Increased hip fracture and mortality in chronic kidney disease individuals: the importance of competing risks.

BACKGROUND Many studies have shown a correlation between chronic kidney disease (CKD) and fracture. However, increased mortality in CKD patients is a competing risk scenario not accounted for in previous studies. Our aim was to investigate the true impact of CKD on hip fracture after accounting for a competing risk with death. METHODS We conducted a population-based cohort study to determine the impact of CKD on hip fractures in individuals aged ≥50years old registered in the SIDIAP(Q) database (representative of 1.9 million people in Catalonia, Spain). Cox regression was used to estimate hazard ratio (HR) for death and hip fracture according to CKD status. A competing risk (Fine and Gray) model was fitted to estimate sub-HR for hip fracture in CKD or CKD-free patients accounting for differential mortality. RESULTS A total of 873,073 (32,934 (3.8%) CKD) patients were observed for 3 years. During follow-up, 4,823 (14.6%) CKD and 36,328 (4.3%) CKD-free participants died (HR, 1.83 [95% CI, 1.78-1.89]), whilst 522 (1.59%) and 6,292 (0.75%) sustained hip fractures, respectively. Adjusted Cox models showed a significantly increased risk of hip fractures for the CKD group (HR, 1.16 [1.06-1.27]), but this association was attenuated in competing risk models accounting for mortality (SHR, 1.14 [1.03-1.27]). CONCLUSIONS Both death and hip fracture rates are increased (by 83% and 16%, respectively) in CKD patients. However, the association between CKD and hip fractures is attenuated when an excess of mortality is taken into account. A competing risk with death must be considered in future analyses of association between CKD and any health outcomes.