Background & Aims: Nonalcoholic fatty liver disease is associated with insulin resistance and diabetes. The purpose of this study was to determine the relationship between intrahepatic triglyceride (IHTG) content and insulin action in liver (suppression of glucose production), skeletal muscle (stimulation of glucose uptake), and adipose tissue (suppression of lipolysis) in nondiabetic obese subjects. Methods: A euglycemic-hyperinsulinemic clamp procedure and stable isotopically labeled tracer infusions were used to assess insulin action, and magnetic resonance spectroscopy was used to determine IHTG content, in 42 nondiabetic obese subjects (body mass index, 36 ± 4 kg/m2) who had a wide range of IHTG content (1%–46%). Results: Hepatic insulin sensitivity, assessed as a function of glucose production rate and plasma insulin concentration, was inversely correlated with IHTG content (r = −0.599; P < .001). The ability of insulin to suppress fatty acid release from adipose tissue and to stimulate glucose uptake by skeletal muscle were also inversely correlated with IHTG content (adipose tissue: r = −0.590, P < .001; skeletal muscle: r = −0.656, P < .001). Multivariate linear regression analyses found that IHTG content was the best predictor of insulin action in liver, skeletal muscle, and adipose tissue, independent of body mass index and percent body fat, and accounted for 34%, 42%, and 44% of the variability in these tissues, respectively (P < .001 for each model). Conclusions: These results show that progressive increases in IHTG content are associated with progressive impairment of insulin action in liver, skeletal muscle, and adipose tissue in nondiabetic obese subjects. Therefore, nonalcoholic fatty liver disease should be considered part of a multiorgan system derangement in insulin sensitivity. Background & Aims: Nonalcoholic fatty liver disease is associated with insulin resistance and diabetes. The purpose of this study was to determine the relationship between intrahepatic triglyceride (IHTG) content and insulin action in liver (suppression of glucose production), skeletal muscle (stimulation of glucose uptake), and adipose tissue (suppression of lipolysis) in nondiabetic obese subjects. Methods: A euglycemic-hyperinsulinemic clamp procedure and stable isotopically labeled tracer infusions were used to assess insulin action, and magnetic resonance spectroscopy was used to determine IHTG content, in 42 nondiabetic obese subjects (body mass index, 36 ± 4 kg/m2) who had a wide range of IHTG content (1%–46%). Results: Hepatic insulin sensitivity, assessed as a function of glucose production rate and plasma insulin concentration, was inversely correlated with IHTG content (r = −0.599; P < .001). The ability of insulin to suppress fatty acid release from adipose tissue and to stimulate glucose uptake by skeletal muscle were also inversely correlated with IHTG content (adipose tissue: r = −0.590, P < .001; skeletal muscle: r = −0.656, P < .001). Multivariate linear regression analyses found that IHTG content was the best predictor of insulin action in liver, skeletal muscle, and adipose tissue, independent of body mass index and percent body fat, and accounted for 34%, 42%, and 44% of the variability in these tissues, respectively (P < .001 for each model). Conclusions: These results show that progressive increases in IHTG content are associated with progressive impairment of insulin action in liver, skeletal muscle, and adipose tissue in nondiabetic obese subjects. Therefore, nonalcoholic fatty liver disease should be considered part of a multiorgan system derangement in insulin sensitivity. See Hanouneh IA et al on page 584 in CGH. See Hanouneh IA et al on page 584 in CGH. Nonalcoholic fatty liver disease (NAFLD) has become an important public health problem in many industrialized countries because of its high prevalence, potential progression to severe liver disease,1Angulo P. Nonalcoholic fatty liver disease.N Engl J Med. 2002; 346: 1221-1231Crossref PubMed Scopus (4293) Google Scholar and association with cardiometabolic abnormalities, including diabetes, the metabolic syndrome, and coronary heart disease.2Marchesini G. Bugianesi E. Forlani G. et al.Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome.Hepatology. 2003; 37: 917-923Crossref PubMed Scopus (2275) Google Scholar, 3Ioannou G.N. Weiss N.S. Boyko E.J. et al.Elevated serum alanine aminotransferase activity and calculated risk of coronary heart disease in the United States.Hepatology. 2006; 43: 1145-1151Crossref PubMed Scopus (198) Google Scholar, 4Adams L.A. Lymp J.F. St Sauver J. et al.The natural history of nonalcoholic fatty liver disease: a population-based cohort study.Gastroenterology. 