You are here

Article Text

Article menu

Download PDFPDF

Cardiac risk factors and prevention
Original article
Non-alcoholic fatty liver disease, metabolic syndrome and subclinical cardiovascular changes in the general population
  1. Nan Hee Kim1,
  2. Juri Park2,
  3. Seong Hwan Kim3,
  4. Yong Hyun Kim3,
  5. Dong Hyuk Kim3,
  6. Goo-Yeong Cho4,
  7. Inkyung Baik5,
  8. Hong Euy Lim6,
  9. Eung Ju Kim6,
  10. Jin Oh Na6,
  11. Jung Bok Lee7,
  12. Seung Ku Lee8,
  13. Chol Shin8
  1. 1Division of Endocrinology, Department of Internal Medicine, Korea University Ansan Hospital, Ansan, Korea
  2. 2Division of Endocrinology, Department of Internal Medicine, Hallym University Kangdong Sacred Heart Hospital, Seoul, Korea
  3. 3Division of Cardiology, Department of Internal Medicine, Korea University Ansan Hospital, Ansan, Korea
  4. 4Division of Cardiology, Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam, Korea
  5. 5Department of Foods and Nutrition, College of Natural Sciences, Kookmin University, Seoul, Korea
  6. 6Division of Cardiology, Department of Internal Medicine, Korea University Guro Hospital, Seoul, Korea
  7. 7Department of Clinical Epidemiology and Biostatistics, Asan Medical Center, Seoul, Korea
  8. 8Institute of Human Genomic Study, Korea University Ansan Hospital, Ansan, Korea
  1. Correspondence to Professor Seong Hwan Kim, Division of Cardiology, Department of Medicine, Korea University Ansan Hospital, 516 Gojan-1-dong, Danwon-gu, Ansan-si, Gyeonggi-do 425-707, Korea; cardioguy{at}korea.ac.kr

Abstract

Objective The effect of non-alcoholic fatty liver disease (NAFLD) on cardiovascular system remains controversial. We investigated the independent contribution of NAFLD to cardiovascular structure and function in the general population.

Methods A total of 1886 participants without known cardiovascular disease were enrolled from the Korean Genome Epidemiology Study. The participants were divided into four groups, based on the presence of NAFLD, metabolic syndrome (MetS), neither or both. NAFLD was diagnosed by CT. Changes in cardiovascular structure and function were assessed by tissue Doppler imaging (TDI) echocardiography, carotid ultrasound and brachial-ankle pulse wave velocity (baPWV).

Results In multivariate analyses, subjects with both NAFLD and MetS had a higher E/Ea ratio and baPWV, as well as a lower TDI Ea velocity (all p<0.001) than those with neither NAFLD nor MetS. Subjects with either NAFLD or MetS also showed significant differences in TDI Ea velocity and baPWV (all p<0.05). However, no significant differences of carotid intima-media thickness (CIMT) values were seen among the four groups. Multivariate linear regression revealed that both NAFLD and MetS were independent predictors of TDI Ea velocity and baPWV (all p<0.001). Both MetS and NAFLD were not a determinant of CIMT.

Conclusions NAFLD was associated with early alterations of cardiovascular system, independent of established cardiovascular risk factors and MetS.

https://doi.org/10.1136/heartjnl-2013-305099

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Non-alcoholic fatty liver disease (NAFLD) is defined as hepatic fat accumulation in the absence of excessive alcohol consumption. Several epidemiological studies have reported an increased incidence of cardiovascular disease (CVD) in NAFLD patients.1 Therefore, the dramatic increase in the prevalence of NAFLD is becoming a major global health problem. A high prevalence of NAFLD was reported in various ethnic groups, ranging from 24% in non-Hispanic blacks and 33% in non-Hispanic whites to 45% in Hispanics.2 NAFLD is also an increasingly common disorder in Koreans, with a prevalence of up to 40%.3 NAFLD is strongly associated with metabolic syndrome (MetS), type 2 diabetes and obesity as these diseases share a common pathophysiology that includes insulin resistance and systemic inflammation.4 ,5 Although NAFLD is considered to be the hepatic manifestation of MetS, some argue that NAFLD should be considered as a separate entity rather than a component of MetS.6 Evidence suggesting that NAFLD contributes to endothelial dysfunction and accelerated atherosclerosis, independent of MetS, supports this hypothesis.7 ,8 However, the independent role of NAFLD for future CVD still remains controversial.9–11 In addition, limited data concerning the effect of NAFLD on LV structure and function are available.12 ,13

