PCOS and your blood test: the markers that matter.

PCOS is a blood test diagnosis

Despite its name, polycystic ovary syndrome is not primarily an ovarian condition. It is a systemic hormonal and metabolic disorder with effects that span reproductive, metabolic, cardiovascular, and psychological health. And the most important tool for diagnosis and management is not an ultrasound — it is your blood work.

The Rotterdam criteria, established in 2003 and still the international diagnostic standard, require at least two of three features: oligo-ovulation or anovulation (irregular or absent periods), clinical or biochemical hyperandrogenism (elevated androgens), and polycystic ovarian morphology on ultrasound[1]. Two of these three can be assessed through blood tests alone.

Yet many women with PCOS report being told their blood work is "normal" — often because the wrong markers were tested, or the right markers were interpreted against reference ranges that are too broad. Here is the complete panel, and how to read it.

Androgens: the diagnostic cornerstone

Total testosterone

Total testosterone is the first-line biochemical marker for hyperandrogenism in PCOS. In women, normal total testosterone typically ranges from 15-70 ng/dL, though reference ranges vary by lab and assay. In PCOS, total testosterone is elevated in approximately 60-80% of cases[2].

The challenge: many women with PCOS have total testosterone in the "high-normal" range (50-70 ng/dL) rather than flagrantly elevated. Labs will not flag this as abnormal. But in the context of symptoms — acne, hirsutism, hair thinning, irregular periods — a total testosterone of 55 ng/dL is very different from 25 ng/dL.

Free testosterone

Free testosterone is the unbound, biologically active fraction. It is a more sensitive marker for hyperandrogenism than total testosterone because it reflects what is actually available to act on tissues. Free testosterone is elevated in up to 80-90% of women with PCOS, even when total testosterone is normal[2].

Why the discrepancy? Because SHBG (sex hormone-binding globulin) is often low in PCOS, meaning more of the total testosterone is free and active. This is why testing total testosterone alone is insufficient.

DHEA-S

Dehydroepiandrosterone sulfate (DHEA-S) is an androgen produced exclusively by the adrenal glands (not the ovaries). It is elevated in approximately 20-30% of women with PCOS, indicating adrenal androgen excess[3]. Very high DHEA-S levels (above 700 mcg/dL) should prompt investigation for adrenal tumors or congenital adrenal hyperplasia, which can mimic PCOS.

SHBG: the amplifier

Sex hormone-binding globulin is a protein made by the liver that binds testosterone, rendering it inactive. SHBG is typically low in PCOS, driven down by insulin resistance (insulin suppresses hepatic SHBG production)[3].

Low SHBG has two consequences: it increases free testosterone (amplifying hyperandrogenism) and serves as an independent marker of insulin resistance. A woman with a total testosterone of 45 ng/dL and SHBG of 20 nmol/L has significantly more bioavailable androgen activity than a woman with the same total testosterone and SHBG of 80 nmol/L.

SHBG below 30 nmol/L in a woman with PCOS symptoms is highly suggestive. Some researchers have proposed using SHBG as a screening marker for PCOS-related metabolic risk because it correlates with insulin resistance, cardiovascular risk, and type 2 diabetes risk[4].

LH and FSH: the ratio that matters

Luteinizing hormone (LH) and follicle-stimulating hormone (FSH) are pituitary hormones that regulate the menstrual cycle. In classic PCOS, LH is disproportionately elevated relative to FSH, producing an LH:FSH ratio above 2:1 or sometimes 3:1[1].

Elevated LH drives ovarian theca cells to produce more androgens. Relatively low FSH means follicles are not adequately stimulated to mature and ovulate — they stall at the small antral stage, producing the characteristic "polycystic" morphology on ultrasound.

Important caveat: the LH:FSH ratio is elevated in only about 60% of women with PCOS, so a normal ratio does not rule out the diagnosis. It is also timing-sensitive — these should be drawn in the early follicular phase (days 2-5 of the menstrual cycle) for accurate interpretation[2].

