Original Article

Blood pressure is associated with the presence and severity of nonalcoholic fatty liver disease across the spectrum of cardiometabolic risk Ehimen C. Aneni a, Ebenezer T. Oni a, Seth S. Martin b, Michael J. Blaha b, Arthur S. Agatston a, Theodore Feldman a, Emir Veledar a,c,d, Raquel D. Conc¸eicao e, Jose A.M. Carvalho e, Raul D. Santos e,f, and Khurram Nasir a,b,c,g

Objectives: To determine the relationship between clinically relevant blood pressure (BP) groups and nonalcoholic fatty liver disease (NAFLD) presence and severity especially in the milieu of other metabolic risk factors. Patients and methods: From a Brazilian cohort of 5362 healthy middle-aged men and women who presented for yearly physical examination and testing, the cross-sectional relationship between BP categories and NAFLD was assessed. BP groups were categorized as normal, prehypertension (PHT), and hypertension (HTN) according to the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure classification. NAFLD was ultrasound diagnosed, excluding persons with alcohol consumption more than 20 g/day. NAFLD severity was estimated using the Fibrosis-4 (FIB-4) risk score. Results: The prevalence of NAFLD was 36.2%. Participants with NAFLD were older (mean 46 vs. 42 years, P < 0.001) and had elevated BMI (mean 29.0 vs. 24.7 kg/m2, P < 0.001). The prevalence of NAFLD among persons with normal BP, PHT, and HTN was 16.5, 37.5, and 59.3%, respectively. In multivariate analyses, PHT and HTN were associated with elevated odds of NAFLD (PHT-adjusted odds ratio 1.3, 95% confidence interval 1.1, 1.6; HTNadjusted odds ratio 1.8, 95% confidence interval 1.4–2.3) compared with normal BP. Among nonobese hypertensive patients, BP control (BP < 140/90 mmHg) was independently associated with 40% lower odds of prevalent NAFLD. Compared with hypertensive patients, both normotensive individuals and prehypertensive patients were more likely to have a low fibrosis risk (FIB-4  1.3). Conclusion: Prevalent NAFLD may be seen early in the development of hypertension, even in the absence of other metabolic risk factors. Controlling BP among nonobese hypertensive patients may be beneficial in preventing or limiting NAFLD.

identification test; BP, blood pressure; CVD, cardiovascular disease; GGT, gamma glutamyl transpeptidase; HDL-c, high-density lipoprotein cholesterol; hsCRP, high-sensitivity C-reactive protein; HTN, hypertension; IDF, International Diabetes Federation; IQR, interquartile range; JNC7, Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure; LDL-c, low-density lipoprotein cholesterol; MetS, metabolic syndrome; NAFLD, nonalcoholic fatty liver disease; PHT, prehypertension; SD, standard deviation

INTRODUCTION

N

onalcoholic fatty liver disease (NAFLD) is one of the most common liver diseases globally and is the most frequently identified cause of abnormal liver function tests in the United States [1]. Apart from the significant risk of end-stage liver disease, NAFLD is emerging as a cardiovascular disease (CVD) risk factor as demonstrated by the significant associations between NAFLD and fatal and nonfatal CVD [2]. Close relationships have been described between NAFLD and metabolic risk factors. In particular, NAFLD closely tracks with obesity, metabolic syndrome (MetS), insulin resistance, and subclinical inflammation, and is associated with the presence of subclinical CVD [3–8]. Both prehypertension (PHT) and hypertension (HTN) are relatively common in the United States with a

Journal of Hypertension 2015, 33:1207–1214 a Center for Prevention and Wellness Research, Baptist Health South Florida, Miami, Florida, bJohns Hopkins Ciccarone Center for Prevention of Cardiovascular Disease, Baltimore, Maryland, cRobert Stempel College of Public Health, Florida International University, Miami, Florida, dSchool of Medicine, Department of Medicine, Emory University, Atlanta, Georgia, USA, eHospital Israelita Albert Einstein, fHeart Institute (InCor), University of Sao Paolo Medical School Hospital, Sao Paolo, Brazil and g Herbert Wertheim College of Medicine, Florida International University, Miami, Florida, USA

Keywords: BP control, hypertension, metabolic syndrome, nonalcoholic fatty liver disease, prehypertension

Correspondence to Khurram Nasir, MD, MPH, 1691 Michigan Ave, Miami Beach, FL 33139, USA. E-mail: [email protected]

Abbreviations: ALT, alanine transaminase; AST, aspartate aminotransferase; AUDIT, alcohol use disorders

J Hypertens 33:1207–1214 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.

