alexa An 1 h-OGLT is an Appropriate Approach for the Determination of Glucose and Insulin Dynamics in Female Functional Androgenization (Including “ Polycystic Ovary Syndrome ”) | Open Access Journals
ISSN: 2161-1017
Endocrinology & Metabolic Syndrome
Like us on:
Make the best use of Scientific Research and information from our 700+ peer reviewed, Open Access Journals that operates with the help of 50,000+ Editorial Board Members and esteemed reviewers and 1000+ Scientific associations in Medical, Clinical, Pharmaceutical, Engineering, Technology and Management Fields.
Meet Inspiring Speakers and Experts at our 3000+ Global Conferenceseries Events with over 600+ Conferences, 1200+ Symposiums and 1200+ Workshops on
Medical, Pharma, Engineering, Science, Technology and Business

An 1 h-OGLT is an Appropriate Approach for the Determination of Glucose and Insulin Dynamics in Female Functional Androgenization (Including “ Polycystic Ovary Syndrome ”)

Wetzka B*, Textor W and Geisthovel F

Centre of Endocrinology and Reproductive Medicine Freiburg, Bismarckallee, Freiburg, Germany

corresponding Author:
Yoko Ozawa, MD, Ph.D
Laboratory of Retinal Cell Biology
Department of Ophthalmology
Keio University School of Medicine
35 Shinanomachi, Shinjuku-ku
Tokyo 160-8582, Japan
Tel: +81-3-5363-3869
Fax: +81-3-5363-3869
E-mail: [email protected]

Received Date: April 27, 2013; Accepted Date: May 17, 2013; Published Date: May 22, 2012

Citation: Wetzka B, Textor W, Geisthovel F (2013) An 1 h-OGLT is an Appropriate Approach for the Determination of Glucose and Insulin Dynamics in Female Functional Androgenization (Including “Polycystic Ovary Syndrome”). Endocrinol Metab Synd S1:011. doi: 10.4172/2161-1017.S1-011

Copyright: © 2013 Ozawa Y, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Visit for more related articles at Endocrinology & Metabolic Syndrome

Abstract

Female functional androgenization syndrome (FAS) is subdivided into 4 groups corresponding to their predominant organ pathology: I (ovary), II (adrenal gland), III (ovary, adipose tissue, liver, pancreas) and IV (residual dysfunctions). Group-specific variable clusters are defined by BMI, hormones, glucose and insulin during oral glucose loading test (OGLT: 0-, 1 h-, 2 h-values), and sonographic ovarian morphology. Abdominal circumference (AC), serum lipids, and blood pressure are used for individual characterization. We investigated which tests assess best the prevalence of pathologic glucose and insulin dynamics in androgenized women. 89 FAS patients were consecutively enrolled. Including a control (n=16), prevalence of insulin resistance (IR) described by OGLT, HOMA-IR, QUICKI, insulin sensitivity index (ISI) or AUC2h-insulin was analyzed. Uni- and multivariate correlation between insulin and additional variables used by the FAS classification system were performed. Regarding the prevalence of IR markers, the 1 h-insulin value was allocated above the individual cut-off value in 66.7%, 62.5% and 22.2% in FAS III, IV, and II patients, respectively. The prevalence of impaired fasting glucose or glucose tolerance was 7 and 14% in FAS III women only. In ROC analysis, insulin 1 h at a cut-off value of 99 mU/L predicted metabolic syndrome with a specificity of 74% and sensitivity of 70%. Furthermore, 1 h-insulin was highly significantly correlated with ISI and AUC2h-insulin. Testosterone correlated significantly with BMI, AC, insulin, HOMA-IR, ISI, glucose and triglycerides. The variance of insulin (0, 1 h, 2 h) was significantly explained by AC, HDLcholesterol and testosterone. In conclusion, the high prevalence of IR in accurately defined androgenized women supports the FAS classification system comprising an OGLT with the analysis of both glucose and insulin. For screening of IR in androgenized patients, an OGLT of 1h (including insulin) appears to be a reliable test approach particularly considering time and cost consumption in this special group of patients at fertile age.

Keywords

Androgenization; Hyperinsulinemia; Cholesterol

Introduction

As described previously [1,2], functional androgenization (FA) consists of 5 groups corresponding to their predominant organ pathology: functional cutaneous androgenization (FCA; skin), functional androgenization syndrome (FAS) I (ovary), II (adrenal gland), III (ovary, obese tissue, pancreas, liver) and IV (residual FA dysfunctions). This systematic classification of FCA and FAS I – III is based on different, strictly determined, group specific variable clusters consisting of primary variables like cutaneous androgenization, body mass index (BMI), luteinizing hormone (LH), testosterone (T), sex hormone-binding globulin (SHBG), free androgen index (FAI), anti-Müllerian hormone (AMH), enlarged polyfollicular ovaries (EPO), dehydroepiandrosterone-sulphate (DHEAS), 17-hydroxyprogesterone (17-OH-P) and an oral glucose loading test (OGLT) including the determination of both glucose and insulin [1]. Secondary variables such as abdominal circumference (AC), cycle irregularities, serum levels of triglycerides, total cholesterol and its lipoprotein fractions as well as blood pressure (BP) are used for patient´s individual characterization [1]. A further sub-classification of FAS I-III into a sub-set “a” showing a full-blown feature and a sub-set “b” reflecting merely a non-classic minimum standard core and/or a miscellaneous constellation (cutaneous - ovarian - adrenal - metabolic) was regarded to be useful. FAS IV patients include women who could not be allocated to neither FCA nor FAS –III. Recent analysis revealed an important sub-group who showed a hepatic-metabolic origin of (indirect) hyperandrogenemia caused by obesity and/or hyperinsulinemia [3].

