Relationships between computed tomography-assessed density, abdominal fat volume, and glucose metabolism after sleeve gastrectomy in Japanese patients with obesity

Abstract

In this study, we investigated the relationships between body weight (BW), computed tomography (CT)-assessed abdominal adipose tissue, and the glycemic metabolic profile in obese Japanese patients following laparoscopic sleeve gastrectomy (LSG). This study analyzed adipose tissue compartments using CT methods before and 1 year after LSG. Thirty obese patients were studied, and variables measured included visceral adipose tissue (VAT), subcutaneous adipose tissue (SAT), density of VAT (VAT-D), and density of SAT (SAT-D). We also examined the parameters in patients according to whether they had type-2 diabetes (T2DM). LSG induced significant losses in BW, SAT, and VAT after LSG. Additionally, SAT-D and VAT-D both increased and fasting plasma glucose (FPG) and HbA1c, but not C-peptide, decreased after surgery. ΔSAT and ΔVAT were positively related, and ΔSAT-D and ΔVAT-D were negatively related to ΔBW and/or FPG. Furthermore, a multivariate regression model showed that total BW loss (TBWL) was closely related to ΔSAT (β = 0.84; p < 0.001) and ΔVAT-D (β = –0.45; p < 0.05) and improvement of FPG was related to ΔVAT (β = 0.61; p < 0.05) after LSG. Finally, ΔFPG was correlated with ΔVAT in 16 T2DM patients (r = 0.58; p < 0.05) but not in non-T2DM patients. TBWL was related to ΔSAT and ΔVAT-D, and improvement of FPG was related to ΔVAT in obese Japanese patients after LSG.

THE POPULARITY of laparoscopic sleeve gastrectomy (LSG), which is performed worldwide as one type of bariatric surgery, has grown in recent years in the context of the low-risk of the technique, the low complication rate compared with gastric bypass surgery, and the fact that LSG provides significant weight reduction and a remarkable improvement in all aspects of type-2 diabetes mellitus (T2DM) [1-4]. LSG is also associated with variable reductions in body fat deposition. In general, the relative degree of adipose tissue loss after weight loss surgery has been studied utilizing different imaging methods, including impedance methods and others [5-12].

Several studies using computed tomography (CT) assessments of abdominal fat distribution have demonstrated associations of visceral adipose tissue (VAT) as well as abdominal subcutaneous adipose tissue (SAT) with various metabolic abnormalities [13-15].

Interestingly, the umbilical fat area assessed by CT is correlated with whole-abdominal fat volume, including SAT and VAT [16, 17]. VAT volume can cause metabolic abnormalities by secreting adipocytokines, which induce insulin resistance and T2DM. VAT is thought to be a more potent contributor to the development of obesity-related metabolic abnormalities compared with SAT volume because of its greater metabolic activity and its access to the circulation system [18-24]. One of the most important improvements in metabolic status after LSG is increased insulin sensitivity or T2DM remission [25-27].

The CT density of VAT (VAT-D) and SAT (SAT-D) is also associated with clinical risks independent of abdominal fat mass [28]. Lower CT abdominal fat density is associated with higher lipid content, whereas higher density reflects increased vascularity [29]. Additionally, VAT-D and SAT-D increase concurrently with fat loss independent of BMI and predict greater total fat loss after bariatric surgery [30].

Obese Japanese patients differ genetically from Caucasians in the distribution of abdominal fat [31-33]. Japanese subjects have higher abdominal VAT relative to abdominal SAT than do Caucasians [33]. This difference may explain, in part, the higher predisposition to T2DM in the Japanese population. However, little is known about the relationship between changes in VAT, SAT, and the glycemic metabolic disorders in obese Japanese patients, especially following LSG.

Therefore, the aim of this study was to evaluate associations of LSG with the CT-assessed abdominal adipose tissue deposition and density levels and the glycemic metabolic profile of obese Japanese patients. Furthermore, we also explored the potential interactions among these variables in T2DM and non-T2DM patients.

Materials and Methods

Patients and study design

We retrospectively evaluated 114 morbidly obese patients (mean age, 43.5 ± 6.8 years) undergoing LSG at Oita University Hospital from Nov 2007 to Feb 2018. The study group consisted of 30 obese patients (17 females and 13 males) who agreed to a CT examination and were referred to our outpatient clinic over a period of 1 year. We also examined the body weight (BW) and physiological metabolic parameters in patients with T2DM (n = 16) and non-T2DM patients (n = 14) diagnosed according to the following procedure. A multidisciplinary team consisting of a family physician, bariatric surgeon, dietician, and psychologist evaluated the obese patients. Selection criteria for bariatric surgery included a body mass index (BMI) greater than 35 kg/m² and the presence of significant comorbidities that could resolve with weight loss. Patients were selected for LSG according to the inclusion criteria developed based on the Japanese Society for Treatment of Obesity. All patients had been obese for at least 1 year and had failed with traditional weight control programs. All patients were screened for major endocrine disorders prior to surgery.

