Following the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines, this study aimed to examine the prevalence and variations of insulin resistance among different PCOS phenotypes.

Study method and participants

This cross-sectional study was conducted between November 2020 and June 2021, focusing on 160 women aged 18 to 45 years diagnosed with polycystic ovary syndrome (PCOS). The participants were recruited from gynecology clinics in Urmia, located in northwestern Iran. The sampling method used was convenient sampling, enrolling all eligible women until the desired sample size was achieved.

To be included in the study, participants had to meet the following criteria: be between 18 and 45 years of age, have a confirmed diagnosis of PCOS by a gynecologist based on clinical, laboratory, and imaging findings, not be pregnant at the time of enrollment, not be undergoing infertility treatment or taking hormonal medications, and not have taken any medications other than over-the-counter (OTC) painkillers in the last 3 months. Additionally, the interval between the onset of menarche and the study enrollment had to be more than 4 years, and participants should not have had severe underlying diseases like malignancy or thalassemia that could impact menstrual cycles. Furthermore, known endocrinopathies like Cushing’s syndrome, untreated thyroid disorders, and similar conditions were also considered as exclusion criteria.

Study procedures

All participants underwent an evaluation using a researcher-developed checklist that encompassed demographic, reproductive, and medical characteristics. The checklist covered a range of factors, including age, level of education, occupation, marital status, number of pregnancies, number of abortions, number of children, age of menarche, economic status, physical activity, diet (both general and specific, such as vegetarianism), and medical history. Following enrollment in the study, participants underwent relevant clinical examinations, paraclinical tests, and ultrasounds as part of the evaluation process.

Anthropometric measurements were conducted following a standard protocol and utilizing calibrated tools. Height was measured without shoes against a fixed strip attached to the wall. Weight was measured using a Seka 755 scale with an accuracy of 500 g, and participants wore light clothing and no shoes during the measurement. BMI was calculated as follows: weight (kg) / height^2 (m).

Waist circumference was determined by using a standard measuring tape aligned parallel to the umbilicus, while hip circumference was measured with a meter at the widest part of the hip region. The waist-to-hip ratio (WHR) was then computed by dividing the waist circumference by the hip circumference.

Clinical indications of hyperandrogenism, such as acne, oily skin, and hirsutism, were observed and assessed. A Gynecologist evaluated all clinical findings, including the Ferriman Gallwey score and galactorrhea. For the diagnosis of hirsutism, a thorough evaluation was conducted by taking a medical history and performing a clinical examination, utilizing the Freeman-Galloway rating score to examine nine body areas for the presence of coarse terminal hairs (including the upper lip, chin, chest, upper and lower abdomen, upper arms, and thighs). The severity of hirsutism was rated on a scale of 1 to 4 for each area, and the total scores were summed. Women with a total score equal to or above 8 were considered to have hirsutism [24].

Participants underwent fasting for 10 to 12 hours before venous blood samples were collected to measure the following parameters: Fasting Blood Sugar (FBS), Fasting Insulin Levels (FIL), Total Serum Testosterone, Thyroid Stimulating Hormone (TSH), Thyroxine (T4), Triiodothyronine (T3), Malonaldehyde (MDA), Sex Hormone-Binding Globulin (SHBG), Follicle-Stimulating Hormone (FSH), Luteinizing Hormone (LH), Hemoglobin (Hb), and Platelets (PLT).

FSH, LH, testosterone, insulin and FBS levels were measured by Abnova commercial kit mIU / ml, Cat.N.DE1288 and GmbH, Germany Demeditec Diagnostics, according to the instructions. The levels of MDA in the follicular fluid were assessed using the Thiobarbituric Acid (TBA) method and the TBARS kit (KA1381) manufactured by Abnova, Taiwan.

Study variablesPolycystic ovary syndrome (PCOS)

PCOS was defined according to Rotterdam criteria, by the presence of 2 of these findings: oligo or anovulation, clinical or biochemical hyperandrogenism, and polycystic ovaries (PCO).

Insulin resistance (IR)

The assessment was performed using the homeostatic-insulin resistance (HOMA-IR) formula, which is calculated as follows: fasting insulin (mg/dL) × fasting blood glucose / 405 (μU/mL) [25].

HOMA-IR > 2.5 suggested insulin resistance.

Body mass index (BMI)

It was determined using the body mass index (BMI) formula, which involves dividing weight by height squared [26].

Polycystic ovaries (PCO)

It was characterized as the presence of 10 or more immature follicles in each ovary and/or an ovarian volume exceeding 10 cm3 on ultrasound [27].

Menstrual disorders/ ovulatory dysfunction (OD)

Included menstrual disorders encompassed amenorrhea, oligomenorrhea, hypomenorrhea, hypermenorrhea, and irregular menstrual intervals, which were determined based on the participants’ medical history. Oligomenorrhea, specifically, was diagnosed when menstrual cycles occurred more than 35 days apart or less than nine times a year [27].

Hyperandrogenism

The definition was established by considering the serum levels of male hormones, including Total Serum Testosterone, SHBG, and FAI, along with the presence of clinical signs such as acne, oily skin, hirsutism, and male pattern hair loss. FAI was determined using the formula: [total testosterone] / SHBG × 100 [28].

PCOS phenotypes

In the study, PCOS phenotypes were classified among the participants through an assessment of their medical history, clinical examination, and paraclinical tests, utilizing the criteria of Hyperandrogenism (H), Ovulatory Dysfunction (OD), and Polycystic Ovaries (PCO) as follows: Phenotype A: OD + PCO + H; Phenotype B: OD + H; Phenotype C: H + PCO; Phenotype D: OD + PCO [6].

Data management and analysis

Our model selection strategy aimed to strike a balance between predictive accuracy and model simplicity. We followed a stepwise approach, beginning with a comprehensive model that included all potentially relevant variables. The data were input into the computer and analyzed using IBM® SPSS® software, version 26. A significance level of 0.05 was set for statistical analyses. Categorical variables were compared using the Chi2 test, while quantitative variables in two groups were analyzed using the Student’s t-test or Mann-Whitney U analysis, depending on the normality of the data distribution. For quantitative variables involving more than two groups, the analysis was conducted using the Anova Test or Kruskal-Wallis analysis, depending on the normality of the variables’ distribution. Linear regression analysis was employed to identify factors influencing HOMA-IR.