AMH in Diagnosis of PCOS
AMH is not currently recommended as a standalone diagnostic test for PCOS and should not replace the established Rotterdam criteria (requiring 2 of 3: oligo/anovulation, hyperandrogenism, polycystic ovarian morphology), though it may serve as a supportive marker when ultrasound is unavailable or unreliable. 1, 2
Current Guideline Position
The international evidence-based guidelines explicitly state that AMH is not clinically applicable for detecting polycystic ovarian morphology (PCOM) or diagnosing PCOS outside of research settings until critical standardization issues are resolved. 1 The Endocrine Society confirms that AMH should not yet replace established diagnostic criteria. 2
Why AMH Cannot Stand Alone
The systematic review informing international guidelines identified fundamental limitations that prevent clinical adoption: 1
- Significant heterogeneity exists across studies with poorly defined PCOS and control populations
- Inconsistent cut-off values ranging from 24.29 to 100 pmol/L across different studies, with no international standard 1
- Assay variability with substantial discrepancies in between-assay conversion factors and differential responses to pre-analytical handling 1
- Significant overlap in AMH levels between women with and without PCOS, limiting standalone diagnostic utility 2
When AMH May Provide Supportive Information
Despite these limitations, AMH demonstrates consistent biological associations with PCOS:
- Serum AMH levels are significantly elevated in women with PCOS compared to normal ovulatory women, with levels 2-3 times higher 1, 3
- AMH correlates primarily with androgen status (testosterone, p<0.001) rather than insulin resistance, adiposity, or gonadotropins 3
- AMH may be particularly useful in adolescents within 8 years of menarche where ultrasound is contraindicated due to poor specificity 1, 2
Diagnostic Performance Data
Research studies show variable but generally good diagnostic accuracy: 1
- Dewailly 2011: 35 pmol/L threshold achieved 92% sensitivity and 97% specificity (AUC 0.973) 1
- Carmina 2016: >33.57 pmol/L showed 79% sensitivity and 96% specificity (AUC 0.952) 1
- Recent study 2023: 3.75 ng/mL (26.78 pmol/L) cutoff yielded 79% sensitivity and 81% specificity (AUC 0.969) 4
- Study 2013: 3.94 ng/mL cutoff achieved 80% sensitivity and 89.8% specificity (AUC 0.916) 5
However, one study found that AMH ≥30 pmol/L correctly identified only 79% of PCOS cases, and elevated AMH in lean women may not indicate PCOS but rather functional hypothalamic hypogonadism. 3, 6
Practical Clinical Algorithm
When evaluating suspected PCOS, follow this sequence:
Apply Rotterdam criteria first - document at least 2 of 3 features: 2, 7
- Oligo/anovulation (cycle length >35 days)
- Clinical/biochemical hyperandrogenism (total testosterone by LC-MS/MS preferred, calculated free testosterone, or androstenedione)
- PCOM on ultrasound (≥20 follicles per ovary or ovarian volume >10 mL)
Consider AMH measurement in specific scenarios: 2, 8
- Adolescents <8 years post-menarche where ultrasound should not be used
- When reliable ultrasound is unavailable or technically inadequate
- When typical clinical/laboratory findings are equivocal
- To support diagnosis when only one Rotterdam criterion is clearly met
Interpret AMH with caution: 1, 2, 6
- Values >30-35 pmol/L support PCOS diagnosis but are not definitive
- Age-specific reference ranges are critical as AMH naturally declines with age
- In lean women with elevated AMH, consider functional hypothalamic hypogonadism (higher SHBG >111 nmol/L suggests FHH rather than PCOS)
Critical Pitfalls to Avoid
- Never diagnose PCOS based solely on elevated AMH - significant overlap exists with normal women and other conditions 2, 6
- Do not use AMH in adolescents as a replacement for clinical judgment - physiological multifollicular ovaries are common in this age group 1, 2
- Recognize assay-specific differences - results are not interchangeable between different AMH assay platforms 1
- Distinguish PCOS from functional hypothalamic hypogonadism in lean women - both show elevated AMH, but FHH has higher SHBG and responds to weight gain 6
Additional Metabolic Assessment Required
Once PCOS is diagnosed (with or without AMH support), perform comprehensive metabolic screening: 2, 7
- Two-hour oral glucose tolerance test for insulin resistance and type 2 diabetes
- Fasting lipid profile for dyslipidemia screening
- Calculate BMI and waist-hip ratio
- Screen for cardiovascular risk factors given long-term health implications