0 (5.0 – 347.0)
|
0.002*
|
|
FSH (IU/L)
|
7.0 (0.1 – 13.8)
|
6.0 (3.3 – 9.9)
|
7.6 (0.2 – 15.0)
|
0.000*
|
Relationship of ovarian reserve parameters to BMI subgroups
The comparisons of ovarian reserve parameters (AMH, FSH, and E2) and BMI subgroups
are shown in [Table 2]. All participants (n = 402) were grouped according to ovarian reserve status: normal
(n = 146), high (n = 112), and poor (n = 144) reserve groups. The mean serum AMH and
FSH concentrations were statistically similar in all ovarian reserve groups and their
BMI subgroups (p > 0.05, for all). By comparison of the mean E2 level, a significant
difference was detected in the normal ovarian reserve group (p = 0.004). In this group,
the women of normal weight (BMI < 25) had the highest E2 (52.0 pg/mL) level. However,
the high and poor ovarian reserve groups did not possessed any significant difference
(p > 0.05, for all).
Table 2 The comparisons of ovarian reserve parameters and BMI in different ovarian reserve
groups.
|
|
Normal ovarian reserve group (n = 146)
|
High ovarian reserve group (n = 112)
|
Poor ovarian reserve group (n = 144)
|
|
BMI (kg/m2)
|
< 25 (n = 77)
|
25.0 – 29.9 (n = 44)
|
≥ 30 (n = 25)
|
p*
|
< 25 (n = 55)
|
25.0 – 29.9 (n = 35)
|
≥ 30 (n = 22)
|
p*
|
< 25 (n = 66)
|
25.0 – 29.9 (n = 47)
|
≥ 30 (n = 31)
|
p*
|
|
Values are shown as median (minimum-maximum). BMI = Body mass index; AMH = Anti-Müllerian
hormone; E2 = Estradiol; FSH = Follicular stimulating hormone.
* p-values indicate statistically significant (p < 0.05).
|
|
AMH (ng/dL)
|
2.3 (0.1 – 9.4)
|
2.3 (0.1 – 11.3)
|
1.9 (0.1 – 7.0)
|
0.485
|
5.5 (1.7 – 20.0)
|
5.4 (1.6 – 19.2)
|
5.6 (1.8 – 14.0)
|
0.687
|
1.3 (0.1 – 10.7)
|
1.02 (0.1 – 8.7)
|
1.1 (0.1 – 6.7)
|
0.394
|
|
E2 (pg/mL)
|
52.0 (5.0 – 263.0)
|
38.3 (8 – 158.0)
|
40.0 (3.0 – 81.0)
|
0.004*
|
35.0 (5.0 – 73.0)
|
40.6 (5.0 – 167.1)
|
37.0 (18.0 – 77.0)
|
0.097
|
46.2 (5.0 – 193.0)
|
47.0 (9.8 – 337.0)
|
40.0 (19.0 – 347.0)
|
0.827
|
|
FSH (IU/L)
|
7.2 (3.1 – 13.0)
|
6.7 (4.4 – 13.4)
|
6.5 (0.1 – 13.8)
|
0.187
|
6.2 (3.3 – 9.1)
|
6.0 (3.4 – 9.9)
|
5.5 (3.7 – 9.5)
|
0.325
|
7.8 (0.2 – 15.5)
|
7.4 (3.6 – 13.4)
|
7.2 (2.9 – 12.4)
|
0.630
|
Relationship of ovarian reserve parameters to WC subgroups
The ovarian reserve parameters of study population were compared according to their
WC subgroups, as demonstrated in [Table 3]. Although the mean FSH levels of normal ovarian reserve group were 7.2 IU/L and
6.7 IU/L respectively, no statistical significance was found (p = 0.177). On the contrary to E2 comparisons in the normal ovarian BMI subgroups, the
mean E2 levels in all ovarian reserve groups and WC subgroups were statistically similar
(p > 0.05, for all). We also demonstrated that the mean AMH levels were not statistically
different from each other in the three ovarian reserve groups (p > 0.05, for all).
Table 3 The comparisons of ovarian reserve parameters and WC in different ovarian reserve
groups.
|
|
Normal ovarian reserve group (n = 146)
|
High ovarian reserve group (n = 112)
|
Poor ovarian reserve group (n = 144)
|
|
WC (cm)
|
< 80 (n = 37)
|
≥ 80 (n = 109)
|
p*
|
< 80 (n = 29)
|
≥ 80 (n = 83)
|
p*
|
< 80 (n = 37)
|
≥ 80 (n = 107)
|
p*
|
|
Values are shown as median (minimum-maximum). WC = Waist circumference; AMH = Anti-Müllerian
hormone; E2 = Estradiol; FSH = Follicular stimulating hormone.