2005; 129: 113-121Abstract Full Text Full Text PDF PubMed Scopus (2380) Google Scholar Obesity is an important risk factor for NAFLD, and the prevalence of NAFLD is linearly associated with body mass index (BMI).5Ruhl C.E. Everhart J.E. Determinants of the association of overweight with elevated serum alanine aminotransferase activity in the United States.Gastroenterology. 2003; 124: 71-79Abstract Full Text Full Text PDF PubMed Scopus (504) Google Scholar The precise mechanism(s) responsible for the link between obesity and its metabolic complications is not known but likely involves insulin resistance, which is a common feature of obesity and NAFLD and is an important risk factor for cardiometabolic disease.6Chitturi S. Abeygunasekera S. Farrell G.C. et al.NASH and insulin resistance: insulin hypersecretion and specific association with the insulin resistance syndrome.Hepatology. 2002; 35: 373-379Crossref PubMed Scopus (1020) Google Scholar, 7Pagano G. Pacini G. Musso G. et al.Nonalcoholic steatohepatitis, insulin resistance, and metabolic syndrome: further evidence for an etiologic association.Hepatology. 2002; 35: 367-372Crossref PubMed Scopus (659) Google Scholar The term “insulin resistance” is most commonly used to describe impaired insulin-mediated glucose uptake in skeletal muscle. However, insulin also has important metabolic effects in other organ systems. Insulin resistance associated with obesity often involves the liver (impaired insulin-mediated suppression of glucose production) and adipose tissue (impaired insulin-mediated suppression of lipolysis).8Groop L.C. Saloranta C. Shank M. et al.The role of free fatty acid metabolism in the pathogenesis of insulin resistance in obesity and noninsulin-dependent diabetes mellitus.J Clin Endocrinol Metab. 1991; 72: 96-107Crossref PubMed Scopus (299) Google Scholar Recently, Gastaldelli et al found that intrahepatic triglyceride (IHTG) content in lean, overweight, and obese subjects with type 2 diabetes mellitus was directly correlated with the severity of insulin resistance in both liver and skeletal muscle.9Gastaldelli A. Cusi K. Pettiti M. et al.Relationship between hepatic/visceral fat and hepatic insulin resistance in nondiabetics and type 2 diabetic subjects.Gastroenterology. 2007; 133: 496-506Abstract Full Text Full Text PDF PubMed Scopus (469) Google Scholar However, the relationship between IHTG content and insulin action in subjects who do not have diabetes is unclear because of conflicting results from different studies, which have reported insulin resistance in liver but not muscle,10Seppala-Lindroos A. Vehkavaara S. Hakkinen A.M. et al.Fat accumulation in the liver is associated with defects in insulin suppression of glucose production and serum free fatty acids independent of obesity in normal men.J Clin Endocrinol Metab. 2002; 87: 3023-3028Crossref PubMed Scopus (877) Google Scholar insulin resistance in muscle but not liver,11Sanyal A.J. Campbell-Sargent C. Mirshahi F. et al.Nonalcoholic steatohepatitis: association of insulin resistance and mitochondrial abnormalities.Gastroenterology. 2001; 120: 1183-1192Abstract Full Text Full Text PDF PubMed Scopus (1833) Google Scholar and insulin resistance in both liver and muscle12Bugianesi E. Gastaldelli A. Vanni E. et al.Insulin resistance in non-diabetic patients with non-alcoholic fatty liver disease: sites and mechanisms.Diabetologia. 2005; 48: 634-642Crossref PubMed Scopus (585) Google Scholar, 13Marchesini G. Brizi M. Bianchi G. et al.Nonalcoholic fatty liver disease: a feature of the metabolic syndrome.Diabetes. 2001; 50: 1844-1850Crossref PubMed Scopus (2022) Google Scholar in subjects with NAFLD. The reason(s) for these apparent discrepancies is not clear but could be related to differences among studies in BMI and IHTG content in the study subjects and in plasma insulin concentration achieved during the clamp procedure used to evaluate insulin sensitivity. The purpose of the present study was to determine the relationship between IHTG content and insulin action in liver (glucose production), skeletal muscle (glucose uptake), and adipose tissue (lipolysis) in nondiabetic obese subjects who had a wide range of hepatic fat accumulation. A euglycemic-hyperinsulinemic clamp procedure, in conjunction with stable isotopically labeled tracer infusions, was used to assess insulin action, and magnetic resonance spectroscopy was used to determine IHTG content. We hypothesized that IHTG content would be inversely correlated with insulin sensitivity in all tissues, showing that NAFLD is part of a multiorgan metabolic disease complex. Forty-two obese subjects (11 men and 31 women; 41 ± 11 years old) participated in this study (Table 1). All subjects completed a comprehensive medical evaluation, which included a history and physical examination, blood tests, the Michigan Alcohol Screening Test,14Selzer M.L. The Michigan alcoholism screening test: the quest for a new diagnostic instrument.Am J Psychiatry. 