| |
Non-alcoholic fatty liver disease is strongly associated with carotid atherosclerosis (increased 13% risk): A systematic review
|
| |
| |
Journal of Hepatology July 2008, advance publication
Uncorrected Proof
Silvia Sookoian13, Carlos J. Pirola2
1Laboratory of Clinical and Molecular Hepatology, Department of Molecular Genetics and Biology of Complex Diseases, Institute of Medical Research, A. Lanari, University of Buenos Aires-CONICET, Ciudad Autonoma de Buenos Aires, Argentina
2Laboratory of Molecular Genetics and Biology of Metabolic Syndrome, Department of Molecular Genetics and Biology of Complex Diseases, Institute of Medical Research, A. Lanari, University of Buenos Aires-CONICET, Ciudad Autonoma de Buenos Aires, Argentina
3Research Council of GCBA, Ciudad Autonoma de Buenos Aires, Argentina
Uncorrected Proof
"....Recent epidemiological studies suggested an increased incidence of major cardiovascular events in patients with ultrasound-diagnosed NAFLD independent of traditional risk factors and components of the metabolic syndrome [7], [8], [24]. Moreover, the 14-year risk of cardiovascular mortality was doubled in 129 patients with biopsy-proven NAFLD compared with that of the reference population [25]......To provide a more objective basis to clinical recommendations and to determine the impact of NAFLD on carotid atherosclerosis, we conducted a meta-analysis......This study showed that carotid IMT is strongly associated with NAFLD, showing that patients with hepatic steatosis have an increase of 13% of IMT in comparison with individuals without fatty liver. This conclusion results from a total of 3497 individuals recruited from 7 heterogeneous studies. Besides, the presence of at least one carotid plaque was also strongly related with the presence of fatty liver disease in 3212 assessed individuals. Notably, meta-regression showed that liver function tests were strongly associated with carotid IMT and the presence of carotid plaques......The main strength of this study is the large sample of individuals screened for both fatty liver disease and carotid IMT.....In summary, because carotid IMT is a marker of early arterial wall change, including atherosclerosis and/or vascular hypertrophy, and patients with NAFLD are at increased risk of higher carotid IMT, routine detection of carotid IMT by B-mode ultrasonography is strongly recommended in patients with fatty liver disease and vice versa. This proposal may provide benefits on primary prevention and in the decision to treat the existing but not diagnosed cardiovascular disease in patients with fatty liver disease....."
Background/Aims
To perform a systematic review of the studies addressing the association between non-alcoholic fatty liver disease (NAFLD) and carotid intima-media thickness (IMT).
Methods
Literature searches identified seven studies that met inclusion criteria: population-based or hospital-based case-control studies about the relation between NAFLD and carotid IMT, in which information on number of subjects in controls and NAFLD patients, and data to evaluate carotid IMT and carotid plaques (measured by carotid ultrasound) could be extracted.
Results
From a total of 3497 subjects (1427 patients and 2070 controls), we found a significant association between NAFLD and carotid IMT either in fixed (D: 0.51, 95% CI: 0.44-0.58, p<1X10-8) or random models (D: 1.44, CI: 95% 0.63-2.24, p<0.0006). Meta-regression analysis showed that mean differences in alanine aminotransferase and ϒ-GT were strongly correlated with those in IMT (p<0.00006 and 0.004, respectively). In addition, 5 reports including 3212 subjects showed that carotid plaques were more frequently observed in NAFLD patients in comparison with controls, fixed model (p<1X10-10), OR: 1.97 95% CI: 1.67-2.32 and random model p<0.0002, OR: 3.13, 95% CI: 1.75-5.58.
Conclusion
Routine measurement of carotid IMT might be implemented in NAFLD patients, as they carry an increase of 13% of carotid IMT.
Introduction
The prevalence of non-alcoholic fatty liver disease (NAFLD) is increasing worldwide because of the rise of obesity and type 2 diabetes prevalence [1]. NAFLD is present in 10-24% of the general population in various countries [2] and is closely related to insulin resistance and markers of oxidative stress and endothelial dysfunction [3], [4]. In fact, NAFLD is considered the hepatic manifestation of the metabolic syndrome (MS) [3], [5].
People with metabolic syndrome are at risk for cardiovascular disease, including coronary heart disease and stroke [6]. The importance of NAFLD and its relationship with MS is now increasingly recognized as recent data suggest that NAFLD is linked to increased cardiovascular risk independently of the broad spectrum of risk factors of MS [7]. Indeed, it is hypothesized that NAFLD is not merely a marker of cardiovascular disease but may also be involved in its pathogenesis [8].
Various non-invasive markers of early arterial wall alteration are currently available, such as arterial wall thickening and stiffening, endothelial dysfunction and coronary artery calcification [9]. Of them, non-invasive assessment of carotid intima-media thickness (IMT) by high-resolution carotid B-mode ultrasonography (US) is widely used as a proxy end point for cardiovascular disease [10]. In addition, carotid US allows the evaluation - in a simple, safe and reproducible way - of lumen diameter, intima-media thickness, and the presence and extent of carotid plaques.
