Oral glucose tolerance test preferred for pre-diabetes screening in PCOS
This study found that among PCOS patients, HbA1C and FPG were poor detectors of pre-diabetes and showed limited agreement with the screening standard, the 2-hour oral glucose tolerance test (OGTT). Additionally, pre-diabetes was found in 7% of normal weight PCOS patients. This finding suggests that the current American Diabetes Association (ADA) recommendation of screening only PCOS women with body mass index (BMI) ≥ 25 kg/m2 would miss a substantial portion of pre-diabetes cases. Additionally, the results of this investigation suggest that OGTT should be the screening test of choice for pre-diabetes in PCOS patients, a population a higher risk for Type 2 Diabetes, rather than the current ADA guidelines which recommend diagnosis via FPG, HbA1c or OGTT.

Strengths include a large sample size, strict definition of PCOS using Rotterdam criteria, evaluation of standard anthropometric data from all subjects and assessment of pre-diabetes through multiple tests. Limitations include a cross-sectional design and a low prevalence of pre-diabetes and T2DM in the study population, who was also younger and leaner (median age 27, median BMI 24) than the average PCOS patient. Reproducibility of results in a more diverse PCOS patient population would confirm study findings and inform management.

Click to read the study in Human Reproduction

Relevant Reading: American Diabetes Association: Standards of Medical Care in Diabetes-2013

In-Depth [cross-sectional study]: This study compared the utility of HbA1C, FPG and the OGTT in identifying pre-diabetes and T2DM in a cohort of 671 Austrian women being evaluated for PCOS from 2006 to 2012. Diagnosis of PCOS was made using the Rotterdam criteria (2 of the 3 following criteria had to be met: clinical or biochemical hyperandrogenism, oligo- or anovulation and sonographic polycystic ovary morphology, defined as ≥12 follicles, 2-9mm in diameter in at least 1 ovary). Pre-diabetes and T2DM were diagnosed using ADA guidelines (elevated FPG and/or HbA1C and/or 2-hour plasma glucose). Standard anthropometric data, hormone levels, serum lipids, HbA1C, FPG and 2-hour OGTT were assessed in all patients.

HbA1C showed weak agreement with OGTT for diagnosing pre-diabetes (κ 0.36, p<0.001) and had poor sensitivity (25%). Among women whose BMI was <25 kg/m2, sensitivity was even poorer at 4.3%. FPG was also demonstrated weak agreement with OGTT for pre-diabetes diagnosis (κ 0.05, p<0.001) with a sensitivity of 40.8%. When screening for Type 2 Diabetes in contrast to pre-diabetes, results from HbA1C and FPG tests demonstrated higher agreement with OGTT results (κ 0.80, p<0.001 for both) and improved sensitivity (66.7% for both).