Carotid Artery Intima-media Thickness in Nonalcoholic Fatty Liver Disease

CLINICAL RESEARCH STUDY

Carotid Artery Intima-media Thickness in Nonalcoholic Fatty Liver Disease Anna Ludovica Fracanzani, MD,a Larry Burdick, MD,a Sara Raselli, PhD,b Paola Pedotti, MD,c Liliana Grigore, MD,b Gennaro Santorelli, MD,a Luca Valenti, MD,a Alessandra Maraschi, MD,a Alberico Catapano, MD,b Silvia Fargion, MDa a

Centro Malattie Metaboliche del Fegato, Dipartimento Medicina Interna, Università di Milano, Milano, Italy; bCentro Studi Aterosclerosi SISA, Dipartimento di Scienze Farmacologiche, Università di Milano, Milano, Italy; cUnità di Genetica ed Epidemiologia Molecolare, Ospedale Maggiore Policlinico, Fondazione IRCCS, Milano, Italy.

ABSTRACT PURPOSE: To evaluate, in patients with nonalcoholic fatty liver disease with no or mild alterations of liver function tests, carotid artery intima-media thickness and the presence of plaques and to define determinants of vascular damage. METHODS: A paired-sample case-control study: 125 patients with nonalcoholic fatty liver disease and 250 controls, without a prior diagnosis of diabetes, hypertension, and cardiovascular disease, matched for sex, age, and body mass index. B-mode ultrasound was used for evaluation of carotid intima-media thickness and presence of small plaques. RESULTS: A significant difference in mean values of intima-media thickness (0.89 ⫾ 0.26 and 0.64 ⫾ 0.14 mm, P ⫽ .0001) and prevalence of plaques (26 [21%] and 15 [6%], P ⬍.001) was observed in nonalcoholic fatty liver disease patients and controls. Variables significantly associated with intima-media thickness higher than 0.64 mm (median value in controls), in both patients and controls were: age (P ⫽ .0001), systolic blood pressure (P ⫽ .004), total and low-density lipoprotein cholesterol (P ⱕ.02 and P ⫽ .01, respectively), fasting glucose (P ⫽ .0001), and cardiovascular risk (P ⫽ .0001) and, only in controls, metabolic syndrome (P ⫽ .0001), HOMA-insulin resistance (P ⫽ .01), and body mass index (P ⫽ .0003). At multivariate logistic regression performed in the overall series of subjects, independent risk predictors of intima-media thickness higher than 0.64 mm were presence of steatosis (odds ratio [OR] ⫽ 6.9), age (OR 6.0), and systolic blood pressure (OR 2.3). CONCLUSION: Patients with nonalcoholic fatty liver disease, even with no or mild alterations of liver tests, should be considered at high risk for cardiovascular complications. © 2008 Elsevier Inc. All rights reserved. • The American Journal of Medicine (2008) 121, 72-78 KEYWORDS: IMT; metabolic syndrome; NAFLD

Nonalcoholic fatty liver disease includes a wide spectrum of liver diseases ranging from fatty liver alone, usually a benign and nonprogressive condition, to nonalcoholic steatohepatitis, which may progress to liver cirrhosis.1,2 Insulin resistance is the underlying condition that facilitates nonalcoholic fatty liver disease occurrence3 and is the key pathogenetic event of obesity, type II diabetes, and dyslipidemia. Recently, nonalcoholic fatty liver disease has been included Supported by grants from FIRST 2004-2005, COFIN 2004, and Ricerca Corrente IRCCS 2004-2005. Requests for reprints should be addressed to Silvia Fargion, MD, Dipartimento Medicina Interna, Pad Granelli Ospedale Maggiore IRCCS, Via F Sforza 35, Milano 20122, Italy. E-mail address: [email protected]

0002-9343/$ -see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2007.08.041

among the components of the metabolic syndrome,1,4,5 at high risk of coronary heart disease and cardiovascular complications.6-9 A high prevalence of fatty liver, evaluated by ultrasound, associated with hyperinsulinemia or insulin resistance and cardiovascular disease, also has been reported in the general population10-14 and in patients with documented coronary artery disease or stroke.15 Carotid B mode ultrasound imaging has been used for evaluation of carotid atherosclerosis.16,17 Intima-media thickness of carotid artery, a validated parameter for detecting subclinical atherosclerosis,18,19 has been shown to predict myocardial infarction and stroke.20,21 Recently, a relationship between nonalcoholic fatty liver disease and atherosclerosis has been reported in healthy men, and it has

