fat in liver/muscle correlates more strongly with insulin sensitivity in rats than abdominal fat

188 VOLUME 17 NUMBER 1 | JANUARY 2009 | www.obesityjournal.org

articles nature publishing group

Methods and techniques

IntroductIonThere is abundant evidence that increased levels of plasma lipids, predominantly free-fatty acids and triglycerides, are causally involved in insulin resistance (1). In the 1960s, Randle et al.(2) introduced the idea that free-fatty acid interfere with insulin action in peripheral tissues, and suggested competition between plasma glucose and free-fatty acids as fuel for energy production, the “glucose fatty-acid cycle.” In animal models, increased insulin resistance has also been observed after lipid infusion or high-fat feeding (3,4), which was accompanied by a rise of the triglyceride content in liver and skeletal muscle (5,6). Currently, the ectopic deposition of fat in nonadipose tissue is considered an important factor in the development of insulin resistance (1,7,8).

Cross-sectional studies in human subjects have shown that insulin resistance correlates more tightly with the intramus-cular lipid (IML) concentration than with any other identified risk factor (9–11). Intrahepatic lipid (IHL) was also shown to

be inversely related to insulin sensitivity (12,13). Considering most ingested (or infused) glucose is taken up by muscle, IML theoretically may better reflect whole-body insulin sensitivity than does IHL, which is expected to correlate with reduced hepatic insulin sensitivity and not necessarily with whole-body insulin action. However, liver lipid content has previously been reported to correlate with measures of whole-body insulin sen-sitivity in individuals with and without diabetes (12,14,15).

Determination of lipid content in muscle or liver was clas-sically only possible by invasive techniques (16–18). Recently, a noninvasive method, volume-localized 1H-magnetic reso-nance spectroscopy (1H-MRS), was established (19–21). This method offers the unique ability to measure IHL or IML con-tent in vivo and has been shown to provide reliable quantifica-tion of IHL/IML content. Recently, noninvasive quantification of liver or muscle lipids with MRS has established that IHL/IML are reliable markers for insulin sensitivity in humans (11,22–27).

Fat in Liver/Muscle Correlates More Strongly With Insulin Sensitivity in Rats Than Abdominal FatSoo Lim1, Kyu R. Son2, In C. Song2, Ho S. Park3, Cheng J. Jin1, Hak C. Jang1, Kyong S. Park3, Young-Bum Kim4 and Hong K. Lee3

Intrahepatic or intramuscular lipid (IHL/IML) content has been reported to be correlated with insulin resistance. Visceral fat has also been shown to be associated with insulin resistance. Thus, we investigated whether IHL/IML or visceral fat content is more closely associated with insulin resistance. Twenty Sprague–Dawley rats were divided into two groups based on regular chow diet (RCD) or high-fat diet (HFD; 40% fat). The insulin-sensitivity index (ISI) was determined by euglycemic glucose clamp study, the amount of visceral fat by computed tomography (CT), and the IHL/IML content by magnetic resonance spectroscopy (MRS). Weight, food, and water intake, physical activity, energy expenditure, lipid profile, adiponectin, and high-sensitivity C-reactive protein (hsCRP) levels were measured. At the study end point, visceral fat, and the IHL/IML content were higher in the HFD group than in the RCD group. The IHL/IML content was more highly correlated with ISI than was visceral fat amount. Stronger correlations were also found between adiponectin or hsCRP level and IML/IHL content than visceral fat, especially in the HFD group. Furthermore, the IHL/IML content was significantly associated with the ISI in the multiple regression models but visceral fat was not. There was clear discrimination between RCD and HFD groups in scatter plots of IML/IHL against the ISI, but substantial overlap in that of visceral fat against the ISI. This result suggests that IHL/IML contents are closely related with insulin resistance or atherosclerosis and is a better metabolic index of insulin sensitivity than the visceral fat.

Obesity (2008) 17, 188–195. doi:10.1038/oby.2008.486

1Department of Internal Medicine, Seoul National University College of Medicine, Seoul National University Bundang Hospital, Seongnam, Korea; 2Department of Radiology, Seoul National University College of Medicine, Seoul, Korea; 3Department of Internal Medicine, Seoul National University College of Medicine, Seoul, Korea; 4Division of Endocrinology, Metabolism and Diabetes, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, USA. Correspondence: Hong Kyu Lee (hkleemd@snu.ac.kr)

Received 10 September 2007; accepted 3 August 2008; published online 23 October 2008. doi:10.1038/oby.2008.486

obesity | VOLUME 17 NUMBER 1 | JANUARY 2009 189

articlesMethods and techniques

Visceral fat has also been shown in many studies to be closely associated with insulin resistance, and an increased deposition of visceral fat plays an important role in the development of type 2 diabetes in humans (28). A computed tomography (CT) scan has been shown to be the most accurate method of meas-uring the amount of visceral fat.

Until now, adipocytokine production and body composi-tion, especially the amount of visceral fat and the IML or IHL content, have not been systematically assessed to determine their combined role in insulin sensitivity and energy expendi-ture. Clearly, studies that define whether lipid accumulations in liver, muscle, or abdomen are most closely associated with insulin resistance would be of interest.

Therefore, in this study we wanted to noninvasively moni-tor in an animal model the lipid content of liver and muscle by MRS and the amount of visceral fat by CT. Our intention was to determine the relationship between IHL/IML levels or the amount of visceral fat and insulin resistance in animals fed a normal or a high-fat diet (HFD). We also aimed to identify whether IHL/IML content or the amount of visceral fat is more closely correlated with insulin resistance.

