C-terminal plasma FGF23 concentrations were measured, and LVM and LVH evaluated by echocardiogram among 2255 individuals ¡Ý65 years in the Cardiovascular Health Study. Linear regression analysis adjusting for demographics, cardiovascular, and kidney related risk factors examined the associations of FGF23 concentrations with LVM. Analyses were stratified by CKD status and adjusted linear and logistic regression analysis explored the associations of FGF23 with LVM and LVH.
Among the entire cohort, higher FGF23 concentrations were associated with greater LVM in adjusted analyses (¦Â?=?6.71 [95 % CI 4.35-9.01]?g per doubling of FGF23). 32 % (n?=?624) had CKD (eGFR <60?mL/min/1.73?m2 and/or urine albumin-to-creatinine ratio >30?mg/g). Associations were stronger among participants with CKD (p interaction?=?0.006): LVM ¦Â?=?9.71 [95 % CI 5.86-13.56]?g per doubling of FGF23 compared to those without CKD (¦Â?=?3.44 [95 % CI 0.77, 6.11]?g per doubling of FGF23). While there was no significant interaction between FGF23 and CKD for LVH (p interaction?=?0.25), the OR (1.46 95 % CI [1.20-1.77]) in the CKD group was statistically significant and of larger magnitude than the OR for in the no CKD group (1.12 [95 % CI 0.97-1.48]).
In a large cohort of older community-dwelling adults, higher FGF23 concentrations were associated with greater LVM and LVH with stronger relationships in participants with CKD.