1971; 127: 1653-1658Crossref PubMed Scopus (2729) Google Scholar and a 2-hour oral glucose tolerance test. Fourteen subjects (33%) had impaired glucose tolerance, based on the results of the 2-hour oral glucose tolerance test. Subjects who had diabetes, had chronic liver disease other than NAFLD, had a Michigan Alcohol Screening Test score ≥4, and had been taking medications known to cause liver abnormalities or affect metabolism were excluded. All subjects were sedentary (ie, participated in regular exercise <1 h/wk and ≤1 time/wk) and weight stable (ie, <2% weight change) for at least 3 months before the study. All subjects provided written informed consent before participating in the study, which was approved by the Human Studies Committee and the General Clinical Research Center Advisory Committee of Washington University School of Medicine in St Louis.Table 1Body Composition of the Study SubjectsMedian (interquartile range)BMI (kg/m2)35 (32–40)Body fat mass (%)41 (37–45)FFM (kg)58 (52–67)Abdominal subcutaneous adipose tissue (cm3)3542 (2663–4280)IAAT volume (cm3)aIAAT volume values available for 39 of 42 subjects because of technical problems in 3 subjects.1443 (857–1792)IHTG content (%)12.1 (3.9–23.3)a IAAT volume values available for 39 of 42 subjects because of technical problems in 3 subjects. Open table in a new tab Fat mass and fat-free mass (FFM) were determined by using dual-energy x-ray absorptiometry (QDR 4500; Hologic, Waltham, MA). Abdominal subcutaneous and intra-abdominal fat volumes were determined by using magnetic resonance imaging (Siemens, Iselin, NJ); the sum of 8 axial images of 1-cm thickness, beginning from the L4–L5 interspace and extending proximally, was used to determine abdominal subcutaneous and intra-abdominal adipose tissue (IAAT) volumes. Magnetic resonance spectroscopy (Siemens, Erlanger, Germany) was used to determine IHTG content; three 2 × 2 × 2–cm voxels were examined in each subject, and the values were averaged to provide an estimate of the percent of total liver volume comprised of fat.15Frimel T.N. Deivanayagam S. Bashir A. et al.Assessment of intrahepatic triglyceride content using magnetic resonance spectroscopy.J Cardiometab Syndr. 2007; 2: 136-138Crossref PubMed Scopus (37) Google Scholar Subjects were admitted to the inpatient unit of the General Clinical Research Center at Washington University School of Medicine on the evening before the clamp procedure. At 7 pm, subjects consumed a standard meal, which provided 12 kcal/kg adjusted body weight and contained 55% of total energy as carbohydrate, 30% as fat, and 15% as protein. Adjusted body weight was calculated as ideal body weight (based on the midpoint of the medium frame of the Metropolitan Life Insurance Tables) plus 0.25 × (Actual Body Weight − Ideal Body Weight). A 240-kcal liquid snack (Ensure; Ross Laboratories, Columbus, OH) was consumed at 8 pm. Subjects then fasted until completion of the clamp procedure the next day. At 5 am the following morning, one catheter was inserted into a forearm vein to infuse stable isotopically labeled tracers (purchased from Cambridge Isotope Laboratories, Andover, MA), dextrose and insulin, and a second catheter was inserted into a radial artery in the contralateral hand to obtain blood samples. Radial artery cannulation was not successful in 5 subjects, so a catheter was inserted into a hand vein, which was heated to 55°C by using a thermostatically controlled box, to obtain arterialized blood samples.16Jensen M.D. Heiling V.J. Heated hand vein blood is satisfactory for measurements during free fatty acid kinetic studies.Metabolism. 