Oxidative Molecular Mechanisms Underlying Liver Diseases: From Systems Biology to the Personalized MedicineView this Special Issue
Research Article | Open Access
Fasting Whole-Body Energy Homeostasis and Hepatic Energy Metabolism in Nondiabetic Humans with Fatty Liver
Background. Fatty liver is believed to be sustained by a higher than normal adipose-derived NEFA flux to the liver. Also, hepatic energy metabolism may be a rate-limiting step of intrahepatic fat (IHF) accumulation. Aims. To assess whole-body energy metabolism and hepatic high-energy phosphates (HEPs) in individuals with fatty liver. Methods. We studied 22 individuals with fatty liver and 22 control individuals matched for anthropometric features by means of (1) hepatic 1H-magnetic resonance spectroscopy (MRS) to measure the IHF content, (2) hepatic 31P-MRS to assess the relative content of HEPs (phosphomonoesters, phosphodiesters, inorganic phosphorus, and ATP), and (3) indirect calorimetry to assess whole-body resting energy expenditure and substrate oxidation. Results. Patients with newly diagnosed fatty liver and controls were not different for anthropometric parameters. Based on HOMA2%-S, individuals with fatty liver were more insulin resistant than controls. Resting energy expenditure and the pattern of substrate oxidation were not different between groups. Relative content of HEPs was not different between groups; in particular, the Pi/γ-ATP ratio, the most important signals in terms of monitoring energy homeostasis, was not different even if it was associated with indirect calorimetry-derived parameters of oxidative substrate disposal. Conclusions. These data demonstrate that fasting whole-body energy metabolism and the relative content of HEPs in nondiabetic patients with fatty liver are not different than those in controls when they are matched for anthropometric features.
Fatty liver is the body composition manifestation of visceral obesity in insulin-resistant subjects [1, 2]. Intrahepatic fat accumulation is thought to be due to increased adipose-derived NEFA flux to the liver  as reported in the fasting state , during euglycemic-hyperinsulinemic clamps [4, 5] and during OGTT . Insulin resistance with respect to lipolysis, therefore, plays a relevant role in patients with fatty liver. Using tracer methodologies, it was found that in patients with NAFLD 60% of liver triglycerides arises from NEFA in the fasting state . In the same conditions, 26% of liver triglycerides arose from de novo lipogenesis , and an apparent increased contribution of de novo lipogenesis vs. NEFA reesterification was reported in patients with NAFLD [7, 8]. Based on this evidence, efforts to treat hepatic steatosis by reducing fatty acids flux through dietary and pharmacological therapy (thiazolidinediones) received strong emphasis [9, 10]. Fatty liver is associated not only with hepatic but also with whole-body insulin resistance which frequently associates with higher intramyocellular lipid contents (IMCL) . IMCL accumulation is associated with impaired muscle energy metabolism , and also, the heart of insulin-resistant subjects with fatty liver showed impaired energy metabolism .
Little data on the association between fatty liver and hepatic energy metabolism are currently available; Sharma et al.  reported that using hepatic 31P-MR spectroscopy, the fasting relative content of high-energy phosphates (HEPs) was altered in obese individuals with NAFLD than in controls. In the present study, we performed hepatic 31P-MRS to assess the liver content of HEPs and indirect calorimetry to assess whole-body resting energy expenditure and substrate oxidation to look for alterations reflecting potential abnormalities of energy metabolism in individuals with newly found fatty liver, when compared to age and BMI-matched individuals used to rule out the effect of overweight and obesity.