In contrast, “polycystic ovary syndrome” (“PCOS”) as defined by Rotterdam consensus (RC) comprises 2 major groups, one with and one without metabolic syndrome (MetS) [4]. These two groups are further divided in up to 12 sub-sets each [5] depending on the combination of the four, partly mis-powered or imprecisely defined variables (“Oligo- and/or anovulation”, “Clinical and/or biochemical signs of hyperandrogenism”, “Polycystic ovaries”), the presence of at least two variables is sufficient for the diagnosis of PCOS. Using these criteria, the diagnosis of PCOS is relatively indefinite not only under clinical but also under scientific conditions of which the presence of two variables is sufficient for the diagnosis of PCOS. Additionally, the RC specifies that no further test of insulin resistance (IR) is necessary for making the diagnosis or for selecting the respective treatment, and recommends that merely obese women (BMI>27 kg/m2) with “PCOS” should be screened for MetS including an OGLT (without the determination of insulin) [4]. Recent studies described a varying, but important prevalence of MetS ranging between 33 and 40% in women diagnosed with “PCOS” [6]. In comparison, the prevalence of MetS in apparently healthy German women is much lower ranging between 3-12.4% in the age group 16-45 y (Prospective Cardiovascular Muenster (PROCAM) study [7]).

The definition of “MetS” according to the Adult Treatment Panel III (ATP III) [8] used by the RC was set up in order to identify individuals at increased risk for cardiovascular disease (CVD). The MetS is based primarily on IR [9,10] which represents a state of impaired metabolic response to insulin. The diagnosis of IR is, however, still controversial: The “gold standard” constitutes the insulin-mediated glucose disposal which is, however, too sophisticated to be used in general practice, and reference values are difficult to define for general population [9]. As surrogate variables, fasting insulin, indices like HOMA-IR (homeostasis model assessment of insulin resistance [11]), QUICKI (quantitative insulin sensitivity check index) [11], the insulin sensitivity index (ISI) or Matsuda-index [12]), insulin during an OGLT [13] as well as area under the curve (AUC) for insulin and glucose levels during an OGLT [14] have been suggested and intensively studied. Most of these tests are primarily developed in order to detect a prediabetic state or a type 2 diabetes mellitus in the older population. However, to date there are no generally accepted normal values of IR test procedures for the special cohort of women being at fertile age where issues regarding reproduction or prevention of future disease are in the main focus of interest. Since women suffering from “PCOS” have e. g. an enhanced risk to develop type 2 diabetes in later life [15,16], relevant biomarkers for the prediction of this disease would be desirable.

The aims of the study were primarily to estimate the prevalence of IR [11,13] and ISI [12] in FA. Possible pathologic links between FAS variables and insulin will be evaluated. Finally, the appropriate tests for assessing and/or predicting best and easiest glucose and insulin dynamics in women presenting with FA will be discussed.

Material and Methods

Study population

Over a period of 2.5 years, 104 premenopausal women being examined at the Centre of Gynaecological Endocrinology and Reproductive Medicine Freiburg (CERF, Freiburg i. Br., Germany) were consecutively enrolled for the study and evaluated retrospectively. The reasons for patients’ clinical visit were cutaneous androgenizing symptoms like acne vulgaris, hirsutism, and androgenetic alopecia, menstrual cycle irregularities (oligo-, amenorrhea), infertility, hyperandrogenemia, obesity, decreased circulating SHBG levels as well as disturbed glucose and insulin metabolism. Patients were grouped into FCA and FAS I–III according to Geisthövel et al. [1]. Due to the enormous heterogenity of FA a clear assignment into a pathogenetic overall concept established in literature was not possible in each patient. Individuals with such variable profiles belonged into group FAS IV, e.g. obese androgenized and hyperinsulinemic patients with an elevated FAI caused by decreased SHBG secretion, but without the visualization of EPOs and hypertestosteronemia [3].

The control group consisted of 16 healthy, normal weight, premenopausal women with no signs of FA recruited by newspaper advertisements. These women had a history of >3 regular menstrual cycles during the last 6 months (21 to 35 days); mean circulating progesterone values obtained around 7 days prior to the expected start of menstrual bleeding were >30 nmol/L, suggesting a sufficient luteal phase.

All individuals had not taken any hormonal or metabolic medication during the preceding 3 months aside from the treatment with iodine and/or thyroxine. The study was approved by the ethical committee of the University of Freiburg, Freiburg i. Br. /Germany, and patients and volunteers gave informed consent.

Clinical examination

Reading of BP was obtained in sitting patients after a rest of 5 to 15 min. The inflatable cuff size applied was adapted to the upper arm circumference. The AC was measured in the standing position, halfway between the lower ribs and the superior anterior iliac spine of the pelvis. Hirsutism was described by applying the Ferriman-Gallwey score [17].

