Reduction in Global Myocardial Glucose Metabolism in Subjects With 1-Hour Postload Hyperglycemia and Impaired Glucose Tolerance

Prediabetes is considered to be an intermediate metabolic condition between normal homeostasis and type 2 diabetes. This condition occurs in individuals with impaired glucose tolerance (IGT), impaired fasting glucose (IFG), and high-risk glycated hemoglobin A_1c (HbA_1c) (1). The International Diabetes Federation estimates that, worldwide, 374 million individuals aged 18–99 years have IGT. This number is projected to increase to 587 million by 2045 (2). Prediabetes not only is associated with an increased risk of progression to type 2 diabetes; it also confers an increased risk for cardiovascular morbidity and mortality (3,4). Several studies showed that a plasma glucose concentration ≥155 mg/dL at hour 1 during an oral glucose tolerance test (OGTT) can identify individuals with normal glucose tolerance (NGT) who are at increased risk for type 2 diabetes (5–11). Notably, 1-h postload hyperglycemia has been associated with cardiovascular organ damage, cardiovascular disease, and all-cause mortality (11–18). From a pathophysiological standpoint, both IGT (19) and NGT with high plasma glucose (≥155 mg/dL) at 1 h (NGT 1-h high) are characterized by impaired peripheral insulin sensitivity, a well-known metabolic defect preceding the development of type 2 diabetes (10,20). Because impaired insulin-stimulated myocardial glucose uptake has been found in patients with type 2 diabetes with or without coronary artery disease (21), it is conceivable that both individuals with IGT and those with NGT 1-h high may exhibit myocardial insulin resistance. Myocardial positron emission tomography (PET) with ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG), a widely used glucose analog, in combination with a euglycemic-hyperinsulinemic clamp is considered to be the gold standard for measuring the myocardial metabolic rate of glucose (MRGlu) under standardized experimental conditions (21–23). Few studies have assessed myocardial MRGlu in individuals with prediabetes (24,25). Hasegawa et al. (24) reported that uptake of ¹⁸F-FDG after glucose loading was not impaired in individuals with IGT and coronary artery disease. Accordingly, Kim et al. (25) found that myocardial ¹⁸F-FDG uptake assessed under fasting state was not decreased in individuals with IFG. Thus, in individuals with IGT or NGT 1-h high characterized by systemic insulin resistance the presence of cardiac insulin resistance remains uncertain. To address this issue, insulin-stimulated myocardial glucose metabolism was measured through the use of cardiac dynamic PET combined with a euglycemic-hyperinsulinemic clamp in individuals with IGT and those with NGT 1-h high; none had coronary artery disease.

We examined 30 subjects participating in the Catanzaro Metabolic Risk Factors (CATAMERI) study, an observational study comprising adult Caucasian outpatients participating in a campaign to assess cardiometabolic risk factors and recruited at a referral university hospital of the University “Magna Graecia” of Catanzaro (9,18). Recruitment mechanisms included word of mouth, fliers, and newspaper advertisements. All subjects were consecutively recruited according to two inclusion criteria: age between 30 and 60 years, and the presence of one or more cardiometabolic risk factors (family history of diabetes, dysglycemia, hypertension, dyslipidemia, and overweight/obesity). Exclusion criteria included previous diagnosis of type 1 or type 2 diabetes; previous cardiovascular disease (determined from the medical history and an electrocardiogram recorded at rest); uncontrolled hypertension; history of malignant or autoimmune diseases; heart or renal failure; chronic liver disease; acute infection; gastrointestinal disease associated with malabsorption; history of alcohol or drug abuse; and treatment with antidiabetes drugs, medicaments known to affect glucose and lipid metabolism (including corticosteroids, estroprogestins, and statins), or medicaments affecting heart function (including ACE inhibitors, angiotensin-receptor blockers, angiotensin receptor–neprilysin inhibitors, β-blockers, and diuretics). All subjects underwent anthropometric evaluation including measurements of BMI and waist circumference; body composition was determined by using bioelectrical impedance. Readings of blood pressure were obtained in the clinic with a sphygmomanometer; measurements were taken from the left arm as the patient lay supine after 5 min of quiet rest. Blood pressure values were calculated as the mean of the last two of three consecutive measurements obtained at 3-min intervals. After a 12-h fast, a 75-g OGTT was performed; samples were obtained at 0, 30, 60, 90, and 120 min for use in the plasma glucose and insulin assays. Glucose tolerance was defined on the basis of OGTT results, according to the American Diabetes Association criteria (1).

