The oral disposition index calculated from a meal tolerance test is a crucial indicator for evaluating differential normalization of postprandial glucose and triglyceride excursions in morbidly obese patients after laparoscopic sleeve gastrectomy

Abstract

To determine the normalization of postprandial blood glucose (PG) and triglyceride (TG) excursions in 30 morbidly obese patients with or without diabetes mellitus (DM) 1-year after they underwent a laparoscopic sleeve gastrectomy (LSG) vs. their pre-surgery data, we administered the 75-g oral glucose tolerance test (OGTT) and a meal tolerance test (MTT) using a 75-g glucose-equivalent carbohydrate- and fat-containing meal. The results were as follows; (i) Postoperative body-weight reduction was associated with DM remission and reduced multiple cardiometabolic risks. (ii) OGTT data showing postprandial hyper-insulinemic hypoglycemia in many post-surgery patients were associated with overdiagnosis of improved glucose tolerance. However, postoperative MTT data without hypoglycemia showed no improvement in the glucose tolerance vs. pre-surgery data. (iii) The disposition index (DI) i.e., [Matsuda index] × (Glucose-induced insulin secretion) was progressively worsened from normal glucose tolerance to DM patients after LSG. These post-surgery DI values measured by the MTT were correlated with 2h-plasma glucose levels and were not normalized in DM patients. (iv) The baseline, 2h-TG, and an increase in 2h-TG values above baseline were correlated with the insulin resistance index, DI, or HbA1c; These TG values were normalized post-LSG. In conclusion, the glucose tolerance curve measured by the MTT was not normalized in T2DM patients, which was associated with impaired normalization of the DI values in those patients 1-year after the LSG. However, the baseline TG and a fat-induced 2h-TG values were normalized postoperatively. The MTT can be used to assess normalization in postprandial glucose and TG excursions after LSG.

BARIATRIC SURGERY is currently the most effective treatment to reduce body weight in individuals with morbid obesity. The reduction in body weight that is achieved following bariatric surgery is closely associated with a marked improvement of metabolic derangement, including the improvement of treatment-resistant hyperglycemia in markedly obese patients with type 2 diabetes mellitus (T2DM) after metabolic surgery [1-3]. As a result, some individuals who have undergone bariatric surgery and who have an HbA1c value <6.5% without the use of any glucose-lowering drugs are classified as having achieved the remission of T2DM [4].

The 75-g oral glucose tolerance test (OGTT) is used in clinical practice as the standard method to identify the normalization of glucose tolerance in T2DM patients. As shown by the 75-g OGTT, individuals with severe obesity who undergo metabolic surgery may experience reactive hypoglycemia due to dumping syndrome, which may modify the glucose tolerance curve [5-9]. This reactive hypoglycemia after the intake of glucose-containing food has been frequently observed in patients who underwent Roux-en-Y gastric bypass (RYGB) surgery [6, 7] and less common in a sleeve gastrectomy [8, 9]. This dumping syndrome modifies glucose metabolism and makes it difficult to accurately evaluate the glucose-tolerance curve based on the present diagnostic criteria.

Our research group have reported that compared to the 75-g OGTT, a new meal tolerance test (MTT) that uses a test meal containing wheat starch equivalent to 75 g glucose, 28.5 g of butter fat, and a low amount of protein is able to reduce postprandial hyperglycemia with a lower postprandial secretion of insulin in morbidly obese individuals with or without glucose intolerance [10]. In that investigation, the MTT was demonstrated to be useful for determining the precise staging of impaired glucose tolerance compared to the 75-g OGTT. We also observed that this MTT meal containing saturated fat can be used to evaluate postprandial hypertriglyceridemia in morbidly obese patients [10].

Another of our studies revealed that the preservation of insulin secretion at the presurgical evaluation is one of the key factors in the determination of the remission of T2DM [11]. In 2015, the ‘ABCD score’ was reported to be an excellent predictive index for screening appropriate candidates for bariatric surgery [12]. Several studies have described improvements of the insulin sensitivity index (ISI) combined with improved β-cell function-dependent glucose disposal activity may contribute to improvement in glycemic control after metabolic surgery [13-15]. Compared to the standard dietary management methods, the body-weight reduction and ISI improvements that are observed after a patient’s metabolic surgery are usually large and sustained for a long period [1, 16-18]. On the other hand, it is also reported that impaired glucose-induced insulin secretion (GIS) from β cells is improved following a strict glycemic control [19, 20]. An increased glucagon-like peptide 1 (GLP-1) secretion after metabolic surgery is reported to be closely associated with improvements of postprandial GIS [14, 21, 22]. A gastric bypass study has shown that preserved insulin secretion and weight loss are the predominant predictors of glycemic control [23]. However, it has also been reported that preintervention β-cell function and its post-surgery changes predict remission but not weight loss or insulin sensitivity [24]. Thus, we speculated that in order to evaluate the normalization of glucose-tolerance curves after metabolic surgery, the post-surgery changes in both the ISI and GIS should be determined and compared to the pre-surgery data.

An investigation using the standard OGTT and meal tolerance tests before and after a laparoscopic sleeve gastrectomy (LSG) is able to calculate the disposition index (DI), which is defined as the product of the ISI and GIS. Thus, the DI, a β-cell function-associated glucose disposal activity may be a crucial determinant that affects the glucose-tolerance curve [25]. In addition, it is generally accepted that postprandial hypertriglyceridemia may be associated with both insulin resistance and hyperglycemia in individuals with or without diabetes [26, 27]. Based on these prior findings, we conducted the present study to establish whether postprandial glucose and/or triglyceride (TG) excursions were normalized in morbidly obese patients with different glucose-tolerance subtypes by improvements of the ISI, GIS, and DI after LSG. Toward this goal, we compared the patients’ pre-and post-surgery data from two glucose-loading tests using the standard 75-g OGTT and the new MTT.

Patients and Methods

Patients

Sixty-one patients with morbid obesity with a body mass index (BMI) >35 kg/m² and age 22–62 years old who were followed up for 1 year after the LSG comprised the study population. Both OGTT and MTT were successfully completed before and after surgery in 37 out of 61 patients since we could not obtain the consent to perform 2 glucose tolerance tests before and after LSG in the 24 patients. Four of 37 patients had hemolysis in their blood samples and the remaining 3 insulin-treated diabetics at the pre-surgery evaluation were excluded from the present analysis. The present 6 diabetes patients were all overt diabetes, where the hyperglycemia was extensively controlled by 6-month pre-surgery nutritional and pharmacological treatments using glucose-lowering drugs (except of insulin) shown in Supplementary Table S1. Two patients were taking a statin, and seven patients were taking blood pressure-lowering medication in the present 30 patients. However, no patient was prescribed with TG-lowering drugs.

All 30 patients (26 females, 4 males) were hospitalized for 6 days for the evaluation of body composition and baseline blood testing under the conditions of planned dietary energy intake before (about 1,400 kcal/day) and after (about 1,200 kcal/day) LSG. In this study, all patients were instructed to consume a standard low-energy diet (25 kcal/kg IBW/day) containing macronutrient composition consisting about 40–60% carbohydrate, 20–35% fat, and 20–25% protein during pre-surgery medical control for 6-months. Each patient’s energy intake was monitored every month and preoperative dietary advices for control of problematic dietary habits were individually given during 30-min nutritional guidance by expert dieticians. Energy intake, and macronutrient energy ratios were assessed by a professional nutritionist using a nutrition and meal management software provided by Chi-technology Co., Ltd (Tokyo, Japan) before and after LSG at every month intervals. The mean values of each patient’s daily energy intake and energy percentage of the three macronutrients were determined at the initial assessment and after surgery, and the mean values of the total 30 patients were compared before and after the surgery.

Informed written consent for study participation and data publication was obtained from all patients. The study was conducted according to the guidelines of the Declaration of Helsinki and was approved by the Institutional Review Board of Kusatsu General Hospital (protocol no. 2017-0317-05, March 24, 2017). Note that Kusatsu General Hospital changed its hospital name to Omi Medical Center on October 1, 2021. The baseline clinical and metabolic characteristics of the patients before and at 1 year after LSG are summarized in Table 1.

