Objective Non-alcoholic fatty liver disease (NAFLD) is the hepatic manifestation of metabolic syndrome and is associated with cardiovascular risk. The aim of this study was to determine the role of fatty liver in predicting coronary artery disease and clinical outcomes in patients undergoing coronary angiogram.
Methods This was a prospective cohort study carried out in a University hospital. Consecutive patients who underwent coronary angiogram had ultrasound screening for fatty liver. Significant cardiovascular disease was defined as ≥50% stenosis in at least one coronary artery. The primary outcome was a composite end point comprising cardiovascular deaths, non-fatal myocardial infarction and the need for further coronary intervention during prospective follow-up.
Results Among 612 recruited patients, 356 (58.2%) had fatty liver by ultrasonography, 318 (52.0%) had elevated serum alanine aminotransferase and 465 (76.0%) had significant coronary artery disease. Coronary artery disease occurred in 84.6% of patients with fatty liver and 64.1% of those without fatty liver (p<0.001). After adjusting for demographic and metabolic factors, fatty liver (adjusted OR 2.31; 95% CI 1.46 to 3.64) and alanine aminotransferase level (adjusted OR 1.01; 95% CI 1.00 to 1.02) remained independently associated with coronary artery disease. At a mean follow-up of 87±22 weeks, 30 (10.0%) patients with fatty liver and 18 (11.0%) patients without fatty liver reached the composite clinical end point (p=0.79).
Conclusions In patients with clinical indications for coronary angiogram, fatty liver is associated with coronary artery disease independently of other metabolic factors. However, fatty liver cannot predict cardiovascular mortality and morbidity in patients with established coronary artery disease.
- myocardial infarction
- alanine aminotransferase
- fatty liver
- liver function test
Statistics from Altmetric.com
- myocardial infarction
- alanine aminotransferase
- fatty liver
- liver function test
Significance of this study
What is already known about this subject?
Non-alcoholic fatty liver disease (NAFLD) is one of the most common chronic liver diseases, and is associated with metabolic syndrome and systemic inflammation.
NAFLD is associated with coronary artery disease in the general population. However, it is unclear if it is a useful marker to predict clinical outcomes in patients with established coronary artery disease.
What are the new findings?
In patients with clinical indications for coronary angiogram, fatty liver and a high serum alanine aminotransferase level were associated with significant coronary artery disease.
The association between NAFLD and coronary artery disease was independent of other demographic and metabolic factors.
Among patients with established coronary artery disease, those with and without fatty liver had a similar incidence of cardiovascular deaths, non-fatal myocardial infarction and need for revascularisation.
How might it impact on clinical practice in the foreseeable future?
Clinicians should be aware of the risk of coronary artery disease in patients with NAFLD. However, fatty liver cannot be used as a marker to predict clinical outcomes in patients with coronary artery disease.
Non-alcoholic fatty liver disease (NAFLD) is one of the most common chronic liver diseases worldwide.1 2 Progressive liver injury is often observed, and some patients may develop cirrhosis and hepatocellular carcinoma.3–5 It is associated with increased mortality from cardiovascular diseases, malignancy and hepatic complications.6–8 Patients with NAFLD usually have metabolic syndrome and diabetes mellitus, and have dysregulated secretion of proinflammatory and anti-inflammatory cytokines.9 10 Recently, NAFLD has also been found to be associated with extrahepatic disorders such as colorectal neoplasm.11
Owing to the close relationship between NAFLD and metabolic syndrome, it is not surprising that patients with NAFLD have increased risk of coronary artery disease as predicted by the Framingham risk score.12 13 Radiological atherosclerosis and endothelial dysfunction are also common in patients with NAFLD.14–17 In selected populations such as patients with diabetes, NAFLD also increases the odds of developing cardiovascular events.18 19
Based on the existing data, it is reasonable to recommend comprehensive metabolic and/or cardiovascular assessment in patients with NAFLD.20 However, it is uncertain if fatty liver provides additional information on the risk of coronary artery disease when other demographic and metabolic risk factors are considered. Recently, there is also hot debate on the association between fatty liver and cardiovascular events because most previous studies did not focus on incident clinical events.21 22 In addition, it is unclear if fatty liver may be a marker for clinical outcome in patients with established coronary artery disease.
