Increased mediastinal fat and impaired left ventricular energy metabolism in young men with newly found fatty liver

Authors

  • Gianluca Perseghin,

    Corresponding author
    1. Department of Internal Medicine, Section of Nutrition/Metabolism, San Raphael Scientific Institute, Milan, Italy
    2. Unit of Clinical Spectroscopy, San Raphael Scientific Institute, Milan, Italy
    3. Faculty of Exercise Sciences, Università degli Studi di Milano, Milan, Italy
    4. Center “Physical exercise for health and wellness of the Faculty of Exercise Sciences,” Università degli Studi di Milano, Milan, Italy
    • Faculty of Exercise Sciences, Università degli Studi di Milano and Internal Medicine/Unit of Clinical Spectroscopy Istituto Scientifico San Raffaele, via Olgettina 60, 20132, Milano, Italy
    Search for more papers by this author
    • fax: (39) 02 2643-2771

  • Guido Lattuada,

    1. Department of Internal Medicine, Section of Nutrition/Metabolism, San Raphael Scientific Institute, Milan, Italy
    Search for more papers by this author
  • Francesco De Cobelli,

    1. Unit of Clinical Spectroscopy, San Raphael Scientific Institute, Milan, Italy
    2. Division of Diagnostic Radiology, San Raphael Scientific Institute, Milan, Italy
    Search for more papers by this author
  • Antonio Esposito,

    1. Division of Diagnostic Radiology, San Raphael Scientific Institute, Milan, Italy
    Search for more papers by this author
  • Elena Belloni,

    1. Division of Diagnostic Radiology, San Raphael Scientific Institute, Milan, Italy
    Search for more papers by this author
  • Georgia Ntali,

    1. Department of Internal Medicine, Section of Nutrition/Metabolism, San Raphael Scientific Institute, Milan, Italy
    Search for more papers by this author
    • She is currently a clinical fellow in the Department of Endocrinology, Polikliniki Hospital, Athens.

  • Francesca Ragogna,

    1. Department of Internal Medicine, Section of Nutrition/Metabolism, San Raphael Scientific Institute, Milan, Italy
    Search for more papers by this author
  • Tamara Canu,

    1. Division of Diagnostic Radiology, San Raphael Scientific Institute, Milan, Italy
    Search for more papers by this author
  • Paola Scifo,

    1. Unit of Clinical Spectroscopy, San Raphael Scientific Institute, Milan, Italy
    2. Division of Nuclear Medicine, Raphael Scientific Institute, Milan, Italy
    Search for more papers by this author
  • Alessandro Del Maschio,

    1. Unit of Clinical Spectroscopy, San Raphael Scientific Institute, Milan, Italy
    2. Division of Diagnostic Radiology, San Raphael Scientific Institute, Milan, Italy
    3. Università Vita e Salute San Raffaele, Milan, Italy
    Search for more papers by this author
  • Livio Luzi

    1. Department of Internal Medicine, Section of Nutrition/Metabolism, San Raphael Scientific Institute, Milan, Italy
    2. Unit of Clinical Spectroscopy, San Raphael Scientific Institute, Milan, Italy
    3. Faculty of Exercise Sciences, Università degli Studi di Milano, Milan, Italy
    4. Center “Physical exercise for health and wellness of the Faculty of Exercise Sciences,” Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