Fracanzani et al

Carotid IMT in Nonalcoholic Fatty Liver Disease

73

study according to a protocol approved by the Senior Staff been suggested to reflect the overall adverse impact of the Committee of our institutions, a board comparable to an metabolic syndrome, in particular of insulin resistance and Institutional Review Board. Clinical, alcohol, and pharmaincreased visceral fat.22 In addition, endothelial dysfunction cologic history, smoking habits, BMI, waist circumference has been reported23 in nonalcoholic fatty liver disease and (measured in a standing position at the level of the umbilirelated to the metabolic syndrome. cus) and arterial blood pressure Carotid atherosclerosis has been (defined as the mean of the second detected in patients with nonalcoand third reading of 3 consecutive holic fatty liver disease in crossCLINICAL SIGNIFICANCE blood pressure measurements), sectional studies22,24,25, and more complete blood count, liver tests recently, carotid intima-media thick● Nonalcoholic fatty liver disease is alanine-aminotransferase and ness has been related to the severstrongly associated with atherosclerotic gamma-glutamyltransferase, fastity of liver damage.26 damage to the carotid artery, even in ing glucose, total and high-density Our aim was to evaluate intipatients without significantly impaired lipoprotein (HDL) cholesterol, trima-media thickness and carotid liver function, hypertension, or glyceride, uric acid, and hepatitis plaques in a case-control study indiabetes. B and C serological analysis were cluding a large series of patients available in all subjects. All tests with nonalcoholic fatty liver dis● Nonalcoholic fatty liver disease might were determined by standard labease with no or mild liver test albe inherently atherogenic. oratory procedures, insulin by a terations and matched controls, ● Patients with this disease are at high commercially purchased radioimand to define the determinants of risk for cardiovascular complications munoassay (RIA, Biochem Immuvascular damage. and should be screened for carotid arnosystems, Bologna, Italy). tery disease, regardless of liver function In all subjects, the cardiovascuMETHODS lar risk (Framingham risk score test results. [FRS]) was calculated according Patients to the third report of the National We studied 125 consecutive paCholesterol Education Program’s tients with nonalcoholic fatty liver Adult Treatment Panel III (ATP III).28 The diagnosis of disease (109 males, 16 females, mean age 49 ⫾ 13 years) metabolic syndrome was based on the ATP III criteria.28 who presented to our liver unit between June 2002 and December 2004. The diagnosis of nonalcoholic fatty liver disease was based on ultrasonography and confirmed by biopsy in 54 patients. Thirty-four (27%) patients were referred because of abnormalities in liver tests, 16 (13%) for dyslipidemia, 45 (36%) for hyperferritinemia, and 30 (24%) for liver steatosis detected by ultrasonography performed for abdominal pain or liver-unrelated causes. In all patients, daily alcohol intake was lower than 20 g (confirmed by at least one family member). Patients with chronic viral hepatitis, autoimmune hepatitis, hereditary hemochromatosis, Wilson’s disease, and drug-induced liver disease were excluded.

Controls Two hundred fifty subjects from the same geographic area of patients, included in a health surveillance program previously submitted to screening for cardiovascular risk in the Progression Lesions Intima Carotid study,27 matched 1:2 (the best possible match) with patients by paired-sample techniques for sex, age (⫾ 5 years), and body mass index (BMI) (⫾ 1.5) were enrolled. All controls had alanine-aminotransferase within normal range (5-39 UI/L), were negative for hepatitis C virus, hepatitis B virus, and human immunodeficiency virus infections or history of liver disease. No patient or control was previously diagnosed with diabetes, hypertension, or cardiovascular disease. All patients gave informed, written consent to participate in the

Insulin Resistance Assessment Insulin resistance was evaluated according to the homeostatic metabolic assessment (HOMA) insulin resistance index (fasting serum insulin ␮IU/mL ⫻ fasting serum glucose mMol/L/ 22.5) and the insulin level at 120 minutes following oral glucose tolerance test (OGTT) performed with 75 g of glucose, according to World Health Organization criteria.29-31 Liver biopsies, performed for persistently abnormal liver biochemical tests or increased ferritin in 54 patients, were processed routinely. The diagnosis of nonalcoholic steatohepatitis was based on Brunt criteria.32 The severity of steatosis was graded 1 to 3 according to the percentage of cells with fatty droplets (1: 10%-33%, 2: 33%-66%, and 3: ⬎66%).

Assessment of Carotid Atherosclerosis Early vascular atherosclerosis evaluated by the measurement of maximum intima-media thickness was determined by high-resolution B-mode ultrasound with a 7.5-MHz transducer, as described previously.16-18,33 Results of intima-media thickness represent the mean intima-media thickness on the right and left sides. For the common carotid artery, bilateral images were obtained 2 cm proximal to the dilatation of the carotid bulb; for each subject, 3 measurements on both sides were performed. Two experienced sonographers unaware of the individuals’ disease status performed intima-media thickness. The inter-rater correla-

74 tion between repeated intima-media thickness measurements was 0.96 mm (P ⬍.001), with similar averages of the 2 sets of readings. The scanning protocol included imaging of the common and internal carotid arteries of both sides in multiple longitudinal and transverse planes. In all regions of interest the presence of the plaques also was recorded. A carotid plaque was defined as a focal thickening ⱖ1.2 mm at the level of carotid artery.

Serum Oxidized Low-density Lipoprotein Serum anti oxidized low-density lipoprotein (Ox-LDL) antibodies concentration was measured by ELISA (ImmunoLisa, IMMCO Diagnostic, Buffalo, NY) and expressed as enzyme units (EU/mL).

Statistical Analysis Results are expressed as means ⫾ standard deviation for continuous variables and as frequencies for categorical variables. Mean values were compared by t test for unequal variances. Frequencies were compared by chi-squared test. Continuous variables were correlated by Spearman test. Multivariate analyses (using a conditional logistic regression model) were performed to assess the association between carotid and metabolic variables. Values of P ⱕ.05 were considered statistically significant. All the statistical analyses were performed with SAS 8.2 software system (SAS Institute Inc., Cary, NC).

