Decreased Non–Insulin-Dependent Glucose Clearance Contributes to the Rise in Fasting Plasma Glucose in the Nondiabetic Range

The participants included 120 normal healthy subjects (67 female and 53 male, aged 38 ± 1 years, BMI 29.1 ± 0.6 kg/m², FPG 93 ± 1 mg/dl, 2-h plasma glucose 115 ± 2, FPI 9 ± 1 μU/ml, and 2-h plasma insulin 65 ± 01). All subjects were of Mexican-American origin and had a normal (75-g) oral glucose tolerance test (OGTT) (FPG <126 mg/dl and 2-h plasma glucose <140 mg/dl).

All subjects had normal liver, cardiopulmonary, and kidney function as determined by medical history, physical examination, screening blood tests, electrocardiogram, and urinalysis. No subject was taking any medication known to affect glucose tolerance. Body weight was stable (±2 kg) for at least 3 months before study in all subjects. The study protocol was approved by the institutional review board of the University of Texas Health Science Center, San Antonio, Texas, and written informed consent was obtained from all subjects before participation. All studies were performed at the general clinical research center of the University of Texas Health Science Center at 0800 h following a 10- to 12-h overnight fast.

Before the start of the OGTT, a small polyethylene catheter was placed into an antecubital vein and blood samples were collected at −30, −15, 0, 30, 60, 90, and 120 min for the measurement of plasma glucose and insulin concentrations. On the day of the OGTT, lean body mass was measured with dual energy X-ray absorptiometry.

Before the start of the insulin clamp, a catheter was placed into an antecubital vein for the infusion of all test substances. A second catheter was inserted retrogradely into a vein on the dorsum of the hand, and the hand was placed into a thermoregulated box heated to 70°C. At 0800 h, all subjects received a primed (25 μCi)-continuous (0.25 μCi/min) infusion of 3-[³H]glucose (DuPont NEN Life Science Products, Boston, MA), which was continued for the 4-h duration of the study. Two hours after the start of tritiated glucose, subjects received a primed-continuous insulin infusion at the rate of 240 pmol (40 mU) · min⁻¹ · m⁻² for 120 min. During the last 30 min of the basal equilibration period (90–120 min), blood samples were taken at 5- to 10-min intervals for the determination of plasma glucose and insulin concentrations and tritiated glucose radioactivity. During the insulin infusion, plasma glucose concentration was measured every 5 min, and a variable infusion of 20% glucose was adjusted, based on the negative feedback principle, to maintain the plasma glucose concentration at each subject's FPG level with a coefficient of variation <5%. Blood samples were collected every 15 min from 120 to 210 min and every 5–10 min from 210 to 240 min for the determination of plasma glucose and insulin concentrations and tritiated glucose radioactivity.

Following an overnight fast, steady-state conditions prevail, and endogenous (primarily reflecting hepatic) glucose production (HGP) was calculated as the tritiated glucose infusion rate (disintegrations per minute) divided by the plasma tritiated glucose specific activity (disintegrations per minute per milligram). During the insulin clamp, non–steady-state conditions for tritiated glucose specific activity prevail, and the rate of glucose appearance (R_a) was calculated with Steele's equation (12). The rate of residual HGP during the insulin clamp was calculated by subtracting the rate of exogenous glucose infusion from the tracer-derived R_a. The insulin-stimulated rate of total glucose disposal was calculated by adding the rate of residual HGP to the exogenous glucose infusion rate. Plasma glucose clearance rate during the fasting state was calculated as the rate of HGP divided by FPG concentration. Hepatic insulin resistance was calculated as the product of HGP and FPI (13).

All values are expressed as means ± SEM. ANOVA was used to compare the difference between the means of the four quantiles. Simple Pearson's correlation was used to assess the relationship between variables. For multivariate regression, FPG was considered the dependent variable, and BMI, FPI, HGP, and glucose clearance were considered independent variables. Non-normally distributed variables (i.e., BMI) were log transformed. Statistical significance was considered at P < 0.05.

