J Infertil Reprod Biol, 2020, Volume 8, Issue 4, Pages: 73-83. https://doi.org/10.47277/JIRB/8(4)/73  
Tetrad of Hormonal and Biochemical Manifestations  
in Phenotypes of Polycystic Ovary Syndrome  
*
Sonal Agarwal , Deepika Krishna, Gautham Praneesh, Kamini Rao  
Milann Fertility Centre, Bengaluru, India  
Received: 21/08/2020  
Accepted: 17/10/2020  
Published: 20/12/2020  
Abstract  
A prospective observational case- control study was conducted to evaluate the relationship and prevalence of subclinical hypothyroidism,  
hyperprolactenemia, impaired glucose metabolism and hyperhomocysteinemia as components of hormonal/biochemical manifestation tetrad  
amongst four phenotypes of polycystic ovary syndrome (PCOS) with respect to controls. 200 women diagnosed as PCOS as per the  
ESHRE/ASRM Rotterdam criteria were taken as cases (group I) and 200 as controls (group II).All women recruited came for subfertility  
treatment to assure homogeneity in study population. Group I was further divided into four phenotypes.Records of demographic (age, BMI,  
duration of infertility), biochemical(homocysteine,fasting blood sugar,HBA1C)and hormonal (thyroid profile and prolactin)parameters of  
patients were taken. Both groups were comparable in age, ethnicity and marital status. Percentages of phenotypes A,B,C and D in our Indian  
population were 25.5%,16%,35.5% and 23 % respectively. BMI was significantly more in PCOS as compared to non-PCOS with highest  
mean in phenotype B subgroup. All parameters measured were significantly increased in PCOS group compared with non-PCOS group with  
prolactin levels similar in all phenotypes. Impaired glucose tolerance and diabetes incidence was more in phenotype C and D. Hypothyroidism  
and hyperhomocystenemia were higher in phenotypes B and C. Phenotypes are affected by hormonal and biochemical manifestations and if  
followed for long term may be associated with metabolic syndrome in future. Thus non classical phenotypes should also be properly  
monitored and treated as this hormonal imbalance and biochemical derangements add to the brunt of management of PCOS.  
Keywords: Hyperhomocysteinemia , Hyperprolactinemia, Hypothyroidism, Impaired glucose metabolism  
Introduction  
a multidisciplinary approach to benefit the patient in a holistic  
manner depending on evidence based medicine practice(7).We  
have evaluated prevalence of various phenotypes and scrutinized  
their hormonal and biochemical characteristics compared to non-  
PCOS women presenting to our center.  
Polycystic ovarian syndrome (PCOS) is the most common  
multisystem endocrinopathy with epigenetic and environmental  
influences affecting early pre-pubertal age to post-menopausal  
age group with varied presentations in each(1). The variability in  
etiopathogenesis and clinical presentation has made the  
appropriate diagnosis of PCOS difficult (2). This endocrine-  
metabolic disorder affects 5%10% of women of reproductive  
age with oligo-anovulation, menstrual disturbances, and/or  
androgen excess as the main features (3).It has been affirmed by  
various studies that hypothyroidism is also a state of insulin  
resistance which is the crux of genesis of PCOS. Individually both  
hypothyroidism and PCOS have changes in lipid metabolism  
For all clinical manifestations, the root cause is  
hyperandrogenemia and impaired insulin response (8).  
Hypothyroidism leads to metabolic derangements such as  
decrease in sex-hormone binding globulin causing increased  
levels of free androgens in blood and decreased metabolic  
clearance, defective glucose disposal and hyperlipidemia, thus  
affecting gonadal function and fertility (9, 10). Increased TSH  
levels along with hyperprolactenemia lead to deposition of  
mucopolysaccharides in various organs such as ovaries affecting  
its function resulting in anovulation and reproductive disruption.  
