Accurate assessment of glomerular filtration rate (GFR) is important in ascertaining health status, especially in older individuals and in populations at risk of chronic kidney disease. Recently we observed that smokers with renal artery stenosis and renovascular hypertension had lower GFR when estimated by cystatin-C, than with creatinine-based estimates. We sought to resolve this issue.Design and method:
Data from the Diabetes Control and Complications Trial (DCCT) were used to estimate GFR in individuals with diabetes. Iohexol clearance was used as the standard, and was compared to GFR estimates using creatinine and cystatin-C. Creatinine and cystatin-C estimates were calculated with the 24-hour urine creatinine clearance and the CKD-EPI cystatin-C. formula, and compared to iohexol based estimates using the Bland-Altman method.Results:
There were 441 patients with iohexol, creatinine and cystatin-C values. The average age was 27.6 ± 6.6 years. The mean iohexol GFR for the population was 127.5 ± 19.2 ml/min. The cystatin-C and creatinine-based estimates were close to the iohexol estimates in non-smokers (differing by −7.7 ± 24.7 and −2.7 ± 31.1 ml/min respectively, p = ns). However in smokers, the cystatin-C estimates were considerably lower than iohexol-based GFR estimates (difference = −21.4 ± 25.1 ml/min p < 0.001), while creatinine based estimates were not (difference = −4.6 ± 21.6 ml/min, p = 0.58).Conclusions:
Cystatin-C based estimates of GFR are accurate in non-smoking diabetics and yet substantially and significantly underestimate GFR in diabetics that smoke cigarettes.