ROLE OF ANTI-MÜLLERIAN HORMONE IN THE DIAGNOSIS OF POLYCYSTIC OVARY SYNDROME IN ADOLESCENCE

Abstract

Background: Diagnosing polycystic ovary syndrome (PCOS) in adolescence is difficult because menstrual irregularity, acne, and polycystic ovarian morphology overlap with normal pubertal physiology. Anti-Müllerian hormone (AMH), a granulosa-cell product that is raised in PCOS, has been proposed as a simple blood test, but current international guidance advises against using it for diagnosis in adolescents. Objective: To evaluate the diagnostic accuracy of serum AMH for identifying PCOS in adolescent girls, to derive an optimal cutoff, and to examine how accuracy varies with age and body mass index. Methods: This hospital-based cross-sectional analytical study analyzed a simulated patient-level dataset of 200 females aged 13–19 years presenting with menstrual irregularity, features of hyperandrogenism, or suspected PCOS, classified as PCOS (n=100) or non-PCOS (n=100) using adolescent-specific criteria. Serum AMH (ng/mL) was compared between groups using the Mann–Whitney U test. Diagnostic accuracy was assessed by receiver operating characteristic (ROC) analysis with DeLong confidence intervals, the Youden index for the cutoff, and Wilson intervals for sensitivity and specificity. AMH was compared with other markers, and its association with PCOS was adjusted for age and BMI by logistic regression. Results: Median AMH was 6.92 ng/mL (IQR 5.28–8.84) in the PCOS group and 3.71 ng/mL (IQR 2.91–4.77) in the non-PCOS group (p<0.001). The area under the ROC curve (AUC) was 0.860 (95% CI 0.808–0.912). The optimal cutoff was ≥5.22 ng/mL (bootstrap 95% interval 4.75–5.95), giving sensitivity 78.0% (95% CI 68.9–85.0), specificity 82.0% (73.3–88.3), positive predictive value 81.3% and negative predictive value 78.8%, with 22 of 100 PCOS cases missed and 18 of 100 non-PCOS girls misclassified. AMH discriminated no better than total testosterone (AUC 0.883) or ovarian volume (AUC 0.862). Each 1 ng/mL rise in AMH was associated with higher odds of PCOS after adjustment for age and BMI (odds ratio 2.36, 95% CI 1.82–3.07). Accuracy was lowest at ages 13–15 years (AUC 0.758, 95% CI 0.620–0.896, 21 PCOS cases) and higher at 16–17 years (0.910) and 18–19 years (0.882). Conclusion: In this simulated cohort, AMH showed good but imperfect discrimination of PCOS in adolescents, with about one in five cases missed at the optimal cutoff and weaker performance in younger girls. AMH did not outperform simpler androgen and ultrasound markers. These findings are consistent with guidance that AMH should not be used as a stand-alone diagnostic test in adolescents and require confirmation in real, prospectively collected clinical data.

Keywords: Anti-Müllerian hormone; polycystic ovary syndrome; adolescent; diagnostic accuracy; ROC curve; hyperandrogenism.