Potential conflict of interest: Nothing to report.
Steatohepatitis/Metabolic Liver Disease
Left ventricular dysfunction in obese children and adolescents with nonalcoholic fatty liver disease
Article first published online: 23 DEC 2013
© 2013 by the American Association for the Study of Liver Diseases
Volume 59, Issue 2, pages 461–470, February 2014
How to Cite
Pacifico, L., Di Martino, M., De Merulis, A., Bezzi, M., Osborn, J. F., Catalano, C. and Chiesa, C. (2014), Left ventricular dysfunction in obese children and adolescents with nonalcoholic fatty liver disease. Hepatology, 59: 461–470. doi: 10.1002/hep.26610
This study was supported by a grant from Sapienza University of Rome (Progetti di Ricerca Universitaria 2011-2012).
See Editorial on Page 372
- Issue published online: 29 JAN 2014
- Article first published online: 23 DEC 2013
- Accepted manuscript online: 11 JUL 2013 04:43AM EST
- Manuscript Accepted: 24 JUN 2013
- Manuscript Received: 9 MAY 2013
Nonalcoholic fatty liver disease (NAFLD) may increase the risk for cardiac dysfunction. The present study aimed to determine whether, in children, NAFLD is associated with subclinical left ventricular (LV) structural and functional abnormalities independently of metabolic risk factors. We performed a complete echocardiographic study including tissue Doppler imaging, magnetic resonance imaging (MRI) for measurement of hepatic fat fraction (HFF) and abdominal fat mass distribution, along with lipid profile, insulin sensitivity, and high-sensitivity C-reactive protein in 108 obese children, 54 with (HFF ≥5%) and 54 without NAFLD, and 18 lean healthy subjects. The three groups were matched for age, gender, and pubertal status, and obese children with NAFLD were matched for body mass index/standard deviation score with those without NAFLD. Forty-one of the children with NAFLD underwent liver biopsy. Compared to controls and children without liver involvement, those with NAFLD had features of LV diastolic dysfunction, including higher E-to-e' ratio and lower e' tissue velocity. The Tei index (reflecting the combined systolic and diastolic LV function) was also significantly higher in NAFLD children. Among children with biopsy-proven NAFLD, 26 had definite nonalcoholic steatohepatitis (NASH) and 15 were not-NASH. Patients with definite-NASH had significantly lower e' velocity and significantly higher E-to-e' and Tei index (P < 0.001, respectively) than those without NASH. In multiple logistic regression analysis, NAFLD was the only statistically significant variable associated with increased E-to-e' ratio, whereas NAFLD and systolic blood pressure were significantly associated with increased Tei index. Conclusion: Asymptomatic obese children with NAFLD exhibit features of early LV diastolic and systolic dysfunction, and these abnormalities are more severe in those with NASH. (Hepatology 2014;59:461–470)
late mitral velocity
late diastolic tissue velocity
analysis of variance
American Society of Echocardiography
body mass index
chronic liver disease
NASH Clinical Research Network
diastolic blood pressure
early mitral velocity
early diastolic tissue velocity
epicardial adipose tissue
high-density lipoprotein cholesterol
hepatic fat fraction
homeostasis model assessment of insulin resistance
high-sensitivity C-reactive protein
isovolumetric contraction time
isovolumetric relaxation time
interventricular septal thickness
interventricular septal thickness at end diastole
interventricular septal thickness at end systole
LV dimension at end diastole
LV dimension at end systole
LV ejection fraction
LV fractional shortening
LV posterior wall thickness at end diastole
LV posterior wall thickness at end systole
magnetic resonance imaging
magnetic resonance spectroscopy
nonalcoholic fatty liver disease
NAFLD activity score
oral glucose tolerance test
relative wall thickness
systolic tissue velocity
systolic blood pressure
standard deviation score
tissue Doppler imaging
whole-body insulin sensitivity index
Nonalcoholic fatty liver disease (NAFLD) has become the most common cause of chronic liver disease (CLD) in children and adolescents. The recent rise in prevalence rates of overweight and obesity is likely to explain the NAFLD epidemic worldwide. Patients with NAFLD, both adults and children, typically meet the diagnostic criteria for metabolic syndrome (MetS; i.e., abdominal obesity, hypertension, atherogenic dyslipidemia, and dysglycemia) and therefore are at risk of developing cardiovascular disease and coronary heart disease. Several studies (including the pediatric population) have reported independent associations between NAFLD and impaired flow-mediated vasodilatation and increased carotid artery intima-media thickness, two reliable markers of subclinical atherosclerosis, after adjusting for cardiovascular risk factors and MetS.[3-5] Also, it has been shown that, in adult subjects, NAFLD is associated with cardiac dysfunction, including myocardial insulin resistance (IR), altered cardiac energy metabolism, abnormal left ventricular (LV) structure, and impaired diastolic function.[8, 9] The duration of these subclinical abnormalities may be important, because treatment to reverse the process is most likely to be effective earlier in the disease. However, very little information is available on whether the presence of NAFLD in children is associated with early evidence of cardiac structural and functional abnormalities.
