Anti-Mullerian Hormone Levels in Adolescence in Relation to Long-term Follow-up for Presence of Polycystic Ovary Syndrome

Context: Anti-Mullerian hormone (AMH) measured in adolescence as biomarker for prediction of adult polycystic ovary syndrome (PCOS) is doubtful but not substantiated. Objective: To investigate whether serum AMH levels and other PCOS-associated features in adolescence can predict the presence of PCOS in adulthood. Design and Setting: A long-term follow-up study based on a unique adolescent study on menstrual irregularities performed between 1990 and 1997. Participants and interventions: AMH was assayed in 271 adolescent girls. Data on PCOS features were combined with AMH levels. In 160 of the 271 (59%) participants, we collected information in adulthood about their menstrual cycle pattern and presence of PCOS (features) by questionnaire 2 decades after the initial study. Results: AMH was higher in adolescent girls with oligomenorrhea compared with girls with regular cycles, median (interquartile range): 4.6 (3.1-7.5) versus 2.6 (1.7-3.8) mu g/L (P<0.001). Women with PCOS in adulthood had a higher median adolescent AMH of 6.0 compared with 2.5 mu g/L in the non-PCOS group (P<0.001). AMH at adolescence showed an area under the receiver operating characteristic curve for PCOS in adulthood of 0.78. In adolescent girls with oligomenorrhea the proportion developing PCOS in adulthood was 22.5% (95% CI, 12.4-37.4) against 5.1% (95% CI, 2.1-12.0) in girls with a regular cycle (P=0.005). Given adolescent oligomenorrhea, adding high AMH as factor to predict adult PCOS or adult oligomenorrhea was of no value. Conclusions: Adolescent AMH either alone or adjuvant to adolescent oligomenorrhea does not contribute as prognostic marker for PCOS in adulthood. Therefore, we do not recommend routine its use in clinical practice.