RESULTS The baseline demographic, clinical, and biochemical characteristics of patients with nonalcoholic fatty liver disease and controls are shown in Table 1. HDL cholesterol (40.2 ⫾ 11 vs 50.2 ⫾ 12 mg/100 mL, P ⱕ.0001), triglycerides (158 ⫾ 96 vs 122 ⫾ 27 mg/100 mL, P ⱕ.0001), fasting glucose (96.6 ⫾ 19 vs 92.5 ⫾ 13 mg/100 mL, P ⱕ.01), HOMA-IR (3.8 ⫾ 2.7 vs 2.9 ⫾ 1.7), alanine-aminotransferase (46.6 ⫾ 35 vs 11.6 ⫾ 5.2 UI/L), gamma-glutamyltransferase (58.8 ⫾ 78 vs 25.2 ⫾ 16 UI/L, and cardiovascular risk (FRS, 10 years %) (14.5 ⫾ 10 vs 9.5 ⫾ 7.6 years, P ⱕ.0001) were significantly different in the 2 groups. Metabolic syndrome was present in 51 (41%) patients and in 26 (10%) controls (P ⱕ.0001). In patients with nonalcoholic fatty liver disease, abnormal glucose metabolism was present in 49 (38%), fasting glucose between 110 and 126 mg/100 mL in 21 (17%), impaired glucose tolerance after OGTT in 26 (21%), 6 of whom also had increased fasting glucose, and diabetes in 10 (8%), 3 diagnosed after OGTT. Because we had no information on steatosis of our control group, we divided subjects according to alanine-aminotransferase quartiles and analyzed the characteristics of controls in the different alanine-aminotransferase quartiles. Subjects in the higher quartile (alanine-aminotransferase 14-39 UI/L, median alanine-aminotransferase 21 ⫾ 4.1, 56 cases) were significantly younger (49 ⫾ 11 years and 53 ⫾ 11 years, P ⫽ .04), had higher total cholesterol (238 ⫾ 43 and 217 ⫾ 38 UI/L, P ⱕ.0001), LDL cholesterol

The American Journal of Medicine, Vol 121, No 1, January 2008 Table 1 Clinical and Laboratory Data of Patients with Nonalcoholic Fatty Liver Disease and Controls Characteristics Sex (M/F) Age (years) BMI (Kg/m2) Waist circumference (cm) Cholesterol total (mg/ 100 mL) Cholesterol HDL (mg/100 mL) Cholesterol LDL (mg/100 mL) Triglycerides (mg/100 mL) Fasting glucose (mg/100 mL) HOMA-IR Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Alanine-aminotransferase (UI/L) Gammaglutamyltransferase (UI/L) (FRS, 10-year %) Alcohol intake (g/day) 0-10 10-25* Current smoker Diabetes Hypertension Metabolic syndrome

Patients (n ⫽ 125)

Controls (n ⫽ 250)

P Value

109/16 50.5 ⫾ 13 26.6 ⫾ 4.5 93 ⫾ 19 217 ⫾ 42

218/32 52 ⫾ 11 26.7 ⫾ 3.0 94.4 ⫾ 8.3 222 ⫾ 40

.2 .5 .4 .3

40.2 ⫾ 10.8

50 ⫾ 13

146 ⫾ 37

148 ⫾ 37

.7

158 ⫾ 96

122 ⫾ 78

.0001

97 ⫾ 18

93⫾13

3.8 ⫾ 2.7 132 ⫾ 16

2.9⫾1.7 133 ⫾ 15

83 ⫾ 12

84 ⫾ 9

46.6 ⫾ 34.8

11.6 ⫾ 5.2

.0001

58.8 ⫾ 78

25.2 ⫾ 16

.0001

14.5 ⫾ 10

9.5 ⫾ 7.6

.0001

99 (40) 151 (60) 90 (35) 24 (9) 123 (47) 26 (10)

.0001 .0001 .5 .7 .3 .0001

88 (70) 37 (30) 50 (41) 10 (7) 54 (43) 51 (41)

-

.0001

.01 .0001 .6 .4

Abbreviations: ( ) ⫽ % values; M ⫽ males; F ⫽ females; BMI ⫽ body mass index; HDL⫽ high-density lipoprotein; LDL ⫽ low-density lipoprotein; FRS ⫽ Framingham risk score; HOMA-IR ⫽ homeostasis model assessment insulin resistance index. Mean ⫾ SD. Alanine-aminotransferase normal range 5-39 UI/L; gammaglutamyl transferase normal range 7-32 UI/L. Metabolic syndrome defined by positivity for 3 or more of the following criteria: 1) fasting glucose ⬎110 mg/dL; 2) central obesity (waist circumference ⬎102 cm [men] and ⬎88 cm [women]); 3) arterial pressure ⬎130/85 mm Hg; 4) triglyceride levels ⬎150 mg/dL or current use of fibrates; 5) HDL cholesterol ⬍40 mg/dL (men) and ⬍50 mg/dL (women). *For patients with nonalcoholic fatty liver disease, maximum alcohol intake was 20 g/day and 25 g/day for controls.

