Visceral adipose tissue area as an independent risk factor for elevated liver enzyme in nonalcoholic fatty liver disease.

Chronic elevations in alanine aminotransferase (ALT) levels are associated with body composition. The aim of this study was to evaluate the relationship between elevated liver enzyme levels and the visceral tissue area in subjects with and without nonalcoholic fatty liver disease (NAFLD).An observational cohort study was conducted with subjects undergoing general health examinations. To evaluate the visceral and subcutaneous abdominal adipose tissue area, a computed tomography scan was performed. NAFLD was diagnosed if a person demonstrated fatty liver on ultrasonography without a history of significant alcohol consumption or chronic liver disease. Abnormal liver enzyme levels were based on ALT elevations according to the updated Asian definition.Of the 5100 subjects, 3712 (72.8%) met the inclusion criteria, and NAFLD was found in 1185 subjects. Elevated ALT values were positively correlated with body mass index, waist circumference, and subcutaneous and visceral adipose tissue area. These relationships were attenuated, although they remained significant in a dose-dependent manner, after adjusting for multiple liver injury risk factors. In addition, when body mass index and subcutaneous and visceral tissue areas were finally considered in combination, only visceral adipose tissue remained independently associated with elevated ALT levels in the ultrasonographically diagnosed NAFLD group (P for trend <0.001 for men and women).Elevated ALT levels were independently and dose-dependently associated with visceral fat accumulation in the healthy general population, especially in ultrasonographically diagnosed NAFLD patients. These results reemphasize the importance of visceral fat in the pathogenesis of NAFLD.

INTRODUCTION

The serum concentration of alanine aminotransferase (ALT), which is considered to be the most specific marker for liver damage,[1] is associated with body mass index (BMI) and progressively increases with increasing BMI values.[2-4] A population-based study using data from the Third National Health and Nutrition Examination Survey (NHANES III) in the United States revealed that central adiposity, hyperleptinemia, and hyperinsulinemia are important factors in the association between overweight and increased ALT levels.[5] Another study reported a role of central adiposity in predicting increased liver enzyme levels.[6] A study using data from a ultrasonography (US) population survey also determined that trunk fat was associated with increased serum ALT levels independently of BMI and waist circumference (WC).[7] However, in this study, trunk fat was not differentiated into visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT). Several studies have used computed tomography (CT) to determine the association between serum ALT levels and VAT area; however, these studies were limited by selection bias and small sample sizes.[8,9]

Nonalcoholic fatty liver disease (NAFLD), a common cause of chronic liver disease, with an increasing prevalence (20%–30%) worldwide,[10,11] is the most common cause of elevated serum ALT levels.[10] An association between hepatic fat and VAT has been suggested,[12] and hepatic steatosis, as measured by proton magnetic resonance spectroscopy, was found to be closely related to central obesity.[13] In addition, the severity of fatty liver has been linked to the VAT area as evaluated by CT.[14] Furthermore, the VAT area assessed by magnetic resonance imaging has also been directly associated with the severity of hepatic inflammation and fibrosis, independent of insulin resistance and hepatic steatosis.[15] However, few studies have investigated the relationship between liver injury and VAT in terms of NAFLD. Therefore, the aim of this study was to determine the relationship between the elevation of liver enzyme levels, VAT, and anthropometric indexes in subjects with and without NAFLD.

PATIENTS AND METHODS

Study Population

For our study, we analyzed the database from a previously described cohort.[16,17] Briefly, a total of 5100 subjects who underwent abdominal ultrasonography, abdominal fat CT scans for adipose tissue areas, and blood samplings for a routine health checkup at the Seoul National University Hospital Gangnam Healthcare Center, Seoul, Korea, were recruited into the study. A total of 280 subjects who were positive for hepatitis B surface antigen, 56 who were positive for hepatitis C antibodies, 949 with alcohol consumption >140 g/wk, and 14 with other hepatitis histories, as identified by a questionnaire, were excluded. Additionally, we excluded 99 subjects who in the past year had taken medications known to provoke fatty liver. Finally, 3712 subjects met the inclusion criteria. Ethical approval for this study was obtained from the institutional review board of the Seoul National University Hospital with an informed consent waiver prior to the study.