Insulin and glucose: the metabolic core

Fasting insulin

Insulin resistance is present in 50-70% of women with PCOS — including lean women — and is considered a central driver of the condition, not just a comorbidity[2]. Elevated insulin stimulates ovarian androgen production, suppresses SHBG, promotes weight gain (particularly visceral fat), and creates a self-reinforcing cycle.

Fasting insulin is the single best screening marker for insulin resistance in PCOS. A fasting insulin above 10-12 mIU/L is suggestive; above 15-20 mIU/L is strongly indicative. However, many labs do not flag insulin until it exceeds 25 mIU/L, missing the early and moderate stages.

HOMA-IR

Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) is calculated as (fasting insulin x fasting glucose) / 405 (using mg/dL for glucose and mIU/L for insulin). A HOMA-IR above 2.0 suggests insulin resistance; above 2.5 is considered diagnostic. In PCOS, HOMA-IR is a more reliable indicator than either fasting glucose or HbA1c alone[5].

Fasting glucose and HbA1c

Fasting glucose and HbA1c are important for assessing diabetes risk, but they are late markers — they do not rise until insulin resistance is already advanced. A woman with PCOS can have a normal fasting glucose (85 mg/dL) and normal HbA1c (5.2%) but a fasting insulin of 22 mIU/L and HOMA-IR of 4.6. This is why testing glucose without insulin is inadequate for PCOS.

AMH: the ovarian reserve marker repurposed

Anti-Mullerian hormone (AMH) is produced by granulosa cells of small antral follicles. In PCOS, the excess of small follicles produces elevated AMH — typically 2-4x higher than age-matched controls[6].

AMH above 4.7 ng/mL has been proposed as a diagnostic criterion for PCOS, potentially replacing ultrasound (which is operator-dependent and difficult in adolescents). The 2023 international evidence-based PCOS guideline acknowledges AMH as an alternative to ultrasound for diagnosis in adults[7].

Higher AMH levels within the PCOS population correlate with more severe phenotypes: greater anovulation, higher androgens, and worse insulin resistance. AMH also declines more slowly with age in PCOS, which is one reason fertility may be relatively preserved into the late 30s compared to the general population.

Thyroid function: the essential rule-out

Hypothyroidism can cause irregular periods, weight gain, fatigue, and hair changes — all symptoms that overlap with PCOS. Hashimoto's thyroiditis is also more common in women with PCOS than in the general population[8].

Every PCOS work-up should include TSH and free T4 at minimum. If TSH is elevated or borderline, check thyroid peroxidase antibodies (TPO-Ab) to assess for Hashimoto's. Treatment of concurrent hypothyroidism can meaningfully improve PCOS symptoms.

Prolactin

Mild hyperprolactinemia (prolactin levels modestly above normal) can cause irregular periods and mimic PCOS. More significantly, a prolactinoma (pituitary tumor) can present with menstrual irregularity and androgen excess. Checking prolactin at diagnosis rules out this differential[1].

Note: prolactin is stress-sensitive. A mildly elevated result (25-40 ng/mL) should be repeated fasting and relaxed before pursuing further investigation. Significantly elevated prolactin (above 100 ng/mL) warrants MRI of the pituitary.

The complete PCOS blood panel

PCOS phenotypes: not all PCOS looks the same

The Rotterdam criteria define four PCOS phenotypes based on which two (or three) of the three criteria are met[2]:

• Phenotype A (classic): Hyperandrogenism + oligo-anovulation + polycystic ovaries. The most metabolically severe phenotype.
• Phenotype B (classic): Hyperandrogenism + oligo-anovulation (without polycystic ovaries). Also metabolically severe.
• Phenotype C (ovulatory): Hyperandrogenism + polycystic ovaries (with regular periods). Milder metabolic profile.
• Phenotype D (non-hyperandrogenic): Oligo-anovulation + polycystic ovaries (without hyperandrogenism). Mildest metabolic profile, but controversial — some experts argue this should not be classified as PCOS.

Your blood markers can help identify your phenotype, which matters for treatment approach and long-term risk assessment.