The study was approved by the ethics committee of the University Magna Graecia of Catanzaro, and informed consent was obtained from each subject in accordance with principles of the Declaration of Helsinki.

Global myocardial MRGlu, expressed as millimoles per minute per 100 mg, was measured from an ¹⁸F-FDG PET scan acquired during the course of a euglycemic-hyperinsulinemic clamp, as previously described (9,20). Subjects received a priming dose of insulin (100 UI/mL) (Humulin R; Eli Lilly) during the initial 10 min of euglycemic hyperinsulinemic clamp to acutely increase plasma insulin to the desired level, followed by continuous insulin infusion fixed at 40 mU/m² per minute. Blood glucose level was maintained at 90 mg/dL for the next 120 min by infusing 20% glucose at varying rates according to blood glucose concentration, which was measured at 5-min intervals (the mean coefficient of variation of blood glucose was <4%). Glucose metabolized by the whole body (M) was calculated as the mean rate of glucose infusion measured during the last 60 min of the clamp (steady state), expressed as milligrams per minute per kilogram of fat-free mass (M_FFM).

The ¹⁸F-FDG PET procedure was performed with a hybrid PET/computed tomography (CT) scanner (Discovery ST8–2D PET scanner; GE), starting 60 min after the insulin infusion. Patients were positioned supine under the camera, and the correct collimation to the chest was checked by using a scout CT image. Dynamic acquisition was started simultaneously with intravenous injection of 370 MBq ¹⁸F-FDG; this acquisition period lasted 60 min. A continuous frame-mode acquisition was performed as follows: 8 × 15 s, 2 × 30 s, 2 × 120 s, 1 × 180 s, 6 × 300 s, 1 × 600 s (26). After a 10-min break, a static image was acquired over 1,200 s (i.e., between 70 and 90 min after injection). PET images were corrected for decay and attenuation, then reconstructed in a 128 × 128 matrix by using an OSEM algorithm, and then CT-based attenuation correction was performed. The insulin-glucose infusion continued during the entire PET acquisition period. Myocardial MRGlu was estimated by using Patlak compartmental modeling (27), a widely diffuse technique performed by a graphical tool specific to cardiac image analysis (PCARD) in PMOD software (version 3.806) (28). PCARD measures global MRGlu and segmental myocardial glucose uptake by using a segmentation algorithm to divide the myocardium into 17 standard segments according to American Society of Nuclear Cardiology guidelines and the American Heart Association (29). Segments are then grouped into three vascular territories corresponding to the left anterior descending (LAD) artery, the left circumflex (LCX) artery, and the right coronary artery (RCA).

Comprehensive echocardiography was performed by a single experienced examiner who was blinded to the patients’ clinical and laboratory data. Patients were placed in a partial left decubitus position, and tracings were taken with a VIVID-7 Pro ultrasound machine (GE Technologies, Milwaukee, WI) with an annular phased-array 2.5-MHz transducer. Only frames showing optimal visualization of cardiac structures were considered for reading. The left ventricular (LV) internal diameter was measured at end-diastole and end-systole, as recommended by the American Society of Echocardiography (30). Standard methods were used to calculate ejection fraction as a measure of LV systolic chamber function (13). LV mass was calculated by using the Devereux formula and was normalized by body surface area (LV mass index) (31). Left atrial volume was determined by using the apical four-chamber and two-chamber views. A pulsed Doppler transmitral flow velocity profile was determined from the apical four-chamber view, and the sample volume was positioned at the tip of the mitral valve leaflets. The following parameters were evaluated in order to determine diastolic function: peak transvalvular flow velocity in early diastole (E wave), peak transvalvular flow velocity in late diastole (A wave), and E/A ratio (14).

Plasma glucose, total and HDL cholesterol, and triglycerides were measured using enzymatic methods (Roche Diagnostics, Mannheim, Germany). HbA_1c was measured with high performance liquid chromatography using an NGSP-certified automated analyzer (Adams HA-8160 HbA_1c analyzer, Menarini, Italy). Plasma insulin concentration was measured with a chemiluminescence-based assay (Immulite, Siemens, Italy).