Journal of Hypertension

Received 27 July 2014 Revised 23 December 2014 Accepted 23 December 2014

DOI:10.1097/HJH.0000000000000532

www.jhypertension.com

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

1207

Aneni et al.

prevalence of about 30% each among adults older than 20 years [9]. Compared with those with normal blood pressure (BP), PHT increases the risk of HTN by two to six-fold over 4 years [10]. PHT is also associated with increased risk of mortality and other major cardiovascular events such as myocardial infarction, stroke, and congestive heart failure [11,12]. Recent studies suggest an independent association between NAFLD and hypertension (HTN) [13,14], and when present among hypertensive patients, NAFLD is associated with CVD complications such as left ventricular hypertrophy, ventricular dysfunction, and cardiovascular dysautonomia [15–17]. The relationship between PHT or HTN and NAFLD among nonobese patients or those without other metabolic abnormalities is not clearly defined. However, some studies have demonstrated that NAFLD can also be seen in nonobese populations and persons without the complete spectrum of MetS [18]. This is particularly relevant as recent studies suggest that there is elevated mortality risk among nonobese persons with NAFLD [19]. NAFLD is often referred to as ‘the hepatic expression of MetS’ because of its close relationship to a spectrum of metabolic abnormalities [20]. As a consequence, much of the scientific literature has focused on understanding the relationship between clinically evident MetS and NAFLD. Thus, it is not known whether elevated BP can identify persons with NAFLD who are otherwise metabolically healthy. Given that the scientific knowledge on the relationship between NAFLD and epidemiologically relevant BP is sparse, we investigated these relationships and described how other metabolic risk factors affect this association in a large population-based cohort.

METHODS This cross-sectional study was conducted among 6464 Brazilian individuals, without a history of CVD or liver disease who underwent mandatory employer-sponsored occupational health clinical examination at the Preventive Medical Center of the Hospital Israelita, Albert Einstein, Sao Paulo, Brazil, between November 2008 and July 2010. The study was approved by the institutional review board and a waiver of consent was granted. The examination protocol included obtaining demographic, lifestyle, and clinical history via questionnaire, physical examination, and laboratory testing. Alcohol use was quantified by the Alcohol Use Disorders Identification Test (AUDIT) [21] as a numeric score, and physical activity was categorized by the short form of the International Physical Activity Questionnaire [22]. All participants underwent anthropometric measurement and physical examination. BP was measured according to the American Heart Association guidelines [23], the first of which was after a 5min rest. The mean of three resting measurements was obtained and used in this analysis. Waist circumference was assessed using a measuring tape placed around the abdomen at the smallest diameter between the iliac crest and costal margin. Fasting blood samples were obtained for laboratory testing. Serum glucose, total cholesterol, triglycerides, 1208