Ultrasonography of the ovaries using a vaginal ultrasound probe (Sonoline SI 200, Siemens, München/Germany; Logiq 200 Pro Series, General Electrics, Solingen/Germany; Sonoace SA 8000, Marl/ Germany) was performed at the time of blood sampling and was described before [1,2]. Briefly, an EPO corresponds to the so called “polycystic ovary” (“PCO”) [4,5] and was defined by a maximum (max.) ovarian diameter of ≥ 31 mm and an antral follicle count (3- 10 mm follicles) per max. ovarian area ≥ 8. The presence of EPOs was evaluated binary, thereby the unilateral visualization/individual was considered positive.

Endocrine and metabolic variables

Venous blood sampling was performed between days 3 to 6 of regular menstrual cycles and at random in oligo-amenorrhea. The functional allocation of an early- to mid-follicular phase was hormonally confirmed by serum levels of estradiol <440 pmol/L and progesterone ≤6 nmol/L, as described previously [1]. After an overnight (12 h) fast, 3 venous blood samples were obtained between 8 a.m. and 9 a.m. at 20 min intervals (baseline value, 0 h). Blood samples were centrifuged and the sera pooled in equal volumes, processed immediately, and an aliquot was stored at –20°C until assayed. Serum LH, FSH, T, DHEAS and 17-OH-P levels were determined using commercially available immunoassay kits (ACS – Automated Chemiluminescence System, Bayer Health Care, Fernwald/Germany). SHBG was analysed by a highly specific radioimmunoassay kit from DSL (Sinsheim/ Germany). Serum glucose was analysed using an enzymatic UV test on the basis of hexokinase and insulin by an electro chemiluminescence immunoassay (ECLIA, Modular E170) (both MVZ Clotten, Freiburg/ Germany). Furthermore, a lipid analysis was performed by measuring total triglycerides and cholesterol and by running a cholesterol electrophoresis (MVZ Clotten, Freiburg/Germany). Estradiol, progesterone, prolactin, cortisol, thyroidea stimulating hormone (TSH) and free thyroxine (analysed on the ACS, Bayer/Germany) were determined as internal controls to confirm the early follicular phase and to exclude other pathologies. Additionally, the free androgen index (FAI=T/SHBG×100) was calculated. Serum AMH was analysed in kryo-preserved sera using the highly specific enzyme immunoassay kit from DSL (Sinsheim/Germany). For all assays, the intra- and interassay variations were <6% and <10%, respectively.

In addition to determining baseline fasting serum levels at 0 h, an OGLT was performed in all patients and controls, thereby an oral administration of 75 g glucose (Dextro® O.G-T.; Roche/Germany) was administered after the last baseline blood sampling, and further blood samples were collected 1 and 2 h afterwards to determine serum glucose and insulin levels. Thereafter, 250 μg ACTH (Synacthen®; Novartis Pharma/Germany) was administered i.v. over 3 minutes, and a last blood sample was collected one hour later to determine serum 17- OH-P (ACTH test) as an endocrine marker of cytochrome peroxidase 21A2 deficiency.

Definitions

Regarding the glucose cut-off values during the OGLT, a modified definition was applied in the present study based on the criteria of the American Diabetes Association which defined IFG as having fasting glucose levels between 100-125 mg/dL [18], and on results of former studies [1,13,19]: the cut-off values of glucose 0 (fasting), 1 h and 2 h (following the OGLT) were 100 mg/dL, 140 mg/dL, and 110 mg/dL, respectively. In addition, insulin was determined during the OGLT according to earlier studies [13,19]. Cut-off values for hyperinsulinemia using the definition “elevation of insulin > 2 SD above the mean of the control group” [1,19] or for IR according to the definition “elevation of insulin > 3rd quartile of normal weight apparently healthy women” [13] were outlined as follows: insulin 0>15 mU/L, insulin 1 h>99 mU/L, and insulin 2 h>85 mU/L, respectively. Both analyses considered obese and non-obese women at fertile age in a gynaecological outpatient clinic, an approach which we thought to be more representative for the FA (“PCOS”) patients than previous studies looking for both male and female patients with CVD at older age.

The homeostasis model assessment of insulin resistance (HOMAIR) was calculated using the following equation: fasting insulin [mU/L] × fasting glucose [mmol/L])/22.5 [11]. An upper limit of 2.7 was used [13,14]. Additionally, the quantitative insulin sensitivity check index (QUICKI) was calculated as follows: 1/(log fasting insulin [mU/L] + log fasting glucose [mg/dl]) [11]. A lower limit of 0.14 was used [13].

The insulin sensitivity index (ISI) [12] includes fasting glucose and insulin as well as their levels during OGLT and is calculated as follows: 10000/√(fasting glucose [mg/dl] × fasting insulin [mU/L]) × (mean glucoseoglt × mean insulinoglt). An ISI<6 was classified as insulin resistant according to Bals-pratsch et al. [20] who used OGLT with the determination of 3 values (fasting, 1 h and 2 h) in women of reproductive age.

