Circulating retinol binding protein 4 levels in coronary artery disease: a systematic review and meta-analysis

Background Retinol binding protein 4 (RBP4) has been proposed to play a role in the pathophysiology of coronary artery disease (CAD), but previous findings on the association of RBP4 levels with CAD are inconsistent. Methods A meta-analysis based on observational studies was conducted to evaluate the association between circulating RBP4 levels and CAD. Databases including PubMed, Web of Science, Embase, Google Scholar and ClinicalTrials.gov database were searched for eligible studies published up to 12 July 2021. Standard mean differences (SMDs) with 95% confidence intervals (CIs) were calculated using the inverse variance heterogeneity (IVhet) and random-effects model for data with moderate and high heterogeneity (I-2 > 30%) and data with low heterogeneity were analysed using a fixed-effects model (I-2 <= 30%). Moreover, a bias-adjusted quality-effects model was generated, and the prediction interval was also calculated under the random-effects model. Results Two nested case-control studies, one cohort study and twelve case-control studies with a total of 7111 participants were included. Circulating RBP4 levels in patients with CAD were comparable to those in the controls under the IVhet model (SMD: 0.25, 95% CI: - 0.29-0.79, I-2: 96.00%). The quality-effects model produced consistent results. However, the association turned to be significant under the random-effect model (SMD: 0.46, 95% CI: 0.17-0.75, I-2: 96.00%), whereas the 95% predictive interval (PI) included null values (95% PI: - 0.82-1.74). Subgroup analyses illustrated a positive relationship between CAD and RBP4 levels in patients with complications (SMD: 1.34, 95% CI: 0.38-2.29, I-2: 96.00%). The meta-regression analysis revealed that the mean BMI of patients (P = 0.03) and complication status (P = 0.01) influenced the variation in SMD. Conclusions There was low-quality evidence that patients with CAD exhibited similar circulating RBP4 levels compared with controls, and high inter-study heterogeneity was also observed. Thus, RBP4 might not be a potential risk factor for CAD. Comparisons among different subtypes of RBP4 with larger sample size are needed in the future.