SEARCH

SEARCH BY CITATION

Keywords:

  • fat distribution;
  • non-alcoholic fatty liver;
  • thigh fat;
  • visceral fat

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

Background and Aim:  Some people have a fatty liver despite having low visceral fat and a low body mass index (BMI). We investigated whether fat distribution, especially thigh subcutaneous fat and thigh intramuscular fat, is associated with non-alcoholic fatty liver disease (NAFLD).

Methods:  The patients consisted of 408 men and women. NAFLD was defined by an ultrasound scan and excluded other liver diseases. Visceral, subcutaneous abdominal, intramuscular, and subcutaneous thigh adipose tissue was measured by computed tomography.

Results:  The frequency of NAFLD decreased over a quartile of thigh fat independently of BMI in the female patients. Additional adjustments for age and visceral fat area did not change the results. This finding was not observed in the male patients. To investigate the relationship between each fat distribution and NAFLD, we performed a logistic regression analysis. Fat distribution was divided into four groups: visceral fat, abdominal subcutaneous fat, thigh subcutaneous fat, and thigh intramuscular fat. All four fat components were chosen as variables for the regression model. Age, BMI, and the homeostasis model assessment (HOMA) index were then adjusted successively. A larger subcutaneous fat area was negatively associated with NAFLD after adjustment for visceral fat and abdominal subcutaneous fat areas in women, but not in men. It did not change even after age adjustment, BMI, and the HOMA index.

Conclusion:  Low femoral subcutaneous fat amounts were shown to be independently associated with fatty liver disease in women. These results show the importance of accurate measurements of other regional body compositions as well as visceral fat amounts when investigating NAFLD.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

Non-alcoholic fatty liver disease (NAFLD) is considered to be one of the phenotypes of metabolic syndrome and is characterized by obesity, insulin resistance, and dyslipidemia. The majority of NAFLD occur in obese or overweight individuals.1 However, not all NAFLD patients are obese. Several studies have reported that less than 50% of patients with NAFLD were obese,2,3 and the median prevalence rate of obesity in NAFLD patients was 71%, with a range of 57–93%.4,5 Even though Asians generally have a low body mass index (BMI), the prevalence of NAFLD was estimated to be 9.3–44.2%.6 Studies in Hong Kong and Singapore showed that the risk of developing cardiovascular disease or diabetes at a lower BMI was higher for people in Hong Kong and Singapore than Western people.7,8 It seems that Asian populations have a high percentage of body fat at a low BMI and have a different fat distribution from Western people.7,9 It means that there are differences in the relationship between BMI and the percentage of body fat or fat distribution, so the BMI itself can not represent the metabolically-active fat amount.

Increasing data suggest that fat distribution can be correlated with the development of metabolic syndrome more than BMI or the total fat amount itself. A recent study shows that thigh intramuscular fat and subcutaneous thigh fat were positively and negatively associated with lipid profile and insulin resistance, respectively.10 However, previous studies focused on metabolic syndrome not NAFLD, and the fat amount was measured by dual energy X-ray absorptiometry method or the waist/hip ratio.11,12 There are few studies on the relationship between fat distributions, including femoral subcutaneous fat, intramuscular fat and visceral fat, and hepatic steatosis using computed tomography (CT). We investigated whether fat distribution, especially thigh subcutaneous, intramuscular fat, and visceral fat, is associated with NAFLD independently of BMI.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

Study population

We studied consecutive patients who visited an obesity clinic and gastroenterology department for a thorough regular check-up or elevated liver enzyme. Of the 1068 patients who visited the hospital, 459 patients agreed to undergo fat CT. Among them, 51 diabetic patients were excluded. In the end, a total of 408 patients were enrolled in our study. Fatty liver was diagnosed based on sonographic findings. The definition of NAFLD is as follows: (i) patients with an alcohol consumption ≤20 g/day; (ii) patients who were not taking drugs, such as any herbal medication, amiodarone, methotrexate, synthetic estrogens, nucleoside analogs, and glucocorticoids within 3 months; (iii) negative study for viral hepatitis, autoimmune hepatitis, primary biliary cirrhosis, drug-induced liver disease, and thyroid disease; and (iv) patients with an aspartate aminotransferase/alanine aminotransferase ratio less than 1.0.13,14 Because the study involved data that were routinely collected, consent was not specifically obtained. The institutional review boards of the hospital approved the study.