Methods and ProceduresanimalsTwenty male Sprague–Dawley rats aged 8 weeks were divided into two groups (10/group). One group was fed with a regular chow diet (RCD) for 5 months; the other group was fed with an RCD for the first 3 months and an HFD (40% by weight) for the last 2 months. The regular chow consisted of 16.0% fat, 63.0% carbohydrate, and 20.0% protein (by calories), and 7.0% corn oil, 10.0% sucrose, 13.2% dex-trose, 40.0% cornstarch, 5.0% cellulose, and 20.0% casein (by weight). The HFD consisted of 64.0% fat, 20.0% carbohydrate, and 14.0% pro-tein (by calories), and 33.0% shortening, 7.0% corn oil, 10.0% sucrose, 13.2% dextrose, 5.0% cornstarch, 5.0% cellulose, and 20.0% casein (by weight). The remaining percentages of the two diets consisted of vita-mins and minerals.

Both groups had the same mean body weight at baseline. All rats were maintained in plastic cages in an air-conditioned room at 22 ± 2 °C and 55 ± 10% humidity. The animals were all allowed free access to diet (Han Sam R&D, Seoul, Korea). Weight, diet, and water intake were measured twice a month during the study period. Animal movement was evalu-ated by a spontaneous motor activity analyzer (IW-800CT, modular test chamber, Korea). Horizontal locomotion and rearing activity were monitored over 4 h twice a week. Animals were handled in compliance with the Guide for Experimental Animal Research from the Laboratory for Experimental Animal Research, Clinical Research Institute, Seoul National University Hospital.

obesity measurement

Abdominal obesity by high resolution CT. Visceral fat in rats can be assessed by CT and is significantly correlated with insulin sensitivity (29–31). In this study, the visceral fat areas were quantified by a non-contrast CT scan (Somatom Sensation 16, Siemens, Munich, Germany) (120 kVp, 150 mAs, 3 mm slice thickness, 3 mm reconstruction interval) under 1.8–2.0% isoflurane in O2 anesthesia at the study end point. With the rats in a supine position, CT slice scans were acquired at the upper margin of the L3 vertebra and over L1–L5 to measure the amount of abdominal and visceral fat at a single level and in the whole abdomen. Adipose tissue attenuation was determined by measuring the mean value of all pixels within the range of −250 to −50 Hounsfield units. The images were converted into files compatible with a commercial software program (Rapidia; 3DMED, Seoul, Korea). To assess visceral

adipose tissue volume, each abdominal image was edited by erasing the image exterior to the innermost abdominal wall muscles with a mouse-driven cursor, and the resulting images were saved in separate files. The amount of visceral fat was calculated by a radiologist.

IML and IHL content by MRS. (Figure 1). The IML and IHL con-tent was measured by MRS, a relatively new approach to noninva-sive assessment of IML or IHL under the same anesthesia. Rats were anesthetized with 2–3 volume percent isoflurane and 1:2 O2:N2O, and their temperature was kept at 37.5 °C. An in vivo 1H-MRS study was performed by a three Tesla clinical unit (Signa Excite; GE, Milwaukee, WI) using an 8-channel head coil. Voxels of 10 × 10 × 10 mm were located in anterior thigh muscle and liver, avoiding vascular structures and gross adipose tissue deposits.

A probe-p sequence (TR/TE = 2,000 ms/35 ms) was used for MRS. Magnetic resonance imaging with the T2-weighted fast spin echo sequences (3,000–3,500/100–120, 18-cm field of view, 256 × 160 matrix, 3 mm slice thickness) in the axial or coronal planes preceded 1H-MR spectra in order to define the volume of interest. All spectra were proc-essed using Mrdx (CAD Impact, Seoul, Korea) based on IDL (Research Systems, Boulder, CO). The water signal was suppressed by a frequency-selective saturation pulse at the resonance of water. A sweep width of 5,000 Hz was used with a data size of 2,048 points. Only the second half of the echo was acquired. Following the zero filling of 8,192 points in all the free induction decay data, an exponential line broadening (center: 0 ms, half time: 150 ms) was performed before a Fourier transformation. Zero-order phase correction was applied to all spectra.

The integral of the IML signal (1.3 ppm) was related to that of total creatine (tCr; 3.05 ppm). The IML/tCr ratio corresponded to the total muscle IML value. The tCr values for the anterior thigh muscle were determined in rats at 5 months of age. Fat content was expressed as a percentage of the ratio of the fat to water signal. The reliability of the 1H-NMR method for determination of IML or IHL content was assessed by performing repeated measurements on the same individual on different study days and was found to be <15%.

Measurement of energy expenditureAn indirect calorimeter (Columbus Instruments, Columbus, OH) was used to measure the basal energy metabolism status. After baseline O2, CO2 and flow were measured, each rat from each group was localized in a metabolic monitor cage for 30 min, and O2 and CO2 were measured again when the values had stabilized. We calculated energy expenditure according to the following formula provided by the manufacturer.Oxygen consumption and CO2 production:

VO2 = ViXi = VoXoVCO2 = VoYo = ViYi

where: Vi and Vo are the test chamber input and output, Xi and Xo are O2 fractions at the input and output test chamber, Yi and Yo are CO2 tractions at the input and output of the test chamber, Vi or Vo is monitored directly by Mass Flow measurement (l/min). Respiratory exchange ratio (RER):

Energy expenditure (EE) (kcal):

hyperinsulinemic euglycemic clampAt the study end point (5 months), hyperinsulinemic euglycemic clamp was performed as described by DeFronzo et al. (32). The clamp was conducted using one rat from each group in pairs. At 8 AM on the study day, a microcatheter (Intramedic polyethylene tubing PE

2

2

VCORER =

VO

22

2

VCOEE = 3.815 + 1.232 VO

VO

× ×

190 VOLUME 17 NUMBER 1 | JANUARY 2009 | www.obesityjournal.org

articlesMethods and techniques

10, PE 50; Clay Adams, Parsippany, NJ) was inserted into the tail vein and artery. Normal saline was infused into the tail artery by syringe pump (Medifusion 2010I; Medexinc, Duluth, GA) at a rate of 15 μl/min.

