INTRODUCTION
Incretin is a hormone that is secreted from the intestine within minutes after food intake and stimulates pancreatic β-cells to produce insulin, which leads to postprandial glucose stabilization. Glucagon-like peptide 1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) are two principal incretin hormones that are secreted from L-cells in the ileum and K-cells in the duodenum and jejunum, respectively [1]. Both incretins are responsible for 50% to 70% of the postprandial insulin secretion in healthy people. This incretin effect falls to <20% in patients with type 2 diabetes mellitus (T2DM) [2,3]. Either reduced insulinotropic efficacy or reduced secretion of incretin can be attributable to the reduced incretin effect in T2DM patients [4,5]. In T2DM, the impaired insulinotropic action of incretin is more pronounced in GIP than in GLP-1 [2]. Those impaired insulinotropic efficacies may be secondary to a general impairment in β-cell function rather than a specific defect in incretin action [6]. In T2DM, meal related GLP-1 secretion have been found to be deficient compared with normal state in some studies [7,8], whereas GIP levels are found to be relatively normal in the majority of studies [7,8,9]. In fasting state, small study reported decreased fasting level of GLP-1 in T2DM patients compared with normal state [10]. It has also been reported that some metabolic markers are related to incretin concentrations in T2DM patients. Two recent meta-analyses suggested that the deterioration of glycemic control (glycosylated hemoglobin [HbA1c]) is associated with reduced GLP-1 and GIP secretions [11,12], and that high body mass index (BMI) is associated with increased GIP secretion, but increasing age is associated with reduced GIP responses [11]. Alssema et al. [13] reported that GIP response was positively related to triglyceride levels.
Intact incretin, a biologically active form of incretin, is degraded by the enzyme dipeptidyl peptidase-4 (DPP-4) to inactive metabolite shortly after its secretion [14]. This inactive form accounts for a substantial part of total incretin. The total form of incretin reflects its secretion from enteroendocrine cells, whereas intact incretin reflects its action on glucose-induced insulin secretion [15]. Incretin-based drugs that increase circulating intact incretin level or mimic the action of intact GLP-1 (iGLP-1) have now emerged as important tools in the treatment of T2DM [16]. Thus, intact incretin is a target in the current strategy for the treatment of T2DM. Considering the insulinotropic action of intact incretin, it can be assumed that there is some relationship between glucose, insulin, and intact incretin levels. However, only a few studies have investigated the determinants or factors associated with intact incretin levels in T2DM patients.
Here, we measured iGLP-1 and intact GIP (iGIP) levels before and after ingestion of a standard mixed meal in patients with T2DM. Then we cross-sectionally analyzed the relationships between intact incretin levels and various clinicobiochemical variables including glycemic parameters.
METHODS
Subjects
This cross-sectional study included 336 patients with T2DM who admitted to the Diabetic Clinic at the Endocrinology Department of Jeju National University Hospital, South Korea. Patients with stage 4 chronic kidney disease (estimated glomerular filtration rate [eGFR] <30 mL/min), those with liver cirrhosis and ascites, and those with infectious or inflammatory disease were excluded. Patients receiving any DPP-4 inhibitor or GLP-1 receptor agonist, which may influence plasma incretin levels, were also excluded. The present study was conducted in compliance with the Declaration of Helsinki and was approved by the Institutional Review Board of Jeju National University Hospital (IRB No. JEJUNUH 2017-02-008). In addition, written consent explaining the purpose and methods of the present study was provided to all the subjects.