The main aims of this study were to test whether obese children with NAFLD have early, subtle changes in LV structure and function independently of metabolic risk factors and whether there is a significant association between severity of liver histology and cardiac abnormalities among NAFLD patients.
Patients and Methods
Study Design and Patients
This observational study included 108 Caucasian obese children and adolescents (body mass index [BMI] above the 95th percentile for age and gender) who were recruited at the Hepatology outpatient Clinic of the Department of Pediatrics, Sapienza University of Rome (Rome, Italy). Fifty-four patients were selected because they met the criteria for diagnosis of NAFLD (i.e., hepatic fat fraction [HFF] ≥5% on magnetic resonance imaging [MRI]). Secondary causes of steatosis, including hepatic virus infections (hepatitis A-E and G, cytomegalovirus, and Epstein-Barr virus), autoimmune hepatitis, metabolic liver disease, α-1-antitrypsin deficiency, cystic fibrosis, Wilson's disease, hemochromatosis, and celiac disease, were ruled out with appropriate tests. Forty-one of the fifty-four obese children with NAFLD also underwent liver biopsy. The other 54 obese patients had no NAFLD and were selected because they were matched (one-to-one basis) with those with NAFLD on age, gender, and pubertal status and, as closely as possible, on BMI/standard deviation score (SDS). All children without NAFLD had normal levels of aminotransferases, HFF <5%, and no evidence of CLDs (see above). None of the obese children, with or without NAFLD, had previously been treated with hepatotoxic drugs, were smokers, had a history of type 1 or type 2 diabetes, renal disease, or chronic alcohol intake.
The study also included 18 healthy children with BMI appropriate for age and gender and who had normal values for biochemical analyses as well as no MRI evidence of fatty liver.
All enrolled subjects had a complete physical examination including measurements of weight, standing height, BMI, waist circumference (WC), determination of pubertal status, systolic (SBP) and diastolic blood pressure (DBP), as reported in detail previously. Degree of obesity was quantified using Cole's least mean-square method, which normalizes the skewed distribution of BMI and expresses BMI as SDS. They all underwent laboratory tests, detailed phenotyping of abdominal and liver fat by MRI, and echocardiographic measurement of epicardial adipose tissue (EAT) thickness, as well as of cardiac dimensions and function.
The research protocol was approved by the hospital ethics committee, and informed consent was obtained from subjects' parents before assessment.
Blood samples were taken from each subject, after an overnight fast, for estimation of glucose, insulin, C peptide, total cholesterol, high-density lipoprotein cholesterol (HDL-C), triglycerides (TGs), alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT), high-sensitivity C-reactive protein (HSCRP), and apolipoprotein (APO) A-1 and B. An oral glucose tolerance test (OGTT) was performed for all obese children with the administration of glucose at 1.75 g/kg body weight (maximum dose: 75 g); blood samples were obtained at 0 minutes and every 30 minutes thereafter for 180 minutes for determination of serum glucose and insulin. Estimates of insulin sensitivity were calculated using homeostasis model assessment of IR (HOMA-IR), defined by fasting insulin and fasting glucose and whole-body insulin sensitivity index (WBISI), based on mean values of insulin and glucose obtained from OGTT and the corresponding fasting values.
Abdominal MRI studies were performed on a 1.5T Siemens Avanto magnet (Siemens AG, Erlangen, Germany). The amount of hepatic fat content (% HFF) was measured by MRI, as previously described and validated. The three-point chemical-shift–based fat-water separation method (fat-only data set) was used to measure abdominal fat mass distribution. MRI results were interpreted by an experienced radiologist who was blinded to clinical, laboratory, and/or histologic findings.