Abstract

Fatty liver is characterized by metabolic abnormalities at the liver, but also at skeletal muscle and adipose tissue sites. It is hypothesized that the heart may be suffering metabolic alterations, and this study was undertaken to ascertain whether individuals with fatty liver have left ventricular (LV) alterations of energy metabolism, structure, and function and abnormal amounts of epicardial fat as a specific marker of visceral fat accumulation. To this end we studied young, nondiabetic men matched for anthropometric features with (n = 21) or without (n = 21) fatty liver by means of (1) cardiac magnetic resonance imaging (MRI); (2) cardiac 31P-MR spectroscopy (MRS); and (3) hepatic 1H-MRS to assess quantitatively the intrahepatic fat (IHF) content. Insulin sensitivity was determined by the updated HOMA-2 computer model. Individuals with fatty liver showed reduced insulin sensitivity, increased serum free fatty acid (FFA), and E-selectin, abnormal adipokine concentrations, and higher blood pressure. LV morphology and systolic and diastolic functions were not different; however, in the scanned intrathoracic region, the intrapericardial (7.8 ± 3.1 versus 5.9 ± 2.5 cm2; P < 0.05) and extrapericardial (11.7 ± 6.1 versus 7.8 ± 3.2 cm2; P < 0.03) fat was increased in men with fatty liver compared with those without fatty liver. The phosphocreatine (PCr)/adenosine triphosphate (ATP) ratio, a recognized in vivo marker of myocardial energy metabolism, was reduced in men with fatty liver in comparison with normals (1.85 ± 0.35 versus 2.11 ± 0.31; P < 0.016). In conclusion, in newly found individuals with fatty liver, fat was accumulated in the epicardial area and despite normal LV morphological features and systolic and diastolic functions, they had abnormal LV energy metabolism. (HEPATOLOGY 2008.)

Nonalcoholic fatty liver (NAFL) is a feature of the insulin resistance syndrome1, 2 and a typical aspect of body composition related to visceral adiposity.3 Multiple metabolic abnormalities in organs and tissues under the influence of insulin were reported: impaired insulin-mediated inhibition of hepatic glucose production was reported at the liver site,1, 4 impaired insulin-stimulated glucose metabolism was shown at the level of the skeletal muscle,1, 5, 6 and impaired insulin-dependent control on lipolysis was detected at the level of the adipose tissue.1, 4 The heart is another organ whose metabolism may be influenced by insulin resistance. Few insights are available about cardiac metabolism in patients with NAFL because of the difficulties of studying noninvasively in vivo cardiac metabolism in humans. Recently, Lautamaki et al.,7 using positron emission tomography and combining 2-deoxy-2-[18F]fluoro-D-glucose and 15O-labeled water, quantified simultaneously cardiac insulin sensitivity and blood flow in patients with type 2 diabetes and coronary artery disease and found that myocardial insulin resistance was more severe in patients with NAFL than in those without it.7 Magnetic resonance spectroscopy (MRS) techniques have been applied to study in vivo metabolic pathways in different organs and tissues8 including the heart.9 In particular, the assessment of the intrahepatic fat (IHF) content as a continuous variable and the assessment of validated markers of cardiac energy metabolism may be performed noninvasively by means of 1H-MRS10, 11 and 31P-MRS,12 respectively. This study was undertaken to test whether abnormalities of (1) left ventricular (LV) morphology, (2) intrapericardial and extrapericardial mediastinal fat content, (3) LV systolic and diastolic functions, and (4) resting LV energy metabolism may be detected in young, nondiabetic men with a newly discovered finding of excessive IHF content when compared with individuals carefully matched for anthropometric features and with a normal IHF content.

Abbreviations

BMI, body mass index; CV, coefficient of variation; ELISA, enzyme-linked immunosorbent assay; FFA, free fatty acid; HOMA2-%S, surrogate index of insulin sensitivity; IHF, intrahepatic fat; LV, left ventricle; MRI, magnetic resonance imaging; MRS, magnetic resonance spectroscopy; NAFL, nonalcoholic fatty liver; PCr/ATP ratio, phosphocreatine/adenosine triphosphate ratio.