The AUC2h for glucose and insulin during OGLT was calculated by the trapezoid rule. An upper cut-off of 7000 mU/L/2 h was applied for insulin according to Lunger et al. [14] investigating women of fertile age.

Statistical analysis

For comparison between FAS groups, ANOVA on ranks (Dunn´s test) was used. Tests were performed using Sigmastat (Jandel Corporation/USA); significance was defined with P<0.05. Linear correlation and multivariate analysis (backward step procedure with P=0.1) was used to analyse relations between parameters of anthropometry, ovarian function and metabolism, especially insulin during OGLT. Analysis of receiver operating characteristics (ROC) was performed for the prediction of MetS (as defined by RC [4]) by insulin. These tests were carried out with SAS V8 for Windows (SAS Institute, Cary/NC).

Results

After exclusion of women presenting with additional pathologies like hyperprolactinemia or missing data (n=13), 91 individuals were left for further study. Since only 2 patients were grouped into FCA, this group was eliminated. Thus, 89 patients who were classified into groups FAS I, II, III or IV, respectively, remained for final analysis. Within the group FAS IV, 5 women (62.5%) belonged to the subgroup of hepatic-metabolic origin of indirect hyperandrogenemia [3].

Anthropometric and basic endocrinological parameters of FA patients and controls necessary for FAS classification are shown in table 1 and were already described earlier [1,2]. Levels of 17-OH-P were not in the pathological range (>30 ng/ml), therefore a significant cytochrome peroxidase 21A2 deficiency (congenital adrenal hyperplasia) could be ruled out. FAS patients were significantly younger vs. control (Table 1). Groups FAS III and IV revealed significantly different levels of BMI and AC, glucose 2 h, AUC2h glucose, insulin 1 h, AUC2h insulin, HOMA-IR, ISI and HDL-cholesterol vs. control as shown in detail in table 2; apart from AUC2h glucose, all these variables were significantly different in FAS III vs. FAS I, too. The comparison between HOMAIR and QUICKI gave comparable results and both variables are highly significantly correlated (R=-1, P<0.0001, Spearman´s correlation). Therefore, only HOMA-IR was considered in further evaluations presented.

Variable FAS I
N=20
FAS II
N=18
FAS III
N=43
FAS IV
N=8
Control
N=16
Age y 23.95 ± 4.2
co
25.5 ± 6.6
co
25.83 ± 5.7
co
20.25 ± 3.7
co
30.75 ± 5.5
Hirsutism 0
FGS 1°

11
6
3
0
6
9
3
0
16
16
9
2
3
3
2
0
16
0
0
0
LH U/L 12.75 ± 5.0
co, FAS II
5.58 ± 2.7 10.0 ± 4.1
co, FAS II
7.39 ± 4.8 5.18 ± 1.8
LH/FSH ratio 2.52 ± 0.9 co, FAS II 1.01 ± 0.4 2.33 ± 1.1 co, FAS II 1.54 ± 0.56 0.93 ± 0.33
DHEAS nmol/L 5.04 ± 2.1 10.66 ± 3.1
co, FAS I,III,IV
6.57 ± 3.1
co
5.89 ± 1.9 4.06 ± 1.6
17-OH-P ng/ml
0 h
1.81 ± 0.74
co
1.26 ± 0.46 1.62 ± 0.54
co
1.21 ± 0.94 0.88 ± 0.30
17-OH-P ng/ml
1 h
2.89 ± 1.03
co
2.81 ± 0.93
co
3.09 ± 1.08
co
2.37 ± 0.76 1.94 ± 0.41
SHBG mmol/L 47.35 ± 20.1 26.11 ± 11.0
co, FAS I
18.6 ± 6.0
co, FAS I
16.08 ± 7.8
co, FAS I
51.68 ± 14.6
Testosterone
nmol/L
2.45 ± 0.82
co
2.14 ± 0.76 2.85 ± 0.90
co, FAS IV
1.64 ± 0.43 1.44 ± 0.42
FAI 6.43 ± 3.98 9.02 ± 3.4 17.09 ± 8.6
co, FAS I,II
11.68 ± 4.6
co
3.06 ± 1.22
AMH ng/mL 14.63 ± 7.04
co, FAS II,IV
5.05 ± 2.76 11.76 ± 6.33
co, FAS II
5.42 ± 3.86 2.57 ±1.48
EPO % 100
18/18*
co, FAS II,IV
25
4/16*
100
39/39*
co, FAS II,IV
17
1/6*
12.5
2/16

Table 1: Anthropometric and basic endocrinological parameters.
Shown as mean ± SD. Significant differences between the 4 functional androgenization syndrome groups FAS I to IV [1,2] and the control (co) are shown in bold figures (ANOVA on ranks + Dunn´s test or Oneway ANOVA), italic letters explain the respective groups (FAS I-IV, co).