Measurement of Clinical and Laboratory Parameters

Briefly, the details of clinical and laboratory measurement were based on the previous description.[17] Each subject completed questionnaires regarding his/her past medical and medication history. For women, questions regarding the presence of menopause and a history of hormone replacement therapy were included. A woman was considered to be menopausal if her menstrual periods had stopped >1 year prior to the study. The participants underwent an anthropometric assessment as well as laboratory and radiological exams on the day of tests. Measurements of height and body weight were measured while the subjects were wearing a light gown, and the BMI was calculated by dividing the weight (kilogram) by the square of the height (square meter). WC was measured at the midpoint between the lower costal margin of the rib cage and the iliac crest using a tape measure. The systolic and diastolic blood pressure was measured twice, and the average values were obtained for data analysis. Hypertension was defined as treatment with an antihypertensive drug, a systolic blood pressure >140 mm Hg, or a diastolic blood pressure >90 mm Hg.

The laboratory tests included serum aspartate aminotransferase, ALT, γ-glutamyl transpeptidase, triglycerides, total cholesterol, high-density lipoprotein (HDL)-cholesterol, glucose, uric acid, hepatitis B surface antigen, and hepatitis C virus antibody, and were performed on an Architect Ci8200 (Abbott Laboratories, Abbott Park, IL). Measurements of serum insulin levels were performed using immunoradiometric assays (Biosource, Nivelles, Belgium), and hemoglobin A1c (HbA1c) was measured with a COBAS INTEGRA 400 (Roche Diagnostics GmbH, Mannheim, Germany).

The presence of diabetes mellitus was defined as a fasting serum glucose level ≥126 mg/dL or as the use of blood glucose-lowering agents. Current smokers were defined as those who had smoked at least 1 cigarette/d during 1 year prior to the study. Former smokers were defined as those who had previously smoked cigarettes regularly.[18] Blood samples were collected before 10:00 am after a 12-hour overnight fast. All of the laboratory tests were performed using standard laboratory methods. Homeostasis model assessment (HOMA)-estimated insulin resistance was used to evaluate insulin resistance, and was obtained by multiplying the fasting insulin by the fasting glucose.[19] Abnormal liver enzyme levels were defined as ALT elevation above the strict cutoff point based on the updated Asian definition by Lee et al.[20] (ALT >33/25 IU/L).

Metabolic syndrome (MS) was diagnosed if patients met 3 or more of the following 5 criteria: central obesity (WC >90 cm [men] or >80 cm [women], based on the definition by the Regional Office for the Western Pacific Region of the World Health Organization criteria), HDL-cholesterol <40 mg/dL (men) or <50 mg/dL (women), a triglyceride level ≥150 mg/dL, pressure ≥130/85 mm Hg or treatment for hypertension, and hyperglycemia (fasting glucose ≥100 mg/dL) or treatment for diabetes.[21]

Measurement of Adipose Tissue Areas

The method for measuring adipose tissue area using CT cross-sectional images has been described previously.[22,23] The subjects were examined with a 16-detector row CT scanner (Somatom Sensation 16; Siemens Medical Solutions, Forchheim, Germany). The cross-sectional surface area (in square centimeter) of the abdominal fat tissue was calculated at the level of the umbilicus using a CT software program (Rapidia 2.8; INFINITT, Seoul, Korea) by specifying the attenuation values for fatty tissues [−250 to −50 Hounsfield unit (HU)]. Because of a lack of a cutoff value for an appropriate healthy amount of abdominal adipose tissue, we used the lowest sex-specific quartile of the VAT and SAT areas as reference values.

Ultrasonographic Assessments and Definition of NAFLD

The ultrasonographic examinations of the liver were carried out by experienced radiologists blinded to the patients’ clinical characteristics. Fatty liver was diagnosed based on the findings of ultrasonography (Acusion, Sequoia 512; Siemens, Mountain View, CA) according to the criteria previously described.[24] NAFLD was diagnosed in subjects who showed the findings of fatty liver on ultrasonography, in the absence of the following: a positive serologic marker for hepatitis B surface antigen or hepatitis C virus serological marker, excessive alcohol intake (>30 g/d for men and >20 g/d for women), medications known to produce fatty liver disease, and other specific hepatic disease.

Statistical Analysis

Comparisons of continuous variables between the 2 groups were performed using Student t test and analysis of variance (ANOVA) or ANOVA with post hoc analysis for multiple comparisons, and categorical data was compared using the χ2 test or Fisher exact test. Variables significant in univariate analysis and previously known risk factors were included in the multivariate models to determine independent predictors of NAFLD. The statistical analyses were performed using SPSS 19.0 (SPSS Inc, Chicago, IL). The statistical significance was achieved at P < 0.05.

