Associations of retinal microvascular caliber with intermediate phenotypes of large arterial function and structure: A systematic review and meta-analysis.

Liu, Mengjiao; Wake, Melissa; Wong, Tien Yin; He, Mingguang; Xiao, Yinzong; Burgner, David; Lycett, Kate

OBJECTIVE Intermediate phenotypes of microcirculation (retinal microvascular caliber) are associated with cardiovascular (CV) risk factors and independently predict CV events. However, the effect of microcirculation variation on the vascular system is unclear. We conducted a systematic review and meta-analysis of observational studies to quantify associations of retinal microvascular caliber (arteriolar, venular caliber, arteriole-to-venule ratio) and preclinical CV measures (large arterial function and structure). METHODS We identified studies in Medline, Embase and Pubmed (1946 to March 2018) studying (a) general population samples and (b) patients with cardiometabolic disease. Study-specific correlation estimates were combined into meta-analysis where possible. RESULTS Of 1,294 studies identified, 26 met inclusion criteria (general population 16, patients 10), of which five studies were included in meta-analysis. Most studied middle-aged adults cross-sectionally, with one childhood study. Large arterial function and structure were predominantly assessed by pulse wave velocity and carotid intima-media thickness, respectively. Only arteriolar caliber was consistently associated with arterial function and structure, with stronger associations observed in cardiometabolic patients. Narrower (worse) arteriolar caliber was associated with faster (poorer) pulse wave velocity (correlation coefficient (r) -0.17, 95% CI -0.25 to -0.10) and greater (poorer) intima-media thickness (r -0.05, 95%CI -0.09 to -0.02) across all adult participants. CONCLUSIONS Retinal arteriolar, but not venular caliber, was modestly associated with large arterial function and weakly associated with large arterial structure, with stronger evidence in patients with cardiometabolic disease. This suggests that preclinical changes in large arteries and the microcirculation have some shared but mainly unique pathway to associate with cardiovascular disease. This article is protected by copyright. All rights reserved.