ORIGINAL CONTRIBUTIONS

nature publishing group

671

Using an Electronic Medical Records Database to

Identify Non-Traditional Cardiovascular Risk Factors in
Nonalcoholic Fatty Liver Disease
Kathleen E. Corey, MD, MPH, MMSc1,2, Uri Kartoun, PhD2,3, Hui Zheng, PhD2,4, Raymond T. Chung, MD1,2 and Stanley Y. Shaw, MD, PhD2,3
OBJECTIVES:

Among adults with nonalcoholic fatty liver disease (NAFLD), 25% of deaths are attributable to
cardiovascular disease (CVD). CVD risk reduction in NAFLD requires not only modication of
traditional CVD risk factors but identication of risk factors unique to NAFLD.

METHODS:

In a NAFLD cohort, we sought to identify non-traditional risk factors associated with CVD. NAFLD
was determined by a previously described algorithm and a multivariable logistic regression model
determined predictors of CVD.

RESULTS:

Of the 8,409 individuals with NAFLD, 3,243 had CVD and 5,166 did not. On multivariable analysis,
CVD among NAFLD patients was associated with traditional CVD risk factors including family history
of CVD (OR 4.25, P=0.0007), hypertension (OR 2.54, P=0.0017), renal failure (OR 1.59, P=0.04),
and age (OR 1.05, P<0.0001). Several non-traditional CVD risk factors including albumin, sodium,
and Model for End-Stage Liver Disease (MELD) score were associated with CVD. On multivariable
analysis, an increased MELD score (OR 1.10, P<0.0001) was associated with an increased risk of
CVD. Albumin (OR 0.52, P<0.0001) and sodium (OR 0.96, P=0.037) were inversely associated with
CVD. In addition, CVD was more common among those with a NAFLD brosis score >0.676 than
those with a score 0.676 (39 vs. 20%, P<0.0001).

CONCLUSIONS: CVD in NAFLD is associated with traditional CVD risk factors, as well as higher MELD scores and

lower albumin and sodium levels. Individuals with evidence of advanced brosis were more likely to
have CVD. These ndings suggest that the drivers of NAFLD may also promote CVD development and
progression.
Am J Gastroenterol 2016; 111:671676; doi:10.1038/ajg.2016.44; published online 1 March 2016

INTRODUCTION
Nonalcoholic fatty liver disease (NAFLD) is the most common
cause of liver disease in the United States, affecting an estimated
80 million adults (1). Nonalcoholic steatohepatitis (NASH) is the
progressive form of NAFLD and can lead to the development of
cirrhosis and hepatocellular carcinoma (25). Although liverrelated complications are frequent among those with NAFLD,
cardiovascular disease (CVD) is the most common cause of mortality, accounting for 25% of deaths (6). NAFLD is associated with
an increased prevalence of aortic and coronary atherosclerosis,
high-risk coronary plaques, and increased coronary artery
calcium scores. Further, NAFLD is associated with increased fatal

and non-fatal CVD events including acute coronary syndromes

(79).
The identification of CVD risk factors among the general population has been the focus of considerable investigation. Identifying
which patient characteristics confer an increased risk of CVD has
contributed to the understanding of CVD pathophysiology. Unlike
in the general population, little attention has been focused on
elucidating non-traditional CVD risk factors in NAFLD. A single
study evaluated the Framingham Risk Score, a composite score
of traditional risk factors including age, gender, cholesterol, highdensity lipoprotein (HDL) level, smoking status, and hypertension,
as a CVD predictor in NAFLD (10). Although the Framingham

1
Liver Center, Gastrointestinal Unit, Massachusetts General Hospital, Boston, MA, USA; 2Harvard Medical School, Boston, MA, USA; 3Center for Systems Biology,
Center for Assessment Technology and Continuous Health, Massachusetts General Hospital, Boston, MA, USA; 4Biostatistics Center, Massachusetts General
Hospital, Boston, MA, USA. Correspondence: Kathleen E. Corey, MD, MPH, MMSc, Liver Center, Gastrointestinal Unit, Massachusetts General Hospital, 55 Fruit
Street, Blake 4, Boston, MA 02114, USA.
E-mail: kcorey@partners.org
Received 4 November 2015; accepted 24 January 2016

2016 by the American College of Gastroenterology

The American Journal of GASTROENTEROLOGY

LIVER

see related editorial on page x

LIVER

672

Corey et al.