In this prospective study, we aimed to study the role of fatty liver in predicting coronary artery disease and clinical outcomes in patients undergoing coronary angiogram.
This was a prospective cohort study. Consecutive adult patients aged ≥18 years who underwent coronary angiogram at the Prince of Wales Hospital, Hong Kong were recruited. We excluded patients with contraindications to coronary angiogram, excessive alcohol intake (>20 g per day in men and 10 g per day in women), and secondary causes of fatty liver (eg, chronic use of systemic corticosteroids or methotrexate). Patients with positive hepatitis B surface antigen, antibody against hepatitis C virus and antinuclear antibody titre >1/160 were also excluded. For logistic reasons, patients undergoing emergency primary percutaneous coronary intervention for acute myocardial infarction were not enrolled. The study protocol was approved by the Clinical Research Ethics Committee of The Chinese University of Hong Kong. All patients provided informed written consent.
Within 1 day before coronary angiogram, all subjects completed a structured questionnaire recording their smoking history, alcohol consumption, past medical history and drug use. Subjects were first classified as current drinkers, ex-drinkers (no alcohol consumption in the last 1 year) and non-drinkers. Current drinkers were asked to report the amount of alcohol consumption in a typical week in the past 1 year. Anthropometric measurements including body weight, body height, waist circumference and hip circumference were made. Body mass index (BMI) was calculated as weight (kg) divided by height (m) squared. Waist circumference was measured at a level midway between the lower rib margin and iliac crest with the tape all around the body in the horizontal position. Blood pressure was measured on both arms in the sitting position after resting for at least 30 min. Liver biochemistry, glucose and lipids were measured after fasting for at least 8 h. The upper limit of normal of the serum alanine aminotransferase level was 30 IU/l in men and 19 IU/l in women.
The diagnosis of fatty liver was based on ultrasonographic features of diffusely increased liver echogenicity greater than that of the kidney or spleen, vascular blurring and deep attenuation of the ultrasound signal.23 The ultrasound scans were performed by two investigators (VWSW and JL) using the GE Logiq E Ultrasound Machine (GE Healthcare, Hong Kong). Both operators were formally trained and had diagnostic accuracy validated against liver histology. During the study recruitment period (October 2007 to November 2008), the two operators performed 93 liver biopsies and prebiopsy ultrasonography examination (65 NAFLD, 20 chronic hepatitis B, eight other liver diseases). Fifty-four (58%) patients had steatosis involving ≥30% of hepatocytes. The sensitivity, specificity, positive predictive value and negative predictive value of ultrasonography in detecting at least 30% steatosis were 94, 97, 98 and 93%, respectively.
The diagnosis of coronary artery disease was based on cardiac catheterisation findings that were reviewed by at least two experienced cardiologists. The degree of stenosis of the left main stem, left anterior descending, left circumflex and right coronary arteries was recorded. Significant coronary artery disease was defined as the presence of at least 50% stenosis at one or more major coronary arteries.24 Triple vessel disease was defined as the presence of at least 50% stenosis at the left anterior descending, left circumflex and right coronary arteries. After the procedure, medical therapy, percutaneous coronary intervention and coronary artery bypass grafting were provided as clinically indicated.
Upon discharge, the patients were followed up at the cardiac clinic every 3–6 months. During each visit, patients were reviewed for recurrence of angina, functional status, blood pressure and latest drug regime.
The primary outcome was a composite end point comprising cardiovascular deaths, non-fatal myocardial infarction and the need for further coronary intervention. Secondary outcomes included overall mortality, cardiovascular deaths, non-fatal myocardial infarction, need for further coronary intervention and liver-related deaths. The diagnosis of myocardial infarction was based on clinical features, electrocardiography and troponin results as recommended by the Joint European Society of Cardiology/American College of Cardiology/American Heart Association/World Heart Federation Task Force for the Redefinition of Myocardial Infarction.25 All clinical events were verified by an independent, masked adjudication committee.