Beta cell dysfunction and insulin resistance can lead to  
development of type 2 diabetes mellitus. Insulin resistance may  
lead to increase in homocysteine levels due to altered methionine  
metabolism (11). High sensitivity C-reactive protein, adiponectin  
and homocysteine are serum biomarkers of cardiovascular  
disease and are found to be abnormal in women with PCOS.  
There is clustering of obesity, impaired glucose tolerance,  
dyslipidemia and hypertension as risk factors (12).  
(
increase in total cholesterol and in low-density lipoprotein  
cholesterol) and endothelial function leading to increased risk of  
arterial hypertension and other cardiovascular problems(4).  
Association of thyroid dysfunction and other hormonal and  
biochemical parameters of PCOS are thus frequently studied to  
establish a relationship between the two.  
PCOS is a constellation of symptoms and signs having  
significant implications on woman’s health and fertility (5).  
Phenotypes are influenced by genetic, racial, geographic,  
environmental and lifestyle factors like diet and over-  
nutrition.This affects them causing more adverse effects in  
respect to reproductive and metabolic outcome (6). PCOS is  
characterized by vast diversity in manifestations, necessitating a  
Measurement of hormonal and biochemical parameters in this  
study will help in providing early interventions in women with  
*
Corresponding author: Sonal Agarwall, Milann Fertility Centre, Bengaluru, India. Email: sonaljaipur28@gmail.com  
7
3
J Infertil Reprod Biol, 2020, Volume 8, Issue 4, Pages: 73-83. https://doi.org/10.47277/JIRB/8(4)/73  
PCOS to prevent progression from pre-diabetic state to full  
blown diabetes and other risk factors.  
significant drug intake suspected to affect metabolic function  
were not included in the study.Women with any renal, hepatic,  
thyroid or cardiac dysfunction,Cushing’s syndrome, congenital  
adrenal hyperplasia and adrenal tumors were excluded. All  
women had to undergo detailed anthropometric assessment in  
form of measurement of height,weight,waist and hip  
circumference,blood pressure and systemic examination .Body  
mass index (BMI) was calculated by the formula: weight in  
kilograms/(height in meters)2 .Hirsutism was assessed using the  
modified Ferriman-Gallwey score, with nine specified body areas  
counted by a single observer.A score of more than/equal to 8 out  
of 36 was considered significant. Biochemical hyperandrogenism  
was defined as a serum total T level of >0.022 nmol/L. Trans-  
vaginal ultrasonography was done by a single clinician to rule out  
any inter-observer variation to see presence of more than 12  
ovarian follicles 2-9 mm in any ovary and increased ovarian  
volume > 10 cc(calculated using the formula 0.5 × length × width  
Materials and methods  
Study design and participants  
This prospective observational case-control study was  
conducted at Milann fertility center which is a tertiary care center.  
Study was conducted from April 2017 to January 2018.Women  
were divided into two groups according to European Society of  
Human Reproduction & Embryology/American Society for  
Reproductive Medicine(ESHRE/ASRM) Rotterdam criteria into  
cases and controls. Rotterdam criteria for the diagnosis of PCOS  
includes presence of any two of the following criteria:clinical or  
biochemical hyperandrogenemia, anovulation/Oligomenorrhea  
and polycystic ovaries on ultrasonography(13). They were  
recruited at the time of diagnosis and who have not yet initiated  
treatment for any biochemical and hormonal derrangements. All  
participants before allocation signed an informed consent form.  
Approval was obtained from the Institutional Ethical Committee.  
Women of group I were divided into four phenotypes -  
hyperandrogenism (HA) + oligo-/anovulation (OA) + polycystic  
ovaries at ultrasound (PCO) (phenotype A,full-blown syndrome);  
HA + OA (phenotype B, former National Institutes of Health  
definition); HA + PCO (phenotype C) and OA + PCO (phenotype  
D). Four different combinations of phenotypes have been  
deciphered by Rotterdam/AEPCOS society based on clinical and  
endocrinological findings (14).Consequently records of  
demographic, hormonal and biochemical parameters of patients  
were taken.  
