Background

The development of obesity, with or without insulin resistance (IR), and with or without β-cell pancreatic functional changes (insulin secretion, IS), demonstrates the extraordinary limitations of HOMA2 IS and IR models and their interpretation.

Methods

To expand understanding of the physiological basis of the relationships between adiposity, glucose, insulin, and the HOMA2 IS and IR models in spontaneously obese/overweight nonhuman primates (NHPs) we have carried out a careful longitudinal study of 92 adult monkeys (rhesus, Macaca mulatta) held under constant conditions (using consistent standardized sampling and analytic methods, including fasting insulin (FI) and glucose levels (FPG), IV glucose tolerance (IVGTT), acute insulin response to IV glucose (AIR), and euglycemic hyperinsulinemic clamps (M value calculated correcting for fat mass)).

Results

In a subset, the 18 subjects studied from the highest FI levels to overt T2DM, FI and AIR were highly correlated, with both showing major declines (r=0.25; p<0.001), but no relationship to excess adiposity. The age of onset of overt diabetes (DM) was identical between overweight (mean±SD, peak weight 15.55±0.9 kg) and obese (19.14±3.6 kg).Insulin dropped from maximal levels and IGT to low levels at T2DM (r=0.16; p<0.001), a finding in accord with the HOMA2 IS and IR models, however, these models had no applicability to the earlier period of weight gain (obesity) and increasing insulin resistance preceding IGT.

Conclusions

The HOMA model has been predicated upon an assertion of a balance between hepatic glucose production (HGP) (which is elevated exactly in parallel with the increase in fasting glucose levels) and insulin secretion (elevated years before the increase in HGP) in a hypothesized feedback loop, which was not supported by this longitudinal study. The hypothesized hyperbolic relationship between insulin secretion and insulin action was absent and physiologically implausible.