www.jhypertension.com

creatinine, and liver enzymes [aspartate aminotransferase (AST); alanine transaminase (ALT); and gamma glutamyl transpeptidase (GGT)] were measured with enzymatic colorimetric assays on a Vitros automated platform (Johnson & Johnson Clinical Diagnostics, New Brunswick, New Jersey, USA) HDL-cholesterol (HDL-c) was obtained using a precipitation method whereas for triglycerides 400 mg/dl or less low-density lipoprotein cholesterol (LDL-c) was calculated using the Friedwald equation [24]. Platelets were counted automatically (Beckman Coulter Act5Diff; Beckman Coulter, Brea, California, USA) and high-sensitivity Creactive protein (hsCRP) was measured by immunonephelometry (Dade-Behring, Newark, Delaware, USA). Elevated hsCRP was defined as hsCRP more than 3 mg/l. The glomerular filtration rate for each participant was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation: Glomerular filtration rate ¼ 141 min (Scr/k,1)a  max(Scr/k,1)1.209  0.993Age  1.018 [if woman]  1.159 [if black] where Scr is serum creatinine (mg/dl), k is 0.7 for women and 0.9 for men, a is 0.329 for women and 0.411 for men, min indicates the minimum of Scr/k or 1, and max indicates the maximum of Scr/k or 1 [25]. The presence of liver fat was assessed with abdominal ultrasound after fasting for at least 6 h. Fatty liver was defined as the presence of increased hepatic echogenicity with the renal parenchyma distinguishable from the liver [26]. The ultrasound images were obtained by an Acuson XP-10 scanner (Acuson, Mountain View, California, USA). Each scan was read by a board certified radiologist. Ultrasound has a sensitivity and specificity of about 85 and 94%, respectively, with an area under the receiver operating characteristic curve of 0.93 for the diagnosis of moderate-to-severe fatty liver [27]. NAFLD was defined as ultrasound diagnosed fatty liver in persons with AUDIT alcohol score less than 8 (scores >8 is associated with harmful hazardous drinking) [21]. In sequence, we excluded persons with AUDIT scores OF 8 at least (N ¼ 854) or missing (N ¼ 180). From the remaining population, we excluded those with missing ultrasound assessment of fatty liver (N ¼ 67) and those with missing BP measurements or history of hypertension (N ¼ 1). After exclusions, the study population consisted of 5362 participants (77% men), with mean age of 44 years. We categorized BP into three groups – normal, PHT [Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC7) normal and high normal BP], and HTN. Normal BP was defined as SBP/DBP less than 120/80 mmHg in the untreated with no known history of hypertension. PHT was defined as untreated SBP of 120–139 or DBP of 80–89 mmHg. Persons with SBP 140 mmHg at least or DBP 90 mmHg at least or on antihypertensive medication were categorized as hypertensive. MetS was defined as the presence OF three or more of the following – truncal obesity (waist circumference 102 cm [40 in.] at least for men and 88 cm [36 in.] at least for women), high BP (BP  130/85 mmHg or the use of antihypertensive medications), hyperglycemia (fasting blood glucose, FBG, 100 mg/dl), low HDL-c (40 mg/dl for men Volume 33  Number 6  June 2015

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

BP and NAFLD across cardiometabolic risk spectrum

and 50 mg/dl for women), and hypertriglyceridemia (150 mg/dl) [28]. Obesity was defined as BMI 30 kg/m2 at least or BMI more than 25 with a waist circumference 102 cm [40 in.] at least for men and 88 cm [36 in.] at least for women. In assessing the relationship between BP and severity of NAFLD, we compared BP with the fibrosis 4 (FIB-4) index, calculated as age (years)  AST [U/l]/(platelets [109/l]  (ALT [U/l])1/2, among those persons with NAFLD [29]. The FIB-4 index is a useful noninvasive marker of hepatic fibrosis among persons with NAFLD with an area under the receiver operating characteristic curve of 0.81 for predicting liver-related events [30]. Persons with FIB-4 less than 1.30 were categorized as having a low risk for hepatic fibrosis, those with FIB-4 more than 2.67 were categorized as high risk whereas those with scores between 1.30 and 2.67 were regarded as having an indeterminate risk [31].

Logistic regression models were also used to assess the relationship between ordinal BP groups and NAFLD. Covariates in the model included age and sex, physical activity, metabolic risk factors such as waist circumference, HDL-c, triglycerides, and blood glucose, liver enzymes including AST, ALT, GGT, and platelets. To ascertain the relationship between BP control and NAFLD, logistic regression analysis in a population that excluded those not categorized as hypertensive patients was conducted. Multivariate models were also generated for the association between HTN and NAFLD among obese and nonobese persons and among those with 0, 1–2, and more than 2 MetS components. Statistical analysis was conducted in Stata Statistical Software, Release 12.0; Statacorp, College Station, Texas, USA [32]. The 95% confidence intervals (CIs) for the odds ratios (ORs) were determined. All P values were two-sided and a P value 0.05 or less was considered statistically significant.

