SEARCH

SEARCH BY CITATION

Keywords:

  • fatty liver;
  • gender difference;
  • liver spleen ratio;
  • visceral fat;
  • waist circumference

Abstract

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

Background and Aim:  Abdominal obesity, a component of metabolic syndrome, is a major risk factor for non-alcoholic fatty liver disease (NAFLD). In recent worldwide definitions of metabolic syndrome, waist measurement has been proposed as a simple and useful estimate of abdominal obesity, taking into account gender differences in waist circumference. The present cross-sectional study investigated the correlation of hepatic fat accumulation and waist circumference in Japanese NAFLD patients to determine if there are gender differences in this relationship.

Methods:  Consecutive patients (n = 2111) who had at least one of two criteria for liver disease (alanine aminotransferase [ALT] level >30 IU/mL and aspartate aminotransferase [AST]/ALT ratio <1) underwent abdominal ultrasonography. Patients positive for hepatitis B virus, hepatitis C virus or autoimmune antibodies and whose alcohol intake was >20 g/day were excluded. Patients with NAFLD underwent abdominal computed tomography. Hepatic fat accumulation was estimated by liver/spleen attenuation ratio (L/S ratio) and visceral adipose accumulation was measured as visceral fat area (VFA) at the umbilical level.

Results:  Of the 221 NAFLD patients, 103 were females. In males, the relationship between L/S ratio and waist circumference was negative (r =−0.356, P < 0.01), and there was no correlation in the female group. The relationship between L/S ratio and VFA was negative in both groups (males: r = −0.269, P < 0.01; females: r = −0.319, P < 0.01). Subcutaneous fat area/total fat area ratio at the umbilical level was larger in females than in males (P < 0.01).

Conclusions:  In NAFLD patients, waist measurement is more susceptible to gender differences than VFA.


Introduction

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

Concomitant with the Westernized lifestyle in Japanese, the incidence of non-alcoholic fatty liver disease (NAFLD) has been increasing in Japan.1,2 It has been suggested that simple fatty liver, the most common type of NAFLD, is a precursor of non-alcoholic steatohepatitis (NASH), which can progress to liver cirrhosis and hepatocellular carcinoma.3,4 NAFLD is considered to be one of the phenotypes of metabolic syndrome, which is characterized by abdominal obesity, type 2 diabetes mellitus, hyperlipidemia and hypertension,5 and it is well recognized that accumulation of visceral fat is a more important risk factor for metabolic syndrome than subcutaneous fat.6,7

NAFLD has few symptoms and is clinically characterized by a slight elevation of the serum aminotransferase, hyperglycemia and insulin resistance with obesity.2,8,9 The diagnosis can be made by liver biopsy and/or by imaging, such as abdominal computed tomography (CT), magnetic resonance imaging (MRI), magnetic resonance spectroscopy and ultrasonography (US).10,11 Liver biopsy is the gold standard, especially for NASH, for evaluation of steatosis, inflammation and fibrosis, although it is invasive and sometimes there is incidental peritoneal bleeding. Non-invasive evaluation of liver fat accumulation is mainly performed with abdominal CT, using liver/spleen attenuation ratio (L/S ratio). Previous studies have indicated that visceral fat area (VFA) evaluated by CT at the umbilical level correlates with visceral fat volume, and that accumulation of risk factors for metabolic syndrome is related to an increase in VFA of more than 100 cm2.12,13

Waist circumference is used as an index of visceral fat deposition in the diagnosis of metabolic syndrome, because of its ease of measurement.14–16 The International Diabetes Federation Consensus of the Worldwide Definition of the Metabolic Syndrome states that ethnic and gender differences in waist circumference require cut-off measurements to be developed for each country.15 We previously reported that the degree of hepatic fat deposition correlates with VFA, as measured using abdominal CT, in patients with NAFLD.17 However, there are temporal and financial considerations to performing mass abdominal CT scan for the measurement of VFA, so the primary aim of the present study was to investigate the correlation between hepatic fat deposition and waist circumference in Japanese patients with NAFLD and, second, whether there is any influence of gender differences in waist measurement.

