THE RELATIVE CONTRIBUTION OF ANDROGENS AND INSULIN IN DETERMINING ABDOMINAL BODY-FAT DISTRIBUTION IN PREMENOPAUSAL WOMEN

To investigate the relative contribution of insulin and sex hormones in determining the abdominal pattern of fat distribution in premenopausal women, five groups of age-matched subjects were examined: Group 1 consisted of 14 normal weight eumenorrheic women (NO); Group 2 of 9 obese eumenorrheic women (OB); Group 3 of 14 normal weight hyperandrogenic women with polycystic ovary syndrome (NO-HA); G roup 4 of 1 0 obese hyperandrogenic women with polycystic ovary syndrome (OB-HA) and, finally, Group 5 of 10 obese hyperandrogenic women with polycystic ovary syndrome and acanthosis nigricans (OB-HA-AN). Both the two normal weight groups and the three obese groups were matched for body mass index values. Sex hormone pattern showed significantly higher LH and testosterone levels in hyperandrogenic women with respect to NO and OB women but obese hyperandrogenic groups (OB-HA and OB-HA-AN) presented significantly lower LH concentrations than NO-HA. Fasting and glucose-stimulated insulin levels were significantly higher in OB than NO, in OB-HA and OB-HA-AN than in OB and NO-HA, and in OB-HA-AN than in OB-HA, without any significant difference between OB and NO-HA. Body fat distribution, expressed by the waist to hip ratio (WHR), showed progressively higher values (p < 0.01) from NO to OB, NO-HA, OB-HA and, particularly, OB-HA-AN women. Determination coefficients r2 obtained from simple regression analysis showed that the sum of insulin values during the glucose tolerance test and testosterone levels had a more significant power in determining WHR variability. Multiple regression analysis showed that more than 30 percent (r2 = 0.322, p < 0.01) of the total variance accounting for WHR was explained by these two variables. Therefore we suggest that hyperinsulinemia and androgens probably represent important factors in determining predominant localization of body fat in the abdominal sites in premenopausal women, since they can explain more than 30 percent of the total variability of WHR.