The spectrum of nonalcoholic fatty liver disease (NAFLD)

Transcription

1 Nonalcoholic Fatty Liver Disease as the Transducer of Hepatic Oversecretion of Very-Low-Density Lipoprotein Apolipoprotein B-100 in Obesity Dick C. Chan, Gerald F. Watts, SengKhee Gan, Annette T.Y. Wong, Esther M.M. Ooi, P. Hugh R. Barrett Objective To examine the association between liver fat content and very low-density lipoprotein (VLDL) apolipoprotein (apo) B-100 kinetics and the corresponding responses to weight loss in obese subjects. Methods and Results VLDL apob-100 kinetics were assessed using stable isotope tracers, and the fat content of the liver and abdomen was determined by magnetic resonance techniques in 25 obese subjects. In univariate analysis, liver fat content was significantly (P 0.05 in all) associated with body mass index (r 0.65), visceral fat area (r 0.45), triglycerides (r 0.40), homeostasis model assessment score (r 0.40), VLDL apob-100 concentrations (r 0.44), and secretion rate (r 0.45). However, liver fat content was not associated with plasma concentrations of retinol-binding protein 4, fetuin A, adiponectin, interleukin-6, and tumor necrosis factor-. Of these 25 subjects, 9 diagnosed as having nonalcoholic fatty liver disease (which is highly prevalent in obese individuals and strongly associated with dyslipidemia) underwent a weight loss program. The low-fat diet achieved significant reduction in body weight, body mass index, liver fat, visceral and subcutaneous fat areas, homeostasis model assessment score, triglycerides, VLDL apob-100 concentrations, and VLDL apob-100 secretion rate. The percentage reduction of liver fat with weight loss was significantly associated with the corresponding decreases in VLDL apob-100 secretion (r 0.67) and visceral fat (r 0.84). Conclusion In patients with obesity, hepatic steatosis increases VLDL apob-100 secretion. Weight loss can help reduce this abnormality. (Arterioscler Thromb Vasc Biol. 2010;30: ) Key Words: obesity liver fat lipoprotein metabolism weight loss cardiovascular disease The spectrum of nonalcoholic fatty liver disease (NAFLD) includes benign lipid accumulation in the liver (simple steatosis) through to steatohepatitis, fibrosis, and cirrhosis. 1 The prevalence of NAFLD in the general population is up to 30%; however, it is much higher among individuals with type 2 diabetes mellitus and obesity, 2 key features of metabolic syndrome. 2 NAFLD is strongly associated with dyslipidemia, particularly hypertriglyceridemia, which may contribute to the increased risk of cardiovascular disease in these subjects. 3 The underlying mechanism for hypertriglyceridemia in obesity chiefly relates to the oversecretion of triglyceride-rich lipoproteins, particularly very low-density lipoprotein (VLDL) particles. 4 These abnormalities may contribute to central obesity and insulin resistance. Hepatic accumulation of lipids (ie, triglyceride and cholesteryl ester) may perturb VLDL apolipoprotein (apo) B-100 metabolism in insulin-resistant obese subjects, sufficient to account for hypertriglyceridemia. However, the precise mechanisms responsible for the close relationship between NAFLD and metabolic syndrome remain uncertain. Visceral fat accumulation markedly increases the flux of free fatty acids (FFAs) in the portal vein to the liver. This stimulates hepatic gluconeogenesis and triglyceride synthesis; it also impairs hepatic extraction of insulin. 5 Hepatic insulin resistance also attenuates hepatic lipogenesis and gluconeogenesis, thereby increasing triglyceride accumulation and subsequent VLDL secretion. 6 The accumulation of intrahepatic fat per se also stimulates VLDL secretion by providing a source of fatty acids for the packaging and processing of VLDL. 7 However, a direct link between intrahepatic fat and VLDL apob-100 metabolism has not yet been conclusively demonstrated in patients with obesity. Recently, there has been intense interest in the potential impact of adipocytokines (eg, adiponectin and retinol-binding protein 4 [RBP-4]) and other inflammatory mediators on glucose and lipid metabolism and their contribution to the pathogenesis of NAFLD and hepatic inflammation. 2,8 However, the precise role of these mediators in modulating the Received on: October 7, 2009; final version accepted on: January 26, From the Metabolic Research Centre (D.C.C., G.F.W., S.G., A.T.Y.W., E.M.M.O., and P.H.R.B.), School of Medicine and Pharmacology, University of Western Australia, Perth. Correspondence to Gerald F. Watts, School of Medicine and Pharmacology, University of Western Australia, Royal Perth Hospital, GPO Box X2213, Perth, Western Australia 6847, Australia. gerald.watts@uwa.edu.au 2010 American Heart Association, Inc. Arterioscler Thromb Vasc Biol is available at DOI: /ATVBAHA

