Abstract: PO0348
Factors Associated with Change in Fibroblast Growth Factor 23 Levels from Midlife to Late-Life: The ARIC Study
Session Information
- Biochemical Aspects of Mineral and Bone Disease
October 22, 2020 | Location: On-Demand
Abstract Time: 10:00 AM - 12:00 PM
Category: Bone and Mineral Metabolism
- 402 Bone and Mineral Metabolism: Clinical
Authors
- Ishigami, Junichi, Johns Hopkins University, Baltimore, Maryland, United States
- Honda, Yasuyuki, Johns Hopkins University, Baltimore, Maryland, United States
- Karger, Amy B., University of Minnesota, Minneapolis, Minnesota, United States
- Selvin, Elizabeth, Johns Hopkins University, Baltimore, Maryland, United States
- Coresh, Josef, Johns Hopkins University, Baltimore, Maryland, United States
- Lutsey, Pamela L., University of Minnesota, Minneapolis, Minnesota, United States
- Matsushita, Kunihiro, Johns Hopkins University, Baltimore, Maryland, United States
Background
Factors associated with change in fibroblast growth factor 23 (FGF23) levels from midlife to late-life are not well-characterized in the general population.
Methods
Among 5,881 participants of the Atherosclerosis Risk in Communities Study who had serum level of FGF23 measured during midlife (visit 3, 1993-1995, mean age 58 years, 58% women, 23% black race) and late-life (visit 5, 2011-2013, mean age 76 years), we explored demographic and clinical factors associated with change in FGF23 levels. Change in FGF23 levels was regressed on pre-specified factors assessed at visit 3 of age (above vs. below median), sex, race, ever smoke, high BMI (≥ vs. <30 kg/m2), hypertension, diabetes, history of CVD, and reduced eGFR (≥ vs. <60 ml/min/1.73m2) using multivariable linear regression models.
Results
The mean FGF23 level increased by 21.0 (95%CI, 20.3-21.6) pg/mL from 39.7 at visit 3 to 60.6 pg/mL at visit 5. Reduced eGFR, diabetes, hypertension, female, older age, and white race were significantly associated with a greater increase in FGF23 levels (Table). Although history of CVD demonstrated a similar magnitude as race, the β coefficient was not significant. We also did not observe significant associations for BMI or smoking. The associations were strongest for reduced eGFR and diabetes with similar degrees of associations (ΔFGF23, 6.7 [95%CI, 2.7 to 10.6] pg/mL for reduced eGFR and 6.7 [4.4 to 9.0] pg/mL for diabetes) independent of each other.
Conclusion
In addition to reduced eGFR, we identified diabetes, hypertension, female, older age, and white race as predictors of an increase in FGF23 levels from midlife to late-life. Among these, the strong association of diabetes independent of kidney function deserves future investigations.