Evidence for Linkage of Postchallenge Insulin Levels With Intestinal Fatty Acid-Binding Protein (FABP2) in Mexican-Americans

Single genes with large effects may contribute to insulin resistance or influence susceptibility to non-insulin-dependent diabetes mellitus (NIDDM). In the Pima Indians, results from sib-pair analysis have suggested that a gene on chromosome 4q influences both fasting insulin levels and maximal insulin action. We conducted sib-pair and logarithm of odds (LOD)-score linkage analysis to seek evidence for linkage between genes influencing insulin levels and chromosome 4q loci. Analyses were conducted on nondiabetic individuals from 28 different families participating in the San Antonio Family Diabetes Study. All subjects received a 2-h oral glucose tolerance test. Fasting insulin levels were measured in 382 nondiabetic individuals, and 2-h insulin levels were measured in 366 individuals. Initial sib-pair linkage analysis revealed a possible association between 2-h post–glucose challenge insulin levels and the intestinal fatty acid–binding protein (FABP2) locus located in the region of chromosome 4q28–31 (P = 0.006). Subsequent sib-pair linkage analysis of 11 additional chromosome 4q markers supported this hypothesis. We next conducted segregation analyses to estimate allele frequencies and other model parameters for the putative locus influencing 2-h insulin levels. Results of LOD-score linkage analysis indicated possible linkage between the major gene described by the segregation model and FABP2. Using combined segregation and linkage analysis, we obtained a LOD-score of 2.80 at recombination frequency of 0.0 between FABP2 and the putative locus influencing 2-h insulin levels. The maximum likelihood estimate of the allele associated with low insulin levels was 0.21. Individuals having one or two copies of this allele had a mean ln(2-h insulin level) equal to 3.484 (back-transformed mean = 298.4 pmol/1), compared with 4.480 (back-transformed mean = 807.8 pmol/1) for individuals in whom this allele was absent. Approximately 32% of the total phenotypic variance in ln(2-h insulin levels) could be attributed to this locus. These results are consistent with the hypothesis that FABP2, or a tightly linked gene, influences 2-h insulin levels. This gene may be associated with insulin resistance.