Table 1

Clinical characteristics of the 30 morbidly obese patients before and 1 year after laparoscopic sleeve gastrectomy (LSG)

	Pre-surgery	Post-surgery	% change
n (males/females)	30 (4/26)	30 (4/26)
Body weight, kg	106.8 (94.7–119.4)	68.2 (59.2–79.7)***	–36.1
Body fat, %	50.6 ± 3.3	33.2 ± 8.4***	–34.4
LBM, %	47.1 ± 3.6	62.9 ± 7.8***	+33.5
SBP/DBP, mmHg	141/85 ± 16/17	115***/67*** ± 13/12	–18/–21
FPG, mg/dL	91.0 (84.0–96.8)	85.5 (80–91.5)*	–6.0
IRI, μU/mL	10.8 (8.1–12.8)	4.4 (3.7–5.2)***	–59.3
HbA1c, %	5.9 (5.7–6.4)	5.5 (5.2–5.7)***	–6.8
HOMA-R	2.3 (1.7–3.0)	1.0 (0.8–1.1)***	–56.5
HOMA-β	154.1 (103.6–211.2)	69.2 (56.3–94.4)***	–55.1
TG, mg/dL	107 (82–134.8)	58.5 (45.3–70.8)***	–45.3
HDL-C, mg/dL	54.0 (45.0–60.0)	66.5 (57.3–69.0)***	+23.1
LDL-C, mg/dL	112.0 (94.5–127.3)	113.5 (98.5–124.5)	ns
AST, IU/L	20.5 (17.0–29.8)	17.0 (16.0–21.0)**	–17.1
ALT, IU/L	26.0 (20–42.5)	15.0 (11.3–19.5)***	–42.3
γ-GTP, IU/L	28.0 (19.0–58.0)	12.0 (9.3–15.5)***	–57.1
Albumin, mg/dL	3.9 (3.8–4.2)	4.0 (3.7–4.2)	ns
Uric acid, mg/dL	5.1 (4.5–5.7)	4.2 (3.6–5.0)***	–17.6
eGFR, mL/min/1.73m2	86.3 (72.0–104.8)	84.5 (71.2–95.8)	ns
s-Creatinine, mg/dL	0.6 (0.5–0.7)	0.6 (0.6–0.7)	ns

Data are mean ± SD or median (Quartile 1 – Quartile 3), n = 30, *** p < 0.001, ** p < 0.01, * p < 0.05 vs. the pre-surgery value. %change indicates post-surgery value × 100/pre-surgery value – 100 (%). %LBM: % lean body mass, HOMA-R: homeostasis model assessment of insulin resistance, HOMA-β: homeostasis model assessment of β, ns: not significant, SBP/DBP: systolic blood pressure/diastolic blood pressure.

Laparoscopic sleeve gastrectomy procedure

Each LSG was performed using a standard laparoscopic technique. The greater curvature of the stomach was vertically resected to leave the tube-shaped lesser curvature (approx. 100 mL) using an automatic linear stapler 3 cm from pylorus to the bundle of His along a 45-F orally inserted gastric tube. This LSG procedure is currently the most frequently performed bariatric procedure in Japan due to its good long-term weight loss and improvement in obesity-related health disorders [28].

The 75-g OGTT and MTT

The 75-g OGTT and MTT protocols were as we described [10]. Briefly, on the 2nd day after their hospital admission, the patients underwent a 75-g OGTT. On the 4th day after admission, they underwent an MTT using the new test meal, which consisted of 75 g of glucose (85% flour starch and 15% maltose), 28.5 g of butter fat (saturated fat, 43.3% of total energy), and 8.0 g of protein with the total energy of 592 kcal in the 12 pieces of cookie meal (Saraya Co., Osaka, Japan). After the patient underwent a 12-h overnight fast, blood samples were taken at 0, 30, 60, 120 min after the oral intake of the test meal. The patient’s plasma glucose (PG), immunoreactive insulin (IRI), and C-peptide reactivity (CPR) levels were measured. The serum triglyceride (TG) concentrations during the MTT were also measured.

Biochemical and body composition analyses

Low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transpeptidase (γ-GTP), albumin, creatinine, and uric acid were also measured at baseline before and after surgery. The estimated glomerular filtration rate (eGFR) was calculated by the following equation from the Kidney Disease: Improving Global Outcomes (KDIGO) 2012 guideline for chronic kidney disease (CKD), using serum creatinine and age: for males = 194 × serum creatinine (Scr)^–1.094 × age^–0.287; for females = GFR (male) × 0.739. Blood samples were centrifuged for 15 min (3,000 g at 4°C) and frozen at –80°C until analysis.

The patients’ HbA1c levels were measured using the standard high-performance liquid chromatography (HPLC) method. The data are expressed as National Glycohemoglobin Standardization Program (NGSP) values. Plasma glucose (PG) was determined using the standard enzymatic method, and plasma insulin (INS) levels were measured using a chemiluminescence immunoassay (CLIA) kit (Abbott, Tokyo). CPR was also measured by a CLIA. Serum LDL-C, HDL-C, TG, ALT, AST, γ-GTP, albumin, creatinine, and uric acid concentrations were measured in the central laboratory by routine methods. Body composition was measured by the bioelectrical impedance method using an InBody 770 device (In Body Japan, Tokyo) [29].

Mathematical modeling and the calculation of indices determining glucose tolerance curves

The glucose tolerance, homeostatic model assessment of insulin resistance (HOMA-R), insulin secretion-beta (HOMA-β), insulinogenic index, and Matsuda index were calculated from the patients’ OGTT and MTT data [10, 30, 31]. In the present study, we measured glucose-induced insulin secretory (GIS) activities by using the following four parameters: the insulinogenic index (ΔINS_0–30/ΔPG_0–30), the acute GIS (INS-AUC_0–30/PG-AUC_0–30), the total GIS (INS-AUC_0–120/PG-AUC_0–120), and theΔCPR_0–30/ΔPG_0–30. The patients’ serum triglyceride (TG) concentrations during the MTT were also measured. The disposition index (DI) was calculated as a product of the [Matsuda index] and [the total GIS] [25, 32]. The relationship between the values of the Matsuda index and the total GIS was extrapolated as a hyperbolic relationship. We also used the same equation for the calculation of the DI from the patients’ MTT data in the present study. We compared the parameters of GIS, the Matsuda index, and the DI calculated from the OGTT and MTT data in the 30 patients classified as having normal glucose tolerance (NGT), impaired glucose tolerance (IGT), and diabetes mellitus (DM) before and at 1-year after the LSG to study the differences between OGTT and MTT data.

Statistical analyses

Continuous variables are presented as the mean (standard deviation [SD]) or median (interquartile range; Quartile 1 [Q1] and Quartile 3 [Q3]). Normally distributed data were analyzed by parametric tests. Non-normally distributed data were analyzed by the Wilcoxon rank-sum test. The total area under the curves (AUC) for insulin and glucose were calculated using the trapezoid rule. Differences between the data from the patients’ OGTTs and MTTs were analyzed using a paired Student t-test. The significance of differences in the non-parametric data between before and after surgery was examined with the Wilcoxson rank-sum test and a 2 × 3 contingency table for the χ²-test. Significant differences between OGTT and MTT in the percentage among the three glucose tolerance subtypes (NGT, IGT, and DM) between before and after LSG were analyzed using Fisher’s test. Significant differences between preoperative and postoperative PG-AUC values for NGT, IGT, and DM patients measured by OGTT and MTT were examined with the Kruskal-Wallis test respectively.

We conducted a non-parametric statistical analysis of the differences in the means of the continuous variables (i.e., the Matsuda index, glucose-stimulated insulin secretion, and the DI) among the NGT, IGT, and DM groups with the Kruskal-Wallis test. The Steel-Dwass method was used to test the statistical significance between two independent variables. We also determined the correlations between two independent variables measured by either the OGTT or the MTT or the pre- and post-surgery data by conducting a Pearson’s correlation analysis. Probability (p)-values <0.05 were considered significant.