Recent studies have shown that NAFLD patients have significantly greater carotid IMT than age and sex-matched patients without NAFLD, independently of the classical risk factors of the metabolic syndrome. However, some degree of variability about the mean carotid IMT values' can be observed among all the published reports that result in a difficult evaluation of the magnitude of the observation. For instance, among the different studies, mean carotid IMT values in NAFLD patients range from 0.64±0.10mm to 1.24±0.13mm. The importance of this is to decide if further recommendations with regard to carotid atherosclerosis screening should be implemented in all NAFLD patients, as currently available epidemiological data indicate that a value of carotid IMT at or above 1mm at any age is associated with a significantly increased risk of myocardial infarction and/or cerebrovascular disease [11].
In view of the evidence mentioned above, our primary purpose was to estimate from the available literature, the strength of the increased carotid IMT and carotid atherosclerosis observed in NAFLD patients. In addition, we systematically evaluated the study characteristics that could be responsible for the association.
Results
We evaluated seven studies that met the selection criteria and that were identified using the search strategy described in Appendix Fig. A1. Studies characteristics are shown in Table 1. Data on one further study was unavailable because in the paper the authors did not disclose the raw data; the authors were contacted but they were unable to provide the data in the required format for evaluation [16]. In addition, the patients included in this study have fatty liver of unspecified aetiology.
Over 100% of the studies met the QUADAS quality-assessment criteria, which are summarized in Table 2. All the studies scored well in terms of adequate descriptions of selection criteria and reference test, blind assessment of the reference test and availability of clinical data.
Study characteristics
Five studies were hospital-based case-control [17], [18], [19], [20], [21], and other two were population-based case-control [22], [23] studies. Study characteristics are shown in Table 1.
In all the cases, carotid IMT was measured by B-mode ultrasound by an operator who was blind to the clinical characteristics of the patients. Carotid IMT measurements were made bilaterally at the level of common carotid artery. Maximum value of carotid IMT was used in some of these studies [19], [21], [22], whereas mean value of carotid IMT readings was used in others [17], [18], [20], [23].
Carotid plaque was defined as a focal thickening >1.2 and >1.3 [23] at the level of carotid artery.
Diagnosis about fatty liver disease was performed in all the studies by abdominal ultrasound referred to as gbrighth liver or diffuse increase of echogenicity of the liver compared to the of the kidney; in some studies fatty liver was confirmed by liver biopsy [17], [19], [22].
Control subjects were matched for age and sex in some studies [17], [18], [21], but for age, sex and BMI in others [19], [20]. One study included only male but age matched subjects [22], and other study was a cross-sectional survey from a random sample of population that was selected using population registries [23].
3.2. IMT
Data regarding of carotid IMT extracted from the 7 studies included 3497 individuals, and showed a significant association of increased carotid IMT with NAFLD either in the fixed (p<1X10-8) or the random model (p<0.0006) (Fig. 1).
We assessed between study heterogeneity by using the Q statistic and observed significant heterogeneity (p<0.0002), I2: 98.4%. Subjects were stratified by ethnicity, study design and associated disease condition but the heterogeneity remains significant.
As we hypothesized that heterogeneity may be due to liver disease severity, where sufficient data were available, we used regression analysis to investigate whether additional variables such liver enzymes (ALT and ϒ-GT) likely to be associated with severity of fat cytotoxicity were associated with the studied outcome effect and hence whether differences in these laboratory variables between the studies accounted for some of the observed heterogeneity. ALT and ϒ-GT were included as putative surrogate indicators of severity of fat cytotoxicity because carotid IMT value according to liver biopsy results (when available) was not disclosure by the authors except for one study [20]. The following six reports included data about ALT and ϒ-GT value [17], [18], [19], [20], [21], [23], which were included in the analysis. According to meta-regression analysis, carotid IMT mean differences showed strong correlation with ALT and ϒ-GT mean differences. This analysis revealed that ALT had a significant impact on the carotid IMT, as D of carotid IMT strongly correlates with ALT D values (slope=0.86, p<0.00006) (Fig. 2A) and carotid IMT D was also correlated with ϒ-GT D values (slope 0.66, p<0.004) (Fig. 2B).
From the Begg and Mazumdar's rank correlation test (Kendall's tau: 0.048, 2-tailed-p<1.0), it seems that the there was no publication bias.
We estimated the NAFLD effect on carotid IMT by calculating, from the included studies, the percentage of increase of carotid IMT in NAFLD patients in relation to controls as follows: the weighted mean of NAFLD patient carotid IMT minus the weighted mean of controls, divided by the weighted mean of controls multiplied by 100. This analysis showed that NAFLD patients carry an estimated increase of 13% in carotid IMT.
Carotid plaque
We found five heterogeneous reports (p<0.001, I2: 77.7) that evaluated carotid plaque [17], [18], [19], [20], [23]. The comparison between cases and controls, including 3212 subjects, showed that carotid plaque was more frequently observed in NAFLD patients than controls by fixed (p<1X10-10) or random effect (p<0.0002) models (Fig. 3). To investigate the source of heterogeneity, we analyzed the data by grouping the reports by study design and after removing one study [23], the heterogeneity was removed. By meta-regression analysis, we observed that MH ORs for carotid plaques was correlated with ALT and ϒ-GT mean differences (slope: 0.451, p<0.001 and slope: 0.395, p<0.003, respectively).