×thickness)suggestive  
of  
PCOM  
(polycystic  
ovary  
morphology).Women were divided into 2 groups based on  
ESHRE/ASRM Rotterdam criteria with group I as PCOS  
women(cases) and group II as non-PCOS women(controls).  
Thyroid stimulating hormone (TSH), free thyroxine (FT4),  
glycosylated hemoglobin (HBA1C), fasting blood glucose (FBS),  
prolactin (PRL) and plasma homocysteine (PH) were measured  
in all women recruited in the study. The blood samples were  
obtained from peripheral vein in early morning after a fasting  
period of atleast 8 hours. TSH, FT4 and PRL and total T levels  
were measured by electrochemiluminescence assay (Cobas e411-  
Hitachi).Plasma homocysteine and fasting blood sugar were  
measured by enzymatic method through Cobas c311.HBA1C was  
measured through BioRad DIO method. Intra- and inter assay  
variations were within the limits permitted by manufacturer  
company.  
Subclinical hypothyroidism is defined as serum TSH levels  
of >2.5mIU/L with no clinical symptoms and signs having normal  
FT4 levels according to ASRM guidelines. In our study a TSH  
cutoff of 2.5 IU/L was taken to define subclinical hypothyroidism  
to determine association of increased TSH levels in all four  
phenotypes of PCOS and also in controls. Plasma homocysteine  
cut off of 8umol/l was taken for hyperhomocystenemia.HBA1C  
Patient population  
Inclusion criteria:  
All PCOS defined as per ESHRE/ASRM Rotterdam criteria  
of 2003 for group I and non-PCOS in group II, aged 21-35 years,  
BMI >18 and <30 kg/m2, no symptoms of hypothyroidism and  
free thyroxine levels (t4) in range of 4.6-12ug/dl, proven  
subfertility and willingness to participate in study.  
Exclusion criteria:  
Patients with chronicdiseases like overt hypothyroidism and  
hyperthyroidism, kidney or liver failure, late-onset adrenal  
hyperplasia and diabetes, severe endometriosis, severe male  
factor infertility, multiple fibroids, previous IVF treatment failure  
and those not willing to participate.  
>5.3%, FBS >100 mg/dl and PRL >30 ng/dl cut off were taken  
for deranged blood sugar levels (marker for insulin resistance)  
and hyperprolactenemia respectively. The results were described  
as mean ± standard deviation. Significance level was defined at  
5
%, and the software used for the analysis was the SAS statistical  
Recruitment and analysis  
software package, version 9.1.  
All patients who came to our center who were unable to  
conceive were interviewed in detail. History taking comprised of  
complaints,past treatment history,menstrual history related to age  
of menarche, regularity, duration, and number of cycles per  
year,marital history,contraceptive and obstetric history.Questions  
regarding past medical, surgical history and family history were  
also asked.Written informed consent was taken from all patients  
recruited for the study.  
Participant flow  
The participant flow is shown in Figure I. Out of 221 PCOS  
patients screened for study, only 200 followed up with blood tests  
and thus were enrolled for study. Out of 214 NON-PCOS women  
only 200 followed up with blood tests whowere enrolled for  
study. Thus 200 women with PCOS in group I and 200 non-PCOS  
women were included in group II for final analysis.  
Oligomenorrhea was defined as an inter-menstrual interval of  
more than 35 days or a total of eight or fewer menstrual cycles  
per year. Duration and extent of abnormal hair growth, weight  
gain, and development of acne or alopecia, along with family  
history of hirsutism, menstrual disorders, and diabetes mellitus or  
glucose intolerance was written in record file. Women giving  
history of any use of hormonal preparation, androgens or any  
Statistical analysis  
Descriptive statistics were presented as means and standard  
deviation for continuous variables .Frequencies and proportions  
were used for categorical variables. Independent sample t test was  
used for continuous variables which were normally distributed  
7
4
J Infertil Reprod Biol, 2020, Volume 8, Issue 4, Pages: 73-83. https://doi.org/10.47277/JIRB/8(4)/73  
and Mann-Whitney U test for data not normally distributed. Chi-  
and PH as associative factors with PCOS infertile women.  
