Associations of FGF23 with change in bone mineral density and fracture risk in older individuals

T Isakova, X Cai, J Lee, R Katz… - Journal of Bone and …, 2016 - academic.oup.com
T Isakova, X Cai, J Lee, R Katz, JA Cauley, LF Fried, AN Hoofnagle, S Satterfield, TB Harris…
Journal of Bone and Mineral Research, 2016academic.oup.com
Elevated levels of the phosphate‐regulating hormone fibroblast growth factor 23 (FGF23)
have been linked to greater risk of fractures in some studies, especially among individuals
with chronic kidney disease (CKD). We evaluated FGF23 as a risk factor for bone loss and
fractures in the Health, Aging, and Body Composition (Health ABC) study, which is a
prospective biracial cohort of well‐functioning adults aged 70 to 79 years recruited at two
clinical centers in the United States. The sample for the bone mineral density (BMD) …
Abstract
Elevated levels of the phosphate‐regulating hormone fibroblast growth factor 23 (FGF23) have been linked to greater risk of fractures in some studies, especially among individuals with chronic kidney disease (CKD). We evaluated FGF23 as a risk factor for bone loss and fractures in the Health, Aging, and Body Composition (Health ABC) study, which is a prospective biracial cohort of well‐functioning adults aged 70 to 79 years recruited at two clinical centers in the United States. The sample for the bone mineral density (BMD) analyses consisted of 2234 participants who had at least two serial total hip areal BMD measures. The fracture analyses included 2786 participants, 567 of whom sustained a fracture during a median follow up of 4.95 years. Linear mixed‐effects models were used for longitudinal measurements of total hip areal BMD and the proportional subdistribution hazard regression model subject to competing risks of death was used for risk of fracture. The median FGF23 was 46.7 (interquartile range [IQR] 36.7 to 60.2) pg/mL. The mean annualized percent change in total hip areal BMD did not vary significantly according to FGF23 quartile in all participants (p for trend = 0.70), but the effect was modified by CKD status (adjusted p for interaction <0.001). Among participants with CKD, the unadjusted mean annualized percent change in total hip areal BMD was greater with higher levels of FGF23 (unadjusted p for trend = 0.02), but the trend was attenuated with adjustment for estimated glomerular filtration rate and parathyroid hormone (adjusted p for trend = 0.30). FGF23 was not significantly associated with fracture risk in crude (hazard ratio [HR] per doubling of FGF23, 0.97; 95% CI, 0.85 to 1.12) or adjusted models (HR per doubling of FGF23, 1.02; 95% CI, 0.86 to 1.22), and these findings were not modified by gender or CKD status. FGF23 levels are not associated with bone loss or fracture risk in older adults with low prevalence of CKD. © 2015 American Society for Bone and Mineral Research.
Oxford University Press