    Increased LV mass index, carotid intima-media thickness (CIMT) and pulse wave velocity (PWV) are accepted as the most useful markers for predicting CVD in asymptomatic subjects. In addition to its use in determining CVD prognosis, tissue Doppler imaging (TDI) has emerged as a useful tool for the early detection of myocardial dysfunction.14 To our knowledge, no epidemiological study has extensively evaluated the relationships between NAFLD, MetS and target organ damage. We therefore examined the independent contribution of NAFLD to cardiovascular structure and function in a community-based cohort.

    Methods

    Study participants

    The study participants were enrolled from the Ansan cohort, which is a population-based cohort included in the Korean Genome and Epidemiology Study. A total of 5015 Ansan cohort members aged ≥40 years have been assessed biennially with scheduled on-site follow-up visits since 2001. Further information and the study design for this ongoing prospective cohort are available in a previous report.15 For this cross-sectional study, 3262 cohort subjects who participated in the fifth cycle of the 2-year follow-up study from 26 May 2009 to 7 December 2010 were included. We excluded (1) participants who did not undergo unenhanced CT, TDI echocardiography, carotid ultrasound and brachial-ankle PWV (baPWV); (2) participants who had known CVD, including congenital heart disease, cardiomyopathy, valvular heart disease, arrhythmia, renal dysfunction (serum creatinine ≥2.0 mg/dL) or LV EF <55%; and (3) participants who had positive viral markers for hepatitis B surface antigen, hepatitis C virus antibody, other chronic liver disease, heavy drinking history (≥140 g/week) or serum alanine transferase (ALT) and aspartate transferase (AST) levels more than three times the upper limit of normal. A total of 1369 subjects were excluded, and the remaining 1886 cohort members (713 men and 1173 women) aged ≥45 years were eligible for the study. This study was approved by the Human Subjects Review Committee at the Korea University Ansan Hospital, and written informed consent was obtained from all participants.

    Clinical and laboratory measurements

    Height (cm) and body weight (kg) were measured to the nearest 0.1 cm or 0.1 kg, and body mass index (BMI) was calculated. Waist circumference (cm) was measured at the narrowest point between the lower rib and the iliac crest. The measurement of alcohol consumption has been described in detail previously.16 We defined participants as hypertensive if they were taking antihypertensive medication, had a systolic blood pressure ≥140 mm Hg and/or a diastolic blood pressure ≥90 mm Hg. Patients were defined as diabetic if they took an oral hypoglycaemic agent or insulin therapy, if their fasting levels of blood glucose were ≥126 mg/dL and/or if they had a 2-h postprandial glucose ≥200 mg/dL. The homeostasis model assessment insulin resistance index (HOMA-IR) was calculated as fasting serum insulin (µU/mL)×fasting plasma glucose (mg/dL)/405.

    Based on a modified version of the National Cholesterol Education Program Adult Treatment Panel III, participants had MetS if they fulfilled any three or more of the following criteria: central obesity with ethnicity-specific values (waist circumference ≥90 cm for men and ≥80 cm for women),17 hypertriglyceridaemia (≥150 mg/dL), low high-density lipoprotein (HDL) cholesterol (<40 mg/dL for men and <50 mg/dL for women), high blood pressure (≥130/85 mm Hg or the use of antihypertensive medication) and either fasting glucose ≥100 mg/dL or the use of antidiabetic medication.

    CT scans (Brilliance 64; Philips, Cleveland, Ohio, USA) were at 120 kV of exposure without using intravenous contrast. Hepatic attenuation was measured by randomly selecting three circular 1 cm2 region of interest (ROI) on three transverse sections. Those transverse sections were selected at the T11 mid-body, T11-T12 interspace and T12 mid-body levels. To provide an internal control, the mean splenic attenuation was calculated by averaging splenic attenuation measurements from two random ROI on each of the three transverse sections used for the evaluation of hepatic attenuation. The liver attenuation index (LAI), derived from the difference between the mean hepatic and splenic attenuation, was used as a parameter to diagnose NAFLD. As an LAI of <5 HUs predicted >5% steatosis18 and a minimum of 5% steatosis is required for histological confirmation of NAFLD,19 we used LAI <5 HUs as a diagnostic cut-off for NAFLD. According to the presence or absence of NAFLD and/or MetS, participants were divided into four groups as follows: neither NAFLD nor MetS, only NAFLD without MetS, only MetS without NAFLD or both NAFLD and MetS.