1991; 40: 406-409Abstract Full Text PDF PubMed Scopus (72) Google Scholar At 6 am, a primed continuous infusion of [6,6-2H2]glucose (priming dose, 22.5 μmol · kg−1; infusion rate, 0.25 μmol · kg−1 · min−1), dissolved in 0.9% NaCl solution, was started and maintained for 5.5 hours, until the end of stage 1 of the euglycemic-hyperinsulinemic clamp procedure. At 8 am, a continuous infusion of [2,2-2H2]palmitate (infusion rate, 0.035 μmol · kg−1 · min−1), bound to human albumin, was started and maintained for 3.5 hours (through the end of stage 1 of the euglycemic-hyperinsulinemic clamp procedure). At 9:30 am (3.5 hours after starting the infusion of glucose tracer), a 2-stage euglycemic-hyperinsulinemic clamp procedure was started and continued for 6 hours. During stage 1 of the clamp procedure (3.5 to 5.5 hours), insulin was infused at a rate of 20 mU · m−2 body surface area (BSA) · min−1 (initiated with a priming dose of 80 mU · m−2 BSA · min−1 for 5 minutes and then 40 mU · m−2 BSA · min−1 for 5 minutes). During stage 2 of the clamp procedure (5.5 to 9.5 hours), insulin was infused at a rate of 50 mU · m−2 BSA · min−1 (initiated with a priming dose of 200 mU · m−2 BSA · min−1 for 5 minutes and then 100 mU · m−2 BSA · min−1 for 5 minutes). These 2 insulin infusion rates were chosen to evaluate adipose tissue insulin sensitivity (low-dose insulin infusion to submaximally suppress lipolysis of adipose tissue triglycerides) and skeletal muscle insulin sensitivity (high-dose insulin infusion to adequately stimulate muscle glucose uptake).17Groop L.C. Bonadonna R.C. DelPrato S. et al.Glucose and free fatty acid metabolism in non-insulin-dependent diabetes mellitus Evidence for multiple sites of insulin resistance.J Clin Invest. 1989; 84: 205-213Crossref PubMed Scopus (743) Google Scholar Euglycemia was maintained at a blood glucose concentration of approximately 5.6 mmol/L (100 mg/dL) throughout stages 1 and 2 by infusing 20% dextrose enriched to 2.5% with [6,6-2H2]glucose at variable rates. The infusion rates of [6,6-2H2]glucose and [2,2-2H2]palmitate were reduced by 50% during stage 1, and [6,6-2H2]glucose infusion was reduced by 75% during stage 2 of the clamp procedure to account for changes in hepatic glucose production and lipolytic rates. Blood samples were obtained before beginning the tracer infusion to determine background glucose and palmitate tracer-to-tracee ratios, and every 10 minutes during the final 30 minutes of the basal period and stages 1 and 2 of the clamp procedure to determine glucose, free fatty acid, and insulin concentrations and substrate kinetics. These blood samples were collected in chilled tubes containing sodium EDTA. Samples were placed on ice; plasma was separated by centrifugation within 30 minutes of collection and then stored at −70°C until final analyses were performed. Blood was also obtained every 10 minutes during insulin infusion to monitor plasma glucose concentrations. These samples were collected in tubes containing heparin and analyzed immediately by using an automated glucose analyzer (Yellow Spring Instruments Co, Yellow Springs, OH). Plasma free fatty acid concentrations were quantified by using gas chromatography (5890-II; Hewlett-Packard, Palo Alto, CA) after adding heptadecanoic acid to plasma as an internal standard.18Patterson B.W. Zhao G. Klein S. Improved accuracy and precision of gas chromatography/mass spectrometry measurements for metabolic tracers.Metabolism. 1998; 47: 706-712Abstract Full Text PDF PubMed Scopus (74) Google Scholar Plasma insulin concentration was measured by using radioimmunoassay (Linco Research, St Charles, MO). Plasma glucose and palmitate tracer-to-tracee ratios were determined by using electron impact ionization gas chromatography/mass spectroscopy (MSD 5973 system with capillary column; Hewlett-Packard) as previously described.19Patterson B.W. Zhao G. Elias N. et al.Validation of a new procedure to determine plasma fatty acid concentration and isotopic enrichment.J Lipid Res. 1999; 40: 2118-2124Abstract Full Text Full Text PDF PubMed Google Scholar Isotopic steady state conditions were achieved during the final 30 minutes of the basal period and stages 1 and 2 of the clamp procedure. Basal endogenous glucose rate of appearance (Ra) in plasma was calculated by dividing the glucose tracer infusion rate by the average plasma glucose tracer-to-tracee ratio during the last 30 minutes of the basal period. It was assumed that glucose rate of disappearance (Rd) was equal to glucose Ra during basal conditions; during the clamp procedure, glucose Rd was assumed to be equal to the sum of endogenous glucose Ra and the rate of infused glucose. Palmitate Ra, an index of adipose tissue lipolysis, was calculated by dividing the palmitate tracer infusion rate by the average plasma palmitate tracer-to-tracee ratio obtained during the final 30 minutes of the basal period and stage 1 of the clamp procedure.20Mittendorfer B. Liem O. Patterson B.W. et al.What does the measurement of whole-body fatty acid rate of appearance in plasma by using a fatty acid tracer really mean?.Diabetes. 