Association between markers of renal function and C‐reactive protein level in the elderly: Confounding by functional status

Abstract

Objective. To analyse the cross‐sectional association between measures of renal function and inflammation in an elderly population and to evaluate the confounding effect of impaired physical functioning on these relationships. Material and methods. Cystatin C and creatinine were measured in serum samples from 1110 elderly subjects in a community‐based cross‐sectional survey (Lieto Study) in southwestern Finland. Glomerular filtration rate (GFR) was estimated by means of the Modification of Diet in Renal Disease (MDRD) equation. Associations between renal measures and sensitive C‐reactive protein (CRP) and the impact of functional status were determined by multivariate linear models. Results. Based on standardized coefficients, cystatin C (β 0.19; p<0.001) showed the strongest association with CRP compared to creatinine (β 0.14; p<0.001) and estimated GFR (β −0.13; p<0.001). Levels of CRP linearly increased across quintiles of cystatin C, whereas for creatinine and estimated GFR the increase was less graded. Impaired physical functioning was strongly associated with elevated levels of cystatin C (p<0.001) and CRP (p<0.001), but not with creatinine (p = 0.45) or estimated GFR (p = 0.38). For persons with impaired physical functioning, the odds ratio for belonging to the highest compared to the lowest cystatin C quintile was 7.04 (95% confidence interval 3.49–14.9; p<0.001), whereas for creatinine and estimated GFR this difference was not significant. Conclusion. The weaker association observed between CRP and creatinine‐based measures, as compared to cystatin C, reflects the misclassification of elderly frail subjects as having normal kidney function rather than suggests cystatin C itself to be a marker of inflammation.