Background: Cardiovascular mortality is extremely high in end-stage renal disease. Cardiovascular mortality risk also is increased in selected (high-risk) individuals with mild to moderate impairment of renal function. It is not clear whether a similar association exists in the general population and, if so, through what mechanisms. We investigated the association of renal function with all-cause and cardiovascular mortality in a population-based cohort and explored potential mechanisms underlying any such relationship.
Methods: An age-, sex-, and glucose-tolerance-stratified sample (N = 631) of a population-based cohort aged 50 to 75 years was followed prospectively. After up to 10.2 years of follow-up, 117 subjects had died (50 of cardiovascular causes). At baseline, renal function was estimated by the serum creatinine level, the Cockcroft-Gault formula and Levey's equation.
Results: At baseline, the mean age was 64 +/- 7 years, 48% were men, 55% had hypertension, and 27% (by design) had type 2 diabetes. Serum creatinine was 91.7 +/- 19.0 micromol/L; creatinine clearance as estimated by the Cockroft-Gault formula was 72.5 +/- 13.7 mL/min/1.73 m(2), and the glomerular filtration rate (GFR) estimated by Levey's equation was 67.8 +/- 12.1 mL/min/1.73 m(2). Renal function was inversely associated with all-cause and with cardiovascular mortality. Relative risks (95% confidence intervals) were 1.08 (1.04 to 1.13) and 1.11 (1.07 to 1.16) per 5 micromol/L increase of serum creatinine; 1.07 (0.98 to 1.17) and 1.15 (1.01 to 1.31) for each decrease of 5 mL/min/1.73 m(2) creatinine clearance; and 1.15 (1.05 to 1.26) and 1.26 (1.12 to 1.42) for each decrease of 5 mL/min/1.73 m(2) of GFR. These associations remained after adjusting for age, sex, glucose tolerance status, hypertension, prior cardiovascular disease, low-density lipoprotein cholesterol, homocysteine, (micro)albuminuria, von Willebrand factor, soluble vascular adhesion molecule-1 and C-reactive protein. Analyses in diabetic and hypertensive subjects gave similar results.
Conclusion: Mild to moderate loss of renal function is strongly associated with an increased risk of cardiovascular mortality. The mechanism behind this association is unclear but does not appear to involve common risk factors such as hypertension, diabetes or hyperhomocysteinemia. Estimation of renal function by relatively simple methods therefore may be a valuable tool for cardiovascular risk assessment over and above that provided by conventional risk factors. Our results were obtained in a general middle-aged to elderly population, and thus have broad applicability.