Does beta cell dysfunction or insulin resistance play a role in the development of adiposity? [abstract]

ABSTRACT

OBJECTIVE: To test the hypothesis that pre-existing beta cell function (BCF) and insulin resistance (IR) are related to the development of adiposity. METHOD: A sample of 493 subjects (mean age 47ñ18 yrs; 55 percent women) was selected from the Spanish Town Survey. Base-line anthropometry, blood pressure, lipid profile, oral glucose tolerance test and fasting insulin levels were done. Anthropometry was repeated 4.5 years later in 319 subjects. Baseline IR and BCF were computed using the method of homeostatic model assessment. Cross-sectional and prospective associations between measures of adiposity (body mass index, waist-hip ratio) and explanatory variables (age, gender, IR, BCF, lipids, blood pressure, glycaemia) were assessed using correlation and multiple linear regression and analyses. RESULTS: Baseline body mass index (BMI) and waisthip ratio (WHR), folow-up BMI and WHR, change in weight and change in BMI, but not change in WHR, were correlated with IR (r=-0.120; p<0.001). BCF was correlated only with baseline WHR (r=-0.21; p<0.001). In multiple regression analyses, gender, HDL-C, LDL-C, BCF and IR were associated with baseline BMI regardless of glycaemic state. Baseline WHR was associated with the same variables and triglycerides (TG), but the association with the BCF and TG lost significance in subjects with normoglycaemia. Follow-up BMI was associated with age, IR, HDL-C and LDL-C, but not after adjustment for baseline BMI. Follow-up WHR was associated with gender, TG and HDL-C. After adjustment for baseline WHR or BMI, only gender and lipids remained significant. IR, TG and baseline BMI were associated with change in weight. Change in BMI was significantly correlated with IR, but not after adjustment for baseline BMI. Change in WHR was associated with baseline WHR and gender. CONCLUSIONS: BCF and IR are independently associated with current adiposity. IR, but not BCF, significantly explains some of the variance of future adiposity. However, this effect is ameliorated if baseline adiposity is considered. (AU)