Results

Improvements in clinical characteristics at 1-year after LSG

As shown in Table 1, at 1-year after the patients underwent the LSG, their body weights decreased by 36.1% from the baseline value with marked improvements of a 34.4% reduction of adiposity and a 33.5% increase in lean body mass. In addition, both mean systolic and diastolic blood pressure were reduced to the normotensive levels. The patients’ HbA1c values significantly improved from 5.9% to 5.5%, with decreased HOMA-β, HOMA-R, ALT, AST, TG, and uric acid levels, but an increased HDL-C compared to the respective pre-surgery values. However, the patients’ LDL-C, fasting PG, eGFR, serum creatinine, and serum albumin levels were not significantly changed after surgery.

As shown in Table 2, BMI was not different among NGT, IGT, and DM patients and the reduction of BMI was similar among 3 different glucose tolerance subtypes. Pre-surgery HbA1c and FPG levels in DM patients were significantly higher than those of NGT or IGT patients. Those pre-surgery values in each glucose tolerance subtype were significantly reduced after LSG. A pre-surgery HOMA-R was similar among 3 glucose tolerance subtypes and those values were significantly reduced after LSG, respectively. Pre-surgery HOMA-β in DM patients were significantly lower than those of either NGT or IGT patients and the post-surgery values of 3 glucose tolerance subtypes were significantly or tended to be lower than that of pre-surgery values, respectively.

Table 2

Changes in clinical parameters at 1 year after LSG compared with the respective pre-surgical values in morbidly obese patients

		NGT		IGT		DM
Patients, n		13		11		6
Female/male		12/1		10/1		4/2
		Pre-surgery	Post-surgery	Pre-surgery	Post-surgery	Pre-surgery	Post-surgery
BMI, kg/m2		40.3 (38.1–50.3)	26.5 (24.5–30.3)**	39.4 (38.3–50.2)	26.5 (24.7–29.7)**	38.6 (37.6–41.0)	28.2 (23.2–29.7)*
% body weight reduction		34.2 (30.9–43.8)		35.2 (30.6–38.1)		37.5 (28.8–41.3)
HbA1c, %		5.7 (5.4–5.9)††	5.4 (5.2–5.6)**	5.8 (5.6–6.0)††	5.4 (5.2–5.7)**	7.0 (6.1–7.2)	5.5 (5.0–5.9)*
OGTT	FPG, mg/dL	82.5 ± 9.9††	82.8 ± 6.1†	91.3 ± 7.2††	84.4 ± 6.7*†	106.0 ± 6.8	95.8 ± 8.6*
	2h-PG,mg/dL	108.9 ± 16.9††	63.5 ± 22.9***	155.8 ± 13.9	80.8 ± 40.5***	209.3 ± 38.8	87.7 ± 40.4**
	IRI, μU/mL	11.3 ± 5.9	4.0 ± 0.8***	12.0 ± 4.6	6.1 ± 3.2**	10.8 ± 3.7	4.7 ± 1.6*
	2h-IRI, μU/mL	55.3 ± 38.0	14.3 ± 14.2**	130.6 ± 68.9	23.4 ± 31.2**	97.8 ± 42.3	16.6 ± 22.9**
	HOMA-R	1.8 (1.5–2.8)	0.8 (0.7–0.9)**	2.3 (2.1–2.9)	1.0 (0.95–1.4)**	3.1 (2.1–3.1)	1.1 (0.9–1.3)*
	HOMA-β	211 (128–299)†	66 (56–90)**	151 (108–176)†	99 (69–121)†	78 (64–122)	54 (38–72)

Data are mean ± SD or median (Q1–Q3). The statistical analyses between pre- and post-surgery in the patients with NGT, IGT, and DM were performed by either a two-way ANOVA for statistical significance using Tukey’s test (for parametric data) or by the Kruskal-Wallis and Steel-Dwass test (for non-parametric data). *** p < 0.001, ** p < 0.01, * p < 0.05 vs. the comparable data before surgery. ^†† p < 0.01, ^† p < 0.05 vs. the comparable DM data.

Changes in glucose tolerance curves measured by the OGTT and MTT after LSG

As shown in Fig. 1, the post-surgery glucose tolerance curves during the OGTT showed a significant reduction of 2-h plasma glucose (PG) levels and the peak level of PG was shifted to 30–60 min after the glucose loading in each glucose tolerance stage (Fig. 1A–C). In addition, 14 patients (47%) showed a 2-h PG level <60 mg/dL including NGT (38.5%), IGT (54.5%), and DM (50%) patients. The area under the curve (AUC) of glycemic excursion (PG-AUC) calculated from OGTT data were not significantly different between before and after the LSGs in the NGT and DM groups except of the IGT data (Fig. 1D). However, the PG-AUC values calculated from OGTT data were statistically different among patients with 3 different glucose tolerance stages both before and after surgery, respectively.

Fig. 1

Comparison of the glucose-tolerance curves obtained during the oral glucose tolerance test (OGTT) between pre- (blue line) and post- (red line) laparoscopic sleeve gastrectomy (LSG) surgery in morbidly obese patients with normal glucose tolerance (NGT) (A), impaired glucose tolerance (IGT) (B), and diabetes mellitus (DM) (C). D: The area under the curve (AUC) of the glucose-tolerance curves in the NGT, IGT, and DM patients which were classified based on the clinical diagnosis before surgery. The percentages of patients in the hypoglycemic range, i.e., 2h-PG <60 mg/dL are shown in the NGT (38.5%), IGT (54.5%), and DM (50%) groups after the LSG. None of the patients were hypoglycemic before surgery. D: The AUCs of plasma glucose (PG-AUC) in the NGT, IGT, and DM patients. Both the pre-surgery (p < 0.0001) and post-surgery (p < 0.005) PG-AUCs during the OGTT differed significantly among the three patient groups (Kruskal-Wallis test). The DM patients were each being treated with two or more glucose-lowering drugs for glycemic control before their LSG surgery and without such drugs after their LSGs. Data are mean ± SD. *** p < 0.001, ** p < 0.01, * p < 0.05 vs. the pre-surgery values.

On the other hand, the glucose tolerance curve in each of the three glucose tolerance subtypes during the MTT was not significantly different between before and after the LSGs, with an exception of a significantly lower 2-h PG value in the NGT patients after the LSG (Fig. 2A–C). In the post-surgery MTT, no patient with 2h-PG <60 mg/dL was observed. There was also no significant difference in the PG-AUC calculated from MTT data between before and after the LSG (Fig. 2D). Similar to the OGTT data, those PG-AUC values among 3 different glucose tolerance stages were significantly different both before and after surgery, respectively.

Fig. 2

Comparison of glucose-tolerance curves during the meal tolerance test (MTT) between pre- (blue line) and post- (red line) LSG surgery in the patients with NGT (A), IGT (B), and DM (C). D: The AUCs of the glucose-tolerance curves in the NGT, IGT, and DM patients classified by presurgical diagnosis are shown. Both the pre-surgery (p < 0.0005) and post-surgery (p < 0.002) PG-AUCs during the MTT also differed significantly among the three patient groups (Kruskal-Wallis test). No patients showed 2h-PG <60 mg/dL after an MTT. Data are mean ± SD. ** p < 0.01 vs. the pre-surgery values.

As shown in Fig. 3, concerning plasma insulin (INS) excursions after the OGTT and MTT, a significant reduction of the IRI levels at the baseline and the 2-h values after LSG were observed in each glucose tolerance subtype. Supplementary Fig. S1 depicted the plasma CPR excursions after OGTT and MTT, revealing significantly lower post-surgery CPR levels at the baseline and at 2-h values in every glucose tolerance subtype compared to the pre-surgery values, which were similar to INS data. Both ΔIRI and ΔCPR levels at 30 min after the post-surgery OGTT were significantly higher than those after the MTT in response to a higher increment of ΔPG after OGTT than that of ΔPG after MTT (Supplementary Fig. S2).

Fig. 3

Comparison of plasma insulin levels during the OGTT in the patients with NGT (A), IGT (B), and DM (C) and during the MTT in the NGT (D), IGT (E), and DM (F) groups pre- (blue line) and post- (red line) LSG surgeries. Data are mean ± SD. *** p < 0.001, ** p < 0.01, * p < 0.05 vs. the pre-surgery values.

