OBJECTIVE: Familial dysbetalipoproteinemia (FD) or Type III hyperlipoproteinemia is closely associated with the ε2ε2 genotype of the common APOE polymorphism although not all ε2 homozygotes develop FD indicating that additional factors play a role including insulin resistance (IR). The current study was undertaken to explore relationships and influences among factors, especially IR, that might elucidate FD progression pathways. METHODS: Bayesian network (BN) modeling, a probabilistic graphical exploratory data analysis tool that portrays relationships and influences among variables as simple diagrams, was applied to 52 e2e2 subjects. An algorithm based on apolipoprotein and lipid values identified 24 subjects having FD. BN modeling parameters included plasma apoE, HDL cholesterol (HDL-C), apolipoprotein A-I (apoA-I), apolipoprotein A-II (apoA-II), apoA-I/HDL-C ratio, apoA-II/HDL-C ratio, insulin, and Homeostatic Model Assessment of Insulin Resistance (HOMA-IR). RESULTS: Modeling resulted in twenty network graphs. Each graph revealed apoE and the apoA-II/HDL-C ratio as sole determinants of FD prevalence. BN results did not demonstrate a direct role for insulin and HOMA-IR. However, multiple graphs in the set did reveal indirect influence of IR on FD prevalence as conveyed through the apoA-II/HDL-C ratio; while all remaining graphs in the set demonstrated the apoA-II/HDL-C ratio as directly influencing insulin levels and HOMA-IR. For apoE, the other determinant of FD prevalence, results revealed no relationship with IR parameters. CONCLUSIONS: In so far as insulin levels and HOMA-IR are associated with IR in e2e2 subjects, IR may act indirectly in FD progression via the apoA-II/HDL-C ratio; and/or the apoA-II/HDL-C ratio acts directly to promote IR.