Methods

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

Study population

Consecutive Japanese patients who visited Eguchi Hospital between January 2004 and December 2005 with suspected liver disease (at least one of two criteria: alanine aminotransferase [ALT] level >30 IU/mL and aspartate aminotransferase [AST]/ALT level ratio <1) were recruited (n = 2111). All patients underwent abdominal US and those with evidence of excessive alcohol intake (>20 g/day) or other causes of liver disease (hepatitis C virus [HCV] antibody or hepatitis B surface antigen [HBs], autoimmune liver disease, biliary disease) or cirrhosis or malignant disease or use of antihypertensive or antidiabetic agents, or very elderly patients (age > 75 years) were excluded. NAFLD was ultrasonographically diagnosed by two experienced doctors (YE and NO) who were unaware of each subject's clinical and laboratory findings. Informed consent had been acquired for the use of data for research purposes, and written informed consent was obtained from eligible patients with NAFLD following explanation of the study regimen. The study protocol was approved and supervised by the ethical committee of Eguchi Hospital.

Definition of NAFLD

A 4 MHz transducer (LOGIQ 7; GE Healthcare, Wauwatosa WI, USA) was used to observe sagittal and transverse views of the left lobe of the liver and the spleen, subcostal views of the right lobe of the liver and the right kidney, and intracostal views of the right lobe of the liver, the portal vein (right branch) and the hepatic vein (right and middle branches). NAFLD was diagnosed according to the following criteria: slight diffuse increase in bright homogeneous echoes (i.e. ‘bright liver’) in the liver parenchyma with normal visualization of the diaphragm and portal and hepatic vein borders, and normal hepato-renal contrast of echogenicity; diffuse increase in bright echoes in the liver parenchyma with slightly impaired visualization of the peripheral portal and hepatic vein borders (i.e. ‘vascular blurring’); marked increase in bright echoes at a shallow depth with deep attenuation and impaired visualization of the diaphragm and marked vascular blurring.10 Patients who were diagnosed as having NAFLD underwent further CT scan.

Abdominal CT protocol and assessment

Unenhanced spiral acquisition through the liver was obtained during a breath-hold at 5.0 mm collimation, 15.0 mm/rotation table speed (HQ mode, pitch 1:3), 120 kV (p), and auto mA (Light speed QXi; GE Healthcare). Images were reconstructed at 10 mm increments. All patients underwent abdominal CT in the morning after a 12 h overnight fast. The CT numbers (in Hounsfield units) of the region of interest (ROI) of 120 mm2 in the periphery of the liver and the spleen, away from major vessels, were measured at five points in each organ, and the mean numbers were used to determine L/S ratio as an index of fat accumulation in the liver.10 In addition, subcutaneous fat area (SFA) and VFA (both in cm2) were measured at the umbilical level and calculated using a computer software program (Fat Scan; N2 System Co., Osaka, Japan).12

Physical examination and serum biochemistry

Bodyweight (kg) and height (m) were obtained with NAFLD patients clothed in a light gown without shoes and used to calculate the body mass index (BMI) (kg/m2). Waist circumference (cm) was measured at the umbilicus. Venous blood samples were taken from all patients before 9.00 h after the 12-h overnight fast to determine the levels of AST (IU/L), ALT (IU/L), total cholesterol (T-CHO; mg/dL), triglyceride (TG; mg/dL), plasma glucose (PG; mg/dL) and plasma insulin (μU/mL; using an enzyme immunoassay [Dainabot, Tokyo, Japan]). All patients underwent a 75-g oral glucose tolerance test and type 2 diabetes mellitus was diagnosed according to World Health Organization criteria.18 Impaired fasting glycemia (IFG) was defined as fasting plasma glucose ≥110 and <126 mg/dL; impaired glucose tolerance (IGT) was defined as 2-h post load ≥140 and <200 mg/dL; diabetes mellitus was defined as fasting plasma glucose ≥126 mg/dL or history of diabetes mellitus or current use of antidiabetic drugs. Insulin resistance was calculated as the homeostasis model (HOMA-IR) using the following formula: HOMA-IR = fasting insulin (μU/mL) × PG (mg/dL)/405.19 Patients who were defined as having diabetes mellitus, IFG, and IGT were not included in this evaluation.

