Osteoprotegerin and Ischemic Stroke Prognosis: A Prospective Multicenter Study and Mendelian Randomization Analysis

Abstract

Background: Osteoprotegerin was implicated in vascular injury and inflammatory responses, but its prognostic value in ischemic stroke remained unclear. We aimed to prospectively investigate the association between plasma osteoprotegerin and ischemic stroke prognosis combined with a Mendelian randomization analysis. Methods: Our prospective study follows the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) reporting guideline. We measured baseline plasma osteoprotegerin levels for 3490 ischemic stroke patients recruited between August 2009 and May 2013 in 26 hospitals across China. The primary outcome was a composite outcome of death and major disability at 3 months after ischemic stroke. Results: After adjustment for age, sex, admission National Institutes of Health Stroke Scale score, and other important covariates, elevated osteoprotegerin levels were associated with increased risks of primary outcome (odds ratio, 1.40 [95% CI, 1.05–1.88]), death (hazard ratio, 2.05 [95% CI, 1.04–4.08]), and composite outcome of death and vascular events (hazard ratio, 2.00 [95% CI, 1.15–3.48]) when 2 extreme quartiles were compared. Each 1-SD higher log-osteoprotegerin was associated with a 18% (95% CI, 6%–32%) increased risk of primary outcome, 69% (95% CI, 31%–118%) increased risk of death, and 53% (95% CI, 24%–89%) increased risk of composite outcome of death and vascular events, respectively. Multiple-adjusted spline regression model showed a linear association of osteoprotegerin with primary outcome ( P for linearity <0.001). Adding osteoprotegerin to conventional risk factors did not significantly improve discriminatory power (C statistics, 0.817 versus 0.818; P =0.232) but did slightly improve the risk reclassification of primary outcome (net reclassification improvement: 13.68%, P <0.001; integrated discrimination improvement: 0.23%, P =0.039). In Mendelian randomization analysis, genetically determined high plasma osteoprotegerin was associated with increased risk of primary outcome (odds ratio, 5.74 [95% CI, 1.12–29.44]; P =0.036). Conclusions: Elevated plasma osteoprotegerin was associated with poor prognosis of ischemic stroke, and genetically determined high plasma osteoprotegerin was associated with an increased risk of primary outcome in Mendelian randomization analysis. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT01840072.