Patients with severe heart and renal disease were excluded from the study. The study was conducted according to the guidelines included in the Declaration of Helsinki. All subjects provided informed consent for their participation in the study, which was approved by the Ethical Committee of Oita University.

Anthropometric measurements and metabolic parameters

The anthropometric and body composition characteristics of the patients were evaluated using the following parameters: height, BW, BMI, and waist circumference. BMI was calculated as weight/height² (kilograms per square meter). The waist circumference was measured midway between the lower rib margin and the iliac crest in standing subjects after normal expiration. Blood pressure (BP) measurements were taken with a digital BP monitor that had an appropriate cuff size.

Blood was extracted from the antecubital vein with the patient in the recumbent position at 8:00 AM after an overnight fast. The patients underwent routine laboratory tests, including assays for plasma triglycerides, LDL-C, HDL-C, fasting plasma glucose (FPG), HbA1c, and C-peptide (CPR). BUN and creatinine (Cr) as well as uric acid levels were also evaluated. Blood samples were collected the morning following overnight fasting. For purposes of this study, we diagnosed diabetes according to the guidelines of the Japan Diabetes Society: FPG >126 mg/dL and HbA1c >6.5%.

Criteria for remission and improvement in T2DM

The criteria for remission and improvement in selected comorbidities were based on a previous study [8]. Complete remission included fasting glucose <100 mg/dL and HbA1c <6% without the use of medication for diabetes. Partial remission included fasting glucose levels of 100–125 mg/dL and HbA1c levels of 6–6.4% without medication. Improvements included a statistically significant reduction in HbA1c, the failure of FPG to meet the criteria for remission, or a decrease in antidiabetic medication requirements (by discontinuing insulin or one oral agent or a 50% reduction in dose).

CT imaging

Imaging was performed on a 16-detector row CT scanner (Canon, Tokyo, Japan). Patients were positioned in a supine position with both arms stretched above the head to avoid beam hardening artifacts. A single-slice CT scan was applied at the umbilical level (L3–L5). CT scans were performed with a 5-m slice thickness, 120 kvP, and 100 mA. Abdominal circumference (cm), cross-sectional areas (cm²), and mean Hounsfield unit (HU) for VAT, SAT, and liver were measured using semi-automated tracings with density thresholds between –250 and –50 HU. All female participants had a negative pregnancy test before the CT scan. All participants underwent a single-slice scan preoperatively and at 1 year postoperatively. All subjects underwent a CT scan at the level of the umbilicus for cross-sectional measurement of abdominal visceral fat areas and were analyzed with Fat scan version 3 software (N2 Systems, Osaka, Japan). Details of the procedures have been previously described [19]. This method was validated by other determinations of VAT and has been widely adopted as a practical method for evaluating regional adiposity. The software was used to calculate the areas of the total abdominal adipose tissue, SAT, and VAT. SAT was calculated by delineating the abdominal subcutaneous area outside the abdominal muscle wall with a graph pen and then computing subcutaneous fat using the software. For each patient with a single-slice CT, SAT and VAT were quantified in cm². The ratios of SAT to VAT (SAT/VAT), SAT-D, VAT-D, liver-D and the ratio of SAT-D to VAT-D (SAT-D/VAT-D) were also calculated. The CT images are presented here in an unidentified random fashion. For each patient with a single-slice CT, the SAT and VAT densities were quantified.

Statistical analysis

Statistical analysis of the data was performed using the JMP 13.2 program (SAS Inc, Irvine, CA, USA). Data are expressed as mean ± SD and were compared using nonparametric tests. A Wilcoxon paired test was used for comparisons between obese patients and between before and after LSG. Spearman’s analysis was performed to determine the relationship among changes in abdominal adiposity, glycemic profile, and the other continuous variables. Multiple linear regression analysis was conducted to assess the associations among FPG and abdominal fat after adjustment for other parameters including BW. Group differences or correlations of p < 0.05 were considered statistically significant.

Results

Baseline clinical characteristics of obese patients and time-course changes in body weight, %TBWL, and %EBWL

Table 1 shows the time-course changes in body weight (BW), %TBWL, %EBWL, and metabolic parameters. A significant reduction in BW was observed after LSG (Table 1). The BWs were 119.2 ± 18.0 kg before surgery and 79.2 ± 15.8 kg 1 year after surgery (p < 0.01). The %TBWL was 33.1 ± 10.5% 1 year after surgery and the %EBWL was 79.4 ± 28.0% 1 year after surgery (Table 1). There were no gender differences in delta-BW, %TBWL, or %EBWL in this study (data not shown).

