Objective: 1) Determine in a sample of American Indians (AI) how well insulin sensitivity (SI) measured by the frequently sampled intravenous glucose test (FSIGT) correlates with a simpler measure of insulin resistance (IR) measured by the homeostasis assessment (HOMA) model; (2) compare insulin sensitivity in a sample of diabetic and non-diabetic Al in the Strong Heart Study (SHS) with that of White, Black, and Hispanic Americans in the Insulin Resistance Atherosclerosis Study (IRAS).
Design: Cross sectional
Setting: Community
Participants: Sixty-one Al participants in SHS MAIN OUTCOME MEASURES: Mean SI measured by FSIGT, a complex protocol to evaluate insulin sensitivity, and mean IR measured by the HOMA model, a method based on measures of fasting glucose and fasting insulin.
Results: Although 70% of sample participants were non-diabetic, only 18% were insulin sensitive by SI. Diabetes status strongly confounded Si among Al in SHS. At non-diabetic levels of fasting glucose (< 126 mg/dL), SI correlated well with HOMA IR (rho = -0.49, P = .0009), but SI did not reflect HOMA IR at levels of fasting glucose that are diagnostic of diabetes (> or = 126 mg/dL; rho = -0.13, P = n.s.). With the exception of some Hispanic participants in IRAS, mean SI of non-diabetic Al in SHS was lower than that of their non-diabetic IRAS counterparts. Diabetic Al participants in SHS had markedly lower mean SI than all diabetic participants in IRAS.
Conclusions: The HOMA model may be a useful tool to identify non-diabetic American Indians who might benefit from early CVD risk factor modification.