Patients and Methods

Subjects

We recruited 42 men within the Istituto Scientifico H San Raffaele. Twenty-one were selected because they were known to have increased IHF content based on their previous participation in a 1H-MRS–based survey to assess the prevalence of fatty liver in our Institute13; the other 21 volunteers served as control group and were selected because they were tightly matched with the fatty liver group for the anthropometric features [age and body mass index (BMI)] and lifestyle habits, and they had normal IHF content, having participated in the same survey. All study subjects had no history of hypertension or endocrine/metabolic disease. Habitual physical activity was assessed by using a questionnaire.14 Body weight had been stable for at least 6 months. We assessed the metabolic syndrome using the Adult Treatment Panel III (ATP-III) definition with the exception of the waist criteria; rather, a BMI greater than 30 kg/m2 was used. History of hepatic disease [including hepatitis B virus or hepatitis C virus infection, genetic liver disease such as hemochromatosis, alpha-1-anti-trypsin deficit, Wilson disease], substance abuse, or daily consumption of more than 1 alcoholic drink daily (<20 g/day) or the equivalent in beer and wine was an exclusion criteria. Five individuals with increased IHF content and 3 normal individuals were smokers. Additional main criteria for the exclusion from the study were (1) history of coronary, cerebral, or peripheral vascular events; (2) history of dilated cardiomyopathy; (3) previous knowledge of a pathological ejection fraction; (4) previous knowledge of resting electrocardiogram markers of cardiac ischemia; and (5) features compatible with the New York Heart Association classes for heart failure. The anthropometric characteristics of the subjects are summarized in Table 1. None was taking medications, and all were in good health as assessed by medical history and physical examination. Informed consent was obtained from all subjects after explanation of purposes, nature, and potential risks of the study. The protocol was approved by the Ethical Committee of the Istituto Scientifico H San Raffaele.

Table 1. Anthropometric and Laboratory Characteristics of Men With (IHF Content > 5% ww) or Without (IHF Content < 5%ww) Excessive Hepatic Fat Accumulation
 IHF Content < 5% Wet WeightIHF Content > 5% Wet WeightP Value
  1. Values are given as mean ± SD.

  2. Independent-samples t test (2-tailed) or Kruskal-Wallis nonparametric test.

  3. Abbreviation: PAI, physical activity index.

N2121 
IHF content (% wet weight)2.20 ± 0.94 range: 0.69–4.6514.59 ± 8.84 range: 5.40–39.140.0001
Age (years)36 ± 735 ± 70.39
Height (cm)176 ± 6177 ± 50.43
Weight (kg)84 ± 786 ± 130.82
BMI (kg/m2)27.1 ± 1.327.5 ± 3.60.42
Systolic BP (mm Hg)122 ± 9128 ± 100.06
Diastolic BP (mm Hg)80 ± 984 ± 80.26
Fasting glucose (mmol/L)4.86 ± 0.565.16 ± 0.580.052
Total cholesterol (mmol/L)4.75 ± 0.775.17 ± 1.410.40
HDL-cholesterol (mmol/L)1.45 ± 0.351.16 ± 0.290.008
Triglycerides (mmol/L)1.00 ± 0.491.85 ± 1.200.028
LDL-cholesterol (mmol/L)2.84 ± 0.623.28 ± 1.300.32
Creatinine (μmol/L)82 ± 982 ± 90.94
TSH (mU/L)1.25 ± 1.221.23 ± 0.990.75
AST (U/L)30 ± 1538 ± 130.35
ALT (U/L)27 ± 954 ± 280.01
PAI8.18 ± 1.217.85 ± 1.510.40

Experimental Protocol

Subjects were instructed to consume an iso-caloric diet and to abstain from exercise activity for 3 days before the magnetic resonance imaging (MRI)–MRS studies. All volunteers underwent the MRI–MRS protocol at 7:30 AM to 9:30 AM in the resting state after a 10-hour overnight fasting period and after the collection of venous blood for the assessment of postabsorptive plasma glucose, total cholesterol, high-density lipoprotein cholesterol, triglycerides, nonesterified fatty acids, insulin, leptin, adiponectin, resistin, E-selectin, hs-c-reactive protein, thyroid-stimulating hormone, and creatinine.

Hepatic 1H-MRS.

1H-MRS was performed at rest and with patients in the supine position with the use of a 1.5-T whole-body scanner (Gyroscan Intera Master 1.5 MR System; Philips Medical Systems, Best, the Netherlands) using a conventional circular superficial coil (C1-coil) as previously described in detail.11, 13 Briefly, T1 in-phase and out-of-phase sequences were obtained to look for a potential loss of signal on out-of-phase images, indicating the presence of IHF accumulation, then 2 1H spectra were collected in the same pre-scanning conditions using a PRESS pulse sequence (inter-pulse delay TR = 3000 msec, spin echo time TE = 25 msec, 1024 data points over a 1000-Hz spectral width and 64 acquisitions) with and without suppression of the water signal.

