Glucose-Induced Insulin Secretion in Dyslipidemic and Normolipidemic Patients With Normal Glucose Tolerance

RESEARCH DESIGN AND METHODS

Twenty dyslipidemic subjects (13 men and 7 women) with BMI >27 kg/m² but having normal glucose tolerance, as documented by a standard oral glucose tolerance test (fasting glycemia 94 ± 7 mg/dl, 2-h glycemia 95 ± 17 mg/dl) (9), were selected for the study. They had a mean (±SD) age of 46.6 ± 7.6 years, body weight of 85.4 ± 10.3 kg, height of 173.2 ± 9.2 cm, BMI of 28.6 ± 4.1 kg/m², waist circumference of 98.9 ± 11.0 cm, waist-to-hip ratio of 0.91 ± 0.08, fasting plasma triglyceride concentration of 2.25 ± 0.84 mmol/l, fasting total plasma cholesterol concentration of 5.66 ± 1.04 mmol/l, fasting HDL cholesterol concentration of 1.06 ± 0.11 mmol/l, and blood pressure of 123 ± 13/78 ± 8 mmHg. They were compared with a group of 20 normolipemic subjects (18 men and 2 women) with normal glucose tolerance (fasting glycemia 96 ± 7 mg/dl, 2-h glycemia 95 ± 22 mg/dl) and similar age (52.6 ± 6.3 years), body weight (91.8 ± 17.2 kg), height (176 ± 10 cm), BMI (29.6 ± 3.9), waist circumference (99.6 ± 12.2 cm), waist-to-hip ratio (0.91 ± 0.07), and blood pressure (126 ± 11/80 ± 7 mmHg) but had normal triglyceride (0.80 ± 0.21 mmol/l), total cholesterol (5.03 ± 0.91 mmol/l), and HDL cholesterol (1.71 ± 0.27 mmol/l) concentrations. All subjects were recruited from the GEMS (Genetic Epidemiology of Metabolic Syndrome) project study population (10). Fourteen dyslipidemic and no normolipemic patients fulfilled the criteria for the metabolic syndrome (2). Six dyslipidemic and three normolipemic subjects were receiving treatment for high blood pressure, and two dyslipidemic patients were being treated with statins.

Each participant took part in a two-step hyperglycemic clamp (target glycemia of 135 mg/dl during 60 min and of 180 mg/dl during the next 60 min) to simultaneously evaluate glucose-induced insulin secretion and glucose/insulin-mediated glucose disposal. Participants received 2 mg/day dexamethasone during the 2 days preceding the clamp procedure and 0.5 mg on the morning of the procedure. [6,6-²H₂]glucose (bolus 2 mg/kg, continuous infusion 20 μg/kg/min for 60 min before the clamp, then exogenous glucose labeled with 1.25% [6,6-²H₂]glucose [hot infusate approach] [11]) was used to calculate whole-body glucose kinetics. The results obtained were compared by means of unpaired t tests.

RESULTS

At their inclusion in the study, dyslipidemic and normolipemic subjects had similar fasting plasma glucose (94 ± 7 vs. 97 ± 7 mg/dl) and insulin (31.5 ± 4.9 vs. 31.7 ± 7.4 mU/l). After administration of dexamethasone for 2 days (Table 1), dyslipidemic subjects had fasting glucose and insulin concentrations that were not significantly different from those of normolipemic subjects. During the clamp procedure, similar rates of exogenous glucose infusion were necessary to maintain target glycemia (Table 1). However, the rates of exogenous infusion divided by steady-state plasma insulin concentration were lower in dyslipidemic than in normolipemic subjects, indicating impaired glucose/insulin-induced glucose metabolism. Compared with normolipemic subjects, dyslipidemic subjects had 41 and 74% higher plasma insulin concentrations during the low and high plateaus of glycemia, respectively. Basal endogenous glucose production and its suppression at both plateaus of glycemia were similar in dyslipidemic and normolipemic subjects.

CONCLUSIONS

This study focused on dyslipidemic patients free of any disorder of glucose homeostasis. Such patients offer the possibility to investigate the contribution of insulin sensitivity and/or secretion in the pathogenesis of dyslipidemia. Glucose homeostasis was evaluated by a two-step hyperglycemic clamp after administration of dexamethasone, a procedure aimed at decreasing insulin sensitivity with compensatory hyperinsulinemia. The rate of exogenous glucose infusion divided by plasma insulin concentration was lower in dyslipidemic than in normolipemic subjects, indicating a lower insulin- and glucose-induced glucose disposal. Although this protocol was not performed without dexamethasone administration, these results are consistent with some degree of insulin resistance in dyslipidemic subjects. Dyslipidemic subjects also had significantly higher plasma insulin concentration at each plateau of glycemia, consistent with impaired insulin sensitivity, but this also indicates that their glucose-induced insulin secretion was not decreased. This hyperinsulinemia cannot be attributed to differences in BMI or body fat distribution, since dyslipidemic and normolipemic patients had similar characteristics. Since both groups had similar plasma glucose and insulin concentrations before dexamethasone, we propose that short-term dexamethasone administration offers a way to detect subclinical insulin resistance in dyslipidemic patients.

In conclusion, our results indicate that dyslipidemic, glucose-tolerant patients have increased glucose-induced insulin secretion to compensate for impaired insulin sensitivity. We propose that these alterations are indicative of insulin resistance in dyslipidemic patients. The chronic hyperinsulinemia consecutive to insulin resistance may possibly contribute to increase plasma triglyceride concentrations by stimulation of hepatic de novo lipogenesis and secretion of VLDL lipoproteins (12,13). Conversely, it is also possible that primary alterations of hepatic lipid metabolism, leading to hypertriglyceridemia and low HDL levels may in the long-term contribute to the development of insulin resistance and hyperinsulinemia (14,15). Only prospective studies in individuals at risk for the metabolic syndrome will provide clarification between these two hypotheses.