Echocardiography was performed with a commercially available echocardiographic system with tissue Doppler imaging (TDI) capabilities equipped with a 4-MHz phased-array transducer (SONOS 5500; Phillips, Andover, MA). Guided by two-dimensional (2D) echocardiography, standard M-mode recordings of the LV measurements, including LV dimension at end diastole (LVDd), LV posterior wall thickness at end diastole (LVPWd), interventricular septal thickness at end diastole (IVSd), LV dimension at end systole (LVDs), LV posterior wall thickness at end systole (LVPWs), interventricular septal thickness at end systole (IVSs), LV ejection fraction (LVEF), and LV fractional shortening (LVFS), were obtained according to the American Society of Echocardiography (ASE). Relative wall thickness (RWT) was calculated using the following formula: IVSd + LVPWd/LVDd. Because normal RWT increases with age, its raw value was adjusted for age by the following equation: (RWT − 0.005) × (age − 10). LV mass, calculated by the ASE method, was normalized for body height in meters to the allometric power of 2.7, which linearizes the relation between LV mass and height (i.e., body growth) and identifies the effect of excess body weight. LV function was evaluated using the following pulse-wave Doppler echocardiographic parameters: early (E) and late (A) mitral velocity; E-to-A ratio, and deceleration time (DT). TDI echocardiography of the septal and lateral mitral annulus was used to measure the early (e') and late (a') annular diastolic and systolic (s') tissue velocities, and the mean values of septal and lateral annulus measurements were used for analysis. E-to-e' and e'-to-a' ratios were also calculated. The Doppler-derived index of combined systolic and diastolic myocardial performance (Tei index), defined as the sum of isovolumetric relaxation time (IVRT) and isovolumetric contraction time (IVCT) divided by ejection time (ET) was used for quantification of the global LV function. Echocardiographically, EAT is generally identified as the relatively echo-free space between the outer wall of the myocardium and the visceral layer of pericardium; its thickness was measured during end systole at the point on the free wall of the right ventricle along the midline of the ultrasound (US) beam, with the best effort to be perpendicular to the aortic annulus, used as an anatomic landmark. The average value from three cardiac cycles was computed and used for statistical analysis.
All echocardiographic readings were made online by a single experienced pediatric cardiologist who was blinded to the metabolic status of the children.
The clinical indication for biopsy was either to assess the presence of nonalcoholic steatohepatitis (NASH) or other likely independent or competing liver diseases. Percutaneous needle liver biopsy was performed as previously described. The main histologic features of NAFLD were scored using the NASH Clinical Research Network (CRN) criteria. Features of steatosis, lobular inflammation, and hepatocyte ballooning were combined to obtain the NAFLD activity score (NAS). As recommended by a recent NASH CRN article, a microscopic diagnosis, based on overall injury pattern (i.e., steatosis, hepatocyte ballooning, and inflammation), as well as the presence of additional lesions (i.e., zonality of lesions, portal inflammation, and fibrosis), has been assigned to each case. Accordingly, biopsies were subdivided into not-NASH and definite-NASH subcategories.
All analyses were conducted by COBAS 6000 (Roche Diagnostics, Indianapolis, IN). Insulin and C-peptide concentrations were measured on cobas e 601 module (electrochemiluminescence technology; Roche Diagnostics), whereas the remaining analytes were measured on cobas e 501 clinical chemistry module (photometric technology), according to the instructions of the manufacturer.
Statistical analyses were performed using the SPSS v. 19 package (IBM SPSS, Chicago, IL). Data are expressed either as frequencies or means with 95% confidence intervals (CIs). Distributions of continuous variables were examined for skewness and kurtosis and were logarithmically transformed, when appropriate. Geometric means are reported for total cholesterol and HDL-C, TGs, APO A-1, APO B, HSCRP, insulin, and HOMA-IR values. Differences among groups were tested for significance using analysis of variance for quantitative variables, with Bonferroni's correction for multiple comparisons, and the chi-square test for qualitative variables. Pearson's correlation and linear regression coefficients were used to examine the relationship between variables. The independence of the associations of NAFLD with cardiac variables, considered as the dependent variable, was assessed by multiple logistic regression analysis. A P value of less than 0.05 was considered statistically significant.