Variables with skewed distribution, such as triglycerides and M_FFM, were natural log transformed for statistical analyses. Continuous variables are expressed as the mean ± SD. Categorical variables were compared by using the χ² test. ANOVA was used to compare differences in continuous variables among groups, with post hoc least significant difference corrections. Differences in metabolic variables and myocardial glucose uptake among groups were tested by using a general linear model after adjusting for age, sex, and BMI. Relationships between variables were determined on the basis of the Pearson correlation coefficient (r). Multivariable linear regression analysis was applied in order to determine the independent association between global myocardial MRGlu and the following parameters: glucose tolerance, age, sex, BMI, systolic and diastolic blood pressure, and diastolic function (measured as the E/A ratio).

Previous studies have reported a 22–41% reduction in myocardial glucose uptake in subjects with type 2 diabetes or prediabetes (21,25,32). In light of this, using an online power calculator (https://clincalc.com/Stats/SampleSize.aspx), we calculated that 10 subjects in each group had 90% power to detect a 36% difference in myocardial glucose uptake, with a level of significance of 5%.

For all analyses, a P value ≤0.05 was considered to be statistically significant. All analyses were performed by using SPSS software version 16.0 for Mac.

Of the 30 individuals without diabetes examined, 10 had IGT. We divided the 20 individuals with NGT into two subgroups: 10 with 1-h postload glucose <155 mg/dL (NGT 1-h low), and 10 with 1-h postload glucose ≥155 mg/dL (NGT 1-h high). Anthropometric and biochemical characteristics of the study participants are shown in Table 1. Significant between-group differences were found with respect to age (subjects in the IGT and NGT 1-h high groups were older than those in the NGT 1-h low group). As expected, those with NGT 1-h high and those with IGT had significantly higher 1-h plasma glucose and HbA_1c than did those with NGT 1-h low, whereas individuals with IGT exhibited higher 2-h postload plasma glucose than did subjects in the NGT groups. No significant differences were found between groups with respect to fasting glucose and insulin levels, total cholesterol, HDL and LDL cholesterol, or triglycerides. Both the NGT 1-h high and the IGT groups showed a significant reduction in insulin-stimulated glucose disposal (M_FFM) (P = 0.02) (Table 1). Plasma glucose concentration during the last hour of the clamp and steady-state plasma insulin levels during the clamp did not differ among the three groups (Table 1).

Echocardiographic characteristics of the study groups are reported in Supplementary Table 1. None of the subjects in the NGT 1-h high and the IGT groups showed significant differences in LV mass, LV mass index, or LV end-systolic and LV end-diastolic diameters compared with NGT 1-h low individuals. In addition, LV systolic function, as reflected by LV ejection fraction, was similar among the three groups, whereas echocardiographic parameters related to diastolic function (measured on the basis of the E/A ratio) tended to be lower in both the NGT 1-h high and IGT groups than in the NGT 1-h low group, although the differences did not reach the threshold for significance.

After adjusting for age, sex, and BMI, both subjects with NGT 1-h high (23.7 ± 6.4 mmol/min/100 mg; P = 0.024) and those with IGT (16.4 ± 6 mmol/min/100 mg; P < 0.0001) exhibited a significant reduction in global myocardial MRGlu compared with subjects with NGT 1-h low (32.8 ± 9.7 mmol/min/100 mg) (Table 1 and Fig. 1A). Representative PET images related to global myocardial MRGlu in the three study groups are shown in Fig. 2. These differences in global myocardial MRGlu remained significant after adjusting for additional confounders including fasting and 2-h postload glucose, blood pressure, total cholesterol, HDL and LDL cholesterol, and triglycerides (P = 0.005, NGT 1-h low vs. NGT 1-h high; P = 0.024, NGT 1-h low vs. IGT).