Risk Score accurately predicted a 10-year CVD risk, none of its

individual components were found to be predictors of CVD, and
no novel risk factors were evaluated. Thus, little is known about the
value of non-traditional CVD risk factors in NAFLD.
CVD events are believed to be rare in individuals with
chronic and end-stage liver disease (11). The systemic vasodilatation and decreased lipid synthesis that accompany liver disease
are thought to decrease CVD risk (12,13). However, NAFLD
is, in many ways, distinct from other causes of liver disease.
Even in late stages, NAFLD is associated with dyslipidemia and
hypertension, which confer increased CVD risk (14). We hypothesize that the same drivers of progressive NAFLD, systemic
inflammation, lipid oxidation, and endothelial dysfunction, may
also drive the development of CVD, making CVD increasingly
prevalent as NAFLD progresses and associated with markers
of liver disease progression. By using a large electronic medical
record (EMR)-based cohort of 8,409 individuals with NAFLD, we
evaluated those with and without CVD to identify unique CVD
risk factors.

ous intervention, or sudden death. For laboratory variables, when

more than one value was present the average of all available values
was used.
NAFLD fibrosis score (NFS) was calculated according to the
published formula (16):

Model for End-Stage Liver Disease (MELD) score was calculated according to the published formula (17):

METHODS
Patients and data for the present study were drawn from a previously described cohort created from the Partners HealthCare EMR utilizing the Partners Research Patient Data Registry
(RPDR). This centralized clinical data registry contains data from
all institutions in the Partners HealthCare System and includes
data on ~10 million patients with ~2.3 billion EMR facts. We utilized data from the Massachusetts General Hospital and Brigham
and Womens Hospital, both in Boston, that serve the greater
Northeast United States.
NAFLD was defined using a previously validated algorithm for
the identification of NAFLD in an EMR database (15). The algorithm calculates a NAFLD probability per patient based on the
most recent triglycerides measurement, the total number of billing
codes for NAFLD (ICD-9 571.8 or 571.9), and the total number
of mentions of NAFLD in clinical narrative notes. The algorithm
incorporates text processing to identify clinical narrative notes
associated with NAFLD. As a first step, the algorithm was applied
to the RPDR cohort to identify all individuals with NAFLD. As a
second step, patients with either a diagnosis of cirrhosis or a nonviral hepatitis were excluded. In total, 8,409 adults aged 18 years
exceeded the NAFLD probability threshold of 0.85 and were considered in our analysis.
CVD was considered present when an individual had 1 ICD-9
or CPT code for myocardial infarction, CVD, ischemic heart disease, angina, or peripheral vascular disease. Comorbidities were
determined by 1 ICD-9 or CPT code for the comorbidity over
their lifetime prior to the diagnosis of CVD. We extracted from
the notes expressions to determine an individuals most recent
smoking status (past, present, never). In addition, to determine
whether the patient had a family history of CVD, we identified in
clinical narrative notes the indication of at least one family member being reported as having myocardial infarction, heart attack,
angina, coronary artery bypass surgery, cardiovascular percutaneThe American Journal of GASTROENTEROLOGY

Statistical analysis

Categorical variables were compared using the 2-test. Continuous

variables were compared using the t-test or the MannWhitneys
test, as appropriate. To determine odds ratio (OR) for the variables associated with the presence of NAFLD, logistic regression was performed. The following variables, based on statistical
significance and clinical relevance, were included: age, gender,
ethnicity, diabetes, hypertension, dyslipidemia, obstructive sleep
apnea, non-HDL cholesterol, renal failure, low-density lipoprotein level (LDL), alanine aminotransferase level, NFS, MELD
score, and family history of CVD. Statistical analysis was performed on SAS 9.4 (SAS Institute, Cary, NC). We examined the
collinearity among covariates in the multivariable model based on
their variance inflation factor. The multivariable model consists
of covariates that do not have overly high variance inflation factor (maximum VIF 3.3). This study was approved by the Partners
Healthcare Human Research Committee that serves as the institutional review board for both Brigham and Womens Hospital and
Massachusetts General Hospital.

