Association between baseline kidney function and change in CRP: an analysis of the cardiovascular health study.

BACKGROUND In cross-sectional analyses, C-reactive protein (CRP) levels are inversely related to levels of kidney function. The relationship between kidney function and subsequent changes in CRP is unknown. METHODS We studied 4,364 individuals from the Cardiovascular Health Study, a longitudinal cohort of community-dwelling older adults. Baseline eGFRcys was estimated using cystatin C. CRP was measured at baseline and after 3 and 7 years of follow-up; slopes of change in CRP were calculated. RESULTS The mean (SD) age of the cohort was 72 (5.2) years; mean (SD) eGFRcys was 78.9 (18.4) ml/min/1.73 m(2). The median (interquartile range IQR) baseline CRP was 2.39 (1.22, 4.33) mg/l; the median (IQR) yearly change in CRP was -0.0051 (-0.020 to 0.27) mg/l/year. After adjustment for demographic characteristics and the initial level of CRP, each standard deviation lower baseline eGFR was associated with a small and non-significant yearly increase in CRP (0.032 mg/l/year; 95% CI: -0.005 to 0.070, p = 0.094). CONCLUSIONS We did not find a relationship between eGFR and subsequent changes in CRP. The association between kidney function and CRP in cross-sectional analyses may reflect unmeasured confounding by atherosclerosis; alternatively, the burden of comorbidity and interval mortality in this population may have masked a stronger longitudinal association between kidney function and change in CRP. Further study in younger populations may clarify whether impaired kidney function leads to change in inflammation over time.