Table 1 Basal and time-course changes in body weight and plasma metabolic parameters

	pre-LSG	12M	p
Age (years)	43.5 ± 6.8
Male/Female	13/17
Body weight (kg)	119.2 ± 18.0	79.2 ± 15.8	<0.001
Total body weight loss (kg)		40.0 ± 15.9
%TBWL		33.1 ± 10.5
%EBWL		79.4 ± 28.0
BMI (kg/m2)	45.1 ± 7.5	29.8 ± 5.4	<0.001
Waist circumstance (cm)	126.6 ± 14.0	99.6 ± 13.8	<0.001
Systolic blood pressure (mmHg)	137.5 ± 16.8	123.7 ± 15.6	<0.001
Diastolic blood pressure (mmHg)	86.4 ± 11.9	77.2 ± 12.6	0.002
Fasting plasma glucose (mg/dL)	123.5 ± 36.2	92.7 ± 16.8	<0.001
HbA1c (%)	6.9 ± 1.4	5.3 ± 0.5	<0.001
Plasma CPR (ng/mL)	3.8 ± 1.2	3.1 ± 2.2	0.09
Triglycerides (mg/dL)	174.6 ± 87.4	76.3 ± 42.4	<0.001
HDL cholesterol (mg/dL)	43.1 ± 8.2	58.0 ± 15.5	<0.001
LDL cholesterol (mg/dL)	125.6 ± 26.5	114.6 ± 33.2	0.053
BUN (mg/dL)	12.1 ± 3.6	13.4 ± 3.9	0.057
Creatinine (mg/dL)	0.65 ± 0.15	0.67 ± 0.16	0.48
AST (IU/L)	44.3 ± 38.1	17.4 ± 5.4	<0.001
ALT (IU/L)	55.4 ± 42.7	13.0 ± 5.3	<0.001
Gamma-GTP (IU/L)	47.9 ± 34.9	16.6 ± 11.6	<0.001

%TBWL, % total body weight loss; %EBWL, % excessive body weight loss; M, month ; mean ± SD

Time-course changes in plasma metabolic parameters

FPG, HbA1c, and triglyceride levels were all significantly decreased after LSG (Table 1). The average FPGs were 123.5 ± 36.2 mg/dL before surgery and 92.7 ± 16.8 mg/dL 1 year after surgery (p < 0.01) (Table 1). The plasma HbA1c levels were 6.9 ± 1.4% before surgery and 5.3 ± 0.5% 1 year after surgery (p < 0.01) (Table 1).

The fasting triglyceride levels were 174.6 ± 87.4 mg/dL before surgery and 76.3 ± 42.4 mg/dL 1 year after surgery (p < 0.01) (Table 1). By contrast, the plasma HDL levels increased from 43.1 ± 8.2 mg/dL before surgery to 58.0 ± 15.5 mg/dL 1 year after surgery (p < 0.01). The plasma LDL, BUN, and Cr levels did not significantly change during the study period (p > 0.05) (Table 1). The plasma CPR levels tended to decrease, from 3.8 ± 1.2 ng/mL before surgery to 3.1 ± 2.2 ng/mL 1 year after surgery, but this change was not statistically significant (p > 0.1). There were 16 patients with diabetes in the sample; 94% (15/16) of these patients underwent complete or partial remission.

Time-course changes in delta-SAT, VAT, SAT-D, VAT-D, and V/S ratio values by CT imaging

The SAT, VAT, and V/S ratio values were markedly decreased 1 year after surgery in comparison to the preoperative values (Fig. 1-A–C). By contrast, there were significant percentage increases in SAT-D and VAT-D levels 1 year after surgery (Fig. 1-D, E). The total volume reduction of SAT was significantly higher than the total volume reduction of VAT 1 year after LSG (p < 0.01). The total percent reduction of VAT significantly changed in comparison to the total percent reduction of SAT 1 year after LSG (p < 0.01), and the mean VAT/SAT ratio decreased from 0.43 ± 0.21 preoperatively to 0.32 ± 0.21 1 year postoperatively (p < 0.01) (Fig. 1-C). The liver density was significantly higher after LSG (pre-liver-D = 36.3 ± 18.2 HU, post-liver-D = 57.9 ± 9.6 HU; p < 0.01). In the present study, the VAT-D in women subjects were not significantly different from VAT-D in men (pre-VAT-D women = –115.2 ± 10.7 HU, and female ΔVAT-D = 9.3 ± 24.6 HU; pre-VAT-D men = –111.4 ± 18.3 HU and male ΔVAT-D = 16.0 ± 19.7 HU; p > 0.1).

Fig. 1

Changes in CT parameters in (A) SAT, (B) VAT, (C) V/S, (D) SAT-D, and (E) VAT-D, before surgery and 1 year after LSG. SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue; SAT-D, CT density of subcutaneous adipose tissue; VAT-D, CT density visceral adipose tissue; LSG, laparoscopic sleeve gastrectomy. * p < 0.01, ** p < 0.001 for pre-surgery.

