Disclosure: The authors declared non conflict of interest.
Beta and alpha cell function in metabolically healthy but obese subjects: Relationship with entero-insular axis†
Version of Record online: 26 MAR 2013
Copyright © 2012 The Obesity Society
Volume 21, Issue 2, pages 320–325, February 2013
How to Cite
Calanna, S., Piro, S., Di Pino, A., Maria Zagami, R., Urbano, F., Purrello, F. and Maria Rabuazzo, A. (2013), Beta and alpha cell function in metabolically healthy but obese subjects: Relationship with entero-insular axis. Obesity, 21: 320–325. doi: 10.1002/oby.20017
- Issue online: 26 MAR 2013
- Version of Record online: 26 MAR 2013
- Accepted manuscript online: 28 AUG 2012 10:57AM EST
- Manuscript Accepted: 18 JUN 2012
- Manuscript Revised: 18 MAY 2012
- Manuscript Received: 24 JAN 2012
Obesity is widely acknowledged as a critical risk factor for metabolic complications. Among obese subjects, there is a phenotype of metabolically healthy but obese (MHO) individuals that shows a favorable cardiometabolic risk profile. We aimed to evaluate the potential mechanisms underlying the metabolic profile of this subset, including alpha and beta cell function and entero-insular axis.
Design and Methods:
One hundred twenty-nine obese and 24 nonobese subjects were studied. Obese participants were defined as MHO or at-risk obese, according to the homeostasis model of assessment-insulin resistance (HOMA-IR) index (MHO: lower tertile of HOMA-IR, n = 43; at-risk: upper tertile of HOMA-IR index, n = 41). Insulin, glucagon, and incretin responses after a 120′ oral glucose tolerance test (75-g OGTT) were investigated.
During OGTT, MHO individuals showed in comparison with at-risk subjects: lower fasting and afterloads plasma levels of glucose, insulin, and C-peptide; higher disposition index; lower fasting (P = 0.004) and at 30′ (P = 0.01) plasma glucose-dependent insulinotropic polypeptide (GIP) levels; lower area under the curve (AUC) (0-30) for GIP (P = 0.008); higher glucagon-like peptide-1 (GLP-1) plasma levels at 90′ (P = 0.02) and 120′ (P = 0.02); lower glucagon plasma levels at baseline (P = 0.04) and at 30′ (P = 0.03); and appropriate glucagon suppression after the oral glucose load.
MHO subjects show, as well as normal-weight individuals, a lower diabetogenic profile by virtue of higher disposition index and unaffected entero-insular axis. At-risk obese individuals present increased GIP levels that might play a role in determining increased glucagon secretion and inappropriate glucagon responses after glucose load, thus contributing to impaired glucose homeostasis.
Obesity is a serious public health problem worldwide. Among US adults, 65% are overweight (body mass index, BMI > 25 kg/m2) and 30% are obese (BMI > 30 kg/m2) (1). Several data indicate obesity as a high-risk condition for metabolic and cardiovascular diseases including type 2 diabetes, hypertension, and dyslipidemia. But, among the whole obese population, there are subjects, defined as metabolically healthy but obese (MHO), who have a prevalence of cardiovascular disease (CVD) events over 3-11 years similar to normal-weight (NW) people (2).
This subset seems to be protected or more resistant to the obesity-related complications, and is characterized by high insulin sensitivity and favorable hormonal, inflammation, and immune profiles (3-5).
MHO prevalence is predicted to be 30-40% of the obese population (6). Although no standardized method to identify MHO individuals exists, the risk of developing cardiovascular disease may be comparable between the different methods (7). At difference with insulin resistance, insulin secretion and its role in maintaining glucose homeostasis has been less widely studied in MHO subjects, as well as the incretin effect, that accounts for 50-70% of insulin secretion after oral glucose load (8). In contrast, other gastrointestinal peptides have been investigated in insulin resistance conditions, like ghrelin, a peptide predominantly produced by the stomach, that is involved in the control of food intake and energy metabolism. It has been shown an acylated ghrelin excess in subjects with metabolic syndrome (9), in insulin-resistant postmenopausal women (10), and in insulin-resistant obese children (11) that could contribute to obesity-associated insulin resistance in metabolic syndrome.