We assumed that 15% of subjects without NAFLD and 30% of those with NAFLD would have coronary artery disease and/or develop the primary outcome. A sample size of 480 subjects in total would achieve 80% power to detect this difference at a 5% significance level. Allowing a yearly drop-out rate of 5%, 600 subjects were required. In the current study, the database was frozen when the last recruited patient had reached 1 year of follow-up. A follow-up analysis was planned when the last recruited patient reached 5 years of follow-up.
Continuous variables were expressed as mean±SD or median (IQR), and compared between patients with and without NAFLD using unpaired t test or Mann–Whitney U test as appropriate. Categorical variables were compared using χ2 test. Binary logistic regression analysis was performed to identify independent factors associated with coronary artery disease. Time-to-event curves of patients with and without NAFLD were compared using the log-rank test. The Cox proportional hazard model was used to identify independent factors associated with the incident cardiovascular event. Factors with p values <0.05 by univariate analysis were included in the multivariate analysis. All statistical analysis was performed using SPSS version 16.0 software. Statistical significance was taken as a two-sided p value of <0.05.
Seven hundred and forty-seven patients were screened for eligibility. After excluding patients undergoing emergency percutaneous coronary intervention, those with viral hepatitis and excessive alcohol consumption, 612 patients were enrolled in the current study (figure 1). Overall, 356 (58.2%) patients had fatty liver. Three hundred and eighteen (52.0%) patients had an elevated serum alanine aminotransferase level. The fatty liver group contained more male patients and patients with diabetes, hypertension, obesity and dyslipidaemia (table 1). Fewer patients in the fatty liver group had coronary angiogram primarily for the evaluation of valvular heart disease. Ninety-three (15.2%) patients were modest drinkers (<20 g per day in men and <10 g per day in women), 30 (4.9%) were ex-drinkers and 489 (79.9%) were non-drinkers.
Fatty liver and coronary artery disease
Fatty liver was detected in 301 of 465 (64.7%) patients with coronary artery disease and 55 of 147 (37.4%) patients without coronary artery disease (p<0.001).
Patients with fatty liver had a higher degree of stenosis at the left anterior descending, left circumflex and right coronary arteries (table 2). Patients with fatty liver had a higher prevalence of significant stenosis at the left anterior descending (65.2% vs 52.7%), left circumflex (49.4% vs 34.4%) and right coronary arteries (52.8% vs 37.1%) than patients without fatty liver. Overall, stenosis of at least 50% at one or more coronary arteries was found in 84.6% of patients with fatty liver and 64.1% of patients without fatty liver (p<0.001). On the other hand, the prevalence of left main stem disease and triple vessel disease was similar between the two groups.
By univariate analysis, fatty liver, older age, male gender, diabetes, higher waist circumference, fasting glucose and alanine aminotransferase, and a lower high-density lipoprotein (HDL)-cholesterol level were associated with coronary artery disease (table 3). By multivariate analysis, fatty liver, older age, male gender, lower HDL-cholesterol and higher alanine aminotransferase level remained as independent factors associated with coronary artery disease.
Fatty liver and clinical outcome
To evaluate the interaction between fatty liver and cardiovascular outcomes, we included 465 patients with coronary artery disease and excluded the 147 patients without coronary artery disease from the subsequent analysis. Among 465 patients with coronary artery disease, 347 (74.6%) had percutaneous coronary intervention, 40 (8.6%) received coronary artery bypass grafting and 78 (16.8%) continued medical therapy. Among patients who had percutaneous coronary intervention, 158 (45.5%) had bare metal stents, 185 (53.3%) had drug-eluting stents and 4 (1.2%) patients did not have stenting. At a mean follow-up of 87±22 weeks, 30 of 301 (10.0%) patients with fatty liver and 18 of 164 (11.0%) patients without fatty liver reached the composite end point of cardiovascular death, non-fatal myocardial infarction or need for further coronary intervention (p=0.79 by log-rank test; figure 2 and table 4). Since the mode of cardiac intervention might have modified the risk of cardiovascular events, analysis was repeated by stratification. The risk of reaching the composite end point in patients with and without fatty liver remained similar in patients on medical therapy alone (p=0.70 by log-rank test), percutaneous coronary intervention (p=0.90) and coronary artery bypass grafting (p=0.088). Similarly, patients with fatty liver did not have increased risk of developing the composite end point after adjusting for the mode of therapy for coronary artery disease by the Cox proportional hazard model (adjusted HR 0.89; 95% CI 0.49 to 1.60; p=0.70).