Relationships between pre-surgery SAT, VAT, SAT-D, VAT-D, and V/S ratio values and ΔBW and physiological metabolic parameters after LSG

Pre-surgery SAT volume, but not VAT volume, was related to ΔBW (r = –0.50; p < 0.01) (Table 2). Additionally, pre-surgery VAT volume, but not SAT volume, was related to ΔFPG (r = –0.36; p < 0.05) (Table 2). ΔHbA1c was not related to pre-surgery SAT, VAT, SAT-D, VAT-D, or the V/S ratio (p > 0.1) (Table 2). We found no association of the pre-surgery SAT, VAT, SAT-D, VAT-D, and V/S ratio values with several liver and lipid metabolic parameters (p > 0.1) (data not shown).

Table 2 Correlation between abdominal fat parameters at pre-surgery and delta-body weight, delta-FPG, delta-HbA1c

Variables (pre)	delta BW		delta FPG		delta HbA1c
	r	p value	r	p value	r	p value
VAT (cm2)	0.05	0.78	–0.36	0.04*	–0.23	0.22
SAT (cm2)	–0.50	0.005**	–0.02	0.89	0.05	0.78
V/S ratio	0.40	0.03*	–0.17	0.36	–0.17	0.38
VAT-D (HU)	–0.08	0.65	–0.03	0.89	0.03	0.9
SAT-D (HU)	0.01	0.96	–0.06	0.76	–0.08	0.68

Variables: variables at pre-surgery: Visceral adipose tissue (VAT), Subcutaneous adipose tissue (SAT), V/S ratio, VAT CT density (VAT-D), SAT CT density (SAT-D), fasting plasma glucose (FPG) and Body weight (BW). r; correlation coefficient, * p < 0.05, ** p < 0.01.

Relationships between SAT, VAT, SAT-D, VAT-D, and V/S ratio values and ΔBW and physiological metabolic parameters after LSG

ΔBW was related to ΔSAT (r = 0.80; p < 0.001), ΔSAT-D (r = –0.40; p < 0.05), and ΔVAT-D (r = –0.44; p < 0.05) from pre-surgery to 1 year after surgery (Table 3). Additionally, ΔFPG was closely related to ΔVAT volume from pre-surgery to 1 year after surgery (r = 0.39; p < 0.05) (Table 3). ΔHbA1c from pre- to 1-year post-surgery were not related to such changes in SAT, VAT, SAT-D, VAT-D, and the V/S ratio (Table 3). There were no relationships between pre-surgery to 1-year changes in SAT, SAT density, the V/S ratio, and several liver and lipid metabolic parameters (p > 0.1) (data not shown). These relationships were observed in both females and males (data not shown).

Table 3 Correlation between delta-abdominal fat parameters and delta-body weight, delta-FPG, delta-HbA1c

Variables (delta)	delta BW		delta FPG		delta HbA1c
	r	p value	r	p value	r	p value
VAT (cm2)	0.34	0.06*	0.39	0.03*	0.3	0.12
SAT (cm2)	0.8	<0.001**	0.25	0.19	0.11	0.59
V/S ratio	–0.15	0.43	–0.10	0.59	0.002	0.99
VAT-D (HU)	–0.44	0.01*	–0.04	0.83	0.06	0.77
SAT-D (HU)	–0.40	0.02*	–0.004	0.98	0.15	0.44

Variables: delta (0–12 month) variables, VAT, Visceral adipose tissue; SAT, Subcutaneous adipose tissue, V/S ratio, VAT CT density (VAT-D), SAT CT density (SAT-D), HU: Hounsfield Unit, fasting plasma glucose (FPG) and Body weight (BW). r; correlation coefficient. *p < 0.05, **p < 0.01.

Additionally, multiple linear regression analysis was conducted to access the associations among BW, FPG, HbA1c, and abdominal fat using CT methods. Multivariate regression modeling showed that TBWL was associated with ΔSAT (β = 0.84; SE = 0.02; p < 0.001) and ΔVAT-D (β = –0.45; SE = 0.13; p < 0.05). Moreover, improvement of FPG was associated with ΔVAT (β = 0.61; SE = 0.17; p < 0.05) (Table 4).

Table 4 Multiple linear regression models with delta-body weight, delta-FPG and delta-HbA1c as the dependent variables

Variables (delta)	delta BW		delta FPG		delta HbA1c
	β	p value	β	p value	β	p value
VAT (cm2)	–0.04	0.78	0.61	0.02*	0.47	0.08
SAT (cm2)	0.84	<0.001**	0.04	0.87	0.05	0.85
V/S ratio	0.08	0.55	–0.35	0.14	–0.17	0.49
VAT-D (HU)	–0.45	0.02*	0.06	0.84	–0.03	0.93
SAT-D (HU)	0.31	0.11	0.16	0.60	0.36	0.31

Variables: variables at pre-surgery: Visceral adipose tissue (VAT), Subcutaneous adipose tissue (SAT), V/S ratio, VAT CT density (VAT-D), SAT CT density (SAT-D), fasting plasma glucose (FPG) and Body weight (BW). r; correlation coefficient, * p < 0.05, ** p < 0.01.