RESULTS
Baseline characteristics

Of the 8,409 individuals, 3,243 individuals had CVD, whereas

5,166 individuals had no evidence of CVD (Table 1). Individuals
with NAFLD and CVD were older (61.9 vs. 52.3 years, P<0.0001),
more likely to be male (55.2 vs. 51.3%, P=0.0006), and Caucasian
(92.6 vs. 87.1%, P<0.0001). There was no difference in the mean
BMI or prevalence of obesity between groups (33.3 vs. 33.4 kg/
m2, P=0.30). All variables considered were calculated based on the
available values or measurements from date of birth of a patient to
the last EMR fact that was available in the cohorti.e., September
2010.
VOLUME 111 | MAY 2016 www.amjgastro.com

Cardiovascular Risk Factors in Fatty Liver Disease

673

250

Table 1. Baseline characteristics

NAFLD +CVD

P value

52.314.1

61.913.1

<0.0001

Mean ages.d.
(years)

Male

2,652 (51.3%)

1,789 (55.2%)

0.0006

Female

2,514 (48.7%)

1,454 (44.8%)

0.0006

3,399 (87.1%)

2,574 (92.6%)

<0.0001

African American

296 (7.6%)

146 (5.2%)

<0.0001

Other

206 (5.3%)

60 (2.2%)

<0.0001

BMIs.d (kg/m2)

33.47.6

33.310.6

0.30

Obesity; no. (%)

1,747 (33.8%)

1,074 (33.1%)

0.51

Diabetes mellitus;
no. (%)

3,340 (64.7%)

2,651 (82.0%)

<0.0001

Hypertension;
no. (%)

2,946 (57.0%)

2,745 (84.6%)

<0.0001

45
40

Non-HDL-C

Cholesterol

HDL

**

35
30
25

NAFLD CVD

20

NAFLD +CVD

15
10
5
0
ESR (mm/h)

Family history of
CVD; no. (%)

3,014 (58.3%)

2,172 (67.0%)

<0.0001

363 (7.0%)

546 (16.8%)

<0.0001

LDLs.d. (mg/dl)

112.538.2

106.338.8

<0.0001

Non-HDL-Cs.d.
(mg/dl)

189.763.6

182.358.8

<0.0001

Albumins.d. (g/dl)

3.90.69

3.70.63

<0.0001

Sodiums.d.
(mmol/l)

139.02.9

138.02.7

<0.0001

MELD scores.d.

10.25.0

12.35.7

<0.0001

Renal failure;
no. (%)

BMI, body mass index; CVD, cardiovascular disease; HDL-C, high-density

lipoprotein cholesterol; MELD, Model for End-Stage Liver Disease; NAFLD,
nonalcoholic fatty liver disease.

100.00%

60.00%

40.00%

20.00%

NAFLD CVD
NAFLD +CVD

on
C

PD

ep
r

/a

es

st
h

si

SA
O
/d
et
y
An

xi

ys

lip

id

em

TN

ia

0.00%

NAFLD +CVD

50

LDL

White

NAFLD CVD

Ethnicity; no. (%)

80.00%

150

100

Gender; no. (%)

Figure 1. Prevalence of comorbidities by CVD status in NAFLD. COPD,

chronic obstructive pulrnonary disease; CVD, cardiovascular disease;
DM, diabetes; HTN, hypertension; NAFLD, nonalcoholic fatty liver disease;
OSA, sleep apnea. *P<0.0001.