RESULTS

Statistical analysis All variables were graphically examined for normality. Normally distributed continuous variables are expressed as mean  standard deviation whereas nonparametric continuous variables are expressed as median  interquartile range (IQR). Categorical variables are expressed as percentages. The mean of continuous variables across two or more groups was compared using analysis of variance. For comparisons of nonparametric variables, the Kruskal–Wallis test was used. Frequencies of categorical variables were compared across two or more groups using the x2 test of independence.

Population characteristics The clinical and laboratory characteristics of those with and without NAFLD are shown in Table 1. Those with NAFLD had significantly higher mean BMI (29 vs. 25 kg/m2, P < 0.001) and a higher prevalence of the MetS (41 vs. 7%). Persons with NAFLD also had significantly higher mean SBP, DBP, mean arterial BP (DBP  1/3[SBP  DBP]) DBP]) and systemic pulse pressure (SBP – DBP). Among those with hypertension, antihypertensive use alone was not associated with NAFLD; however; those who were well controlled (BP < 140/90 mmHg) had lower prevalence of

TABLE 1. Baseline characteristics of the study population between nonalcoholic fatty liver disease groups Variable Age, mean  SD (years) % Male BMI, mean  SD Waist circumference, mean  SD FBG, mean  SD Total cholesterol, mean  SD HDL-c, mean  SD LDL-c, mean  SD AST, mean  SD ALT, mean  SD GGT, median (IQR) eGFRb, mean  SD Triglycerides, median (IQR) hsCRP, mean (IQR) % Obese % with MetSa % with HTN SBP, mean  SD DBP, mean  SD Systemic pulse pressurec, mean  SD Mean arterial BPd, Mean  SD Platelet count (103), mean  SD % On antihypertensive treatment (among hypertensive patients) % BP controlled (among hypertensive patients)

N

All participants

NAFLD ()

NAFLD (þ)

P

5362 5362 5362 5346 5343 5343 5343 5342 5344 5345 5345 4908 5344 4588 5353 5356 5362 5362 5362 5362 5362 5332 1059 1059

43.5  9.3 77.0 26.2  4.0 91.1  12.1 88.9  9.9 205.1  36.5 48.2  13.1 130.6  33.0 30.7  10.9 37.0  20.5 29 (21, 43) 103.8  13.6 113 (82, 161) 1.2 (0.6, 2.4) 22.2 19.1 19.8 118.3  12.4 76.6  7.9 41.7  7.6 90.5  9.0 242.0  56 59.0 69.0

42.1  9.3 68.3 24.7  3.1 86.3  10.2 86.8  8.4 202.3  35.5 54.4  13.5 128.4  32.7 28.9  10.0 31.4  16.2 26 (19, 35) 105.1  13.7 98 (74, 134) 1.7 (0.9, 2.9) 10.0 7.4 12.3 115.4  11.7 74.8  7.6 40.6  7.2 88.3  8.5 240.9  53.3 59.4 73.8

46.0  8.9 93.2 29.0  3.9 100.1  10.0 92.9  11.9 210.3  37.8 42.0  9.7 134.7  33.3 34.1  11.8 47.4  23.4 38 (29, 52) 101.2  13.1 152 (109, 209) 1.0 (0.5, 2.0) 45.2 41.0 33.7 123.6  12.1 80.0  7.4 43.6  7.9 94.5  8.4 242.5  57.3 58.5 65.8

Blood pressure is associated with the presence and severity of nonalcoholic fatty liver disease across the spectrum of cardiometabolic risk.

To determine the relationship between clinically relevant blood pressure (BP) groups and nonalcoholic fatty liver disease (NAFLD) presence and severit...
340KB Sizes 0 Downloads 10 Views