Baseline and time-course changes in plasma HbA1c, plasma CPR, VAT-D, and SAT-D in the T2DM group and non-T2DM group

Table 5 shows the time-course changes in plasma HbA1c, plasma CPR, VAT-D, and SAT-D in the T2DM group and non-T2DM group. Pre-operative HbA1c in the T2DM group was significantly higher than that in the non-T2DM group, and the HbA1c values decreased in both groups after LSG (Table 5). SAT-D increased in both in the T2DM and non-T2DM groups after LSG. VAT-D also increased in the non-T2DM group after LSG. The plasma CPR of the T2DM and non-T2DM groups did not significantly differ after LSG (Table 5).

Table 5 Basal and time-course changes in fat density and plasma metabolic parameters in type-2 diabetic and non type-2 diabetic patients

	pre-LSG	12M	p
Type-2 diabetic patients
HbA1c	7.9 ± 1.3	5.4 ± 0.6	<0.001
Plasm C-peptide (ng/mL)	3.9 ± 1.2	3.3 ± 2.3	0.25
VAT-D (HU)	–111.0 ± 17.6	–101.9 ± 15.4	0.078
SAT-D (HU)	–116.0 ± 11.7	–105.7 ± 11.6	0.036
Non-type-2 diabetic patients
HbA1c	5.8 ± 0.2	5.2 ± 0.3	0.003
Plasm C-peptide (ng/mL)	3.7 ± 1.3	2.5 ± 2.1	0.14
VAT-D (HU)	–116.4 ± 9.2	–100.6 ± 26.2	0.036
SAT-D (HU)	–119.6 ± 9.0	–100.3 ± 22.8	0.013

VAT-D, Visceral adipose tissue density; SAT, Subcutaneous adipose tissue density

Relationships between SAT, VAT, SAT-D, VAT-D, and V/S ratio values and ΔBW and metabolic parameters in T2DM and non-T2DM patients

The BW and physiological metabolic parameters of the patients with and without T2DM are presented in Table 6. The VAT-D before LSG and the ΔVAT-D of T2DM subjects were not significantly different from those of non-T2DM subjects (T2DM: pre-VAT-D = –111.0 ± 17.5 HU and ΔVAT-D = 9.1 ± 18.3 HU; non-T2DM: pre-VAT-D = –116.4 ± 9.1 HU and ΔVAT-D = 15.8 ± 26.6 HU; p > 0.1 for each). Similar to the results of the analysis of all patients, the ΔBW was related to the ΔSAT in both T2DM and non-T2DM obese patients. By contrast, the ΔFPG was related to the ΔVAT volume only in T2DM (r = –0.58; p < 0.05) but not in non-T2DM obese patients (p > 0.1; Table 6).

Table 6 Correlation between delta-abdominal fat parameters and delta-body weight, delta-FPG, delta-HbA1c in type-2 diabetic and non type-2 diabetic patients

Variables (delta)	delta BW		delta FPG		delta HbA1c
	r	p value	r	p value	r	p value
Type-2 diabetic patients
VAT (cm2)	0.36	0.17	0.58	0.02*	0.47	0.06
SAT (cm2)	0.79	<0.001**	0.36	0.17	0.13	0.64
V/S ratio	–0.05	0.84	–0.02	0.95	0.22	0.42
VAT-D (HU)	–0.08	0.76	–0.22	0.40	–0.22	0.42
SAT-D (HU)	–0.32	0.22	–0.19	0.47	–0.09	0.76
Non-type-2 diabetic patients
VAT (cm2)	0.34	0.35	–0.28	0.45	0.47	0.18
SAT (cm2)	0.88	<0.001**	–0.20	0.58	–0.03	0.94
V/S ratio	–0.40	0.26	–0.25	0.49	0.06	0.88
VAT-D (HU)	–0.75	0.01*	0.35	0.34	0.42	0.23
SAT-D (HU)	–0.65	0.04*	0.26	0.48	0.37	0.30

Variables: delta (0–12 month) variables: VAT, Visceral adipose tissue; SAT, Subcutaneous adipose tissue, V/S ratio, VAT CT density (VAT-D), SAT CT density (SAT-D), HU: Hounsfield Unit, fasting plasma glucose (FPG) and Body weight (BW). r; correlation coefficient. * p < 0.05, ** p < 0.01.

Discussion

The aim of this study was to explore changes in the distribution and density of abdominal adipose tissue after LSG and to assess the importance of such changes as determinants of improvements in the metabolic profiles of patients in response to LSG. With its excellent resolution of adipose tissue, CT offers a direct method for assessing visceral and subcutaneous fat deposition and density [19]. The level of umbilicus was chosen in terms of achieving reproducibility and reliability [16, 17].