(162 ⫾ 38 and 143 ⫾ 35 mg/100 mL, P ⱕ.001), lower HDL cholesterol (47 ⫾ 12 and 51 ⫾ 13 mg/100 mL, P ⱕ.02), higher triglycerides (142 ⫾ 68 and 112 ⫾ 73 mg/100 mL, P ⱕ.004), HOMA-IR (3.3 ⫾ 2.3 and 2.8 ⫾ 1.5, P ⱕ.03), and prevalence of metabolic syndrome (17% and 8%, P ⱕ.03).

Vascular Damage The mean value of carotid intima-media thickness was 0.89 ⫾ 0.26 mm in patients and 0.64 ⫾ 0.14 mm in controls

Fracanzani et al

Carotid IMT in Nonalcoholic Fatty Liver Disease

75

Table 2 Variables Significantly Associated with Carotid Intima-media Thickness (IMT) ⬎ OR ⱕ0.64 mm in the Overall Series of Subjects and in Nonalcoholic Fatty Liver Disease and Controls Evaluated Separately

Variables Age (years) BMI Cholesterol total (mg/100 mL) Cholesterol LDL (mg/100 mL) Fasting glucose (mg/100 mL) Alanine-amino-transferase (UI/l) Gamma-glutamyl-transferase (UI/L) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) HOMA-IR Impaired glucose tolerance Diabetes FRS,10-year % Metabolic syndrome

Overall series

Patients

Controls

IMT ⱕ0.64 mm (n ⫽ 164) Mean ⫾ SD

IMT ⬎0.64 mm (n ⫽ 211) Mean ⫾ SD P Value

IMT ⱕ0.64 mm (n ⫽ 24) Mean ⫾ SD

IMT ⬎0.64 mm (n ⫽ 101) Mean ⫾ SD P Value

IMT ⱕ0.64 mm (n ⫽ 140) Mean ⫾ SD

IMT ⬎0.64 mm (n ⫽ 110) Mean ⫾ SD P Value

45 ⫾ 10 26 ⫾ 3 215 ⫾ 43

56 ⫾ 10 26 ⫾ 4 225 ⫾ 39

.0001 .09 .02

34 ⫾ 7 25.4 ⫾ 3 183 ⫾ 32

51 ⫾ 11 25.2 ⫾ 5 213 ⫾ 36

.0001 .5 .03

45 ⫾ 9 25.8 ⫾ 2 210 ⫾ 38

56.6 ⫾ 8 27 ⫾ 3 221 ⫾ 33

.0001 .0003 .03

142 ⫾ 39

151 ⫾ 35

.01

115 ⫾ 29

142 ⫾ 32

.03

137 ⫾ 35

147 ⫾ 29

.04

90 ⫾ 12

97 ⫾ 17

.0001

83 ⫾ 7

95 ⫾ 17

.0001

89 ⫾ 10

92 ⫾ 12

.02

18.3 ⫾ 20

27 ⫾ 29

.001

41 ⫾ 55

37 ⫾ 44

.2

11 ⫾ 12

10 ⫾ 11

.5

29 ⫾ 23

42 ⫾ 61

.01

29 ⫾ 47

45 ⫾ 61

.4

23 ⫾ 25

22 ⫾ 25

.8

129 ⫾ 13

136 ⫾ 16

127 ⫾ 12

135 ⫾ 14

83 ⫾ 8

85 ⫾ 10

2.8 ⫾ 1.8 10 (8) 8⫾7 11 (7)

119 ⫾ 8

131 ⫾ 15

.02

86 ⫾ 9

88 ⫾ 14

.4

84 ⫾ 10

87 ⫾ 9

.03

3.6 ⫾ 2.4

.0004

2.3 ⫾ 2.6

2.9 ⫾ 3.2

.01

.2 .0001 .0001

3.4 ⫾ 3.9 26 (21) 8 (8) 18 ⫾ 11 45 (44)

.6

22 (10) 12 ⫾ 9 66 (31)

2.7 ⫾ 3.6 0 2 (8) 8⫾7 6 (25)

7 (5) 9⫾8 5 (4)

16 (14) 13 ⫾ 9 21 (19)

.01 .001 .0001

.0001

.004

1.0 .0001 .07

.0001

Abbreviations: HDL⫽ high-density lipoprotein; LDL ⫽ low-density lipoprotein; HOMA-IR ⫽ homeostasis model assessment insulin resistance index; FRS ⫽ Framingham risk score. Mean ⫾ SD (%). Metabolic syndrome defined by positivity for 3 or more of the following criteria: 1) fasting glucose ⬎110 mg/dL; 2) central obesity (waist circumference ⬎102 cm [men] and ⬎88 cm [women]); 3) arterial pressure ⬎130/85 mm Hg; 4) triglyceride levels ⬎150 mg/dL or current use of fibrates; 5) HDL cholesterol ⬍40 mg/dL (men) and ⬍50 mg/dL (women).