2 1044 Arterioscler Thromb Vasc Biol May 2010 association between intrahepatic fat and VLDL apob-100 metabolism remains to be elucidated. In the present study, we used a stable isotope technique to test the hypothesis that increased intrahepatic fat content is associated with oversecretion of VLDL apob-100 in subjects with visceral obesity. We also explored these relationships before and after a period of moderate weight loss. Methods Subjects A total of 25 nonsmoking, centrally obese, white subjects (15 men and 10 women) consuming weight-maintaining diets ad libitum were recruited for the study. None of the subjects had diabetes mellitus (excluded by an oral glucose tolerance test result), the apo E2/E2 genotype, macroproteinuria, creatinemia ( 120 mol/l), hypothyroidism, or abnormal liver enzyme levels. No subject consumed more than 20 g/d of alcohol. There was also no report of a history of cardiovascular disease. Finally, no subject was taking medication or other agents known to affect lipid metabolism. The study was approved by the Ethics Committee of Royal Perth Hospital, and informed consent was obtained before the study was started. Clinical Protocols All subjects were admitted to the metabolic ward in the morning after a 14-hour fast. They were studied in a semirecumbent position and allowed to drink only water. A venous blood sample was collected for measurements of biochemical analytes. Arterial blood pressure was recorded after 3 minutes in the supine position using a monitor (Dinamap1846 SX/P; Critikon Inc, Tampa, Fla). Dietary intake was assessed for energy and major nutrients using at least two 24-hour dietary diaries and subsequently analyzed using computer software (DIET 4 Nutrient Calculation Software; Xyris Software, Queensland, Australia). Body composition was estimated at rest in the supine position using an analyzer (Holtain Body Composition Analyser; Holtain Ltd, Dyfed, Wales) from which total body fat and fat-free mass (FFM) were derived; FFM was calculated using the following formula by Lukaski et al 9 : FFM (0.85 [H 2 /Z]) 3.04, where H is height (in centimeters) and Z is impedance. For this measure, subjects were asked to fast overnight and to refrain from alcoholic beverages for 24 hours; they were then studied in the morning, after emptying their bladder, in a temperature-controlled room. The technical error for FFM was less than 3%, calculated from 3 repeated measurements by the same operator. A single bolus of deuterated leucine (d3-leucine), 5 mg/kg of body weight, was administered intravenously within a 2-minute period into an antecubital vein via a 21G butterfly needle. Blood samples were taken at baseline and after injection of the isotope at 5, 10, 20, 30, and 40 minutes and at 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0, 6.0, 8.0, and 10.0 hours. Subjects were then given a snack and allowed to go home. Additional fasting blood samples were collected in the morning on the following 4 days of the same week (24, 48, 72, and 96 hours). Liver and Abdominal Fat Measurements Proton magnetic resonance spectroscopy was performed with a Sonata 1.5T system (Siemens, Bayswater, Australia). Sequences and quantification procedures for liver fat were described previously. 10 Briefly, a single-volume stimulated echo acquisition mode sequence was used, with parameters for repetition time (1500 milliseconds), echo time (15 milliseconds), voxel size (4 3 4 cm), 7 signal averages, breath holding, and no water saturation; voxels were positioned to avoid hepatic or portal vessels. The manufacturer s magnetic resonance spectroscopy processing software was used to obtain the liver fat to liver water signal ratio, which was converted to liver fat percentage, assuming that the total liver signal equals the fat signal plus the water signal. Magnetic resonance imaging of 8 transaxial segments (field of view, cm; 10-mm thickness) at intervertebral disc levels from T11 to S1 was performed using a 1.0-T scanner (Picker MR; Picker International, Cleveland, Ohio) and a T1-weighted fast-spin echo sequence with a high fat to water signal ratio. Visceral and subcutaneous abdominal adipose tissue areas were measured at vertebra L3. Diet Intervention A total of 9 subjects (6 men and 3 women) were selected at random from those with NAFLD, defined as a liver fat content of more than 5%. 