Metabolically obese but normal-weight (MONW) individuals present metabolic disturbances typical of obese individuals. Additionally, metabolically healthy but obese (MHO) individuals have been identified who are relatively insulin sensitive and have a favorable cardiovascular risk profile. We compared insulin secretion patterns of MONW and MHO with those of two age-matched groups comprising nonobese individuals or obese insulin-resistant subjects, respectively. To this end, 110 nonobese subjects and 87 obese subjects were stratified into quartile based on their insulin-stimulated glucose disposal (MFFM). Insulin secretion was estimated by acute insulin response (AIR) during an intravenous glucose-tolerance test (IVGTT), and the disposition index was calculated as AIR × MFFM. We found that, as defined, MFFM was lower in MONW, who exhibited higher triglycerides, free-fatty acid (FFA), and 2-h postchallenge glucose levels compared to normal nonobese group. Insulin secretion was higher in MONW than in normal nonobese subjects, but disposition index was lower in MONW. Disposition index did not differ between MONW and insulin-resistant obese. MFFM was higher in MHO who exhibited lower waist circumference, blood pressure (BP), triglycerides, FFA, insulin levels, and higher high-density lipoprotein (HDL) cholesterol compared to insulin-resistant obese. Insulin secretion did not differ between insulin-resistant obese and MHO, but disposition index was lower in the former group. In conclusion, MONW and insulin-resistant obese showed decreased compensatory insulin secretion compared to normal nonobese and MHO subjects, respectively. Because these subjects also exhibited a worse metabolic risk profile, these findings may account for their increased risk for type 2 diabetes.
Obesity and type 2 diabetes are serious threats to public health that are reaching epidemic proportions worldwide (1,2,3,4). Increased BMI has been associated with metabolic and cardiovascular risk factors including type 2 diabetes, hypertension, dyslipidemia, but there is increasing evidence that sub-phenotypes of obesity exist that appear to deviate from the apparent linear relationship between increased BMI and its adverse clinical outcomes (5,6,7,8,9,10). Obese subjects, referred to as metabolically healthy but obese (MHO) individuals, have been identified who, despite having BMI exceeding 30 kg/m2, are relatively insulin sensitive and have a rather favorable cardiovascular risk profile (5,6,7,8,9,10,11,12,13). On the other hand, metabolically obese but normal-weight (MONW) individuals who, despite having a normal-weight BMI, present metabolic disturbances typical of obese individuals including impaired insulin sensitivity, increased visceral adiposity, low levels of high-density lipoprotein (HDL) cholesterol, elevated levels of fasting glucose and triglycerides, and hypertension have been described (8,14,15,16,17,18,19,20,21). Recently, it has been shown that the MONW-like phenotype is associated with a three- to fourfold higher risk for type 2 diabetes as compared with control subjects, and that the MHO phenotype is associated with a three- to fourfold lower risk for type 2 diabetes as compared with obese insulin-resistant subjects (22). For obesity and insulin resistance to be associated with type 2 diabetes, pancreatic β-cells must be unable to compensate for reduced insulin sensitivity (23,24). Although MONW individuals have been recognized as one “at high risk” category and MHO as one “at low risk” category for type 2 diabetes, there is little or no evidence, to the best of our knowledge, on the relationship between insulin secretion and insulin sensitivity in these subjects. We hypothesized that MONW individuals, who are at high risk of developing type 2 diabetes even when their glucose levels are still normal, would have β-cell dysfunction as well as that MHO individuals would have a higher capability to compensate for the enhanced insulin demand compared with obese insulin-resistant individuals at high risk of developing type 2 diabetes.
In this study, we identified a group of MONW individuals and a group of MHO individuals from a cohort of nondiabetic Italian whites on the basis of the insulin sensitivity determined by the hyperinsulinemic-euglycemic clamp technique, and compared their clinical characteristics including insulin secretion, with those of two age-matched groups comprising nonobese individuals or obese insulin-resistant subjects, respectively.
Methods and Procedures
The study group consisted of 110 nonobese individuals (BMI ≤ 25 kg/m2) and 87 obese individuals (BMI ≥ 30 kg/m2). Subjects were included in the study if they met the following criteria: age ranging from 19 to 54 years, absence of diabetes mellitus or impaired glucose tolerance (fasting plasma glucose <126 mg/dl, and 2-h postchallenge glucose <140 mg/dl), and absence of known inflammatory disease or pathologies able to modify glucose metabolism.