2. Materials and Methods
Study subjects were selected among a large group of otherwise healthy employees of the San Raffaele Scientific Institute previously studied to assess the relationship between habitual physical activity and hepatic triglyceride content using hepatic 1H-MRS . Within that original population, twenty-two individuals known to have excessive IHF content and twenty-two control individuals known to have normal IHF content and selected to be comparable for the anthropometric features to those with fatty liver (in order to minimize the confounding effect of obesity on the parameters of interest) performed hepatic 31P-MRS to assess HEP content and indirect calorimetry to assess whole-body energy metabolism. The 1 : 1 procedure of matching was based using the criteria of age within 3 years and BMI within 1 unit. This strategy has been similarly used in the past to assess whether a difference in cardiac HEP metabolism could be associated with the excessive IHF content , and thirteen patients out of the twenty-two with fatty liver and ten control subjects out of the twenty-two who participated to the cardiac protocol  performed this additional measurement of HEP content also at the level of the liver. According to the American Association for the Study of Liver Diseases (AASLD), normal or higher than normal IHF content was set at 5% ww . Outpatients admitted to the Center of Nutrition/Metabolism of the San Raffaele Scientific Institute were studied. In order to be recruited, body weight had to be stable for at least six months. Patients with history of hepatic disease, substance abuse, or daily consumption of more than one alcohol drink daily (<20 g/day) or the equivalent in beer and wine were excluded from the study. Table 1 summarized the anthropometric characteristic of the recruited subjects, which, based on the medical history, physical examination, blood and urinary tests, were in good health. An informed written consent was collected for each subject who participated in the study. Recruited subjects gave their informed written consent after the explanation of purposes, nature, and potential risks of the study. The Ethical Committee of the Istituto Scientifico H San Raffaele approved the studies.
; independent-samples -test (2-tailed). HOMA: HOMA2: homeostatic model assessment; HOMA2-S: insulin sensitivity; HOMA2-B: β-cell sensitivity.
2.2. Experimental Protocol
In three days before the study, the recruited subjects were asked to consume isocaloric diet and to abstain from exercise activity. They were studied after an 8-10-hour overnight fast by means of 31P-MRS and 1H-MRS for the assessment of the hepatic relative high-energy phosphates and the IHF content, respectively. Assessment of whole-body energy metabolism by means of indirect calorimetry and blood drawing to measure serum insulin, C-peptide, plasma glucose, NEFA, lipid profile, and biochemical parameters was performed the same morning or a few days apart.
2.2.1. 31P-MR Spectroscopy
Hepatic 31P-MRS was performed with volunteers in the supine position within a 1.5 T whole-body scanner (Gyroscan Intera Master 1.5 MR System; Philips Medical Systems, Best, the Netherlands). 31P spectra were obtained by means of a 10 cm diameter surface coil used for the transmission and detection of radio frequency signals at the resonance frequency of 31P (at 1.5 T, 25.85 MHz). A small sample container built in the coil center, containing an aqueous solution of methyl-phosphonate, served as geometrical reference. The surface coil was secured in place with a Velcro band around the abdomen and chest, helping to minimize breathing artifacts. MR imaging was performed to acquire scout images, to establish the exact position of the 31P surface coil, and eventually to reposition it. Localized homogeneity adjustment was performed using the body coil by optimizing the 1H-MR spectroscopy water signal. Shim volumes were planned on the transverse and sagittal scout images. The transmitter-receiver was then switched without time delay to the 31P frequency. Manual tuning and matching of the 31P surface coil was performed to adjust for different coil loading. The radio frequency level was adjusted to obtain a 180° pulse of 40 ms for the reference sample at the center of the 31P-surface coil. The acquisition of 31P-MR spectra was performed with a recycle time of 3.6 s. ISIS volume selection in three dimensions (3D-ISIS) was employed. It was based on 192 averaged free induction decays. The Volume of Interest (VOI) was oriented perpendicular to the abdomen wall, avoiding inclusion of abdominal wall muscle and diaphragm muscle. The volume size was approximately 3 cm3. The acquisition time was 11 min. Adiabatic frequency-modulated hyperbolic secant pulses and adiabatic half-passage detection pulses were used to achieve inversion and excitation over the entire VOI. The examination time was 40-45 min. 3D-ISIS was employed after testing that uses higher spatial resolution (2D-ISIS+1D SI using a one-dimensional phase encoding bar with 32 rows of 1 cm thickness each angulated perpendicular to the abdominal wall for the anterior-posterior direction and lateral and craniocaudal dimensions dependent on the patients’ liver size); the Pi/γ-ATP ratios were in agreement and showed absence of muscle PCr signal.
2.2.2. 1H-MR Spectroscopy
Hepatic 1H-MR spectroscopy was performed in all volunteers with the same MR system as previously described .
2.2.3. Indirect Calorimetry
After lying quietly for 30 min, REE was measured by continuous indirect calorimetry with a ventilated hood system (SensorMedics 2900, Metabolic Measurement Cart) performed for 45 min as previously described . The mean coefficient of variation (CV) within the session for both O2 (%) and CO2 (%) measurements was below 5%.
2.3. Analytical Determinations
Glucose (Beckman Coulter Inc., Fullerton, CA), FFA, triglycerides, total cholesterol, and HDL cholesterol were measured as previously described . Plasma levels of insulin (sensitivity 2 μU/ml; intra- and interassay CV <3.1% and 6%, respectively) were measured with RIA (Linco Research, Missouri, USA). Plasma C-peptide was measured (CVs: %, %) using a double-antibody RIA kit (Diagnostic Product Corporation, Los Angeles, CA).
Fasting-based indices of insulin sensitivity (HOMA2-%S) and β-cell responsivity (HOMA2-%B) were determined by the updated HOMA2 method  available from http://www.OCDEM.ox.ac.uk. 31P-MR spectra were transferred to a remote SUN-SPARC workstation for analysis. The spectra were quantified automatically by model function analysis in the time domain. The spectral fitting routine was based on a nonlinear least-squares Gauss-Newton implementation for exponential damping as previously performed for the postprocessing of spectra obtained from transplanted kidney . A typical 31P spectrum is depicted in Figure 1 with description of the signals in the legend. γ-ATP and Pi are considered the most important signals in terms of monitoring energy homeostasis because the γ-ATP phosphorous is the one which is released as Pi in the reaction of ATP hydrolysis to ADP. For this reason, in the present work, we have chosen the Pi/γ-ATP and the Pi/ATP (mean of the three phosphorous signals) ratios as markers of intrahepatic ATP metabolism. We also calculated the PME/Pi, PME/γ-ATP, and PME/ATP ratios as comparison with previously published data in hepatic diseases. The percent IHF was calculated as previously described . REE was calculated by Weir’s standard equation  from the O2 consumption rate and the CO2 production rates measured by means of indirect calorimetry (excluding the first 10 min of data acquisition) and from the urinary nitrogen excretion. Predicted REE was calculated using the Harris-Benedict equations . Glucose, lipid, and protein oxidation was estimated as previously described .
2.5. Statistical Analysis
Data in text tables are . Analyses were performed using the SPSS software (ver. 13.0; SPSS Inc., Chicago). Comparison between groups was performed using the 2-tailed independent-samples -test, and a value less than 0.05 was considered to be significant. Variables with skewed distribution assessed using the Kolmogorov-Smirnov test of normality were log-transformed before the analysis. Correlation analysis was performed using two-tailed Pearson’s correlation.
Sample size calculation was based on preliminary acquisitions in a small () group of normal-weight subjects. 11 individuals for each group provided a power of 85% using a -test at a one-side α of 0.05 when assuming a 20% difference between groups in the Pi/γ-ATP ratio with a standard deviation of 0.50.
3.1. Anthropometric and Laboratory Characteristics of Study Subjects
Individuals with or without fatty liver ( deviation and ranges of the IHF content are summarized in Table 1) were not different with respect to age and BMI (Table 1). Subjects with excessive IHF content had higher serum triglycerides and plasma C-peptide concentrations. HOMA2-S, as a surrogate index of insulin sensitivity, was lower in subjects with fatty liver than in normals.
3.2. Parameters of Whole-Body Energy Metabolism
Resting energy expenditure (REE) and the percent of predicted REE were not different between groups. Also, whole-body glucose and lipid oxidation was not different between groups as reflected by the similar respiratory quotient (Table 2).
The peak area of each metabolite (PME, Pi, PDE, γ-ATP, α-ATP, and β-ATP) was expressed as a percentage of the total 31P-MR signal (the sum of all the resonances). ; independent-samples -test (2-tailed). REE: resting energy expenditure.
3.3. Intrahepatic High-Energy Phosphates
The intrahepatic high-energy phosphate content is summarized in Table 2. The peak area of each metabolite (PME, Pi, PDE, γ-ATP, α-ATP, and β-ATP) was expressed as a percentage of the total 31P-MR signal (the sum of all the resonances). No difference was detected for each of the metabolites between the two groups of study. Also, the Pi/γ-ATP, Pi/ATP, PME/Pi, PME/γ-ATP, and PME/ATP ratios were not different between groups.
