Article

Background and objectivesHigh levels of fibroblast growth factor 23 are associated with accelerated progression of CKD. Whether high fibroblast growth factor 23 levels also predict incident CKD is uncertain. Design, setting, participants, & measurementsA prospective case-cohort study was conducted within the Action to Control Cardiovascular Risk in Diabetes Trial. The analytic sample consisted of a random subcohort of 590 patients with type 2 diabetes without prevalent CKD at baseline, 124 of whom developed incident CKD during follow-up, and 520 additional patients with incident CKD outside the random subcohort. The association between serum intact fibroblast growth factor 23 and incident CKD, defined as the new onset of eGFR,60 ml/min per 1.73m2 that represented a$25%decrease frombaseline in an individualwith eGFR$60ml/min per 1.73m2 and no microalbuminuria (,30 mg/g creatinine) at baseline, was tested. Results The mean baseline eGFR in the random subcohort was 90.9622.7 ml/min per 1.73 m2. During a median follow-up of 4.7 years, therewas a total of 644 patients with incident CKD. Themedian baseline fibroblast growth factor 23 level was modestly higher among patients with incident CKD versus controls (43.5, interquartile range=34.7–55.1 versus 39.8, interquartile range=31.9–49.5 pg/ml; P,0.001). Higher baseline fibroblast growth factor 23 levels were associated with higher risk of incident CKD in unadjusted and demographics-adjusted models, but the effect was attenuated after additional adjustment for clinical risk factors and baseline eGFR (hazard ratio per SD of natural log fibroblast growth factor 23, 1.09; 95 % confidence interval, 0.94 to 1.27), which was the strongest predictor of incident CKD. Consistent with the results of primary analyses, baseline fibroblast growth factor 23 was not associated with eGFR slope. Conclusions Higher fibroblast growth factor 23 levels are not independently associated with higher risk of incident CKD in patients with type 2 diabetes