Clinical assessments

Complete blood cell count, liver profile, lipid profile, fasting insulin, hepatitis B and C viral markers, and the thyroid hormone level were routinely checked. The antinuclear antibody, antimitochondrial antibody, serum iron, and total iron binding capacity were also checked to rule out other liver disease in patients with abnormal liver enzyme. BMI was calculated by dividing weight in kilograms by height in meters squared. Waist circumference, defined as the midpoint between the lower rib margin and the iliac crest, was measured as an index of regional fat distribution. Information on daily alcohol consumption, medication history within 3 months, and other lifestyle characteristics was obtained from the questionnaires given to all participants. The participants reported on how many standard glasses of beer or distilled spirits they had consumed per day or per week during the 12 months prior to the recruitment. Ethanol intake was calculated on the basis of average glass volume, and ethanol content for each type of alcoholic beverage.

Following a 12 h fast, blood was obtained in the morning. Fasting glucose was measured using a glucose oxidase method, and total cholesterol, triglycerides, and high-density lipoprotein–cholesterol (HDL-C) levels were measured using enzymatic colorimetric procedures with an autoanalyzer (Hitachi-747; Hitachi, Tokyo, Japan). The serum insulin concentration was measured by chemiluminescent immunoassay (Immulite 2000, Diagnostic Products, Los Angeles, CA, USA; coefficient of variation, <7%). Insulin resistance was measured using the homeostasis model assessment (HOMA) of the insulin resistance index (fasting glucose [mmol/L] × fasting insulin [uU/mL]/22.5).

Fatty liver was diagnosed by characteristic echo-patterns, according to conventional criteria of at least two of the following: increased contrast of the hepatic compared with the renal parenchyma, vascular blurring, focal sparing, or narrowing of the lumen of the hepatic veins. All sonographic findings were blinded and evaluated twice regarding grading of fatty liver by two radiologists.

Measurement of abdominal visceral and femoral adiposity

CT scans of the abdomen and thighs were acquired using Light speed ultra 16 (General Electric Medical Systems, Milwaukee, WI, USA), as described previously.15,16 Briefly, the scans were completed at 120 kVp, 200–250 mAs, and the slice thickness was set at 10 mm. The scan of the abdomen was performed at the level of umbilicus in the supine position. The scan at mid-thigh level was performed at the midpoint between the medial edge of the greater trochanter and the intercondylar fossa. CT images were reviewed on an Advantage workstation (General Electric Medical Systems, USA) for the determination of adipose and muscle tissue areas using Voxtool 3.0.58 software (General Electric Medical Systems, USA) by a radiologist. Fat tissue, muscle tissue, and bone were distinguished by their particular range of tissue density in Hounsfield units (HU). Visceral fat tissue was manually separated from subcutaneous fat tissue by tracing along the fascial plane defining the internal abdominal wall. In the thigh, the total subcutaneous fat area was calculated by the previous method.17 To estimate the intermuscular fat amount, HU of the thigh muscle were used. The total area of non-adipose, non-bone tissue within the deep fascial plane was used as a measure of muscle area.

Statistical methods

Data were expressed as mean ± SD. Differences between groups were examined for statistical significance using the Student's t-test, ANOVA, and the Pearson correlation test. A multivariate analysis was performed using a logistic regression analysis. All analyzes were performed by use of SPSS for Windows version 10.1 (SPSS, Chicago, IL, USA). A P < 0.05 was considered significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

Patient characteristics

A total of 408 patients were included in the study. Table 1 shows the background characteristics of the 408 patients. The mean age of the patients was 46.9 ± 10.9 years; 176 (43.1%) patients s were male and 232 (56.9%) were female; the mean BMI was 25.9 ± 3.4 kg/m2. Despite being of similar BMI, the male and female patients showed vastly different characteristics. In the female patients, the abdominal subcutaneous fat area (306.4 ± 102.0 vs 263.8 ± 89.0, P = 0.001) and thigh subcutaneous fat area (84.7 ± 29.0 vs 46.3 ± 19.1, P = 0.001) were greater than that of the male patients. However, the abdominal visceral fat area (111.3 ± 44.4 vs 88.7 ± 43.0, P = 0.001) and thigh muscle area (157.2 ± 25.6 vs 109.2 ± 18.0, P = 0.001) were greater in the male patients than the female patients. Table 1 presents the characteristics of the participants, with and without NAFLD separately, according to sex. Most of the metabolic parameters, including total cholesterol, HDL-C, triglyceride, free fatty acid, apolipoprotein B level, and the HOMA index showed more unfavorable values in NAFLD participants than in non-fatty liver participants.