Variable FAS I
N=20
FAS II
N=18
FAS III
N=43
FAS IV
N=8
Control
N=16
BMI kg/m2 21.45 ±1.8 27.22 ± 6.9
co, FAS I
31.26 ± 5.9
co, FAS I
28.78 ± 5.4
co, FAS I
21.4 ± 1.9
AC cm 78.3 ± 6.9 89.3 ± 15.3 100.6 ± 14.6
co, FAS I
97.5 ± 15.3
co, FAS I
77.75 ± 5.6
BP systolic mmHg 109 ± 9.0 119.2 ± 15.0 125.4 ± 13.4
co, FAS I
123.1 ± 14.9 109.1 ± 11.2
BP diastolic mmHg 68.75 ± 7.2 76.9 ± 12.5 84.6 ± 11.4
co, FAS I
80.63 ± 7.8 74.1 ± 7.1
Glucose 0 mg/dL 83.1 ± 7.2 81.6 ± 8.3 86.6 ± 10.8
co
82.8 ± 4.6 79.3 ± 5.3
Glucose 1h mg/dL 106.6 ± 32.9 111.4 ± 36.9 131.9 ± 43.1
co
116.8 ± 27.5 79.9 ± 18.5
Glucose 2h mg/dL
85.95 ± 22.6 85.4 ± 16.0 102.4 ± 33.7
co, FAS I
96.9 ± 13.3 co 74.2 ± 15.1
AUC2h Glucose
mg/dL/2h
11468 ± 2395 11693 ± 2642 13398 ± 3751
co
12394 ± 1713
co
9398 ± 1411
Insulin 0 µU/mL
6.21 ± 2.5 9.91 ± 8.4 18.8 ± 21.0
co, FAS I
17.7 ± 14.9 5.96 ± 3.0
Insulin 1 h µU/mL 53.96 ± 22.1 93.04 ± 96.5
co
171.4 ± 141.4
co, FAS I
155.5 ± 121.2
co
32.56 ± 16.9
Insulin 2 h µU/mL
34.46 ± 22.5 67.36 ± 97.5 147.1 ± 202.8
co, FAS I
99.6 ± 68.2
co
26.7 ± 15.1
AUC2h Insulin
µU/mL/2 h
4457 ± 1822 7901 ± 8892
co
15258 ± 14518
co, FAS I
12845 ± 9302
co
2935 ± 1283
HOMA-IR 1.29 ± 0.56 2.06 ± 1.9 4.07 ± 4.66
co, FAS I
3.56 ± 2.79
co
1.17 ± 0.59
QUICKI 0.163 ± 0.013 0.153 ± 0.017 0.144 ± 0.017
co, FAS I
0.145 ± 0.019 0.166 ± 0.013
ISI 9.89 ± 4.49 7.29 ± 3.37 4.45 ± 3.33
co, FAS I
5.45 ± 5.15
co
13.0 ± 4.49
Triglycerides mg/dL 76.75 ± 25.0 90.78 ± 36.6 143.77 ± 63.2
co, FAS I
91.63 ± 32.2 80.75 ± 32.6
Cholesterol mg/dL 185.45 ± 32.7 192.44 ± 40.8 198.7 ± 38.2 162.6 ± 30.3 182.6 ± 27.4
HDL-Cholesterol
mg/dL
63.4 ± 16.0 61.2 ± 29.0 49.0 ± 14.7
co, FAS I
44.6 ± 18.4
co
66.6 ± 14.2
LDL-Cholesterol
mg/dL
110.7 ± 32.8 115.5 ± 35.6 130.6 ± 36.7 104.6 ± 36.0 101.25 ± 25.2
VLDL-Cholesterol
mg/dL
11.35 ± 7.9 15.82 ± 9.3 19 ± 10.7 13.25 ± 7.5 14.7 ± 5.4

Table 2: Metabolic parameters. Shown as mean ± SD. Significant differences between the 4 functional androgenization syndrome groups FAS I to IV and the control (co) are shown in bold figures.

The different prevalences of values above the respective cut-off levels (s. “Definitions”) of insulin and glucose regarding HOMA-IR [11,13], OGLT [13], AUC2h insulin [14], and ISI [12,20] in groups FAS II-IV are shown in figure 1. The prevalence of pathological values was in a comparable range for both FAS III and IV patients. FAS I was not considered in this context because pathologies of insulin and glucose dynamics are - per definition em - not present in this group. Only insulin and glucose 1 h append additional pathologic cases in comparison with HOMA-IR, insulin 0/2 h and glucose 0/2 h. Correlation analysis of the IR indices showed that insulin 1 h correlated highly significantly with AUC2h insulin and ISI and to a lesser extent with HOMA-IR (Figure 2).

endocrinology-metabolic-syndrome-insulin-resistance

Figure 1: Prevalence of the various indices for insulin resistance (IR).
For definitions and cut-off levels see Materials and Methods. Results are shown as percentage of values above the respective upper cut-off level from total number in each group (FAS II: n=18; FAS III: n=43; FAS IV: n=8). Results for insulin 1h were repeated in panel C for easier comparison with the indices shown there.Panel A: insulin 0/ h/2 h during OGLT. Panel B: glucose 0/1 h/2 h during OGLT. Panel C: insulin 1 h and indices for IR.
AUC: area under the curve; FAS: functional androgenization syndrome; HOMA-IR: homeostasis modell assessment of insulin resistance; ISI: insulin sensitivity index.

endocrinology-metabolic-syndrome-glucose-loading

Figure 2: Analysis of correlations between insulin 1 h after oral glucose loading test (OGLT).
And the area under the curve (AUC2h) for insulin during a 2 h OGLT, the insulin sensitivity index (ISI) and the homeostasis modell assessment of insulin resistance (HOMA-IR). Linear correlation analysis was performed for the left panels, Spearman correlation analysis was performed for the right panels, n=105. Correlation coefficient R and significance level P are shown.