The clamp study was begun at 2 PM after at least 5 h fasting. Blood samples (200 μl) were drawn from the tail artery to measure baseline glucose and insulin. Insulin (Novolin-R; Novo Nordisk, Bagsvaerd, Denmark) was infused by syringe pump into one tail vein at a rate of 12 mU/kg/min and venous blood (50 μl) was drawn from the tail vein every 10 min to measure blood glucose. Twenty-five percent glucose was infused into the other tail vein, and the infusion rate was calibrated to maintain the baseline glucose level. Insulin was dissolved in normal saline with 0.2% bovine serum albumin (Sigma) to the concentration of 0.24 U/ml and was used after filtering through a 0.45 μm filter. Blood samples were drawn at 90 and 120 min after clamping to measure the insulin level. Insulin was measured by a rat insulin–specific kit (Linco Research, St. Charles, MO). Serum glucose level reached stability after 70–80 min of clamp study. The glucose infusion rate was measured during the clamp study when steady state was achieved (90–120 min). The insulin sensitivity index (ISI) was calculated as the glucose infu-sion rate divided by the mean insulin concentration at 90–120 min after clamping.

Intravenous glucose tolerance test (IVGtt)An IVGTT was performed in the light period between 1:00 and 2:00 PM. Animals were food-deprived for 5 h before the experiment started. To facilitate stress-free blood sampling, two infusion catheters (PE 10, Intramedic; Clay Adams, Parsippany, NJ) were placed in the tail veins of rats on the evening before the experiment. A bolus dose of 0.5 g/kg body weight glucose was injected into the right tail vein immediately after blood sampling from the left tail vein for measurement of serum concentrations of glucose and insulin (0 min). Blood samples were col-lected again from the left tail vein at 2, 4, 6, 10, 20, 30, and 60 min for measurements of serum concentrations of glucose and insulin. Glucose levels were measured by YSI 2300 (Yellow Springs Instrument, Yellow Springs, OH). Blood samples were stored at −70 °C until insulin meas-urement. Plasma levels of insulin were measured using commercial

radioimmunoassay kits (Linco, St. Louis, MO). Areas under the curve for glucose and insulin (AUCglucose and AUCinsulin) were calculated using the trapezoid rule for insulin data from 0 to 60 min.

Measurement of biochemical parametersAfter 12 h of overnight fasting at the study end point, blood samples were drawn from the abdominal aorta. Plasma was separated immedi-ately by centrifugation (3,000 rpm, 10 min, 4 °C). Plasma glucose levels were measured using a glucose oxidase method (YSI 2300-STAT; Yellow Springs Instrument) immediately after blood was drawn. Serum insulin was measured using insulin-specific radioimmunoassay kits (Linco). Homeostasis model assessment insulin resistance index (HOMA-IR) and β-cell function (HOMA-β) were obtained to evaluate insulin resist-ance and β-cell function. Total cholesterol and triglyceride concentra-tions were determined by enzymatic procedures (Hitachi 747 chemistry analyzer; Hitachi, Tokyo, Japan).

Measurement of adiponectin and hscrP concentrationsThe adiponectin level was measured using an enzyme-linked immuno-sorbent assay (Adipogen, Seoul, Korea). The intra- and interassay coef-ficients of variation were 3.3% and 7.4%, respectively. High-sensitivity C-reactive protein (hsCRP) concentration was measured by immuno-radiometric assay as previously described (33).

Western blot assays for akt and pakt in liver and muscleLiver and muscle tissues were harvested at the end of the study for west-ern blot analysis. Tissue segments were homogenized in lysis buffer and protein concentrations were determined using a protein assay kit (Pierce Biotechnology, Rockford, IL). Proteins (20 μg aliquots) were separated on sodium dodecyl sulfate polyacrylamide electrophoresis gels and trans-ferred to a polyvinylidene difluoride membrane (Millipore, Billerica, MA). Membranes were blocked with phosphate-buffered saline plus 0.3% Tween-20 (T-PBS) containing 5% dry milk and incubated with primary antibody overnight at 4 °C. After three washes with phosphate-buffered saline plus 0.3% Tween-20, membranes were reblocked and incubated with secondary antibody for 1 h at room temperature. The primary antibodies used were anti-Akt antibody (1:500 dilution; Santa Cruz Biotechnology, Santa Cruz, CA) and anti-pAkt antibody (Ser473) (1:500 dilution; Santa Cruz Biotechnology). The secondary antibodies were anti-rabbit immunoglobulin G/horseradish peroxidase conjugate (1:4,000 dilution; Promega, Madison, WI). Band density was quantified using a densitometer (Bio image analyzer, CN-115, ETX-20MX, Vilber Laurmat, Marne-La-Valle, France) and normalized to that of Akt.

statisticsAll data are presented as mean ± s.d. We used Student’s t-test to com-pare parameters between the two groups. The Pearson correlation coefficient was used for analysis of the simple correlation between the content of visceral fat, IML or IHL, and various parameters including insulin resistance index, adiponectin, and hsCRP. The value of energy expenditure was log-transformed before analysis because it showed skewed distribution. Values at P < 0.05 were considered statistically significant.

resultsThe amount of abdominal visceral fat, as measured by CT scan or the amount of fat in liver or muscle measured by MRS, and body weight were higher in rats fed an HFD than in those fed an RCD (P < 0.01). The amount of epididymal fat was also slightly higher in the HFD group, in the RCD group with bor-derline significance (Table 1). As expected, HFD rats showed poorer insulin sensitivity than the RCD group; higher plasma insulin, the AUCinsulin and AUCglucose, and HOMA-IR, lower

a b

Figure 1 Water-suppressed magnetic resonance spectra of the (a) hepatic tissue and (b) anterior thigh muscle using a probe-p (TR/TE = 2,000 ms/ 35 ms) on a 3.0T magnetic resonance unit. Both magnetic resonance spectra show larger lipid magnetic resonance signals at the range from 0.9 to 1.6 ppm.

obesity | VOLUME 17 NUMBER 1 | JANUARY 2009 191

articlesMethods and techniques

glucose infusion rate and ISI. HFD rats also showed lower β-cell function assessed by the HOMA model than the RCD group but this difference was not statistically significant. The HFD group showed poor lipid profiles compared to the RCD group although only triglyceride levels had borderline statis-tical significance. Adiponectin was lower and an inflamma-tory marker, hsCRP, was higher in HFD rats than in RCD rats. No significant changes in horizontal/spontaneous locomotor activities were observed at any point.