By univariate analysis, diabetes, total cholesterol, low-density lipoprotein (LDL)-cholesterol and creatinine were associated with the composite end point (table 5). By multivariate analysis, only serum creatinine remained as an independent factor associated with the composite end point. Although LDL-cholesterol had a negative association with the composite end point by univariate analysis, patients with LDL-cholesterol <2.6 mmol/l had a higher prevalence of diabetes (45% vs 24%; p<0.001) and hypertension (74% vs 61%; p=0.003). The association became insignificant after adjusting for other factors.
Overall, 11 (3.7%) patients with fatty liver and 13 (7.9%) patients without fatty liver died (p=0.055 by log-rank test). In both groups, cardiovascular disease was the leading cause of death (table 4). Similarly, there was no significant difference in the incidence of non-fatal myocardial infarction and the need for further coronary intervention between patients with and without fatty liver.
In this large prospective study, fatty liver diagnosed by ultrasonography was associated with coronary artery disease independently of other demographic and metabolic factors. On the other hand, in patients with established coronary artery disease, fatty liver was not associated with adverse outcomes including cardiovascular death, non-fatal myocardial infarction and the need for further coronary intervention.
Since NAFLD and metabolic syndrome are closely related, a number of studies have suggested that patients with NAFLD have increased cardiovascular risk. In the general population, fatty liver and an elevated serum alanine aminotransferase level are associated with higher Framingham risk score.12 13 26 Fatty liver is also associated with increased carotid artery calcium and intimal media thickness.14 16 In a small study of 61 subjects, fatty liver was also associated with coronary artery stenosis of at least 50% as detected by CT.17 In the Valpolicella Heart Diabetes Study, 2103 patients with type 2 diabetes were followed for a median of 6.5 years.18 The risk of cardiovascular disease was doubled in patients with fatty liver at baseline. Recently, the same group of investigators confirmed the association between NAFLD and vascular diseases in 250 patients with type 1 diabetes.27 In long-term studies, NAFLD patients had an increased mortality rate, with cardiovascular disease being one of the leading causes of death.6–8 The interaction with metabolic factors also affects the risk of complications and mortality in patients with different chronic liver diseases.28–30 Our study was different from previous studies in that most patients have already had clinical suspicion of coronary artery disease. Even so, fatty liver remained as one of the most important factors associated with coronary artery disease by coronary angiogram.
One common criticism of studies investigating the association between fatty liver and coronary artery disease is that fatty liver may be just a marker of obesity and metabolic syndrome and provide no additional information.21 In particular, only a few studies investigated if fatty liver would improve disease prediction when other metabolic factors are considered. As such, our study showed that fatty liver was associated with coronary artery disease independently of other demographic and metabolic factors.
On the other hand, in patients with confirmed coronary artery disease, fatty liver was not associated with cardiovascular outcomes. Although this seems to deviate from previous observations of the association between NAFLD and cardiovascular mortality, the finding could be explained by the different study population.6–8 In previous studies, most patients did not have coronary artery disease. In contrast, cardiovascular events and deaths were assessed in patients with established coronary artery disease in our study. By combining the findings of our study and previous works, one might conclude that NAFLD correlates with incident coronary artery disease, but cannot be used as a prognostic marker in patients with established coronary artery disease. In the latter case, the prognosis may be governed by other factors.