    Standard echocardiography (Vivid 7, GE Vingmed, Horton, Norway) was used to determine structural and functional cardiac changes. The left atrial (LA) anteroposterior diameter was measured using the M-mode of parasternal long-axis view. The LV EF was calculated by the modified biplane Simpson's method. The LV mass was determined according to the Devereux formula, normalised for height to the power of 2.7. The early (E) and late (A) wave velocities, the deceleration time of early filling and the mitral E/A ratio were all measured from the mitral inflow profile. The systolic (Sa), early diastolic (Ea) and late diastolic (Aa) velocities were measured at the septal mitral annulus. The mitral E/Ea ratio as an index of LV diastolic filling pressure was subsequently calculated. All measurements were performed by one observer.

    The right and left CIMT values were measured using B-mode ultrasonography (SonoSite TITAN, SonoSite, Bothell, Washington, USA) with a 7.5 MHz linear-array transducer. Near and far walls were examined approximately 1 cm proximal to the bulb. The mean CIMT value was analysed using semiautomated software (M'ATH SR V.2.0, Metris, France) and was calculated as the average of the left and right CIMT scores obtained from the near and far walls. The coefficients of variation between and within sonographers were less than 5%.

    baPWV was measured in the supine position following a minimum of 5 min of rest using an automatic volume-plethysmographic device (VP-2000, Colin, Komaki, Japan). The PWV is calculated as the distance between the two sites over the pulse transit time. The time interval between the brachium and the ankle was defined as the time interval between the wave front of the brachial waveform and that of the ankle waveform. The distance between sampling points of the baPWV was calculated automatically according to the height of the subjects, and the mean baPWV of the left and right arm was used for analysis. All measurements were performed by the same examiner, and intraobserver coefficient of variation was 3.6%.

    Statistical analyses

    Data are expressed as means±SD for continuous variables and as percentages for categorical variables. Variables without normal distribution are given as median and IQR and were log-transformed for use in statistical analyses. The analyses of continuous and categorical variables to assess differences among the four groups in this study were determined by one-way ANOVA using either the Scheffe post hoc test for comparison or the χ2 test. The comparison of CV structure and function between the four groups was performed by one-way ANCOVA of general linear models. The least-square means and 95% CIs for LV mass index, TDI Ea velocity, CIMT and baPWV were measured. Using the stepwise method, multivariate linear regression models were used to determine the independent predictors of LV mass index, TDI Ea velocity, CIMT and baPWV. On the basis of clinical relevance and univariate analyses, all multivariate models were adjusted for age, gender, BMI, heart rate, alcohol consumption, current smoking, high-sensitivity C-reactive protein (hsCRP), antihypertensive medications, antidiabetic medications, lipid-lowering drugs and family history of CVD. A p<0.05 was considered statistically significant. All statistical analyses were performed using SAS statistical software (SAS V.9.1.3, SAS Institute, Cary, North Carolina, USA).

    Results

    Clinical and cardiometabolic profile

    Clinical and biochemical characteristics of 1886 study participants categorised by the presence or absence of NAFLD and MetS are presented in table 1. In our cohort of a middle-aged and elderly Korean population, the prevalence of NAFLD was 22.3% (29.0% in males and 17.9% in females) and the prevalence of MetS was 31.9% (29.6% in males and 33.6% in females). The 12.8% of study participants had both NAFLD and MetS, 19.1% had only MetS and 9.5% had only NAFLD. Subjects with both NAFLD and MetS had the highest values in BMI, hsCRP, AST and ALT, HOMA-IR, and the MetS components except HDL-cholesterol. Participants with either NAFLD or MetS showed intermediate values in cardiometabolic parameters. Both hsCRP and HOMA-IR levels were also significantly increased in individuals with NAFLD only (p<0.001) and with MetS only (p<0.001) in comparison with those who had neither.