Traditional CVD risk factors

Traditional risk factors for CVD were more prevalent in individuals with NAFLD and CVD compared with those with
2016 by the American College of Gastroenterology

CRP (mg/l)

Figure 2. Comparison between laboratory variables by CVD status in

NAFLD. (a) Lipid levels by CVD status in NAFLD. (b) ESR and CRP by CVD
status in NAFLD. CRP, c-reactive protein; CVD, cardiovascular disease;
ESR, erythrocyte sedimentation rate; HDL, high-density lipoprotein; LDL,
low-density lipoprotein; NAFLD, nonalcoholic fatty liver disease; non-HDL-C,
non-high-density lipoprotein cholesterol. *P<0.0001, **P= 0.007.

NAFLD alone on univariate analysis (Figure 1). Type 2 diabetes (82.0 vs. 64.7%, P<0.0001) was more frequent and median
HbA1C (7.3 vs. 7.08%, P=0.0009) was significantly higher in
those with both CVD and NAFLD compared with those with
NAFLD alone. Hypertension (84.6 vs. 57.0%, P<0.0001), family
history of CVD (67.0 vs. 58.3%, P<0.0001), and current or past
tobacco use (53.7 vs. 41.1%, P<0.0001) were associated with
the presence of CVD in NAFLD patients. Other comorbidities including obstructive sleep apnea, anxiety and depression,
chronic obstructive pulmonary disease, and asthma were more
frequent in NAFLD and CVD when compared with NAFLD
alone (Figure 1).
Dyslipidemia and statin use were more frequent in individuals with both NAFLD and CVD than those with NAFLD alone
(75.4 vs. 55.5%, P<0.0001 and 49.0 vs. 26.1%, P<0.0001, respectively). Mean LDL (106.3 vs. 112.5 mg/dl, P<0.0001), total cholesterol (214.5 vs. 221.9 mg/dl, P<0.0001), and non-HDL cholesterol
(182.3 vs. 189.7 mg/dl, P<0.0001) were lower in those with NAFLD
and CVD compared with those with NAFLD alone (Figure 2a).
HDL levels were lower in those with CVD (35.9 vs. 37.9 mg/dl,
P<0.0001), although there was no difference in triglyceride levels. Other risk markers of CVD disease including ESR (40.7 vs.
33.2 mm/h, P<0.0001) and C-reactive protein (37.3 vs. 32.9 mg/l,
P=0.007) were higher in those with CVD and NAFLD (Figure 2b
and Figure 3).
A diagnosis of renal failure was more common in CVD and
NAFLD compared with those with only NAFLD (16.8 vs. 7.0%,
P<0.0001). Individuals with CVD and NAFLD had higher serum
The American Journal of GASTROENTEROLOGY

LIVER

NAFLD CVD

mg/dl

Variable

*
200

674

Corey et al.

12

Table 2. Factors associated with CVD in NAFLD on multivariable

analysisa

LIVER

10
8

Variable

4
2

Hypertension

2.54 (1.424.58)

0.0017

NAFLD +CVD

Renal failure

1.59 (1.012.49)

0.04

0
INR

Albumin
(g/dl)

Bilirubin
(mg/dl)

NAFLD
fibrosis
score

P value

NAFLD CVD

OR (95% CI)a

MELD score

Figure 3. Liver function by CVD status in NAFLD. CVD, cardiovascular

disease; INR, international normalized ratio; MELD, Model for End-Stage
Liver Disease; NAFLD, nonalcoholic fatty liver disease. *P<0.0001.

creatinine levels (1.35 vs. 1.16 mg/dl, P<0.0001) and lower estimated glomerular filtration rates (57.8 vs. 62.9 ml/min per 1.73 m2,
P<0.0001).

MELD score

1.10 (1.061.14)

<0.0001

Age (years)

1.05 (1.031.06)

<0.0001

Non-HDL-C (mg/dl)

1.01 (1.0011.012)

0.026

LDL (mg/dl)

0.99 (0.980.99)

0.008

Family history of CVD

4.25 (1.849.83)

0.0007

Albumin

0.52 (0.440.63)

<0.0001

Sodium

0.96 (0.930.99)

0.037

CVD, cardiovascular disease; HDL-C, high-density lipoprotein cholesterol; LDL,

low-density lipoprotein; MELD, Model for End-Stage Liver Disease; NAFLD,
nonalcoholic fatty liver disease.
a
Adjusted for gender, ethnicity, diabetes, dyslipidemia, ALT, family history of
CVD, and OSA.