Further variables representing highest prevalence of anthropometric, metabolic or vascular pathology in the FAS groups were AC (n=51) followed by HDL-cholesterol (n=37), elevated BP (n=35) and triglycerides (n=21). The prevalence of IFG and IGT was lower (Figure 1B), and was found only in group FAS III including one woman presenting with diabetes mellitus type II. The prevalence of MetS itself as defined by RC [4] was 44%, 38% and 28% in FAS III, FAS IV and FAS II, respectively. In ROC analysis, insulin 1 h at a cutoff value of 99 mU/L predicted MetS with a specificity of 74% and a sensitivity of 70% in approximately the same way as AUC2h insulin (cut-off 7000 mU/L/2 h; specificity 71%, sensitivity 76%).

Linear correlation analysis showed significantly positive correlations between T and metabolic parameters, namely BMI (R=0.32; P=0.0008), AC (R=0.31, P=0.001), insulin 0/1 h/2 h (Table 3), AUC2h insulin (R=0.30, P=.0016), HOMA-IR (R=0.33, P=0.0006), ISI (R=0.345, P=0.0003), glucose 2 h (R=0.32, P=0.001), AUC2h glucose (R=0.237, P=0.015) and triglycerides (R=0.33, P=0.0006), whereas DHEAS correlated significantly positively only with SHBG (R=0.336, P=0.0004). Significantly positive correlations were found between insulin and BMI, AC, systolic and diastolic BP, T, and triglycerides, whereas SHBG and HDL cholesterol correlated negatively (Table 3). The AC is in a highly significant way positively correlated with the BMI (R=0.8884, P<0.0001) which is a primary variable for FAS classification. Multivariate analysis showed that out of 7 variables with highest correlation coefficient, namely BMI, AC, systolic BP, T, SHBG, triglycerides and HDL-cholesterol, only AC, HDL-cholesterol and T could predict insulin levels to a significant extent (Table 4). The variability of insulin is explained to the greatest proportion by AC (between 21.7% and 31.5%).

  Insulin 0 Insulin 1 h Insulin 2 h
BMI 0.84 <.0001 0.533 <.0001 0.459 <.0001
Abdominal circumference 0.507 <.0001 0.561 <.0001 0.466 <.0001
Blood pressure systolic 0.288 .003 0.369 .0001 0.241 .0132
Blood pressure diastolic 0.269 .0055 0.288 .003 0.208 .033
Triglycerides 0.443 <.0001 0.395 <.0001 0.402 <.0001
HDL-Cholesterol -0.465 <.0001 -0.496 <.0001 -0.440 <.0001
SHBG -0.354 .0002 -0.435 <.0001 -0.331 .0006
Testosterone 0.317 .001 0.259 .0075 0.343 .0003
LH 0.008 .933 0.037 .705 0.024 .81
Anti-Müllerian Hormone 0.0563 .568 0.046 .64 0.0752 .446
DHEAS 0.008 .94 0.058 .556 0.058 .557

Table 3: Linear correlation analysis between insulin levels (fasting = 0, 1 h and 2 h) during oral glucose loading test and anthropometric, endocrinological and metabolic parameters. The table shows the correlation coefficient (in bold letters if significant) and the significance level P (italic letters). N=105.

  Insulin 0 Insulin 1h Insulin2h
Abdominal circumference P
Explains variance
.004
25.7%
<.0001
31.5%
.02
21.7%
HDL-Cholesterol P
Explains variance
.002
21.6%
.003
24.6%
.0024
19.3%
Testosterone P
Explains variance
.018
10.1%
.14
6.7%
.005
11.8%

Table 4: Multivariate analysis of endocrinological and metabolic variables for the dependency of insulin levels (fasting = 0, 1 h and 2 h after oral glucose loading test) on other variables (BMI, abdominal circumference, systolic blood pressure, sexual hormone binding globulin, testosterone, triglycerides and HDL-cholesterol). Only significant effects are shown. Stepwise backward elimination at the P=0.01-level. The determination of variance is shown in % (“explains variance”).

Concerning AMH, we did not found any correlation with metabolic parameters. However, a significantly positive correlation between AMH and T (R=0.335, P=0.0005) and LH (R=0.548, P<0.0001) was observed.

Discussion

The data reported here show for the first time that the 1h-OGLT is both an appropriate as well as a time- and cost-saving test procedure for the assessment of glucose and insulin dynamics in systematically defined androgenized female adolescents and women. These results go in line with former results of our group [19] and Moltz [13] suggesting that the OGLT including 1 h-glucose (cut-off: 140 mg/dL) and 1 h-insulin (cut-off: 99 mg/dL) values is a sufficient tool to diagnose an IR in apparently healthy normal weight or obese women in up to 90%. According to the RC [4], the American Diabetes Association [18] or the International Diabetes Federation [21], a 2 h-OGLT is recommended for assessing glucose metabolism. However, these data are based on investigations including individuals of both gender mainly in age groups >30 years, and the aim is to detect a diabetic state [18,21]. The individuals enrolled in the present study were all female and relatively young (mean age 25 y), and the predominant aim here was to discover discrete or moderate, nevertheless significant alterations of the glucose and insulin dynamics as early risk markers, since androgenized women have increased risks for gestational diabetes mellitus [20,22] and type 2 diabetes in later life [15,16].

