FAMILIAL CLUSTERING OF INSULIN SENSITIVITY

This study's objective was to determine whether there is familial clustering of insulin sensitivity (S(I)) or insulin-independent glucose uptake (S(G)), which would be evidence that they are genetically determined traits. Outpatients had a 3-h intravenous glucose tolerance test. Nondiabetic individuals (n = 183), ranging in age from 16 to 60 yr, were from 105 families that had 2 parents with non-insulin-dependent diabetes mellitus. Of these families, 62 contributed 1 offspring, 21 contributed 2, 13 contributed 3, 6 contributed 4, and 2 and 1 contributed 5 and 6, respectively. The minimal model of glucose disposal and the glucose and insulin values from the intravenous glucose tolerance tests were used to estimate S(I) and S(G). The intraclass correlation coefficient was used to compare the within-family variability of S(I) and S(G) with the respective between-family distributions. The intraclass correlation coefficients were 0.26 (P = 0.008) for S(I) and 0.081 (P = 0.45) for S(G). S(I) and S(G) were uncorrelated (r = -0.059, P = 0.42). The intraclass correlation of S(I) could not be explained by familial clustering of fasting insulin or ideal body weight. Finally, the 10 families with the lowest values of S(I) had a significantly higher within-sibship variability of S(I) than the other 33 families (P < 0.001, F test). S(I) but not S(G) showed familial clustering, which is consistent with a polygenic determinant of S(I). In addition, a large within-family variability of S(I) in some families is compatible with a major gene eff ect with a dominant mode of inheritance. Thus, defects in genes affecting S(I) are plausible candidates for determinants of familial clustering of non-insulin-dependent diabetes mellitus.