On the first day, after 12-h fasting, all subjects underwent anthropometrical evaluation including BMI, waist circumference, and body composition evaluated by bioelectrical impedance. Readings of clinic blood pressure (BP) were obtained in the supine patients, after 5 min of quiet rest, with a mercury sphygmomanometer. Values were calculated as the average of the past two of three consecutive measurements obtained at 3-min intervals. A 75-g oral glucose-tolerance test (OGTT) was performed with 0, 30, 60, 90, and 120 min sampling for plasma glucose, and insulin. On the second day, individuals were subjected to intravenous GTT (IVGTT) and to hyperinsulinemic-euglycemic clamp. After a 12-h overnight fast, a bolus of glucose (300 mg/kg in a 50% solution) was given (within 30 s) into the antecubital vein to increase the blood glucose level acutely. Samples for the measurement of blood glucose and plasma insulin were drawn at 2, 4, 6, 8, and 10 min. Acute insulin response (AIR) was calculated as the mean plasma insulin concentration during the first 10 min after glucose administration by the trapezoidal method. The degree of insulin sensitivity was evaluated using the euglycemic hyperinsulinemic clamp technique. Subjects received a priming dose of insulin (Humulin; Eli Lilly, Indianapolis, IN) during the initial 10 min to raise the serum insulin concentration acutely to the desired level (mean 102 ± 20 μU/ml), where it was maintained by continuous insulin infusion fixed at 40 mU/m2 × min. The blood glucose level was maintained constant at 90 mg/dl for the next 120 min by infusing 20% glucose at varying rates according to blood glucose measurements performed at 5-min intervals (mean coefficient of variation of blood glucose was <4%). Fat-free mass (FFM) was measured using electrical bioimpedance. Glucose disposal (M) was calculated as the mean rate of glucose infusion measured during the past 60 min of the clamp examination (steady-state) and is expressed as milligrams per minute per kilogram FFM (MFFM). Glucose disposal was also normalized for steady-state plasma insulin during the past 60 min of the insulin clamp (mg × min−1 × kg FFM−1 × μU × ml−1). To evaluate β-cell function, the so-called disposition index was calculated as AIR × MFFM (22,23).
The protocol was approved by the ethical committee and informed written consent was obtained from all participants. All the investigations were performed in accordance with the principles of the Declaration of Helsinki.
Plasma glucose was measured in duplicate by the glucose oxidation method (Beckman Glucose Analyzer II; Beckman Instruments, Milan, Italy). Total and HDL cholesterol concentrations and triglycerides were measured by enzymatic methods (Roche Diagnostics GmbH, Mannheim, Germany). Loe-density lipoprotein (LDL) cholesterol level was calculated by the Friedewald formula: total cholesterol—HDL cholesterol—(triglycerides/5). Plasma insulin concentration was determined by radioimmunoassay (Adaltis, Montreal, Quebec, Canada).
Continuous data are expressed as means ± s.d. and median. Differences of continuous variables between two groups were compared using unpaired Student's t -test. One-way ANOVA was used to compare differences of continuous variables across the four groups of individuals with Bonferroni correction for multiple comparisons. Categorical variables were compared by χ2-test. A P value <0.05 was considered statistically significant. All analyses were performed using SPSS software programme Version 12.0 for Windows (SPSS, Chicago, IL).
Because values for insulin-stimulated glucose disposal are distributed continuously in the general population, there is no objective definition of insulin resistance. Therefore, the 110 nonobese subjects and the 87 obese subjects who constitute the present study population were separately stratified into quartiles based on their MFFM values, according to previous studies (11,21). In the obese group, subjects were defined as being MHO if their MFFM value was in the upper quartile (MFFM = >12.3 mg/min × kg FFM; n = 22), whereas obese subjects with MFFM values in the two lower quartiles (MFFM = <8.7 mg/min × kg FFM; n = 43) were defined as being insulin-resistant obese. Analogously, in nonobese group, subjects were defined as being MONW if their MFFM value was in the lower quartile (MFFM = <10.2 mg/min × kg FFM; n = 27), whereas nonobese subjects with MFFM values in the two upper quartiles (MFFM = >12.3 mg/min × kg FFM; n = 55) were defined as being normal nonobese subjects. The clinical characteristics and laboratory data for the four study groups are shown in Table 1.