Because of the numerically, but not statistically significant, between-group differences in steady-state plasma insulin levels achieved during the insulin clamp, we reanalyzed differences in global myocardial MRGlu after matching the three groups for steady-state plasma insulin levels during the clamp (57 ± 11 mU/mL for NGT 1-h low [n = 6], 55 ± 12 mU/mL for NGT 1-h high [n = 6], and 55 ± 9 mU/mL for IGT [n = 6]). Both subjects with NGT 1-h high (21.4 ± 5.8 mmol/min/100 mg; P = 0.02), and those with IGT (17.8 ± 6.1 mmol/min/100 mg; P = 0.003) exhibited a significant reduction in global myocardial MRGlu; this value was 29.7 ± 3.5 mmol/min/100 mg for those with NGT 1-h low.

Additionally, differences in global myocardial MRGlu between the three groups remained significant after adjusting for morphofunctional parameters of the left chambers, including LV mass index, LV ejection fraction, and E/A ratio (P = 0.02, NGT 1-h low vs. NGT 1-h high; P < 0.0001, NGT 1-h low vs. IGT). Global myocardial MRGlu was weakly correlated with E/A ratio in an unadjusted univariate correlation analysis (r = 0.345; P = 0.09). In order to evaluate whether glucose tolerance was an independent contributor to global myocardial MRGlu, a multivariable regression analysis was performed, running a model including age, sex, BMI, blood pressure, and diastolic function (measured on the basis of E/A ratio). The association between glucose tolerance and global myocardial MRGlu remained significant (β = −0.694; P = 0.001), and the E/A ratio was not independently associated with global myocardial MRGlu (β = −0.193; P = 0.440).

Univariate analyses showed that MRGlu was positively correlated with insulin-stimulated glucose disposal (M_FFM) (r = 0.441; P = 0.019) (Fig. 3A) and negatively correlated with 1-h postload (r = −0.422; P = 0.025) (Fig. 3B) and 2-h postload (r = −0.374; P = 0.05) (Fig. 3C) glucose levels. By contrast, no significant correlation was detected between MRGlu and fasting plasma glucose (r = −0.046; P = 0.818).

As compared with NGT 1-h low subjects, both those in the NGT 1-h high and IGT groups showed a significant reduction in segmental myocardial MRGlu corresponding to the vascular territories sprayed by the LAD artery (mean values 32.1 ± 9.1 vs. 21.4 ± 7.4 and 15.2 ± 5.6; P = 0.018 and P < 0.0001, respectively) and the RCA (34.1 ± 12 vs. 23.6 ± 6 and 15.9 ± 5.6; P = 0.036 and P < 0.0001, respectively) (Fig. 1B and C). Subjects with IGT showed a significant reduction in segmental MRGlu corresponding to the LAD artery (P = 0.035) and the RCA (P = 0.014) (Fig. 1D). Furthermore, subjects with IGT showed a significant reduction in MRGlu in the vascular territory of the LCX artery (18.3 ± 7.3 vs. 27.2 ± 8.3 for NGT 1-h high and 32.4 ± 9.5 for NGT 1-h low; P = 0.024, P = 0.001, respectively) (Fig. 1D).

The salient finding of this study is that both individuals with IGT and those with NGT 1-h high exhibit insulin resistance throughout whole body and in myocardial muscle. Differences in global myocardial MRGlu between groups remained significant after adjusting for LDL and HDL cholesterol and triglycerides, thus arguing against the possibility that small, non–statistically significant differences in lipid profile between groups might be responsible for the observed differences in global myocardial MRGlu. Interestingly, cardiac insulin resistance was determined in noninfarcted, normally contractile myocardium, and impaired myocardial glucose metabolism was found in the vascular territories sprayed by the main coronary arteries. These findings are consistent with studies of patients with type 2 diabetes with or without coronary artery disease (21,30–32). By contrast, two previous studies of individuals with IGT and coronary artery disease or with IFG found comparable rates of myocardial glucose uptake in patients with these prediabetic conditions (24,25). Possible explanations for this disparity include differences in the methods used to assess myocardial glucose uptake (fasting state vs. oral glucose loading vs. euglycemic-hyperinsulinemic clamp) and the population analyzed (IGT with coronary artery disease vs. IFG with no history of myocardial infarction or heart failure vs. IGT or NGT 1-h high [both without coronary artery disease]).