    View this table:
    Table 1

    Clinical characteristics of study participants according to different combinations of NAFLD and MetS

    NAFLD, MetS and cardiovascular changes

    Table 2 shows the result of unadjusted analyses from TDI echocardiography, CIMT and baPWV studies. Compared with subjects with neither NAFLD nor MetS, those with both disorders showed the most significant differences in structural and functional LV parameters, such as LA diameter, LV mass index, transmitral Doppler indices, systolic/diastolic TDI values and E/Ea ratio. In a similar fashion, NAFLD or MetS individually also showed statistically significant differences in LV structure and function. After multivariate analyses including BMI, any significant differences in LA diameter and LV mass index were not observed among the study groups. However, NAFLD and MetS were each significantly and independently associated with lower levels of TDI Ea velocity (figure 1).

    View this table:
    Table 2

    Measurements of cardiovascular structure and function according to different combinations of NAFLD and MetS

    Figure 1
    Figure 1

    Multivariate-adjusted mean values of (A) LV mass index; (B) tissue Doppler imaging Ea velocity; (C) common carotid intima-media thickness; (D) brachial-ankle pulse wave velocity after adjustment for age, gender, body mass index, heart rate, alcohol consumption, current smoking, high-sensitivity C-reactive protein, antihypertensive medications, antidiabetic medications, lipid-lowering drugs and family history of cardiovascular disease. Bars represent 95% CI of mean.

    We further analysed the effects of NAFLD and MetS on vascular structure and function using CIMT and baPWV. Values of CIMT and baPWV gradually increased and reached statistical significance across groups (p<0.001 for trend) (table 2). However, after adjustment for CV confounders including BMI, no significant differences were detected in CIMT values of the four study groups (figure 1). Even those with both NAFLD and MetS did not show statistically significant difference of CIMT value in comparison to those with neither NAFLD nor MetS. On the other hand, significantly higher baPWV values were found in the groups of NAFLD and/or MetS compared with those who had neither NAFLD nor MetS.

    There were no interactions for the majority of the parameters investigated.

    Independent predictors of LV changes, CIMT and baPWV

    To investigate whether NAFLD and MetS were independent predictors of structural and functional CV parameters, we used multivariate linear regression analyses with LV mass index, TDI Ea velocity, CIMT and baPWV as dependent variables (table 3). In this model, LAI and MetS were taken together as independent variables, along with age, gender, BMI, heart rate, alcohol consumption, current smoking, hsCRP, antihypertensive medications, antidiabetic medications, lipid-lowering drugs and family history of CVD. The results showed that both LAI and MetS were independent factors affecting TDI Ea velocity and baPWV. However, both NAFLD and MetS were not independent predictors for LV mass index and CIMT in a multivariate model.

    View this table:
    Table 3

    Independent correlates of cardiovascular parameters: multivariate analysis

    Discussion

    In this middle-aged and elderly population cohort, we found independent effects of NAFLD and MetS on CV system. As expected, MetS alone was closely associated with subclinical changes in LV diastolic function in addition to arterial stiffness. NAFLD alone also demonstrated subtle deterioration in arterial stiffness and LV diastolic function, independent of MetS. Interestingly, the presence of NAFLD additively led to subclinical deterioration in LV diastolic function in subjects with MetS. As a result, NAFLD was an independent predictor of LV diastolic function and arterial stiffness, assessed by TDI Ea velocity and bPWV, even after adjusting for MetS as well as traditional CV risk factors including BMI.

    Previous epidemiological studies have reported an increased risk of CV mortality in subjects with NAFLD. Additionally, a correlation between NAFLD and adverse CV events existed independent of MetS and other CV confounders.20 However, some argue that the evidence of this association is limited because of incomplete adjustment of CV confounders and age interaction.9 Accordingly, they recommended that more caution should be taken to consider subjects with NAFLD as being at high risk for CVD with the current evidence base. Considering that NAFLD may be an independent risk factor of CV events, its presence appears to be associated with relevant target organ damage. However, there is limited information and inconsistent data regarding the independent effect of NAFLD on LV structure and function, CIMT and arterial stiffness.