Statistical analysis

Continuous variables were summarized as the mean ± SD. Pearson's correlation coefficient analysis was used to compare L/S ratio, waist circumference, VFA and other variables. Differences were considered significant at P < 0.05. All analyses were carried out using Statview for Windows (Version 5.0; SAS Institute, Cary, NC, USA).

Results

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

Patients characteristics

Two hundred and twenty-one patients (118 males, 103 females) with NAFLD, who met the inclusion criteria, were included in the analysis. Table 1 shows the background characteristics of the 221 patients. In male patients, the mean age was 53.0 ± 12.7 years (range: 19.0–75.0 years); the mean BMI was 25.8 ± 3.2 kg/m2 (range: 18.6–36.3 kg/m2); the mean VFA was 150.2 ± 53.9 cm2 (range: 39.9–312.3 cm2); the mean ALT was 44.0 ± 54.6 IU/L (range: 4.0–543.0 IU/L); the mean AST/ALT ratio was 0.9 ± 0.4 (range: 0.4–2.8) and L/S ratio was 1.0 ± 0.2 (range: 0.3–1.4). IGT and/or IFG were present in 18 patients (8.1%) and diabetes mellitus was present in 11 patients (5.0%). In female patients, the mean age was 57.7 ± 10.5 years (range: 24.0–75.0 years); the mean BMI was 25.7 ± 3.2 kg/m2 (range: 19.9–35.2 kg/m2); the mean VFA was 112.3 ± 41.0 cm2 (range: 11.0–266.9 cm2); the mean ALT was 33.9 ± 33.2 IU/L (range: 9.0–198.0 IU/L); the mean AST/ALT ratio was 1.0 ± 0.4 (range: 0.2–2.0) and L/S ratio was 1.1 ± 0.2 (range: 0.3–1.5). IGT and/or IFG were present in 12 patients (5.4%), and diabetes mellitus was present in 12 patients (5.4%), respectively. Waist circumference was positively correlated to VFA in both gender groups and this correlation was higher in males (males: r = 0.723, P < 0.01; females: r = 0.663, P < 0.01).

Table 1.  Characteristics of the patients having at least one of two criteria for liver disease
 Overall (n = 221)Male (n = 118)Female (n = 103)
  1. Data are mean ± SD. Statistical significance of male group compared with female group: *P < 0.01; **P < 0.05. ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; HOMA-IR, homeostasis assessment model of insulin resistance; L/S ratio; liver/spleen attenuation ratio; PG, plasma glucose; SFA, subcutaneous fat area; T-CHO, total cholesterol; TG, triglyceride; VFA, visceral fat area.

Age (years)55.2 ± 11.953.0 ± 12.757.7 ± 10.5*
Height (cm)160.2 ± 9.1166.1 ± 7.2153.4 ± 5.6*
Weight (kg)66.2 ± 11.271.2 ± 10.660.5 ± 8.8*
BMI (kg/m2)25.7 ± 3.225.8 ± 3.225.7 ± 3.2
Waist circumference (cm)95.3 ± 10.194.1 ± 9.696.7 ± 10.4
VFA (cm2)132.6 ± 51.8150.3 ± 53.9112.3 ± 41.0*
SFA (cm2)194.7 ± 77.2153.4 ± 53.3242.0 ± 73.2*
AST (IU/L)31.8 ± 25.035.0 ± 30.728.1 ± 15.7**
ALT (IU/L)39.3 ± 46.044.0 ± 54.633.9 ± 33.2*
AST/ALT ratio1.0 ± 0.40.9 ± 0.41.1 ± 0.4*
ALP (IU/L)239.6 ± 81.8236.7 ± 80.0243.1 ± 84.2
T-CHO (mg/dL)214.6 ± 41.9212.8 ± 42.7216.6 ± 41.2
TG (mg/dL)138.3 ± 85.4155.4 ± 100.5118.8 ± 59.0**
PG (mg/dL)106.3 ± 18.9107.7 ± 18.5104.8 ± 19.4**
Insulin (μU/mL)10.1 ± 6.610.2 ± 7.010.0 ± 6.2
HOMA-IR2.2 ± 1.32.2 ± 1.12.2 ± 1.5
L/S ratio1.1 ± 0.21.0 ± 0.21.1 ± 0.2*