In the FAS classification system presented, group FAS I comprises lean patients with ovarian hypertestosteronemia caused by EPOs without metabolic dysfunctions (Tables 1 and 2); this group corresponds in principle to the first RC group “PCOS without MetS” [4]. On the opposite, group FAS III is characterized by anthropometric, ovarian, pituitarian, fat-tissue-allocated, pancreatic and hepatic pathologies; the ovarian dysfunction which appears to be equivalent to that of FAS I [1,2] is merely a part of the overall pathogenesis resembling the second RC-group “PCOS with MetS” [4]. The majority of FAS IV patients depict a feature markedly overlapping with that of FAS III, however, the ovaries is in the normal range or even small and oligofollicular; consequently, circulating T shows normal levels [3]. In this group, cutaneous androgenization is caused indirectly by the suppression of hepatic SHBG secretion induced by hyperinsulinemia [23] and obesity which results in increased FAI levels. This sub-group seems to be alike to the “PCOS” sub-group “oligo-/amenorrhea and hyperandrogenism without PCO” [4] or “non-PCO PCOS”. This sub-group shows a high prevalence of metabolic pathologies [24,25] similarly to FAS IV. Th e majority of women belonging to group FAS III and IV are characterized by typical signs of IR suggesting that they have a high risk of endothelial dysfunction resulting in microor macro-angiopathy [9,10,13]. Using the RC definition [4], a MetS was diagnosed in approx. 40% of the FAS III and IV patients in line with an earlier review [6]. However, the limited RC definition of MetS omitting the determination of insulin [4] does not include women with “PCOS” who are mainly hyperinsulinemic, a fact that must be considered as a marked weakness of the RC [4,5]. Unfortunately, the classification system of the Androgen Excess Society suggesting 10 different phenotypes of “PCOS” based on the presence or absence of 4 main symptoms “hyperandrogenism”, “hirsutism”, “oligomenorrhea” and “polycystic ovaries” [26] did not really show any basic advantages vs. the RC [4] for identifying androgenized women with IR [27]. Missing the notice of a preexisting IR which is the basis of MetS and its consecutive disorders [9,10,13] may lead to a considerable failure at the assignment for e. g. preventive medical measures in these relatively young ladies. For example, FAS IV women who visited our clinic at a mean age of 20 y presented with hyperinsulinemia and signs of IR at a prevalence >50%, but none of them showed a pathological fasting glucose or glucose tolerance (Figure 1). Although the group FAS IV in this study is relatively small (further research is in progress), the data are extremely relevant for counseling these women.

In accordance with earlier studies, we observed significantly positive correlations between T and many metabolic parameters [19,28-30]. This suggests that the pathognomonic hyperandrogenemia of FAS may interact or even promote android fat distribution leading to an elevated AC followed by an augmentation of IR and finally MetS [25,31]. Enhanced serum insulin levels may stimulate ovarian T synthesis and decrease hepatic SHBG secretion resulting in a vicious cycle [19,28]. This underlines again the impact of insulin on the pathology of FAS III and IV. Considering AMH, we did not observe any correlations with metabolic parameters suggesting that AMH is a specific marker of ovarian function [2,30]. On the contrary to FAS III and IV, group FAS II reflects predominantly an adrenal hyperandrogenemia [1]. In consequence, DHEAS did not correlate with metabolic variables neither in the present investigation nor in earlier studies [31,32]. In accordance, we observed less metabolic pathologies in FAS II women than in FAS III or IV (Table 2, Figure 1) fitting to the assumption of Cussons et al. [6] that this group of patients “is different”. In the FAS classification algorithm, sub-group “b” represents miscellaneous constellations [1]. The majority of FAS IIb patients is characterized by the combination of adrenal hyperandrogenemia and metabolic pathologies which constitutes the sub-group which is finally responsible on both IR and metabolic dysfunctions in FAS II.

Central obesity is the most important independent risk factor for CVD [21,33]. In the present study, AC was the strongest predictor of insulin variance (Table 4) followed by HDL-cholesterol and T. There was a high prevalence (>50%) of metabolic dysfunctions in FAS III and FAS IV. A long-term follow-up showed an increased incidence of MetS [34], coronary heart disease and stroke [35] in women with PCOS compared with controls. Furthermore, it would be of great clinical relevance to predict the risk for the development of type 2 diabetes in later life. A recent study reported that glucose 1 h was a strong predictor for type 2 diabetes in men and women (mean age 47 years), whereas MetS alone was found to be less reliably predictive [36]. In the present study, between 4 and 14% of FAS III women showed IFG or IGT (Figure 1). This observation was described before in “PCOS” patients [6,37]. Regarding glucose 1 h, however, 19-38% of women showed levels >140 mg/dL [13,19] (Figure 1). The assessment of insulin adds the information of pathologically elevated insulin levels as an early response to restore serum glucose concentrations to physiological ranges, and is therefore more sensitive [10,13,38]. Hyperinsulinemia precedes overt type 2 diabetes mellitus [13,38] and constitutes an early signal for impaired glucose metabolism corresponding in principle the term “IR”.