(P ⱕ.0001). No difference in intima-media thickness was observed in controls in different alanine-aminotransferase quartiles. Carotid plaques were detected in 26 (21%) patients and in 15 (6%) controls (P ⱕ.0001). None had clinically relevant carotid stenosis (ie, ⱖ60%). Variables significantly associated with intima-media thickness higher than 0.64 mm (median value observed in controls) are shown in Table 2. Age, total and LDL cholesterol, fasting glucose, systolic blood pressure, and FRS were significantly associated with intima-media thickness higher than 0.64 mm in the overall series of subjects, as well as in patients with nonalcoholic fatty liver disease and in controls evaluated separately, whereas metabolic syndrome, BMI, diastolic pressure, and HOMA-IR were significantly associated with intima-media thickness higher than 0.64 mm in the overall series but, when groups were considered separately, only in controls. A significant correlation between alanine-aminotransferase, gamma-glutamyltransferase, and intima-media thickness as continuous variables (Spearman coefficient for alanine-aminotransferase 0.33 [P ⫽ .0001] and gammaglutamyltransferase 0.19 [P ⫽ .0002]) was found in the

overall series, but not when patients with nonalcoholic fatty liver disease and controls were analyzed separately. Multivariate logistic regression analysis was performed to evaluate independent factors associated with intima-media thickness ⬎0.64 mm in the overall series of subjects. The model included intima-media thickness ⬎0.64 mm as dependent variable and several risk factors (age, sex, BMI, cholesterol, triglycerides, fasting glucose, HOMA-IR, FRS, smoke, diabetes, blood pressure, metabolic syndrome, steatosis) as fully adjusted multiple logistic regression model variables (Table 3). The strongest independent risk predictors of intima-media thickness ⬎0.64 mm, considering the overall series (375 subjects), was the presence of steatosis. This also, was confirmed excluding control subjects in the upper alanine-aminotransferase quartile (14-39 UI/L, median 21 ⫾ 4.1 [56 cases]) who could possibly have steatosis. Analyzing separately patients and controls, the strongest variables associated with intima-media thickness higher than ⬎0.64 mm in both groups were age and blood pressure. The Figure depicts the age distribution of subjects with nonalcoholic fatty liver disease and controls with intima-

76

The American Journal of Medicine, Vol 121, No 1, January 2008

Table 3 Independent Variables Significantly Associated with Carotid Intima-media Thickness ⬎0.64 mm in the Overall Series of Subjects (Multivariate Analysis) Variables

Overall Subjects

Steatosis Age Systolic blood pressure FRS, 10-year % Metabolic syndrome HOMA-IR

OR* 6.9 6.0 2.3 1.6 2.1 1.5

95% CI 3.5-13.6 3.2-11.5 1.2-4.4 0.95-2.8 0.9-5.5 0.9-2.8

Variables significantly associated with intima-media thickness at multivariate analysis excluding normal subjects in the upper alanineaminotransferase quartile (56 cases): steatosis (OR .6.8, 95% CI, 3.1-14, P ⫽ .0001), age (OR 7.1, 95% CI, 3.4-14, P ⫽ .0001), systolic blood pressure (OR 2.7, 95% CI, 1.2-5.7, P ⫽ .008). *Adjusted odds ratios (OR) are shown; adjusted for sex, smoking habits, total low-density lipoprotein and high-density lipoprotein cholesterol, triglycerides, fasting glucose, metabolic syndrome, diabetes, body mass index, and alanine-aminotransferase.

media thickness ⬎0.64 mm. The peak of the curve in patients is reached 10 years earlier than in controls. Liver biopsy performed in 54 patients who did not differ from nonbiopsied patients for cardiovascular risk factors showed the presence of macrovesicular steatosis in all of them (grade 1 in 32 [60%], grade 2 in 10 [19%], grade 3 in 12 [22%]), with a mean percentage of cells with fatty droplets of 23%, 42%, and 68%, respectively. Twenty-five (46%) patients had nonalcoholic steatohepatitis, none had cirrhosis, there was advanced fibrosis (stage ⬎3) in 5 (9%), stage 2 in 4 (8%), stage 1 in 17 (31%), and absence of fibrosis in 28 (52%). No significant difference in intimamedia thickness values was observed between patients with pure fatty liver and nonalcoholic steatohepatitis, as well as biopsied and nonbiopsied. Ox-LDL values were measured in available sera of 32 patients. Those with higher intima-media thickness (1.05 ⫾ 0.26 mm) had higher ox-LDL values compared with those with lower intima-media thickness (0.7 ⫾ 0.1 mm) (17 cases, 65.26 ⫾ 45 EU/mL, and 15 cases 29.5 ⫾ 17 EU/mL, P ⫽ .01).

A first problem we faced was that in different studies,7,8,18,34 values of carotid intima-media thickness in normal populations markedly differed, with increasing evidence that even low intima-media thickness values were associated with carotid atherosclerosis and increased risk of cardiovascular events. According to this we arbitrarily used the median value of intima-media thickness of our large series of asymptomatic subjects included in a health surveillance program for cardiovascular events as cutoff value. Steatosis was the strongest variable independently associated with intima-media thickness. Our findings confirm previous results in patients with nonalcoholic fatty liver disease in cross-sectional studies,22,24,25 and in a case-control study in which patients with nonalcoholic fatty liver disease were enrolled only if they had increased alanine-aminotransferase.26 Given that nonalcoholic fatty liver disease is the hepatic manifestation of the metabolic syndrome, our results are in line with the increased risk of atherosclerosis in subjects with metabolic syndrome7,8,35 and with the relation between early atherosclerosis and insulin resistance, present in almost all patients with nonalcoholic fatty liver disease.12-14,36 At logistic regression analysis performed in the overall series of subjects, the strongest independent risk predictors of vascular damage was steatosis followed by age and systolic blood pressure. We found a significant association between intima-media thickness higher than 0.64 mm and HOMA-IR in the overall series and in controls, but only a trend in patients, possibly for the evenly increased HOMA-IR. Consistent with the role of glucose metabolism abnormalities is the significant relation between increased carotid intima-media thickness and increased fasting glucose and impaired glucose tolerance, as previously reported.37,38 Interestingly, a moderately increased risk for cardiovascular events in patients with type 2 diabetes and associated nonalcoholic fatty liver disease compared with those without nonalcoholic fatty liver disease was reported in a large prospective study.39 The same authors found a higher intima-media thickness in patients with nonalcoholic fatty liver disease and increased alanine-aminotransferase