Table 1.  Descriptive characteristics by NAFLD status
 Men (n = 176)Women (n = 232)
NAFLD (n = 119)Control (n = 57)P*NAFLD (n = 91)Control (n = 141)P*
  • *

    Significant at P < 0.05 by the Student t-test. Values are expressed as mean ± SD. HDL-C, high-density lipoprotein–cholesterol; HOMA, homeostatic model assessment; NAFLD, non-alcoholic fatty liver disease.

Age (year)45.1 ± 9.045.8 ± 12.30.6651.0 ± 11.346.3 ± 10.9<0.01
Body mass index (kg/m2)26.8 ± 2.724.7 ± 2.7<0.0126.9 ± 4.125.1 ± 3.4<0.01
Waist circumference (cm)92.4 ± 6.785.1 ± 7.7<0.0187.9 ± 9.682.7 ± 9.3<0.01
Hip circumference (cm)102.7 ± 5.5100.2 ± 5.80.08101.0 ± 7.899.2 ± 7.30.21
Waist/hip ratio0.90 ± 0.040.89 ± 0.040.250.89 ± 0.040.86 ± 0.05<0.01
Abdomen
 Total fat area (cm2)291.5 ± 80.8205.9 ± 76.9<0.01339.4 ± 102.4285.1 ± 96.1<0.01
 Visceral fat area (cm2)124.5 ± 41.783.7 ± 36.5<0.01110.8 ± 44.074.4 ± 35.7<0.01
 Subcutaneous fat area (cm2)166.9 ± 59.0122.2 ± 52.6<0.01228.6 ± 79.4210.6 ± 79.40.09
Thigh
 Thigh fat area (cm2)48.9 ± 18.640.7 ± 19.0<0.0182.1 ± 30.188.6 ± 28.10.24
 Thigh muscle area (cm2)161.1 ± 24.9148.7 ± 25.0<0.01113.5 ± 18.9106.3 ± 16.8<0.01
 Hounsfield unit of thigh muscle48.8 ± 3.049.0 ± 3.30.6843.4 ± 4.245.1 ± 3.9<0.01
Laboratory findings
 Alanine aminotransferase (IU)51.6 ± 36.527.8 ± 24.2<0.0128.3 ± 15.223.0 ± 18.10.02
 Aspartate aminotransferase (IU)32.8 ± 15.824.7 ± 8.7<0.0133.5 ± 25.620.9 ± 10.6<0.01
 Total cholesterol (mg/dL)206.3 ± 37.4188.5 ± 27.0<0.01206.1 ± 37.6196.4 ± 36.40.05
 Triglyceride (mg/dL)181.9 ± 101.1114.5 ± 65.1<0.01159.0 ± 89.8109.5 ± 59.4<0.01
 HDL-C (mg/dL)43.1 ± 8.049.9 ± 10.5<0.0148.4 ± 9.154.5 ± 10.4<0.01
 Free fatty acid (mg/dL)798.7 ± 348.2605.7 ± 265.6<0.01772.1 ± 289.1674.1 ± 267.7<0.01
 Apolipoprotein B (mg/dL)111.9 ± 29.786.5 ± 22.7<0.0197.1 ± 37.486.4 ± 24.20.01
 HOMA3.50 ± 2.02.19 ± 1.1<0.013.80 ± 2.72.20 ± 0.90.07

Relationship between fat distribution and fatty liver

Based on the subcutaneous thigh fat and abdominal visceral fat areas, the patients were divided into four groups, and BMI was divided into three groups, respectively. It was clearly shown that the frequency of NAFLD increased with BMI and visceral fat area. The frequency of NAFLD increased with the visceral fat area, even in the normal BMI patients (Fig. 1). Figure 2 shows the frequency of NAFLD by combined stratification of thigh fat area and BMI. In women, but not in men, the frequency of NAFLD decreased over a quartile of thigh fat independently of BMI. Additional adjustments for age and visceral fat area did not change the results (data not shown). This finding was not observed in the male patients.

image

Figure 1. Frequency of non-alcoholic fatty liver disease (NAFLD) according to body mass index (BMI) and visceral fat quartiles.