Table 1. Anthropometric and biochemical characteristics of the study subjects
Comparison between MONW and normal nonobese subjects
No differences in age, BMI, waist circumference, FFM, fat mass, systolic BP (SBP), diastolic BP (DBP), total cholesterol, HDL and LDL cholesterol, fasting glucose, and insulin levels were observed between MONW and normal nonobese subjects (Table 1). By design, insulin-stimulated glucose disposal was lower in the MONW subjects who also exhibited significantly higher triglycerides, free-fatty acids (FFAs), and 2-h postchallenge plasma glucose levels. Insulin secretion, measured as AIR during an IVGTT, was higher in MONW as compared with normal nonobese subjects, although this difference was of marginal significance after Bonferroni correction for multiple comparisons (P = 0.07). As insulin secretion is dependent on actual insulin sensitivity, we compared the disposition index, calculated as the product of the insulin-stimulated glucose disposal and AIR index between the two groups. The disposition index was significantly lower in the MONW group as compared with normal nonobese subjects even after correction for gender, age, and fat mass (P < 0.03, Figure 1).
Comparison between MHO and insulin-resistant obese individuals
No differences in age, BMI, FFM, fat mass, SBP, total and LDL cholesterol, fasting and 2-h postchallenge glucose were observed between MHO, and insulin-resistant obese individuals. By design, insulin-stimulated glucose disposal was higher in the MHO subjects who also exhibited significant lower waist circumference, DBP, triglycerides, FFAs, fasting insulin levels, and higher HDL cholesterol levels. Insulin secretion during an IVGTT did not differ between MHO and insulin-resistant obese subjects. The disposition index was lower in the insulin-resistant obese group as compared with the MHO subjects even after correction for gender, age, and fat mass (P < 0.003, Figure 1) or Bonferroni correction for multiple comparisons (P = 0.001).
Comparison between MHO and normal nonobese subjects
We next compared MHO and normal nonobese subjects who display similar insulin-stimulated glucose disposal although having, by definition, marked differences in BMI, fat mass, and waist circumference. No differences in age, total cholesterol, FFAs, fasting plasma glucose and insulin levels were observed between MHO and normal nonobese subjects. MHO subjects exhibited significant higher SBP, DBP, LDL cholesterol, triglycerides, 2-h postchallenge plasma glucose, and lower HDL cholesterol levels (Table 1). Insulin secretion during an IVGTT was higher in the MHO subjects as compared with normal nonobese subjects even after Bonferroni correction for multiple comparisons (P = 0.01). The disposition index was higher in the MHO group as compared with normal nonobese subjects even after correction for gender, age, and fat mass (P < 0.001, Figure 1) or Bonferroni correction for multiple comparisons (P = 0.0001). The differences in AIR and disposition indexes were not statistically significant after adjustment for BMI, thus indicating that the higher insulin secretion in MHO individuals is because of an adaptative mechanism to compensate for increased body surface demand.
Comparison between MONW and insulin-resistant obese individuals
Finally, we compared MONW and insulin-resistant obese individuals, the two groups which display lower insulin-stimulated glucose disposal. By design, BMI, FFM, fat mass, and waist circumference were lower in the MONW subjects who also exhibited significant higher insulin-stimulated glucose disposal (Table 1). No differences in total and LDL cholesterol, FFAs, insulin secretion during an IVGTT, fasting and 2-h postchallenge plasma glucose levels were observed between MONW and insulin-resistant obese individuals. MONW subjects exhibited significant lower SBP, DBP, triglycerides and fasting insulin levels, and higher HDL cholesterol levels. No differences in the disposition indexes were observed between MONW and insulin-resistant obese individuals even after correction for gender, age, and fat mass.
