SP292EFFECT OF VERY LOW-PROTEIN DIET VERSUS STANDARD LOW-PROTEIN DIET ON RENAL DEATH IN PATIENTS WITH CHRONIC KIDNEY DISEASE: A PRAGMATIC, RANDOMIZED, CONTROLLED, MULTICENTER TRIAL

Abstract

Introduction and Aims: Whether marked dietary protein restriction delays start of dialysis and affects survival in patients with advanced chronic kidney disease (CKD) remains an unsolved issue. We investigated the effect of prolonged very low-protein diet on the risk of renal death in unselected patients stage 4-5 CKD. Methods: We enrolled patients naïve for very low protein diet, without acute kidney injury in the previous three months and followed from at least six months in ten Italian outpatient nephrology clinics. Patients were randomly assigned to very low-protein diet (VLPD, 0.35 g/kg ideal Body wt/day supplemented with ketoacids) or standard low-protein diet (LPD, 0.60 g/kg ideal Body wt/day). Randomization was centralized and stratified by center and CKD stage. We registered start of chronic dialysis (end stage renal disease, ESRD) and all-cause death occurring before and after dialysis start. Primary outcome was time to renal death, defined as the first event between ESRD or patient death. Secondary outcomes were ESRD, overall mortality and compliance to diet. Trial is registered with ClinicalTrials.gov number NCT00323713. Results: We randomized 227 patients and 223 were eligible for the analysis (107 VLPD and 116 LPD). Male gender was 61%, diabetes mellitus 35%, age 64 years, creatinine clearance 31.8 mL/min/1.73 m2 (CKD-4, 65%; CKD-5, 35%), protein intake 0.89 g/kg ideal Body wt/day. Overall median follow-up was 6.7 years without differences between arms; 180 renal deaths (141 dialysis) and 89 all-cause deaths (39 before dialysis and 50 during dialysis) occurred. Time to renal death was similar in VLPD and LPD (Log rank, p=0.28) with an hazard ratio (HR) of 1.17 (VLPD vs LPD, 95%CI 0.88-1.57). No difference was observed for ESRD (Gray's test p=0.51). Also overall mortality did not differ (Log rank, p=0.82) and HR was 0.95 (VLPD vs LPD, 95%CI 0.62-1.45). Mean protein intake reduction (24 mths -baseline, g/kg ideal Body wt/day) was -0.22 in VLPD and -0.02 in LPD (p<0.001, adjusted for baseline). HR of renal death in patients adherent to protein prescription was 0.66 (VLPD vs LPD, 95%CI 0.40-1.09). Conclusions: In unselected population of patients with advanced CKD under nephrology care, very low-protein diet, as compared with standard low-protein diet, does not change the risk of renal death as well as the risk of death before and after the start of dialysis treatment.