11 They entered a weight-reducing hypocaloric diet (low-fat and low-calorie) program for 16 weeks, immediately followed by a 6-week weight stabilization period. For the first 16 weeks of weight loss, a dietitian provided dietary counseling based on calculated basal energy expenditure, from which 2500 kj was subtracted to estimate the required dietary energy intake. We attempted to maintain the composition of the diet during the end of the study, when the weight stabilization period was similar to that in the run-in phase. All kinetic and imaging procedures were repeated after interventions. Measurements of VLDL apo-b100 Enrichments and Calculation of Kinetic Parameters VLDL was isolated from 2 ml of plasma by ultracentrifugation (Optima XL-100K; Beckman Coulter, Sydney, NSW, Australia) at a density of The procedures for isopropanol precipitation, delipidation, hydrolysis, and derivatization of leucine to the oxazolinone derivative were previously described. 12 Plasma-free leucine was also isolated by cation-exchange chromatography using resin (AG 50 W-X8; BioRad, Richmond, Calif) after removing plasma proteins with 60% perchloric acid. Isotopic enrichment was determined using gas chromatography mass spectrometry, with selected ion monitoring of samples at a mass to charge ratio of 212 (derived from d 3 -leucine) and 209 (derived from unlabeled leucine) and negative ion chemical ionization. Tracer to tracee ratios were derived from isotopic ratios for each sample. The isotopic enrichment of VLDL apob-100 after infusion of d 3 -leucine before and after weight loss is shown in Appendix 1. The multicompartmental model used to describe VLDL apob100 leucine tracer to tracee ratios were previously described (Appendix 2). 13 The SAAM II program (SAAM Institute, Seattle, Wash) was used to fit the model to the observed tracer to tracee ratios. The fractional catabolic rate (FCR) of VLDL apob-100 was derived from the model parameters giving the best fit. Plasma volume was determined after adjusting for the decrease in relative plasma volume associated with an increase in body weight. 14,15 The VLDL apob-100 secretion rate was calculated as the product of VLDL apob-100 FCR and the corresponding pool size, and expressed as milligrams per kilogram per day. Biochemical Analytes All biochemical analyses were measured in samples obtained at baseline. Plasma cholesterol and triglyceride concentrations were determined by standard enzymatic methods (Hitachi, Tokyo, Japan; Roche Diagnostic GmbH, Mannheim, Germany). The high-density lipoprotein cholesterol level was measured by an enzymatic calorimetric method using a commercial kit (Boehringer Mannheim, Mannheim). The LDL cholesterol level was calculated by the Friedewald calculation. Plasma nonesterified fatty acids were measured by an enzymatic colorimetric assay (Wako Pure Chemical, Osaka, Japan). The VLDL apob100 concentration was measured using a modification of the method described by Beghin et al. 16 Plasma total apoa-i and apob-100 concentrations were determined by immunonephelometry (Behring Diagnostics, Kingsgrove, Australia). The plasma insulin level was measured by a solid-phase 2-site sequential chemiluminescent immunometric assay (Diagnostic Products Corporation, Los Angeles, Calif) and a glucose-by-hexokinase method (Hitachi). Insulin resistance was estimated by the homeostasis model assessment (HOMA) score (ie, fasting insulin level [mu/l] fasting glucose level [mmol/l]/22.5). 17 Plasma RBP-4, fetuin A, adiponectin, interleukin 6, and tumor necrosis factor concentrations were determined using enzyme immunoassay kits (Quantikine, Research & Development Systems, Minneapolis,

3 Chan et al Liver Fat and VLDL Metabolism in Obesity 1045 Table 1. Anthropometric and Biochemical Characteristics and VLDL apob-100 Kinetic Parameters of the 25 Subjects Studied Characteristic Value, Mean SD (Range) Age, y 57 8 (42 74) Weight, kg (74 127) Body mass index* ( ) Fat Visceral, cm (63 458) Subcutaneous, cm (72 510) Total body, kg (20 73) Fat-free mass, kg (37 75) Liver fat content, % ( ) Nonesterified fatty acid, meq/l ( ) Cholesterol, mmol/l ( ) Triglycerides, mmol/l ( ) Cholesterol, mmol/l HDL ( ) LDL ( ) ApoA-I, g/l ( ) ApoB-100, g/l ( ) Lathosterol, mol/l ( ) Campesterol, mol/l ( ) Glucose, mmol/l ( ) Insulin, U/L (5 25) HOMA score ( ) Retinol-binding protein 4, mg/l (13 42) Fetuin A, mg/l ( ) Adiponectin, mg/l ( ) Interleukin 6, ng/l ( ) TNF-, ng/l ( ) VLDL apob-100 Concentration, mg/l (49 240) Fractional catabolic rate, pools/d ( ) Secretion rate, mg/kg/d ( ) Apo indicates apolipoprotein; HDL, high-density lipoprotein; HOMA, homeostasis model assessment; LDL, low-density lipoprotein; TNF, tumor necrosis factor; VLDL, very low-density lipoprotein. *Calculated as weight in kilograms divided by height in meters squared. Minn). Plasma lathosterol and campesterol concentrations were assayed