Cardiac 31P-MRS.

Cardiac 31P-MRS was performed at rest using the same system.1231P spectra were obtained by means of a 10-cm-diameter surface coil used for transmission and detection of radiofrequency signals as previously described.12 Briefly, electrocardiogram-triggered MRI was performed to acquire scout images, to establish the exact position of the 31P surface coil, and eventually to reposition the coil center. Localized homogeneity adjustment was performed by optimizing the 1H-MRS water signal. Manual tuning and matching of the 31P surface coil was performed to adjust for different coil loading. The radiofrequency level was adjusted to obtain a 180° pulse of 40 milliseconds for the reference sample at the center of the 31P-surface coil. The acquisition of 31P-MR spectra was triggered to the R-wave of the electrocardiogram, with a trigger delay time of 200 milliseconds and a recycle time of 3.6 seconds. ISIS volume selection in 3 dimensions (3D-ISIS) based on 192 averaged free induction decays was employed. The volume size was typically 5 (caudo-cranial) × 6 × 5 cm3 avoiding inclusion of chest wall muscle and diaphragm muscle. Acquisition time was 10 minutes. Adiabatic frequency-modulated hyperbolic secant pulses and adiabatic half-passage detection pulses were used to achieve inversion and excitation over the entire volume of interest. Examination time was 40 to 45 minutes.

Cardiac MRI.

MRI studies were performed using standard MRI methodology as previously described in detail,12 with the previously described scanner, using an enhanced gradient system with a maximum gradient strength of 30 mT/m and a maximum gradient slew rate of 150 mT/m/sec. The Cardiac Research software patch (operating system 9) was used.

Analytical Determinations

Glucose concentration was measured with standard glucose oxidase method on a glucose analyzer (Beckman Coulter, Fullerton, CA). Free fatty acid (FFA), triglycerides, total cholesterol, and high-density lipoprotein cholesterol were measured as previously described.11, 13 Plasma levels of insulin and leptin were measured with radioimmunoassay (Linco Research, MO), following the manufacturer's assay protocols. Plasma c-peptide was also measured using a radioimmunoassay kit (Diagnostic Product Corporation, Los Angeles, CA). Serum resistin [enzyme-linked immunosorbent assay (ELISA) kit; BioVendor Laboratory Medicine, Inc, Brno, Czech Republic] and adiponectin (ELISA kit B-Bridge International, Inc., Sunnyvale, CA) were measured as previously described.11 Serum hs c-reactive protein was measured using an ELISA kit (Diagnostic Biochem Canada Inc., London, Ontario, Canada) with a sensitivity of 10 ng/mL and intra-assay and interassay coefficient of variation (CV) < 5% and < 7%, respectively. Serum E-selectin was measured by ELISA kit (R&D Systems, Inc., Minneapolis, MN); the sensitivity was of 0.1 ng/mL, the intra-assay CV was <5%, and interassay <9%. Serum creatinine was measured using an enzymatic method on a Hitachi 747.11 Thyroid-stimulating hormone was measured by immunofluorimetric method. Blood pressure was measured twice (at the end of the MRI–MRS session) with volunteers in the lying position.

Calculations

IHF Content.

The IHF was calculated as previously described,10, 11, 13 and absolute concentrations expressed as percent fat by weight of volume using equations validated by Longo et al.15 Traditionally, liver fat content greater than 50 mg/g (5% by wet weight and equivalent to 6.5% of ratio methylene/methylene + water × 100 in our setting) is diagnostic of hepatic steatosis,16, 17 and study subjects were diagnosed as individuals with normal IHF content (<5% wet weight) or higher than normal IHF content (>5% wet weight).

31P-MRS Analysis.

31P-MR spectra were analyzed automatically in the time domain using Fitmasters, as previously described,12 based on correction for the adenosine triphosphate (ATP) contribution from blood18 and for partial saturation.19, 20 An estimate of the signal-to-noise ratio of each spectra was obtained from the relative Cramer-Rao standard deviation calculated for the PCr/ATP, which is a commonly reported index of accuracy of the spectral quantification.19

MRI Analysis.