Interestingly, the impairment of insulin-mediated myocardial glucose metabolism that we found in individuals who were at risk for type 2 diabetes (i.e., those with NGT 1-h high and those with IGT) was of the same of magnitude as that observed in patients with type 2 diabetes, ranging between 22 and 41% (21,32–34). These data suggest that myocardial insulin resistance may be an early defect in the pathogenesis of type 2 diabetes, occurring before steady elevation of plasma glucose levels and already detectable in dysglycemic conditions associated with an increased risk of type 2 diabetes. On the other hand, the finding that both 1-h and 2-h postload plasma glucose levels, but not fasting glycemia, were inversely associated with myocardial glucose metabolism points to a causal role for postprandial hyperglycemia. In this respect, 1-h postload hyperglycemia has been considered as an indicator of augmented glycemic variability (35), and a number of studies have suggested that excessive postprandial glucose excursions are associated with adverse cardiovascular outcomes (36). The mechanism by which elevated 1-h and 2-h postload plasma glucose levels negatively affect myocardial glucose uptake remains to be established. Elevation of glucose levels may have a causative role in cardiac and skeletal muscle insulin resistance by affecting insulin signaling and action (33,37). Accordingly, reduced insulin-stimulated glucose transport and impaired insulin signaling have been found in hearts of Goto-Kakizaki rats, a polygenic model of spontaneous type 2 diabetes (38). Also, chronic subclinical inflammation found in both individuals with IGT and those with NGT 1-h high (39) could be a mechanistic factor, as it affects insulin sensitivity in vivo and in vitro (40). Clearly, further studies are needed to elucidate the mechanism by which elevated postload plasma glucose levels negatively affect myocardial glucose uptake.

The results presented here are clinically relevant because evidence has shown that myocardial insulin resistance may antedate—and possibly cause—the development of coronary artery disease (21).

This study has several strengths. To the best of our knowledge, it is the first to assess insulin-stimulated myocardial and whole-body glucose metabolism using the gold standard cardiac PET in combination with a euglycemic-hyperinsulinemic clamp in individuals with IGT and in those with NGT 1-h high. Moreover, we analyzed not only global MRGlu but also the segmental myocardial glucose metabolism corresponding to the vascular territories of the main coronary arteries. Additionally, all tests, including anthropometric measures, the OGTT, and ¹⁸F-FGD PET, combined with the euglycemic-hyperinsulinemic clamp, were performed at the same time by skilled staff after standardized training.

This study also has some limitations. Glucose tolerance was only measured once, through an OGTT. According to current guidelines (1), one positive test is not sufficient to firmly establish the presence or absence of glucose tolerance, which should be defined on the basis of at least two separate measurements; therefore, some participants might have been misclassified. Furthermore, the results are based only measurement from Caucasian subjects who were recruited in a homogenous geographical region of southern Italy. Although this could also be considered a strength of the study, it remains to be addressed whether our findings can be extended to other ethnic groups. Moreover, the steady-state plasma insulin levels achieved during the insulin clamp are numerically, but not statistically, different between groups, and this may have contributed, at least in part, to the higher cardiac glucose uptake measured. Additionally, the cross-sectional design of the study, which lacks prospective data, impedes any assumptions about the casual relationships between impaired myocardial glucose uptake and cardiovascular outcomes.

In conclusion, we show that both individuals with IGT and those with NGT 1-h high, all without coronary artery disease, exhibit reductions in myocardial glucose uptake as assessed by dynamic ¹⁸F-FDG PET combined with a euglycemic-hyperinsulinemic clamp. Impaired insulin-stimulated myocardial glucose metabolism was paralleled by a decrease in whole-body glucose disposal. Myocardial glucose metabolism was inversely associated with elevated 1-h and 2-h postload plasma glucose levels. Further studies are required to elucidate the mechanism by which postload hyperglycemia affects myocardial glucose metabolism, and to determine the clinical impact of impaired insulin-stimulated myocardial MRGlu on the development of abnormal cardiac function.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. E.S. researched and analyzed data, and wrote and edited the manuscript. E.P. researched and analyzed data. F.A., A.P., P.Vi., and T.V.F. researched data and reviewed the manuscript. F.P., P.Ve., and G.L.C. contributed to the discussion and reviewed the manuscript. G.S. conceived and designed the study and wrote, reviewed, and edited the manuscript. All authors read and approved the final manuscript. G.S. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.