2003; 52: 1641-1648Crossref PubMed Scopus (96) Google Scholar Hepatic insulin sensitivity was determined by the reciprocal of the hepatic insulin resistance index, which was calculated as the product of the basal endogenous glucose production rate (in μmol · kg FFM−1 · min−1) and fasting plasma insulin concentration (in mU/L).21Matsuda M. DeFronzo R.A. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp.Diabetes Care. 1999; 22: 1462-1470Crossref PubMed Scopus (4519) Google Scholar, 22Gastaldelli A. Ferrannini E. Miyazaki Y. et al.Beta-cell dysfunction and glucose intolerance: results from the San Antonio metabolism (SAM) study.Diabetologia. 2004; 47: 31-39Crossref PubMed Scopus (270) Google Scholar Adipose tissue insulin resistance was assessed by calculating the relative decrease from basal in palmitate Ra during stage 1 of the clamp procedure. Skeletal muscle insulin resistance was assessed by calculating the relative increase from basal in glucose Rd during stage 2 of the clamp procedure. The statistical significance of the effect of insulin infusion on substrate metabolism was analyzed by using Student t test for paired samples. The statistical significance of relationships between variables was evaluated by using the Pearson product moment correlation coefficient. Multiple stepwise linear regression analyses (forward and backward with age, sex, BMI, percent body fat, FFM, IHTG content, and IAAT volume as independent variables) were performed to identify predictors of insulin action in liver, skeletal muscle, and adipose tissue. All reported P values are 2 sided, and a P value of < .05 was considered statistically significant. Statistical analyses were performed by using SPSS (Windows) version 14.0 (SPSS Inc, Chicago, IL). Results are expressed as means ± SD unless otherwise stated. There was a 65-fold range in IHTG content (from 0.7% to 45.5%) and a 10-fold range in IAAT volume (from 349 cm3 to 3759 cm3) but only a 2-fold range in percent body fat (from 28% to 52%) and a 1.5-fold range in BMI (from 30 kg/m2 to 46 kg/m2) (Table 1). No significant relationships were detected between IHTG content and BMI (r = 0.258; P = .10) or percent body fat (r = −0.063; P = .69). In contrast, IAAT volume correlated directly with IHTG content (Figure 1; r = 0.317; P < .05). Three subjects had very high values for IAAT volume; eliminating these outliers improved the correlation between IHTG and IAAT (r = 0.543; P = .001). Eight of the 42 subjects (19%) had elevated serum alanine aminotransferase (ALT) concentration (>53 IU/mL) (Table 2). There was a direct correlation between IHTG content and serum ALT concentration (r = 0.647; P < .001). Serum ALT concentration was also inversely correlated with hepatic insulin sensitivity (r = −0.340; P = .03), adipose tissue insulin sensitivity (r = −0.493; P = .001), and skeletal muscle insulin sensitivity (r = −0.379; P = .01).Table 2Metabolic Characteristics of the Study SubjectsMedian (interquartile range)Plasma glucose (mg/dL)94 (90–102)Plasma insulin (μU/mL)15.0 (10.7–21.3)Plasma triglycerides (mg/dL)118 (88–173)Plasma high-density lipoprotein cholesterol (mg/dL)44 (36–56)Plasma low-density lipoprotein cholesterol (mg/dL)95 (76–108)Serum ALT (U/L)30 (23–45) Open table in a new tab Basal glucose Ra ranged from 10.2 to 17.6 μmol · kg FFM−1 · min−1 and basal palmitate Ra ranged from 1.16 to 3.88 μmol · kg FFM−1 · min−1. Basal glucose Ra and palmitate Ra did not correlate with IHTG content (r = 0.141, P = .37 and r = 0.130, P = .41, respectively). IHTG content was directly correlated with basal plasma insulin concentrations (r = 0.598; P < .001) and inversely correlated with hepatic insulin sensitivity (r = −0.599; P < .001) (Figure 2, top panel). The euglycemic-hyperinsulinemic clamp procedure was used to assess adipose tissue (data from stage 1) and skeletal muscle (data from stage 2) insulin sensitivity. Mean plasma insulin concentrations increased from 17 ± 8 μU/mL at baseline to 48 ± 14 μU/mL during stage 1 and to 106 ± 26 μU/mL during stage 2 of the euglycemic-hyperinsulinemic clamp procedure. Palmitate Ra decreased by 66% ± 8% during stage 1 of the euglycemic-hyperinsulinemic clamp procedure (from 2.1 ± 0.6 μmol · kg FFM−1 · min−1 during basal conditions to 0.7 ± 0.3 μmol · kg FFM−1 · min−1 during stage 1) (P < .001) (Table 3). The relative suppression of palmitate Ra by low-dose insulin infusion during stage 1 was inversely correlated with IHTG content (r = −0.590, P < .001) (Figure 2, middle panel). Glucose Rd increased 3-fold above baseline during stage 2 of the euglycemic-hyperinsulinemic clamp procedure (from 14.4 ± 1.8 μmol · kg FFM−1 · min−1 during basal conditions to 46.7 ± 16.1 μmol · kg FFM−1 · min−1 during stage 2) (P < .001) (Table 3). The relative increase in glucose Rd by high-dose insulin infusion