O besity is associated with metabolic and cardiovascular risk factors that include type 2 diabetes, hypertension, and dyslipidemia (1-4). A subset of obese subjects has been identified that appears to be protected from obesity-related metabolic abnormalities (5-11). These subjects, termed metabolically healthy but obese (MHO), are relatively insulin sensitive and have a rather favorable cardiovascular risk profile (5-11). Although the existence of MHO individuals has been recognized, only a few studies have examined in detail the metabolic characteristics associated with their protective profile (5-13). Whereas MHO individuals appear to have a more favorable cardiovascular risk profile than insulinresistant obese (IRO) individuals, they show early signs of atherosclerosis compared with lean subjects, which could not be explained by alterations in cardiovascular risk factors (12). Among the factors that may account for the early atherosclerosis, insulin-like growth factor (IGF)-1 is a plausible candidate because low plasma IGF-1 concentrations are associated with type 2 diabetes, insulin resistance (14 -16), and increased risk of coronary artery disease (17)(18)(19)(20)(21)(22). To further characterize the protective profile of MHO individuals, we compared clinical characteristics, including cardiovascular risk factors, plasma IGF-1 levels, and intima-media thickness (IMT) of the common carotid artery, of a group of MHO women from a cohort of nondiabetic Italian Caucasians with those of two age-matched groups comprising healthy nonobese or IRO women.
RESEARCH DESIGN AND METHODS -The study group consisted of 73 nonobese (BMI Ͻ27 kg/m 2 ) and 80 obese (BMI Ͼ30 kg/m 2 ) women, recruited by announcements in the Universities of Rome and Catanzaro areas. The inclusion criteria included the following: aged 19 -48 years, absence of diabetes, and absence of known inflammatory disease or pathologies affecting glucose metabolism. A total of 68 women in the cohort used oral contraceptives, whereas the remaining had regular menses. Subjects underwent anthropometrical evaluation, a 75-g oral glucose tolerance test, and a euglycemic-hyperinsulinemic clamp, as previously described (23). Glucose disposal (M) was calculated as the mean rate of glucose infusion during the last 60 min of the examination and was expressed as milligrams per minute per kilogram fatfree mass (M FFM ). IMT of the common carotid artery was measured by highresolution B-mode ultrasound using an ATL-HDI 3000 system equipped with a 5-MHz linear array transducer, as previously described (24). The protocol was approved by the ethics committees, and informed written consent was obtained from all participants. The investigations were performed in accordance with the principles of the Declaration of Helsinki. Plasma insulin and IGF-1 concentrations were determined by a chemiluminescence-based assay. Relationships between variables were sought by stepwise multivariate linear regression analysis.
RESULTS -Because insulin sensitivity is a continuous trait, there is n...