Insulin sensitivity of blood glucose versus insulin sensitivity of blood free fatty acids in normal, obese, and obese-diabetic subjects.

We calculated insulin sensitivity indices (ISI) concerning the insulin effect on both glycemia and blood free fatty acids (FFA), named ISI(gly) and ISI(ffa), respectively, in 34 normal, 27 obese, and 11 obese-diabetic subjects by using the following formulas: ISI(gly)= 2/[(INSp x GLYp) +1], and ISI(ffa)= 2/[(INSp x FFAp)+1], in which INSp, GLYp, and FFAp = insulinemic, glycemic, and FFA areas during oral glucose tolerance test (OGTT) (75 g glucose, suggested sampling time: 0, 1, and 2 hours) of the person studied. A slight modification of these formulas allows the calculation of insulin resistance indices (IRI), ie, IRI(gly) and IRI(ffa). ISI and IRI are complementary, as their sum is always equal to 2, so that IRI can be deduced from ISI and vice versa. By using basal levels instead of areas, insulin sensitivity (or resistance) in the basal state can also be measured. Basal levels and areas are expressed by taking the mean normal value as 1, so that in normal subjects ISI(gly) and ISI(ffa), as well as IRI(gly) and IRI(ffa), are always around 1, with maximal variations comprised between 0 and 2. ISI(ffa) was markedly reduced in both the obese (mean, 0.47 +/- 0.04) and the obese-diabetic (mean, 0.41 +/- 0.06) subjects, whereas ISI(gly) was less reduced in the obese (mean, 0.57 +/- 0.04) than in the obese-diabetic (mean, 0.40 +/- 0.03) subjects. ISI(gly)-basal was less affected than ISI(ffa)-basal in both groups. Multiple regression showed that ISI(gly) and ISI(ffa) were significantly inversely correlated with age, body mass index (BMI), and diastolic (but not systolic) blood pressure. Meta-analysis of data from the literature showed that ISI(gly) was significantly correlated with the hyperinsulinemic-euglycemic clamp data. However, the "clamp" is performed under artificial, persistent hyperinsulinemia (which entails FFA suppression) as never occurs in the life of patients, whereas our indices are performed under physiologic conditions, and represent simple tools suitable for clinical or epidemiologic studies, allowing assessment of whole-body insulin sensitivity with regard to both glycemia and blood FFA.