Very low-protein diet plus ketoacids in chronic kidney disease and risk of death during end-stage renal disease: an historical, cohort, controlled study.

Abstract BACKGROUND: Very low-protein intake during chronic kidney disease (CKD) improves metabolic disorders and may delay dialysis start without compromising nutritional status, but concerns have been raised on a possible negative effect on survival during dialysis. This study aimed at evaluating whether a very low-protein diet during CKD is associated with a greater risk of death while on dialysis treatment. METHODS: This is an historical, cohort, controlled study, enrolling patients at dialysis start previously treated in a tertiary nephrology clinic with a very low-protein diet supplemented with amino acids and ketoacids (s-VLPD group, n = 184) or without s-VLPD [tertiary nephrology care (TNC) group, n = 334] and unselected patients [control (CON) group, n = 9.092]. The major outcome was survival rate during end-stage renal disease associated to s-VLPD treatment during CKD. The propensity score methods and Cox regression model were used to match groups at the start of dialysis to perform survival analysis and estimate adjusted hazard ratio (HR). RESULTS: In s-VLPD, TNC and CON groups, average age was 67.5, 66.0 and 66.3 years, respectively (P = 0.521) and male prevalence was 55, 55 and 62%, respectively (P = 0.004). Diabetes prevalence differed in the three groups (P < 0.001), being 18, 17 and 31% in s-VLPD, CON and TNC, respectively. A different prevalence of cardiovascular (CV) disease was found (P < 0.001), being similar in TNC and CON (31 and 25%) and higher in s-VLPD (41%). Median follow-up during renal replacement therapy (RRT) was 36, 32 and 36 months in the three groups. Adjusted HR estimated on matched propensity patients was 0.59 (0.45-0.78) for s-VLPD versus CON. Subgroup analysis showed a lower mortality risk in s-VLPD versus matched-CON in younger patients (<70 years) and those without CV disease. No significant difference in HRs was found between s-VLPD and TNC. CONCLUSION: s-VLPD during CKD does not increase mortality in the subsequent RRT period.