Non-traditional risk factors for CVD

Several non-traditional risk factors for CVD and NAFLD were

identified on univariate analysis. CVD was associated with
decreased albumin levels (3.7 vs. 3.9 g/dl, P<0.0001) and platelet
counts (246.1 vs. 259.6 th/cumm, P<0.0001) compared with those
with NAFLD alone. In addition, patients with NAFLD and CVD
had increased total bilirubin (0.8 vs. 0.7 mg/dl, P<0.0001) and INR
(1.4 vs. 1.2, P<0.0001). NFSs (2.11.52 vs. 1.31.45, P<0.0001)
and mean MELD scores (12.35.7 vs. 10.25.0, P<0.0001) were
also significantly higher in those with both CVD and NAFLD
compared with those with NAFLD alone. No difference was seen
in mean AST levels between groups, although those without CVD
had slightly increased alanine aminotransferase level (45.8 vs.
45.5 U/l, P<0.0001) when compared with those with CVD.
Factors associated with CVD on regression analysis

On multivariable analysis after controlling for gender, ethnicity,

diabetes, dyslipidemia, alanine aminotransferase, and obstructive
sleep apnea, family history of CVD and hypertension were most
strongly associated with the presence of CVD (Table 2). Age, nonHDL cholesterol, and renal failure remained directly associated
with the presence of CVD. LDL was inversely associated with the
presence of CVD. We assessed the correlation between LDL level
and use of lipid lowering medication. Lipid lowering medication
use was inversely correlated with the LDL level (r=0.09, P<0.0001),
indicating that statin use was likely associated with CVD.
In addition, the non-traditional CVD risk factors MELD
score, albumin level, and sodium level were associated with
increased risk of CVD. MELD score was associated with an
increased risk of CVD after adjustment with an OR 1.10, 95%
confidence interval (CI) 1.061.14. Albumin and sodium levels
were inversely associated with risk of CVD, demonstrating that
low albumin and sodium levels confer an increased risk of CVD
(OR 0.52, 95% CI 0.440.63 and OR 0.96, 95% CI 0.930.99,
respectively).
The American Journal of GASTROENTEROLOGY

CVD by NFS

Histologic diagnosis of NASH was not available in the present

cohort. To evaluate whether CVD prevalence differed in those
with NASH and advanced fibrosis, we evaluated the CVD prevalence by the NFS. Individuals with NAFLD and a NFS>0.676 had
a significantly higher prevalence of CVD compared with those
with NFS0.676 (39 vs. 20%, P<0.0001). This finding further suggests an association between advanced fibrosis and CVD.

DISCUSSION
The present study demonstrates that among individuals with
NAFLD, MELD score, albumin, and sodium are non-traditional
predictors of CVD. Further, we confirm the validity of our model
by demonstrating that several known risk factors for CVD in the
general population are associated with CVD in NAFLD.
We found that MELD score, albumin, and sodium levels were
associated with a diagnosis of CVD. Each of these factors is
known to be independently associated with disease progression
in chronic liver disease (1826). Further, we demonstrated that
those with advanced fibrosis as predicted by NFS had a higher
prevalence of CVD, also suggesting that advanced liver disease
is associated with increased risk of CVD. Our findings demonstrate that CVD may be associated with progressive liver disease
among those with NAFLD and suggest that similar processes
may drive the development of atherosclerosis, steatohepatitis,
and liver fibrosis. This finding is counter to the widely held belief
that CVD events are less frequent in end-stage liver disease in
part due to systemic vasodilatation and impaired lipid synthesis
(11). However, NAFLD may likely be an exception to this rule
secondary to the associated systemic inflammatory response,
endothelial dysfunction, and lipid peroxidation that accompanies
advanced NAFLD histology and can impact the development of
atherosclerotic disease (2730). Furthermore, hypercoagulablity
VOLUME 111 | MAY 2016 www.amjgastro.com