The improvement in glucose intolerance staging after LSG differed between the OGTT and MTT

In the diagnosis of the glucose intolerance stage based on the patients’ OGTT results, the pre-surgery classifications were: NGT, n = 13 (43%); IGT, n = 11 (37%) according to the current Japanese diagnostic criteria [33] and all six DM patients (20%) received pre-surgical medical treatments with at least two glucose-lowering drugs. Thus, glucose tolerance curve of two DM patients showed IGT pattern and 4 patients showed DM pattern. Thus, pre-surgical evaluation of glucose tolerance staging showed 43.3% NGT, 43.3% IGT, and 13.4% DM as shown in Table 3. At 1-year post-LSG, the classification using the OGTT data was significantly changed (p < 0.001) to: NGT, n = 28 (93.3%); IGT, n = 2 (6.7%); and DM, n = 0 (0%). However, following the diagnosis of glucose tolerance using MTT, the pre-surgery classifications were: NGT, n = 16 (53.3%); IGT, n = 12 (40%); and DM, n = 2 (6.6%), which was not significantly different from the pre-surgical classification by the OGTT data. At 1- year post-LSG, the number of patients classified using MTT changed as follows: NGT, n = 23 (76.7%); IGT, n = 5 (16.7%); and DM, n = 2 (6.6%), showing only a tendency for improvement compared with the pre-surgery values (Table 3).

Table 3

Changes in the number of different glucose intolerance subtypes postoperatively compared with the pre-surgery numbers of the patients with NGT, IGT, and DM measured by the OGTT and MTT

Clinical glucose intolerance		NGT	IGT	DM	Total
	Patients, n	13	11	6	30
Pre-surgery n (%)	OGTT	13 (43.3%)	13 (43.3%)	4 (13.4 %)	30
	MTT	16 (53.3%)	12 (40%)	2 (6.6%)	30
Post-surgery n (%)	OGTT	28 (93.3%)	2 (6.7%)	0	30***
	MTT	23 (76.7%)	5 (16.7%)	2 (6.6%)	30

The data are n (%). Changes in the patient’s numbers of normal glucose tolerance (NGT), impaired glucose tolerance (IGT), and diabetes mellitus (DM) were compared with the pre- and post-surgery OGTT and MTT data. We used a 2 × 3 contingency table in Excel for the χ²-test. Significant changes in the percentage of glucose tolerance subtypes between before and after LSG were analyzed using Fisher’s test. *** p < 0.001 vs. comparable data before surgery. The MTT data were not significantly different between before and after LSG.

We further investigated the parameters to regulate glucose tolerance curves. The Matsuda index, glucose-induced insulin secretion (GIS), and disposition index (DI) values were well correlated between the pre-surgical data of the OGTT and MTT as shown in Supplementary Fig. S3. As depicted in Supplementary Fig. S4, the post-surgery data of both GIS and disposition index were also each well correlated between the OGTT and MTT. However, the correlation for the post-surgery Matsuda index values between the OGTT and MTT data did not reach statistical significance (r = 0.35, p < 0.056).

The comparison of the Matsuda index, GIS, and disposition index (DI) in morbidly obese patients with NGT, IGT, and DM between before and after LSG

As shown in Table 4, the post-surgery Matsuda index and HOMA-R calculated from OGTT data in NGT, IGT, and DM patients were significantly improved compared to the corresponding pre-surgery data, respectively. Similar changes in both Matsuda index and HOMA-R measured by MTT were also found in NGT and IGT patients. However, the post-surgery Matsuda index and HOMA-R values calculated from MTT data in the DM patients were not significantly different from the pre-surgery values. Furthermore, these post-surgery data in DM patients were significantly lower than those of NGT patients. An increase in the Matsuda index calculated from OGTT data was significantly correlated with %body weight reduction (r = 0.47, p < 0.01) (Fig. 4A) and with plasma insulin levels (r = –0.78, p < 0.001) (Fig. 4B).

Table 4

Changes in parameters of insulin sensitivity and glucose-induced insulin secretion (GIS) calculated from the OGTT and MTT in morbidly obese patients with three glucose tolerance subtypes at 1 year post-surgery vs. pre-surgery

Glucose-loading test	Glucose tolerance sub-type	Pre-/Post-surgery	Insulin sensitivity		Glucose-induced insulin secretion (GIS)
			Matsuda index	HOMA-R	insulinogenic index ΔINS0–30/ΔPG0–30	Total GIS INS-AUC0–120/PG-AUC0–120
OGTT	NGT	Pre	4.6 (3.9–5.6)	1.8 (1.5–2.8)	0.9 (0.4–1.4)	39.9 (31.7–72.6)
		Post	7.6 (6.6–8.5)**	0.8 (0.7–0.9)**	1.0 (0.9–1.7)†	44.3 (38.8–64.6)†
	IGT	Pre	3.0 (2.2–3.3)	2.3 (2.1–2.9)	0.8 (0.6–1.3)†	51.0 (44.2–54.4)†
		Post	5.1 (4.0–8.4)**	1.0 (0.95–1.4)**	1.1 (0.7–1.4)†	50.8 (26.6–78.1)
	DM	Pre	2.9 (2.7–4.1)	3.1 (2.1–3.1)	0.4 (0.3–0.5)	26.7 (20.7–30.6)
		Post	5.6 (4.6–6.2)*	1.1 (0.9–1.3)*	0.5 (0.2–0.7)	35.6 (25.7–37.3)
MTT	NGT	Pre	5.0 (4.3–6.0)	1.7 (1.3–2.0)	1.7 (1.1–2.6)†	42.3 (36.2–65.8)†
		Post	9.5 (7.2–12.4)*†	0.7 (0.6–0.9)**†	1.2 (0.9–2.0)†	40.4 (29.5–49.2)
	IGT	Pre	3.7 (3.4–4.2)	2.3 (2.1–2.6)	1.1 (0.7–1.5)	41.0 (35.7–58.7) †
		Post	6.1 (5.5–8.0)**	0.9 (0.9–1.3)**	0.7 (0.6–1.3)	34.3 (24.0–37.1)
	DM	Pre	4.0 (3.4–4.6)	2.2 (1.9–2.6)	0.6 (0.3–0.8)	24.4 (22.1–31.6)
		Post	4.9 (4.2–6.6)	1.4 (1.1–1.9)	0.6 (0.5–0.8)	27.5 (19.6–28.0)

The non-parametric analysis of the significant difference of continuous variables (Matsuda index, HOMA-R, and GIS) among the three independent groups (NGT, IGT, and DM patients) was performed using the Kruskal-Wallis test. Data are median (Quartiles 1–3) (n = 30), ^† p < 0.05 vs. the corresponding values in DM patients, where the other group (NGT, IGT) data were compared using the Steel-Dwass test. ** p < 0.01, * p < 0.05 vs. pre-surgery values using the Wilcoxson rank-sum test.

Fig. 4

The correlation between the Matsuda index calculated from the patients’ OGTT data was positively correlated with their %body weight reduction and negatively correlated with the post-LSG surgery AUC of plasma insulin levels. The Matsuda index was positively correlated with the %body weight reduction (r = 0.47, p = 0.007) and negatively correlated with the AUC-IRI (r = –0.78, p < 0.001) after the LSG.

However, the patients’ glucose-induced insulin secretion (GIS) based on the data of the insulinogenic index or total GIS calculated from both the OGTT and MTT data did not differ significantly between before and after the LSG in the patients with glucose tolerance subtypes (Table 4). Concerning the differences in insulinogenic index among the 3 glucose tolerance subtypes, DM patients showed significantly lower or tended to be lower than that of the NGT or IGT patients both before and after LSG.

As shown in Supplementary Fig. S5, the Matsuda index (y-axis) and the total GIS (INS-AUC_0–120/PG-AUC_0–120) (x-axis) using the pre-surgery (A) and post-surgery (B) OGTT data in NGT, IGT, and DM patients were separately plotted. We found that each curve from OGTT data of all glucose tolerance subtypes was extrapolated to a hyperbolic curve, respectively. Pre-surgery hyperbolic curve in DM patients was plotted as numerically lower Matsuda index values than that of pre-surgery NGT patients. In a post-surgery OGTT data, the hyperbolic curve in each glucose intolerant subtype tended to be higher in Matsuda index compared to that of the pre-surgery data. Furthermore, the hyperbolic curve in DM patients was numerically lower compared with that of NGT patients. A similar hyperbolic relation between Matsuda index and GIS calculated from MTT data was also shown in NGT and IGT patients before (C) and after (D) LSG. DM patients was not fitted to hyperbolic curves. Similarly, DM patients also showed numerically lower than that of NGT patients.