using gas chromatography mass spectrometry. Genomic DNA was extracted from whole blood, and apoe genotypes were determined using the TaqMan assay on PRISM 7000 sequence detection analyzer (Applied Biosystems, Mulgrave, Victoria, Australia). All statistical analyses were performed using software (SPSS 15; SPSS Inc, Chicago, Ill). Associations were examined by simple and multivariate linear regression methods. The paired t test was used to determine the impact of weight loss treatment. Statistical significance was defined at the 5% level using a 2-tailed test. Results The anthropometric and biochemical characteristics of the 25 middle-aged and normotensive subjects are summarized in Table 1. Twenty-one subjects had NAFLD as a diagnosis of hepatic steatosis (intrahepatic triglyceride content 5% of Table 2. Associations (Pearson Correlation Coefficients) of Liver Fat Content With the Clinical and Biochemical Characteristics and the Kinetic Indexes for VLDL apob-100 in the Subjects Studied Characteristic Pearson Correlation Coefficient Age 0.08 Weight 0.70* Body mass index 0.65* Fat Visceral 0.45 Subcutaneous 0.29 Total body 0.72* Fat-free mass 0.03 Energy intake 0.41 Fat intake 0.42 Nonesterified fatty acid 0.03 Cholesterol Triglyceride 0.40 Cholesterol HDL 0.15 LDL 0.03 ApoA-I 0.10 ApoB Lathosterol 0.17 Campesterol 0.27 Glucose 0.27 Insulin 0.38 HOMA score 0.40 Alanine aminotransferase 0.51* Retinol-binding protein Fetuin A 0.18 Adiponectin 0.05 Interleukin TNF VLDL apob-100 Concentration 0.44 Fractional catabolic rate 0.15 Secretion rate 0.47 Apo indicates apolipoprotein; HDL, high-density lipoprotein; HOMA, homeostasis model assessment; LDL, low-density lipoprotein; TNF, tumor necrosis factor; VLDL, very low-density lipoprotein. *P P the liver volume). Compared with 15 (9 men and 6 women) nonobese age- and sex-matched subjects, given as mean SD (aged years) selected from a previous report, 18 our obese subjects were insulin resistant and mildly dyslipidemic. The 25 obese subjects had a significantly (P 0.05 for all) higher VLDL apob-100 concentration and secretion rate and a reduced FCR (Appendix 3). The average daily energy and nutrient intake was as follows: kj, 37 6% energy from fat, 39 5% energy from carbohydrates, 21 4% energy from protein, and 3 3% energy from alcohol.

4 1046 Arterioscler Thromb Vasc Biol May rate 25 secretion /day) VLDL-a pob-100 (m g/kg r= 0.45 P<0.05 Figure 1. Association between the VLDL apob-100 secretion rate and liver fat content in 25 obese subjects Liver fat content (%) Correlational Analysis Table 2 shows the correlation between liver fat content and the anthropometric and biochemical characteristics in the subjects studied. In a univariate analysis, liver fat content was significantly (P 0.05 for all) and positively associated with body mass index, visceral fat area at vertebra L3, total body fat, energy intake, fat intake, plasma triglycerides, HOMA score, alanine aminotransferase level, and VLDL apob-100 concentrations. As shown in Figure 1, liver fat content was also significantly associated with the VLDL apob-100 secretion rate (r 0.45, P 0.02). The association between liver fat content and VLDL apob-100 secretion rate remained significant (P 0.05 for both) after adjusting for dietary fat or carbohydrate intake (r 0.45 or r 0.44, respectively). Liver fat content was not associated with plasma concentrations of RPB-4, fetuin A, adiponectin, interleukin 6, and tumor necrosis factor concentrations. In a multiple regression analysis including visceral fat and HOMA score, liver fat content was an independent predictor of the VLDL apob- 100 secretion rate ( coefficient, 0.53; P 0.02). Inclusion of dietary fat or carbohydrate intake instead of visceral fat did not alter the significance of the association between liver fat content and VLDL apob-100 secretion rate ( coefficient, 0.57; P 0.006). The HOMA score was not significantly (P 0.05 for both) associated with the VLDL apob-100 secretion rate (r 0.04) and the VLDL apob-100 FCR (r 0.27). Similarly, neither fat nor carbohydrate intake was associated with the VLDL apob-100 secretion rate (r 0.13 and r 0.28, respectively) and the VLDL apob-100 FCR (r 0.26 and r 0.23, respectively). Weight Loss Table 3 shows the dietary composition and nutrient intake of subjects during the study. As noted, the weight loss group participants significantly reduced their total energy and fat and their saturated fat intake, and significantly increased their carbohydrate consumption, during the active weight loss period. Table 4 shows the clinical and biochemical characteristics and the kinetic indexes for VLDL apob-100 metabolism after a 16-week dietary