In this study, LSG reduced overall abdominal VAT as well as SAT. Strong correlations between TBWL and SAT decrease were identified. We also found that SAT rather than VAT was related to TBWL after LSG. Additionally, high SAT before surgery was related to TBWL in obese patients. This might be because the reduction in the total fat volume of the SAT was significantly higher than the reduction in the VAT volume 1 year after LSG. In fact, the SAT volume of the morbidly obese patients was 2–3 times higher than their VAT volume. Thus, the reduction in BW was more dependent on pre-surgery SAT than VAT volume. EBWL was not related to SAT change or SAT-D in this study as the actual volume of TBWL differs from EBWL. The total percent reduction of VAT significantly changed in comparison to the total percent reduction of SAT, and TBWL was associated with ΔVAT-D rather than ΔVAT in this study. This is not surprising, as lower CT density in adipose tissue suggests lipid-rich fat and adipocyte hypertrophy [30]. The results showing that TBWL was associated with ΔVAT-D rather than ΔVAT indicate that fat volume may not be enough for a full assessment of the effects of LSG. Interestingly, the association between delta BW and ΔVAT-D, ΔSAT-D was confirmed only in the non-T2DM group. This observation suggested the possibility that several glycemic or hormonal changes, and/or medications in the T2DM group, might have influenced the results. Further studies are needed to clarify the mechanism underpinning the association between delta BW and fat density in the T2DM and non-T2DM groups.

Consistent with other studies, our results reflect significant improvements in the anthropometric variables in obese patients after LSG [5-12]. Moreover, our study confirms that LSG induced a substantial improvement in several glucose parameters. Specifically, high VAT before surgery was linked to ΔFPG after LSG. VAT is a metabolically active fat deposition that may differentially contribute to the metabolic consequences of T2DM. VAT is also thought to play an important role in the expression of the metabolic complications of diabetes due to its unique position with respect to portal circulation and its secretory function for various adipocytokines [18-24]. VAT is also an important part of the endocrine system that is involved in the complex interrelationship between obesity and T2DM. VAT, along with other ectopic fat depositions, has been associated with glucose homeostasis and insulin resistance [18, 34]. In the context of these observations, it is quite reasonable to suspect that VAT is closely related to the level of FPG. Markers of glucose metabolism other than FPG tended to correlate with CT-assessed parameters in this study, but these relationships did not reach statistical significance. We also obtained results regarding pre-VAT–FPG relationships in T2DM patients. In contrast to FPG, the absence of a relationship between ΔHbA1c and body-fat parameters may imply the operation of body-fat-independent metabolic effects on HbA1c and/or the influence of postprandial glucose changes after LSG. Additionally, the plasma CPR of the T2DM group did not significantly differ after LSG. These findings suggest that glucose and hormonal changes other than plasma CPR may have influenced VAT in the T2DM group. Nonetheless, a large-scale analysis might reveal relationships between CT-assessed parameters and glucose metabolism in greater detail.

To our knowledge, this study may be the first to use CT systems to examine relationships between CT-assessed fat distribution/density and glucose metabolic parameters in obese Japanese individuals who have undergone LSG. Obese Japanese subjects have different body compositions than Caucasians. In particular, Japanese people have higher abdominal VAT than Caucasians [33]. The results indicated different impacts of abdominal adipose tissue on glucose metabolism. Multiple regression analysis demonstrated that high VAT was closely related to improved FPG in obese Japanese patients following LSG.

We also examined the BW and physiological metabolic parameters in patients with and without T2DM. The ΔBW were related to ΔSAT in both T2DM and non-T2DM patients. Interestingly, ΔFPG were related to ΔVAT volume only in T2DM patients. We also found that high average levels of FPG at baseline in patients were related to VAT and glucose metabolic parameters.

This study has several limitations, including its small sample size, its duration, the possibility that it involved X-ray exposure, and its use of a longitudinal observational study design. Thus, this study does not allow for the identification of causal relationships, and future prospective studies that include appropriate control patients are needed to confirm our results. Additionally, this study did not fully assess glucose homeostasis, as there was no assessment of markers of insulin resistance, insulin secretion, postprandial glucose, or insulin levels. Although 94% of medication-free T2DM patients had diabetic remissions after LSG, we cannot fully exclude the influence of medication. Furthermore, as weight loss continued for at least 1 year after surgery, we did not investigate the effect of weight loss on abdominal lipid deposits during the period of greatest weight loss.

Taken together, our results show that significant changes in fat deposition and density occurred in these obese Japanese patients after LSG. Additionally, the TBWL was strongly related to ΔSAT and ΔVAT-D. The improvement in FPG is dependent on ΔVAT in obese Japanese T2DM patients after LSG.

Abbreviations

CT, computed tomography; LSG, laparoscopic sleeve gastrectomy; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue; SAT-D, CT density of subcutaneous adipose tissue; VAT-D, CT density visceral adipose tissue.