DI is calculated by the formula of [Insulin sensitivity index, ISI] × [glucose-induced insulin secretion; GIS] [25, 31], which has been reported to be an indicator of β-cell function-associated glucose disposal activity [34]. In the present study, the pre-surgery DI calculated from OGTT (Fig. 5A) and MTT (Fig. 5C) data showed a significant stepwise reduction from the NGT, to IGT, and to DM groups. On the other hand, the post-surgery DI calculated from OGTT data was significantly higher than the corresponding pre-surgery values in all three glucose tolerance subtypes (Fig. 5B). In contrast, the post-surgery DI calculated from MTT data in the NGT patients also showed significantly higher values than that of the pre-surgery data, whereas the DIs in the patients with either IGT or DM after LSG were not significantly higher than each of the respective pre-surgical values (Fig. 5D).

Fig. 5

Comparison of the disposition index (DI) values in the patients with NGT, IGT, and DM calculated from the OGTT (A, B) and MTT (C, D) data between before (A, C) and after LSG surgery (B, D). Data are median and interquartile range (Q1–Q3). The significance of differences in the DI values among the three glucose-tolerance group was examined in a nonparametric analysis (Kruskal-Wallis test), and the Steel-Dwass method was used to test the significance. ** p < 0.01, * p < 0.05 between pairs of glucose-tolerance subtypes. Wilcoxon’s signed rank test was used to compare the significance of DI data before and after LSG surgery. ^†† p < 0.01, ^† p < 0.05 vs. the corresponding pre-surgery values.

In the present OGTT data, 14 patients (47%) showed a 2-h PG level <60 mg/dL after LSG. In our analysis performed to identify any factors related to hypoglycemia, we found only a negative linear correlation between the 2-h PG values and the DI (r = –0.525, p < 0.005) (Fig. 6C), without a significant correlation between the 2-h PG levels and either total GIS (Fig. 6A) or Matsuda index (IS) (Fig. 6B). In the present post-surgery MTT data, there were no patients with 2h-PG <60 mg/dL. On the other hand, we also found a significant negative linear correlation between the 2h-PG values and not only DI (Fig. 6F) (r = –0.673, p < 0.005), but also total GIS (r = –0.436, p < 0.01) (Fig. 6D) and Matsuda index (r = –0.487, p < 0.01), respectively (Fig. 6E).

Fig. 6

The correlation between PG₁₂₀ values during the OGTT (A–C) and MTT (D–F) and the total glucose-induced insulin secretion (GIS) (INS-AUC_0–120/PG-AUC_0–120) (A, D), Matsuda index (B, E), and disposition index (DI) (C, F) in patients with morbid obesity after an LSG. Among the total of 30 patients, the hypoglycemic range (i.e., PG₁₂₀ <60 mg/dL) after OGTT was detected in 14 postoperative patients. At 1 year after the LSG, the DI values (C) were negatively correlated (r = –0.525, p < 0.005) with the PG₁₂₀ values after the OGTT in all patients with morbid obesity. However, the 2h-PG levels were not significantly correlated with the total GIS (A) or the Matsuda index (B). Hypoglycemia was not observed in any patients during the post-surgery MTT. However, there were significant correlations between the post-MTT PG₁₂₀ values and the total GIS (D) (r = –0.436, p < 0.01), the Matsuda index (r = –0.487, p < 0.01) (E), and the DI (r = –0.673, p < 0.005) (F).

The differential regulation of the baseline and postprandial TG levels after MTT between before and after LSG

As shown in Fig. 7, the pre-surgery baseline TG values were not significantly different among NGT, IGT, and DM patients (A) and post-surgery TG values were significantly lower than each pre-surgery levels (B). The pre-surgery 2h-TG values progressively increased from NGT, to IGT, and to DM patients (C). Those values were significantly reduced in IGT and DM patients after LSG, resulting in no difference among 3 different glucose tolerance subtypes (D). Similarly, increases in 2h-TG values above the baseline after MTT (ΔTG_0–120) progressively increased from NGT, to IGT, and to DM patients (E). However, those values were not significantly suppressed after LSG compared to pre-surgery values and those post-surgery ΔTG_0–120 values were not different among 3 glucose tolerance subtypes (F).

Fig. 7

The baseline triglyceride (TG₀) and 2h-TG (TG₁₂₀) values and a meal-induced increase in the TG level (ΔTG_0–120) in the NGT, IGT, and DM groups were compared between before and after the patients’ LSG surgeries. The baseline TG (TG₀) (A, B), TG₁₂₀ values (C, D), and the meal-induced increase in TG levels (ΔTG_0–120) (E, F) were compared between before (A, C, E) and after (B, D, F) the LSG surgeries in the three glucose-tolerance subtype groups. A non-parametric analysis was used to examine the differences in TG₀, TG₁₂₀, and ΔTG_0–120 results among the glucose-tolerance groups by the Kruskal-Wallis test. The p-values were assessed by the Steel-Dwass test. ** p < 0.01, between pairs of glucose-tolerance subtypes. ^††† p < 0.001, ^†† p < 0.01, ^† p < 0.05 vs. each comparable pre-surgery value by Wilcoxon’s rank-sum test.

As shown in Table 5, the baseline TG₀ values in pre-surgery data were significantly correlated with HOMA-R, Matsuda index, and DI, but not correlated with HbA1c levels. In contrast, pre-surgery ΔTG_0–120 values were not correlated with the HOMA-R and Matsuda index but were correlated with the HbA1c and DI values. Pre-surgery TG₁₂₀ values were significantly correlated with all parameters (HOMA-R, HbA1c, Matsuda index, and DI). However, the post-surgery baseline TG₀ values were significantly correlated with the HbA1c values and post-surgery ΔTG_0–120 values were significantly correlated with Matsuda index. The post-surgery TG₁₂₀ values were significantly correlated with both HbA1c and Matsuda index values.

Table 5

Association of TG₀, ΔTG_0–120, and TG₁₂₀ with the Matsuda index (insulin sensitivity), insulin secretion, and HbA1c values in pre- and post- surgery patients

Pre-surgery	HOMA-R	HbA1c, %	Matsuda index	DI
TG0	r = 0.451, p < 0.01	ns	r = –0.526, p < 0.001	r = –0.460, p < 0.005
ΔTG0–120	ns	r = 0.495, p < 0.01	ns	r = –0.557, p < 0.001
TG120	r = 0.427, p < 0.01	r = 0.379, p < 0.05	r = –0.567, p < 0.001	r = –0.606, p < 0.001
Post-surgery	HOMA-R	HbA1c (%)	Matsuda index	DI
TG0	ns	r = 0.494, p < 0.01	ns	ns
ΔTG0–120	ns	ns	r = –0.415, p < 0.02	ns
TG120	ns	r = 0.432, p < 0.05	r = –0.310, p < 0.05	ns

A significant correlation between TG levels (TG₀, ΔTG_0–120, or TG₁₂₀) and HOMA-R, HbA1c, or Matsuda index was calculated pre- and post-surgery and depicted in this table, respectively. DI; disposition index, ns; not significant

Discussion

Our analyses revealed the following three major results. (i) Compared to the preoperative data in 30 morbidly obese patients with or without diabetes, a significant weight loss was established and then effectively maintained at 1 year after LSG. The LSG surgery was closely associated with the remission of diabetes as well as significant improvements of other cardiometabolic risks. (ii) The diagnosis of glucose intolerance by the OGTT data at 1-year post-LSG was affected by reactive hypoglycemia due to LSG-induced dumping syndrome. However, the new test meal containing 75 g of glucose and fat was useful for reaching the correct diagnosis of glucose-tolerance subtypes in morbidly obese patients after an LSG. The DI, which is an index of β-cell function-associated glucose-disposal capacity, showed a significant stepwise reduction from NGT, to IGT and to DM both before and after the LSGs. These impaired post-surgery DI values calculated from MTT data in the IGT and DM patients did not significantly increase compared to the pre-surgery values, and they were significantly lower than those of the NGT patients. (iii) However, the baseline and postprandial TG_0–120 levels in the post-surgery DM patients were normalized to the NGT patients’ data.

