Insulin Action and Age: European Group for the Study of Insulin Resistance (EGIR)

Evidence that age is associated with insulin resistance is discordant. We analyzed euglycemic insulin clamp (1 mU x min(-1) x kg(-1)) data collected at 20 centers throughout Europe from 1,146 men and women with normal glucose tolerance, ranging in age from 18 to 85 years. In the whole group, insulin action (as the M value) declined slightly with age (at a rate of 0.9 micromol x min(-1)-kg(-1) per decade of life, 95% CI = 0.4-1.3, P = 0.0002). When adjusted for BMI, this relationship was no longer statistically significant. The same result was obtained whether insulin action was expressed per kilogram of body weight or per kilogram of fat-free mass, expressed as the M:I ratio, or estimated from fasting plasma insulin concentrations. Subgroup analysis showed that a significant BMI-adjusted decrease in insulin action with age was present only in lean (BMI <25 kg/m2) women (a rate of 1.6 micromol x min(-1) x kg(-1) per decade, 95% CI = 0.6-2.5, P = 0.001), in whom percentage fat mass also increased with age (by 0.38% body weight per decade, P = 0.0007). Insulin action was positively associated with insulin suppression of circulating free fatty acids (FFAs) (+1.5 micromol x min(-1) x kg(-1) for each 10% increase in FFA suppression, P < 0.0001) in a multivariate model accounting for sex, BMI, age, and fasting FFA levels. Furthermore, insulin suppression of FFAs improved with age in men (2% per decade, P < 0.0001) but not in women. In the subgroup of lean women in whom insulin action declined with age, adding FFA suppression to a multiple regression equation canceled the association between age and insulin action. Thus, the small effect of age on insulin action could be adequately explained on the basis of age-related changes in body composition and substrate competition. We conclude that in healthy Europeans, age per se is not a significant cause of insulin resistance of glucose metabolism or lipolysis.