Image analysis was performed using an image-processing workstation (EasyVision; Philips Medical Systems) by using the cardiac analysis software package as previously described.12

Cardiac Fat Compartmentation.

The thoracic region was also imaged by MRI using a standard protocol with electrocardiogram triggering and patients in breath-hold (10-12 seconds) during the acquisition time as reported by Sironi et al.21 Epicardial adipose tissue scans were obtained by fast-spin echo T1-weighted sequences with oblique axial orientation, for a correct study of horizontal long axes of the heart (TE, 42 ms; echo train length, 23; bandwidth, 62.50; slice thickness, 8 mm; slice gap, 0 mm; field of vision, 38 cm; matrix, 288 × 224; phase field of vision, 0.75; NEX 1; Trigger delayminimum 8-mm-thick section with 0-mm intersection gap, field of view, and a 256 × 256 matrix). Extrapericardial and intrapericardial fat areas were measured using a semiautomatic program and are expressed as centimeters squared.

Insulin Sensitivity.

We estimated insulin sensitivity and secretion using the updated computer model HOMA-2 indexes,22 available from http://www.OCDEM.ox.ac.uk.

Statistical Analysis

Data in text, tables, and figures are mean ± standard deviation. Analyses were performed using the SPSS software (ver. 13.0; SPSS Inc, Chicago). When parameters showed a skewed distribution (Kolmogorov-Smirnov test of normality), they were log-transformed before the analysis (PCr/ATP ratio, IHF content, diastolic blood pressure, homeostasis model assessment [HOMA] percent insulin activity). Comparison between groups was performed using 2-tailed independent-samples t test or Kruskal-Wallis nonparametric test, depending on the distribution of the data, and a P value less than 0.05 was considered statistically significant. Two-tailed Pearson's correlation was performed to establish partial correlation coefficients between variables, and the nonparametric correlation coefficient was obtained using Spearman's rho when appropriate. Statistical significance was defined as a P value < 0.05. A prior power calculation analysis indicated that 17 subjects per group were required to provide a power of 90% to detect a 20% difference in PCr/ATP ratio between groups; 13 subjects per group were required to provide a power of 90% to detect a 10% difference in ejection fraction, 21 subjects per group were required to provide a power of 90% to detect a 20% difference in E/A peak flow ratio, and 16 subjects per group were required to provide a power of 91% to detect a 25% difference in end-diastolic wall mass.

Results

Intrahepatic Fat Content.

IHF content assessed by means of 1H-MRS (Fig. 1) ranged between 5.40 and 39.14%ww in individuals with fatty liver and between 0.69 and 4.65%ww in the control group.

Figure 1.

1H spectra of the liver. Two representative 1H spectra are shown. The spectrum on top shows the water signal (4.7 ppm) and the the CH2- residues peak of intrahepatic triglycerides (1.35 ppm) of a healthy volunteer with a normal IHF content (1.3% ww). On the bottom the spectrum of an individuals with fatty liver (36.1% ww) is shown.

Anthropometric and Biochemical Characteristics of Study Groups (Table 1).

The anthropometric features of the 2 groups are summarized in Table 1. Age and BMI were not different between groups; study subjects were, on average, a population of overweight individuals. Arterial blood pressure was within the normal range even if men with increased IHF content had a trend for higher systolic values (P = 0.07). Fasting plasma glucose was higher in subjects with increased IHF content (P = 0.025). Total cholesterol was not different between groups; meanwhile, high-density lipoprotein cholesterol was reduced (P = 0.018), and triglycerides were slightly increased (P = 0.05) in men with increased IHF content in comparison with normal subjects. The prevalence of the metabolic syndrome was higher in men with increased IHF content (43%) than in normal subjects (9.5%; Pearson chi-squared = 16.81; P < 0.01). Creatinine and thyroid-stimulating hormone were not different between groups. Physical activity index (data not shown) was not different between groups.

Insulin Sensitivity and Adipocytes' Hormones.

