Background Estimates of glomerular filtration rate (GFR), including creatinine and creatinine based formulae, are inaccurate in extremes of GFR and substantially biased in patients with chronic heart failure (CHF).

Objective To investigate whether serum cystatin C levels would be a better, more accurate and simple alternative for estimation of GFR and prognosis in CHF.

Design Cohort study.

Setting Chronic heart failure.

Patients, interventions and main outcome measure In 102 CHF patients, the correlation between GFR as estimated by ¹²⁵I-iothalamate clearance (GFR_IOTH), the modification of diet in renal disease formula (GFR_MDRD) and cystatin C was investigated. The combined endpoint consisted of the first occurrence of all cause mortality, heart transplantation or admission for CHF within 24 months.

Results Mean age was 58±12 years; 77% were male. Mean left ventricular ejection fraction was 28±9%. Mean GFR_IOTH was 75±27 ml/min/1.73 m², while median cystatin C levels were 0.80 (0.69–1.02) mg/l. GFR_IOTH was strongly correlated with all renal function estimates, including 1/cystatin C (r=0.867, p<0.001). GFR_IOTH was better predicted by 1/cystatin C compared to 1/serum creatinine (z=3.12, p=0.002), but equally predicted compared to GFR_MDRD (z=0.92, p=0.356). Serum 1/cystatin C was a strong independent predictor of prognosis (HR: 2.27 per SD increase, 95% CI 1.12 to 4.63), comparable to GFR_MDRD.

Conclusions Cystatin C is an accurate and easy estimate of renal function with prognostic properties superior to serum creatinine and similar to creatinine based formulae in patients with CHF.