intervention in 9 obese subjects with NAFLD. The low-fat diet achieved significant reduction (P 0.05) in body weight ( 6%), body mass index ( 6%), and liver ( 29%), visceral ( 18%), and subcutaneous ( 10%) fat areas; total body fat ( 13%); HOMA score ( 15%); triglyceride ( 22%), RBP-4 ( 14%), fetuin ( 40%), and VLDL apob- 100 ( 22%) concentrations; and the VLDL apob-100 secretion rate ( 23%). There was a significant increase (P 0.05) in plasma adiponectin concentration (16%). The percentage reduction of liver fat with weight loss was significantly associated with the corresponding changes in the VLDL apob-100 secretion rate (r 0.67) (Figure 2) and the visceral fat concentration (r 0.84). There was no significant association (P 0.05 for all) between the percentage reduction of Table 3. Dietary Composition and Nutrient Intake in the Weight Reduction Group* Weight Reduction Group (n 20) Variable 0 wk (Baseline) 8 wk wk Total energy, kj Fat, % energy Saturated fat, g Carbohydrate, % energy Protein, % energy Alcohol, % energy Fiber, g *Data are given as the mean SEM. Differences at 8 weeks and after weight stabilization (16 22 weeks) relative to baseline were analyzed using paired t tests. P

5 Chan et al Liver Fat and VLDL Metabolism in Obesity 1047 Table 4. Changes in Anthropometric and Biochemical Characteristics and in VLDL apob-100 Kinetics in the Subjects Before and After Weight Loss Characteristic Before Weight Loss* After Weight Loss* Change (SD) P Value Weight, kg (4.2) Body mass index (1.3) Fat Visceral, cm (66.0) 0.04 Subcutaneous, cm (41.0) 0.04 Total body, kg (3.7) Fat-free mass, kg (2.3) 0.21 Liver fat content, % (5.2) Nonesterified fatty acid, meq/l (0.23) 0.15 Cholesterol, mmol/l (0.73) 0.35 Triglyceride, mmol/l (0.37) 0.02 Cholesterol, mmol/l HDL (0.24) 0.94 LDL (0.60) 0.72 ApoA-I, g/l (0.16) 0.48 ApoB-100, g/l (0.21) 0.17 Lathosterol, mol/l (1.75) Campesterol, mol/l (2.70) 0.57 Glucose, mmol/l (0.34) 0.16 Insulin, U/L (2.36) 0.03 HOMA score (0.55) 0.03 Retinol-binding protein 4, mg/l (3.7) 0.02 Fetuin A, mg/l (113) 0.01 Adiponectin, mg/l (1.20) 0.05 Interleukin 6, ng/l (0.32) 0.64 TNF-, ng/l (0.50) 0.37 VLDL apob-100 Concentration, mg/l (20) 0.02 Fractional catabolic rate, pools/d (2.80) 0.36 Secretion rate, mg/kg/d (6.6) 0.03 Apo indicates apolipoprotein; HDL, high-density lipoprotein; HOMA, homeostasis model assessment; LDL, low-density lipoprotein; TNF, tumor necrosis factor; VLDL, very low-density lipoprotein. *Data are given as mean SEM. Significant differences before and after treatment were compared using paired t tests. Calculated as weight in kilograms divided by height in meters squared. VLDL apob-100 secretion rate and the changes in energy (r 0.11) and fat (r 0.32) intake, visceral fat concentration (r 0.41), HOMA score (r 0.22), or lathosterol concentration (r 0.14). The percentage decline in liver fat content just failed to be significantly associated with the corresponding reduction in the VLDL apob-100 secretion rate after adjusting for corresponding changes in HOMA score (r 0.70, P 0.05), dietary fat intake (r 0.66, P 0.07), and lathosterol (r 0.67, P 0.07) or visceral fat (r 0.66, P 0.08) concentration. The partial correlation coefficients for changes in VLDL apob-100 secretion rate with changes in HOMA score, dietary fat intake, and lathosterol and visceral fat concentration were as follows: 0.35 (P 0.40), 0.29 (P 0.48), 0.09 (P 0.83), and 0.38 (P 0.35), respectively. As expected, the change in VLDL apob-100 FCR was not associated with the corresponding changes in visceral fat (r 0.16, P 0.70), subcutaneous fat (r 0.21, P 0.62), HOMA score (r 0.44, P 0.28), liver fat (r 0.44, P 0.44), lathosterol concentration (r 0.45, P 0.27), fat intake (r 0.09, P 0.84), and adiponectin (r 0.33, P 0.42), RBP-4 (r 0.56, P 0.14), or fetuin A (r 0.22, P 0.59) level. Discussion To our knowledge, we present new evidence in subjects with visceral obesity that elevated liver fat content was predictive of VLDL apob-100 oversecretion, and that the reduction of liver fat with weight loss, was significantly associated with a corresponding decrease in VLDL apob-100 secretion. Few studies have examined the association between liver fat content and VLDL apob-100 metabolism in obese subjects. Adiels et al 19 reported that liver fat content was significantly associated with the VLDL 1 apob-100 secretion