Table 2 shows the correlation between visceral fat meas-ured by CT, the amount of fat in liver and muscle measured by MRS, and various parameters related to insulin resistance. When both groups were combined, the amount of visceral fat and the IHL and IML content were all correlated with the ISI as measured by hyperinsulinemic euglycemic clamp, as well as with surrogate indices of insulin resistance. IHL and IML con-tent were negatively correlated with the adiponectin level and positively with hsCRP levels.

After separating the RCD and HFD groups, the negative corre-lation between IHL/IML content and adiponectin or ISI, and the positive correlation between IHL/IML content and HOMA-IR or hsCRP remained in the HFD group (Table 3). In addition, significant correlations were found between insulin resistance parameters and other related factors including ISI, HOMA-IR,

adiponectin, hsCRP and various parameters related with obes-ity, lipids and energy metabolism.

Using multiple regression models, we further investigated the independent association of IHL/IML content or visceral fat amount with ISI. In this analysis, weight, fasting insulin and glucose, total cholesterol, triglyceride, HDL-cholesterol, diet intervention, energy expenditure, hsCRP, and IHL/IML contents were selected as independent variables. Finally, IHL/IML contents were significantly associated with ISI after adjustment for other variables (P = 0.046 and P = 0.005, respectively). Diet intervention was also found to affect ISI significantly (P < 0.05). This significant association between IHL/IML contents and ISI disappeared when HOMA index or adiponectin level was added in the regression model. We also conducted another regression model including vis-ceral fat amount and same variables. Interestingly, visceral fat amount was not associated with ISI in this regression model.

Figure 2 shows the correlation between the ISI measured by hyperinsulinemic euglycemic clamp and the amount of intra-hepatic and intramuscular fat measured by proton MRS (1H-MRS). There was a strong correlation between ISI and IHL/IML. Although there was substantial overlap between the RCD and HFD groups in the scatter plot of visceral fat vs. ISI, there

table 1 characteristics of weight, amount of fat, and biochemical parameters in rats after 2 months of diet intervention

Regular chow diet (n = 10) High-fat diet (n = 10) P

Weight (g) 545.5 ± 38.5 603.0 ± 46.7 0.003

Visceral fat (mm3) at a single level 2,055.2 ± 726.2 3,397.3 ± 1077.8 0.004

Total visceral fat (mm3) 25,005.7 ± 6,845.9 35,365.4 ± 6,025.7 0.001

Intrahepatic lipid contenta 0.133 ± 0.037 0.266 ± 0.074 0.001

Intramuscular lipid contenta 0.069 ± 0.025 0.134 ± 0.027 0.001

Epididymal fat (g) 6.6 ± 0.9 7.9 ± 2.0 0.087

Plasma glucose (mmol/l) 4.51 ± 0.25 4.81 ± 0.56 NS

Plasma insulin (pmol/l) 4.86 ± 2.08 7.64 ± 4.86 0.076

Insulin-to-glucose ratio 1.01 ± 0.48 1.64 ± 1.13 NS

AUCinsulin 1.5 ± 0.33 2.3 ± 0.33 0.001

AUCglucose 139.3 ± 9.0 155.0 ± 16.5 0.017

HOMA-IR 0.131 ± 0.057 0.239 ± 0.144 0.048

HOMA-β 14.0 ± 7.5 21.6 ± 17.5 NS

Total cholesterol (mmol/l) 1.25 ± 0.48 1.52 ± 0.43 NS

Triglyceride (mmol/l) 0.75 ± 0.20 1.06 ± 0.47 0.077

HDL-cholesterol (mmol/l) 0.51 ± 0.13 0.51 ± 0.13 NS

LDL-cholesterol (mmol/l) 0.40 ± 0.39 0.53 ± 0.32 NS

Adiponectin (μg/ml) 6.2 ± 0.8 4.7 ± 0.7 0.001

hsCRP (mg/l) 0.03 ± 0.01 0.05 ± 0.02 0.018

Glucose infusion rate (mg/kg/min) 43.0 ± 7.3 35.5 ± 10.2 0.061

Insulin sensitivity index 7.6 ± 1.2 4.6 ± 1.5 0.001

Energy expenditure (kcal/kg/day) 67.9 ± 7.0 58.1 ± 10.1 0.023

AUCinsulin and AUCglucose, area under the curve of insulin and glucose; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance; HOMA-β, homeostasis model assessment of β-cell function; hsCRP, high-sensitivity C-reactive protein; LDL, low-density lipoprotein.aArbitrary unit.

192 VOLUME 17 NUMBER 1 | JANUARY 2009 | www.obesityjournal.org

articlesMethods and techniques

was relatively clear discrimination between RCD and HFD groups in the plot of IML/IHL against ISI.