RESULTS

A total of 3712 subjects were analyzed. The mean age was 51.6 ± 9.7 years, and 55.5% of the subjects were men. NAFLD was found in 1185 (31.9%) of the 3712 subjects. In men, elevations of ALT were positively correlated with obesity, MS, higher BMI, WC, systolic blood pressure, triglyceride, fasting insulin, uric acid, HOMA index, SAT and VAT values, lower HDL-cholesterol levels, and younger age in the ultrasonographically normal and NAFLD groups. In men, obesity, BMI, WC, triglycerides, fasting insulin, HOMA index, MS, SAT, and VAT were significantly increased in a dose-dependent manner in the ultrasonographically diagnosed normal group with normal ALT levels compared with the ultrasonographically diagnosed normal group with elevated ALT levels, the ultrasonographically diagnosed NAFLD group with normal ALT levels, and the ultrasonographically diagnosed NAFLD group with elevated ALT levels (P < 0.001, Table 1). In women, elevated ALT levels showed a significant association with higher BMI, WC, triglycerides, HbA1c, fasting insulin, HOMA index, MS, uric acid, VAT and SAT in the ultrasonographically diagnosed normal and NAFLD groups. The presence of diabetes mellitus, hypertension and MS, BMI, WC, triglyceride level, total adipose tissue, SAT, and VAT were higher, and showed a dose-dependent increase in the ultrasonographically diagnosed normal group with normal ALT levels, ultrasonographically diagnosed normal group with elevated ALT levels, ultrasonographically diagnosed NAFLD group with normal ALT levels, and the ultrasonographically diagnosed NAFLD group with elevated ALT levels (Table 2).

TABLE 1

Clinical Characteristics of Study Population (Men)

TABLE 2

Clinical Characteristics of Study Population (Women)

Relationship Between Abdominal Adipose Tissue Area and Elevated ALT Levels

We evaluated the risk for ALT elevation according to body measurement indices such as VAT, SAT, WC, and BMI. The associations between elevated ALT and the body measurement indices that were observed in the univariate analysis were attenuated in the multivariate analysis; however, the associations were still significant. After adjusting for multiple liver injury risk factors including age, smoking status, systolic blood pressure, fasting glucose, triglycerides and HDL-cholesterol, and postmenopausal status and use of hormone replacement therapy (in women), elevation of ALT showed positive relationships with BMI, WC, and VAT in ultrasonographically diagnosed NAFLD individuals and in the ultrasonographically normal group. Though the odds ratios of the VAT quartiles were similar between the ultrasonographically diagnosed normal and NAFLD groups in men, the odds ratios of the VAT quartile in the ultrasonographically diagnosed NAFLD group were higher than in the normal group in women (Tables 3 and 4).

TABLE 3

Risk of the Body Measure Indices and Abdominal Adipose Tissue Areas for Elevated Serum ALT in Men (Cutoff 33)

TABLE 4

Risk of the Body Measure Indices and Abdominal Adipose Tissue Areas for Elevated Serum ALT in Women (Cutoff 25)

In addition, when SAT and VAT were considered together, VAT was significantly associated with elevated ALT in men and women. However, the dose dependence was stronger in the ultrasonographically diagnosed NAFLD group than in the normal group. SAT showed a significant association only in men in the ultrasonographically diagnosed normal and NAFLD groups and not in women (Table 5). Finally, after a further adjustment for BMI as a surrogate marker of general obesity, a statistically significant association remained between VAT and elevated ALT in the ultrasonographically diagnosed NAFLD group (P for trend < 0.001 for men and women). This association was attenuated and no longer reached statistical significance in the ultrasonographically normal group (P for trend 0.079 for men and 0.114 for women, Table 6). The SAT had no significant association with ALT elevation, and BMI was associated with elevated ALT only in men in the ultrasonographically diagnosed NAFLD group (Table 6). These results indicate that VAT is an independent risk factor for elevated ALT in ultrasonographically diagnosed NAFLD.