What explains the lack of association between metabolic diseases and cardiovascular outcomes? In fact, in patients with established coronary artery disease, traditional cardiovascular risk factors such as dyslipidaemia and obesity are often found to be not associated with subsequent mortality and recurrent events.31 32 Paradoxically, aspirin use and lower BMI have been shown to be associated with recurrent coronary events in some studies.33 34 This is because the risk factors for index and recurrent events are often different due to index event bias.35 If a factor is important in the development of coronary artery disease, it may cause disease even when the burden of other risk factors is lower. For example, in our study, patients who had coronary artery disease despite a low LDL-cholesterol level actually had higher prevalence of diabetes and hypertension.
Serum transaminases are surrogate markers of fatty liver after the exclusion of other liver diseases. In population studies, alanine aminotransferase, aspartate aminotransferase and γ-glutamyltransferase are associated with cardiovascular events.36 37 Among various liver enzymes, γ-glutamyltransferase has the strongest association with cardiovascular mortality and events as shown by meta-analyses.21 38 39 That said, the association is weak in older populations and there are limited data to suggest that γ-glutamyltransferase improves disease prediction beyond other clinical factors.21 In our study, serum alanine aminotransferase had weak association with coronary artery disease independently of other metabolic factors, but did not predict cardiovascular events in patients with established coronary artery disease.
The mechanism linking NAFLD and coronary artery disease is not completely understood. NAFLD is associated with metabolic syndrome and central obesity, which are conditions closely related to cardiovascular risk.9 40 Its association with dyslipidaemia and hyperglycaemia was also independent of the amount of visceral fat.41 However, our data indicate that the association between NAFLD and coronary artery disease is independent of these metabolic factors. This suggests that NAFLD may contribute directly to the development of atherosclerosis, or NAFLD may be associated with other unmeasured cardiovascular risk factors. NAFLD is characterised by increased level of proinflammatory cytokines and marked insulin resistance.9 10 42 43 Systemic inflammation leads to activation of the nuclear factor-κB and c-Jun N-terminal kinase pathways, and may contribute to accelerated atherosclerosis.44 45 In addition, patients with non-alcoholic steatohepatitis, the active form of NAFLD, have increased production of procoagulant factors such as plasminogen activator inhibitor 1, fibrinogen and factor VII.18 Postprandial hyperglycaemia and dyslipidaemia are also common in patients with NAFLD.10 46 47 Previous studies have confirmed the association between postprandial hyperglycaemia and coronary artery disease.48
Our study had a few limitations. First, an ultrasound scan is an imperfect investigation for fatty liver. However, it was difficult to arrange sophisticated investigations such as magnetic resonance spectroscopy in this study because some patients had an unstable cardiovascular condition and needed an early coronary angiogram. Instead, an ultrasound examination was performed by trained doctors with performance matched with liver biopsy. Previous studies also showed that ultrasound scans had >95% sensitivity in detecting steatosis that affected >33% of hepatocytes.49 The use of ultrasound to diagnose fatty liver is also more feasible than magnetic resonance spectroscopy or liver biopsy in real-life clinical practice. Secondly, the number of clinical events during follow-up was relatively small and the follow-up duration after coronary angiogram was short. However, there was no trend of increased mortality or clinical events in the fatty liver group for up to 120 weeks. The second part of this project is ongoing and the analysis will be repeated after the last recruited patient has reached 5 years of follow-up. Finally, the mode of therapy, such as medical treatment, percutaneous coronary intervention and coronary artery bypass grafting, would have modified the disease outcome of the patients. However, the incidence of cardiovascular outcomes in patients with and without fatty liver remained similar after stratification by the mode of therapy or adjustment by multivariate analysis.
In conclusion, in patients with clinical indications for coronary angiogram, ultrasonographic fatty liver is associated with coronary artery disease, and the association is independent of other metabolic factors. On the other hand, fatty liver cannot predict cardiovascular mortality and morbidity in patients with established coronary artery disease.
Funding The study was partially supported by the research fund of the Department of Medicine and Therapeutics, The Chinese University of Hong Kong.
Competing interests None.
Ethics approval This study was conducted with the approval of the Clinical Research Ethics Committee, The Chinese University of Hong Kong (CRE-2007.317).
Provenance and peer review Not commissioned; externally peer reviewed.
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.