Download figure to PowerPoint

image

Figure 2. Frequency of non-alcoholic fatty liver disease (NAFLD) in men (a) and women (b) within quartiles of thigh subcutaneous fat area and body mass index (BMI).

Download figure to PowerPoint

To investigate the relationship between each fat distribution and NAFLD, we performed a logistic regression analysis. We focused on fat distribution and NAFLD. Fat was divided into four groups: visceral fat, abdominal subcutaneous fat, thigh subcutaneous fat, and thigh intramuscular fat. The amount of intramuscular thigh fat was estimated by the HU of the thigh muscle. At first, all four fat compartments were chosen as variables for the Model 1 regression mode, then age, BMI, and the HOMA index were adjusted successively (Models 2 & 3). A larger thigh subcutaneous fat area was negatively associated with NAFLD after adjustment for visceral fat and abdominal subcutaneous fat area in women (Model 1). It did not change even after adjusted age, BMI, and the HOMA index (Models 2 & 3). Subcutaneous thigh fat was negatively associated with NAFLD independently of the visceral fat area in women. It means that the low thigh fat area was strongly associated with NAFLD independently of visceral fat, age, and the HOMA index (Table 2). There was a clear positive relationship between the frequency of NAFLD and quartile of visceral fat area in both the male and female patients after age and BMI-adjustment. However, there was a negative relationship between the odd ratios and the quartile of thigh fat area in the female patients (Table 3).

Table 2.  Associations of fat distributions and other factors with NAFLD
  MenWomen
  • *

    < 0.05, odds ratios with 95% confidence intervals from logistic regression analysis. HOMA, Homeostatic model assessment; NAFLD, non-alcoholic fatty liver disease.

Model 1Visceral fat area (cm2)1.38* (1.20–1.58)1.28* (1.15–1.43)
Abdominal subcutaneous fat area (cm2)1.17* (1.04–1.32)1.05 (0.99–1.12)
Thigh fat area (cm2)0.83 (0.59–1.16)0.77* (0.66–0.91)
Hounsfield unit of thigh muscle7.74* (1.89–31.66)0.74 (0.32–1.72)
Model 2Visceral fat area (cm2)1.45* (1.23–1.72)1.21* (1.06–1.37)
Abdominal subcutaneous fat area (cm2)1.12 (0.98–1.26)1.01 (0.923–1.08)
Thigh fat area (cm2)0.76 (0.52–1.12)0.71* (0.59–0.86)
Hounsfield unit of thigh muscle7.35* (1.57–34.41)0.75 (0.29–1.87)
Age (year)0.96 (0.92–1.01)1.00 (0.96–1.04)
Body mass index (m2/kg)1.15 (0.92–1.45)1.26* (1.03–1.52)
Model 3Visceral fat area (cm2)1.38* (1.14–1.66)1.10 (0.96–1.25)
Abdominal subcutaneous fat area (cm2)1.09 (0.94–1.26)1.00 (0.92–1.09)
Thigh fat area (cm2)0.94 (0.644–1.39)0.73* (0.59–0.89)
Hounsfield unit of thigh muscle9.46* (1.66–53.83)1.11 (0.39–3.09)
Age (year)0.98 (0.93–1.04)1.02 (0.97–1.06)
Body mass index (m2/kg)1.11 (0.85–1.43)1.19 (0.97–1.47)
HOMA index1.04* (1.00–1.08)1.02* (1.02–1.09)
Table 3.  Multivariate regression analysis of NAFLD according to abdominal visceral fat and thigh fat quartile stratified by sex
Model 2 FirstSecondThirdFourth*P for trend
  • *

    < 0.05, Hounsfield unit of thigh muscle, abdominal subcutaneous fat area, age, and body mass index-adjusted odds ratios with 95% confidence intervals from logistic regression analysis (Model 2). NAFLD, non-alcoholic fatty liver disease.

