Osteoprotegerin concentration and risk of cardiovascular outcomes in nine general population studies: Literature-based meta-analysis involving 26,442 participants

Abstract

Background Recent experimental and epidemiological studies have suggested that osteoprotegerin, a key regulator in bone metabolism, may be involved in vascular calcification and atherosclerosis. Our aim was to reliably quantify the associations of osteoprotegerin concentration and incidence of first-ever cardiovascular disease outcomes in the general population. Methods Using the electronic databases MEDLINE, EMBASE and Web of Science (January 1975 and April 2017, no language restrictions), nine relevant studies were identified involving a total of 26,442 participants recruited from the general population. Over a mean follow-up of 8.5 years, 2,160 cardiovascular disease, 2,123 coronary heart disease, and 1,102 stroke outcomes were recorded. Study-specific risk ratios were combined with random-effects meta-analysis. Results When comparing individuals in the top with those in the bottom third of osteoprotegerin concentration, the combined risk ratio was 1.83 (95% confidence interval: 1.46, 2.30; P<0.001; I2 = 76.8%) for cardiovascular disease, 1.72 for coronary heart disease (1.26, 2.37; P = 0.001; I2 = 83.5%), and 1.58 for stroke (1.18, 2.12; P = 0.002; I2 = 65.2%). Associations appeared stronger at younger age (P = 0.018 for cardiovascular disease), in studies that did not employ statistical adjustment (P = 0.023 for cardiovascular disease and 0.018 for coronary heart disease), and potentially in studies that measured osteoprotegerin in plasma rather than in serum (P = 0.005 for cardiovascular disease and 0.018 for coronary heart disease). Magnitudes of associations did not differ according to the proportion of males, geographical region, or osteoprotegerin assay manufacturer. There was no evidence for publication bias for any of the outcomes assessed (all P>0.05). Conclusions Elevated osteoprotegerin concentration is associated with an increased risk of incident cardiovascular disease in the general population. The mechanisms underlying this observation deserve further investigation.