Apolipoprotein E isoform polymorphisms are not associated with insulin resistance: the Framingham Offspring Study

Insulin resistance and the apolipoprotein (apo) allele e4 have both been associated with coronary heart disease (CHD). We examined the relationship between insulin resistance and apo(e) polymorphisms among participants in the Framingham Offspring Study.

During 1991-1995, subjects underwent a clinical examination and an oral glucose tolerance test with measurement of fasting and 2-h glucose, insulin levels, and fasting lipid levels. We measured insulin resistance using the homeostasis model, in which insulin resistance (HOMA-IR) = fasting insulin x glucose/22.5. Apo(e) isoforms and phenotypes were determined in 1983-1987 using isoelectric focusing of plasma VLDL. Of the 2,120 subjects with complete HOMA-IR and apo(e) data, 204 with type 2 diabetes were excluded. The remainder were classified with features of the insulin resistance syndrome including impaired glucose tolerance (1997 American Diabetes Association criteria), hypertension (criteria from the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure [6th report]), obesity (BMI >85th percentile), high waist-to-hip ratio (>85th percentile), and high triglyceride and low HDL levels (NCEP-2 criteria). We analyzed data with contingency tables and age- and sex-adjusted logistic regression models.

Among the 1,916 subjects, the mean age was 55 (range 28-83 years) and 51% were women. Median HOMA-IR was 6.4 (interquartile range 5.2-8.2), and allele frequencies were 7.8, 79.9, and 12.4% for apo(e) alleles e2, e3, and e4, respectively There were no differences in proportions of apo(e) isoforms or alleles across increasing quintiles of HOMA-IR. A less dramatic increase in proportions occurred with elevated triglycerides associated with increasing HOMA-IR among those with apo(e) isoforms 2/2 and 2/3 compared with the others (P < or = 0.01 for interaction). Otherwise, apo(e) did not substantially modify associations between insulin resistance and features of the insulin resistance syndrome.

There is no association between apo(e) polymorphisms and insulin resistance. These appear to represent 2 completely independent risk factors for CHD.