Clinical, Fat Distribution, and Metabolic Characteristics
Clinical and metabolic characteristics of the study population are summarized in Tables 1 and 2, respectively. Whereas BMI-SDS were similar in obese patients with and without NAFLD, WC and abdominal fat were significantly greater in obese subjects with NAFLD than in those without NAFLD. NAFLD patients had higher values for TGs, fasting insulin, C peptide, HOMA-IR, 2-hour insulin, and HSCRP, and lower WBISI values, compared to subjects with no liver involvement and to healthy controls. As expected, children with NAFLD had higher liver enzymes, compared to those without NAFLD and to healthy children.
|Characteristics||Lean (n = 18)||Obese||P Value|
|Without NAFLD (n = 54)||NAFLD (n = 54)|
|Age, yearsa||12.5 (11.3-13.8)||12.6 (11.3-13.8)||12.6 (11.3-13.8)|
|Male gender, n (%)a||8 (53.3)||30 (55.6)||30 (55.6)|
|Prepubertal, n (%)a||2 (13.3)||7 (13)||7 (13)|
|BMI-SDS||0.47 (−0.13-1.0)||2.0 (1.95-2.15)a||2.10 (2.0-2.21)a|
|Waist circumference, cm||77 (70-84)||89 (85-92)b||94 (91-97)d, e||<0.0001|
|SBP, mmHg||100 (94-106)||106 (102-110)||114 (110-118)c, g||<0.001|
|DBP, mmHg||60 (57-65)||65 (62-68)||68 (65-71)d||<0.001|
|Abdominal fat, cm2|
|Visceral fat||199 (128-271)||433 (368-498)c||521 (460-582)d, e||<0.0001|
|Subcutaneous fat||793 (345-1,241)||1,939 (1,609-2,268)c||2,432 (2,014-2,850)d, e||<0.01|
|HFF, %||0.81 (0.55-1.10)||1.83 (1.51-2.14)d||17.8 (14.0-21.6)d, h||<0.0001|
|Variables||Lean (n = 18)||Obese||P Value|
|Without NAFLD (n = 54)||NAFLD (n = 54)|
|TGs, mg/dL||84 (50-119)||82 (70-96)||103 (86-120)c, e||<0.05|
|Total cholesterol, mg/dL||151 (129-172)||161 (153-170)||151 (142-160)||0.19|
|HDL-C, mg/dL||62 (52-73)||47 (43-51)c||47 (43-51)c||<0.05|
|APO A-1, g/L||1.64 (1.44-1.85)||1.35 (1.31-1.40)b||1.30 (1.22-1.38)b||<0.01|
|APO B, g/L||0.85 (0.66-1.04)||0.79 (0.74-0.84)||0.73 (0.65-0.81)||0.55|
|AST, U/L||19 (16-21)||24 (22-26)||32 (25-40)c, e||<0.001|
|ALT, U/L||12 (11-13)||21 (17-24)a||45 (32-60)c, g||<0.0001|
|GGT, U/L||9 (8-10)||16 (14-19)b||22 (19-26)c, g||<0.0001|
|Fasting glucose, mg/dL||84 (79-88)||84 (83-86)||82 (80-84)||0.20|
|2-hour glucose, mg/dL||88 (77-98)||91 (87-97)||91 (87-96)||0.27|
|Fasting insulin, µU/mL||9 (7-11)||14 (12-16)a||23 (19-27)c, e||<0.001|
|2-h insulin, µU/mL||35 (24-46)||38 (31-45)||80 (60-103)c, f||<0.001|
|Fasting C peptide, pmol/L||596 (449-744)||755 (686-823)||1,100 (988-1,217)d, h||<0.0001|
|HOMA-IR values||1.69 (1.05-2.33)||2.76 (2.24-3.28)a||4.60 (3.76-5.43)d, g||<0.001|
|WBISI||8.13 (4.55-11.7)||6.67 (5.50-7.72)||4.22 (3.30-5.15)c, g||<0.001|
|HbA1c, %||5.2 (5.0-5.4)||5.2 (5.1-5.3)||5.3 (5.2-5.5)||0.24|
|HSCRP, µg/L||672 (300-1,044)||3,414 (1,678-5,110)b||4,854 (2,560-7,147)d, e||<0.0001|
LV Dimensions and Function
Compared to lean subjects and to obese children without liver involvement, patients with NAFLD had higher values for SBP and greater EAT thickness, but only very mild changes in cardiac geometry, manifested by higher interventricular septal thickness (IVS; Table 3). LV mass index was slightly increased in obese children, regardless of the presence of NAFLD, compared to healthy controls. Age-adjusted RWT was similar in the three groups. In contrast, significant differences in parameters of diastolic function were obtained in children with NAFLD. NAFLD patients had increased IVRT, higher E-to-e' ratio, and lower e' tissue velocity, compared to lean controls as well as to obese subjects without NAFLD. The e'-to-a' ratio showed a tendency to be lower in obese children with NAFLD (P = 0.06). Peak velocities of early (E) and late diastolic filling (A), IVCT, and a' tissue velocity were similar in the three groups. Among the systolic function parameters, only s' tissue velocity was significantly lower in NAFLD children, whereas LVEF and LVFS were similar in the three groups. The Tei index (which reflects the combined systolic and diastolic LV function) was also significantly higher in children with NAFLD, compared to lean controls as well as to obese subjects without NAFLD.
