Correlation between reduction of superior interventricular groove epicardial fat thickness and improvement of insulin resistance after weight loss in obese men

We prospectively enrolled 40 men in total, who were diagnosed with MetS using the ATP-III criteria, from outpatient clinics at Taichung Veterans General Hospital (Taichung, Taiwan) in April to September 2008 (Figure 1). Parts of the results have been published previously[23, 25, 26]. In brief, the study enrollment criteria included central obesity, which was defined as waist circumference ≥90 cm, in addition to at least two other ATP-III MetS criteria[32]. Patients with diabetes mellitus requiring oral anti-diabetic agents or insulin shots were excluded from the study. The patients undertook a 3-month weight reduction program. The dietitian instructed the study subjects to consume a 1200 Kcal diet per day comprising 55% carbohydrate, 30% fat, and 15% protein. Patients were educated how to estimate the calories and nutritional components per unit amount of food[25, 33, 34]. They all received an instruction brochure, which stated all the nutritional facts and calories per unit amount of food for instant check. The patients were asked to keep a very detailed record of each meal. A dietitian would estimate the calories based on the record in order to keep the daily requirement of 1200Kcal per day. In addition, the dietitian closely followed them with phone-call consultations. We encouraged peer competition for weight loss and showed weight and waist circumference changes in bar graphs at each gathering, which was held once every week in the first month, then biweekly in the 2nd and 3rd month. The patients were encouraged to have at least two to three light physical exercise periods per week. We provided a 60-minute fitness program with a physical trainer at the biweekly gatherings. The study subjects underwent anthropometric data check and blood sampling again at the 3^rd, 12^th and 18^th month after the beginning of weight loss program and then every 24 months thereafter if they are compliant with our recalls. Thirty-two of the obese non-diabetic men with MetS agreed undergoing MRI examination for measurement of subcutaneous fat, abdominal visceral fat, and EAT before and after weight loss (Figure 1). All subjects provided written informed consent. The authors confirm that all ongoing and related trials for this study/intervention are registered. The study protocol was approved by the Human Research Ethics Review Committee of Taichung Veterans General Hospital (Taichung, Taiwan). The protocol and case enrollment were conducted after getting the ethics committee approval; however, the registration at http://​www.​clinicaltrials.​gov (Multi-faceted Evaluations Following Weight Reduction in Subjects with Metabolic Syndrome NCT 01065753) was delayed till Feb 8, 2010 because clinical trial registration was not compulsory in our country back then.

All obese MetS subjects underwent blood tests and oral glucose tolerance test (OGTT) after an overnight fast to exclude those with unknown diabetes. After a fasting blood sample was collected, glucose load of 75 g was ingested over 5 minutes. Blood samples were collected at 30 minutes, 60 minutes, 90 minutes, and 120 minutes after the test load. Blood glucose and insulin concentrations were measured in each sample. Serum insulin was determined by a commercially available assay kit (IMMULITE, I-2000, EURO/Diagnostic Products Corporation, Gwynedd, UK). The inter- and intra-assay coefficients of variation for insulin (range 10.7 to 439 μU/ml) were 4.3% and 5.4%, respectively. Insulin-resistance was estimated using the homeostasis model assessment of IR (HOMA-IR), defined as fasting glucose mg/dl × fasting insulin μU/mL/405[35, 36]. OGTT-derived Matsuda insulin-sensitivity index (ISI) was defined as 10,000/sqrt [fasting glucose (mM) × fasting insulin (μU/mL) × mean glucose during OGTT (mM) × mean insulin during OGTT (μU/mL)][37]. OGTT glucose and insulin areas under the curve (AUC) were defined as areas under the curve for glucose and insulin vs. time during the OGTT, respectively[8]. Serum hs-CRP was determined by particle-enhanced immunoturbidimetry (Latex microparticles sensitized with duck anti-CRP IgY kit, provided by Good Biotech Corp., Taichung, Taiwan). The intra- and inter-assay coefficients of variance were 1.4% and 1.42%, respectively[8]. Serum monocyte chemotactic protein (MCP)-1 was determined by enzyme-linked immunosorbent assays (R&D Systems, Inc., Minneapolis, MN, USA)[38, 39]. The intra- and inter-assay coefficients of variation for MCP-1 were 5.80% and 5.70%, respectively, with a minimum detectable concentration of <5.0 pg/mL.

Serum adiponectin, leptin, and plasma plasminogen activator inhibitor (PAI)-1 concentrations were determined by enzyme-linked immunosorbent assays (R&D Systems, Inc., Minneapolis, MN, USA). The intra-assay and inter-assay coefficients of variation of PAI-1 were 6.73% and 7.33%, respectively, with a minimum detectable concentration of 0.014-0.142 ng/mL. The intra- and inter-assay coefficients of variation for adiponectin were 3.53% and 6.50%, respectively, with a minimum detectable concentration of 0.079-0.891 ng/mL. The intra- and inter-assay coefficients of variation for leptin were 3.17% and 4.37%, respectively, with a minimum detectable concentration of <7.8 pg/mL.

