Background
Type 2 diabetes mellitus (T2DM) is a chronic and increasing worldwide disorder that is characterized by hyperglycemia, insulin resistance, and dyslipidemia, and its independent association with dysfunctional adiposity, such as excess visceral fat has been reported [
1,
2]. Several studies demonstrated that T2DM also provokes long-term dysfunction and damage to various organs including the eye, kidney, and heart [
3]. In general, the heart requires constant sources of energy that mainly include free fatty acids (FFA) and glucose for continuous pumping, and has high flexibility with regards to the energy substrate metabolism [
4]. However, in the T2DM heart, alterations in myocardial substrate metabolism, characterized by increased myocardial fatty acid metabolism and concurrently decreased glucose metabolism, have been reported [
4‐
6].
Positron emission tomography (PET) can monitor increased uptake of the glucose analogue [
18F]-fluorodeoxyglucose (
18FDG), that is taken up by tissues via glucose transporter proteins [
7]. Recent studies have reported that disturbances of carbohydrate, fat, and protein metabolism altered biodistribution of
18FDG in patients with T2DM [
8]. In addition, regarding myocardial insulin sensitivity, myocardial glucose uptake can differ according to the status of whole body insulin resistance, such as prediabetes and T2DM, compared with normal glucose tolerance (NGT) controls. However, few studies have investigated the relationship(s) of myocardial glucose uptake using PET, adiposity, and other metabolic profiles in subjects with NGT, prediabetes, and T2DM [
8‐
10].
In the current study, we therefore investigated myocardial glucose uptake using [18F]-fluorodeoxyglucose-positron emission tomography (18FDG-PET), and characterized its associations with various clinical and laboratory parameters according to glycemic status.
18FDG-PET and image analysis
Whole body PET-CT was performed using either one of two combined PET-CT scanners: a Biograph TruePoint 40 (Siemens Medical Solutions, Hoffman Estates, IL, USA) or a Discovery 600 (General Electric Medical Systems, Milwaukee, WI, USA). All patients fasted for at least 8 h, and blood glucose levels were recorded before the injection of 18FDG. Approximately 5.5 mBq of 18FDG per kilogram of body weight were administered intravenously. PET-CT scanning was conducted from the skull base to the mid-thigh at 60 min after injection. For the Biograph TruePoint 40 scanner, a spiral CT scan with a 0.5 s rotation time, 35 mA, 120 kVp, and 5 mm section width with arms raised, was used. For the Discovery 600 scanner, a spiral CT scan with a 0.8 s rotation time, 60 mA, 120 kVp, 3.75 mm section thickness, 1.25 mm collimation, and 27.5 mm table feed per rotation with arms raised, was used. PET image acquisition followed CT scanning using the following parameters: 2.5 min per bed position of 21.6 cm in a three-dimensional acquisition mode (Biograph TruePoint 40) or 2 min per bed position of 15.7 cm in a three-dimensional acquisition mode (Discovery 600). Reconstructions of PET images were acquired using a 128 × 128 matrix with ordered subset expectation maximization and attenuation correction.
The standardized uptake value (SUV) was calculated by nuclear medicine experts who were blind to the subjects’ clinical and laboratory data as follows: SUV = (decay-corrected activity [kBq] per mL of tissue volume)/(injected
18FDG activity [kBq]/body mass [g]). Two-dimensional regions of interest (ROIs) were drawn through the transaxial images to measure the SUV
max of the left ventricular myocardium. Patients with striking focal FDG uptake in the left ventricle that could be caused by ischemic change were excluded [
17]. We also obtained the liver SUV, which was quite stable over time, from the circular ROI along the periphery of the right lobe, 1 cm from the margin by averaging at least three times of these values. The values of SUV of the heart to liver FDG uptake ratio (SUV
Heart/SUV
Liver) were used to estimate the myocardial glucose uptake to minimize variability [
18‐
20].
Statistical analysis
All continuous variables were expressed as the mean ± standard deviation, and categorical variables were expressed as proportions. Differences were analysed using the analysis of variance (ANOVA) for continuous variables and the Chi-square test for categorical variables. Comparisons of myocardial glucose uptake relative to the status of diabetes were calculated with the Jonckheere–Terpstra trend test. Pearson’s correlation coefficients were calculated to examine the relationships between myocardial glucose uptake and metabolic variables. Multiple linear regression analysis was performed to determine the independent relationships of the studied variables, and standardized β was represented as the coefficient β. The odds ratios (ORs) and 95 % confidence intervals (CIs) for the factors associated with T2DM were calculated using the multiple logistic regression analysis. In the Pearson’s correlation, multiple linear regression, and multiple logistic regression analysis, values of myocardial glucose uptake (SUVHeart/Liver) were log-transformed to achieve normal distribution. A value of p < 0.05 was considered statistically significant. Statistical analyses were performed using PASW Statistics software, version 20.0 for Windows (SPSS Inc., Chicago, IL, USA).
