HYPERINSULINEMIA IN THE POLYCYSTIC-OVARY-SYNDROME CONFIRMED WITH A SPECIFIC IMMUNORADIOMETRIC ASSAY FOR INSULIN

OBJECTIVE Hyperinsulinaemia in the polycystic ovary syndrome (PCOS) has previously been defined using polyclonal radioimmunoassays (RIA) in which partially processed insulin-like molecules cross-react. This study aimed to reassess hyperinsulinaemia in women with PCOS using specific immunoradiometric assays (IRMA) for insulin, proinsulin and 32-33 split proinsulin. DESIGN Patients attended for 75 g oral glucose tolerance tests and were divided into groups depending on their degree of obesity and fasting insulin status determined by RIA. IRMA measurements for insulin-like molecules in plasma from patients with PCOS and controls were compared. PATIENTS Thirty-four patients with ultrasound diagnosed PCOS presented to a reproductive endocrinology clinic. A control group comprised women with normal ovaries on ultrasound. Four groups were constructed, two with normal fasting insulin concentrations (lean PCOS and controls) and two with hyperinsulinaemia (lean and obese PCOS). MEASUREMENTS Plasma glucose, insulin (RIA and IRMA), proinsulin and 32-33 split proinsulin concentrations were measured at time 0, 30 and 120 minutes of an oral glucose tolerance test. RESULTS Hyperinsulinaemia determined by RIA in lean and obese women with PCOS was confirmed using a specific IRMA assay for insulin. Plasma proinsulin and 32-33 split proinsulin concentrations were higher in hyperinsulinaemic women with PCOS compared with women with normal insulin concentrations. The proportion of circulating insulin-like molecules represented by proinsulin and 32-33 split proinsulin was similar in all groups studied. CONCLUSIONS Hyperinsulinaemia in PCOS is likely to reflect insulin resistance because the raised concentrations of proinsulin and 32-33 split proinsulin were in proportion to the raised insulin concentrations. Hyperinsulinaemia in PCOS, defined by RIA, therefore differs from that in non-insulin dependent diabetes mellitus where it is largely accounted for by disproportionate hyperproinsulinaemia.