MenAbdominal visceral fat1.019.58 (1.76–217.45)57.94 (5.10–658.28)93.78 (7.13–1232.7)<0.01
Thigh fat1.00.73 (0.25–2.07)0.01 (0.01–3.65)0.08 (0.01–2.28)0.53
WomenAbdominal visceral fat1.03.73 (1.42–9.81)4.45 (1.53–12.93)10.08 (2.88–35.26)<0.01
Thigh fat1.00.33 (0.07–1.56)0.13 (0.03–0.615)0.07 (0.01–0.40)<0.01
TotalAbdominal visceral fat1.05.16 (2.27–11.72)9.26 (3.88–22.11)16.96 (6.35–45.30)<0.01
Thigh fat1.00.94 (0.46–1.91)0.55 (0.25–1.24)0.16 (0.05–0.46)<0.01

Relationship between fat distribution and metabolic risk factors

Table 4 shows the correlation between body fat parameters and the metabolic risk factors: triglyceride, fasting glucose, free fatty acid, HOMA, γ-glutamyltransferase, and apolipoprotein B. In both men and women, the visceral fat area correlated most strongly with HOMA among the different risk factors. In general, the visceral fat area was more highly correlated univariately with risk factors than the total abdominal fat area. However, the thigh fat area showed a negative correlation with most of the metabolic risk factors in both men and women. Table 4 also gives partial correlation coefficients for each after adjustment for BMI and age. Partial correlation for the thigh fat area often showed modest additional and significant negative correlations with risk factors after adjustment for BMI and age.

Table 4.  Univariate and partial correlations between fat distributions and metabolic risk factors
Body fat distributionTGFBSFFAHOMAGGTApo B
  1. Pearson correlation coefficients (partial correlation coefficients adjusted for body mass index and age), *< 0.05, < 0.01. Apo B, apolipoprotein B (mg/dL); FBS, Fasting blood sugar; FFA, Free fatty acid; GGT, γ-glutamyltransferase; HOMA, homeostatic model assessment; TG, triglyceride.

Men (n = 176)
 Body mass index0.11−0.090.16*0.370.15*0.37
 Abdominal total fat area (cm2)0.13 (0.19)0.01 (0.17)0.19 (0.04)0.42 (0.30)0.21 (0.21*)0.36 (0.10)
 Visceral fat area (cm2)0.15* (0.18)0.11 (0.21*)0.14 (−0.04)0.39 (0.33)0.23 (0.21*)0.30 (0.09)
 Abdominal subcutaneous fat area (cm2)0.09 (0.10)−0.06 (0.03)0.18* (0.10)0.34 (0.10)0.15* (0.09)0.31 (−0.05)
 Thigh fat area (cm2)0.05 (−0.12)−0.14* (−0.23*)0.09 (−0.11)0.17* (−0.04)0.09 (−0.13)0.25* (0.17)
 Hounsfield unit of thigh muscle−0.01 (−0.01)−0.12 (−0.02)−0.14 (0.04)−0.16* (−0.04)0.01 (0.03)−0.07 (−0.13)
 Waist/hip ratio0.09 (−0.05)0.02 (0.02)−0.05 (−0.13)0.15 (0.19)0.17 (0.14)0.14 (0.04)
Women (n = 232)
 Body mass index0.180.010.15*0.390.070.07
 Abdominal total fat area (cm2)0.18 (−0.05)0.01 (0.14)0.22 (0.22)0.37 (−0.10)0.07 (0.04)0.03 (−0.10)
 Visceral fat area (cm2)0.39 (0.29*)0.15* (0.35)0.20 (0.08)0.54 (0.46)0.12 (0.12)0.24* (0.24*)
 Abdominal subcutaneous fat area (cm2)0.02 (−0.24*)−0.08 (−0.08)0.17 (0.13)0.20 (−0.40)0.03 (−0.04)−0.09 (−0.25*)
 Thigh fat area (cm2)−0.09 (−0.37*)−0.17 (−0.21*)0.01 (−0.18)0.10 (−0.43)−0.05 (−0.16)−0.14 (−0.27*)
 Hounsfield unit of thigh muscle−0.10 (0.08)−0.11 (−0.01)−0.20 (−0.06)−0.14 (0.01)−0.05 (0.05)−0.15 (0.03)
 Waist/hip ratio0.28 (−0.18)0.23 (0.19)0.19 (−0.13)0.32 (0.11)0.13 (0.12)0.30 (−0.08)