    Whereas a number of studies have shown the effect of MetS on LV structure and function, few studies have mentioned that subjects with NAFLD had impaired LV function and increased LV mass index with relatively small sample sizes or with only type 2 diabetic patients.12 ,21 ,22 The results of our study, which were obtained from a large epidemiological study, are in line with previous studies, and provide evidence that NAFLD may contribute to early abnormalities of LV diastolic function, independent of the MetS. In particular, early and subclinical change in LV function could be detected in subjects with NAFLD by using TDI technique, but not by conventional echocardiography. Contrary to the positive findings of these studies, two studies investigating the alterations of LV energy metabolism in NAFLD did not demonstrate significant differences in LV structure and function compared with controls.23 ,24 However, because neither of these studies used TDI echocardiography to detect subtle LV changes, it seems that abnormal myocardial energy metabolism precedes overt LV dysfunction detected by conventional echocardiography.

    Though a recent systemic review including 3497 subjects reported an increase of 13% of CIMT in NAFLD patients,8 the data regarding this association contain contradictions.10 ,11 Interestingly, studies that failed to demonstrate an association between NAFLD and CIMT were all conducted in a population of type 2 diabetes mellitus patients, while another study demonstrated a strong association between NAFLD and CIMT in diet-controlled type 2 diabetic individuals.25 Based on these conflicting data, it is possible that the severity of type 2 diabetes mellitus may play a role in the relationship between NAFLD and CIMT thickness. However, no association between carotid atherosclerosis and NAFLD was observed in our study, even after excluding type 2 diabetic patients (data not shown), which imply that hepatic steatosis could not be a direct mediator of subclinical vascular disease in the carotid artery regardless of type 2 diabetes mellitus.

    The PWV is recognised as a valuable surrogate marker for arterial stiffness. Contrary to the studies pertaining to CIMT in subjects with NAFLD, previous literature on arterial stiffness has consistently reported that NAFLD is strongly associated with PWV, independent of MetS,26 ,27 and that NAFLD rather than MetS was a major contributor.26 Our results are in agreement with these studies and demonstrated that NAFLD and MetS independently accelerate arterial stiffness. However, further studies should be needed to explain the different role of NAFLD in CIMT and arterial stiffness.

    The mechanism by which NAFLD leads to an increased risk of CVD or causes target organ damage is unknown. Proposed mechanisms include oxidative stress, chronic inflammation, insulin resistance and adipocytokines.28 ,29 hsCRP, as a circulating marker of systemic inflammation, is known to be a strong predictor of future CVD and to be closely linked with NAFLD, independent of MetS and obesity. A recent study demonstrated that the combinations of NAFLD, obesity and MetS were associated with an additive increase in high hsCRP levels measured in subjects with NAFLD alone.5 In accordance with these studies, the present study found NAFLD was associated with hsCRP levels, even after adjusting for MetS, obesity and other CV risk factors (data not shown). Accordingly, it seems that the elevation of hsCRP levels in NAFLD may play an important role in subclinical changes of the CV system.

    Study limitation

    Although the current study has several strengths, such as relatively large sample size, the use of non-enhanced CT for a better quantitative assessment of NAFLD, the application of a novel TDI technique for detection of early changes in LV function and a comprehensive approach for assessment of CV target organ changes, some limitations should be addressed. The cross-sectional study design did not allow us to establish causal relationship. The results of the current study conducted by a single ethnic group cannot be generalised to other ethnic groups. NAFLD diagnosis was made based on LAI using CT instead of histological confirmation by liver biopsy. However, unenhanced CT could be a very useful non-invasive modality for the diagnosis of NAFLD, especially in large epidemiological studies.30 In addition, the radiation exposure related to CT scans may limit their use in clinical practice. Finally, our data might have unidentified confounding factor such as potential coronary artery disease because the participants did not undergo quantitative coronary angiography.

    Conclusions

    This study demonstrated that the presence of NAFLD was associated with subclinical CV changes assessed by TDI and baPWV, independent of MetS and other CV confounders. In addition, the presence of NAFLD in subjects with MetS additively contributed to a subclinical deterioration in LV diastolic function. Additional studies exploring whether these alterations of target organs by NAFLD, independent of MetS, progress to overt heart failure, hypertension or stroke, and whether these changes are reversible after the improvement of NAFLD, are needed.