Fasting plasma insulin was not different between groups; meanwhile, c-peptide was higher in individuals with excessive IHF content in comparison with normals (P = 0.023). Despite accurate BMI matching, insulin sensitivity [surrogate index of insulin sensitivity (HOMA2-%S); P = 0.047) was impaired in men with excessive IHF content in comparison with normals (Table 2). Fasting plasma FFA concentration was higher in men with excessive IHF content in comparison with normals (P = 0.036). Plasma leptin was increased in subjects with excessive IHF content in comparison with normals in contrast with plasma adiponectin and resistin concentrations, which were reduced. Serum high-sensitivity C-reactive protein was not different between groups. Serum E-selectin was higher in patients with (47 ± 19 ng/mL) than in individuals without fatty liver (38 ± 14 ng/mL; P < 0.05).

Table 2. Parameters of Insulin Sensitivity, Circulating FFA, Adipokines, and hsCRP in Individuals With (IHF Content > 5% ww) or Without (IHF Content < 5% ww) Excessive Hepatic Fat Accumulation
 IHF Content < 5% Wet WeightIHF Content > 5% Wet WeightP Value
  1. Values are expressed as mean ± SD.

  2. Independent-samples t test (2-tailed) or Kruskal-Wallis nonparametric test.

Fasting plasma insulin (pmol/L)85 ± 28105 ± 400.18
Fasting plasma c-peptide (nmol/L)0.71 ± 0.290.96 ± 0.310.018
HOMA2-%B152 ± 43152 ± 460.81
HOMA2-%S63 ± 1851 ± 190.047
FFA and adipokines   
Free fatty acids (μmol/L)559 ± 184711 ± 1940.014
Leptin (ng/mL)5.7 ± 2.88.5 ± 3.60.015
Adiponectin (μg/mL)7.4 ± 2.65.3 ± 1.90.015
Resistin (ng/mL)4.0 ± 1.03.3 ± 0.80.019
hsCRP (mg/L)0.90 ± 0.851.36 ± 1.140.19

LV Anatomical and Functional Features.

Morphological parameters of the LV were not different between the 2 groups of study (Table 3). Also, systolic and diastolic functions were not different in patients with excessive IHF content.

Table 3. Morphologic Parameters and Functional Features of the LV in Men With (IHF Content > 5% ww) or Without (IHF Content < 5%ww) Excessive Hepatic Fat Accumulation
 IHF Content < 5% Wet WeightIHF Content > 5% Wet WeightP Value
  1. Values are expressed as mean ± SD.

  2. Independent-samples t test (2-tailed) or Kruskal-Wallis nonparametric test.

Heart rate (beats/min)63 ± 761 ± 80.51
Morphological features   
End diastolic volume (mL)140 ± 29140 ± 220.97
End systolic volume (mL)53 ± 1550 ± 100.44
End diastolic wall mass (g)145 ± 20144 ± 200.88
Systolic function   
Stroke volume (mL)86 ± 1689 ± 140.57
Cardiac output (L/min)5.4 ± 0.75.5 ± 1.00.76
Ejection fraction (%)62 ± 464 ± 30.07
Diastolic function   
Early peak filling rate (mL/sec)432 ± 106440 ± 720.80
Atrial peak filling rate (mL/sec)207 ± 96204 ± 750.93
E/A peak flow2.0 ± 0.62.2 ± 0.60.36
Deceleration time (msec)198 ± 36194 ± 350.77

LV PCr/ATP Ratio.

The PCr/ATP ratio, assessed by means of 31P-MRS (Fig. 2), was reduced in patients with excessive IHF content in comparison with normals (1.84 ± 0.34 versus 2.11 ± 0.32, respectively; P = 0.016, Fig. 3). Despite the overweight condition affecting the study subjects, accuracy was excellent: the mean relative Cramer-Rao standard deviation was not different among groups (16 ± 6 and 16 ± 7%; P = 0.96). The intra-examination variability in our setting in overweight patients was already reported (CV: 7 ± 4%), as well as the interexamination variability (CV: 16 ± 5%).23 As for consistency of the acquisition protocol, the average distance of the coil center from the LV wall (4.7 ± 0.6 versus 4.6 ± 0.8 cm; P = 0.91) and the size of the volume of interest (149 ± 30 versus 164 ± 57 cm3, P = 0.39) were not different between groups.