during stage 2 was inversely correlated with hepatic fat content (r = −0.656; P < .001) (Figure 2, bottom panel).Table 3Substrate Kinetics During Basal Conditions and During Stage 1 (Low-Dose Insulin) and Stage 2 (High-Dose Insulin) of the Hyperinsulinemic-Euglycemic Clamp Procedure in the Entire Cohort of Study SubjectsBasalStage 1Stage 2Endogenous glucose Ra (μmol · kg FFM−1 · min−1)14.0 ± 1.84.2 ± 1.71.6 ± 1.9Glucose Rd (μmol · kg FFM−1 · min−1)14.4 ± 1.821.0 ± 7.346.7 ± 16.1Palmitate Ra (μmol · kg FFM−1 · min−1)2.1 ± 0.60.7 ± 0.30.4 ± 0.2NOTE. Values are expressed as means ± SD. Open table in a new tab NOTE. Values are expressed as means ± SD. Multivariate linear regression analyses, which included age, sex, BMI, percent body fat, FFM, IHTG, and IAAT volume as independent variables, found that IHTG content was the best predictor of insulin action in liver, skeletal muscle, and adipose tissue, accounting for 34%, 42%, and 44% of the variability in these tissues, respectively (P < .001 for each model). In addition, IAAT volume was also identified as an independent predictor of insulin action in liver and skeletal muscle, so IHTG and IAAT accounted for 42% and 57%, respectively (P < .001 for each model). None of the other measures of body composition or fat distribution (FFM, percent body fat, and BMI) was an independent predictor of insulin action in any tissue. Obesity is an important risk factor for NAFLD and insulin resistance. In the present study, we evaluated the relationship between IHTG content and insulin sensitivity in obese subjects who did not have type 2 diabetes mellitus to avoid the potential confounding influence of advanced insulin resistance and beta cell failure on our measures of insulin action. In addition, we studied subjects who had a large range in IHTG content (1%–46% of liver volume) but a small range in percent body fat to increase our ability to determine the relationship between hepatic fat and insulin action, independent of adiposity. Insulin action was assessed in liver, skeletal muscle, and adipose tissue, because these are the major organs involved in the metabolic pathophysiology of obesity. Our data show that the amount of IHTG in obese persons without diabetes is directly correlated with impaired insulin action in liver (suppression of glucose production), skeletal muscle (stimulation of glucose uptake), and adipose tissue (suppression of lipolysis), independent of percent body fat and IAAT volume. These results suggest that NAFLD should be considered part of a multiorgan system derangement in insulin sensitivity and help explain why NAFLD is so closely linked with diabetes23Amarapurka D.N. Amarapurkar A.D. Patel N.D. et al.Nonalcoholic steatohepatitis (NASH) with diabetes: predictors of liver fibrosis.Ann Hepatol. 2006; 5: 30-33Crossref PubMed Google Scholar and the metabolic syndrome2Marchesini G. Bugianesi E. Forlani G. et al.Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome.Hepatology. 2003; 37: 917-923Crossref PubMed Scopus (2275) Google Scholar and is an important risk factor for coronary heart disease.4Adams L.A. Lymp J.F. St Sauver J. et al.The natural history of nonalcoholic fatty liver disease: a population-based cohort study.Gastroenterology. 2005; 129: 113-121Abstract Full Text Full Text PDF PubMed Scopus (2380) Google Scholar, 24Villanova N. Moscatiello S. Ramilli S. et al.Endothelial dysfunction and cardiovascular risk profile in nonalcoholic fatty liver disease.Hepatology. 2005; 42: 473-480Crossref PubMed Scopus (530) Google Scholar Our study cannot determine whether NAFLD actually causes or is simply a consequence of insulin resistance. In fact, it is possible that excessive IHTG is both a cause and a manifestation of insulin resistance, resulting from a sequence of events initiated by adipose tissue insulin resistance and propagated by excessive IHTG, whereby (1) adipose tissue insulin resistance increases the rate of release of free fatty acids into the bloodstream and increases free fatty acid delivery to the liver25Horowitz J.F. Klein S. Whole body and abdominal lipolytic sensitivity to epinephrine is suppressed in upper body obese women.Am J Physiol Endocrinol Metab. 2000; 278: E1144-E1152Crossref PubMed Google Scholar; (2) inadequate hepatic oxidization and/or secretion (as very-low-density lipoprotein triglyceride) of the increased fatty acid load results in fatty acid esterification and IHTG accumulation26Cassader M. Gambino R. Musso G. et al.Postprandial triglyceride-rich lipoprotein metabolism and insulin sensitivity in nonalcoholic steatohepatitis patients.Lipids. 