DISCUSSION In the present study, we evaluated intima-media thickness and looked for carotid plaques in patients with nonalcoholic fatty liver disease with no or only mild liver function test alterations, and investigated the relationship between vascular damage and metabolic alterations. In our paired-sample case-control study performed in a large series of subjects, we demonstrated that patients with nonalcoholic fatty liver disease, independently of the severity of liver damage, had higher intima-media thickness values and prevalence of carotid plaques than matched controls.

Figure Age distribution of subjects with intima-media thickness higher than 0.64 mm in patients with nonalcoholic fatty liver disease (white bars) and in controls (black bars). Difference of age distribution between patients and controls P ⫽ .004. yr ⫽ years.

Fracanzani et al

Carotid IMT in Nonalcoholic Fatty Liver Disease

than in controls and reported the highest values in nonalcoholic steatohepatitis.26 Differently from this study, we did not observe higher intima-media thickness in nonalcoholic steatohepatitis than in pure fatty liver. The different characteristics of the patients included in our study, no or mild alterations of liver tests versus chronically elevated alanineaminotransferase and prevalence of nonalcoholic steatohepatitis (46 % vs 81%), may account for these apparent discrepancies.26 A possible limitation of our study is the small number of patients with liver biopsy due to the normality of biochemical tests in the majority of our cases. Epidemiological data suggest that cardiovascular mortality is increased in patients with nonalcoholic fatty liver disease.40,41 Our results indirectly support this hypothesis, because the presence of nonalcoholic fatty liver disease was the strongest variable independently associated with intimamedia thickness higher than 0.64 mm and possibly widespread atherosclerosis. This suggests that nonalcoholic fatty liver disease could be atherogenetic by itself, as hypothesized by Brea et al25 analyzing a small series of subjects. In favor of the proatherogenetic role of nonalcoholic fatty liver disease is the finding that despite similar BMI, prevalence of diabetes, and hypertension, patients with nonalcoholic fatty liver disease had higher intima-media thickness and prevalence of plaques than controls. Interestingly, we found a significant association between intima-media thickness and metabolic syndrome only in controls (data not shown), suggesting that in patients with nonalcoholic fatty liver disease, steatosis itself has a prevalent role in inducing vascular damage. The earlier age of intima-media thickness ⬎0.64 mm in patients than in controls further supports the proatherogenetic role of nonalcoholic fatty liver disease, not only in middle age but also in young adults. Recently, on the basis of a 2-fold increased atherosclerosis in autopsies of children with fatty liver, it was hypothesized that atherosclerosis begins in childhood and progresses through adulthood, leading to lesions responsible for later cardiovascular disease.42 It was recently shown that patients with nonalcoholic fatty liver disease have endothelial dysfunction potentially responsible for cardiovascular disease in the long term.23 Enhanced oxidative stress, inflammation with release of inflammatory cytokines, and abnormal lipoprotein metabolism could account for the proatherogenic effect of nonalcoholic fatty liver disease.23,24,43,44 In keeping with this are the increased levels of ox-LDL in patients with higher intima-media thickness, as previously reported.45,46 Visceral fat, even in normal weight or slightly overweight subjects with steatosis, could play a direct role in the development of atherosclerosis through its multiple secreted factors.22,47 In the present study, a moderate alcohol intake was more prevalent in controls than in patients. Although controversial data suggest that alcohol reduces the risk of coronary heart disease, we did not find a significant correlation between alcohol intake and intima-media thickness.48,49

77

A further limitation of this study is the lack of information on the presence of steatosis in controls. We assumed that subjects in the high alanine-aminotransferase quartile, in which alanine-aminotransferase values were between 14 and 39 UI/L (median 21 ⫾ 4.1), might have steatosis.50 Intima-media thickness of this subgroup was similar to that of subjects in the other alanine-aminotransferase quartiles, although metabolic parameters differed. Results of multivariate analyses in which subjects in the upper alanineaminotransferase quartile were included or not included, were similar. In controls, we also evaluated the fatty liver index, recently proposed as a simple and accurate predictor of hepatic steatosis in the general population,51 and only 10 (4%) subjects fulfilled the criteria for the diagnosis of steatosis. Finally, the possible inclusion of controls with nonalcoholic fatty liver disease may have led to underestimation of the differences in the intima-media thickness between the 2 groups rather than the opposite. In conclusion, our study suggests that a careful evaluation not only of the liver but also of the cardiovascular system should be performed in nonalcoholic fatty liver disease to prevent later morbidity related to atherosclerosis. Clinical implication of the results of our study is that patients with nonalcoholic fatty liver disease, independently of severity of liver disease, should be considered at high risk for cardiovascular complications and screened for carotid artery disease.

