Relationship of hepatic and peripheral insulin resistance with plasminogen activator inhibitor-1 in Pima Indians

Abstract

Plasminogen activator inhibitor-1 (PAI-1) is related to insulin resistance and several components of the insulin resistance syndrome, and PAI-1 levels are elevated in subjects with non—insulin-dependent diabetes mellitus. Many Pima Indians are obese, insulin-resistant, and hyperinsulinemic, and they have high rates of diabetes but a low risk of ischemic heart disease. In contrast to whites and Asians, PAI-1 activity is similar between nondiabetic and diabetic Pima Indians. We therefore examined the association of PAI-1 with hepatic and peripheral insulin action measured using the hyperinsulinemic-euglycemic clamp. To investigate if insulin per se has any effect on PAI-1 in vivo, we also assessed the effects of endogenous (during a 75-g oral glucose load) and exogenous (during hyperinsulinemic clamp) insulin on PAI-1 antigen. Twenty-one (14 men and seven women; mean age, 26.3 ± 4.8 years) Pima Indians underwent a 75-g oral glucose tolerance test (OGTT) and a sequential hyperinsulinemic-euglycemic clamp. Peripheral insulin action was measured as absolute glucose uptake (M value) and normalized to estimated metabolic body size (EMBS). Hepatic insulin action was measured as percent suppression of basal hepatic glucose output during hyperinsulinemia. PAI-1 antigen was determined using a two-site enzyme-linked immunosorbent assay that detects only free PAI-1. PAI-1 antigen concentrations were significantly related to body mass index ([BMI] r s = .54, P = .012), waist ( r s = .52, P = .016) and thigh ( r s = .63, P = .002) circumference, and fasting plasma insulin concentration ( r s = .59, P = .004). PAI-1 antigen concentrations were not significantly associated with peripheral glucose uptake (M value) during either low-dose ( r s = −.01, P = NS) or high-dose ( r s = −.11, P = NS) insulin infusion. PAI-1 antigen was negatively correlated with basal hepatic glucose output ( r s = −.57, P = .013) and percent suppression of hepatic glucose output during hyperinsulinemia ( r s = −.69, P = .005). However, this relationship was largely due to the confounding effects of BMI, waist and thigh girth, fasting insulin, and 2-hour postload glucose concentrations, and was not significant when controlled for these variables (partial r s = −.30, P = NS). There was no significant relationship of PAI-1 antigen concentration with glucose storage or glucose oxidation. Despite a threefold increase in plasma insulin concentrations during the OGTT, there were no significant changes in PAI-1 antigen concentrations (median, 57, 61, 55, and 44 ng/mL at 0, 60, 120, and 180 minutes, respectively; P = NS by ANOVA). During the hyperinsulinemic clamp, mean plasma insulin concentrations at the end of low-dose (240 pmol/m 2/min) and high-dose (2,400 pmol/m 2/min) infusions were 1,005 and 14,230 pmol/L, respectively. However, PAI-1 antigen concentrations at the end of low-dose and high-dose insulin infusions were similar to those at baseline (median, 63, 43, and 58 ng/mL, respectively; P = NS by ANOVA). PAI-1 antigen in Pima Indians is related to several components of the insulin resistance syndrome. However, direct measurement of insulin resistance indicates that hepatic but not peripheral insulin resistance is related to PAI-1 antigen. Neither endogenous nor exogenous hyperinsulinemia for short periods had any significant effect on PAI-1 antigen concentrations. Short-term hyperinsulinemia is unlikely to be an important regulator of PAI-1 in Pima Indians. The relationship of PAI-1 antigen to hepatic insulin resistance is largely dependent on the relationship of PAI-1 to indices of obesity and fasting insulin concentrations.