Aims: Vitamin D deficiency has been associated with risk of overt cardiovascular disease (CVD), but associations with subclinical disease are not well characterized. Hence, we examined associations of circulating vitamin D concentrations and left ventricular (LV) geometry and function by echocardiography at baseline and after 5 years in a community-based study.
Methods and results: In the PIVUS study, we measured serum 25-dihydroxyvitamin-D (25-OH D) at age 70 and performed echocardiography including LV mass, wall thickness, end-diastolic diameter, end-systolic diameter (LVESD), left atrial diameter, fractional shortening, ejection fraction, isovolumic relaxation time, and E/A ratio at both age 70 and 75. We included 870 participants (52% women) without prior myocardial infarctions, heart failure, or prevalent valvular disease. After adjusting for potential confounders, 25-OH D at baseline was found to be significantly associated with LVESD, fractional shortening, and ejection fraction (β, -0.42 mm, P = 0.03; β, 0.70%, P = 0.03; and β, 0.91% P = 0.01, respectively), per 1 SD increase in 25-OH D (SD = 20 nmol/L) at baseline. In longitudinal analyses, vitamin D levels at baseline were not significantly associated with change in LV geometry and function after 5 years.
Conclusion: In our community-based study among the elderly, we found higher circulating vitamin D concentrations to be associated cross-sectionally with better LV systolic function and smaller LVESD at baseline. The association persisted after adjusting for several potential confounders, including cardiovascular risk factors and calcium, phosphate, and parathyroid hormone levels. Randomized clinical trials are needed to establish firmly or refute a causal relationship between vitamin D levels and changes in LV geometry and function.