We aimed to investigate beta cell function and the interaction with the entero-insular axis and glucagons secretion in this subset of obese subjects to better understand the mechanisms underlying their lower risk for diabetes.
Methods and Procedures
The study group consisted of 129 obese individuals (BMI ≥ 30 kg/m2) and 24 nonobese individuals (BMI < 25 kg/m2). They were consecutively enrolled among patients attending our university hospital for cardiovascular risk evaluation. Subjects were included in the study if they met the following criteria: fasting plasma glucose ≤ 100 mg/dl; absence of diabetes mellitus (previous history of diabetes and 2-h postchallenge glucose ≥ 200); no use of medications known to affect glucose metabolism; absence of previous cardiovascular events, clinical evidence of advanced liver or renal disease, and/or a recent history of acute illness. The local Ethics Committee approved the study, and after the nature of the study had been explained to each participant, informed written consent was obtained.
All participants underwent standard blood testing after a 12-h overnight fast (plasma glucose, insulin, and lipids), anthropometric measurements, and blood pressure. On the basis of the previous study (12), subjects were defined as metabolically healthy (MHO) or at-risk obese, according to the HOMA-IR index (MHO: lower tertile of HOMA-IR; at-risk: upper tertile of HOMA-IR index). Thus, 43 MHO+, 41 MHO−, and 24 NW subjects were included in this study.
These subjects underwent a 75-g oral glucose tolerance test (OGTT): samples for glucose, insulin, C-peptide, glucagon, glucagon-like peptide-1 (GLP-1), and glucose-dependent insulinotropic polypeptide (GIP) measurements were drawn at baseline and every 30 min after glucose ingestion up to and including 120 min, and distributed into chilled tubes containing EDTA with addition of aprotinin (500 kIU/ml blood), for analysis of plasma glucagon, GLP-1, and plasma GIP. For the analysis of glucose, insulin, and C-peptide, blood was distributed into chilled tubes containing heparin. All tubes were immediately cooled on ice and centrifuged at 4°C for 20 min. Plasma was stored at −20°C until analysis. The diabetic patients, according to American Diabetes Association criteria (13) based on their 2-h post-OGTT glucose levels, were excluded by this study.
Body weight and height were measured, and BMI was calculated as weight (kg)/(height [m])2. Waist circumference was measured in a standing position at the level of the umbilicus. Blood pressure was measured with a calibrated mercury sphygmomanometer when the subjects had rested in the supine position for 10 min. Plasma glucose was measured with the glucose oxidase method. Serum insulin was measured using a microparticle enzyme immunoassay, Axsym System, Abbott Laboratories (Wiesbadem, Germany). C-peptide was measured using an ELISA Kit (Millipore Corporation, Billerica, MA; coeffiecient of variation % (CV%): interassay: 5.0-8.7, intra-assay: 1.6-4). High-sensitivity C-reactive protein was measured using immunoturbidometry (Multigent C-reactive protein (CRP) Vario, Abbott Laboratories, Chicago, IL; CV%: interassay: 0.4 ± 0.02, intra-assay: 0.4 ± 0.01). Plasma glucagon was measured by a radioimmunoassay, RIA (Millipore Corporation; CV%: inter-assay: 11.7 ± 3.0, intra-assay: 4.9 ± 1.3). GLP-1 (active) was measured by an ELISA Kit (Millipore Corporation; CV%: interassay: 8 ± 4.8, intra-assay: 7.4 ± 1.1). Total GIP was measured by an ELISA Kit (Millipore Corporation; CV%: interassay: 1.8-6.1, intra-assay: 3.0–8.8). Serum total cholesterol, triglycerides, and high-density lipoprotein (HDL) cholesterol were measured by available enzymatic methods. LDL cholesterol was calculated using the Friedewald formula.
The area under the insulin, C-peptide, glucose, GLP-1, and GIP was calculated by the trapezoidal method.
Early-phase insulin release was assessed by the insulinogenic index (IG30) calculated as the change of insulin (pmol/l) divided by the change of glucose (mmol/l) during the first 30 min of the OGTT ΔI(0-30)/ΔG(0-30) (14).
Insulin resistance was estimated using HOMA-IR calculated as previously described (15).
To evaluate the relationship between beta cell function and insulin resistance, disposition index was calculated as IG30/HOMA-IR (16).