17-hydroxyprogesterone: ruling out congenital adrenal hyperplasia

Non-classic congenital adrenal hyperplasia (NCAH) due to 21-hydroxylase deficiency can closely mimic PCOS — it causes androgen excess, irregular periods, and sometimes polycystic ovaries. It affects 1-10% of hyperandrogenic women depending on ethnicity[9].

An early-morning 17-hydroxyprogesterone (17-OHP) level below 200 ng/dL effectively rules out NCAH. A level above 200 ng/dL warrants an ACTH stimulation test. This is particularly important in women of Ashkenazi Jewish, Hispanic, Mediterranean, or South Asian descent, where NCAH prevalence is higher.

Lipid panel and cardiovascular risk

Women with PCOS have a significantly elevated lifetime cardiovascular risk. Dyslipidemia — typically elevated triglycerides, low HDL, and sometimes elevated LDL — is present in 70% of women with PCOS[4].

A full lipid panel (total cholesterol, LDL, HDL, triglycerides) should be part of every PCOS assessment. The triglyceride-to-HDL ratio is a particularly useful surrogate marker for insulin resistance and atherogenic dyslipidemia: a ratio above 3.0 is concerning, and above 3.5 is strongly indicative of metabolic syndrome[5].

Vitamin D

Vitamin D deficiency is significantly more prevalent in women with PCOS compared to controls, with prevalence estimates of 67-85%[2]. Low vitamin D is associated with worse insulin resistance, higher androgens, and lower SHBG in PCOS. Supplementation studies show modest but consistent improvements in insulin sensitivity and androgen levels[10].

When to test: timing matters

For the most interpretable results, blood should be drawn:

• Early follicular phase (days 2-5): LH, FSH, estradiol, testosterone. If periods are absent or highly irregular, draw at any time but note this.
• Fasting (8-12 hours): Glucose, insulin, lipid panel. This is critical — non-fasting insulin is uninterpretable.
• Morning (before 10 AM): Testosterone and DHEA-S have diurnal variation and are highest in the morning.
• Off hormonal contraceptives: Ideally, hormonal contraceptives should be stopped for 3 months before diagnostic blood work, as they suppress androgens and raise SHBG, masking the underlying picture. If stopping is not practical, interpret results with this limitation in mind.

Monitoring over time

PCOS is a lifelong condition. Regular monitoring serves two purposes: assessing treatment response and screening for metabolic progression.

• Every 6-12 months on treatment: Testosterone, free testosterone, SHBG, fasting insulin, fasting glucose, HbA1c
• Annually: Full lipid panel, vitamin D, TSH
• Every 2-3 years: Oral glucose tolerance test (more sensitive than fasting glucose for detecting impaired glucose tolerance, which is more common in PCOS)[7]
• If planning pregnancy: Full hormonal panel, AMH, glucose tolerance test, TSH, iron/ferritin, folate, B12

The inflammation connection

PCOS is increasingly recognized as a state of chronic low-grade inflammation. hs-CRP is elevated in women with PCOS independent of BMI[11]. While hs-CRP is not diagnostic for PCOS, it adds to the cardiovascular risk assessment and can track response to lifestyle interventions.

Lifestyle interventions and what to expect on blood work

A 5-10% reduction in body weight, when applicable, is the most effective intervention for PCOS, improving nearly every blood marker[7]:

• Testosterone and free testosterone decrease
• SHBG increases
• Fasting insulin and HOMA-IR improve
• Lipid profile improves (triglycerides fall, HDL rises)
• hs-CRP decreases
• Ovulatory cycles may resume

Inositol (myo-inositol 4g + D-chiro-inositol 0.1g daily) has moderate evidence for improving insulin sensitivity and ovulation in PCOS, with improvements typically visible on blood work within 3-6 months[12].

Metformin reduces fasting insulin, improves HOMA-IR, and modestly reduces testosterone. Spironolactone reduces free testosterone and improves acne and hirsutism but takes 3-6 months for clinical effect[2].