In this phase II dose-finding study, 54 patients with CKD stage 5 were randomized to either low, medium or high doses of ND (50, 100 or 200 mg/week for 24 weeks, respectively, in males; doses halved in females), while 7 patients acted as non-randomized controls. The primary outcome measure was appendicular lean mass (ALM) by dual-energy X-ray absorptiometry. Fluid overload (hydration of the fat-free mass) and indicators of physical functioning were secondary measures. Harms were also recorded. Data were analysed using Quade's (1967) non-parametric analysis of covariance.