When evaluating metabolic risks in women with PCOS, androgenic status, especially abdominal obesity and age, should be taken into account, which would allow tailored management of the syndrome from early adulthood on.

Context: Women with polycystic ovary syndrome (PCOS) have increased androgen secretion throughout fertile life; however, the data on the effect of menopause on hyperandrogenemia in these women are scarce. Nevertheless, large comprehensive comparative studies on age-related androgen levels in women with PCOS are lacking. Objective:The objective of the study was to investigate the effect of age on serum androgen levels in women with PCOS and to determine cutoff values for androgens and SHBG associated with a PCOS diagnosis.Design: This was a case-control study. Setting:The study was conducted in five university sites in the Nordic countries.Patients: In all, 681 women with PCOS and 230 referent women were grouped according to age into seven age groups (18 to Ͼ 50 y).Interventions: There were no interventions.Main Outcome measures: T, SHBG, free androgen index (FAI), calculated free T (cFT), androstenedione (A4), and dehydroepiandrosterone sulfate were measured.Results: Androgen levels in women with PCOS decreased with age toward menopause. The difference between women with PCOS and the referent women narrowed and individual variation increased as they approached menopause. T levels, FAI, and cFT were significantly higher in women with PCOS aged 18 -44 years (P Ͻ .001, adjusted for body mass index). The best predictive factors for having PCOS were cFT (Ն0.40 ng/dL, odds ratio [OR] 7.90), FAI (Ն2.0, OR 6.71), and A4 (Ն277.94 ng/dL, OR 6.16). Conclusions:Women with PCOS had elevated serum androgen levels also after menopause. The parameters that best predicted PCOS at all ages were cFT, A4, and FAI. (J Clin Endocrinol Metab 100: 3400 -3407, 2015)

STUDY QUESTIONSCan serum anti-Müllerian hormone (AMH) levels measured in female adolescents predict polycystic ovary syndrome (PCOS)-associated features in adolescence and early adulthood?SUMMARY ANSWERAMH levels associated well with PCOS-associated features (such as testosterone levels and oligoamenorrhoea) in adolescence, but was not an ideal marker to predict PCOS-associated features in early adulthood.WHAT IS KNOWN ALREADYSeveral studies have reported that there is a strong correlation between antral follicle count and serum AMH levels and that women with PCOS/PCO have significantly higher serum AMH levels than women with normal ovaries. Other studies have reported an association between AMH serum levels and hyperandrogenism in adolescence, but none has prospectively assessed AMH as a risk predictor for developing features of PCOS during adulthood.STUDY DESIGN, SIZE, DURATIONA subset of 400 girls was selected from the prospective population-based Northern Finland Birth Cohort 1986 (n = 4567 at age 16 and n = 4503 at age 26). The population has been followed from 1986 to the present.PARTICIPANTS/MATERIAL, SETTING, METHODSAt age 16, 400 girls (100 from each testosterone quartile: 50 with oligo- or amenorrhoea and 50 with a normal menstrual cycle) were selected at random from the cohort for AMH measurement. Metabolic parameters were also assessed at age 16 in all participants. Postal questionnaires enquired about oligo- or amenorrhoea, hirsutism, contraceptive use and reproductive health at ages 16 and 26.MAIN RESULTS AND ROLE OF CHANCEThere was a significant correlation between AMH and testosterone at age 16 (r = 0.36, P < 0.001). AMH levels at age 16 were significantly higher among girls with oligo- or amenorrhoea compared with girls with normal menstrual cycles (35.9 pmol/l [95% CI: 33.2;38.6] versus 27.7 pmol/l [95% CI: 25.0;30.4], P < 0.001). AMH at age 16 was higher in girls who developed hirsutism at age 26 compared with the non-hirsute group (31.4 pmol/l [95% CI 27.1;36.5] versus 25.8 pmol/l [95% CI 23.3;28.6], P = 0.036). AMH at age 16 was also higher in women with PCOS at age 26 compared with the non-PCOS subjects (38.1 pmol/l [95% CI 29.1;48.4] versus 30.2 pmol/l [95% CI 27.9;32.4], P = 0.044). The sensitivity and specificity of the AMH (cut-off 22.5 pmol/l) for predicting PCOS at age 26 was 85.7 and 37.5%, respectively. The addition of testosterone did not significantly improve the accuracy of the test. There was no significant correlation between AMH levels and metabolic indices at age 16.IMPLICATIONS, REASONS FOR CAUTIONAMH is related to oligo- or amenorrhoea in adolescence, but it is not a good marker for metabolic factors. The relatively low rate of participation in the questionnaire at age 26 may also have affected the results. AMH was measured in a subset of the whole cohort. AMH measurement is lacking international standardization and therefore the concentrations and cut-off points are method dependent.WIDER IMPLICATIONS FOR THE FINDINGSUsing a high enough cut-off value of AMH to predict which...