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

It has already been reported that women have a more favorable metabolic profile than men having similar BMI.18 Moreover, in epidemiological studies, adiposity is associated with increased prevalence of cardiovascular risk factors, including hypertension, type 2 diabetes mellitus in men than women.19,20 The mechanisms underlying this sex difference in metabolic syndrome are not clear, even though there are some possible mechanisms that have been identified. First, men and women have a different body shape. Women typically have a lower body adipose distribution (gynoid) characterized by fat deposition in the gluteofemoral region. However, men have an upper body adipose distribution (android), which is characterized by central fat deposition in the abdominal intraperitoneal regions. In our study, women showed a larger femoral subcutaneous fat area and a smaller visceral fat area than men, and it seemed that visceral fat and thigh subcutaneous fat demonstrated different effects on NAFLD incidence independently of BMI and visceral fat mass. It has been suggested that fat might have opposite effects depending on its location. It might originate from differences in lipolytic activity between abdominal subcutaneous fat and subcutaneous thigh fat. Subcutaneous thigh fat is more likely to take up non-esterified fatty acid (NEFA). Therefore, it has been proposed that subcutaneous thigh fat acts as a “metabolic sink” for circulating NEFA.21 These hypotheses were supported by the Hong Kong study which found that females with a relatively small hip circumference had an increased risk of developing diabetes and dyslipidemia, whereas those with a larger hip circumference had more protection.22

The second possible mechanism is the “insulin–leptin axis”. Oda et al.23 explained the mechanism of this sex difference as the insulin–leptin axis. Leptin resistance correlates better with subcutaneous fat and women show distinctly higher leptin levels than men. Moreover, insulin resistance highly correlates with visceral fat. In our results, the women had smaller visceral fat area and a larger abdominal and thigh subcutaneous fat area than men. This means that the visceral and subcutaneous fat ratio might affect the insulin–leptin axis. Fatty liver was divided into three groups (mild, moderate, and severe) according to the degree of steatosis. We also evaluated the relationship between fat distribution and the degree of sonographic steatosis. The visceral fat area showed a strong positive association with the degree of steatosis in both males and females. The femoral subcutaneous fat area tended to negatively correlate with the degree of steatosis, but no statistically significant difference was observed.

Few studies have reported evidence for an association between intramuscular adipose tissue and metabolic abnormalities.24,25 One potential mechanism of action linking intramuscular fat with insulin signaling is through triacylglycerol metabolites interfering with insulin signaling transduction, thereby altering whole-body glucose and lipid metabolism. We also checked the thigh muscle density in HU at the mid-thigh level to assess the intramuscular fat amount. The low HU of muscle is correlated with the lipid content of muscle fibers;26 the lower HU of the thigh muscles indicates a higher intramuscular lipid content. In this study, we did not find any relationship between the intramuscular fat amount and fatty liver in men and women. We used the single CT mid-thigh slice method. It has methodological limitations because it is not clear how well total body intramuscular fat is represented by a single CT slice. A comprehensive evaluation of whole body intramuscular adipose tissue is required, such as whole-body magnetic resonance imaging (MRI) or CT scan.

The major limitation of this study is that the diagnosis of NAFLD was made on clinical history and ultrasound findings only. The gold standard for the diagnosis of NAFLD is liver biopsy. At present, MRI may be the most reliable, quantitative technique and could be preferable. However, ultrasound scanning, when positive, can give a degree of diagnostic certainty. Four studies have examined the sensitivity and specificity of ultrasound imaging for recognizing fat; the sensitivity and specificity of ultrasound images for recognizing fatty liver were 60–94% and 84–95%, respectively, as assessed by liver biopsy.27–30 Moreover, a good correlation has also been shown between quantitative ultrasound attenuation and the quantitative histological assessment of liver fat using calibrated phantoms.31 In the present study, all sonographic findings were reviewed by two radiologists to minimize interobserver variation. The agreement of fatty liver diagnosis between the two radiologists was 82% (κ = 0.64).

Our main finding in this study was that femoral subcutaneous fat amounts were shown to be independently associated with fatty liver disease in women. This result indicates the importance of accurate measurements of other regional body compositions as well as visceral fat amounts when investigating NAFLD.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References