Degree Name

Doctor of Philosophy


School of Medicine


Chronic kidney disease (CKD) is a major public health problem with significant clinical, societal and psychosocial burdens. Nutrition has been an integral part of medical management in patients with CKD for more than a century; its importance has shifted away from the pre-dialysis stages since the 1970s due to technological advances in renal replacement therapy (RRT), namely dialysis and transplantation. However, nutrition abnormalities can emerge before or during dialysis and continue to be associated with poor outcomes. This suggests a need to revisit nutrition management in end stage kidney disease (ESKD) to gain a broader insight into its effects on clinical outcomes.

The aim of this thesis was to examine the relationships between nutritional factors and clinical outcomes in people with ESKD, addressing the question, “Is nutrition management good enough only when it starts at or near dialysis initiation?”

This thesis was built on a “research in practice” framework. A series of studies was conducted using retrospective data on clinical cohorts attending the renal unit at the St. George Hospital, Sydney.

Study I examined the available data from all attending patients at the initiation of dialysis from 2000 to 2010. This study examined the association between nutritional parameters at the initiation of dialysis and mortality in the clinical cohort (2000-2010, n =167). The hypothesis was that poor nutrition at the start of dialysis predicted high mortality risk. The mean glomerular filtration rate (GFR) at the start of dialysis was 8.0±2.7 mL/min/1.73m2; about half (52.1%) of the patients were rated as malnourished or scored B or C using subjective serum albumin (s-albumin) and malnutrition (SGA score B or C) independently predicted mortality over the 10 year study period (P <0.0001, P P =0.01 and P =0.02 respectively). Overweight and obesity defined as body mass index (BMI) ≥26 kg/m2 did not show any advantage on survival (P =0.73). The combination of malnutrition and overweight/or obesity (SGA B and C + BMI ≥26 kg/m2) was global assessment (SGA). Older age (>65 years), presence of peripheral vascular disease, reduced associated with a three-fold increase in mortality risk with adj. HR 2.96, 95% CI: 1.12–7.33, P =0.02 compared to being well nourished with a BMI /m2 (referent). Being well nourished (SGA = A) was found to be associated with lower mortality risk irrespective of the levels of s-albumin and BMI. Thus, malnutrition at the start of dialysis was found to be an independent predictor of mortality. In addition, no statistical difference in survival was observed between the early and late start groups, which commenced dialysis with a GFR equal and above, or less than, 7 mL/min/1.73m2. Therefore, the results supported the recommendations in the literature that with careful clinical management of ESKD, including nutritional inputs, dialysis can be started at lower levels of GFR. This has tremendous healthcare cost implications.

A multidisciplinary pre-dialysis assessment clinic was established in 2002 following the preliminary analysis of data on the clinical cohort after two years. Study II of the thesis focused on the cohort attending the pre-dialysis clinic. This study was conducted in two parts.

Study IIa was a cross-sectional study of the nutritional characteristics of patients first attending the new multidisciplinary pre-dialysis assessment clinic during the period 2002 to 2008 (n =210). The hypothesis was that a high prevalence of nutritional abnormalities was present before starting dialysis. The mean GFR was 17.3±6.5 mL/min/1.73m2 with 40.5% of patients rated as malnourished (SGA score B or C). Energy and protein intakes correlated positively with GFR, being r =0.17, P r=0.29, P

In Study IIb, the reference data collection period was extended for a further 4 years to enable analyses on baseline data for a ten year period (April 2002 to March 2012, n=501). In the preliminary clinic evaluation eighteen months after it started, patients had low GFR on presentation and the prevalence of malnutrition was high. We hypothesized as the clinic became more established, earlier referral with higher levels of GFR would occur over time and better nutritional status in the first clinic assessment would be achieved. For ease of comparison, the data were divided into two halves, or two 5-year periods, comparing patients referred between April, 2002 and March 2007 (period 1) to those referred between April 2007 and March 2012 (period 2). GFR was 16.7±6.7 vs. 22.1±9.1 mL/min/1.73m2, P

In conclusion, nutrition abnormalities merged during the decline of kidney function before the initiation of dialysis. Nutritional factors, along with older age (>65 years), and co-morbidities at the initiation of dialysis independently predict mortality. To answer the question “Is nutrition management good enough only when it starts at or near dialysis initiation?” the answer is NO. Through research in practice, the results of this thesis suggested structured nutrition management should be implemented well before dialysis is required and even before the pre-dialysis stage to improve health outcomes in patients with ESKD.