Figure 3 shows total and phosphorylated Akt (pAkt) level in muscle and liver by western blot. Fold increase of the pAkt level

0.0 2.5 5.0 7.5 10.0 12.5

0.0

0.1

0.2

0.3

0.4

0.5

Regular chow

r = −0.597P = 0.068

r = −0.586P = 0.075

Insulin-sensitivity index

0.0 2.5 5.0 7.5 10.0 12.5

0

1,000

2,000

3,000

4,000

5,000

6,000

Regular chow

High fat

r = −0.446P = 0.196

r = −0.546P = 0.103

Insulin-sensitivity index

Fat

am

ount

at L

3 by

CT

sca

n

Fat

am

ount

ove

r L1

–L5

by C

T s

can

0.0 2.5 5.0 7.5 10.0 12.5

0

10,000

20,000

30,000

40,000

50,000

Regular chow

High fat

r = −0.676P = 0.032

r = −0.571P = 0.085

Insulin-sensitivity index

0.0 2.5 5.0 7.5 10.0 12.5

0.00

0.05

0.10

0.15

0.20

r = −0.620P = 0.056

High fat High fat

r = −0.478P = 0.163

Regular chow

Insulin-sensitivity index

Intr

amus

cula

r lip

id c

onte

nt b

y M

RS

Intr

ahep

atic

lipi

d co

nten

t by

MR

S

a

b

Figure 2 (a) Correlation between insulin sensitivity index measured by euglycemic hyperinsulinemic clamp and intrahepatic and intramuscular fat amount by proton magnetic resonance spectroscopy (1H-MRS), and (b) between insulin sensitivity index and the amount of visceral fat at L3 or over L1~L5 by CT scan.

Regular chow

Regular chowHigh fat

Regular chowHigh fat

Regular chow

a b

5

2

3

4

P < 0.05

0

1

2

Pho

spho

ryla

tion

(fol

d in

crea

se)

5

2

3

4

0

1

2

Pho

spho

ryla

tion

(fol

d in

crea

se)

3

4

3

4

Regular chow

1

2

1

2

0.10 0.15 0.20 0.0 0.1 0.2 0.3 0.4 0.5

0

Intramuscular lipid content by MRS Intrahepatic lipid content by MRS

pAkt

leve

l (fo

ld in

crea

se)

pAkt

leve

l (fo

ld in

crea

se)

0

pAkt

Akt

pAkt

Akt

pAkt

Akt

pAkt

Akt

High fat High fat

P < 0.05

High fat

r = −0.676P = 0.001

r = −0.534P = 0.015

Regular chow High fat

0.00 0.05

Muscle Liver

Figure 3 Western blot analysis of phosphorylated Akt (pAkt) in (a) muscle and (b) liver (upper), quantified band densities of pAkt in each tissue (middle), and correlation analysis between intramuscular/intrahepatic fat amount and pAkt (lower).

table 2 correlation between visceral fat by ct and amount of fat in liver and muscle by Mrs, and various parameters related to insulin resistance

Visceral fat (single level) Total visceral fatIntrahepatic lipid

contentIntramuscular lipid

content

Weight (g) 0.683** 0.541* 0.583** 0.664**

Epididymal fat (g) 0.261 0.232 0.277 0.417

Plasma glucose (mmol/l) 0.374 0.288 0.436 0.469

Plasma insulin (pmol/l) 0.636** 0.624** 0.684** 0.648**

Insulin-to-glucose ratio 0.581** 0.611** 0.614** 0.576**

AUCinsulin 0.669** 0.683** 0.762** 0.799**

AUCglucose 0.586** 0.590** 0.583 0.623**

HOMA-IR 0.646** 0.604** 0.711** 0.676**

HOMA-β 0.211 0.340** 0.282 0.237

Total cholesterol (mmol/l) 0.191 0.257 0.175 0.178

Triglyceride (mmol/l) 0.169 0.353 0.337 0.433

HDL-cholesterol (mmol/l) 0.109 0.131 –0.116 –0.130

LDL-cholesterol (mmol/l) 0.133 0.119 0.104 0.065

Adiponectin (μg/ml) –0.541* –0.657** –0.772** –0.795**

hsCRP (mg/l) 0.409 0.410 0.738** 0.682**

GIR (mg/kg/min) –0.571** –0.568** –0.524* –0.686**

Insulin sensitivity index –0.486 –0.556* –0.576* –0.707**

Log EE (kcal/kg/day) –0.214 –0.291 –0.302 –0.364

AUCinsulin and AUCglucose, Area under the curve of insulin and glucose; CT, computed tomography; EE, energy expenditure; GIR, glucose infusion rate; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance; HOMA-β, homeostasis model assessment of β-cell function; hsCRP, high-sensitivity C-reactive protein; LDL, low-density lipoprotein; MRS, magnetic resonance spectroscopy.*P < 0.05. **P < 0.01.

obesity | VOLUME 17 NUMBER 1 | JANUARY 2009 193

articlesMethods and techniques

in muscle and liver tissue was presented in each animal (the first one was used as a reference). The pAkt expression levels in both muscle and liver were found to be less in HFD group than those in RCD group. The IML/IHL contents were negatively correlated with pAkt expression levels of muscle/liver (index of tissue-specific insulin sensitivity) (P < 0.05).

dIscussIonIn this study, we found that the content of IHL or IML was more closely associated with ISI than was the content of abdominal visceral fat. Furthermore, there was a relatively clear discrimi-nation between animals fed an RCD or an HFD in the scatter plot of IML/IHL against ISI, but substantial overlap in that of visceral fat against ISI. However, there was no significant dif-ference between IHL and IML content in their relationship to ISI or other related factors.

In humans, a negative correlation between insulin sensitiv-ity and lipid content in muscle has been shown (9–11). Skeletal muscle is the main destination for insulin-stimulated glucose disposal9 and therefore is considered a principal determi-nant of systemic insulin resistance. A number of studies have demonstrated that accumulation of intramyocellular triglyc-erides accompanies the development of insulin resistance in prediabetic and/or obese adolescents (34,35), type 1 or type 2 diabetic patients (17,36), women with prior gestational diabetes

(37), and children of type 2 diabetes patients (17,24). A further study of healthy individuals showed that subjects with higher intramyocellular lipid (IMCL) content had decreased insulin receptor autophosphorylation and decreased signaling through insulin receptor substrate-1 and PI-3-kinase, suggesting that the efficiency of the postinsulin receptor tyrosine kinase cascade is impaired in subjects with higher IMCL (23). Dietary interven-tion has been shown to deplete IMCL stores in a selective man-ner in obese subjects, resulting in improved insulin sensitivity (38). Thus, IML becomes a marker for insulin resistance.