6 1048 Arterioscler Thromb Vasc Biol May ion rate 100 secreti ay) hange in VLDL-apoB- (mg/kg/da % c r=0.67 P< % change in liver fat content Figure 2. Association between changes in the VLDL apob-100 secretion rate and liver fat content with weight loss. rate in nonobese men. In that study, no significant association was found between liver fat content and VLDL apob-100 secretion rate in 10 type 2 diabetic men who were treated with diet and/or antidiabetic medication. The insulin-sensitizing effect of such diabetic treatments might have confounded the association between liver fat content and VLDL apob-100 secretion in these subjects. Fabbrini et al 20 found that liver fat content was significantly associated with VLDL-triglyceride secretion, but not with VLDL apob-100 secretion, in obese subjects, most of whom were women; however, they did not study these associations after weight loss. In contrast, we found that liver fat content was significantly associated with VLDL apob-100 secretion; however, we mostly studied obese men and the effects of weight loss on hepatic fat and VLDL apob-100 secretion. In women, the liver secretes less but more triglyceride-enriched VLDL particles than men. 21 There could be a sex difference in the impact of hepatic fat content on the relative secretion of VLDL apob-100 and triglycerides by the liver; this requires further investigation. The liver fat content reflects the balance between FFA flux to the liver, de novo lipogenesis, and diet on the one hand and fatty acid oxidation and triglyceride export on the other hand. 7 Intra-abdominal adipocytes are lipolytically active, and visceral fat accumulation in individuals with central obesity results in markedly increased flux of FFAs in the portal vein to the liver; this can stimulate de novo lipogenesis. 5 Hepatic insulin resistance also upregulates the expression of sterol regulatory element-binding protein 1c and leads to activation of key enzymes for lipogenesis. 6,22 The increased uptake of dietary cholesterol not only activates liver X receptor, but also the gene expression for de novo lipogenesis. 22 The net effect of these processes is increased accumulation of fat in the liver, which may subsequently drive the oversecretion of VLDL apob-100. Consistent with this notion, we found that liver fat content was significantly associated with visceral fat, insulin resistance, dietary fat (or total energy) intake, and the VLDL apob-100 secretion rate. A prolonged fasting state ( 36 hours) could potentially increase FFA release from adipose tissue and delivery to the liver, thereby stimulating triglyceride synthesis and affecting hepatic fat and VLDL apob-100 secretion. 23 However, there is evidence that VLDL-triglyceride and VLDL apob-100 concentrations remain steady within a 20-hour period of fasting, 24 implying that our intermediate term of fasting used in the stable isotope study does not affect hepatic fat content or assembly and secretion of VLDL apob-100. We did not observe a significant association between liver fat content and plasma nonesterified fatty acid concentration, suggesting that systemic nonesterified fatty acid concentration may not necessarily reflect portal FFA flux to the liver. Also, we could not fully exclude the possibility that the observed association between liver fat content and VLDL apob-100 secretion is secondary to the overall effects of visceral fat, insulin resistance, or increased dietary fat intake on VLDL transport. However, we found that liver fat content remained a significant predictor of the VLDL apob-100 secretion rate in regression models that included these metabolic and dietary variables. These data suggest that liver fat content could play an important role in trafficking various sources of fatty acids for the packaging and processing of VLDL. Interestingly, Charlton et al 25 found that subjects with nonalcoholic steatohepatitis, an advanced form of NAFLD, had reduced apob- 100 secretion. Although the mechanism is unknown, the researchers concluded that reduced apob synthesis is a determinant of the development of hepatic steatosis. However, it is possible that changes in the structural integrity and chemical function of hepatocytes in subjects with nonalcoholic steatohepatitis are likely to interfere with the apob secretory pathway. By increasing hepatic endoplasmic reticulum stress, increased FFA supply to the liver may paradoxically decrease apob-100 secretion 26 ; however, the contribution of this mechanism to our principal findings in subject with mild steatosis remains conjectural. The development of inflammation and fibrosis in NAFLD mainly involves 2 stages, the so-called 2-hit hypothesis. 27 Hepatic fat accumulation represents the first hit ; and insulin resistance, oxidative stress, and increased cytokine release induce the second hit. RBP-4, fetuin-a, and adipocytokines have all been implicated in the second hit for the development of inflammation and fibrosis in subjects with NAFLD.