3.4. Correlation Analysis with Parameters of Hepatic High-Energy Phosphates
The Pi/γ-ATP ratio was not associated with anthropometric features or with the features of the metabolic syndrome. It was not associated with the IHF content (; ). It was not associated with parameters of lipid metabolism and fasting plasma NEFA or with fasting plasma glucose, insulin, and HOMA indices. Interestingly, it was associated with indirect calorimetry-derived parameters of oxidative substrate disposal; in fact, it correlated with the respiratory quotient (; ), the fasting whole-body glucose oxidation (; ), and the fasting lipid oxidation (; ). This association with the respiratory quotient was detectable in a more robust fashion in individuals with fatty liver (; ) than in the control individuals (; ). A similar pattern of correlation was found when the Pi/ATP ratio was used. The PME/Pi and PME/ATP ratios were not correlated with any of the metabolic parameters summarized in Table 1.
The present study demonstrates that in nondiabetic individuals with newly found fatty liver, the fasting whole-body energy metabolism and the relative amount of hepatic HEPs are not different than in age- and BMI-matched nondiabetic individuals without fatty liver.
Little data about whole-body energy metabolism were available in patients with fatty liver. Bugianesi et al.  reported that during hyperinsulinemic conditions, whole-body lipid oxidation rate was higher in a small group of individuals with biopsy proven NAFLD () when compared to 6 control subjects. As discussed in the introduction section, this finding may be secondary to the higher plasma NEFA characterizing these patients [4–6]. In our study, we did not assess insulin sensitivity with the clamp procedure, but based on the higher insulin resistance estimated using the fasting-derived index (HOMA2-S in Table 1), it is likely that also our cohort of subjects in the same insulin-stimulated condition could be characterized by higher lipid oxidation rate. The other facet of energy metabolism is the postabsorptive and resting state. In our cohort of individuals with newly diagnosed fatty liver, postabsorptive whole-body energy metabolism was not different than that of controls in absolute terms and when normalized to the predicted values based on sex, age, race, body weight, and height. In addition, substrate oxidative disposal was remarkably similar to that of controls based on semiquantitative parameters such as the respiratory quotient and also based on the oxidative fluxes (Table 2). This may be substantially in agreement with the report of Bugianesi et al. showing only a trend () for higher fasting lipid oxidation rate in their 12 patients when compared to 6 controls. It is possible that a more severe hepatic disease in Bugianesi’s cohort may explain this little discrepancy. In fact, we have also observed that in more serious hepatic diseases, such as in individuals with cirrhosis and hepatocarcinoma, a hypercatabolic state sustained by higher fasting lipid oxidation rate could be detected and that liver transplantation in the long-term normalized this abnormality .
We may conclude that in individuals with newly diagnosed fatty liver, whole-body energy metabolism is not different than in age- and BMI-matched individuals during the postabsorptive and resting state.
With respect to the local hepatic energy metabolism as assessed by means of 31P-MRS, Sharma et al.  reported higher PME/Pi, PME/γ-ATP, PME/β-ATP, and PME/ATP in overweight/obese Asian Indians with NAFLD when compared with normal weight individuals, either with or without NAFLD. We believe that the different HEP contents were not due to the fatty liver per se but rather to the different degrees of obesity. Several observation may support our thinking. In our study, when we compared the overweight subjects with fatty liver with a group of similarly overweight subjects without fatty liver, a difference was not detectable (Table 2). Taking a careful look to the data by Sharma et al. , it may be observed that when anthropometric parameters were matched between individuals with and without NAFLD (in their study the comparison may be performed at a normal range of BMI: 22 kg/m2), the difference was not significant also in their own set of data. In support of this hypothesis, a pilot study using 31P-MRS of the liver suggested that in 8 obese individuals with NASH, ATP recovery within the liver after the administration of fructose e.v. was severely blunted, but the authors reported that the recovery became progressively less efficient as BMI increased also in the group of healthy controls . This study would also confirm that this impairment is present in association with obesity per se, and this suggestion is also given by another study in which ATP was monitored in normal weight and obese individuals in the fasting state and after fructose administration .
An alternative potential explanation for this controversial finding is related to a limitation of our and Sharma’s works. In both studies, there was a lack of histological characterization of the hepatic disease; therefore, no information about the inflammatory status, necrosis, and fibrosis was available. It may not be excluded that different histological features of the disease may be able to explain the reported differences in HEP relative content. In fact, increased PME ratios have been repeatedly found in patients with more severe liver diseases including chronic hepatitis C , liver cirrhosis , or tumors , all conditions in which extensive membrane remodeling may have an effect on the membrane phosphocholine and phosphoethanolamine kinetics and ultimately in PME content.