Acknowledgements

The authors thank RD Yuko Hirota and Kazuyo Adachi for their superb technical assistance. This work was supported by JSPS KAKENHI grant number JP17K01854.

Author Contributions

Y.O., T.M., Y.E., M.Ohta, M.I. and H.S. conceived and designed the study; Y.O., T.M., Y.Y., M.Okamoto and M.A. contributed to the analysis; Y.O., T.M., and K.G. contributed to the interpretation of data; Y.O., T.M., and H.S. wrote the paper.

Compliance with Ethical Standards

A statement of informed consent was obtained from all participants in the study.

Conflict of Interest

There was no conflict of interest related to the present study.

Statement of Human Rights

The study was approved by the Ethics Committee of Oita University, and it complied with the Declaration of Helsinki.

References

• 1   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.
• 2   Gill  RS,  Birch  DW,  Shi  X,  Sharma  AM,  Karmali  S (2010) Sleeve gastrectomy and type 2 diabetes mellitus: a systematic review. Surg Obes Relat Dis 6: 707–713.
• 3   Seki  Y,  Kasama  K,  Haruta  H,  Watanabe  A,  Yokoyama  R, et al. (2017) Five-year-results of laparoscopic sleeve gastrectomy with duodenojejunal bypass for weight loss and type 2 diabetes mellitus. Obes Surg 27: 795–801.
• 4   Naitoh  T,  Kasama  K,  Seki  Y,  Ohta  M,  Oshiro  T, et al. (2018) Efficacy of sleeve gastrectomy with duodenal-jejunal bypass for the treatment of obese severe diabetes patients in Japan: a retrospective multicenter study. Obes Surg 28: 497–505.
• 5   Zhang  C,  Yuan  Y,  Qiu  C,  Zhang  W (2014) A meta-analysis of 2-year effect after surgery: laparoscopic Roux-en-Y gastric bypass versus laparoscopic sleeve gastrectomy for morbid obesity and diabetes mellitus. Obes Surg 24: 1528–1535.
• 6   Murphy  R,  Clarke  MG,  Evennett  NJ,  John Robinson  S,  Lee Humphreys  M, et al. (2018) Laparoscopic sleeve gastrectomy versus banded Roux-en-Y gastric bypass for diabetes and obesity: a prospective randomised double-blind trial. Obes Surg 28: 293–302.
• 7   Friedrich  AE,  Damms-Machado  A,  Meile  T,  Scheuing  N,  Stingel  K, et al. (2013) Laparoscopic sleeve gastrectomy compared to a multidisciplinary weight loss program for obesity—effects on body composition and protein status. Obes Surg 23: 1957–1965.
• 8   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.
• 9   Ricci  C,  Gaeta  M,  Rausa  E,  Macchitella  Y,  Bonavina  L (2014) Early impact of bariatric surgery on type II diabetes, hypertension, and hyperlipidemia: a systematic review, meta-analysis and meta-regression on 6,587 patients. Obes Surg 24: 522–528.
• 10   Schiavo  L,  Scalera  G,  Pilone  V,  De Sena  G,  Iannelli  A, et al. (2017) Fat mass, fat-free mass, and resting metabolic rate in weight-stable sleeve gastrectomy patients compared with weight-stable nonoperated patients. Surg Obes Relat Dis 13: 1692–1699.
• 11   Otto  M,  Elrefai  M,  Krammer  J,  Weib  C,  Kienle  P, et al. (2016) Sleeve gastrectomy and Roux-en-Y gastric bypass lead to comparable changes in body composition after adjustment for initial body mass index. Obes Surg 26: 479–485.
• 12   Ozeki  Y,  Masaki  T,  Yoshida  Y,  Okamoto  M,  Anai  M, et al. (2018) Bioelectrical impedance analysis results for estimating body composition are associated with glucose metabolism following laparoscopic sleeve gastrectomy in obese Japanese patients. Nutrients 10: 1456.
• 13   Galanakis  CG,  Daskalakis  M,  Manios  A,  Xyda  A,  Karantanas  AH, et al. (2015) Computed tomography-based assessment of abdominal adiposity changes and their impact on metabolic alterations following bariatric surgery. World J Surg 39: 417–423.
• 14   Keidar  A,  Appelbaum  L,  Schweiger  C,  Hershkop  K,  Matot  I, et al. (2014) Baseline abdominal lipid partitioning is associated with the metabolic response to bariatric surgery. Obes Surg 24: 1709–1716.
• 15   Umemura  A,  Sasaki  A,  Nitta  H,  Otsuka  K,  Suto  T, et al. (2014) Effects of changes in adipocyte hormones and visceral adipose tissue and the reduction of obesity-related comorbidities after laparoscopic sleeve gastrectomy in Japanese patients with severe obesity. Endocr J 61: 381–391.