7 Chan et al Liver Fat and VLDL Metabolism in Obesity 1049 However, none of these markers was associated with liver fat content in our study. The reason for this remains unclear. It is likely that our nondiabetic obese subjects are in the early stage of NAFLD and that the impact of these metabolic factors might be secondary to other factors, such as obesity, dietary intake, and insulin resistance. Weight loss by dietary restriction is the most common approach to correct NAFLD and dyslipidemia. 28 It has 3 potential effects that may reduce the hepatic availability of fatty acids for triglyceride storage and export. Weight loss reduces the release of FFAs from visceral fat to the liver. It also improves insulin sensitivity, thereby potentially impeding de novo lipogenesis. Consistent with the notion, we found that weight loss effectively reduces visceral fat and improves insulin sensitivity, leading to a significant reduction in liver fat content. Previously, in obese men, we demonstrated that intensive weight loss ( 10% of body weight) with a low-fat and low-calorie diet resulted in a greater reduction in visceral fat mass ( 27%) and hepatic VLDL apob-100 secretion ( 50%) than in the present study; the change in visceral fat was also significantly correlated with the change in hepatic VLDL apob-100 secretion. 29 We did not confirm this association, probably owing to differences in the degree of weight loss between the 2 studies. In this study, we also specifically selected subjects with NAFLD and tested the association between liver fat content and VLDL apob-100 secretion with weight loss. We only observed a significant relationship between changes in VLDL apob-100 secretion and liver fat content, but not changes in visceral fat, insulin resistance, or dietary intake. These data further support a more direct role for liver fat content in the regulation of VLDL transport, consistent with our earlier correlation findings. The existence of a weaker correlation between changes in liver fat content and the VLDL apob-100 secretion rate after adjusting for changes in visceral fat, insulin resistance, lathosterol, or dietary fat intake probably reflects the impact of these metabolic and dietary factors on VLDL transport, consistent with a complex relationship among metabolic risk factors in patients with NAFLD. Despite our findings that weight loss reduced VLDL apob-100 FCR by 20%, the difference was not significant (P 0.36). The lack of effect of weight loss on VLDL apob-100 FCR is consistent with previous studies. 29,30 Changes in body fat, dietary intake, visceral fat, HOMA score, and adipocytokines were also not correlated with VLDL apob-100 FCR. Our study does have limitations. Our weight loss study was not a controlled observation. However, it provides evidence to support the direct relationship between liver fat content and the VLDL apob-100 secretion rate. As in previous protocols, 29,30 we did not use a control dietary period before the first isotope study to avoid alternating the nutritional steady state. On this ad libitum diet, there were no significant associations at stable body weight at baseline between either VLDL apob-100 secretion or liver fat content and nutrient intake. We did not perform liver biopsies and, therefore, cannot make inferences concerning steatohepatitis before and after weight loss; however, the short-term duration of weight loss is unlikely to influence hepatic fibrosis. We did not measure plasma volume directly, which could affect the calculation of the VLDL apob-100 secretion rate. However, our method for estimating plasma volume from body weight has been well validated against more direct methods for measuring plasma volume. 14,15 The sample size in this study was small. Therefore, the findings need to be confirmed in a larger sample size. Our kinetic findings could be clinically important. NAFLD is a common feature of lipid disorders, seen in insulin-resistant and centrally obese subjects. Our kinetic studies of VLDL apob-100 metabolism provide new knowledge of the underlying mechanisms of how liver fat is responsible for hypertriglyceridemia in central obesity. Our data support the role of liver fat as the driver of VLDL apob-100 secretion and highlight the importance of therapies that regulates NAFLD in the management of dyslipidemia and cardiovascular risk. We found that a reduction in liver fat with a moderate weight reduction may account for reduced VLDL apob-100 secretion and plasma concentration. These effects appear to be independent of inflammatory mediators of NAFLD. Given that liver fat content may also play an important role in other triglyceride-rich lipoprotein pathways (such as fatty acid and VLDL-triglyceride metabolism), further studies should examine these pathways and explore the effect of other agents (such as fibrates, fish oils, insulin sensitizers, and cholesterol absorption inhibitors) alone or added to weight loss on liver fat content and the kinetics of triglyceride-rich lipoproteins in these subjects. Sources of Funding This study was supported by research grants from the National Health and Medical Research Council (NHMRC) and the National Heart Foundation of Australia (NHF). Dr Chan is a Career Development Fellow of the NHMRC, Dr Ooi is an NHF postdoctoral fellow, and Prof Barrett is an NHMRC Senior Research Fellow. None. Disclosures References 1. Angulo P. Nonalcoholic fatty liver disease. N Engl J Med. 2002;346: Kotronen A, Yki-Järvinen H. Fatty liver: a novel component of the metabolic syndrome. Arterioscler Thromb Vasc Biol. 2008;28: Targher G, Arcaro G. Non-alcoholic fatty liver disease and increased risk of cardiovascular disease. Atherosclerosis. 