Study procedure
Blood samples were collected after night fasting (over 8 hours) in heparinized tubes to measure glucose, HbA1c, C-peptide, total cholesterol, triglyceride, high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C), alanine aminotransferase (ALT), and creatinine levels. Fasting iGLP-1 and iGIP levels were measured from blood samples that were stored in chilled ethylenediaminetetraacetic acid (EDTA) tubes, to which aprotinin (250 KIU/mL blood; Sigma-Aldrich, St Louis, MO, USA) and DPP-4 inhibitor (10 µL/mL blood; Merck Millipore, Darmstadt, Germany) were added. Subsequently, the subjects were provided a standard mixed meal (480 kcal; carbohydrate: protein: fat=2.8:1:1). The standard mixed meal consisted of boiled rice, soup, three side dishes, and kimchi, and the ratio of the nutrients was always kept constant. The subjects were asked to complete their meal within 20 minutes. Additional blood samples were collected 30 minutes after the start of meal in heparinized tubes to measure glucose and C-peptide levels, and post-meal iGLP-1 and iGIP levels were measured from blood samples in chilled EDTA tubes, to which aprotinin and DPP-4 inhibitor were added. The EDTA tubes containing the blood were immediately stored in ice pouches and then centrifuged. The separated plasma were stored at −20℃, until iGLP-1 and iGIP levels were analyzed. The differences between glucose, C-peptide, iGLP-1, and iGIP levels at 30 minutes after the meal and the corresponding fasting values were indicated as incremental glucose (Δglucose), C-peptide (ΔC-peptide), iGLP-1 (ΔiGLP-1), and iGIP (ΔiGIP) levels, respectively.
Data on age, duration of diabetes mellitus, comorbid diseases, and antidiabetic agents taken by patient were collected from patient histories and relevant electronic medical records from the Jeju National University Hospital. The height and weight of each subject were measured using an electronic meter. Systolic and diastolic blood pressure levels were recorded as the mean of two measurements after each subject rested for more than 10 minutes. The waist circumference was measured (in cm to one decimal place) at the midpoint between the costal margin and iliac crest when each subject had both legs spread at shoulder width.
Measurement of biochemical markers
Blood glucose level was measured using the TBA-200FR chemical analyzer (Toshiba, Tokyo, Japan), with a glucose oxidase method. HbA1c level was measured using the HLC-723G8 (Tosoh, South San Francisco, CA, USA) with an ion-exchange high-performance liquid chromatography. HbA1c standardization was performed in accordance with the guidelines of the Korean Association of Quality Assurance for Clinical Laboratory. C-peptide level was measured by the Modular Analytics E170 electrochemiluminescence immunoassays (Hitachi, Tokyo, Japan). Total cholesterol, triglyceride, HDL-C, LDL-C, creatinine, and ALT levels were measured using the TBA-200FR chemical analyzer. The eGFR values were calculated using the Modification of Diet in Renal Disease equation [17]. iGLP-1 (GLP-1 [7-36] amide, GLP-1 [7-37]) and iGIP (GIP [1-42]) levels were measured by a commercially available kit (Code No. 27784 and 27201, respectively; Immuno-Biological Laboratories Co. Ltd., Gunma, Japan) with a sandwich enzyme immunoassay.
Statistical analysis
All values are presented as mean±standard deviation or percentage or median (interquartile range). The difference between glucose, C-peptide, iGLP-1, and iGIP levels at 30 minutes after meal and fasting were analyzed by paired t-tests or Wilcoxon signed rank test. Spearman's rho were calculated to analyze the relationship between incretin levels and continuous clinicobiochemical variables. Wilcoxon rank sum test was used to compare incretin levels according to the presence or absence of categorical variables. Multiple linear regression analyses were used to determine independent relationships between intact incretin levels and continuous clinicobiochemical variables. Regression analyses were performed using fasting iGLP-1, fasting GIP, ΔiGLP-1, or ΔiGIP levels as dependent variables and age, sex, and factors with significant differences in the univariate analysis (Spearman's correlation and Wilcoxon rank sum test) as independent variables. All analyses were performed using the SPSS version 14.0 software program (SPSS Inc., Chicago, IL, USA), and P values of <0.05 were defined as significant.
RESULTS
Clinicobiochemical characteristics and fasting and postmeal incretin levels in the subjects
The levels of clinicobiochemical variables of the subjects are summarized in Table 1. A total of 336 patients with T2DM were enrolled in this study. Their average age was 58 years old, and there were more men than women. The subjects had relatively long durations of diabetes and poor glycemic control. About half of all subjects were taking sulfonylureas, about two-thirds were on metformin, and only a few subjects were taking thiazolidinediones or α-glucosidase inhibitors. About one-third of all subjects were injecting insulins (Table 1).
Fasting and 30-minute postmeal glucose, C-peptide, and intact incretin levels in the study subjects are presented in Table 2. Glucose and C-peptide levels were significantly increased after ingestion of a standard mixed meal. In comparison with the fasting iGLP-1 and iGIP levels, those 30-minute levels increased by 190% and 570%, respectively. The increment of iGIP level was thus larger than that of iGLP.
Univariate analyses of intact incretin levels with clinical and laboratory variables
We performed Spearman's correlation anaylsis and Wilcoxon rank sum test for univariate analysis of the relationships between intact incretin levels and clinical and laboratory variables (Tables 3 and 4). In the Spearman's correlation analysis, fasting iGLP-1 levels were significantly positively correlated with glucose, creatinine, and triglyceride levels, and negatively correlated with eGFR. Other variables were not significantly correlated with fasting iGLP-1 levels. ΔiGLP-1 levels were only positively correlated with ΔC-peptide. Fasting iGIP levels showed significant positive correlations with HbA1c, fasting glucose and negative correlations with ΔC-peptide. Incremental iGIP (ΔiGIP) levels were significantly negatively correlated with diabetes duration and HbA1c, and positively correlated with Δglucose, ΔC-peptide levels and eGFR (Table 3). Fasting level and incremental level of iGIP showed significant negative correlation.
The relationships between intact incretin levels and categorical variables were analyzed by Wilcoxon rank sum test. There were no sex differences regarding fasting iGLP-1, ΔiGLP-1, and fasting iGIP-1 levels. However, ΔiGIP levels were significantly higher in male subjects than females. Intact incretin levels did not differ with the use of sulfonylurea, metformin, thiazolidinedione, and α-glucosidase inhibitor. However, insulin users had lower ΔiGIP levels compared with noninsulin users. No differences were found in fasting iGLP-1, ΔiGLP-1, and fasting iGIP levels with the use of insulin (Table 4).
Multivariate analyses predicting intact incretin levels
Multiple linear regression analyses were performed to examine the variables independently predicting intact incretin levels. The variables included were age, sex, and significant variables in the univariate analyses. In the regression model with fasting iGLP-1 as the dependent variable, only fasting glucose remained significantly related factors. Age, sex, eGFR, and triglycerides failed to maintain those significant correlations. Creatinine was excluded due to collinearity with age and eGFR. The model with ΔiGLP-1 showed that ΔC-peptide was a significantly related factor after adjustment for age and sex. In the model with fasting iGIP, fasting C-peptide and ΔiGIP remained independent correlated factor after adjustment for age, sex, HbA1c, fasting glucose levels. ΔC-peptide was excluded from the variables because it occurred later than fasting iGIP. The model with ΔiGIP showed that only Δglucose, ΔC-peptide, and fasting iGIP were significant correlated factor after adjustment for age, sex, disease duration, and the use of insulin (Table 5).
DISCUSSION
In this study, we cross-sectionally analyzed the relationships between intact incretin levels and various clinicochemical variables in 336 T2DM patients. Univariate and multivariate analyses showed that fasting iGLP-1 levels were related to fasting glucose levels. ΔiGLP-1 levels were significantly correlated with ΔC-peptide levels, fasting iGIP levels were related to fasting C-peptide levels, and ΔiGIP levels were significantly correlated with ΔC-peptide and Δglucose levels. Incretins are released in response to food ingestion and regulate postprandial glucose levels rather than fasting glucose. Those postprandial increments are physiologically more important than the fasting levels. Therefore, our study results suggest that intact incretin levels are primarily related to C-peptide and glucose levels.
Few studies have investigated the predictors of incretin levels in T2DM patients. Toft-Nielsen et al. [7] reported that BMI and male sex negatively influenced the variability of both GLP-1 and GIP responses, and insulin area under the curve negatively influenced that of only GLP-1 response. Vollmer et al. [9] reported that free fatty acid levels and sex were significantly related to GLP-1 response, but glucose control or insulin secretion was not related to incretin levels. Another study showed that GLP-1 and GIP responses to oral glucose were related to fasting triglycerides [13]. There is no clear reason why our results differ from other findings. Differences in methods of measuring incretin may be one of the reasons for the discrepancy. In our study, we measured intact incretin, in contrast to the above-mentioned studies that measured total incretin. Intact incretin is a biologically active form, whereas total forms of incretin are composed of intact incretin plus its metabolites produced by enzyme DPP-4. It is intact and not total incretin that stimulates insulin secretion and influences blood glucose levels. Therefore, we argue that our study showed the significant relationship with C-peptide and glucose levels—not with other parameters—because we measured intact incretin levels.