Statistical comparison of clinical and biomedical parameters was performed using Stat View version 6.0 for Windows. Data are given as means ± SE. When necessary, numerical variables were logarithmically transformed to reduce skewness, and values were expressed as arithmetic means. Statistical analysis included unpaired t-test, χ2 test, and ANOVA for continued variables. A P-value less than 0.05 was considered statistically significant.
Comparison between MHO and at-risk obese subjects
Anthropometric and biochemical characteristics are shown in Table 1. No differences in age, BMI, sex distribution, triglycerides, systolic and diastolic blood pressure, total and LDL cholesterol, and PCR were observed between MHO and at-risk obese subjects. MHO subjects had lower waist circumference (P = 0.006), higher HDL cholesterol levels (P = 0.01), and lower fasting plasma glucose (P < 0.0001). By definition, HOMA-IR was significantly lower in MHO (P < 0.0001) than in at-risk obese subjects.
|MHO (1)||At-risk obese (2)||P 2 vs. 1||NW (3)||P 3 vs. 1|
|Number (female/male)||43 (29/14)||41 (22/19)||NS||24 (20/4)||NS|
|Age (years)||42.8 ± 2||42 ± 2||NS||38.3 ± 2||NS|
|BMI (kg/m2)||34.6 ± 0.6||35.9 ± 0.5||NS||22.09 ± 0.4||<0.0001|
|Fasting glucose (mmol/l)||4.8 ± 0.07||5.36 ± 0.1||<0.0001||4.7 ± 0.1||NS|
|Waist circumference (cm)||108 ± 1.6||113.35 ± 1.4||0.006||81.7 ± 1.5||<0.0001|
|HDL cholesterol (mmol/l)||1.22 ± 0.05||1.06 ± 0.05||0.01||1.39 ± 0.06||0.03|
|Triglycerides (mmol/l)||1.33 ± 0.1||1.6 ± 0.1||NS||0.94 ± 0.08||0.03|
|Systolic blood pressure (mmHg)||126 ± 2.1||130.8 ± 2.3||NS||114.16 ± 1.5||0.0005|
|Diastolic blood pressure (mmHg)||78.2 ± 1.7||81.34 ± 1.5||NS||71.6 ± 1.1||0.009|
|Total cholesterol (mmol/l)||5.11 ± 0.16||5.11 ± 0.16||NS||4.83 ± 0.24||NS|
|LDL cholesterol (mmol/l)||3.28 ± 0.13||3.32 ± 0.14||NS||3.01 ± 0.2||NS|
|CRP (mg/dl)||0.40 ± 0.59||0.44 ± 0.05||NS||0.11 ± 0.01||NS|
|HOMA-IR||1.5 ± 0.07||5.2 ± 0.3||<0.0001||0.9 ± 0.06||NS|
Time course for plasma glucose during the OGTT is depicted in Figure 1A. MHO subjects showed significantly lower glucose plasma levels at all individual time points, although at-risk obese individuals presented plasma glucose levels within normal limits. Accordingly, AUC (0-120) for glucose was significantly lower in MHO (P < 0.0001), as shown in Table 2.
|MHO (1)||At-risk obese (2)||P 2 vs. 1||NW (3)||P 3 vs. 1|
|IG30 (pmol/mmol)||207.6 ± 41.9||201.6 ± 23.2||NS||103.9 ± 70.4||0.006|
|Disposition index (IG30/HOMA-IR)||153.8 ± 28.9||44.5 ± 5.8||0.0001||126.4 ± 23.4||NS|
|AUC (0-120) glucose mmol/l*min||800.5 ± 30.7||994.9 ± 30.5||<0.0001||789.2 ± 32||NS|
|AUC (0-120) insulin 103*pmol/l*min||48.7 ± 5.1||100.4 ± 8.4||<0.0001||28.7 ± 2.3||0.002|
|AUC (0-120) C-peptide nmol/l*min||228 ± 20.6||330.2 ± 37.3||0.01||173.8 ± 17.6||0.04|
|AUC (0-30) GIP pmol/l*min||804.6 ± 125.4||1,467.6 ± 222.8||0.008||783.8 ± 223.6||NS|
|AUC (0-120) GIP 103*pmol/l*min||4.9 ± 0.6||7.1 ± 0.8||0.055||4.8 ± 1.1||NS|
|AUC (0-30) GLP-1 pmol/l*min||104.1 ± 12.4||74.46 ± 9.1||NS||115.2 ± 3||NS|
|AUC (0-120) GLP-1 pmol/l*min||454.9 ± 5||318.7 ± 3||NS||422.6 ± 8||NS|
Insulin and C-peptide
MHO subjects had lower insulin plasma levels at baseline (P < 0.0001) and during the entire course of OGTT (Figure 1B). Likewise, C-peptide (Figure 2C) was lower in MHO subjects at baseline (P = 0.0002), at 30′ (P = 0.01), at 60′ (P = 0.04), at 90′ (P = 0.048), and at 120′ (P = 0.005). Accordingly, AUC (0-120) for insulin and C-peptide was significantly lower in MHO (P < 0.0001 and P = 0.01, respectively), as shown in Table 2.
