233. Characteristics of Adolescents With PCOS As Diagnosed By The Rotterdam Criteria

Abstract

Polycystic ovarian syndrome (PCOS) is challenging to diagnose during adolescence. Diagnostic criteria for adults are difficult to apply to adolescents in part because ultrasound criteria are based on adult data. This retrospective study aimed to assess and compare clinical, metabolic, and hormonal characteristics of adolescents diagnosed with PCOS who required ultrasound for diagnosis according to Rotterdam criteria to those of adolescents who did not. Two hundred forty-two females presenting to an outpatient adolescent medicine practice with complaints of irregular menses, hirsutism, or hyperandrogenism from June 2012-June 2016 were evaluated as a part of a quality improvement project. Inclusion criteria for the current study included age 13-25 years, physician diagnosis of PCOS, and availability of at least one laboratory result (n=176). Patients were divided into 3 groups: those diagnosed by clinical or biochemical hyperandrogenism and irregular menstruation (HA/IM, n=148); by hyperandrogenism and polycystic ovary morphology (PCOM) on ultrasound (HA/PCOM, n=10); or by irregular menstruation and PCOM (IM/PCOM, n=18). Data included clinical characteristics (age at evaluation, age at menarche, body mass index (BMI), blood pressure, acne score (range: 0-5), Ferriman-Gallwey (FG) score (range: 0-36), and hormonal measures (FSH, LH, prolactin, TSH, total testosterone, free testosterone, DHEAS, 17-hydroxyprogesterone). Metabolic measures obtained included non-fasting total cholesterol and HDL for those classified as low-risk and HbA1c with fasting total cholesterol, HDL, LDL, triglycerides, insulin, and glucose for those classified as high-risk (BMI > 30 kg/m2 and/or personal/family history of diabetes). Given the non-normal distribution of most measures, statistical comparisons between groups were performed using the non-parametric Wilcoxon rank sum test, with significance adjusted for multiple comparisons at p=0.017. Median age at evaluation was 16.0 years [inter-quartile range (IQR) = 2 years]. Median age at menarche was 12 years (IQR = 2 years). Median BMI for the entire cohort was 28.9 kg/m2; BMI was not significantly different between groups (HA/IM=29.3 kg/m2, HA/PCOM=25 kg/m2, and IM/PCOM=24 kg/m2). Significant differences were found between groups HA/IM and IM/PCOM for median acne scores (2 vs 1, p<0.01), FG scores (6 vs 3, p=0.01), median total testosterone (43 ng/dL vs 30 ng/dL, p<0.01) and free testosterone (7.9 pg/mL vs 5.2 pg/mL, p<0.01). There was a significant difference in median HbA1c for group HA/IM versus IM/PCOM (5.3% vs 5.1%, p=0.01) and a marginally significant difference in median HbA1c for group HA/PCOM versus HA/IM (5.9% vs 5.3%; p=0.04). The majority of adolescents (84%) evaluated did not require ultrasound for diagnosis of PCOS as they had already met 2 of the 3 Rotterdam criteria. Limited differences were found between diagnostic groups. Acne scores, FG scores, and testosterone concentrations were elevated in those with HA as expected. Additionally, those with HA had significantly higher HbA1c and trended towards greater BMIs, suggesting elevated cardiovascular risk in adolescents with HA. Further longitudinal studies looking into future cardiovascular health by PCOS diagnosis type in adolescence should be considered.