Figure 2.

31P MR spectra of the LV. Two representative 31P spectra are shown. The spectrum on top shows resonance peaks of phosphocreatine (PCr), adenosine tri-phosphate (ATP), phosphodiesters (PDE), and the combined signals of 2,3-diphosphoglycerate (2,3-DPG) originating from blood and inorganic phosphate (Pi) of a healthy volunteer with normal IHF content, and the spectrum in the lower panel shows the resonance peaks of an individual with fatty liver.

Figure 3.

LV PCr/ATP ratio and intrapericardial and extrapericardial fat. (A) The PCr/ATP ratio (mean ± SD) was lower in patients with fatty liver (empty columns) than in the controls (filled columns)(P = 0.016; Kruskal-Wallis nonparametric test). (B,C) The intrapericardial and extrapericardial fat was higher in men with fatty liver (empty columns) compared with those without fatty liver (filled columns) (P < 0.05 and P < 0.03, respectively; 2-tailed independent-samples t test).

Cardiac Fat Compartmentation.

In the scanned intrathoracic region, the intrapericardial (7.8 ± 3.1 versus 5.9 ± 2.5 cm2; P < 0.05) and extrapericardial (11.7 ± 6.1 versus 7.8 ± 3.2 cm2; P < 0.03) fat was increased in men with fatty liver compared with those without fatty liver.

Correlation Analysis.

Spearman's rho correlation analysis showed that the PCr/ATP ratio was associated with the diastolic blood pressure (r = −0.48; P < 0.008) and with fasting plasma glucose (r = −0.35; P < 0.05) and insulin (r = −0.37; P < 0.03) but was not significantly associated with the IHF content (r = −0.33; P < 0.08), serum FFA or adipokines and E-selectin concentrations, intrapericardial, or mediastinal fat. Both the intrapericardial and extrapericardial mediastinal fat were associated with many anthropometric and metabolic typical factors of the metabolic syndrome, but in general the extrapericardial mediastinal fat was more closely associated with BMI, IHF, total cholesterol and triglycerides, glucose, diastolic blood pressure, and metabolic syndrome than the intrapericardial fat. Even if they were not correlated with the PCr/ATP ratio, we detected a significant association between the early/atrial peak filling rate ratio with the intra-pericardial (r = −0.45; P < 0.005) and extrapericardial mediastinal (r = −0.36; P < 0.03) fat.

Discussion

This study reports the novel finding that in the resting state the high-energy phosphates' metabolism of the LV of individuals with excessive IHF accumulation is impaired. This alteration was detected despite normal morphology or systolic and diastolic function, supporting the hypothesis that the metabolic defect may be an early alteration in the LV of these individuals. The imaging data suggest that an associated alteration was a greater amount of epicardial fat as a feature of visceral adiposity.

This work was designed as an attempt to find a relationship between hepatic and cardiac metabolism in vivo in humans. Strength of the current work was the use of combined advanced noninvasive techniques; quantitation of the IHF content as a continuous variable was performed by means of 1H-MRS, which can be used in the clinical setting, and its accuracy and safety make it an ideal methodology to assess and monitor changes in the IHF content.10 In parallel, LV high-energy phosphate metabolism was studied by means of 31P-MRS, and the detection of potential, parallel alterations in the LV anatomy and function were investigated by means of cardiac MRI in the clinical setting. An additional strength of this study was the choice of studying young and matched individuals, avoiding the confounding effects of other factors including age, anthropometric features, exercise habits, ethnicity, diabetes, hypertension, and drug therapy. Based on this assumption, the significance of the LV abnormality of energy metabolism appears to be an early event in young men with excessive IHF content, in the absence of any other morphological or functional cardiac alteration. The isolated detection of reduced PCr/ATP ratio in these subjects would support the hypothesis that cardiac metabolic remodeling is taking place and may precede the development of functional and structural remodeling of the heart.24 In the past, a reduced PCr/ATP ratio was reported in patients with ischemic cardiac disease,25 cardiac syndrome X,26 hypertension,27 type 1 diabetes, end-stage renal failure,12 and type 2 diabetes,28, 29 but in all studies LV hypertrophy and systolic/diastolic dysfunctions were also observed in parallel with the metabolic alteration. Moreover, depletion of creatine was associated with the progression of heart failure in patients with dilated and hypertrophic cardiomyopathy.30 A limit of the current work is obviously its cross-sectional nature, which does not allow conclusions regarding the potential link between the impairment of the PCr/ATP ratio and the development of heart failure in these patients; this issue remains to be ascertained, even if it is reported that patients with NAFL have endothelial dysfunction31 and increased risk of atherosclerosis.32