Begg and Mazumdar's rank correlation test (Kendall's tau: 0.001, 2-tailed p<0.81) showed that there seems not to be publication bias.
Discussion
Recent epidemiological studies suggested an increased incidence of major cardiovascular events in patients with ultrasound-diagnosed NAFLD independent of traditional risk factors and components of the metabolic syndrome [7], [8], [24]. Moreover, the 14-year risk of cardiovascular mortality was doubled in 129 patients with biopsy-proven NAFLD compared with that of the reference population [25].
A strong association between NAFLD and endothelial dysfunction as measured by brachial artery flow-mediated vasodilation - a reliable marker of early atherosclerosis, was recently described [26], [27]. However, despite several previous studies demonstrated the association between NAFLD and carotid IMT and/or carotid plaque, no general consensus exists on the systematic screening of carotid atherosclerosis in patients with fatty liver disease. In fact, the only general recommendation for management of NAFLD patients to date is related to lifestyle changes and an attempt at gradual weight loss along with appropriate control of serum glucose and lipid levels [28], [29].
To provide a more objective basis to clinical recommendations and to determine the impact of NAFLD on carotid atherosclerosis, we conducted a meta-analysis.
This study showed that carotid IMT is strongly associated with NAFLD, showing that patients with hepatic steatosis have an increase of 13% of IMT in comparison with individuals without fatty liver. This conclusion results from a total of 3497 individuals recruited from 7 heterogeneous studies. Besides, the presence of at least one carotid plaque was also strongly related with the presence of fatty liver disease in 3212 assessed individuals. Notably, meta-regression showed that liver function tests were strongly associated with carotid IMT and the presence of carotid plaques.
The main strength of this study is the large sample of individuals screened for both fatty liver disease and carotid IMT. In addition, this is the first study showing evidence-based data that may result in the formulation of management guidelines for NAFLD elsewhere in relation with prevention of cardiovascular events in individuals who until now are not currently included as carriers of potential risk factors for carotid atherosclerosis.
A note of caution should be added as the presence of heterogeneity may potentially restrict the interpretation of the pooled risk estimates. Heterogeneity in a meta-analysis is mostly produced by differences in study design and background characteristics of the subjects and the extent of heterogeneity might influence the conclusions. However, the random effect model, where heterogeneity is no longer a main issue, provided a significant result about both carotid features.
Although heterogeneity was addressed statistically by applying a random effects model, we aimed to further investigate potential sources of it where possible. Thus, the full dataset was utilized for investigation of heterogeneity by meta-regression. When main outcomes were analyzed according to liver enzymes (ALT and ϒ-GT values) and entered into a meta-regression, they proved to be a statistically significant source of heterogeneity. The main hypothesis forwarded to explain these results is the observation that in NAFLD group were included patients with different stages of disease severity, and consequently patients with more severe fatty liver disease are those who show elevated liver function test results. These data are in agreement with previously published epidemiological studies that have related serum liver enzymes with atherothrombotic risk profile and elevated risk of cardiovascular disease [30], [31], [32], [33].
A critique concerns the observation that few studies showed data about liver disease severity. The limited data on this issue may to some extent be attributable to the difficulties in performing liver biopsy as it is an invasive and costly procedure, and even in the most experienced hands it is prone to complications. Only one study disclosed the data of carotid IMT according to liver histology, and showed that carotid IMT was significantly different between patients with nonalcoholic steatohepatitis, patients with simple steatosis and controls subjects [20]. Therefore, this lack of information remains to be addressed.
The results of this meta-analysis have important clinical implications. First, this study shows that carotid atherosclerosis disease should be suspected when there are at least characteristic changes on hepatic ultrasonography that show fatty liver disease. Besides, recent data suggest that NAFLD is strongly associated with carotid atherosclerosis even in childhood [34]. Carotid plaques were more common in NAFLD patients. These data reinforce the importance of the described association as there is strong relation between plaque prevalence and stroke suggesting that the presence of any plaque on ultrasound increases the likelihood of clinical disease in the future [35].
Second, our analysis supports the association between liver enzymes and carotid atherosclerosis, and the potentially strong relation between ALT and ϒ-GT with carotid IMT. Finally, although this issue cannot be solved from the current study, our observation supports the notion that subjects with carotid atherosclerosis should be assessed for fatty liver disease.
In summary, because carotid IMT is a marker of early arterial wall change, including atherosclerosis and/or vascular hypertrophy, and patients with NAFLD are at increased risk of higher carotid IMT, routine detection of carotid IMT by B-mode ultrasonography is strongly recommended in patients with fatty liver disease and vice versa. This proposal may provide benefits on primary prevention and in the decision to treat the existing but not diagnosed cardiovascular disease in patients with fatty liver disease.
|
|
| |
| |
|
|
|