IHL has also been known to be inversely related to insu-lin sensitivity (12,13). The accumulation of lipid in the liver reduces insulin clearance and leads to peripheral insulin resist-ance via a downregulation of insulin receptors (12,39). Thus, the accumulation of lipid in the liver might be a result of an increased flux of free-fatty acids into liver and reduced insulin clearance, and lead to peripheral insulin resistance.

Interestingly, drug intervention with rosiglitazone induced a clear reduction of lipid content in liver and muscle despite increased body weight, and a marked improvement of insulin sensitivity. It also increased adiponectin levels in human and animal models. Thus, the insulin sensitizing properties of ros-iglitazone were consistent with a redistribution of lipids from nonadipose tissues (skeletal muscle and liver) back into fat tis-sue (40) and with increasing levels of adiponectin (41).

table 3 correlation according to diet group between visceral fat by ct, amount of fat in liver and muscle by Mrs, and various parameters related to insulin resistance

Regular chow diet group High-fat diet group

Visceral fat (single level)

Total visceral fat IHL content

IML content

Visceral fat (single level)

Total visceral fat IHL content IML content

Weight (g) 0.438 0.206 0.504 0.410 0.516 0.229 0.100 0.293

Epididymal fat (g) –0.226 –0.241 –0.545 0.166 0.111 0.050 0.078 0.178

Plasma glucose (mmol/l) –0.029 –0.283 –0.172 –0.030 0.284 0.270 0.351 0.434

Plasma insulin (pmol/l) 0.468 0.551 0.618* 0.493 0.571 0.579 0.625 0.651*

Insulin/glucose 0.436 0.543 0.589 0.463 0.527 0.585 0.567 0.578

AUCinsulin 0.460 0.452 0.324 0.389 0.370 0.297 0.561 0.595

AUCglucose 0.150 0.378 0.347 0.037 0.500 0.427 0.326 0.554

HOMA-IR 0.493 0.547 0.620 0.520 0.574 0.544 0.649* 0.680*

HOMA-β 0.328 0.491 0.500 0.380 0.064 0.208 0.191 0.122

Total cholesterol (mmol/l) 0.050 0.027 0.138 –0.076 –0.021 0.161 –0.217 –0.162

Triglyceride (mmol/l) 0.312 0.402 0.020 0.516 –0.269 –0.021 0.045 0.073

HDL-cholesterol (mmol/l) –0.301 0.042 –0.107 –0.308 0.383 0.197 –0.371 –0.363

LDL-cholesterol (mmol/l) 0.086 –0.076 0.201 –0.117 0.003 0.162 –0.185 –0.131

Adiponectin (μg/ml) –0.103 –0.434 –0.210 –0.376 –0.253 –0.179 –0.750* –0.655*

hsCRP (mg/l) 0.236 0.295 0.287 0.311 0.081 –0.055 0.617* 0.630*

GIR (mg/kg/min) –0.362 –0.409 –0.226 –0.531 –0.473 –0.485 –0.376 –0.715*

Insulin-sensitivity index –0.446 –0.476 –0.586 –0.473 –0.546 –0.571 –0.597* –0.620*

Log EE (kcal/kg/day) 0.103 −0.037 –0.321 –0.176 0.142 0.052 –0.402 –0.541

AUCinsulin and AUCglucose, Area under the curve of insulin and glucose; CT, computed tomography; EE, energy expenditure; GIR, glucose infusion rate; HOMA-β and HOMA-IR, homeostasis model assessment of β-cell function and insulin resistance; hsCRP, high-sensitivity C-reactive protein; LDL, low-density lipoprotein; MRS, magnetic resonance spectroscopy.*P < 0.05.

194 VOLUME 17 NUMBER 1 | JANUARY 2009 | www.obesityjournal.org

articlesMethods and techniques

Visceral fat has also been shown to be closely correlated with insulin sensitivity in both human and animal models (29,42). In this study, we too found that the amount of visceral fat was highly correlated with the ISI as measured by hyperinsuline-mic euglycemic clamp, and with various surrogate markers of insulin resistance. Although the amount of visceral fat both at L3 level and over the area between L1–L5 correlated with the ISI, the amount over L1–L5 showed a stronger relationship with ISI than that of the single L3 level (Table 2).

In this study, lipid content in liver or muscle had a similar or higher correlation coefficient with ISI than visceral fat meas-ured by CT scan. Furthermore, IHL/IML content was found to be significantly associated with insulin sensitivity in the mul-tiple regression model but visceral fat amount was not. These data suggest that ectopic fat accumulation may be more clearly related with insulin resistance than visceral fat.

Interestingly, the variability in intramuscular or IHL accu-mulation seems to be higher than that of visceral fat accumu-lation (Figure 2). In the rats examined, this probably explains the fact that the former variables better predict insulin resist-ance than latter. Moreover, after rats were divided into two groups based on their diet, there was relatively clear discrimi-nation between RCD and HFD groups in the scatter plot of IML/IHL against ISI, while there was substantial overlap in that of visceral fat against ISI. This finding implies that IML or IHL might be more vulnerable to diet intervention than visceral fat.

As seen in Figure 3, the pAkt expression levels in both mus-cle and liver were found to be less in HFD group than in RCD group. Importantly, the IML/IHL contents were negatively correlated with pAkt expression levels in each tissue. This find-ing suggests that IML/IHL content is closely associated with an index of tissue-specific insulin sensitivity.

