Associations of different molecular forms of AMH and biomarkers of PCOS and normal women
Article In Press
Marie Louise Wissing, Ph.D., Anne Lis Mikkelsen, DMSc, Ajay Kumar, Ph.D., Bhanu Kalra, Ph.D., Susanne Elisabeth Pors, Ph.D., Esben Meulengracht Flachs, Ph.D., Claus Yding Andersen, D.M.Sc.
To study different antimüllerian hormone (AMH) isoforms in women with polycystic ovary syndrome (PCOS) and healthy regularly cycling women and to investigate whether levels of AMH isoforms combined with baseline characteristics can predict PCOS.
Eighty-eight women with PCOS and 24 age- and body mass index (BMI)–matched normal control subject women recruited from April 2010 to February 2013. AMH isoforms were analyzed in biobanked serum samples collected at Holbaek Fertility Clinic, Denmark. All study participants went through a baseline examination including gynecologic history, objective examination, transvaginal ultrasound, and blood sampling. Each woman was characterized by measurement of total T, free T, SHBG, A, DHEAS, FSH, LH, E2, PRL, TSH, serum insulin, plasma glucose, and C-peptide.
Main Outcome Measure(s)
Serum levels of three different AMH isoforms.
Levels of AMH measured with each of three AMH ELISAs were significantly higher in women with PCOS compared with control women. The ratio between AMH isoforms showed significant associations with metabolic parameters (BMI, SHBG, C-peptide, cholesterols, triglycerides, and the modified homeostasis-model assessment). Prediction of PCOS showed a high precision with areas under the receiver operating characteristic curve of 97% when AMH measurements were combined with androgens and BMI.
Three ELISAs detecting different parts of the AMH molecule all detected significantly higher levels in women with PCOS compared with control women. The relative distribution of AMH isoforms did not differ between women with PCOS and control women. AMH isoforms alone and in combination with baseline characteristics predicted PCOS with close to 100% area under the receiver operating characteristic curve.