Histograms have been created to show association between  
different variables amongst cases and controls.All tests were two-  
sided with p-value of less than 0.05 considered as statistically  
significant.  
square test or Fisher’s exact test was used for categorical variables  
where appropriate. Odds ratio (OR) with 95% confidence  
intervals (CIs) was calculated. For post-hoc analysis one way  
ANOVA test was used.  
In addition, receiver operating  
characteristic analysis was used to evaluate TSH, HBA1C, PRL  
Figure 1. Participant flow chart  
Results  
to 5.605 ±2.99 in controls (group II). Table 3 illustrates various  
hormonal and biochemical parameters such as plasma  
homocysteine, AMH, TSH, PRL, FBS and HBA1C amongst  
PCOS group and non-PCOS group. In the PCOS group, all  
hormonal and biochemical parameters were observed to be higher  
as compared to NON-PCOS group and were statistically  
significant.  
Both groups were divided into two categories based on their  
HBA1C and FBS levels into pre-diabetic state and type 2 diabetes  
mellitus.  
Demographic profile of PCOS and non-PCOS women was  
compared and different phenotypes of PCOS women were also  
compared through sub-group analysis as illustrated in table 1 and  
2
.In our study in PCOS group maximum prevalence was of  
phenotype C (35.5%) followed by phenotype A (25.5%),  
phenotype D (23%) and least for phenotype B (16%). Mean age  
of women of phenotype A, B, C and D was 29.5 ±3.7, 28.9±3.7,  
2
9.2±3.8 and 29.4±3.3 years respectively. Mean age of group II  
women was 30.1±4.1 years. Age was similar in both groups and  
sub-groups. Women with PCOS had a mean BMI of 25.9±4.3  
Kg/m2 and non-PCOS women had BMI of 24.5±3.6 kg/m2.  
Higher BMI was observed in all phenotypes of PCOS with respect  
to controls (p=0.001).Obese and overweight women were seen  
maximum in phenotype B, followed by phenotype D, A and C  
and were least in controls. Both groups are similar regarding  
number of women presenting with primary or secondary  
infertility. Duration of infertility was noticed to be significantly  
more in cases (group I) with mean of 6.065 ±3.045 as compared  
CATEGORY 1 (PRE-DIABETES): HBA1C 6-6.4% AND  
FBS 100-125 MG/DL  
CATEGORY 2 (TYPE 2 DIABETES MELLITUS): HBA1C  
>=6.45 AND FBS >=126 MG/DL  
Population histogram was created for each category for both  
groups.Our study has deciphered prevalence of IGT as 17.5% and  
type 2 DM as 2% in PCOS women and 5% IGT and 1.5% T2DM  
prevalence in non-PCOS population as depicted in table 4 which  
7
5
J Infertil Reprod Biol, 2020, Volume 8, Issue 4, Pages: 73-83. https://doi.org/10.47277/JIRB/8(4)/73  
is quite similar to previous studies done across the world.  
different phenotypes of PCOS and histogram was formed for sub-  
group analysis and main group analysis both. 55 % (n=105)  
women in group I had TSH>= 2.5 m IU/L and only 39.5% (n=79)  
of group II had TSH>= 2.5 m IU/L. There was statistically  
significant difference between the two groups (P=0.041).Our  
study showed higher prevalence of SCH in women with PCOS  
compared with non-PCOS women. Bar graph with standard error  
of mean in figure 3 (C) depicts that TSH levels are highest in  
women with phenotype B followed by phenotype C and D and  
least in phenotype A. 6.5%(n=13) women of group I and  
5.5%(n=11) women of group II had prolactin levels of more  
than/equal to 30 ng/ml. There was a statistically significant  
association of PRL with PCOS (p<0.001). It is portrayed in figure  
3 (B) that all four phenotypes of PCOS have similar levels of  
prolactin. 68 % (n=136) of PCOS women and only 49 %(n-98) of  
NON-PCOS group of women had plasma homocysteine levels of  
more than/equal to 8. The mean serum homocysteine  
concentration was significantly higher in the PCOS group  
(11.45±6.563 mmol/l) versus in NON-PCOS group (9.72±7.404  
mmol/L) with p value <0.001. Our study has outlined higher  
homocysteine levels in phenotype C followed by phenotype B.  