Based on our own data, we therefore would like to conclude that the relative content of hepatic HEPs is not different in individuals with newly found fatty liver that in controls when study groups are comparable for body mass and that the previously reported differences may be secondary to body fatness rather than fatty liver per se.
In summary, our set of data would support the hypothesis that fasting whole-body and hepatic energy metabolism is not abnormal in individuals with fatty liver when compared to matched controls in agreement with findings generated using different methodological tools , in contrast with the finding within the skeletal muscle  and the heart of insulin-resistant individuals .
Since in the postabsorptive and resting state the liver, along with the skeletal muscle, is the tissue contributing most to energy expenditure, we wondered whether the acquired parameters could be somehow related to each other. We noticed that a significant correlation could be detected between the whole-body substrate oxidative disposal (expressed as respiratory quotient, glucose, and lipid oxidation) and the Pi/γ-ATP ratio. The correlation would imply that a relatively higher hepatic content of Pi (a more catabolic condition if the hepatocellular ATP content is postulated to be comparable) is present when lipids are the preferential substrate for the oxidative processes. This correlation was not a peculiar feature of the individuals with fatty liver but was detectable also in the control individuals without fatty liver even if by a minor extent.
Even if these 31P-MRS studies do not allow the assessment of the absolute hepatic HEP concentrations, which represents a methodological limitation, the expression of the signal intensities as ratios would imply that the intracellular hepatic ATP concentration is constant and not different between groups. A higher Pi/γ-ATP ratio would therefore imply for a catabolic energetic state of the hepatocytes, in which a reduced phosphorylation status would be reflected by an increase of Pi levels relative to ATP. The lack of difference of the Pi/γ-ATP ratio between groups we are reporting in the present work would therefore imply a normal hepatic energy metabolism, even if we cannot exclude that the absolute Pi and ATP hepatocellular content may be consensually higher and/or lower in parallel resulting in no change in the ratio. We believe that the lack of differences in the Pi/γ-ATP ratio between the groups in the present work implies a normal hepatic metabolism because we also analyzed our data as the relative percent contribution of each single signal to the total 31P detectable signals, as summarized in Table 2, and once more we found no significant difference between groups. It is important to emphasize that future studies will require the determination of the absolute substrate concentration. With this respect, Chmelík et al.  have validated a spectroscopic absolute quantification of 31P metabolites in human livers based on the use of an external phantom with known concentrations of the substrates of interest.
In conclusion, the present work demonstrates that fasting and resting whole-body energy metabolism and the relative content of HEPs in nondiabetic patients with newly diagnosed fatty liver are not different than those in controls when they are matched for anthropometric features. Absolute quantification of these substrates remains a mandatory step to firmly establish whether alteration of hepatic energy metabolism is present in individuals with fatty liver.
|31P-MRS:||Phosphorus-31 magnetic resonance spectroscopy|
|1H-MRS:||Proton magnetic resonance spectroscopy|
|NEFA:||Nonesterified fatty acids|
|OGTT:||Oral glucose tolerance test|
|BMI:||Body mass index|
|3D-ISIS:||ISIS volume selection in three dimensions|
|REE:||Resting energy expenditure|
|HOMA:||Homeostatic model assessment|
|IMCL:||Intramyocellular lipid contents|
|NAFLD:||Nonalcoholic fatty liver disease.|
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors have no potential competing interests and conflicts to declare.
G.L. and G.P. designed and conducted the research, contributed to the interpretation of the results, drafted the manuscript, and approved the final version. M.G.R., D.C.F., E.A., and D.M.A. contributed to the study design, conducted the research, performed the statistical analysis, reviewed the manuscript, and approved the final version. G.M., S.P., and G.C. contributed to the interpretation of the results, performed the statistical analysis, contributed to the manuscript writing, and approved the final version.
This study was supported by grants by (1) the Italian Ministry of Health (RF98.49, RF99.55, and RF01.1831): this grant was received in order to develop MRI/MRS techniques for the noninvasive assessment of organ and tissue metabolism in diabetes and related disorders; (2) FIRST 2007: this was a small grant from Università degli Studi di Milano for the study of the metabolic derangements in individuals with NAFLD; and (3) the European Association for the Study of Diabetes (EFSD): this grant is a given support to Gianluca Perseghin and is devoted to the study of alterations of cardiac metabolism in insulin-resistant subjects by means of the use of cardiac 31P-MRS. The work was independent of it, but the cardiac 31P-MRS methodology opened the possibility to apply the technique to the study of liver energy metabolism which was applied in this work.