Type 2 diabetes results from a combination of insulin resistance at the level of skeletal muscle and adipose tissue, increased hepatic glucose production, and failure of pancreatic β-cells to compensate for the enhanced insulin demand (23,24). There is evidence that the MONW-like phenotype is associated with a higher risk for type 2 diabetes as compared with control nonobese individuals, whereas the MHO-like phenotype is associated with a lower risk for type 2 diabetes as compared with insulin-resistant obese subjects (22). Because pancreatic β-cell dysfunction is a sine qua non feature for the development of type 2 diabetes, we compared insulin secretion indexes between a group of MONW individuals and a group of MHO individuals with those of two age-matched groups comprising nonobese individuals or obese insulin-resistant subjects, respectively. We observed that first-phase glucose-stimulated insulin secretion did not differ between MONW and normal nonobese subjects. Because the amount of insulin secreted by the β-cell is strongly dependent on the prevailing degree of insulin sensitivity, assessment of β-cell function corrected for differences in insulin sensitivity is a critical point when evaluating β-cell function (23,24,25,26). Thus, adjusting insulin secretion for the level of insulin sensitivity using the disposition index (insulin sensitivity × insulin secretion) may be a better measure of β-cell function (23,24,26). Using this approach, we found that MONW subjects showed an impaired compensatory first-phase insulin secretion as compared with normal nonobese individuals. Because these subjects also exhibited a worse metabolic risk profile as compared with normal nonobese subjects (21), these findings may account for the increased risk for type 2 diabetes that has been associated with a MONW-like phenotype (22). Interestingly, no differences in the disposition index were observed between MONW and insulin-resistant obese individuals, two groups of insulin-resistant subjects at high risk of developing type 2 diabetes (22). These data are consistent with those observed in other groups at increased risk of developing type 2 diabetes including women with a history of gestational diabetes or polycystic ovarian syndrome, and offspring of type 2 diabetic patients (27,28,29,30). The inability of either MONW or insulin-resistant obese individuals to adequate their insulin secretion to the degree of insulin insensitivity raises the possibility of a link between the two pathophysiological processes. Traditionally, these two defects have been considered as discrete pathophysiological lesions; however, over the past several years, evidence has accumulated indicating that the insulin-signaling cascade may play an important functional role in pancreatic β-cell function (31). Thus, the two main pathophysiological features of type 2 diabetes should no longer be viewed as two discrete phenomena, but rather as different facets of the same molecular defects.
We also provide evidence that MHO subjects have a better metabolic risk profile and higher disposition index as compared with obese insulin-resistant individuals. After adjustment for BMI, MHO subjects exhibited a β-cell function comparable to normal nonobese subjects. The present data may help to explain the lower risk for type 2 diabetes observed in the MHO-like phenotype as compared with obese insulin-resistant subjects. Furthermore, these results suggest that obesity per se does not lead to an impairment in β-cell function, which may be due to other metabolic and hormonal factors. A metabolic derangement that might contribute to defects in compensatory β-cell response observed in MONW and obese insulin-resistant subjects is elevated plasma FFA concentrations. Although FFAs are important for normal insulin secretion, chronic exposure to elevated concentrations of FFAs in vitro and in vivo is associated with a marked impairment in glucose-stimulated insulin release (32,33,34). In addition, elevation in FFA levels induced by a lipid infusion in vivo contribute to the development of both insulin resistance and impaired compensatory β-cell response, a phenomenon referred to as “lipotoxicity” (35,36). Accordingly, we observed that both MONW and obese insulin-resistant subjects have increased levels of plasma FFA as compared with normal nonobese individuals and MHO individuals, respectively, thus making the lipotoxicity hypothesis plausible. On the other hand, it is also possible that impaired insulin secretion resulting in lower insulin concentrations may contribute to increase FFA levels as a consequence of reduced anti-lipolytic effect of insulin. Obviously, we cannot exclude that other factors such as the so-called “adipokines” may play an important role in defects in compensatory β-cell response observed in MONW and obese insulin-resistant subjects.
There are a number of potential limitations to our study. First, insulin-mediated glucose disposal is a continuous trait, and, therefore, the present results are based on an operational definition of low insulin sensitivity that is arbitrary given the lack of a generally agreed upon cut-point for defining low insulin sensitivity. On the other hand, there is evidence from prospective studies suggesting a statistically significant increase in clinical outcomes in individuals in the lowest quartile of insulin sensitivity (11,37), and the lowest quartile of insulin sensitivity has been included by some expert groups among the diagnostic criteria for metabolic syndrome (38,39). Second, we used a cross-sectional approach, which does not allow us to make any causal associations between impaired β-cell function and incidence of type 2 diabetes in our cohort. Despite this limitation, our results are strengthened by using gold standard techniques as well as by studying a well-characterized cohort with the measurement of various metabolic variables. In addition, insulin sensitivity and glucose-stimulated insulin secretion were determined only once; therefore, intra-individual variation in levels of these variables cannot be taken into account. Finally, all subjects in this study were whites. Thus, these results cannot be readily generalized to other ethnic groups because differences between ethnic groups in insulin-resistance and β-cell function have been reported in a number of studies. This cross-sectional study should be considered hypothesis generating and requiring confirmation by both cross-sectional and prospective studies in other ethnic populations.
This study was supported by the European Community's FP6 EUGENE2 no. LSHM-CT-2004-512013 grant (to G.S.).