Discussion
In this study, we firstly investigated fasting myocardial glucose uptake by using 18FDG-PET and visceral/subcutaneous adipose tissue areas by abdominal CT in a total of 346 individuals, who were stratified based on glucose tolerance (NGT, IFG, and T2DM). Our results demonstrated that fasting myocardial glucose uptake was markedly decreased in patients with T2DM compared to the other two groups. Reduced myocardial glucose uptake was related with a greater visceral fat area, higher concentration of circulating FFA, and uric acids, which could be related to systemic insulin resistance. Furthermore, the alteration of myocardial glucose uptake was strongly associated with T2DM, in along with visceral adiposity.
Previously, hyperglycemia, hyperinsulinemia, and disturbances of carbohydrates, fatty acids, and protein metabolism have all been correlated with prediabetes and T2DM [
3,
22,
23]. Therefore, impaired glucose uptake and metabolism in not only the skeletal muscle, but also in the heart, which requires sources of energy mostly from FFA and glucose, could be correlated with insulin resistance for prediabetes and T2DM, contributing to the development of hyperglycemia [
24]. The data of this study support that systemic insulin resistance is strongly related to decreased myocardial glucose uptake [
25,
26]. The estimation of whole body and adipose tissue insulin resistance by HOMA-IR and Adipo-IR all showed significant negative correlations with myocardial glucose uptake in this large number of study population. Although previous studies revealed that myocardial fatty acid metabolism increased with obesity and female sex [
26,
27], relationship between visceral adiposity and myocardial glucose uptake has not been studied yet. To note, this study demonstrates that greater visceral fat area, not subcutaneous is significantly associated with decreased myocardial glucose uptake as well as the presence of T2DM, even after adjustment of other confounding factors including sex and BMI. Visceral adipose tissue has proven to be causally linked to insulin resistance much greater than subcutaneous adipose tissue, by paracrine and endocrine effects from a set of cytokines, particularly high levels of tumor necrosis factor-alpha (TNF-α), low levels of adiponectin, increased macrophage accumulation, and excess of circulating FFA [
28‐
30]. These findings are consistent with the results of the current study, showing that myocardial glucose uptake was significantly decreased, while visceral adiposity was increased with elevated levels of plasma FFA, in patients with T2DM.
In agreement with previous studies, we showed that fasting FFA was an independent predictor for myocardial glucose uptake [
9,
10]. There have been conflicting findings on the relationship between the direct effect of myocardial insulin resistance on myocardial glucose uptake and the independence of increased plasma FFA [
9,
25]. For example, there were reports that serum FFA concentrations suppressed by acipimox, a potent nicotinic acid derivative, affected glucose uptake in the myocardium via inhibition of lipolysis [
9], and FFA levels decreased by rosiglitazone therapy were associated with the improvement in myocardial glucose uptake [
31]. However, Yokoyama et al. showed that the whole body glucose disposal rate (GDR) was independently related to myocardial FDG uptake, whereas FFA was not [
25]. The more prominent relation between GDR and myocardial FDG uptake than between myocardial FDG uptake and FFA in patients with diabetes suggested that insulin resistance regulates the myocardial cellular glucose FFA cycle, the so-called Randle cycle [
32], and/or levels of plasma FFA. In a similar manner, Hicks et al. reported that the correlation between myocardial FDG uptake and GDR was greater than that between myocardial FDG uptake and FFA in diabetic patients [
33]. In addition, the current data which visceral adiposity or uric acid as well as fasting FFA was an independent determinant for myocardial glucose uptake, also suggest that not only direct effect of elevated fasting FFA concentration but also insulin resistance may connect with myocardial metabolism.