    References

    1. ↵
      1. Ruttmann E,
      2. Brant LJ,
      3. Concin H,
      4. et al
      . Gamma-glutamyl transferase as a risk factor for cardiovascular disease mortality: an epidemiological investigation in a cohort of 163,944 Austrian adults. Circulation 2005;112:2130–7.
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Browning JD,
      2. Szczepaniak LS,
      3. Dobbins R,
      4. et al
      . Prevalence of hepatic steatosis in an urban population in the United States: impact of ethnicity. Hepatology 2004;40:1387–95.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Kim D,
      2. Choi SY,
      3. Park EH,
      4. et al
      . Nonalcoholic fatty liver disease is associated with coronary artery calcification. Hepatology 2012;56:605–13.
      OpenUrlCrossRefPubMed
    4. ↵
      1. Marchesini G,
      2. Brizi M,
      3. Morselli-Labate AM,
      4. et al
      . Association of nonalcoholic fatty liver disease with insulin resistance. Am J Med 1999;107:450–5.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Ndumele CE,
      2. Nasir K,
      3. Conceiçao RD,
      4. et al
      . Hepatic steatosis, obesity, and the metabolic syndrome are independently and additively associated with increased systemic inflammation. Arterioscler Thromb Vasc Biol 2011;31:1927–32.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Smits MM,
      2. Ioannou GN,
      3. Boyko EJ,
      4. et al
      . Non-alcoholic fatty liver disease as an independent manifestation of the metabolic syndrome: results of a US national survey in three ethnic groups. J Gastroenterol Hepatol 2013;28:664–70.
      OpenUrlCrossRefPubMed
    7. ↵
      1. Villanova N,
      2. Moscatiello S,
      3. Ramilli S,
      4. et al
      . Endothelial dysfunction and cardiovascular risk profile in nonalcoholic fatty liver disease. Hepatology 2005;42:473–80.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Sookoian S,
      2. Pirola CJ
      . Non-alcoholic fatty liver disease is strongly associated with carotid atherosclerosis: a systematic review. J Hepatol 2008;49:600–7.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Ghouri N,
      2. Preiss D,
      3. Sattar N
      . Liver enzymes, nonalcoholic fatty liver disease, and incident cardiovascular disease: a narrative review and clinical perspective of prospective data. Hepatology 2010;52:1156–61.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. McKimmie RL,
      2. Daniel KR,
      3. Carr JJ,
      4. et al
      . Hepatic steatosis and subclinical cardiovascular disease in a cohort enriched for type 2 diabetes: the Diabetes Heart Study. Am J Gastroenterol 2008;103:3029–35.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Petit JM,
      2. Guiu B,
      3. Terriat B,
      4. et al
      . Nonalcoholic fatty liver is not associated with carotid intima-media thickness in type 2 diabetic patients. J Clin Endocrinol Metab 2009;94:4103–6.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Bonapace S,
      2. Perseghin G,
      3. Molon G,
      4. et al
      . Nonalcoholic fatty liver disease is associated with left ventricular diastolic dysfunction in patients with type 2 diabetes. Diabetes Care 2012;35:389–95.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Mantovani A,
      2. Zoppini G,
      3. Targher G,
      4. et al
      . Non-alcoholic fatty liver disease is independently associated with left ventricular hypertrophy in hypertensive Type 2 diabetic individuals. J Endocrinol Invest 2012;35:215–18.
      OpenUrlPubMed
    14. ↵
      1. Yu CM,
      2. Sanderson JE,
      3. Marwick TH,
      4. et al
      . Tissue Doppler imaging a new prognosticator for cardiovascular diseases. J Am Coll Cardiol 2007;49:1903–14.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Lee SK,
      2. Kim SH,
      3. Cho GY,
      4. et al
      . Obesity phenotype and incident hypertension: a prospective community-based cohort study. J Hypertens 2013;31:145–51.
      OpenUrlCrossRefPubMed
    16. ↵
      1. Baik I,
      2. Shin C
      . Prospective study of alcohol consumption and metabolic syndrome. Am J Clin Nutr 2008;87:1455–63.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      World Health Organization Western Pacific Region. International Association for the Study of Obesity, International Obesity Task Force. The Asia-Pacific perspective: redefining obesity and its treatment. Health Communications Australia Pty Limit, 2000.