Early-phase insulin secretion, estimated by IG30, was similar among the groups (P = NS). However, the disposition index was significantly higher in MHO group when compared with at-risk subjects (P = 0.0001) (Table 2).
Plasma GIP (Figure 2A) and GLP-1 (Figure 2B) levels increased significantly in response to oral glucose load. Plasma GIP levels were significantly lower in MHO subjects at baseline and at 30′ (P = 004 and P = 0.01, respectively). Also, AUC (0-30) for GIP was lower in these subjects (P = 0.008, Table 2). Moreover, AUC (0-120) for GIP was lower, although this difference did not reach the statistical significance (P = 0.055).
GLP-1 plasma levels were significantly higher at 90′ (P = 0.02) and 120′ (P = 0.02) in MHO group. AUC for GLP-1 did not differ among the groups (Table 2).
MHO subjects showed lower glucagon levels at baseline (P = 0.04). Moreover, plasma glucagon decline after the glucose load was different in the two groups (Figure 2C). MHO subjects reached 50% of maximum glucagon's suppression at 30′ after OGTT, whereas at-risk subjects showed a delayed glucagon suppression reaching only 12% and the same time (P < 0.05). Accordingly, glucagon levels after 30′ of the OGTT were lower in MHO subjects (P = 0.03).
Comparison between MHO and NW subjects
Anthropometric and biochemical characteristics are shown in Table 1. No differences in age and sex distribution, fasting plasma glucose, total and LDL cholesterol, and PCR were observed among the groups. NW subjects had lower BMI (P < 0.0001), waist circumference (P < 0.0001), triglycerides (P = 0.03), and higher HDL cholesterol (P = 0.03) than MHO subjects.
HOMA-IR index was slightly higher in MHO, but not statistically different (P = 0.08). Glucose, GIP, GLP-1, and glucagon fasting plasma levels and responses did not differ among MHO and NW subjects (Figures 1 and 2). Accordingly, the AUCs for the above parameters were similar among the groups (Table 2).
Insulin and C-peptide plasma levels during 120′ OGTT were higher in MHO than in NW individuals (Figure 1). Correspondingly, AUCs (0-120) and IG30 were increased in MHO subjects (Table 2), although these differences disappeared when HOMA-IR was used as covariate. Therefore, the disposition index did not differ among the groups (P = NS).
This study demonstrates that MHO subjects show several differences in comparison to at-risk obese when insulin, glucagon, and incretin hormones were investigated after OGTT.
An important and original finding of our study is that MHO subjects and NW individuals exhibit a similar disposition index, suggesting a preserved beta cell function. In contrast, beta cell function of at-risk obese is unable to compensate their insulin resistance.
A second finding of our study is that in MHO subjects, both glucagon basal levels and glucagon suppression during OGTT were similar to control subjects, and different to at-risk obese. We also find a significant difference in GIP levels between the two groups of obese. As GIP has been shown to have glucagonotropic effects in human subjects, the increased GIP levels in at-risk obese might contribute to the increased glucagon basal levels and to the inappropriate glucagon's suppression after glucose load.
Although it has been accepted that excess weight does not affect every subject the same way, there is no universal definition of metabolic health in obese people. We used HOMA-IR index to identify MHO subjects, to focus on insulin resistance, that is considered as a key feature of the pathophysiology of obesity, metabolic syndrome, and diabetes and also as the link to the development of atherosclerosis and cardiovascular events. MHO subjects had lower glucose and insulin profiles during OGTT. Early-phase insulin secretion, evaluated by IG30, was similar in the two groups.