Adiponectin has been shown to be closely related to glu-cose metabolism (43,44). We found that there was a higher correlation coefficient between adiponectin level and IML/IHL than between adiponectin and the amount of visceral fat, especially in the HFD group. In addition, we found that there was a significant negative correlation, especially in the HFD group, between IML/IHL content and levels of hsCRP, a well-known inflammatory marker (45). Thus, IML and IHL content was directly associated with adipocytokine concentration or low-grade inflammation, suggesting that IML and IHL content plays a key role in insulin resistance and atherosclerosis. In this study, insulin sensitivity was determined by a clamp study, and was correlated with lipid content in liver, muscle, and abdomen. This result was in accord with a study of nonobese, nondiabetic individuals (22).

Sinha et al. reported IMCL and extramyocellular lipid sep-arately in their human study using MRS, and showed a sig-nificant relationship between extramyocellular lipid and IMCL (46). Jacob et al. suggested that IMCL rather than extramyo-cellular lipid was closely related with insulin resistance (24). However, in our study, a clear distinction between extramyo-cellular lipid and IMCL was impossible because of the small size of the volume of interest.

On a cautionary note, it must be reminded that this is an ani-mal study and cross-sectional nature. Thus, caution is needed when translating this study result into humans.

In summary, a closer association was found between insulin sensitivity and the lipid content of metabolically active tissue, particularly muscle, than the amount of visceral fat. Although the measurement of lipid content in liver or muscle is not rou-tinely performed, we suggest that IHL and IML contents are closely related with insulin sensitivity, and might be a better metabolic index indicating insulin sensitivity than the amount of visceral fat.

acknoWledGMentsWe gratefully acknowledge the technical assistance of Cherl Namkoong. This study was partly supported by grants from the Korea Ministry of Health and Welfare (M10642140004-06N4214-00410), the IT R&D program of Ministry of Information and Communication/Institute for Information Technology Advancement (2006-S075-01, Development of a early diagnostic system of metabolic syndrome based on nanosensor integrated network computing), and the American Diabetes Association, 7-05-PPG-02 to Y.B.K.

© 2008 The Obesity Society

reFerences1. McGarry JD. Banting lecture 2001: dysregulation of fatty acid metabolism in

the etiology of type 2 diabetes. Diabetes 2002;51:7–18.2. Randle PJ, Garland PB, Hales CN, Newsholme EA. The glucose fatty-acid

cycle. Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1963;1:785–789.

3. Chalkley SM, Hettiarachchi M, Chisholm DJ, Kraegen EW. Five-hour fatty acid elevation increases muscle lipids and impairs glycogen synthesis in the rat. Metabolism 1998;47:1121–1126.

4. Oakes ND, Cooney GJ, Camilleri S, Chisholm DJ, Kraegen EW. Mechanisms of liver and muscle insulin resistance induced by chronic high-fat feeding. Diabetes 1997;46:1768–1774.

5. Storlien LH, James DE, Burleigh KM, Chisholm DJ, Kraegen EW. Fat feeding causes widespread in vivo insulin resistance, decreased energy expenditure, and obesity in rats. Am J Physiol 1986;251:E576–E583.

6. Storlien LH, Jenkins AB, Chisholm DJ et al. Influence of dietary fat composition on development of insulin resistance in rats. Relationship to muscle triglyceride and omega-3 fatty acids in muscle phospholipid. Diabetes 1991;40:280–289.

7. Unger RH, Orci L. Diseases of liporegulation: new perspective on obesity and related disorders. FASEB J 2001;15:312–321.

8. Kelley DE, Mandarino LJ. Fuel selection in human skeletal muscle in insulin resistance: a reexamination. Diabetes 2000;49:677–683.

9. Phillips DI, Caddy S, Ilic V et al. Intramuscular triglyceride and muscle insulin sensitivity: evidence for a relationship in nondiabetic subjects. Metabolism 1996;45:947–950.

10. Forouhi NG, Jenkinson G, Thomas EL et al. Relation of triglyceride stores in skeletal muscle cells to central obesity and insulin sensitivity in European and South Asian men. Diabetologia 1999;42:932–935.

11. Perseghin G, Scifo P, De CF 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 1999;48:1600–1606.

12. Banerji MA, Buckley MC, Chaiken RL et al. Liver fat, serum triglycerides and visceral adipose tissue in insulin-sensitive and insulin-resistant black men with NIDDM. Int J Obes Relat Metab Disord 1995;19:846–850.

13. Goto T, Onuma T, Takebe K, Kral JG. The influence of fatty liver on insulin clearance and insulin resistance in non-diabetic Japanese subjects. Int J Obes Relat Metab Disord 1995;19:841–845.

14. Larson-Meyer DE, Heilbronn LK, Redman LM et al. Effect of calorie restriction with or without exercise on insulin sensitivity, beta-cell function, fat cell size, and ectopic lipid in overweight subjects. Diabetes Care 2006;29:1337–1344.

15. Ryysy L, Hakkinen AM, Goto T et al. Hepatic fat content and insulin action on free fatty acids and glucose metabolism rather than insulin absorption are

obesity | VOLUME 17 NUMBER 1 | JANUARY 2009 195

articlesMethods and techniques

associated with insulin requirements during insulin therapy in type 2 diabetic patients. Diabetes 2000;49:749–758.

16. Pan DA, Lillioja S, Kriketos AD et al. Skeletal muscle triglyceride levels are inversely related to insulin action. Diabetes 1997;46:983–988.

17. Levin K, Daa SH, Alford FP, Beck-Nielsen H. Morphometric documentation of abnormal intramyocellular fat storage and reduced glycogen in obese patients with Type II diabetes. Diabetologia 2001;44:824–833.

18. He J, Watkins S, Kelley DE. Skeletal muscle lipid content and oxidative enzyme activity in relation to muscle fiber type in type 2 diabetes and obesity. Diabetes 2001;50:817–823.

