What hormonal investigations are indicated for a reproductive‑age woman with suspected polycystic ovary syndrome (PCOS)?

Hormonal Investigation in PCOS

First-Line Essential Hormonal Tests

Measure total testosterone and free testosterone using liquid chromatography-tandem mass spectrometry (LC-MS/MS) as your primary androgen assessment, as these provide the best diagnostic accuracy with total testosterone showing 74% sensitivity and 86% specificity, while free testosterone demonstrates superior sensitivity of 89% with 83% specificity. 1, 2, 3

• LC-MS/MS is mandatory as the preferred assay method, showing superior specificity (92%) compared to direct immunoassays (78%), which have unacceptably high false-positive rates 1, 2, 3
• If LC-MS/MS is unavailable, calculate Free Androgen Index (FAI) using total testosterone and SHBG as an alternative 2
• Important caveat: Total testosterone is abnormal in only 70% of women with confirmed PCOS, meaning 30% have normal testosterone levels despite having the condition 2, 3

Measure thyroid-stimulating hormone (TSH) to exclude thyroid disease as a cause of menstrual irregularity. 1, 2, 3

Measure prolactin using morning resting serum levels to exclude hyperprolactinemia, with levels >20 μg/L considered abnormal. 1, 2, 3

• Confirm prolactin elevation with 2-3 samples at 20-60 minute intervals via indwelling cannula to exclude stress-related spurious elevation 2

Second-Line Hormonal Tests (When Clinical Suspicion Remains High Despite Normal Testosterone)

If total testosterone and free testosterone are normal but clinical suspicion remains high, measure androstenedione (A4) and DHEAS as adjunctive tests. 2, 3

• A4 shows sensitivity of 75% and specificity of 71% 2
• DHEAS shows sensitivity of 75% and specificity of 67% 2
• These have poorer specificity than testosterone and should only be used as second-line tests 2
• Androstenedione >10.0 nmol/L suggests adrenal/ovarian tumor 1, 2
• Elevated DHEAS (age 20-29: >3800 ng/ml; age 30-39: >2700 ng/ml) requires ruling out non-classical congenital adrenal hyperplasia 1, 2, 3

Additional Hormonal Assessment for Ovulatory Function

Measure LH and FSH between days 3-6 of the menstrual cycle (if cycles occur), calculating based on an average of three estimations taken 20 minutes apart. 1, 2, 3

• An LH/FSH ratio >2 suggests PCOS, though this is abnormal in only 35-44% of women with PCOS, making it a poor standalone diagnostic marker 2
• FSH levels are typically normal or low-normal in PCOS patients 2

Measure mid-luteal phase progesterone (blood taken during mid-luteal phase according to menstrual cycle) to confirm anovulation, with levels <6 nmol/L indicating anovulation. 1, 2, 3

Mandatory Metabolic Screening (Part of Comprehensive PCOS Evaluation)

Perform a two-hour oral glucose tolerance test with 75g glucose load to screen for glucose intolerance and type 2 diabetes, as PCOS women have increased risk for these conditions. 2, 3

• Fasting glucose >7.8 mmol/L is suggestive of diabetes 1
• Measure fasting glucose and insulin levels, with glucose/insulin ratio >4 suggesting reduced insulin sensitivity 1, 2, 3

Order a fasting lipid panel including total cholesterol, LDL, HDL, and triglycerides to assess dyslipidemia. 2, 3

• Target LDL <100 mg/dL, HDL >35 mg/dL, and triglycerides <150 mg/dL 2

Calculate BMI and measure waist-hip ratio (WHR), with WHR >0.9 indicating truncal obesity and increased metabolic risk. 1, 2, 3

Critical Diagnostic Considerations and Pitfalls

If the patient has both irregular menstrual cycles AND clinical/biochemical hyperandrogenism, ultrasound is not necessary for PCOS diagnosis, though it will identify the complete PCOS phenotype. 2

The presence of clinical hyperandrogenism (hirsutism, acne, or alopecia) plus irregular menstrual cycles can establish PCOS diagnosis even with normal laboratory values, as the Rotterdam criteria require only two of three features. 2, 3

Do not use AMH levels as an alternative for detecting polycystic ovarian morphology or as a single test for PCOS diagnosis due to lack of standardization, no validated cut-offs, and significant overlap between women with and without PCOS. 1, 2, 3

• While AMH ≥35 pmol/L shows 92% sensitivity and 97% specificity, it requires further validation in different populations 1, 2

If the patient is on hormonal contraception, hormone testing is unreliable and should be deferred until after removal or expiration of the contraceptive. 3

Exclusion of Mimicking Conditions

Screen for Cushing's syndrome if patient has buffalo hump, moon facies, hypertension, abdominal striae, central fat distribution, easy bruising, or proximal myopathies. 2, 3

Consider androgen-secreting tumors if rapid onset of symptoms, severe hirsutism, or very high testosterone levels. 2, 3

If DHEAS is elevated, consider non-classical congenital adrenal hyperplasia. 1, 2, 3

Special Considerations for Adolescents

In adolescents, PCOS diagnosis requires hyperandrogenism (clinical or biochemical) in the presence of persistent oligomenorrhea lasting 2-3 years beyond menarche. 3, 4

Do not use ultrasound for PCOS diagnosis in adolescents with gynecological age <8 years post-menarche due to high false-positive rates from normal multifollicular ovaries. 2, 3, 4

• Large, multicystic ovaries are a common finding in adolescents, so ultrasound is not a first-line investigation in women <17 years of age 4