|Characteristics||Lean (n = 18)||Obese||P Value|
|Without NAFLD (n = 54)||NAFLD (n = 54)|
|LVDd, mm||43.2 (42.2-44.3)||45.1 (42.3-48.0)||45.7 (44.3-47.1)||0.16|
|LVPWd, mm||8.4 (7.3-10.0)||8.3 (7.5-9.1)||8.6 (7.8-9.4)||0.50|
|IVSd, mm||6.6 (5.6-7.7)||7.0 (6.6-7.4)a||7.7 (7.1-8.4)a, e||<0.05|
|LVDs, mm||26.3 (24.9-27.7)||26.6 (25.9-27.4)||28.3 (27.1-29.5)||0.56|
|LVPWs, mm||13.5 (11.6-15.8)||13.3 (12.5-14.1)||13.9 (13.0-14.7)||0.60|
|IVSs, mm||11.0 (10.0-13.3)||11.3 (10.1-12.5)||12.8 (12.0-13.6)a, e||<0.05|
|Age-adjusted RWT, mm||0.34 (0.32-0.36)||0.36 (0.33-0.38)||0.36 (0.33-0.38)||0.45|
|LV mass index, g/height2.7||29.1 (26.6-32.6)||35.4 (32.2-37.8)a||35.7 (33.0-38.3)a||<0.05|
|LVEF, %||68.5 (64.4-72.4)||68.0 (66.6-69.4)||67.2 (65.1-69.4)||0.91|
|LVFS, %||38.5 (36.5-41.6)||37.6 (36.2-38.9)||39.5 (37.1-41.9)||0.79|
|E, cm/s||94.4 (88.7-100.1)||101.6 (97.9-104.2)||101.0 (96.8-104.2)||0.28|
|A, cm/s||50.2 (44.1-56.4)||51.2 (48.6-53.7)||51.9 (49.0-54.8)||0.81|
|E-to-A ratio||1.94 (1.72-2.15)||2.10 (1.92-2.20)||1.99 (1.90-2.10)||0.72|
|IVRT, ms||51.9 (48.4-55.7)||64.8 (52.9-96.7)a||75.9 (57.5-101.8)b, f||<0.01|
|IVCT, ms||49.0 (43.4-54.6)||65.2 (41.6-88.7)||69.2 (42.6-99.8)||0.54|
|ET, ms||272 (240-290)||280 (270-293)||282 (272-294)||0.06|
|DT, ms||0.149 (0.125-0.161)||0.151 (0.142-0.160)||0.148 (0.129-0.139)||0.58|
|Tei index||0.22 (0.21-0.25)||0.23 (0.21-0.24)||0.25 (0.24-0.26)c, g||<0.001|
|s' velocity, cm/s||8.70 (7.82-9.60)||8.64 (8.02-9.28)||7.81 (7.42-8.20)b, f||<0.01|
|e' velocity, cm/s||15.8 (14.8-16.8)||14.9 (10.6-18.0)||14.6 (13.7-15.4)b, f||<0.01|
|a' velocity, cm/s||5.52 (4.04-7.09)||5.69 (5.11-6.27)||6.08 (5.59-6.57)||0.51|
|e'-to-a' ratio||2.96 (2.60-3.29)||2.86 (1.95-3.57)||2.61 (2.29-2.93)||0.062|
|E-to-e' ratio||5.96 (4.89-7.02)||6.33 (5.86-6.79)||7.26 (6.78-7.75)b, f||< 0.01|
|EAT, mm||8.0 (6.6-9.4)||10.4 (9.8-10.9)c||11.6 (11.0-12.3)d, g||<0.0001|
Findings in Children With Biopsy-Proven NAFLD
Histopathological features associated with NAFLD are reported in Table 4. Twenty-six children (63.4%) had definite-NASH, whereas 15 (36.5%) were not-NASH. Clinical, metabolic, and echocardiographic characteristics of this cohort are summarized in Supporting Table 1. Compared to children without NASH, patients with NASH had significantly lower s' velocity (mean, 7.46 [95% CI: 7.01-7.90] versus 8.73 [95% CI: 7.98-9.47] cm/s; P < 0.001) and e' velocity (mean, 13.4 [95% CI: 12.4-14.3] versus 15.8 [95% CI: 14.6-17.7] cm/s; P < 0.001) as well as significantly higher E-to-e' ratio (mean, 8.10 [95% CI: 7.49-8.69] versus 6.01 [95% CI: 5.41-6.62]; P < 0.001] and Tei index (mean, 0.26 [95% CI: 0.25-0.27] versus 0.23 [95% CI: 0.22-0.25]; P < 0.001; (Figs. 1-4). All other cardiac parameters were similar.
