Background: Cross-sectional studies have shown an inverse correlation between serum C-reactive protein (CRP) and serum albumin concentration in hemodialysis patients. The net effects of inflammation and dietary protein intake on nutritional markers over time are unknown.
Methods: To explore the effects of CRP and normalized protein catabolic rate (nPCR) on serum albumin and creatinine, we analyzed six consecutive months of laboratory data from 364 hemodialysis patients, using a multivariable Mixed model with conservative biases.
Results: The overall trend over time in serum albumin was slightly positive (0.039 g/dL/month) and in serum creatinine slightly negative (-0.052 mg/dL/month). With increasing CRP, serum albumin declined significantly (-0.124 g/dL/month per unit increase in log CRP, adjusted for age, gender, race, diabetes, and nPCR, P < 0.0001). Serum albumin increased with increasing nPCR (0.021 g/dL/month per 0.1 g/kg/day, P < 0.0001). The effect of CRP on albumin was attenuated in African Americans and at a higher nPCR. Corresponding values for creatinine mirrored those for albumin. With increasing CRP, creatinine declined significantly [-0.142 mg/dL/month per unit increase in log CRP, adjusted for age, gender, race, diabetes (time since initiation of dialysis; vintage), Kt/V, and nPCR, P = 0.002]. Serum creatinine increased with increasing nPCR (0.183 mg/dL/month per g/kg/day, P < 0.0001).
Conclusions: Proxies of inflammation and dietary protein intake exert competing effects on serum albumin and creatinine in hemodialysis patients. These data provide a rationale for prospective testing of dietary protein supplementation in hemodialysis patients with biochemical evidence of ongoing inflammation and "malnutrition."