Abstract P004: Estimating the Mortality Benefits of Ideal Cardiovascular Health: A Dose-response Meta-analysis
Background: Studies have shown that individuals with higher levels of Ideal Cardiovascular Health (ICVH) metrics have favorable health outcomes. However, none of these studies quantify the reduction in the risk of death per unit increase in ICVH. In this study, we estimated the reduction in CVD and all-cause mortality attributable to per unit improvement in ICVH.
Methods: Five studies identified through a literature search were included. Major differences in the re-categorization of the 7 ICVH metrics were noted, with studies having between 3-6 groups. Reference groups also ranged from 0 to ≤2 ideal CVH metrics. Because of these differences, risk reduction per unit increase in ICVH was not comparable across studies. To overcome these challenges, we quantified association between ICVH metrics and all-cause/CVD mortality using a dose-response meta-analysis, so all exposure categories contribute to a defined pooled association and heterogeneity is reduced by accounting for values assigned to these qualitative metrics.
Results: Original data and meta-analysis model of studies assessing CVD mortality are shown in the figure. There was a significant linear trend ( p < 0.001) with moderate heterogeneity (Q test = 12, p = 0.02 and I2 = 66 %). Each unit increase in ICVH metric was associated with 18% (95% CI: 12, 23) reduction in HR. In all-cause mortality analysis (graph not shown), there was also a significant linear trend (p < 0.001) with moderate heterogeneity (Q test = 18, p = 0.003, and I2 = 72 %). Each unit increase in ICVH metric was associated with a 10% (95% CI: 6, 13) reduction in HR for all cause mortality.
Conclusion: Our study is the first meta-analysis to demonstrate a dose-response impact of ICVH on mortality. The novel dose-response design allowed us to investigate strength, direction and type of association as well as overcomes challenges posed by heterogeneity ICVH metrics categorization while preserving the statistical control for confounding variables that was conducted in the individual studies.
Author Disclosures: E.C. Aneni: None. A. Crippa: None. A. Younus: None. C.U. Osondu: None. E. Veledar: None. K. Nasir: None.
- © 2016 by American Heart Association, Inc.