Abstract
Background:
Numerous studies have assessed the association of Nonalcoholic fatty liver disease (NAFLD) withcardiovascular disease (CVD). However, results have been conflicting due to variability in definitionsof NAFLD and ascertainment of CVD, often combining clinical and surrogate endpoints. We therefore systematically reviewed published literature to assess the association between NAFLD and clinical cardiovascular events.
Methods:
We searched Medline, Cochrane, google scholar, CINAHL, and Web of Sciencedatabasesusing terms “nonalcoholic fatty liver disease”, “cardiovascular disease”, and their combinations to identify studies published through March 2015. Data from selected studies was extracted and meta-analysis was then performed using Random effects model following the PRISMA guidelines. Publication bias and heterogeneity wereassessed. The main outcome measure was Odds ratio (OR) with 95% CI. Clinical CVD was defined as symptomatic coronary artery disease, myocardial infarction, coronary or peripheral intervention, ischemic stroke, and symptomatic peripheral vascular disease.
Results:
A total of 7 studies with 14634 patients (NAFLD: 4204; controls: 10430) were included in the final analysis. 3 studies were cross- sectional reporting prevalence, while 4 studies were prospective cohort studies reporting incidence. Patients with NAFLD had a significantly higher risk of clinical CVD compared to controls [OR: 3.17; 95% CI: 1.89-5.30, P<0.01) (figure 1A). There was significant heterogeneity (I2=93%). Funnel plot and Begg’s test did not reveal significant publication bias. Separate analyses of the cohort and cross sectional studies and exclusion sensitivity analysis did not alter the findings (figure 1B).
Conclusion:
NAFLD is associated with a three fold increase in the risk of clinical CVD compared to controls without NAFLD. These results need to be conformed in large prospective studies.