Based on our data, the pathogenesis of the altered LV energy metabolism in men with fatty liver is obscure. Ectopic fat accumulation within the liver is believed to be mainly sustained by an increased adipose-derived FFA flux.33 It is well established that the primary fuel for myocardium in the fasting state are FFAs34 and that a reduced myocardial fatty acids oxidation is an independent predictor of dysfunction in hypertensive patients.35 In our patients with fatty liver, the fasting plasma FFA concentration was increased (Table 2); therefore, the excessive FFA flux to the heart may be important, as recently suggested by Kankaanpää et al.36 When the supply of FFAs exceeds the rate of their oxidative disposal, oxidative stress is induced, and intracellular lipid intermediates (triglycerides, ceramides, long chain fatty acyl-CoA) concentrations increase with a consequent lipotoxicity that in the short run may determine impairment of energy homeostasis and in the long run may determine contractile dysfunction.37 Against this hypothesis, the dependence on myocardial fatty acid metabolism could not be completely explained by increased serum FFA level in insulin-resistant, obese, and young women,38 and in our study subjects the fasting FFA level, even if higher, was not correlatively associated with impairment of LV energy metabolism. Conversely, one main finding of the current work is that the amount of epicardial fat was increased (Fig. 3B,C), and we cannot exclude that this visceral adipose tissue may be the source of an extra amount of FFA locally, not detectable based on the measurement within the peripheral blood. There is also the possibility that the adipose tissue may influence the hepatic and cardiac metabolism through the release of adipokines,39 and for this reason we measured the fasting serum concentration of leptin, adiponectin, and resistin. The serum concentrations of these adipokines were different between groups (Table 2), but a statistical relationship with the PCr/ATP ratio was not found.

Low-grade inflammation has been linked to both NAFL40 and heart failure41 and therefore, because it is an additional potential underlying mechanism common to hepatic steatosis and impaired cardiac energy homeostasis, we measured serum high-sensitivity C-reactive protein concentration. We failed to find (1) a correlative relationship between its serum concentration and the PCr/ATP ratio, or (2) a difference between the 2 groups, but it should be noted that that high-sensitivity C-reactive protein was slightly but significantly associated with the IHF content (Spearman's rho: r = 0.40; P = 0.038). We cannot therefore exclude the possibility that hepatic steatosis per se may cause low-grade inflammation, which in turn may affect myocardial metabolism.

Insulin resistance itself may be involved in the metabolic dysfunction, as we have recently suggested studying a population of obese individuals23; in fact, in this study both fasting plasma glucose and insulin concentrations were significantly associated in an inverse relationship with the PCr/ATP ratio as well as the surrogate index of insulin sensitivity. Other possible factors involved in the altered PCr/ATP ratio may not be related to classical metabolic factors; patients with NAFL have impaired flow-mediated vasodilation,32 and we found that serum E-selectin concentration was higher in patients with fatty liver than in control patients, supporting a potential role for an underlying endothelial dysfunction in patients with NAFL. We also found that, even if not pathological, the diastolic blood pressure in these patients was higher than in the controls.

In conclusion, the current study provides evidence of an abnormal energy metabolism affecting the LV of young and newly diagnosed individuals with fatty liver. This finding was accompanied by another metabolic alteration, excessive fat accumulation in the epicardial area. Whether this myocardial metabolic alteration is attributable to an excessive flux of substrates (FFA) or inflammatory mediators (adipocytokines) locally released by the accumulated epicardial fat is a fascinating hypothesis that needs further study.

Ancillary