- G. Marchesini, M. Brizi, A. M. Morselli-Labate et al., “Association of nonalcoholic fatty liver disease with insulin resistance,” The American Journal of Medicine, vol. 107, no. 5, pp. 450–455, 1999.
- G. Marchesini, E. Bugianesi, G. Forlani et al., “Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome,” Hepatology, vol. 37, no. 4, pp. 917–923, 2003.
- M. Roden, “Mechanisms of disease: hepatic steatosis in type 2 diabetes—pathogenesis and clinical relevance,” Nature Clinical Practice Endocrinology & Metabolism, vol. 2, no. 6, pp. 335–348, 2006.
- G. Marchesini, M. Brizi, G. Bianchi et al., “Nonalcoholic fatty liver disease: a feature of the metabolic syndrome,” Diabetes, vol. 50, no. 8, pp. 1844–1850, 2001.
- A. Gastaldelli, K. Cusi, M. Pettiti et al., “Relationship between hepatic/visceral fat and hepatic insulin resistance in nondiabetic and type 2 diabetic subjects,” Gastroenterology, vol. 133, no. 2, pp. 496–506, 2007.
- H. B. Holt, S. H. Wild, P. J. Wood et al., “Non-esterified fatty acid concentrations are independently associated with hepatic steatosis in obese subjects,” Diabetologia, vol. 49, no. 1, pp. 141–148, 2006.
- K. L. Donnelly, C. I. Smith, S. J. Schwarzenberg, J. Jessurun, M. D. Boldt, and E. J. Parks, “Sources of fatty acids stored in liver and secreted via lipoproteins in patients with nonalcoholic fatty liver disease,” The Journal of Clinical Investigation, vol. 115, no. 5, pp. 1343–1351, 2005.
- F. Diraison, P. Moulin, and M. Beylot, “Contribution of hepatic de novo lipogenesis and reesterification of plasma non esterified fatty acids to plasma triglyceride synthesis during non-alcoholic fatty liver disease,” Diabetes & Metabolism, vol. 29, no. 5, pp. 478–485, 2003.
- R. Belfort, S. A. Harrison, K. Brown et al., “A placebo-controlled trial of pioglitazone in subjects with nonalcoholic steatohepatitis,” The New England Journal of Medicine, vol. 355, no. 22, pp. 2297–2307, 2006.
- V. Ratziu, P. Giral, S. Jacqueminet et al., “Rosiglitazone for nonalcoholic steatohepatitis: one-year results of the randomized placebo-controlled Fatty Liver Improvement with Rosiglitazone Therapy (FLIRT) trial,” Gastroenterology, vol. 135, no. 1, pp. 100–110, 2008.
- G. Perseghin, P. Scifo, F. de Cobelli et al., “Intramyocellular triglyceride content is a determinant of in vivo insulin resistance in humans: a 1H-13C nuclear magnetic resonance spectroscopy assessment in offspring of type 2 diabetic parents,” Diabetes, vol. 48, no. 8, pp. 1600–1606, 1999.
- K. F. Petersen, S. Dufour, and G. I. Shulman, “Decreased insulin-stimulated ATP synthesis and phosphate transport in muscle of insulin-resistant offspring of type 2 diabetic parents,” PLoS Medicine, vol. 2, no. 9, article e233, 2005.
- G. Perseghin, G. Lattuada, F. de Cobelli et al., “Increased mediastinal fat and impaired left ventricular energy metabolism in young men with newly found fatty liver,” Hepatology, vol. 47, no. 1, pp. 51–58, 2008.
- R. Sharma, S. Sinha, K. A. Danishad et al., “Investigation of hepatic gluconeogenesis pathway in non-diabetic Asian Indians with non-alcoholic fatty liver disease using in vivo (31P) phosphorus magnetic resonance spectroscopy,” Atherosclerosis, vol. 203, no. 1, pp. 291–297, 2009.
- G. Perseghin, G. Lattuada, F. de Cobelli et al., “Habitual physical activity is associated with intrahepatic fat content in humans,” Diabetes Care, vol. 30, no. 3, pp. 683–688, 2007.