Marked cardiometabolic risk reduction and remission of diabetes in morbidly obese patients after LSG

An epidemiology investigation has reported that insulin resistance is associated with more prevalent cardiometabolic disorders than β-cell dysfunction; β-cell dysfunction was directly associated with the progression of diabetes but not with obesity-associated events [27]. In the present study, the maintenance of long-term weight loss 1-year after an LSG resulted in both diabetes remission and concomitant improvements of other cardiometabolic risks. The HbA1c levels of the patients with DM were well controlled from a median value of 7.0% to 5.5% without the use of any glucose-lowering drugs after the patients’ LSGs (Supplementary Table S1), which indicated that the remission was established after the LSG based on the present criteria of diabetes remission [4, 12, 35]. All of the present patients were overt diabetes at the first visit to our obesity clinic. The present all T2DM patients with morbid obesity had undergone an LSG after achieving a weight loss of ≥5% of the baseline values through pre-surgery medical treatment for ≥6 months. Thus, the DM patients’ pre-surgery values of hyperglycemia, blood pressure, TG, and ALT levels were lower than the respective values recorded before the initial visit’s data.

In addition, the patients’ mean total energy intake decreased significantly from 2,589 ± 655 kcal/day at the initial visit to our hospital to 914 ± 294 kcal/day (–65%) (p < 0.001) at 1 year after their LSGs. Such a low energy intake after surgery might be related to the maintenance of body weight loss and diabetes remission as well as the improvements in the other cardiometabolic risks.

Clinical usefulness of the new MTT for correctly evaluating changes in postprandial glucose in morbidly obese patients after LSG

The benefits of using the new MTT for evaluation of the postprandial glucose excursion in morbidly obese patients after metabolic surgery can be summarized as follows. (i) In our previous study, the diagnosis of glucose-tolerance subtypes as indicated by MTT results was comparable to that of the OGTT data before surgery [10]. In addition, the DI values calculated from the OGTT data were significantly correlated with those of the MTT data in the pre- and post-surgery patients, respectively (Supplementary Figs. S3 and S4). Furthermore, DI, Matsuda index, and GIS values calculated from the post-surgery MTT data were associated with the 2h-PG values, respectively (Fig. 6). Since the 2h-PG value after glucose load was used as an important diagnostic criterion for glucose intolerance, glucose tolerance curves might be closely determined by the GIS, Matsuda index, and DI values, respectively in post-surgery patients. (ii) A new MTT was useful for the correct diagnosis of the glucose-tolerance state in the post-LSG patients. In this study, the stimulation of insulin secretion by the MTT was at a lesser degree than that by the OGTT, and we thus observed that none of the patients showed post-surgery 2h-PG levels <60 mg/dL by the MTT. In contrast, we observed that hypoglycemia after the OGTT was present in 47% of the post-surgery patients, which might be overdiagnosis of normalization of glucose tolerance curve. This postprandial hypoglycemia during OGTT was induced by excessive insulin secretion stimulated by dumping syndrome after the LSG (Supplementary Fig. S2). GLP-1 secretion might be associated with this postprandial hypoglycemic event [36, 37].

Improvements in the DI as a crucial factor to evaluate the improvements of glucose-tolerance curves in morbidly obese patients after LSG

The disposition index (DI) is calculated by multiplying the insulin sensitivity index (ISI) by the GIS [25, 31] and has been reported to be an indicator of β-cell function-associated glucose disposal activity [34]. In the present study, the DI values calculated using the patients’ OGTT data and their MTT data both progressively decreased from the NGT, to IGT, and to DM groups before and after the LSG. These impaired pre- and post-surgery DI values in DM patients were consistent with the differences in PG-AUC values among the NGT, IGT, and DM patients in both OGTT and MTT data. The post-surgery DI values calculated using the MTT data of the IGT and DM patients were not significantly increased compared to the respective pre-surgery data. These results suggested that the impaired post-operative β-cell function-dependent glucose disposal activity evaluated by the DI in T2DM patients might be associated with impaired normalization of glucose tolerance after the LSG surgery. These results were coincident with the fact that DI values were closely negatively correlated with 2h-PG values after the MTT (Fig. 6F). Interestingly, in the present study, the 2h-PG values during the post-surgery MTT as well as OGTT were significantly correlated with the DI values, respectively. These results indicate that the 2-hr PG value obtained during the MTT and OGTT were closely associated with the post-surgery DI values.

Dutia et al. also reported limited improvements of β-cell function after RYGB surgery [38]. A reduced DI value in a patient with DM reflected by insufficient β-cell compensatory function may therefore be a major cause of impaired post-surgery normalization of glucose intolerance compared to NGT patients. Rothberg et al. reported short-term weight loss improves not only insulin sensitivity but also β-cell function with the use of a very-low-energy diet (VLED) in obese patients with or without glucose intolerance. However, the long-term preservation of the β-cell function was generally difficult in the DM patients [39]. Steven et al. reported that at a 6-month follow-up, T2DM patients with stable weight reduction after a VLED exhibited a reduction in their HbA1c values from 7.1% to 5.8%. The responders were characterized by improvements of the first-phase insulin response, which progressively deteriorated over 2 years in the DM patients [40]. Similarly, Nannipieri et al. reported that the remission of diabetes was closely related to the pre-surgery β-cell function, which progressively worsened from early remitters, to late remitters, and to non-remitters [41]. Consistently, it has also been shown that preintervention β-cell function and changes in gut hormones after surgery predict the remission of diabetes in obese individuals after metabolic surgery [42]. These results suggest that pre- and post-surgery preservation of β-cell function in response to glucose loading is a crucial factor to establish diabetes remission after body-weight reduction. Even though our present patients with diabetes achieved clinical remission in accord with the currently used criteria, they showed impaired normalization of the glucose-intolerance curve in relation to impaired post-surgery DI values measured by the MTT.

In the present study, the patients’ insulin secretory activity was analyzed with the use of mainly their plasma insulin levels in response to glucose loading. It has been reported that hepatic insulin clearance is enhanced in patients shortly after RYGB surgery, which affects plasma concentrations [43]. Although we could not determine whether our patients’ in vivo insulin clearance was modified by the LSG surgery, the blood kinetics of the patients’ plasma insulin and their CPR responses showed similar in vivo excursions in both of the glucose-loading tests before and after the LSG. We thus concluded that the measurement of pancreatic insulin secretory potency could be correctly evaluated by determining the insulin-secretory pattern in patients’ OGTT and MTT test results at 1 year after they have undergone an LSG.

Normalization of baseline values and postprandial increase in triglyceride (TG) levels in the MTT after LSG

Both insulin resistance and β-cell dysfunction are associated with postprandial hyperglycemia and postprandial hyper-TG levels. Thus, patients with impaired glucose intolerance and hypertriglyceridemia have shown higher HOMA-R values and an impaired insulinogenic index [27]. In the present study, the patients’ baseline pre-surgery TG values reflecting endogenous TG metabolism at the fasting state were significantly correlated with their HOMA-R, Matsuda index, and DI values, but not with their HbA1c values. The baseline HOMA-R values were not significantly different among the three glucose-tolerance subtypes, which was coincident with no significant difference in the baseline TG values among those pre-surgery patients.

It was recently reported that the postprandial TG response to a high fat meal is partially explained by visceral adiposity, insulin resistance, and physical activity [44]. It is also generally understood that poor glycemia control in diabetes with insulin deficiency leads to abnormal lipoprotein metabolism and hypertriglyceridemia [26]. Consistently, our present findings demonstrated that postprandial TG₁₂₀ values were linearly correlated with the HOMA-R, Matsuda index, DI, and HbA1c levels, respectively. Thus, the TG₁₂₀ values were significantly increased from the NGT to IGT to DM groups, and these values were significantly suppressed to similar levels among those groups postoperatively, due to marked reductions of the Matsuda index and HbA1c values.