To date, it is still unclear which individual marker apart from patch-clamp analysis is best reflecting IR. HOMA-IR derives from a mathematical model describing glucose and insulin metabolism, but includes only fasting levels of insulin and glucose which may be in the normal range, especially in young women [39,40]. These women representing the majority of patients in gynaecological endocrinology and reproductive medicine exhibited individually a pathological metabolic response during OGLT as shown in the present study and former investigations [13,40]. The calculation of AUC for insulin or glucose shows the concentration changes of both variables over the time of testing (mainly 2 or 3 h) and is supposed to mirror better the dynamics of metabolism than single values [14]. The ISI exhibited very good correlations with patch-clamp results and includes insulin and glucose over time reflecting liver and peripheral organ glucose metabolism [12]. Since ISI and AUC2h insulin are very closely correlated, the determination of one of these indices seems to be sufficient (Figure 2). We could also show a close relation between insulin 1 h and ISI and AUC2h insulin, respectively (Figure 2). Insulin 1 h in combination with insulin 0 showed highly significant positive (+) or negative (-) correlations with AC (+), blood pressure (+), SHBG (-), T (+), triglycerides (+) and HDL-cholesterol (-) (Table 3). Insulin 2 h did not add any exceeding information in this regard. Furthermore, insulin 0/1 h levels in combination with glucose 1 h showed the highest individual presence of pathologic values concerning glucose/insulin dynamics (Figure 1). We therefore suggest that an 1 h OGLT including the analysis of both glucose and insulin at the two time points (fasting, 1 h) is a reliable diagnostic tool concerning the individual glucose and insulin metabolism in all androgenized women in accordance with recent studies [22,37].

In consequence, these data support the systematic FA classification algorithm comprising metabolic parameters, especially insulin as a primary variable, in women who are adversely affected by functional androgenization [1,2]. Th e OGLT restricted to 1 h values and including the analysis of not only glucose but also insulin is in line with recent studies [41,42] and appears to be the most reliable test approach being advantageous notably under the consideration of time and cost consumption. This conclusion is particularly important under the aspect of early preventive measures in this special cohort of female individuals at fertile age.

Acknowledgements

We express our thanks to Marina Tetyusheva (Dr. Köhler GmbH Freiburg/ Germany) for her great support regarding the statistical analyses. In addition, the financial support by MSD SHARP & DOHME GMBH, Haar, Germany, is greatly appreciated.

References

Select your language of interest to view the total content in your interested language
Post your comment

Share This Article

Relevant Topics

Recommended Conferences

Article Usage

  • Total views: 11613
  • [From(publication date):
    June-2013 - Aug 24, 2017]
  • Breakdown by view type
  • HTML page views : 7840
  • PDF downloads :3773
 

Post your comment

captcha   Reload  Can't read the image? click here to refresh

Peer Reviewed Journals
 
Make the best use of Scientific Research and information from our 700 + peer reviewed, Open Access Journals
International Conferences 2017-18
 
Meet Inspiring Speakers and Experts at our 3000+ Global Annual Meetings

Contact Us

Agri, Food, Aqua and Veterinary Science Journals

Dr. Krish

[email protected]

1-702-714-7001 Extn: 9040

Clinical and Biochemistry Journals

Datta A

[email protected]

1-702-714-7001Extn: 9037

Business & Management Journals

Ronald

[email protected]

1-702-714-7001Extn: 9042

Chemical Engineering and Chemistry Journals

Gabriel Shaw

[email protected]

1-702-714-7001 Extn: 9040

Earth & Environmental Sciences

Katie Wilson

[email protected]

1-702-714-7001Extn: 9042

Engineering Journals

James Franklin

[email protected]

1-702-714-7001Extn: 9042

General Science and Health care Journals

Andrea Jason

[email protected]

1-702-714-7001Extn: 9043

Genetics and Molecular Biology Journals

Anna Melissa

[email protected]

1-702-714-7001 Extn: 9006

Immunology & Microbiology Journals

David Gorantl

[email protected]

1-702-714-7001Extn: 9014

Informatics Journals

Stephanie Skinner

[email protected]

1-702-714-7001Extn: 9039

Material Sciences Journals

Rachle Green

[email protected]

1-702-714-7001Extn: 9039

Mathematics and Physics Journals

Jim Willison

[email protected]

1-702-714-7001 Extn: 9042

Medical Journals

Nimmi Anna

[email protected]

1-702-714-7001 Extn: 9038

Neuroscience & Psychology Journals

Nathan T

[email protected]

1-702-714-7001Extn: 9041

Pharmaceutical Sciences Journals

John Behannon

[email protected]

1-702-714-7001Extn: 9007

Social & Political Science Journals

Steve Harry

[email protected]

1-702-714-7001 Extn: 9042

 
© 2008-2017 OMICS International - Open Access Publisher. Best viewed in Mozilla Firefox | Google Chrome | Above IE 7.0 version
adwords