The increase in FPG over the nondiabetic range of FPG levels (75–125 mg/dl) in the present study correlated positively (r = 0.32, P < 0.0001) with the increase in BMI (range 20–50 kg/m²) (Fig. 1A). However, over this same range, HGP decreased with the increase in BMI (r = −0.51, P < 0.0001) (Fig. 1B). We also examined the relationship between HGP, expressed per lean body mass, and BMI. The negative relationship between HGP and BMI remained when HGP was expressed per lean body mass (r = −0.31, P = 0.004).

To further examine the relationships between obesity, FPG, and HGP, subjects were divided into four quartiles. The metabolic and anthropometric characteristics of subjects in the four quartiles are shown in Table 1. The four groups were comparable in age and sex, but FPG progressively increased from quartiles 1–4 (P < 0.001 with ANOVA). However, HGP, whether expressed per total body weight or lean body mass, progressively decreased (not increased) (P < 0.001 with ANOVA).

Since FPI is the primary regulator of HGP, we examined the relationship between FPI and BMI. As expected, FPI concentration progressively increased in subjects in quartiles 1–4 (P < 0.0001) (Table 1). FPI correlated strongly and positively with BMI (r = 0.62, P < 0.0001; Fig. 1C) and negatively with HGP (r = −0.34, P < 0.0001). The product of HGP and FPI, an index of hepatic insulin resistance, progressively increased from quartiles 1–4 (P = 0.0002) and correlated closely with BMI (r = 0.52, P < 0.0001; Fig. 2A).

However, despite the marked increase in FPI, the glucose clearance rate was decreased with increasing BMI in quartiles 1–4 (P < 0.0001), and the plasma glucose clearance rate correlated inversely with BMI (r = −0.56, P < 0.0001; Fig. 2B).

To assess the contribution of obesity to the relationship between FPG and glucose clearance, we compared the correlation coefficient between the two variables in lean (BMI <27 kg/m², n = 48) and overweight/obese subjects (BMI >27 kg/m², n = 72). The correlation coefficient was −0.43 in lean subjects and −0.27 in obese subjects. In a multivariate model, using FPG as the dependent variable and BMI, FPI, HGP, and glucose clearance as the independent variables, all four independent variables significantly correlated with FPG, which was explained by the following equation:

Collectively, BMI, HGP, FPI, and glucose clearance explained ∼60% of the variability in FPG, where more than half of this was explained by glucose clearance and basal HGP.

When related to FPG, both HGP (r = 0.17, P = 0.05) and glucose clearance rate (r = −0.42, P < 0.0001) displayed a negative correlation (Fig. 3A and B). A positive correlation between the glucose clearance rate and total body glucose disposal during the insulin clamp (r = 0.39, P < 0.0001) was observed.

The results of the present study demonstrate that in nondiabetic subjects, the increase in BMI is associated with an increase in FPG and, paradoxically, with a decrease in HGP. Because HGP is the main contributor to the elevated FPG concentration in type 2 diabetic subjects (8), one might have expected that the increase in FPG observed with increasing BMI would be associated with a rise in HGP. However, the results of this study demonstrate the opposite. The inverse relationship between FPG and HGP is most striking when one compares subjects in the highest BMI quartile with subjects in the lowest BMI quartile (Table 2), and it excludes the possibility that an increase in HGP is responsible for the increase in FPG in the nondiabetic range. Because under postabsorptive conditions steady-state conditions exist with respect to the FPG concentration, the elevated FPG concentration in obese subjects must be explained by a decrease in tissue glucose clearance. Indeed, when the glucose clearance in obese subjects is compared with that in lean subjects, there is a 34% decrease in glucose clearance rate. Furthermore, the glucose clearance rate correlates negatively with the increase in BMI. These results indicate that the increase in FPG, which accompanies the increase in BMI, primarily results from the decline in glucose clearance and not from excess production of glucose by the liver.

Obese subjects, as expected, had a 2.5-fold increase in FPI concentration compared with lean individuals, and the FPI concentration rose progressively with increasing BMI (Table 1, Fig. 1C). The plasma insulin concentration is the main regulator of HGP (9,10). Thus, the rise in FPI with increasing BMI leads to a progressive decrease in HGP from quartiles 1–4, and HGP was strongly and inversely correlated with the FPI (r = −0.34, P < 0.0001).

