ASSOCIATION BETWEEN NONALCOHOLIC FATTY LIVER DISEASE AND CARDIO-METABOLIC OUTCOMES: A SYSTEMATIC REVIEW AND META-ANALYSIS EVALUATING THE RISK OF DIABETES, CARDIOVASCULAR EVENTS, AND MORTALITY

Abstract

Background:

Nonalcoholic fatty liver disease (NAFLD) is being recognized as a multi-system metabolic disorder that is closely related to insulin resistance and cardiovascular risk. While often independent of obesity, NAFLD has been linked to poor cardio-metabolic outcomes, and the extent and consistency of associations of the disease with incident type 2 diabetes mellitus (T2DM), cardiovascular events and mortality in population-based cohorts is uncertain.

Objectives:

To systematically review and quantitatively synthesize evidence of the association of NAFLD with risk of incident T2DM, CVD events and mortality in adult populations.

Methods:

PubMed, Scopus and Web of Science were also systematically searched for literature on the inception of data bases assessing the association of NAFLD with the occurrence of T2DM, cardiovascular events or mortality. Studies with multivariable-adjusted hazard ratios (HRs) of the risk of having versus not having NAFLD compared to non-NAFLD were included in the quantitative meta-analysis. HRs were pooled by use of random-effects models (DerSimonian-Laird). Heterogeneity was tested using the I² statistic. Articles that had reported different contrasts (e.g., NAFLD regression VS persistence, severity-based comparisons, phenotype-specific analyses, etc.) or incomparable effect measures were synthesized in the form of narrative synthesis. The review was conducted using the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Results:

Ten cohort studies were eligible for the systematic review with five of them providing comparable multivariable-adjusted HRs and eligible for quantitative synthesis. A significantly higher risk of incident T2DM was linked to NAFLD (pooled HR 2.04, 95% CI 1.52-2.74; I²= 38.2%). The combined effect of NAFLD and cardiovascular events demonstrated non-significant positive trend of risk increase with a high degree of heterogeneity (pooled HR 1.09, 95% CI 0.98-1.21; I² = 82.4%). Mortality quantitative pooling could not be done because there were only a few eligible studies; nevertheless, biopsy-confirmed NAFLD was linked to very high all-cause and cardiovascular mortality on a large nationwide cohort. Narrative synthesis of non-poolable studies supported a graded cardio-metabolic risk with severity of the disease, persistence of hepatic steatosis and adverse metabolic phenotypes, including lean NAFLD.

Conclusion:

NAFLD is linked with a significantly higher risk of incident T2DM and is heterogeneously related to cardiovascular events, with some NAFLD phenotypes and risk severity strata having a higher risk. Evidence from large cohorts suggests that there is an increase in mortality in individuals with biopsy-confirmed NAFLD. These results indicate that NAFLD is a significant cardio-metabolic risk factor and supports evidence-based combined metabolic and cardiovascular risk evaluation in NAFLD patients.