Subcutaneous fat and abdominal visceral fat areas were quantified using MRI. Data were acquired using a 1.5 T whole body Siemens Sonata system (Siemens Medical Systems, Erlangen, Germany) using the body coil. For the quantification of abdominal fat, one axial slice of the abdomen was obtained at the level of umbilicus. The scan parameters used were TR =209 ms, TE =4.1 ms, slice thickness =10 mm, and flip angle =70°. The field of view was 400 mm with an in-plane image resolution of 2.4 mm × 1.6 mm[40]. Areas of abdominal visceral fat and subcutaneous fat (millimeters squared) were measured on the computer screen using a track-ball[41]. All measurements were analyzed by a single radiologist who was blinded to the metabolic status of the study subjects.

For EAT measurement, electrocardiogram-gated cine images were acquired with a dedicated 4-element, phased array cardiac coil, using a segmented steady-state free precession [fast imaging with steady-state precession (TrueFISP)] sequence (time to echo/time of repetition 1.6/3.2 ms, temporal resolution 35 ms, in-plane spatial resolution 1.4*1.8 mm, slice thickness 5 mm, inter-slice gap 5 mm). The measurements were done in end-diastolic phase on horizontal long-axis plane and basal short-axis plane (at the level of the end of papillary muscles). Fat thickness in right atrioventricular groove, left atrioventricular groove, and anterior interventricular groove were measured on horizontal long-axis plane; fat thickness in superior interventricular groove, inferior interventricular groove and right ventricular free wall were measured on basal short-axis plane, as described in our previous studies and by other investigators[15, 23, 26, 42].

Continuous variables are expressed as mean ± S.D. Differences in continuous variables before and after weight loss in obese MetS subjects were compared by the paired Student’s t test. Univariate correlation analysis between percentage changes of HOMA-IR or ISI-Matsuda after weight loss vs. percentage changes of obesity-associated inflammatory markers, adipokines, and MRI-measured fat index in obese MetS were analyzed using Pearson correlation analysis. Multivariate linear regression analysis was performed to investigate the independent variables correlated with percentage changes of HOMA-IR or ISI-Matsuda after weight loss. The SPSS 17.0 statistical software package (SPSS, Inc., Chicago, IL, USA) was used for all calculations. A two-tailed p value less than 0.05 was considered statistically significant.

After the 3-month weight reduction program, twenty-seven of the obese non-diabetic men with MetS achieved weight loss of more than 5% (97 ± 14 to 87 ± 14 kg, with an average decrease of 10.7%, p < 0.001) (Figure 1). The fasting HOMA-IR, OGTT-derived Matsuda ISI, and OGTT glucose or insulin AUC changed significantly after weight loss (Table 1). Lipid profiles improved and inflammatory markers such as MCP-1 declined after weight loss (Table 1). In terms of circulating adipokines, the obese non-diabetic men had significantly lower values of PAI-1 after weight loss. There were non-significant changes of adiponectin and leptin levels after weight loss (Table 1).

Significant reductions of subcutaneous fat and abdominal visceral fat areas as well as decreases of regional EAT thickness were detected in the 27 obese men who achieved significant weight loss (all p < 0.001) (Table 2).

The percentage improvements in HOMA- IR index after weight loss were significantly correlated with the MRI-measured decrement ratio of superior interventricular groove (SIVG) EAT thickness (r =0.408, p = 0.035), but not with the subcutaneous fat or abdominal visceral fat changes (Table 3). Moreover, the percentage improvement of HOMA-IR index was also significantly correlated with decrement ratio of secretory leptin after weight loss (r = 0.685, p < 0.001) (Table 3).

The percentage improvements of OGTT-derived Matsuda ISI were significantly correlated with MRI-measured decrement ratio of SIVG EAT thickness (r = −0.501, p = 0.008) as well as with the subcutaneous fat (r = −0.642, p < 0.001), but not with abdominal visceral fat changes after weight loss (Table 3). Moreover, the percentage improvement of OGTT-derived Matsuda ISI was also significantly correlated with decrement ratio of secretory leptin after weight loss (r = −0.439, p = 0.022) (Table 3). In a separate analysis, the decrement ratio of SIVG EAT also correlated well with the decrement ratio of OGTT glucose AUC (r =0.520, p = 0.005). The correlation between decrement ratio of SIVG EAT and the decrement ratio of OGTT insulin AUC was fair but not reaching statistical significance (r =0.327, p = 0.095). Moreover, there was no correlation between absolute difference of MR measured left ventricular ejection fraction vs. HOMA-IR (r = −0.035, p = 0.863) or OGTT-derived Matsuda ISI (r =0.067, p = 0.742) changes after weight loss.

In a multivariate linear regression analysis, the decrement ratio of circulating leptin and SIVG EAT were independent variables correlated with improvements of HOMA-IR index after weight loss, while changes in subcutaneous or abdominal visceral fat were not (Table 4).

In a multivariate linear regression analysis, the decrement ratio of SIVG EAT thickness (r = −0.370, p = 0.017) and subcutaneous fat area (r = −0.673, p = 0.006) were independent variables correlated with the increment ratio of OGTT-derived Matsuda ISI after weight loss (Table 5).