Previous studies showed that in the normal heart under fasting conditions, FDG uptake showed variable myocardial glucose uptake because FFA is a primary source of energy, whereas glucose utilization is relatively low for the myocardial oxidative metabolism compared to glucose-loading conditions [
34,
35]. However, in the T2DM heart, regulation of glucose metabolism differed from the normal heart, therefore prior studies showing that myocardial FDG uptake was significantly decreased in diabetic patients compared to normal subjects are consistent with our results [
8,
34]. The underlying mechanism behind association with systemic insulin resistance and myocardial glucose metabolism has been still investigated. Cardiac myocytes utilize glucose mostly via insulin-sensitive glucose transporters (GLUT4) that are responsible for more than 50 % of all glucose uptakes in the body [
36], and reduced expression and mutations of GLUT4 have been associated with diabetes [
37,
38]. Recent study showed that increased insulin receptor substrate 1 (IRS1)-phosphoinositide 3-kinase (PI3K) activity with a concurrent activation of the insulin receptor was occurred with a diminished translocation of GLUT4 to the sarcolemmal membrane in the heart even in fasting status of diabetes. Also, the increase in expression of GLUT4 trafficking and docking components turned out to be dysfunction of GLUT4 vesicles in diabetic heart [
39]. Therefore, whole body insulin resistance maybe connected with myocardial insulin resistance, in the condition of down-regulated sarcolemmal GLUT4, thus resulting in decreased fasting myocardial glucose uptake in this current study.
In this study, patients with T2DM had relatively low levels of HbA1c (6.8 %) and a low proportion of antidiabetic drug users (41.0 %), and most of them were newly diagnosed or well controlled diabetic patients. However, we found that myocardial glucose uptake showed a marked gradual decrease in patients with insulin resistant prediabetes and even well controlled T2DM. The relationship between hyperglycemia and development of ischemic heart disease has been well known [
40,
41], but the effects of diabetes on myocardial metabolism still remain uncertain [
42,
43]. Several studies have reported that a chronic shift of myocardial substrate preference in the diabetic heart resulted in a prominent decrease in glucose and lactic acid oxidation, and an increase in fatty acid oxidation [
4,
5]. The effects of diabetes on myocardial metabolism are very complex, including systemic metabolic disturbances of hyperglycemia, increased FFA, down-regulation of glucose transporters, increased insufficient energy utilization of fatty acid oxidation, lipid accumulation, and lipid toxicity in cardiomyocytes [
44,
45]. Therefore, the current study results could emphasize the association of diabetes and myocardial metabolism in connection with insulin resistance, and suggest that beneficial effects of an adequate glycemic control on myocardial metabolic disturbances in diabetes. These metabolic disturbances may lead to diabetic cardiomyopathy, however, further studies of the relationship between myocardial glucose uptake and cardiac function will be needed to determine the exact mechanisms [
46‐
48].
This study had several distinguishable strengths. To our knowledge, it was the first study to investigate myocardial glucose uptake using
18FDG-PET in the largest study population together with the subcutaneous and visceral adiposity by abdominal CT, a gold standard method for quantification, and determinations of other metabolic parameters like HOMA-IR and Adipose-IR, which evaluated whole body insulin sensitivity, according to the glycemic status. Second, because the induction of myocardial ischemia could affect myocardial metabolism, we excluded patients not only having history of known coronary artery disease and heart failure but also having CACS over 400, who could have subclinical atherosclerosis [
11]. Also, striking focal FDG uptake in the left ventricle was investigated to exclude ischemic change on PET image analysis [
17]. Therefore, as the incidence of coronary artery disease in patients with diabetes is higher compared to normal healthy people, asymptomatic subjects with coronary artery disease could be excluded in this current study [
49]. Finally, on PET image analysis, myocardial glucose uptake was estimated as ratio of maximum value of SUV in the myocardium and mean value of SUV in the liver. In FDG-PET for malignancy, the liver has been used as an internal standard for grading FDG uptake of whole body lesions because SUV in liver but not in other tissues stays stable over time even in patients with diffuse fatty liver disease when measured as a mean in the right lobe of the liver [
19,
20,
50]. Also, the mean SUV of liver in the early images after injection (50–70 min) as same as the present study shows no dependency on blood glucose level [
51]. In this current study, we obtained three mean SUVs of the right lobe of the liver, and used the values as comparators for increased FDG uptake in the heart. Therefore, the value of myocardial glucose uptake as ratio of maximum value of SUV in the myocardium and mean value of SUV in the liver is relatively consistent and reliable in this study.
The current study also had several limitations. The cross-sectional study design was insufficient to determine a causal relationship in the development of impaired myocardial glucose uptake. In addition, we did not assess cardiac function in our participants; therefore, possible correlations between myocardial glucose uptake and cardiac function would be an important topic for future studies.