2001; 36: 1117-1124Crossref PubMed Scopus (83) Google Scholar; (3) hyperinsulinemia and skeletal muscle insulin resistance also increase IHTG content by stimulating hepatic de novo lipogenesis and hepatic triglyceride synthesis27Petersen K.F. Dufour S. Savage D.B. et al.Inaugural article: the role of skeletal muscle insulin resistance in the pathogenesis of the metabolic syndrome.Proc Natl Acad Sci U S A. 2007; 104: 12587-12594Crossref PubMed Scopus (548) Google Scholar; (4) excessive IHTG releases fatty acids into the cytoplasm, which can cause hepatic insulin resistance and inflammation28Savage D.B. Petersen K.F. Shulman G.I. Disordered lipid metabolism and the pathogenesis of insulin resistance.Physiol Rev. 2007; 87: 507-520Crossref PubMed Scopus (802) Google Scholar; and (5) localized intrahepatic inflammation can contribute to peripheral insulin resistance.29Cai D. Yuan M. Frantz D.F. et al.Local and systemic insulin resistance resulting from hepatic activation of IKK-beta and NF-kappaB.Nat Med. 2005; 11: 183-190Crossref PubMed Scopus (1863) Google Scholar It is also possible that a fatty liver secretes cytokines, α2-Heremans-Schmid glycoprotein, and other, as yet unknown, products into the systemic circulation that can cause peripheral insulin resistance.30Stefan N. Hennige A.M. Staiger H. et al.Alpha2-Heremans-Schmid glycoprotein/fetuin-A is associated with insulin resistance and fat accumulation in the liver in humans.Diabetes Care. 2006; 29: 853-857Crossref PubMed Scopus (410) Google Scholar Traditionally, excessive IHTG, or steatosis, has been defined chemically when IHTG content exceeds 5% of liver volume or liver weight, or histologically when 5% of hepatocytes contain visible intracellular triglycerides.31Kleiner D.E. Brunt E.M. Van Natta M. et al.Design and validation of a histological scoring system for nonalcoholic fatty liver disease.Hepatology. 2005; 41: 1313-1321Crossref PubMed Scopus (7827) Google Scholar, 32Neuschwander-Tetri B.A. Brunt E.M. Wehmeier K.R. et al.Improved nonalcoholic steatohepatitis after 48 weeks of treatment with the PPAR-gamma ligand rosiglitazone.Hepatology. 2003; 38: 1008-1017Crossref PubMed Scopus (727) Google Scholar Recently, data obtained from 2 studies, which evaluated IHTG content by using magnetic resonance spectroscopy in large numbers of subjects, provide additional insights into defining “normal” IHTG content.33Szczepaniak L.S. Nurenberg P. Leonard D. et al.Magnetic resonance spectroscopy to measure hepatic triglyceride content: prevalence of hepatic steatosis in the general population.Am J Physiol Endocrinol Metab. 2005; 288: E462-E468Crossref PubMed Scopus (1263) Google Scholar, 34Petersen K.F. Dufour S. Feng J. et al.Increased prevalence of insulin resistance and nonalcoholic fatty liver disease in Asian-Indian men.Proc Natl Acad Sci U S A. 2006; 103: 18273-18277Crossref PubMed Scopus (310) Google Scholar The results from one study, conducted in a cohort of Hispanic and non-Hispanic white and black subjects who were considered to be at low risk for NAFLD (ie, BMI <25 kg/m2, no diabetes, and normal fasting serum glucose and alanine aminotransferase concentrations), suggest the threshold for a normal amount of IHTG should be 5.6% of liver volume, because this value represented the 95th percentile for this “normal” population.33Szczepaniak L.S. Nurenberg P. Leonard D. et al.Magnetic resonance spectroscopy to measure hepatic triglyceride content: prevalence of hepatic steatosis in the general population.Am J Physiol Endocrinol Metab. 2005; 288: E462-E468Crossref PubMed Scopus (1263) Google Scholar Data from the second study found the 95th percentile for IHTG content was 3% in lean, young adult, and white men and women who had normal oral glucose tolerance.34Petersen K.F. Dufour S. Feng J. et al.Increased prevalence of insulin resistance and nonalcoholic fatty liver disease in Asian-Indian men.Proc Natl Acad Sci U S A. 2006; 103: 18273-18277Crossref PubMed Scopus (310) Google Scholar None of the cut points that have been proposed for diagnosing steatosis is based on the relationship between IHTG and a rigorous assessment of either metabolic or clinical outcome. The results from our study show that the relationship between insulin sensitivity and IHTG content is monotonic, without evidence of an obvious threshold that can be used to define normality. The data from our study show that IHTG content is a marker of insulin sensitivity in multiple tissues, independent of BMI and percent body fat. In addition, IHTG content correlated directly with IAAT volume, which has also been observed in other populations, including patients with type 2 diabetes mellitus,9Gastaldelli A. Cusi K. Pettiti M. et al.Relationship between hepatic/visceral fat and hepatic insulin resistance in nondiabetics and type 2 diabetic subjects.Gastroenterology. 