Difference between cystatin C- and creatinine-based estimated glomerular filtration rate, diabetes status, and incident peripheral artery disease: A population-based cohort study.

Background And Aims: Difference between cystatin C and creatinine-based estimated glomerular filtration rate (eGFRdiff) has been suggested to reflect factors that are associated with vascular risk, independent of kidney function. We aimed to prospectively evaluate the association between eGFRdiff and peripheral artery disease (PAD), as well as the potential modifying role of diabetes.

Methods: This prospective cohort study included 466,245 participants with concurrent measured serum creatinine and cystatin C and free of PAD at baseline (2006-2010) from the UK Biobank. eGFRdiff was calculated as absolute difference (eGFRabdiff) and the ratio (eGFRrediff) between cystatin C- and creatinine-based eGFRs. The incidence of PAD was ascertained using electronic health records. Cox proportional hazards regression models were used to evaluate the associations of eGFRdiff with incident PAD. Potential modification by diabetes was examined. The relative importance and the additive value of eGFRdiff in predicting PAD was evaluated.

Results: During a median follow-up of 13.8 years, PAD developed in 7210 participants. Each standard deviation increment of eGFRabdiff was associated with a 33 % lower risk of PAD. For each 10 % increment in eGFRrediff, the hazard ratio (95 % confidence interval) was 0.78 (0.77, 0.80) for PAD. Consistent associations were observed between eGFRdiff and PAD, irrespective of the presence of diabetes. Adding eGFRdiff into the established model could improve the performance of PAD prediction.

Conclusion: eGFRdiff was significantly associated with risk of incident PAD irrespective of the presence of diabetes and provided additional value in predicting PAD.