|Histopathological Features||Degree or Stage||NO NASH (n = 15)||NASH (n = 26)|
|Steatosis, n (%)||1||8 (53.3)||4 (15.4)|
|2||5 (33.3)||8 (30.8)|
|3||2 (13.3)||14 (53.8)|
|Lobular inflammation, n (%)||0||1 (6.7)||0|
|1||10 (66.6)||8 (30.8)|
|2||4 (26.6)||13 (50.0)|
|Ballooning, n (%)||0||4 (26.6)||0|
|1||10 (66.6)||19 (73.1)|
|2||1 (6.7)||7 (26.9)|
|Fibrosis, n (%)||0||3 (20.0)||8 (30.8)|
|1||8 (53.3)||8 (30.8)|
|2||4 (26.6)||9 (34.6)|
|NAS, mean (SD)||2.60 (0.41)||5.46 (0.65)|
Linear Regression Analysis
In the population of obese children, we investigated the association of E-to-e' ratio as well as of Tei index with clinical and metabolic characteristics after adjustment for age, gender, and pubertal status. HFF (standardized coefficient B: 0.264; P < 0.01) and WBISI (−0.235; P < 0.01) were the only parameters significantly associated with E-to-e' ratio. BMI-SDS (standardized coefficient B: 0.260; P < 0.01), abdominal visceral (0.330; P < 0.001) and subcutaneous adipose tissue (0.370; P < 0.01), and HFF (0.302; P < 0.01), as well as SBP (0.346; P < 0.01), were significantly associated with Tei index. Among the metabolic variables, TGs (standardized coefficient B: 0.210; P < 0.05), insulin (0.310; P < 0.001), C peptide (0.250; P < 0.05), HOMA-IR values (0.340; P < 0.01), and WBISI (−0.465; P < 0.0001) were also significantly associated with Tei index.
Multiple Logistic Regression Analysis
A multiple stepwise logistic analysis was used to assess the independent association of NAFLD and of NASH with increased E-to-e' ratio as well as increased Tei index in the whole obese population and in biopsy-proven NAFLD, respectively. For this purpose, subjects were categorized into two groups according to the median value of E-to-e' ratio (i.e., 6.83) or Tei index (i.e., 0.24). After adjustment for age, gender, pubertal status, and clinical and metabolic variables (including BMI-SDS, abdominal fat, SBP and DBP, TGs, HDL-C, WBISI, and NAFLD), NAFLD (odds ratio [OR]: 3.13 [95% CI: 1.12-8.72]; P < 0.05) was the only statistically significant variable associated with increased E-to-e' ratio, whereas NAFLD (5.60 [95% CI: 1.66-18.9]; P < 0.01) and SBP (1.10 [95% CI: 1.03-1.14; P < 0.01) were significantly associated with increased Tei index. NASH, within NAFLD cases, was significantly associated with increased E-to-e' ratio (OR, 5.70 [95% CI: 1.43-23.1]; P < 0.05) and increased Tei index (7.50 [95% CI: 1.17-28.0]; P < 0.01) after adjustment for anthropometric, clinical, and metabolic variables.
This study provides evidence that (1) asymptomatic obese children with NAFLD exhibit features of early LV diastolic and systolic dysfunction, as measured by 2D echocardiography using TDI, which is available in most of the current echocardiographic systems, being a simple and reliable imaging method for quantitative assessment of cardiac function,[22, 23] and (2) children with more severe liver histology have worse cardiac dysfunction than those with more mild liver changes. We are unaware of any previous studies that have evaluated signs of subclinical cardiac dysfunction in pediatric patients with biopsy-proven NAFLD.