    18. ↵
      1. Limanond P,
      2. Raman SS,
      3. Lassman C,
      4. et al
      . Macrovesicular hepatic steatosis in living related liver donors: correlation between CT and histologic findings. Radiology 2004;230:276–80.
      OpenUrlCrossRefPubMed
    19. ↵
      1. Hübscher SG
      . Histological assessment of non-alcoholic fatty liver disease. Histopathology 2006;49:450–65.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Wannamethee G,
      2. Ebrahim S,
      3. Shaper AG
      . Gamma-glutamyl transferase: determinants and association with mortality from ischemic heart disease and all causes. Am J Epidemiol 1995;142:699–708.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Goland S,
      2. Shimoni S,
      3. Zornitzki T,
      4. et al
      . Cardiac abnormalities as a new manifestation of nonalcoholic fatty liver disease: echocardiographic and tissue Doppler imaging assessment. J Clin Gastroenterol 2006;40:949–55.
      OpenUrlCrossRefPubMed
    22. ↵
      1. Fallo F,
      2. Dalla Pozza A,
      3. Sonino N,
      4. et al
      . Non-alcoholic fatty liver disease is associated with left ventricular diastolicdysfunction in essential hypertension. Nutr Metab Cardiovasc Dis 2009;19:646–53.
      OpenUrlCrossRefPubMed
    23. ↵
      1. Perseghin G,
      2. Lattuada G,
      3. De Cobelli F,
      4. et al
      . Increased mediastinal fat and impaired left ventricular energy metabolism in young men with newly found fatty liver. Hepatology 2008;47:51–8.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Rijzewijk LJ,
      2. Jonker JT,
      3. van der Meer RW,
      4. et al
      . Effects of hepatic triglyceride content on myocardial metabolism in type 2 diabetes. J Am Coll Cardiol 2010;56:225–33.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Targher G,
      2. Bertolini L,
      3. Padovani R,
      4. et al
      . Non-alcoholic fatty liver disease is associated with carotid artery wall thickness in diet-controlled type 2 diabetic patients. J Endocrinol Invest 2006;29:55–60.
      OpenUrlPubMedWeb of Science
    26. ↵
      1. Vlachopoulos C,
      2. Manesis E,
      3. Baou K,
      4. et al
      . Increased arterial stiffness and impaired endothelial function in nonalcoholic Fatty liver disease: a pilot study. Am J Hypertens 2010;23:1183–9.
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Salvi P,
      2. Ruffini R,
      3. Agnoletti D,
      4. et al
      . Increased arterial stiffness in nonalcoholic fatty liver disease: the Cardio-GOOSE study. J Hypertens 2010;28:1699–707.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Targher G,
      2. Bertolini L,
      3. Rodella S,
      4. et al
      . NASH predicts plasma inflammatory biomarkers independently of visceral fat in men. Obesity (Silver Spring) 2008;16:1394–9.
      OpenUrlCrossRefPubMed
    29. ↵
      1. Pagano C,
      2. Soardo G,
      3. Esposito W,
      4. et al
      . Plasma adiponectin is decreased in nonalcoholic fatty liver disease. Eur J Endocrinol 2005;152:113–18.
      OpenUrlAbstract/FREE Full Text
    30. ↵
      1. Pickhardt PJ,
      2. Park SH,
      3. Hahn L,
      4. et al
      . Specificity of unenhanced CT for non-invasive diagnosis of hepatic steatosis: implications for the investigation of the natural history of incidental steatosis. Eur Radiol 2012;22:1075–82.
      OpenUrlCrossRefPubMed
    View Abstract

    Footnotes

    • NHK and JP contributed equally to this study.

    • Contributors NHK and SHK were involved in conception of the study design, interpretation of data and drafting the manuscript. YHK, DHK and G-YC reviewed the manuscript. IB, HEL, JON and JP contributed to discussion and edited the manuscript. EJK and JBL were the primary statistical analysts. SKL and CS revised the manuscript for important intellectual content.

    • Funding This study was supported by grants (2009-E00454-00 and 2010-E71001-00) from the Korean Centers for Disease Control and Prevention, Ministry for Health and Welfare, and was partially supported by a grant (Z1300161) from the Korea University Medical Center.

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval This study was approved by the Human Subjects Review Committee at the Korea University Ansan Hospital.

    • Provenance and peer review Not commissioned; externally peer reviewed.