Prognostic implications for insulin-sensitive and insulin-resistant normal-weight and obese individuals from a population-based cohort

Background: There are few prospective data on the prognosis of insulin-sensitive and insulin-resistant normal-weight (NW) or obese individuals. Objectives: The estimated liver fat content, incidences of hyperglycemia and cardiovascular disease, and all-cause and cardiovascular mortality rates were investigated in a population-based cohort of 1658 individuals who were categorized according to BMI and insulin resistance as defined by HOMA-IR values >= 2.5 and the presence of metabolic syndrome. Design: This was a prospective cohort study with a 9-y follow-up. Anthropometric values, blood pressure, and blood metabolic variables were measured, and information on vital status was collected from demographic files at follow-up. Results: A total of 137 of 677 NW individuals (20%) were classified as insulin resistant and normal weight (IR-NW), and 72 of 330 obese individuals (22%) were classified as insulin sensitive and obese (IS-obese). Incidences of diabetes, impaired fasting glucose, and cardiovascular events were 0.4%, 6.3%, and 3.3%, respectively, in insulin-sensitive and normal-weight (IS-NW) individuals (reference category); 5.8%, 10.2%, and 6.6%, respectively, in IR-NW individuals; and 5.6%, 8.3%, and 8.3%, respectively, in IS-obese individuals. In a multiple logistic regression model, risks of incident hyperglycemia and cardiovascular events increased in both groups compared with in the reference category [HR (95% CI): 2.54 (1.42, 4.55) and 1.98 (0.86, 4.54) in IR-NW subjects; 2.16 (1.01, 4.63) and 2.76 (1.05, 7.28) in IS-obese subjects]. The estimated liver fat content significantly increased during follow-up only in the IR-NW group in the same model. Cardiovascular mortality was 2-3-fold higher in IR-NW and IS-obese than in IS-NW individuals in a Cox regression model. Conclusions: Our data refute the existence of healthy obese phenotypes because IS-obese individuals showed increased cardiometabolic risk. The existence of unhealthy NW phenotypes is supported by their increased risk of incident hyperglycemia, fatty liver, cardiovascular events, and death. Am J Clin Nutr 2012;96:962-9.