Relationship of L/S attenuation ratio with other clinical parameters

Table 2 shows the relationships between L/S ratio and various clinical parameters overall and in both gender groups. Overall and in the male group, the relationship between L/S ratio and waist circumference was negative, and there was no correlation in the female group (Fig. 1). The relationship between L/S ratio and VFA or aminotransferase was negative overall and in both gender groups. However, the relationship with AST/ALT ratio was positive overall and in both groups. The relationship with HOMA-IR was negative in the male group, but neither overall nor in the female group. The ratios between VFA and SFA, and VFA and TFA (VFA + SFA), as measured at the umbilicus were markedly higher in males than in females (P < 0.01), whereas the SFA/TFA ratio was markedly higher in females than in males (P < 0.01).

Table 2.  Relationships between liver/spleen ratio and other clinical parameters
 Overall (n = 221)Male (n = 118)Female (n = 103)
rPrPrP
  1. Peason's correlation coefficient was used to correlate continuous variables. ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; HOMA-IR, homeostasis assessment model of insulin resistance; NS, not significant; PG, plasma glucose; SFA, subcutaneous fat area; T-CHO, total cholesterol; TG, triglyceride; VFA, visceral fat area.

Age (years)0.208<0.010.276<0.01NS 
Height (cm)−0.315<0.01−0.225<0.05NS 
Weight (kg)−0.354<0.01−0.369<0.01NS 
BMI (kg/m2)−0.178<0.05−0.284<0.01NS 
Waist circumference (cm)−0.191<0.01−0.356<0.01NS 
VFA (cm2)−0.350<0.01−0.269<0.01−0.319<0.01
SFA (cm2)NS −0.358<0.01NS 
AST (IU/L)−0.320<0.01−0.318<0.01−0.293<0.01
ALT (IU/L)−0.381<0.01−0.344<0.01−0.437<0.01
AST/ALT ratio0.389<0.010.243<0.050.543<0.01
ALP (IU/L)NS NS NS 
T-CHO (mg/dL)NS NS NS 
TG (mg/dL)−0.195<0.05NS −0.239<0.05
PG (mg/dL)NS NS −0.241<0.05
Insulin (μU/mL)NS NS NS 
HOMA-IRNS −0.484<0.01NS 
image

Figure 1. Relationship between liver/spleen attenuation ratio (L/S ratio) and waist circumference/visceral fat area (VFA) overall and in both gender groups. The relationship between (a) waist circumference and (b) VFA evaluated by abdominal plain computed tomography (CT) shows a significant negative correlation in the 221 patients with non-alcoholic fatty liver disease (NAFLD) (L/S ratio and waist circumference: r = −0.191, P < 0.01; L/S ratio and VFA: r = −0.350, P < 0.01). In males (n = 118), there is a significant negative correlation between (c) L/S ratio and waist circumference and (d) VFA evaluated by abdominal plain CT (L/S ratio and waist circumference: r = −0.356, P < 0.01; L/S ratio and VFA: r = −0.269, P < 0.01). In females (n = 103), there is a significant negative correlation between L/S ratio and VFA (f; r = −0.319, P < 0.01), but not with (e) waist circumference.

Download figure to PowerPoint

A total 103 subjects were categorized into 61 (59.2%) cases of premenopause and 42 (40.8%) of postmenopause. Relationship between L/S ratio and waist circumference + VFA was not different in both groups, indicating that menopause did not influence these factors (data not shown).

