Abstract 12529: The Association of the Different Vitamin D Metabolites to Death, Myocardial Infarction, Heart Failure, and Stroke: Results From the Intermountain Heart Collaborative Study
Introduction: Many epidemiological studies have shown that total circulating levels of 25-hydroxyvitamin D (VitD) are strongly associated with poor cardiovascular (CV) outcomes. However, 85-90% of circulating VitD is bound to vitamin D binding protein (BP) or albumin, and levels obtained may not be truly reflective of the VitD available to act on target cells. Whether more direct markers of VitD activity such as bioavailable VitD (BIO) or free VitD (FREE) are more effective in predicting future CV risk is not known.
Hypothesis: The more direct markers of VitD activity are more strongly associated with adverse CV outcomes.
Methods: Patients (N=4285) undergoing angiography for CAD determination were studied. Levels of VitD, BP and albumin were directly measured and BIO and FREE were calculated using standard formulae. Multivariable Cox hazard regression was utilized to determine quartile (Q) associations with death (all-cause, CAD, and cardiac), MI, HF hospitalization, and stroke. Median length of follow-up was 10.9 years.
Results: Patients averaged 64±12 years, 66% male, 23% diabetic, 90% Caucasian, and 70% diagnosed with CAD. Median and interquartile ranges of VitD, BP, BIO, and FREE were 24.9 (20.0, 28.7) ng/mL, 168.3 (103.7, 245.9) ug/mL, 7.1 (3.9, 15.2) ng/mL, and 15.7 (9.2, 33.3) pg/mL, respectively. Multivariable hazard ratios for the various metabolites with each of the outcomes are shown in Table. BP was not associated with any of the outcomes. VitD, BIO, and FREE were all associated with death (all-cause, CAD, and cardiac). VitD continued to be significantly associated with HF hospitalization and stroke, with its association to MI being attenuated.
Conclusion: Among this predominately Caucasian population, VitD was found to consistently be the strongest predictor of all of the outcomes over BP, BIO, and FREE. Future studies are needed to evaluate whether obtaining thresholds of VitD levels are associated with a reduction in CV risk.
Author Disclosures: H.T. May: None. O. Galenko: None. J.F. Carlquist: None. L.K. Weaver: None. S. Knight: None. T. Barker: None. K.U. Knowlton: None. J.B. Muhlestein: None.
- © 2016 by American Heart Association, Inc.