Tissue insulin sensitivity and body weight in polycystic ovary syndrome.

OBJECTIVE: Polycystic ovarian syndrome (PCOS) and obesity both affect insulin sensitivity. This study was designed to investigate the biochemical indices of PCOS and tissue insulin sensitivity in groups of lean and obese women with clinically equivalent degrees of the syndrome, relative to control subjects.
DESIGN: A prospective study of in vivo parameters and in vitro study of adipocytes to assess insulin sensitivity. PATIENTS: Six lean and 14 overweight patients fulfilling formal diagnostic criteria for PCOS were studied. The degree of hirsutism and amenorrhoea was similar in each group. Eight control subjects were also studied. MEASUREMENTS: Endocrine and metabolic parameters were measured in lean and overweight patients with PCOS and control subjects. In vitro studies of adipocyte insulin receptor binding and adipocyte insulin action were performed.
RESULTS: The mean plasma LH level was elevated in both groups of PCOS but was significantly higher in the lean group (LH levels were 25.1 +/- 3.1 and 14.5 +/- 1.6 iu/l in lean PCOS and obese PCOS, respectively (P = 0.01)). There was a strong inverse correlation between BMI and LH levels (R = - 0.70, P = 0.001). Fasting insulin levels were elevated in both lean and obese groups (11.5 +/- 2.8 and 26.8 +/- 8.1 mU/l, respectively; P = 0.068). Mean serum testosterone and serum androstenedione levels were also elevated in PCOS compared to control subjects but there was no difference between the two groups of PCOS subjects. Insulin receptor binding in amenorrhoeic subjects with PCOS was low in both lean and obese patients with PCOS but was not significantly different between the two groups (0.79 +/- 0.17% and 0.66 +/- 0.07% per 10 cm2 cell membrane, respectively). Maximally insulin-stimulated rates of 3-O-methylglucose transport were low in both groups compared to previously studied normal subjects (0.96 +/- 0.21 and 0.64 +/- 0.07 pmol per 10 cm2 membrane in lean and obese PCOS subjects, respectively; P = NS).
CONCLUSIONS: Lean subjects with a given phenotypic expression of PCOS have an equivalent degree of tissue insulin resistance compared to obese PCOS subjects. This implies that the insulin resistance may be a primary feature of PCOS. If this is so, a similar clinical degree of the syndrome may be brought about by genetically determined insulin resistance in lean subjects or by insulin resistance which is secondary to obesity.