References 1. Angulo P. Nonalcoholic fatty liver disease. N Engl J Med. 2002;346: 1221-1231. 2. McCullough AJ. Update on nonalcoholic fatty liver disease. J Clin Gastroenterol. 2002;34:255-262. 3. Chiturri S, Abeygunasekera S, Farrell GC, et al. Nonalcoholic steathepatitis and insulin resistance: insulin hypersecretion and specific association with the insulin resistance syndrome. Hepatology. 2002; 35:373-379. 4. Marchesini G, Bugianesi E, Forlani G, et al. Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome. Hepatology. 2003;37: 917-923. 5. Angelico F, Del Ben M, Conti R, et al. Insulin resistance, the metabolic syndrome, and nonalcoholic fatty liver disease. J Clin Endocrinol Metab. 2005;90:1578-1582. 6. Lakka HM, Laaksonen DE, Lakka TA, et al. The metabolic syndrome and total and cardiovascular disease mortality in middle age men. JAMA. 2002;288:2709-2716. 7. Bonora E, Kiechl S, Willeit J, et al. M. Carotid atherosclerosis and coronary hearth disease in the metabolic syndrome. Diabetes Care. 2003;26:1251-1257. 8. Scuteri A, Najjar SS, Muller DC, et al. Metabolic syndrome amplifies the age-associated increases in vascular thickness and stiffness. J Am Coll Cardiol. 2004;43:1387-1395. 9. Eckel RH, Grundy SM, Zimmet PZ. The metabolic syndrome. Lancet. 2005;365:1415-1428. 10. Bedogni G, Miglioli L, Masutti F, et al. Prevalence of and risk factors for non alcoholic fatty liver disease: the Dionysos nutrition and liver study. Hepatology. 2005;42:44-52. 11. Despres JP, Lamarche B, Mauriege P, et al. Hyperinsulinemia as an independent risk factor for ischemic hearth disease. N Engl J Med. 1996;334:952-957.

78 12. Bonora E, Kiechl S, Oberhollenzer F, et al. Impaired glucose tolerance, type II diabetes mellitus and carotid atherosclerosis: prospective results from the Bruneck study. Diabetologica. 2000;43:156-164. 13. Golden SH, Folsom AR, Coresh J, et al. Risk factor groupings related to insulin resistance and their synergistic effects on subclinical atherosclerosis: the atherosclerosis risk in communities study. Diabetes. 2002;51:3069-3076. 14. Bokemark L, Wikstrand J, Attvall J, et al. Insulin resistance and intima-media thickness in the carotid and femoral arteries of clinically healthy 58-year-old men. The Atherosclerosis and Insulin Resistance Study (AIR). J Intern Med. 2001;249:59-67. 15. Kerman WN, Inzucchi SE, Viscoli CM, et al. Insulin resistance and risk for stroke. Neurology. 2002;59:809-815. 16. Mukherjee D, Yadav JS. Carotid artery intima-media thickness: indicator of atherosclerosis burden and response to risk factor modification. Am Heart J. 2002;144:753-759. 17. de Groot E, Hovingh GK, Wiegman A, et al. Measurement of arterial wall thickness as surrogate marker for atherosclerosis. Circulation. 2004;109:III33-III38. 18. Stein JH, Fraiser MC, Aeschlimann SE, et al. Vascular age: integrating carotid intima-media thickness measurements with global coronary risk assessment. Clin Cardiol. 2004;27:388-392. 19. Bots ML, Hoes AW, Koudstaal PJ, et al. Common carotid intimamedia thickness and risk of stroke and myocardial infarction: the Rotterdam Study. Circulation. 1997;96:1432-1437. 20. O’Leary DH, Polak JF, Kronmal RA, et al. Carotid-artery intima and media thickness as a risk factors for myocardial infarction and stroke in older adults. Cardivascular Health Study Collaborative Research Group. N Engl J Med. 1999;340:14-22. 21. Hollander M, Hak AE, Koudstaal PJ, et al. Comparison between measurements of atherosclerosis and risk of stroke. Stroke. 2003;34: 2367-2373. 22. Targher G, Bertolini L, Padovani R, et al. Relation of nonalcoholic hepatic steatosis to early carotid atherosclerosis in health men. Diabetes Care. 2004;27:2498-2500. 23. Villanova N, Moscatiello S, Ramilli S, et al. Endothelial dysfunction and cardiovascular risk profile in nonalcoholic fatty liver disease. Hepatology. 2005;42:473-480. 24. Volzke H, Robinson DM, Kleine V, et al. Hepatic steatosis is associated with an increased risk of carotid atherosclerosis. World J Gastroenterol. 2005;11:1848-1853. 25. Brea A, Mosquera D, Martin E, et al. Nonalcoholic fatty liver disease is associated with carotid atherosclerosis: a case-control study. Arterioscler Thromb Vasc Biol. 2005;25:1045-1050. 26. Targher G, Bertolini L, Padovani R, et al. Relations between carotid artery wall thickness and liver histology in subjects with nonalcoholic fatty liver disease. Diabetes Care. 2006;29:1325-1330. 27. Norata GD, Garlaschelli K, Ongari M, et al. Effect of the Toll-like receptor 4 (TLR-4) variants on intima-media thickness and monocytederived macrophage response to LPS. J Intern Med. 2005;258(1):2127. 28. Third report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 2002;106:3143-3421. 29. Haffner SM, Miettinen H, Stern MP. The homeostasis model in the San Antonio heart study. Diabetes Care. 1997;20:1087-1092. 30. Rajala U, Laakso M, Paivansalo M, et al. Low insulin sensitivity measured by both quantitative insulin sensitivity check index and homeostasis model assessment method as a risk factor of increased intima media thickness of carotid artery. J Clin Endocrinol Metab. 2002;87:5029-5097. 31. Bonora E, Targher G, Alberiche M, et al. Homeostasis model assessment closely mirrors the glucose clamp technique in the assessment of insulin sensitivity. Diabetes Care. 2000;23:57-63.