* p-values indicate statistically significant (p < 0.05).
|
|
AMH (ng/dL)
|
2.3 (0.4 – 9.2)
|
2.3 (0.1 – 9.4)
|
0.775
|
5.0 (1.6 – 19.2)
|
5.5 (1.8 – 20.0)
|
0.132
|
0.8 (0.1 – 4.3)
|
1.2 (0.1 – 10.7)
|
0.529
|
|
E2 (pg/mL)
|
41.3 (11.0 – 230.0)
|
43.0 (3.0 – 263.0)
|
0.898
|
35.4 (15.0 – 65.0)
|
37.5 (5.0 – 167.1)
|
0.692
|
46.0 (27.0 – 193)
|
46.7 (9.8 – 123.0)
|
0.963
|
|
FSH (IU/L)
|
7.2 (4.7 – 12.6)
|
6.7 (0.14 – 13.4)
|
0.177
|
6.3 (3.4 – 8.7)
|
6.0 (3.3 – 9.9)
|
0.880
|
8.3 (4.7 – 12.6)
|
7.2 (2.1 – 15.5)
|
0.178
|
Discussion
The overall aim of this study was to investigate whether obesity has an impact on
the measures of ovarian reserve. Specifically, we conducted a comprehensive evaluation
of this impact by assigning participants to three ovarian reserve groups. We hypothesized
that obesity and body size would influence ovarian reserve markers. However, our findings
did not support the notion that ovarian reserve is impaired in obese women.
In an earlier study, higher BMI levels were associated with lower inhibin B levels
in premenopausal women, while the opposite results were observed in postmenopausal
women [15]. A cross-sectional study by Su et al. explored the association between obesity and
serum and ultrasound measures of ovarian reserve in women of late reproductive age.
This group reported that AMH was lower in obese women, while antral follicle count
did not differ in accordance with body size [27].
De Pergola et al. proposed that overweight and fertile women with obesity have lower
FSH, LH, inhibin B, and E2 concentrations. They concluded that a possible inhibitory
effect of obesity on gonadotropins and E2 production may be one of the elucidative
mechanisms [28]. Other authors have also reported similar results. Freeman et al. noted a negative
correlation between BMI and AMH in later reproductive-age women. They found that obese
women had 65% lower AMH levels compared to women of normal weight [20]. Steiner et al. also reported that obese women had 34% lower levels of AMH [29]. However, other authors have suggested that lower AMH levels in late reproductive-age
obese women result from physiological processes other than decreased ovarian reserve
[20], [27].
In contrast to the studies in which ovarian reserve has been negatively affected by
high BMI levels, several studies have demonstrated that serum and ovarian markers
of diminished ovarian reserve (DOR) did not exhibit consistent changes with body size
[30], [31], [32]. In parallel with our study, some authors have reported no significant relationships
between circulating AMH levels and obesity. Nardo et al. and Halawaty et al. did not
detect any differences in FSH, AMH, AFC, and ovarian volume [22], [33]. Furthermore, in a prospective case-controlled study, Shaw et al. found no correlation
between AMH levels and increased BMI levels in 135 Caucasian premenopausal women,
16% of whom were obese women younger than 45 years of age [34]. In a cross-sectional study from Turkey, Sahmay et al. also evaluated 259 premenopausal
women, 14% of whom were were obese women under 45 years of age. They concluded there
was no significant association between AMH and obesity [35].
Moy et al. sought to determine whether the association between obesity and AMH would
be dependent on racial factors. They performed a study that included 350 women, 99
of whom were African-Americans, 59 were Caucasians, 58 were Hispanics, and 34 were
Asians. They concluded that obesity is inversely correlated with AMH in Caucasian
women, but not in other racial groups. Interestingly, when women with DOR were examined
independently, an association between BMI and AMH was reported. However, this same
study demonstrated no association between BMI and AMH in women with normal ovarian
reserve [36]. Moreover, all ovarian reserve patterns, including normal, high, and poor patterns,
were not affected by increased BMI and body size. However, in our study there was
no such relationship between obesity and AMH in 402 Caucasian premenopausal women.
Moreover, we did not demonstrate any association between BMI and AMH in women with
normal, high, and poor reserves.
In contrast to the present study, a very recent systematic review and meta-analysis
conducted by Moslehi et al. reported that ovarian reserve markers of AMH and FSH were
significantly lower in obese than in non-obese women [37].
Interestingly, another finding is that the mean E2 level was higher only in the women
of normal weight of the normal ovarian reserve group. In a study, it was demonstrated
that higher serum levels of leptin in obese women correlate with higher levels of
leptin in the follicular fluid [38]. In vitro studies have shown that leptin affects steroidogenic pathways in granulose
cells, decreasing estrogen and progesterone production in a dose dependent manner
[39], [40]. This mechanism for the adipokines might be the explanation of the highest E2 level
in the group of women of normal weight. Nevertheless, it is essential to accomplish
further studies in order to reveal the causal relationship between E2 and BMI subgroups
in other ovarian reserve groups.
The main strength of our study is the large sample size compared to many prior studies
evaluating the effects of obesity on ovarian reserve markers. Another strength is
that we included WC as another marker of abdominal obesity. In addition to these strengths,
three groups of ovarian reserve (normal, high, and poor) were evaluated to determine
the effect of obesity on different levels of ovarian reserve. A potential limitation
of this study is that the study population primarily consisted of women who were being
treated for infertility, and we did not have any fecundity data for the study group.
In summary, our findings do not support any effect of obesity on ovarian reserve markers.
In other words, the parameters of ovarian reserve do not seem to be affected by increased
BMI and WC. Nevertheless, the decrease in fecundity in infertile obese women may arise
from other multiple factors including folliculogenesis and endometrial receptivity.
Because there is no relationship between obesity and AMH, AMH may be considered as
a reliable marker of ovarian reserve. The etiologies for diminished ovarian reserve
other than obesity also may be necessary to investigate. Further clinical and basic
research studies are needed to elucidate the role of obesity on ovarian reserve and
reproductive functions.
Funding
This research was supported by the Scientific Research Unit of Hitit University, Corum,
Turkey (TIP19001.14.005).