and impaired fibrinolysis found in NASH may also contribute

to CVD progression (31,32). This finding has several important
implications. First, a relationship between worsening liver disease
and CVD suggests that individuals with advanced liver disease
from NAFLD and those being evaluated for liver transplantation
should undergo rigorous evaluation of CVD and CVD risk management. In addition, this finding further strengthens the link
between NAFLD and CVD and suggests that treatment of one
condition could positively impact the other.
In the present study, we confirm that traditional CVD risk
factors including age, family history, hypertension, and renal
failure are risk factors for CVD in NAFLD. These findings
demonstrate the ability of our algorithm (15) and the cohort
that we created to accurately identify CVD and comorbidities.
Family history of CVD was most strongly associated with CVD
in individuals with NAFLD (OR 4.25, 95% CI 1.849.83), a
previously unreported finding. Family history of CVD in the
general population is a predictor of CVD-related death in men
and women but has not been evaluated in a NAFLD population
(3335). This association suggests that a genetic component to
CVD risk in NAFLD patients may exist, and further evaluation
is needed in a prospective cohort. Traditional risk factors of age,
renal failure, and hypertension were also associated with CVD
in NAFLD (36).
The LDL level is a known risk factor for the development of
CVD. However, in the present study, LDL was inversely associated
with CVD (OR 0.99, 95% CI 0.980.99). Although the inverse relationship between LDL and CVD prevalence in NAFLD may seem
to contradict data in non-NAFLD cohorts, we believe that this
finding is due to the significantly higher frequency of lipid lowering medication use in those with CVD compared with those with
NAFLD alone (49.0 vs. 26.1%, P<0.0001).
Our study has several important limitations. The cross-sectional design of our study allows for assessment of factors associated with the presence of CVD in NAFLD but does not allow
us to comment on causality of those risk factors in the pathogenesis of CVD in NAFLD. However, it does allow for the identification of several novel CVD-associated variables that can be
further assessed in prospective cohorts. Our study also uses a
validated algorithm to identify NAFLD, but liver histology was
not available to differentiate between steatosis and NASH or to
determine fibrosis stage. To address this, NFSs, which serve as a
proxy for the presence of NASH and advanced fibrosis, were calculated (16). Our study assessed MELD scores in a population
that was not confined to those with cirrhosis. As a result, other
causes of an elevated MELD score (e.g., anti-coagulation leading to an increased INR) are possible. However, the MELD has
been demonstrated to have predictive value outside of a cirrhotic
population and in our study was accompanied by positive correlations between CVD and other markers of liver disease including
albumin and platelet count (19,37). In addition, comorbidities in
our study were defined by the presence of one or more diagnostic
codes of that condition. Further, patients with CVD may more
frequently attend medical appointments, have more frequent laboratory tests, and may confound our results.
2016 by the American College of Gastroenterology

In conclusion, MELD score, sodium, and albumin levels are

predictors of CVD in NAFLD. Further evaluation is needed to
further elucidate the relationship between progressive CVD and
NAFLD.
ACKNOWLEDGMENTS

We acknowledge Dr Mary Rinella for her critical review of this

manuscript.
CONFLICT OF INTEREST

Guarantor of the article: Kathleen E. Corey, MD, MPH, MMSc.

Specific author contributions: Study planning, conducting the
study, data collection and interpretation, and drafting of the
manuscript: Kathleen E. Corey and Uri Kartoun; data analysis and
interpretation: Hui Zheng; study planning, data interpretation, and
editing the manuscript: Raymond T. Chung and Stanley Y. Shaw. All
the authors have approved the final submitted draft.
Financial Support: This study was funded in part by grants from the
NIH K23 DK099422 (KEC), NIH U54 LM008748 (SYS), and NIH
DK 078772 (RTC).
Potential competing interests: None.