The American Journal of Medicine, Vol 121, No 1, January 2008 32. Brunt EM, Janney CG, Di Bisceglie AM, et al. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol. 1999;94:2467-2474. 33. Folsom AR, Eckfeldt JH, Weitzman S, et al for the Atheroclerosis Risk in Communities ARIC Study Investigation. Relation of carotid artery wall thickness to diabetes mellitus fasting glucose and insulin, body size, and physical activity. Stroke. 1994;25:66-73. 34. Grundy SM, Cleeman JI, Daniels SR, et al. Diagnosis and management of the metabolic syndrome. An America Hearth Association/National Heart, Lung, and Blood Institute Scientific Statements. Circulation. 2005;112:2735-2752. 35. Sun Y, Lin CH, CJ Lu, et al. Carotid atherosclerosis, intima media thickness and risk factors—an analysis of 1781 asymptomatic subjects in Taiwan. Atherosclerosis. 2002;164:89-94. 36. Reilly MP, Wolfe ML, Rhodes T, et al. Measures of insulin resistance add incremental value to the clinical diagnosis of metabolic syndrome in association with coronary atheroschlerosis. Circulation. 2004;110: 803-809. 37. Temelkova-Kurktschiev T, Henkel E, Schaper F, et al. Prevalence and atherosclerosis risk in different type of non-diabetic hyperglycemia: is mild an hyperglycemia underestimated evil? Exp Clin Endocrinol Diabetes. 2000;108:93-99. 38. Tuomilehto J, Qiao Q, Salonen R, et al. Ultrasonographic manifestations of carotid atherosclerosis and glucose intolerance in elderly astern Finnish men. Diabetes Care. 1998;21:1349-1352. 39. Targher G, Bertolini L, Poli F, et al. Nonalcoholic fatty liver disease and risk of future cardiovascular events among type 2 diabetic patients. Diabetes. 2005;54:3541-3546. 40. Jepsen P, Vilstrup H, Mellemkjaer L, et al. Prognosis of patients with a diagnosis of fatty liver: a registry-based cohort study. Hepatogastroenterology. 2003;50:2101-2104. 41. Sanyal A, Banas C, Sargeant C, et al. Similarities and differences in outcomes of cirrhosis due to nonalcoholic steatohepatitis and hepatitis C. Hepatology. 2006;43:682-689. 42. Schwimmer JB, Deuthsch R, Kahen T, et al. Prevalence of fatty liver in children and adolescent. Pediatrics. 2006;118:1388-1393. 43. Valenti L, Dongiovanni P, Fracanzani AL, et al. Increased susceptibility to non alcoholic fatty liver disease in heterozygotes for the mutation responsible for hereditary hemochromatosis. Dig Liver Dis. 2003;35:172-178. 44. Valenti L, Fracanzani AL, Dongiovanni P, et al. Tumor necrosis factor a promoter polymorfisms and insulin resistance in nonalcoholic fatty liver disease. Gastroenterology. 2002;122:274-280. 45. Chalasani N, Deeg MA, Crabb DW. Systemic levels of lipid peroxidation and its metabolic and dietary correlates in patients with nonalcoholic steatohepatitis. Am J Gastroenterol. 2004;99:1497-1502. 46. Yamagishi S, Matsuoka H, Kitano S, et al. Elevated circulating oxidized LDL levels in Japanese subjects with metabolic syndrome. Int J Cardiol. 2007;118:270-272. 47. Hui JM, Hodge A, Farrel GC, et al. Beyond insulin resistance in nonalcoholic steatohepatitis: TNF-alpha or adiponectin? Hepatology. 2004;40:46-54. 48. Agewall S, Wright S, Doughty RN, et al. Does a glass of red wine improve endothelial function? Eur Heart J. 2000;21:74-78. 49. Wu D, Zhai Q, Shi X. Alcohol induced oxidative stress and cell responses. J Gastroenterol Hepatol. 2006;21(Suppl 3):S26-S29. 50. Prati D, Taioli E, Zanella A, et al. Updated definitions of healty ranges for serum alanine aminotransferase levels. Ann Intern Med. 2002;137: 1-9. 51. Bedogni G, Bellentani S, Miglioli L, et al. The Fatty Liver Index: a simple and accurate predictor of hepatic steatosis in the general population. BMC Gastroenterol. 2006;6:33.