- B. Neuschwander-Tetri and S. H. Caldwell, “Nonalcoholic steatohepatitis: summary of an AASLD single topic conference,” Hepatology, vol. 37, no. 5, pp. 1202–1219, 2003.
- G. Perseghin, V. Mazzaferro, S. Benedini et al., “Resting energy expenditure in diabetic and nondiabetic patients with liver cirrhosis: relation with insulin sensitivity and effect of liver transplantation and immunosuppressive therapy,” The American Journal of Clinical Nutrition, vol. 76, no. 3, pp. 541–548, 2002.
- T. M. Wallace, J. C. Levy, and D. R. Matthews, “Use and abuse of HOMA modeling,” Diabetes Care, vol. 27, no. 6, pp. 1487–1495, 2004.
- P. Fiorina, G. Perseghin, F. de Cobelli et al., “Altered kidney graft high-energy phosphate metabolism in kidney-transplanted end-stage renal disease type 1 diabetic patients: a cross-sectional analysis of the effect of kidney alone and kidney-pancreas transplantation,” Diabetes Care, vol. 30, no. 3, pp. 597–603, 2007.
- J. B. V. Weir, “New methods for calculating metabolic rate with special reference to protein metabolism,” Journal of Physiology, vol. 109, no. 1-2, pp. 1–9, 1949.
- M. M. Johnson, R. Chin, and E. F. Haponik, “Nutrition, respiratory function, and disease,” in Modern Nutrition in Health and Disease, M. E. Shils, J. A. Olson, M. Shike, and A. C. Ross, Eds., pp. 1473–1490, Williams & Wilkins, 1999.
- K. N. Frayn, “Calculation of substrate oxidation rates in vivo from gaseous exchange,” Journal of Applied Physiology, vol. 55, no. 2, pp. 628–634, 1983.
- E. Bugianesi, A. Gastaldelli, E. Vanni et al., “Insulin resistance in non-diabetic patients with non-alcoholic fatty liver disease: sites and mechanisms,” Diabetologia, vol. 48, no. 4, pp. 634–642, 2005.
- H. Cortez-Pinto, J. Chatham, V. P. Chacko, C. Arnold, A. Rashid, and A. M. Diehl, “Alterations in liver ATP homeostasis in human non-alcoholic steatohepatitis. A pilot study,” JAMA, vol. 282, no. 17, pp. 1659–1664, 1999.
- S. Nair, V. P. Chacko, C. Arnold, and A. M. Diehl, “Hepatic ATP reserve and efficiency of replenishing: comparison between obese and nonobese normal individuals,” The American Journal of Gastroenterology, vol. 98, no. 2, pp. 466–470, 2003.
- A. K. P. Lim, N. Patel, G. Hamilton, J. V. Hajnal, R. D. Goldin, and S. D. Taylor-Robinson, “The relationship of in vivo 31P MR spectroscopy to histology in chronic hepatitis C,” Hepatology, vol. 37, no. 4, pp. 788–794, 2003.
- R. Jalan, J. Sargentoni, G. A. Coutts et al., “Hepatic phosphorus-31 magnetic resonance spectroscopy in primary biliary cirrhosis and its relation to prognostic models,” Gut, vol. 39, no. 1, pp. 141–146, 1996.
- G. Brinkmann, U. H. Melchert, L. Emde et al., “In vivo P-31-MR-spectroscopy of focal hepatic lesions: effectiveness of tumor detection in clinical practice and experimental studies of surface coil characteristics and localization technique,” Investigative Radiology, vol. 30, no. 1, pp. 56–63, 1995.
- K. F. Petersen, D. E. Befroy, S. Dufour, D. L. Rothman, and G. I. Shulman, “Assessment of hepatic mitochondrial oxidation and pyruvate cycling in NAFLD by 13C magnetic resonance spectroscopy,” Cell Metabolism, vol. 24, no. 1, pp. 167–171, 2016.
- M. Chmelík, A. I. Schmid, S. Gruber et al., “Three-dimensional high-resolution magnetic resonance spectroscopic imaging for absolute quantification of31P metabolites in human liver,” Magnetic Resonance in Medicine, vol. 60, no. 4, pp. 796–802, 2008.
Copyright © 2019 Guido Lattuada et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.