Our obese children with NAFLD were characterized by significantly lower e' tissue velocity and higher E-to-e' ratio, compared with age-, gender-, pubertal status-, and BMI-SDS-matched obese children without NAFLD. Moreover, our NAFLD patients had a significant reduction in s' tissue velocity and a significant increase in Tei index, compared to those without NAFLD, implying a potential effect of NAFLD also on systolic function. A major finding of the current work was that the above-described echocardiographic features of early LV diastolic and systolic dysfunction were significantly associated with NAFLD independently of several metabolic variables. Theoretically, the increased risk of LV dysfunction in patients with NAFLD might reflect the clustering of underlying metabolic risk factors characteristic of this population. Alternatively, as shown in the present study, NAFLD per se, in particular, NASH, might confer a risk of cardiac dysfunction above and beyond that associated with the individual components of MetS.
An increasing number of studies in adults have recently evaluated the effects of NAFLD on cardiac dysfunction.[7-9, 25-27] Goland et al. have shown a markedly impaired diastolic function and mild alterations in LV structure in 38 adult patients with (US diagnosed, n = 27; biopsy-proven: n = 11) NAFLD in the absence of diabetes, hypertension, and morbid obesity. On multivariate analysis, e' on TDI was the only independent index able to characterize patients with NAFLD. Similar findings have been later reported by Fotbolcu et al. in 35 nondiabetic, normotensive adult patients with US-diagnosed NAFLD. However, independent predictors of LV impairment were not determined. Very recently, in a study examining cardiac status by high-resolution MRI in a clinical group of 19 adult patients with NAFLD (defined as >5% intrahepatic lipid on 1H-MRS [magnetic resonance spectroscopy]), Hallsworth et al. demonstrated significant changes in cardiac structure and evidence of early diastolic dysfunction, compared to the 19 age-, gender-, and BMI-matched controls, in the absence of cardiac metabolic changes or overt cardiac disease. There was no correlation between blood pressure and cardiac parameters. Conversely, Perseghin et al. showed that 21 men with higher intrahepatic fat content, as measured by 1H-MRS, had excessive fat accumulation in the epicardial area and abnormal LV energy metabolism, compared to the 21 men matched for anthropometric features with lower intrahepatic fat content. These alterations were detected despite normal LV morphology and function by cardiac MRI. Similarly, in a study involving 61 diabetic male subjects, Rijzewijk et al. found that, compared with the 29 men with lower intrahepatic fat content on 1H-MRS, the 32 patients with higher intrahepatic fat content had decreased myocardial perfusion, glucose uptake, and high-energy phosphate metabolism, but similar values of LV function and morphology by cardiac MRI.
Currently, only two published articles address the relationship between NAFLD and cardiac function in the pediatric population.[28, 29] In a study including 93 obese children with US-diagnosed NAFLD, 307 obese subjects without liver involvement, and 150 age- and gender-matched healthy controls, Alp et al. showed that subclinical systolic and diastolic impairment could be detected by TDI in obese children with NAFLD. Also, cardiac dysfunction was correlated with the increase in grades of liver steatosis. The major limitation of that study was that hepatic ultrasonography, an operator-dependent procedure, was used to diagnose and grade the severity of NAFLD. This technique cannot identify fatty infiltration below 33% and is subject to interpretation or interobserver variation when assessing categories of steatosis severity. Recently, Singh et al. measured, by 2D speckle tracking echocardiography, myocardial function in three groups of age-, gender-, and Tanner-matched adolescents (lean [n = 14]; obese with normal [n = 15] or increased [n = 15] intrahepatic triglyceride [IHTG] content [≥5.6%]). The researchers showed that obese adolescents with increased IHTG had greater impairment of systolic and diastolic function, manifested by decreased systolic and diastolic myocardial strain and strain rate than BMI-SDS-matched obese adolescents with normal IHTG content. The cardiac functional abnormalities were independently associated only with HOMA-IR, after adjustment for BMI, conventional cardiovascular risk factors, and intra-abdominal, -cardiac, and -hepatic fat content. However, given the small number of adolescents included in their study, the possibility of a type 2 error raised by the researchers is well taken. Our results extend the observations of these previous reports in which diagnosis of NAFLD in obese children was exclusively based on US or MRI imaging, but was not confirmed by liver biopsy. In our study, NASH children had more severe abnormalities in LV function, compared with those without NASH.