In addition, MHO subjects have higher disposition index than at-risk obese individuals. Disposition index is the product of measures of insulin sensitivity and first-phase insulin secretion and it has been shown as determinant for the conversion to diabetes (17). This finding suggests a relative insulin deficiency and an inappropriate beta cell response for the level of insulin sensitivity in at-risk obese subjects. In contrast, MHO and NW subjects showed similar glucose levels, insulin secretion, after adjustment for HOMA-IR, and disposition indices.
To our knowledge, no previous difference on glucagon and incretin levels in MHO or at-risk obese subjects has been reported. We found both fasting hyperglucagonemia and lack of suppression of glucagon after oral glucose load in at-risk obese subjects. These are key features of the pathophysiology of T2DM and altered glucose tolerance (18) and contribute to hyperglycemia observed in these patients. In contrast, MHO had both basal plasma glucagon levels and suppression's kinetics similar to NW subjects.
In normal condition, GLP-1 and GIP stimulate adenylyl cyclase (AC) through a Gs-coupled process resulting in localized increases in cyclic AMP that result in the activation of both protein kinase A (PKA) and cAMP-specific guanine nucleotide exchange factor/exchange protein directly activated by cAMP, Epac (19). PKA and Epac mediate the cAMP-dependent potentiation of large dense core secretory vesicle exocytosis and small synaptic vesicle of insulin, respectively. In addition to the pancreas, incretins have potential effects in adipose tissue. Although GLP-1 shows a dual action of lipolysis and lipogenesis in human adipocytes (20), GIP plays a crucial role in adipocyte biology and lipid metabolism. Some studies suggest that increased GIP secretion contributes to the type 2 diabetic phenotype by increasing glucagon secretion and promoting storage of fat (21). It has been described an increased K-cell density and circulating GIP levels in obese animals and high-fat-fed animals (22) and in obese humans (23). Recently, it has been well elucidated that in GIPR-overexpressing adipocytes, GIP promptly activates inflammatory responses, by stimulation of the production of inflammatory cytokines, chemokines, and proinflammatory nuclear factors, mostly through cAMP-PKA pathway (24). Hence, GIP could be a major link between at-risk obesity and insulin resistance, because, in adipose tissue of at-risk obese subjects but not of MHO, it could contribute to a chronic low-degree inflammation that contributes to the development of insulin resistance and T2DM.
GIP secretion has been reported as increased also in healthy offspring of patients with T2DM (25) and in IGT (26, 27). MHO and NW subjects present similar plasma GIP levels supporting the hypothesis that increased GIP secretion may be linked to insulin resistance. Despite the increased GIP levels, an impaired insulinotropic effect of GIP has been reported in diabetic patients (28). Similarly, in our study, the lower disposition index of at-risk obese subjects could be liked to an impaired insulinotropic effect of GIP, in spite of the higher plasma levels.
This alteration might be due to a specific defect in the responsiveness of pancreatic beta cell to GIP. Alternatively, a reduced expression of GIP receptors on islets in response to chronic hyperglycemia has been described (29).
MHO subjects present higher GLP-1 plasma levels at 90′ and 120′ than at-risk obese individuals and no differences in comparison to NW subjects. It has previously been suspected that type 2 diabetic patients have an impaired secretion of GLP-1 particularly in the late phase of the meal response (30, 31), and it has been hypothesized that this reduction could be a consequence of impaired glycemic control (30). The reduction of GLP-1 secretion in at-risk obese subjects could be linked to a chronic state of insulin resistance and/or reduced glucose tolerance.
In conclusion, MHO subjects, as well as nonobese individuals, show a lower diabetogenic profile by virtue of higher disposition index and unaffected entero-insular axis. At-risk obese individuals present increased GIP levels that might play a role in determining increased glucagon secretion and inappropriate glucagon responses after glucose load, and in inducing insulin resistance, thus contributing to impaired glucose homeostasis.
- 25Twenty-four-hour insulin secretion rates, circulating concentrations of fuel substrates and gut incretin hormones in healthy offspring of Type II (non-insulin-dependent) diabetic parents: evidence of several aberrations. Diabetologia 1999; 42: 1314-1323., , , et al.