2007; 133: 496-506Abstract Full Text Full Text PDF PubMed Scopus (469) Google Scholar, 35Adiels M. Taskinen M.R. Packard C. et al.Overproduction of large VLDL particles is driven by increased liver fat content in man.Diabetologia. 2006; 49: 755-765Crossref PubMed Scopus (525) Google Scholar, 36Kelley D.E. McKolanis T.M. Hegazi R.A. et al.Fatty liver in type 2 diabetes mellitus: relation to regional adiposity, fatty acids, and insulin resistance.Am J Physiol Endocrinol Metab. 2003; 285: E906-E916Crossref PubMed Scopus (364) Google Scholar nondiabetic overweight and obese adults,37Westerbacka J. Corner A. Tiikkainen M. et al.Women and men have similar amounts of liver and intra-abdominal fat, despite more subcutaneous fat in women: implications for sex differences in markers of cardiovascular risk.Diabetologia. 2004; 47: 1360-1369Crossref PubMed Scopus (269) Google Scholar and overweight adolescents (S. Klein, unpublished observations, June 2007),38Fishbein M.H. Mogren C. Gleason T. et al.Relationship of hepatic steatosis to adipose tissue distribution in pediatric nonalcoholic fatty liver disease.J Pediatr Gastroenterol Nutr. 2006; 42: 83-88PubMed Google Scholar and helps explain the metabolic interrelationships often observed between subjects with abdominal obesity and NAFLD.2Marchesini G. Bugianesi E. Forlani G. et al.Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome.Hepatology. 2003; 37: 917-923Crossref PubMed Scopus (2275) Google Scholar, 3Ioannou G.N. Weiss N.S. Boyko E.J. et al.Elevated serum alanine aminotransferase activity and calculated risk of coronary heart disease in the United States.Hepatology. 2006; 43: 1145-1151Crossref PubMed Scopus (198) Google Scholar, 4Adams L.A. Lymp J.F. St Sauver J. et al.The natural history of nonalcoholic fatty liver disease: a population-based cohort study.Gastroenterology. 2005; 129: 113-121Abstract Full Text Full Text PDF PubMed Scopus (2380) Google Scholar, 5Ruhl C.E. Everhart J.E. Determinants of the association of overweight with elevated serum alanine aminotransferase activity in the United States.Gastroenterology. 2003; 124: 71-79Abstract Full Text Full Text PDF PubMed Scopus (504) Google Scholar However, IHTG content was a better predictor of insulin action in liver, skeletal muscle, and adipose tissue than was IAAT volume and accounted for 34%–44% of the variability in insulin sensitivity in these tissues. The detection of insulin resistance in our subjects was not simple and required the use of isotope tracer infusion to assess substrate kinetics and determine the metabolic response to a physiologic challenge of insulin infusion. Therefore, these data suggest that quantification of IHTG content by using magnetic resonance spectroscopy could become a useful diagnostic test by identifying insulin-resistant patients who might not be detected by a standard clinical evaluation. Although we found a relationship between hepatic insulin sensitivity, measured by using the hepatic insulin sensitivity index, and IHTG content, we did not observe a significant correlation between the relative suppression of endogenous glucose production during step 1 (low-dose insulin infusion) of the hyperinsulinemic-euglycemic clamp procedure and IHTG. However, stage 1 of our clamp was designed to evaluate adipose tissue, not hepatic insulin sensitivity. The short duration of stage 1 (2 hours) is not long enough to reliably achieve near steady-state conditions needed for accurate assessment of insulin-mediated suppression of glucose production at low plasma insulin concentrations.39Bergman R.N. Finegood D.T. Ader M. Assessment of insulin sensitivity in vivo.Endocrinol Rev. 1985; 6: 45-86Crossref PubMed Scopus (989) Google Scholar In summary, the findings from this study show that NAFLD in nondiabetic obese persons is part of a multiorgan disease complex, manifested by insulin resistance in the liver, skeletal muscle, and adipose tissue. Moreover, there was a direct monotonic relationship between IHTG content and insulin resistance in all 3 tissues, across a large range of percent liver fat. Therefore, even small amounts of IHTG were associated with metabolic dysfunction, without an obvious threshold effect. Even though IHTG content correlated with IAAT volume, IHTG content was a better predictor of insulin resistance, independent of percent body fat and IAAT volume. These results suggest that measurement of IHTG content could be a useful clinical tool to identify patients at increased risk for cardiometabolic disease.