In contrast, a postprandial TG increase above the basal level under a high-fat meal (ΔTG_0–120) reflected mainly exogenous fat intake though intestinal lipid handling [45]. In the present study, the pre-surgery ΔTG_0–120 values in IGT and DM patients were significantly higher than the respective pre-surgery data of NGT patients. This result could be explained by our observation that the pre-surgery ΔTG_0–120 values were significantly correlated with HbA1c or DI values but not with the HOMA-R values or the Matsuda index. Interestingly, our patients’ post-surgery ΔTG_0–120 values were not suppressed compared with their pre-surgery values, although these values were not significantly different among the NGT, IGT, and DM patients.

To summarize these TG data, we observed that TG_0–120 values after the MTT were useful for evaluation of normalization of the postprandial TG metabolism after an LSG, since those values were weakly associated with both the HbA1c and Matsuda index values, but were not associated with the post-surgery DI values in the present study. These results suggest that the postprandial TG metabolism are not tightly related to the normalization of the DI values in post-surgery patients.

Study limitations

The results of this study demonstrated that a new MTT using a meal containing 75 g of glucose and fat was more useful than the standard 75-g OGTT for a precise assessment of glucose-tolerance subtypes and triglyceride metabolism after an LSG. However, this study’s patients were recruited exclusively from our outpatient clinic, and the 30 patients completed two different glucose-tolerance tests both before and 1 year after an LSG. These results were obtained for relatively small numbers of patients: 13 patients with NGT, 11 with IGT, and six with T2DM. The patients with T2DM had good preoperative glycemic control and were not on insulin therapy for the glycemic control of their diabetes. In this observational follow-up study, the postprandial blood glucose and lipid abnormality data obtained before and after the patients’ LSGs were very consistent, showing significant differences between the OGTT and MTT results. However, it would be appropriate to further test our findings in diabetic patients with more widely differing clinical characteristics.

The patients’ plasma concentrations of various incretins and other hormones were measured during the OGTT and MTT both pre- and post-LSG, but this study was designed to analyze only the utility of the new MTT compared to the standard OGTT as a diagnostic tool for postprandial glycemic and triglyceride elevations after metabolic surgery, and thus the focus of this report is the detailed analyses of the pre- and post-surgery differences in insulin secretion, GIS, and the DI calculated using OGTT versus MTT data. The precise relationship between the changes in hormone values in relation to the changes in both the insulin sensitivity and GIS values will be provided in a separate report.

In conclusion, the impaired glucose tolerance curve measured by the MTT in T2DM patients was associated with impaired normalization of the DI values in the patients 1-year after the LSG. In contrast, the postprandial hypertriglyceridemia at 2h after the MTT was normalized post-surgery. These results indicate that the 2h-PG value obtained during the MTT was finely regulated by both insulin sensitivity and insulin secretion in post-LSG patients, which related to the correct diagnosis of glucose tolerance curves under the no dumping syndrome after MTT. The exact reason for the impaired normalization of glucose tolerance curve compared to the normalization of postprandial TG levels after MTT in post-surgery DM patients was unclear in the present study. However, one explanation was that the 2h-PG value after the glucose loading, a major determinant of a glucose tolerance curve was directly correlated with the post-surgery DI values, while postprandial TG_0–120 values were not correlated with those values in the present study.

Acknowledgements

We thank all of the study participants and the staff at Omi Medical Center for their valuable contributions, and we sincerely appreciate the financial support from Sunstar Co., Ltd.

Disclosure

A. Kashiwagi has acted as a medical consultant for Sunstar Co., Ltd. The Omi Medical Center is funded by Sunstar Co., Ltd., and a portion of this funding was used for the present study. T. Kitamura is a member of Endocrine Journal’s editorial board. The remaining authors have no competing interests to disclose.

Internal Ethics Guideline

The Institutional Review Board at the Kusatsu General Hospital approved the study protocol (no. 2017-0317-05, March 24, 2017). Note that Kusatsu General Hospital changed its hospital name to Omi Medical Center on October 1, 2021.

