Abstract
BACKGROUND: We conducted a systematic review and meta-analysis to study the association between non-alcoholic fatty liver disease (NAFLD) and incident cardiovascular disease (CVD). METHODS: We searched Medline, Embase, Cochrane database and TRIP database. Random-effects model meta-analyses were used to obtain pooled effect sizes and 95% confidence intervals. The certainty in evidence was rated using the GRADE tool. RESULTS: Altogether 36 studies including a total of 7,068,007 participants were included in the systematic review and meta-analysis. Pooled data from 19 cohort studies demonstrated a significant increase in the risk of non-fatal CVD events in patients with NAFLD (HR 1.57, 95% CI 1.33-1.85, I(2) = 95%). Pooled data from eight studies showed a significant increase in fatal CVD (HR 1.40, 95% CI 1.24-1.57, I(2) =27%), and eight cohort studies suggested a significant increase in combined non-fatal and fatal CVD (HR 1.41, 95% CI 1.13-1.76, I(2) =80%). Meta-analysis of studies reporting adjusted estimates in NAFLD patients with fibrosis revealed a significant increase in CVD events with acceptable level of heterogeneity (HR 1.64, 95% CI 1.25-2.16, I(2) = 31%). The anticipated absolute increase in the risk of combined fatal and non-fatal CVD was estimated to be 29 more per thousand with NAFLD; that of fatal CVD events 16 more per thousand and that of non-fatal CVD events 19 more per thousand with NAFLD. The GRADE rating ranged from very low to low for overall and subgroup analyses. CONCLUSION: The present systematic review suggests that NAFLD increases the risk of incident CVD. Cohort studies with the ability to analyze subgroup effects based on severity, along with randomized controlled trials that provide experimental evidence demonstrating a decrease in cardiovascular disease events through the treatment of non-alcoholic fatty liver disease, are necessary to validate and reinforce these findings.