Study Highlights
WHAT IS CURRENT KNOWLEDGE

Nonalcoholic fatty liver disease (NAFLD) is an independent

risk factor for cardiovascular disease (CVD).
Risk factors for CVD among those with NAFLD are not welldocumented.

WHAT IS NEW HERE

CVD in those with NAFLD is associated with traditional

CVD risk factors including age, hypertension, renal failure,

and family history of CVD.

MELD score, albumin, and sodium are also associated with

CVD in NAFLD.
Progressive NAFLD may be associated with worsening CVD.

REFERENCES
1. Browning JD, Szczepaniak LS, Dobbins R et al. Prevalence of hepatic
steatosis in an urban population in the United States: impact of ethnicity.
Hepatology 2004;40:138795.
2. Matteoni CA, Younossi ZM, Gramlich T et al. Nonalcoholic fatty liver
disease: a spectrum of clinical and pathological severity. Gastroenterology
1999;116:14139.
3. Soderberg C, Stal P, Askling J et al. Decreased survival of subjects with
elevated liver function tests during a 28-year follow-up. Hepatology
2010;51:595602.
4. Bhala N, Angulo P, van der Poorten D et al. The natural history of nonalcoholic fatty liver disease with advanced fibrosis or cirrhosis: an international
collaborative study. Hepatology 2011;54:120816.
5. Rafiq N, Bai C, Fang Y et al. Long-term follow-up of patients with nonalcoholic fatty liver. Clin Gastroenterol Hepatol 2009;7:2348.
6. Adams LA, Lymp JFSt, Sauver J et al. The natural history of nonalcoholic
fatty liver disease: a population-based cohort study. Gastroenterology
2005;129:11321.
7. Puchner SB, Lu MT, Mayrhofer T et al. High-risk coronary plaque at
coronary CT angiography is associated with nonalcoholic fatty liver disease,

The American Journal of GASTROENTEROLOGY

675

LIVER

Cardiovascular Risk Factors in Fatty Liver Disease

LIVER

676

Corey et al.

independent of coronary plaque and stenosis burden: results from the

ROMICAT II trial. Radiology 2015;274:693701.
8. Stepanova M, Younossi ZM. Independent association between nonalcoholic
fatty liver disease and cardiovascular disease in the US population. Clin
Gastroenterol Hepatol 2012;10:64650.
9. Targher G, Bertolini L, Rodella S et al. Nonalcoholic fatty liver disease is
independently associated with an increased incidence of cardiovascular
events in type 2 diabetic patients. Diabetes Care 2007;30:211921.
10. Treeprasertsuk S, Leverage S, Adams LA et al. The Framingham risk score
and heart disease in nonalcoholic fatty liver disease. Liver Int 2012;32:94550.
11. Marchesini G, Ronchi M, Forlani G et al. Cardiovascular disease in
cirrhosisa point-prevalence study in relation to glucose tolerance. Am J
Gastroenterol 1999;94:65562.
12. Albillos A, Rossi I, Cacho G et al. Enhanced endothelium-dependent
vasodilation in patients with cirrhosis. Am J Physiol 1995;268:G459G464.
13. Bosch J, Garcia-Pagan JC. Complications of cirrhosis. I. Portal hypertension. J Hepatol 2000;32:14156.
14. Targher G. Non-alcoholic fatty liver disease, the metabolic syndrome and the
risk of cardiovascular disease: the plot thickens. Diabet Med 2007;24:16.
15. Corey K, Kartoun U, Zheng H et al. Development and validation of an
algorithm to identify nonalcoholic fatty liver disease in the electronic
medical record. Dig Dis Sci 2015, 17.
16. Angulo P, Hui JM, Marchesini G et al. The NAFLD fibrosis score: a noninvasive system that identifies liver fibrosis in patients with NAFLD.
Hepatology 2007;45:84654.
17. Kamath PS, Kim WR. Advanced Liver Disease Study Group. The model for
end-stage liver disease (MELD). Hepatology 2007;45:797805.
18. Kamath PS, Wiesner RH, Malinchoc M et al. A model to predict survival in
patients with end-stage liver disease. Hepatology 2001;33:46470.
19. Dunn W, Jamil LH, Brown LS et al. MELD accurately predicts mortality in
patients with alcoholic hepatitis. Hepatology 2005;41:3538.
20. Telem DA, Schiano T, Goldstone R et al. Factors that predict outcome of
abdominal operations in patients with advanced cirrhosis. Clin Gastroenterol Hepatol 2010;8:4517.
21. Berman K, Tandra S, Forssell K et al. Incidence and predictors of 30-day
readmission among patients hospitalized for advanced liver disease.
Clin Gastroenterol Hepatol 2011;9:2549.
22. Bhangui P, Laurent A, Amathieu R et al. Assessment of risk for non-hepatic
surgery in cirrhotic patients. J Hepatol 2012;57:87484.
23. Singal AG, Rahimi RS, Clark C et al. An automated model using electronic
medical record data identifies patients with cirrhosis at high risk for readmission. Clin Gastroenterol Hepatol 2013;11:133541.