References

• 1   Schauer  PR,  Bhatt  DL,  Kirwan  JP,  Wolski  K,  Brethauer  SA, et al. (2014) Bariatric surgery versus intensive medical therapy for diabetes—3-year outcomes. N Engl J Med 370: 2002–2013.
• 2   Nielsen  HJ,  Nedrebo  BG,  Foss  A,  Andersen  JR,  Assmus  J, et al. (2021) Seven-year trajectories of body weight, quality of life and comorbidities following Roux-en-Y gastric bypass and sleeve gastrectomy. Int J Obes (Lond) 46: 739–749.
• 3   Rubino  F,  Nathan  DM,  Eckel  RH,  Schauer  PR,  Alberti  KGMM, et al. (2017) Delegates of the 2nd diabetes surgery summit. Metabolic surgery in the treatment algorithm for type 2 diabetes: a joint statement by international diabetes organizations. Obes Surg 27: 2–21.
• 4   Riddle  MC,  Cefalu  WT,  Evans  PH,  Gerstein  HC,  Nauck  MA, et al. (2021) Consensus report: Definition and interpretation of remission in type 2 diabetes. J Clin Endocrinol Metab 107: 1–9.
• 5   Roslin  M,  Damani  J,  Oren  J,  Andrews  R,  Yatco  E, et al. (2011) Abnormal glucose tolerance testing following gastric bypass demonstrates reactive hypoglycemia. Surg Endosc 25: 1926–1932.
• 6   Kim  SH,  Liu  TC,  Abbasi  F,  Lamendola  C,  Morton  JM, et al. (2009) Plasma glucose and insulin regulation is abnormal following gastric bypass surgery with or without neuroglycopenia. Obes Surg 19: 1550–1556.
• 7   Lee  CI,  Clark  JM,  Schweitzer  M,  Magnuson  T,  Steele  K, et al. (2015) Prevalence of and risk factors for hypoglycemic symptoms after gastric bypass and sleeve gastrectomy. Obesity (Silver Spring) 23: 1079–1084.
• 8   Capristo  E,  Panunzi  S,  De Gaetano  A,  Spuntarelli  V,  Bellantone  R, et al. (2018) Incidence of hypoglycemia after gastric bypass vs. sleeve gastrectomy: a randomized trial. J Clin Endocrinol Metab 106: 2136–2146.
• 9   Eisenberg  D,  Azagury  DE,  Ghiassi  S,  Grover  BT,  Kim  JJ (2017) ASMBS position statement on postprandial hyperinsulinemic hypoglycemia after bariatric surgery. Surg Obes Relat Dis 13: 371–378.
• 10   Yamamoto  Y,  Ozamoto  Y,  Kobayashi  M,  Tezuka  Y,  Azum  C, et al. (2022) Effects of a new 75 g glucose- and high fat-containing cookie meal test on postprandial glucose and triglyceride excursions in morbidly obese patients. Endocr J 69: 689–703.
• 11   Ugi  S,  Morino  K,  Yamaguchi  T,  Yamamoto  H,  Kaida  S, et al. (2019) Preserving β-cell function is the major determinant of diabetes remission following laparoscopic sleeve gastrectomy in Japanese obese diabetic patients. Endocr J 66: 817–826.
• 12   Lee  WJ,  Almulaifi  A,  Tsou  JJ,  Ser  KH,  Lee  YC, et al. (2015) Laparoscopic sleeve gastrectomy for type 2 diabetes mellitus predicting the success by ABCD score. Surg Obes Relat Dis 11: 991–996.
• 13   Bojsen-Møller  KN,  Dirksen  C,  Jørgensen  NB,  Jacobsen  SH,  Serup  AK, et al. (2014) Early enhancements of hepatic and later of peripheral insulin sensitivity combined with increased postprandial insulin secretion contribute to improved glycemic control after Roux-en-Y gastric bypass. Diabetes 63: 1725–1737.
• 14   Mallipedhi  A,  Prior  SL,  Barry  JD,  Caplin  S,  Baxter  JN, et al. (2014) Temporal changes in glucose homeostasis and incretin hormone response at 1 and 6 months after laparoscopic sleeve gastrectomy. Surg Obes Relat Dis 10: 860–869.
• 15   Papamargaritis  D,  Tzovaras  G,  Sioka  E,  Zachari  E,  Koukoulis  G, et al. (2017) Comparison of glucose homeostasis parameters between patients with high and low risk of diabetes at 6 weeks and 6 months after sleeve gastrectomy. Surg Obes Relat Dis 13: 1016–1024.
• 16   Hall  KD,  Kahan  S (2018) Maintenance of lost weight and long-term management of obesity. Med Clin North Am 102: 183–197.
• 17   Mingrone  G,  Panunzi  S,  De Gaetano  A,  Guidone  C,  Iaconelli  A, et al. (2012) Bariatric surgery versus conventional medical therapy for type 2 diabetes. N Engl J Med 366: 1577–1585.
• 18   Schauer  PR,  Bhatt  DL,  Kirwan  JP,  Wolski  K,  Aminian  A, et al. (2017) Bariatric surgery versus intensive medical therapy for diabetes — 5-year outcomes. N Engl J Med 376: 641–651.
• 19   Takahara  M,  Shiraiwa  T,  Maeno  Y,  Yamamoto  K,  Shiraiwa  Y, et al. (2022) Improvement of beta-cell function in conjunction with glycemic control after medical nutrition therapy in newly-diagnosed type 2 diabetes mellitus. BMC Endocr Disord 22: 150.
• 20   Harrison  LB,  Adams-Huet  B,  Raskin  P,  Lingvay  I (2012) β-cell function preservation after 3.5 years of intensive diabetes therapy. Diabetes Care 35: 1406–1412.
• 21   Nosso  G,  Griffo  E,  Cotugno  M,  Saldalamacchia  G,  Lupoli  R, et al. (2016) Comparative effects of Roux-en-Y gastric bypass and sleeve gastrectomy on glucose homeostasis and incretin hormones in obese type 2 diabetic patients: a one-year prospective study. Horm Metab Res 48: 312–317.
• 22   Morínigo  R,  Lacy  AM,  Casamitjana  R,  Delgado  S,  Gomis  R, et al. (2006) GLP-1 and changes in glucose tolerance following gastric bypass surgery in morbidly obese subjects. Obes Surg 16: 1594–1601.
• 23   Nguyen  KT,  Billington  CJ,  Vella  A,  Wang  Q,  Ahmed  L, et al. (2015) Preserved insulin secretory capacity and weight loss are the predominant predictors of glycemic control in patients with type 2 diabetes randomized to Roux-en-Y gastric bypass. Diabetes 64: 3104–3110.
• 24   Prasad  M,  Mark  V,  Ligon  C,  Dutia  R,  Nair  N, et al. (2022) Role of the gut in the temporal changes of beta-cell function after gastric bypass in individuals with and without diabetes remission. Diabetes Care 45: 469–476.
• 25   Retnakaran  R,  Shen  S,  Hanley  AJ,  Vuksan  V,  Hamilton  JK, et al. (2008) Hyperbolic relationship between insulin secretion and sensitivity on oral glucose tolerance test. Obesity (Silver Spring) 16: 1901–1907.
• 26   Goldberg  IJ (2001) Clinical review 124. Diabetic dyslipidemia: causes and consequences. J Clin Endocrinol Metab 86: 965–971.
• 27   Wang  T,  Zhao  Z,  Xu  Y,  Qi  Lu,  Xu  M, et al. (2018) Insulin resistance and β-cell dysfunction in relation to cardiometabolic risk patterns. J Clin Endocrinol Metab 103: 2207–2215.
• 28   Haruta  H,  Kasama  K,  Ohta  M,  Sasaki  A,  Yamamoto  H, et al. (2017) Long-term outcomes of bariatric and metabolic surgery in Japan: results of a multi-institutional survey. Obes Surg 27: 754–762.
• 29   Tewari  N,  Awad  S,  Macdonald  IA,  Lobo  DN (2018) A comparison of three methods to assess body composition. Nutrition 47: 1–5.
• 30   Matthews  DR,  Hosker  JP,  Rudenski  AS,  Naylor  BA,  Treacher  DF, et al. (1985) Homeostasis model assessment: insulin resistance and β cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28: 412–419.
• 31   Matsuda  M,  DeFronzo  RA (1999) Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 22: 1462–1470.
• 32   Kahn  SE,  Prigeon  RL,  Mcculloch  DK,  Boyko  EJ,  Bergman  RN, et al. (1993) Quantification of the relationship between insulin sensitivity and β-cell function in human subjects: Evidence for hyperbolic function. Diabetes 42: 1663–1672.
• 33   Seino  Y,  Nanjo  K,  Tajima  N,  Kadowaki  T,  Kashiwagi  A, et al. (2010) Report of the committee on the classification and diagnostic criteria of diabetes mellitus. J Diabetes Investig 1: 213–228.
• 34   Bergman  RN,  Ader  M,  Huecking  K,  Van Citters  G (2002) Accurate assessment of beta-cell function: the hyperbolic correction. Diabetes 51 (Supp 1): S212–S220.
• 35   Lee  WJ,  Hur  KY,  Lakadawala  M,  Kaasama  K,  Wong  SK, et al. (2013) Predicting success of metabolic surgery: age, body mass index, c-peptide, and duration. Surg Obes Related Dis 9: 389–484.
• 36   Jorgensen  NB,  Dirksen  C,  Bojsen-Moller  KN,  Jacobsen  SH,  Worm  D, et al. (2013) Exaggerated glucagon-like peptide 1 response is important for improved β-cell function and glucose tolerance after Roux-en-Y gastric bypass in patients with type 2 diabetes. Diabetes 62: 3044–3052.
• 37   Salehi  M,  Gastaldelli  A,  D’Alessio  DA (2014) Blockade of glucagon-like peptide 1 receptor corrects postprandial hypoglycemia after gastric bypass. Gastroenterology 146: 660–680.
• 38   Dutia  R,  Brakoniecki  K,  Bunker  P,  Paultre  F,  Homel  P, et al. (2014) Limited recovery of β-cell function after gastric bypass despite clinical diabetes remission. Diabetes 63: 1214–1223.
• 39   Rothberg  AE,  Herman  WH,  Wu  C,  IglayReger  HB,  Horowitz  JF, et al. (2020) Weight loss improves β-cell function in people with severe obesity and impaired fasting glucose: a window of opportunity. J Clin Endocrinol Metab 105: e1621–e1630.
• 40   Steven  S,  Hollingsworth  KG,  Al-Mrabeh  A,  Avery  L,  Aribisala  B, et al. (2016) Very low-calorie diet and 6 months of weight stability in type 2 diabetes: pathophysiological changes in responders and nonresponders. Diabetes Care 39: 808–815. Erratum in: Diabetes Care 41: 1321.
• 41   Nannipieri  M,  Mari  A,  Anselmino  M,  Baldi  S,  Barsotti  E, et al. (2011) The role of β-cell function and insulin sensitivity in the remission of type 2 diabetes after gastric bypass surgery. J Clin Endocrinol Metab 96: E1372–E1379.
• 42   Nannipieri  M,  Baldi  S,  Mari  A,  Colligiani  D,  Guarino  D, et al. (2013) Roux-en-Y gastric bypass and sleeve gastrectomy: mechanisms of diabetes remission and role of gut hormones. J Clin Endocrinol Metab 98: 4391–4399.
• 43   Bojsen-Moller  KN,  Dirksen  C,  Jorgensen  NB,  Jacobsen  SH,  Hansen  DL, et al. (2013) Increased hepatic insulin clearance after Roux-en-Y gastric bypass. J Clin Endocrinol Metab 98: E1066–E1071.
• 44   Wilson  SM,  Maes  AP,  Yeoman  CJ,  Walk  ST,  Miles  MP (2021) Determination of the postprandial triglyceride response to a high-fat meal in healthy overweight and obese adults. Lipids Health Dis 20: 107.
• 45   Wit  M,  Trujill-Viera  J,  Strohmeyer  A,  Klingenspor  M,  Haankir  M, et al. (2022) When fat meets the gut-focus on intestinal lipid handling in metabolic health and disease. EMBO Mol Med 14: e14742.