19. Boesch C, Slotboom J, Hoppeler H, Kreis R. In vivo determination of intra-myocellular lipids in human muscle by means of localized 1H-MR-spectroscopy. Magn Reson Med 1997;37:484–493.

20. Petersen KF, Dufour S, Befroy D, Garcia R, Shulman GI. Impaired mitochondrial activity in the insulin-resistant offspring of patients with type 2 diabetes. N Engl J Med 2004;350:664–671.

21. Szczepaniak LS, Babcock EE, Schick F et al. Measurement of intracellular triglyceride stores by H spectroscopy: validation in vivo. Am J Physiol 1999;276:E977–E989.

22. Krssak M, Falk PK, Dresner A et al. Intramyocellular lipid concentrations are correlated with insulin sensitivity in humans: a 1H NMR spectroscopy study. Diabetologia 1999;42:113–116.

23. Virkamaki A, Korsheninnikova E, Seppala-Lindroos A et al. Intramyocellular lipid is associated with resistance to in vivo insulin actions on glucose uptake, antilipolysis, and early insulin signaling pathways in human skeletal muscle. Diabetes 2001;50:2337–2343.

24. Jacob S, Machann J, Rett K et al. Association of increased intramyocellular lipid content with insulin resistance in lean nondiabetic offspring of type 2 diabetic subjects. Diabetes 1999;48:1113–1119.

25. Boden G, Lebed B, Schatz M, Homko C, Lemieux S. Effects of acute changes of plasma free fatty acids on intramyocellular fat content and insulin resistance in healthy subjects. Diabetes 2001;50:1612–1617.

26. Bachmann OP, Dahl DB, Brechtel K et al. Effects of intravenous and dietary lipid challenge on intramyocellular lipid content and the relation with insulin sensitivity in humans. Diabetes 2001;50:2579–2584.

27. Petersen KF, Shulman GI. Etiology of insulin resistance. Am J Med 2006;119:S10–S16.

28. Fujimoto WY, Newell-Morris LL, Grote M, Bergstrom RW, Shuman WP. Visceral fat obesity and morbidity: NIDDM and atherogenic risk in Japanese American men and women. Int J Obes 1991;15(Suppl 2):41–44.

29. Yamashita S, Nakamura T, Shimomura I et al. Insulin resistance and body fat distribution. Diabetes Care 1996;19:287–291.

30. Yoshida S, Inadera H, Ishikawa Y et al. Endocrine disorders and body fat distribution. Int J Obes 1991;15(Suppl 2):37–40.

31. Ross R, Leger L, Guardo R, De GJ, Pike BG. Adipose tissue volume measured by magnetic resonance imaging and computerized tomography in rats. J Appl Physiol 1991;70:2164–2172.

32. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979;237:E214–E223.

33. Koenig W, Sund M, Frohlich M et al. C-Reactive protein, a sensitive marker of inflammation, predicts future risk of coronary heart disease in initially healthy middle-aged men: results from the MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg Cohort Study, 1984 to 1992. Circulation 1999;99:237–242.

34. Weiss R, Dufour S, Groszmann A et al. Low adiponectin levels in adolescent obesity: a marker of increased intramyocellular lipid accumulation. J Clin Endocrinol Metab 2003;88:2014–2018.

35. Weiss R, Dufour S, Taksali SE et al. Prediabetes in obese youth: a syndrome of impaired glucose tolerance, severe insulin resistance, and altered myocellular and abdominal fat partitioning. Lancet 2003;362:951–957.

36. Perseghin G, Lattuada G, Danna M et al. Insulin resistance, intramyocellular lipid content, and plasma adiponectin in patients with type 1 diabetes. Am J Physiol Endocrinol Metab 2003;285:E1174–E1181.

37. Kautzky-Willer A, Krssak M, Winzer C et al. Increased intramyocellular lipid concentration identifies impaired glucose metabolism in women with previous gestational diabetes. Diabetes 2003;52:244–251.

38. Greco AV, Mingrone G, Giancaterini A et al. Insulin resistance in morbid obesity: reversal with intramyocellular fat depletion. Diabetes 2002;51: 144–151.

39. Bjorntorp P. Liver triglycerides and metabolism. Int J Obes Relat Metab Disord 1995;19:839–840.

40. Kuhlmann J, Neumann-Haefelin C, Belz U et al. Intramyocellular lipid and insulin resistance: a longitudinal in vivo 1H-spectroscopic study in Zucker diabetic fatty rats. Diabetes 2003;52:138–144.

41. Yu JG, Javorschi S, Hevener AL et al. The effect of thiazolidinediones on plasma adiponectin levels in normal, obese, and type 2 diabetic subjects. Diabetes 2002;51:2968–2974.

42. Jensen MD. Is visceral fat involved in the pathogenesis of the metabolic syndrome? Human model. Obesity (Silver Spring) 2006;14(Suppl 1): 20S–24S.

43. Weyer C, Funahashi T, Tanaka S et al. Hypoadiponectinemia in obesity and type 2 diabetes: close association with insulin resistance and hyperinsulinemia. J Clin Endocrinol Metab 2001;86:1930–1935.

44. Hotta K, Funahashi T, Arita Y et al. Plasma concentrations of a novel, adipose-specific protein, adiponectin, in type 2 diabetic patients. Arterioscler Thromb Vasc Biol 2000;20:1595–1599.

45. Lim S, Jang HC, Lee HK et al. The relationship between body fat and C-reactive protein in middle-aged Korean population. Atherosclerosis 2006;184:171–177.

46. Sinha R, Dufour S, Petersen KF et al. Assessment of skeletal muscle triglyceride content by (1)H nuclear magnetic resonance spectroscopy in lean and obese adolescents: relationships to insulin sensitivity, total body fat, and central adiposity. Diabetes 2002;51:1022–1027.