Obesity per se seems has a small effect on FPG. Consistent with this, the inverse relationship between FPG and glucose clearance also was observed in lean subjects (BMI <27 kg/m²). Further, the contribution of BMI (multivariate analysis) to the increase in FPG was much smaller compared with the contributions of glucose clearance and HGP (see equation 1), indicating that the primary determinants of FPG are the HGP and tissue glucose clearance. Moreover, the impact of obesity to increase the FPG is due primarily to the decrease in tissue glucose clearance.

We previously have shown that the basal insulin secretion rate increases with the increase in FPG in the nondiabetic range (14). However, when the insulin secretory rate (ISR) was related to FPG, the ratio of ISR to FPG remained constant across the entire nondiabetic range of FPG levels (14). These results indicate that 1) although glucose-stimulated insulin secretion is markedly impaired with increasing FPG, basal insulin secretion is not affected by fasting hyperglycemia, and 2) fasting hyperinsulinemia is a compensatory β-cell response to fasting hyperglycemia. The resultant fasting hyperinsulinemia that accompanies fasting hyperglycemia inhibits HGP and explains the present observation that HGP declines as FPG increases within the nondiabetic range. Thus, the decrease in HGP associated with the increase in BMI can be viewed as a compensatory physiological response to fasting hyperglycemia, which aims to ameliorate the rise in FPG. However, the increased hepatic insulin resistance in obese subjects, which also strongly correlates with increasing BMI, renders the liver more resistant to the action of insulin and results in an incomplete suppression of HGP.

We previously have shown that the increase in insulin secretion rate in response to the rise in FPG peaks at an FPG concentration of ∼140 mg/dl and declines thereafter (15). This results in an inverted U-shaped curve relating the FPI and FPG concentrations (15). It is noteworthy that the increase in HGP becomes evident only when FPG exceeds 140–160 mg/dl (8,15). Collectively, these observations indicate that at FPG <140 mg/dl, the increase in FPG primarily is due to a decrease in tissue glucose clearance. Once the FPG exceeds ∼140–160 mg/dl, the decline in insulin secretion, in the presence of hepatic insulin resistance, results in an increase in HGP resulting in a rise in FPG in type 2 diabetic individuals.

A decrease in non–insulin-dependent glucose clearance previously has been reported in subjects with type 2 diabetes (16). During the postabsorptive state, ∼50% of glucose uptake occurs in the brain, ∼25% in the splanchnic tissues, and ∼25% in skeletal muscle. Glucose uptake by the brain and splanchnic tissues is insulin independent (1). Therefore, it is unlikely that decreased brain or splanchnic glucose uptake can account for the decline in glucose clearance observed with rising FPG levels in the present study. Glucose uptake in skeletal muscle occurs via both insulin-dependent (GLUT4) and insulin-independent (GLUT1) mechanisms (17). Skeletal muscle is well known to be resistant to the action of insulin (18), and this insulin resistance could contribute, in part, to the decline in basal tissue glucose clearance observed in the present study. Similarly, a decrease in GLUT1, the insulin-independent glucose transporter, in skeletal muscle, could contribute to the decrease in tissue glucose clearance observed in normal glucose-tolerant obese subjects in the present study. One previous study has reported a decrease in the amount of GLUT1 protein in skeletal muscle in subjects with type 2 diabetes compared with healthy control subjects (19,20). It is unclear whether the decrease in GLUT1 expression in subjects with type 2 diabetes represents a primary defect or occurs secondary to hyperglycemia. It is also unclear at what FPG level the decrease in GLUT1 expression becomes evident. Nonetheless, a decrease in GLUT1 expression in subjects with elevated FPG concentrations within the nondiabetic range could explain, in part, the decrease in glucose clearance observed with rising FPG in the present study.

Insulin-stimulated glucose uptake in skeletal muscle correlated well with non–insulin-dependent glucose clearance during the fasting state. Thus, it is possible that the same defect responsible for the impairment in insulin-stimulated glucose uptake could explain the decrease in basal glucose clearance, since the majority (>80%) of glucose disposal during the euglycemic insulin clamp occurs in muscle (21).

In summary, the results of the present study demonstrate that the decrease in non–insulin-dependent glucose clearance is the primary factor that contributes to the increase in FPG concentration within the nondiabetic range. The decline in basal HGP observed with rising BMI is explained by the increase in FPI concentration that represents a compensatory response to the obesity-related insulin resistance.