The present study showed that ΔC-peptide levels were positively correlated with both ΔiGLP-1 and ΔiGIP levels, which suggests that both incretins augmented insulin secretion. Specifically, the ΔC-peptide levels had a stronger correlation with ΔiGLP-1 than ΔiGIP levels. This finding accords with previous research reporting that in T2DM patients the insulinotropic effect of GLP-1 is largely preserved more than that of GIP [18,19]. The present study also demonstrated that Δglucose was significantly associated with ΔiGIP but not with ΔiGLP-1. This result could be attributed to the downregulation of the β-cell GIP receptor in response to hyperglycemia [20,21]. The postmeal increment in iGIP level could compensate for decreased expression of the GIP receptor. Morever, in pancreatic α-cells, GIP enhances postprandial glucagon secretion and GLP-1 suppresses it [22]. The opposite effect of GIP and GLP-1 on glucagon may be the reason for the difference in the association between ΔiGIP and ΔiGLP-1 to Δglucose. The lack of positive association between Δglucose and ΔiGLP-1 might be due to the glucose-lowering effect of GLP-1. We measured the early-phase (0 to 30 minutes) glucose and incretin responses to ingestion of a meal. If we had measured their late responses, there would have been a negative relationship between Δglucose and ΔiGLP-1 levels.
Fasting iGIP and ΔiGIP, showed significant negative correlation in univariate and multivariate analysis. It could be explained by while the secretion of GIP increase to overcome the downregulated receptor in pancreatic β-cell, the K-cell has limits to secret GIP.
Fasting iGLP-1 and iGIP levels, unlike their postmeal increments, were significantly related to fasting glucose and C-peptide levels, respectively. We do not know the reason for this discrepancy. On the other hand, female patients had higher fasting iGIP levels than males. It is known that women generally have a slower gastric emptying rate than men [23], which could explain the higher fasting iGIP levels of female patients in our study. Indeed, because incretin hormones participate in the regulation of postprandial glucose rather than fasting glucose levels, fasting incretin levels were two to six times lower than their postmeal levels. Therefore, fasting iGLP-1 and iGIP levels are less important than their incremental levels after consumption of a meal in the physiological aspects of incretin hormones.
Our study has some limitations. First, we did not perform serial measurements of incretin. Postprandial incretin response is generally evaluated using area under the curve or peak concentration of serum incretin. Theoretically, the area under the curve reflecting the integral of concentration—time is more desirable than peak concentration. However, to minimize inconvenience to the study subjects, we chose a peak concentration for assessing the response. Our unpublished data show that intact incretin levels rapidly increased, peaked at 30 minutes after ingestion of a meal, and gradually decreased with time (data not shown). Thus, we measured fasting and 30-minute postmeal values for evaluating the response. Second, we did not measure glucagon levels. iGLP-1 may exert its glucoregulatory action mainly via the suppression of glucagon secretion [24]. Therefore, the measurement of glucagon would clarify the relationship between iGLP-1 and glucose levels. Third, total incretin levels and DPP-4 activity were not assessed in our study. If we had assessed total incretin and DPP-4 activity, we would have understood how intact and total incretin levels related differently to clinical and laboratory parameters. Lastly, most subjects were taking various antidiabetic drugs during the study. We excluded subjects taking DPP-4 inhibitors or GLP-1 agonists that directly influence the serum level of incretin. However, it is known that α-glucosidase inhibitor [25], metformin [26], and insulin [27] can also increase blood incretin levels. Here, we showed that iGLP-1 and iGIP levels did not differ between users and nonusers of α-glucosidase inhibitor or metformin (Table 4). Although ΔiGIP levels were significantly different with the use of insulin, multivariate analysis adjusting for covariates revealed that the use of insulin did not predict ΔiGIP levels (Tables 4 and 5).
In conclusion, postmeal increments in intact incretin levels were primarily related to incremental glucose and C-peptide levels in patients with T2DM. These results suggest that glycemia and insulin secretion are the main factors associated with blood levels of intact incretin hormones. However, to understand the interrelation between intact incretin, glucose, and insulin secretion, serial measurements of glucagon, DPP-4 activity, and total and intact incretin levels are required.
PDF Links
PubReader
XML Download
Full text via DOI
Full text via PMC
CrossRef TDM
Print
Cited