The American Journal of GASTROENTEROLOGY

24. Younossi ZM, Henry L, Stepanova M. A new comorbidity model for predicting mortality in patients with cirrhosis: does it work? Gastroenterology
2014;146:1924.
25. Sharma P, Schaubel DE, Goodrich NP et al. Serum sodium and survival
benefit of liver transplantation. Liver Transplant 2015;21:30813.
26. Biggins SW, Rodriguez HJ, Bacchetti P et al. Serum sodium predicts
mortality in patients listed for liver transplantation. Hepatology 2005;41:
329.
27. De Vito R, Alisi A, Masotti A et al. Markers of activated inflammatory cells
correlate with severity of liver damage in children with nonalcoholic fatty
liver disease. Int J Mol Med 2012;30:4956.
28. Sookoian S, Castano GO, Burgueno AL et al. Circulating levels and hepatic
expression of molecular mediators of atherosclerosis in nonalcoholic fatty
liver disease. Atherosclerosis 2010;209:58591.
29. Thuy S, Ladurner R, Volynets V et al. Nonalcoholic fatty liver disease in
humans is associated with increased plasma endotoxin and plasminogen activator inhibitor 1 concentrations and with fructose intake. J Nutr
2008;138:14525.
30. Targher G, Bertolini L, Rodella S et al. NASH predicts plasma inflammatory
biomarkers independently of visceral fat in men. Obesity 2008;16:13949.
31. Tripodi A, Anstee QM, Sogaard KK et al. Hypercoagulability in cirrhosis:
causes and consequences. J Thromb Haemost 2011;9:171323.
32. Targher G, Chonchol M, Miele L et al. Nonalcoholic fatty liver disease as
a contributor to hypercoagulation and thrombophilia in the metabolic
syndrome. Semin Thromb Hemost 2009;35:27787.
33. Barrett-Connor E, Khaw K. Family history of heart attack as an independent predictor of death due to cardiovascular disease. Circulation
1984;69:10659.
34. Lloyd-Jones DM, Nam BH, D'Agostino RBSr. et al. Parental cardiovascular disease as a risk factor for cardiovascular disease in middle-aged
adults: a prospective study of parents and offspring. J Am Med Assoc
2004;291:220411.
35. Andresdottir MB, Sigurdsson G, Sigvaldason H et al. Fifteen percent of
myocardial infarctions and coronary revascularizations explained by family
history unrelated to conventional risk factors. The Reykjavik Cohort Study.
Eur Heart J 2002;23:165563.
36. Jousilahti P, Vartiainen E, Tuomilehto J et al. Sex, age, cardiovascular risk
factors, and coronary heart disease: a prospective follow-up study of 14 786
middle-aged men and women in Finland. Circulation 1999;99:116572.
37. Wong VW, Chim AM, Wong GL et al. Performance of the new MELD-Na
score in predicting 3-month and 1-year mortality in Chinese patients with
chronic hepatitis B. Liver Transplant 2007;13:122835.

VOLUME 111 | MAY 2016 www.amjgastro.com