Discussion

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

Non-alcoholic fatty liver disease, the precursor of NASH, is the most frequent cause of chronic liver disease worldwide. Studies of the molecular basis of this condition and possible treatments have been published,1,20 but gender differences in the clinical condition of NAFLD have not been previously addressed. The present results show a correlation between the degree of fat deposition in the liver, as determined by L/S ratio measured with abdominal CT scan, and waist circumference in males with NAFLD, but not in females with the same condition.

Subcutaneous adipose and visceral adipose are both considered to be a central feature of metabolic syndrome and, in particular, adipokines secreted by visceral adipose have been identified as a key player.21–23 Visceral fat accumulation is also thought to play an important role in the development of NAFLD,24,25 and we have previously found that the degree of hepatic fat deposition in patients with NAFLD correlates with VFA at the level of the umbilicus as measured by CT.17 In the present study, the correlation between waist circumference and the degree of fat deposition in the liver of males with NAFLD indicates that waist circumference does reflect visceral adipose accumulation, which agrees with the results of previous studies conducted in patients with a variety of metabolic abnormalities including NAFLD.26–29

Waist circumference is a useful method of evaluating abdominal obesity, and is accorded great importance in both the WHO diagnostic guidelines for metabolic syndrome and in the National Cholesterol Education Program Adult Treatment Panel (NECP-ATP) III diagnostic guidelines.14 Gender differences are most definitely seen in waist circumference, with the NCEP guidelines stating a measurement of 102 cm in males and 88 cm in females for the diagnosis of metabolic syndrome, mainly in reference to Caucasian patients. The IDF guidelines also recommend using different cut-off levels for males and females, with different criteria for different races.15 In a study of Japanese patients undergoing routine health checks, maximal sensitivity plus specificity for the diagnosis of metabolic syndrome were achieved using a cut-off of 85 cm in males and 78 cm in females.30

Unlike in males in the present study, in females, the degree of fat deposition in the liver correlated well with VFA at the umbilicus, although waist circumference did not correlate with hepatic fat deposition. The several studies with abdominal US indicated the same results as the present study: waist circumference correlated with hepatic fat accumulation in males, but not in females.31,32 One reason for this may be that in females, not only visceral adipose accumulation, but also subcutaneous adipose accumulation contributes to waist circumference at the umbilical level. The relationship between waist circumference and VFA was higher in males than in females in this study. There are opposing views on the significance of waist circumference in females, but some have raised doubts as to whether waist circumference can be used as an index of visceral adipose accumulation in females,33 as have the results of the present study, and it may be necessary to re-examine this question, including in Caucasian females. Previous studies have not examined in detail the influence of gender differences in waist circumference in patients with NAFLD, but the present results indicate differences in the significance of waist circumference in males and females.

A liver biopsy examination is required for accurate assessment of the degree of hepatic fat deposition, but is not feasible in large numbers of patients, so we used L/S ratio measured with abdominal CT scan to evaluate hepatic fat deposition. However, it is not possible to distinguish between NAFLD and NASH using the L/S ratio, so it is possible that some patients with NASH have been included in this study. In addition, there may have been different waist circumferences because of ethnic differences, and there is the influence of age on visceral adipose accumulation.34–36 Although our results indicate that waist circumference may be used as an indicator of fatty liver in males, further analyses are required to gain a better understanding of gender differences. There are several candidates to affect fat distribution and insulin resistance including estrogens and sex-hormone-binding globulin.37–39 Several studies indicated that menopause could affect waist circumference and VFA,40,41 but menopause had no influence on the outcome of this study. The biochemical mechanism of the gender difference was not clarified in this study, which warrants further exploration regarding gender differences.

It is possible to estimate the degree of fat deposition in the liver by measuring the waist circumference of males with NAFLD. However, in females, there is no correlation, casting doubt on the significance of this measurement in females.

Acknowledgments

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

This study was supported in part by a Grant-in-Aid for Scientific Research (16590606 to TM) from the Ministry of Education, Culture, Sports, Science and Technology, Japan. The authors would like to thank all co-medicals in Eguchi Hospital and Professor Kyuichi Tanikawa (International Institute for Liver Research) for excellent advice.

References

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