• 16   Furukawa  K,  Katabami  T,  Nakajima  Y,  Sato  T,  Kato  H, et al. (2010) Evaluation of whole-abdominal fat volume by 700-slice CT scanning and comparison with the umbilical fat area anthropometric indices. Obes Res Clin Pract 4: e83–e162.
• 17   Kobayashi  J,  Tadokoro  N,  Watanabe  M,  Shinomiya  M (2002) A novel method of measuring intra-abdominal fat volume using helical computed tomography. Int J Obes Relat Metab Disord 26: 398–402.
• 18  Ibrahim MM (2010) Subcutaneous and visceral adipose tissue: structural and functional differences. Obes Rev 11: 11–18.
• 19   Masaki  T,  Anan  F,  Yoshimatsu  H (2011) Visceral fat accumulation is associated with circadian blood pressure in Japanese patients with impaired glucose tolerance. Diabetes Care 34: e32.
• 20   Matsushita  Y,  Nagakawa  T,  Yamamoto  S,  Takahashi  Y,  Yokoyama  T, et al. (2012) Effect of longitudinal changes in visceral fat area and other anthropometric indices to the changes in metabolic risk factors in Japanese men: the Hitachi Health Study. Diabetes Care 35: 1139–1143.
• 21   Marinou  K,  Hodson  L,  Vasan  SK,  Fielding  BA,  Banerjee  R, et al. (2014) Structural and functional properties of deep abdominal subcutaneous adipose tissue explain its association with insulin resistance and cardiovascular risk in men. Diabetes Care 37: 821–829.
• 22   Oh  YH,  Moon  JH,  Kim  HJ,  Kong  MH (2017) Visceral-to-subcutaneous fat ratio as a predictor of the multiple metabolic risk factors for subjects with normal waist circumference in Korea. Diabetes Metab Syndr Obes 10: 505–511.
• 23   Mardinoglu  A,  Heiker  JT,  Gartner  D,  Bjornson  E,  Schon  MR, et al. (2015) Extensive weight loss reveals distinct gene expression changes in human subcutaneous and visceral adipose tissue. Sci Rep 5: 14841.
• 24   Kagawa  M,  Kerr  D,  Uchida  H,  Binns  CW (2006) Differences in the relationship between BMI and percentage body fat between Japanese and Australian–Caucasian young men. Br J Nutr 95: 1002–1007.
• 25   Carroll  JF,  Franks  SF,  Smith  AB,  Phelps  DR (2009) Visceral adipose tissue loss and insulin resistance 6 months after laparoscopic gastric banding surgery: a preliminary study. Obes Surg 19: 47–55.
• 26   Kashyap  SR,  Bhatt  DL,  Wolski  K,  Watanabe  RM,  Abdul-Ghani  M, et al. (2013) Metabolic effects of bariatric surgery in patients with moderate obesity and type 2 diabetes: analysis of a randomized control trial comparing surgery with intensive medical treatment. Diabetes Care 36: 2175–2182.
• 27   Eickhoff  H,  Guimaraes  A,  Louro  TM,  Seica  RM,  Castro  E, et al. (2015) Insulin resistance and beta cell function before and after sleeve gastrectomy in obese patients with impaired fasting glucose or type 2 diabetes. Surg Endosc 29: 438–443.
• 28   Rosenquist  KJ,  Pedley  A,  Massaro  JM,  Therkelsen  KE,  Murabito  U, et al. (2013) Visceral and subcutaneous fat quality and cardiometabolic risk. JACC Cardiovasc Imaging 6: 762–771.
• 29   Hu  HH,  Chung  SA,  Nayak  KS,  Jackson  HA,  Gilsanz  V (2011) Differential computed tomographic attenuation of metabolically active and inactive adipose tissues: preliminary findings. J Comput Assist Tomogr 35: 65–71.
• 30   Torriani  M,  Oliveira  AL,  Azevedo  DC,  Bredella  MA,  Yu  EW (2015) Effects of Roux-en-Y gastric bypass surgery on visceral and subcutaneous fat density by computed tomography. Obes Surg 25: 381–385.
• 31   Deurenberg  P,  Deurenberg-Yap  M,  Guricci  S (2002) Asians are different from Caucasians and from each other in their body mass index/body fat per cent relationship. Obes Rev 3: 141–146.
• 32   Sampei  MA,  Novo  NF,  Juliano  Y,  Sigulem  DM (2001) Comparison of the body mass index to other methods of body fat evaluation in ethnic Japanese and Caucasian adolescent girls. Int J Obes Relat Metab Disord 25: 400–408.
• 33  Bays H (2014) Central obesity as a clinical marker of adiposopathy; increased visceral adiposity as a surrogate marker for global fat dysfunction. Curr Opin Endocrinol Diabetes Obes 21: 345–351.
• 34   Chen  X,  Wu  Y,  Wang  L (2013) Fat-resident tregs: an emerging guard protecting from obesity-associated metabolic disorders. Obes Rev 14: 568–578.