The relationship between NAFLD and retinol-binding protein 4 - an updated systematic review and meta-analysis

PURPOSE: Retinol-binding protein 4 (RBP4) has been considered to be related to metabolic related diseases, such as hyperuricemia, obesity, and diabetes mellitus. However, whether nonalcoholic fatty liver disease (NAFLD) is related to RBP4 is unclear. Previous studies on the relationship between NAFLD and RBP4 levels have yielded inconsistent results. Hence, this meta-analysis was aimed to clarify whether circulating RBP4 levels are in relation to the risk of NAFLD. METHODS: A meta-analysis was performed by applying observational studies to evaluate circulating RBP4 levels and NAFLD. Eligible studies published up to September 23, 2022, were searched in Embase, PubMed, and Cochrane databases. RESULTS: In this study, 17 cross-sectional studies involving 8423 participants were included. Results from a random effects model showed that circulating RBP4 levels were higher in NAFLD patients than non-NAFLD (standardized mean difference (SMD) 0.28; 95% confidence intervals (CI): 0.11–0.46, I(2): 89.8%). This association was confirmed in the Yellow race. However, no significant association was noted in the Caucasian race. After excluding the morbidly obese Population from the weight loss study (n = 2), the results of the comparison remained largely unchanged (SMD 0.28; 95% CI: 0.10–0.47, I(2): 90.8%). Remarkable publication bias was not found. Although considerable heterogeneity was observed among the studies, no potential sources of heterogeneity were found in the subgroup analysis. Diagnostic methods for NAFLD were determined to be a potential source of statistical heterogeneity in meta-regression. CONCLUSION: The findings provide evidence that NAFLD patients exhibit higher levels of circulating RBP4 compared with controls, but high heterogeneity was observed. Thus, a high RBP4 level is probably a potential risk factor for NAFLD. To confirm the causal link between NAFLD and RBP4 level of causality, further prospective cohort studies are needed.