TABLE 5

Multivariate Analysis for Risk of the Abdominal Adipose Tissue Areas for Elevated Serum ALT

TABLE 6

Risk of the Various Adipose Tissue Areas for Elevated Serum ALT (Including BMI)

Next, we analyzed the data regarding the relationship with elevated ALT levels defined by higher cutoff values. Using the cutoff value of ALT >40, a significant association was found between VAT and elevated ALT in the ultrasonographically diagnosed NAFLD group (P for trend 0.017 for men and <0.001 for women, Supplementary Table 1, http://links.lww.com/MD/A212). When the cutoff value of ALT was set at >2 times the upper normal limit, VAT had an association with elevated ALT only in women in the ultrasonographically diagnosed NAFLD group (Supplementary Table 2, http://links.lww.com/MD/A212).

DISCUSSION

In the present study, visceral fat accumulation was significantly associated with elevated ALT levels even after adjusting for metabolic components in the ultrasonographically diagnosed normal and NAFLD groups. The relationship between VAT and elevated ALT remained after adjusting for BMI, a surrogate marker of general obesity, in the ultrasonographically diagnosed NAFLD group.

A large population-based study using NHANES III data in the US showed that central adiposity, as measured by the waist-to-hip circumference ratio, was more strongly associated with elevated serum ALT levels than with BMI.[5] In addition, a study of recent NHANES III data revealed that trunk fat, as measured by dual x-ray absorptiometry, was associated with increased serum ALT levels, and this association was independent of BMI, WC, and other risk factors for liver injury.[7] Previous methods of measuring central fat could not differentiate between VAT and SAT.

Several studies that have assessed the relationship between ALT levels and VAT area using CT scans showed results that are consistent with our study.[8,9] However, previous studies had limitations including small sample sizes and no consideration of the presence of fatty liver. In the present study, the visceral fat area was significantly associated with elevated ALT in subjects with NAFLD, suggesting a role for visceral fat in enhancing fatty infiltration and inflammation in NAFLD. In accordance with our findings, a previous study performed in Japan demonstrated that the severity of fatty liver, assessed by ultrasonography, was correlated with visceral fat accumulation and insulin resistance.[14]

Chronic ALT elevation is considered to be a marker of hepatocyte damage. Increased ALT activity was associated with the MS, and this association was graded across a number of metabolic components.[25,26] In addition, higher ALT levels were previously known to be associated with increased insulin resistance and the development of type 2 diabetes.[27,28] In accordance with previous studies, ALT elevations were significantly correlated with metabolic variables and insulin resistance in our study.

Adipocytes in the visceral fat promote the increased release of free fatty acids and the subsequent production of cytokines, such as adiponectin, interleukin-6, tumor necrosis factor-α, and leptin, and these adipocytokines flow directly into the liver because abdominal fat has a circulatory communication pathway to the liver via the portal vein.[29-31] In addition, these adipocytokines have been known to induce insulin resistance and enhance hepatic steatosis, and they may also induce nonalcoholic steatohepatitis.[32] Furthermore, visceral fat was directly associated with hepatic inflammation and fibrosis in a dose-dependent manner, independent of insulin resistance.[15]

A strength of our study is the relatively large sample size, which allowed for a high-powered analysis of associations. Moreover, the subjects in this study are representative of the healthy population due to the characteristics of a health checkup. Finally, an abdominal fat measurement by CT scan was performed on the same day that blood samples were collected, which allowed for the accurate evaluation of the association between visceral fat and ALT elevations.

However, there are limitations to our study. First, because of the cross-sectional design of this study, we could not identify the causal relationship between elevated ALT levels and visceral fat accumulation. Second, the method of hepatic ultrasonography used to diagnose NAFLD could not identify fatty infiltration below 30%.[33] Other blood parameters, as well as noninvasive methods such as magnetic resonance imaging or the controlled attenuation parameter, could be used to diagnose NAFLD. However, ultrasonography has the advantages of low cost, safety, satisfactory sensitivity, and specificity.[34] Therefore, hepatic ultrasonography is considered the first-line technique in clinical practice guidelines.[11,35] Third, the interpretation of the results is limited due to the lack of data on the serum levels of adiopocytokines. Finally, a single measurement of ALT levels could be influenced by acute liver injury with variable causes. However, we evaluated the presence of fatty liver with ultrasonography to overcome the limitation of a single ALT measurement.

In conclusion, we have demonstrated that the elevation of ALT was independently and dose-dependently associated with visceral fat accumulation in the healthy general population, especially in ultrasonographically diagnosed NAFLD patients. These findings emphasize the importance of visceral fat in the pathogenesis and inflammation of NAFLD.