The pathophysiological mechanisms of cardiac dysfunction in NAFLD cannot be determined from our study, but we might speculate that they involve IR, an abnormal lipid profile, and a low-grade inflammatory state.[32-35] Hepatic steatosis is associated with hepatic IR (i.e., impaired hepatic glucose production), which, in turn, leads to hyperglycemia and compensatory hyperinsulinemia and may worsen systemic as well as cardiac IR. In this study, we found that WBISI was significantly associated in an inverse relationship with diastolic and systolic dysfunction. However, this association disappeared when NAFLD was taken into account as an independent variable, suggesting a more significant role of NAFLD over WBISI. The liver plays a pivotal role in controlling the amount of lipids reaching the circulation. In patients with NAFLD, the increased free fatty acids may induce myocardial lipid accumulation, which is detrimental to LV function.[34-36] In fact, myocardial steatosis may cause alterations in myocardial substrate metabolism and efficiency (i.e., cardiac work/myocardial oxygen consumption) that occur early in the cascade of events leading to impaired LV contractility.[32, 33] Rijzewijk et al. showed that intramyocardial fat content, as detected by 1H-MRS, was significantly higher in uncomplicated type 2 diabetic men than in nondiabetic control subjects and was associated with impaired cardiac metabolism. Although we did not measure directly the myocardial fat content, we found that obese children with NAFLD had a greater EAT thickness than those without liver involvement. Epicardial fat is a metabolically active organ that generates proatherogenic, -inflammatory, and -thrombotic adipocytokines.[37, 38] Its anatomic location, without any barrier to the adjacent myocardium, enables local paracrine interaction between EAT and the myocardium. Using cardiac MRI and P-MRS, Perseghin et al. demonstrated that individuals with fatty liver had an increased amount of fat in the epicardial area and displayed abnormal cardiac metabolism. Thus, epi- and myocardial fat represent abnormal ectopic fat storage and may be a marker of the cumulative effects of NAFLD and IR in the setting of pathological adiposity,[36, 40] with consequent adverse associated cardiovascular outcome. Emerging evidence also suggests that NAFLD, especially in its necroinflammatory form, might be involved in the pathogenesis of cardiac function abnormalities through the systemic release of several mediators from the steatotic and inflamed liver (including C-reactive protein, interleukin-6, tumor necrosis factor-α, and other inflammatory cytokines).[4, 42] Therefore, we measured serum HSCRP. We found a significant difference in serum HSCRP levels between children with and those without NAFLD, but we failed to find an association between its serum concentration and cardiac dysfunction.
A few limitations are worth noting here. First, the cross-sectional design of our study precludes the establishment of causal or temporal relationships between NAFLD/NASH and cardiac abnormalities. Follow-up studies should be helpful in elucidating the cause-and-effect relations and the underlying mechanism. Second, there may be selection bias toward children with increased morbidity because this is a cohort typical of a specialized tertiary care referral center. Third, although our results have been adjusted for WBISI, a reliable method for estimating insulin sensitivity, we did not measure insulin sensitivity by euglycemic clamp in our population, so we cannot be certain that identical results could be obtained after adjustment for the clamp-measured insulin sensitivity.
In summary, this study demonstrates a significant association between NAFLD and early LV dysfunction that can be detected by TDI in obese children. Early intervention during childhood to recognize NAFLD, as well as to prevent its progression to NASH, might represent a crucial step in averting an unfavorable cardiac phenotype.
- 9Impairment of the left ventricular systolic and diastolic function in patients with non-alcoholic fatty liver disease. Cardiol J 2010;17:457-463., , , , , , et al.
- 14American Society of Echocardiography's Nomenclature and Standards Committee; Task Force on Chamber Quantification; American College of Cardiology Echocardiography Committee; American Heart Association; European Association of Echocardiography, European Society of Cardiology. Recommendations for chamber quantification. Eur J Echocardiogr 2006;7:79-108., , , , , , et al.
- 20NASH Clinical Research Network (CRN). Nonalcoholic fatty liver disease (NAFLD) activity score and the histopathologic diagnosis in NAFLD: distinct clinicopathologic meanings. HEPATOLOGY 2011;53:810-820., , , , ;
- 28Association between nonalcoholic fatty liver disease and cardiovascular risk in obese children and adolescents. Can J Cardiol 2012 Oct 2. doi: 10.1016/j.cjca.2012.07.846., , , , , .
Additional Supporting Information may be found in the online version of this article.
|hep26610-sup-0001-supptable1.docx||21K||Supplementary table 1. Characteristics of obese children with and without NASH|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.