Phenotype A and D has similar prevalence with similar  
homocysteine levels as illustrated in figure 3 (D).  
Association of pre-diabetes and T2DM was calculated with  
homocysteine and prolactin levels. High HBA1C and increased  
fasting glucose level is more prevalent in PCOS women with  
hyperhomocystenemia.This shows association between these  
biochemical factors in PCOS resulting in the acquisition of the  
endocrinological tetrad. There is higher prevalence of  
hyperprolactenemia in PCOS group as compared to non-PCOS  
group as illustrated in population histogram in figure 2. 69.5%  
(
n=139) of PCOS women had HBA1C levels of more than /equal  
to 5.3 % as compared to 44.5% (n=89) of non-PCOS women.  
With regard to glucose metabolism, mean HBA1C values were  
found to be higher in women with PCOS with statistically  
significant difference (p<0.001). Highest HBA1C levels was seen  
in phenotype D,then phenotype C and B followed by least in  
phenotype A as depicted in figure 2 (A).Fasting blood sugar was  
found highest in phenotype D, then phenotype C and B with  
similar levels and least in phenotype A as shown in figure 2 (B).  
Measurement of anti-mullerian hormone (AMH) was also  
assessed in different PCOS women as depicted in figure 2 (C)  
which shows higher range in phenotype B, then phenotype D and  
A, followed by phenotype C with least level.  
We studied association of subclinical hypothyroidism in  
Table 1: demographic data of PCOS and non-PCOS women  
Variable  
Age (years)  
Group I (PCOS women)  
29.3±3.6  
Group II (non-PCOS women)  
30.1±4.1  
P value  
0.025  
BMI(kg/m2)  
25.9±4.3  
24.5±3.6  
0.001  
Duration of infertility (years)  
Primary infertility(n)  
Secondary infertility(n)  
Data are expresses as mean± SD  
6.065±3.045  
144(72%)  
56(28%)  
5.605±2.99  
138(69%)  
62(31%)  
-
-
-
Table 2: demographic data of different phenotypes of PCOS women (group I)  
Variable  
Age (years)  
BMI(kg/m2)  
Duration of infertility (years)  
Data are expresses as mean± SD  
Phenotype A  
29.5±3.7  
25.4±3.9  
6.5±3.5  
Phenotype B  
28.9±3.7  
27.3±4.8  
5.9±3.3  
Phenotype C  
29.2±3.8  
24.9±3.5  
5.6±2.7  
Phenotype D  
29.4±3.3  
26.4±4.4  
6.3±2.8  
Table 3: hormonal and biochemical parameters  
Variable  
Plasma homocysteine  
AMH  
TSH  
PRL  
HBA1C  
FBS  
Group I (PCOS women)  
Group II (non-PCOS women)  
9.72±7.404  
P value  
<0.001  
<0.001  
0.041  
<0.001  
<0.001  
<0.001  
11.45±6.563  
6.41±3.699  
2.74± 1.496  
17.49±6.723  
5.50±0.628  
87.78±19.02  
2.51±2.270  
2.57±1.580  
14.59±7.794  
5.31±0.458  
77.40±13.61  
Data are expresses as mean± SD  
Table 4: prevalence of pre-diabetes and type 2 diabetes melllitus in PCOS and non-PCOS women  
Variable  
Pre0diabetes  
Type 2 diabetes mellitus  
Group I (PCOS women)  
35(17.5%)  
Group II (non-PCOS women)  
10(5%)  
3(1.5%)  
4(2%)  
7
6