2007;191: Chan DC, Barrett PH, Watts GF. Lipoprotein transport in the metabolic syndrome: pathophysiological and interventional studies employing stable isotopy and modelling methods. Clin Sci. 2004;107: Wajchenberg BL. Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocrine Rev. 2000;21: Ginsberg HN, Huang LS. The insulin resistance syndrome: impact on lipoprotein metabolism and atherothrombosis. J Cardiovasc Risk. 2000; 7: Goldberg IJ, Ginsberg HN. Ins and outs modulating hepatic triglyceride and development of nonalcoholic fatty liver disease. Gastroenterology. 2006;130: Tilg H, Diehl AM. Cytokines in alcoholic and nonalcoholic steatohepatitis. N Engl J Med. 2000;343: Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Assessment of fat-free mass using bioelectrical impedance measurements of the human body. Am J Clin Nutr. 1985;41: Chan DC, Watts GF, Ng TW, Hua J, Song S, Barrett PHR. Measurement of liver fat by magnetic resonance imaging: relationships with body fat distribution, insulin sensitivity and plasma lipids in healthy men. Diabetes Obes Metab. 2006;8:

8 1050 Arterioscler Thromb Vasc Biol May Szczepaniak LS, Nurenberg P, Leonard D, Browning JD, Reingold JS, Grundy S, Hobbs HH, Dobbins RL. Magnetic resonance spectroscopy to measure hepatic triglyceride content: prevalence of hepatic steatosis in the general population. Am J Physiol Endocrinol Metab. 2005;288: Dwyer KP, Barrett PH, Chan D, Foo JI, Watts GF, Croft KD. Oxazolinone derivative of leucine for GC-MS: a sensitive and robust method for stable isotope kinetic studies of lipoproteins. J Lipid Res. 2002;43: Chan DC, Watts GF, Nguyen MN, Barrett PHR. Apolipoproteins C-III and A-V as predictors of very-low-density lipoprotein triglyceride and apolipoprotein B-100 kinetics. Arterioscler Thromb Vasc Biol. 2006;26: Retzlaff JA, Tauxe WN, Kiely JM, Stroebel CF. Erythrocyte volume, plasma volume, and lean body mass in adult men and women. Blood. 1969;33: Feldschuh J, Enson Y. Prediction of the normal blood volume: relation of blood volume to body habitus. Circulation. 1977;56: Beghin L, Duhal N, Poulain P, Hauw P, Lacroix B, Lecerf JM, Bonte JP, Fruchart JC, Luc G. Measurement of apolipoprotein B concentration in plasma lipoproteins by combining selective precipitation and mass spectrometry. J Lipid Res. 2000;41: Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28: Chan DT, Dogra GK, Irish AB, Ooi EM, Barrett PHR, Chan DC, Watts GF. Chronic kidney disease delays VLDL apob-100 particle catabolism: potential role of apo C-III. J Lipid Res. 2009;50: Adiels M, Taskinen MR, Packard C, Caslake MJ, Soro-Paavonen A, Westerbacka J, Vehkavaara S, Hakkinen A, Olofsson SO, Yki-Jarvinen H, Boren J. Overproduction of large VLDL particles is driven by increased liver fat content in man. Diabetologia. 2006;49: Fabbrini E, Mohammed BS, Magkos F, Korenblat KM, Patterson BW, Klein S. Alterations in adipose tissue and hepatic lipid kinetics in obese men and women with nonalcoholic fatty liver disease. Gastroenterology. 2008;134: Magkos F, Patterson BW, Mohammed BS, Klein S, Mittendorfer B. Women produce fewer but triglyceride-richer very low-density lipoproteins than men. J Clin Endocrinol Metab. 2007;92: Horton JD, Goldstein JL, Brown MS. SREBPs: activators of the complete program of cholesterol and fatty acid synthesis in the liver. J Clin Invest. 2002;109: Moller L, Stodkilde-Jorgensen H, Jensen FT, Jorgensen JO. Fasting in healthy subjects is associated with intrahepatic accumulation of lipids as assessed by 1H-magnetic resonance spectroscopy. Clin Sci. 2008;114: Magkos F, Patterson BW, Mittendorfer B. Reproducibility of stable isotope-labeled tracer measures of VLDL-triglyceride and VLDLapolipoprotein B-100 kinetics. J Lipid Res. 2007;48: Charlton M, Sreekumar R, Rasmussen D, Lindor K, Nair KS. Apolipoprotein synthesis in nonalcoholic steatohepatitis. Hepatology. 2002;35: Ota T, Gayet C, Ginsberg HN. Inhibition of apolipoprotein B100 secretion by lipid-induced hepatic endoplasmic reticulum stress in rodents. J Clin Invest. 2008;118: Day CP, James OF. Steatohepatitis: a tale of two hits? Gastroenterology. 1998;114: Ahmed MH, Byrne CD. Current treatment of non-alcoholic fatty liver disease. Diabetes Obes Metab. 2009;11: Riches FM, Watts GF, Hua J, Stewart GR, Naoumova RP, Barrett PHR. Reduction in visceral adipose tissue is associated with improvement in apolipoprotein B-100 metabolism in obese men. J Clin Endocrinol Metab. 1999;84: Ng TWK, Watts GF, Barrett PHR, Rye KA, Chan DC. Effect of weight loss on LDL and HDL kinetics in the metabolic syndrome: associations with changes in plasma retinol-binding protein-4 and adiponectin levels. Diabetes Care. 2007;30: