!

"

$%%&

Abstract
Malnutrition is present in up to 48% of chronic kidney disease patients on the initiation of renal
1

replacement therapy (dialysis) . At this time, malnutrition is an independent and significant

2

predictor of morbidity and mortality . As a consequence of progressive deterioration in kidney

function, symptoms of decreased appetite and reduced intake are common factors leading to
3

the decline in nutritional status . However, at present there is little evidence to inform nutrition
assessment and intervention for pre-dialysis chronic kidney disease (CKD).

The purpose of this study was to provide evidence for the nutritional management of CKD
patients prior to dialysis with an aim to optimise nutritional status.

To address this, an

investigation comprising of two phases examining nutrition assessment and intervention in a

sample of pre-dialysis Stage IV and V CKD patients was undertaken.

Both phases of the study were conducted through Royal Brisbane and Womens Hospital
(RBWH) Department of Renal Medicine pre-dialysis clinic. Participants met the following criteria:
adult (>18 years) Glomerular Filtration Rate (GFR) <30ml/min CKD, not previously seen by a
dietitian for Stage IV CKD, absence of communication or intellectual impairment inhibiting their
ability to undertake the intervention and an absence of malnutrition from a cause other than
CKD. Phase I was a cross-sectional investigation into the performance of a range of tools
assessing nutrition status, conducted at baseline of Phase II.

Phase II was a randomised-

controlled trial designed to determine if providing individual nutrition counselling with regular
telephone follow-up resulted in improved body composition, nutritional status, dietary intake and
quality of life, compared with standard care.

A range of intermediate, clinical and patient-centred outcome measures were collected at

baseline and twelve weeks. Body composition was measured by total body potassium counting
(TBK), considered a gold-standard measure of body cell mass (BCM, the bodys functional
metabolising tissue). Nutritional status was measured using Subjective Global Assessment
(SGA) and a number of modified versions of SGA, 7-point SGA, Malnutrition Inflammation Score
i

(MIS) and the scored Patient-Generated Subjective Global Assessment (PG-SGA). Dietary
intake was measured using 3-day food records.

Quality of life was measured by Kidney

Disease Quality of Life Short Form version 1.3 (KDQOL-SF

TM

v1.3 RAND University),

4

combining the Short Form-36 (SF-36), with a kidney disease-specific module .

Statistical analysis was carried out using SPSS Version 13 (SPSS Inc, Chicago, IL, USA).
Phase I analysis was based on descriptive and bi-variate statistics, including chi-square, t-test
and ANOVA. For phase II, change variables (Week 12 Week 0) were created for the outcome
measures (BCM, SGA tools, dietary intake (energy and protein) and the 18 KDQOL-SF

TM

sub-

scales). The assessment of change in outcome measures by treatment group was undertaken
by ANCOVA, adjusting for baseline values.

Further multivariate analysis (ANCOVA and

MANCOVA models) were created for outcome variables when confounding variables were
identified and adjusted for.

In Phase I, 56 patients (Male n=34; age mean (SD) 70.7 (14.0); GFRMDRD 22.4 (6.5) mL/min)
underwent baseline assessment. In this population the prevalence of malnutrition was 19.6%
(n=11, SGA B; no C ratings). Malnutrition was associated with lower body cell mass (mean
BCM, 26.3 vs. 33.4 kg p=0.007), body weight (64.8 vs. 76.1 kg p=0.042), BMI (23.7 vs. 27.6
kg/m2 p=0.015) and greater weight loss over previous 6 months (-6.2 vs. -0.1 kg p=0.004).
3.5

Body cell mass indexed for height (BCM-I kg/m ) had a relationship with MIS (r=-0.27 p=0.063)
and scored PG-SGA (r=-0.27 p=0.060), but not with 7-point SGA (F(4) 2.24 p=0.080). PG-SGA
best discriminated malnutrition based on a BCM-I cut-off of <5.25kg/ m

3.5

of all the modified

SGA tools. The scored PG-SGA including the global SGA rating is recommended for use in
pre-dialysis CKD.

In Phase II, 50 patients, (Male n=31 (62.0%); age 69.7 (12.0) years; GFRMDRD 22.1 (6.9)
ml/min) completed the 12 week study period (intervention n=24; standard care n=26). At 12
weeks, there was a clinically significant improvement in all outcome measures in the
intervention group. There was a 3.9% (95% CI, -1.0 to 8.7%) mean difference in change for
Body Cell Mass between the treatment groups, represented by a significant decrease in the

ii

standard care group and maintenance in the intervention group. Nutritional status measured by
SGA improved or was maintained (24/24) in the intervention group, however, decreased in 14%
(4/26) of the standard care group. Energy intake significantly improved in the intervention group
resulting in a mean difference in change of 17.7kJ/kg (8.2 to 27.2 kJ/kg). Quality of life improved
significantly in 10 of the 18 sub-scales in the intervention group. Significant effect modification
for gender was apparent for many of the outcome variables, with females responding most
significantly to the intervention treatment. This study concluded that, overall, structured nutrition
intervention limits the deterioration in nutritional status, improves dietary intake and quality of life
in patients with CKD prior to the onset of renal replacement therapy.

This thesis makes a significant contribution to the evidence base for nutritional management of
pre-dialysis Stage IV CKD. The use of SGA for nutrition assessment and including PG-SGA to
measure change is recommended for routine nutrition assessment of pre-dialysis CKD. The
provision of individual nutrition counselling with regular follow-up, with a focus on promoting
intake provides beneficial patient outcomes supporting optimal nutritional status in pre-dialysis
CKD patients.

iii

Publications by the candidate

Referred journal articles
Campbell KL, Ash S, Bauer J, Davies PSW Evaluation of nutrition assessment tools compared
with body cell mass for the assessment of malnutrition in chronic kidney disease Journal of
Renal Nutrition 17:189-195, 2007

Campbell KL, Ash S, Bauer J, Davies PSW (2007) Critical review of nutrition assessment tools
to measure malnutrition in chronic kidney disease Nutrition and Dietetics Vol 64 (1): p 23-30

Ash S, Campbell K, MacLaughlin H et al (2006) Evidence-based practice guidelines for the

nutritional management of chronic kidney disease Nutrition and Dietetics Vol 63 Suppl 2 S35-45

Conference presentations
Refereed conference presentations from results relating to this thesis
Campbell KL, Ash S, Bauer J, Davies PSW (2007) Randomised-controlled trial of stuctured
th

nutrition intervention in chronic kidney disease, 25 Dietitians Association of Australia National

Conference, Oral presentation

Hays-St Clair S, Campbell KL, Ash S, Bauer J (2007) Long-term outcomes following a
th

randomised-controlled trial of nutrition intervention in chronic kidney disease, 25 Dietitians

Association of Australia National Conference, Oral presentation

Campbell KL, Ash S, Bauer J, Davies PSW (2006) Performance of assessment tools to
measure nutrition status in pre-dialysis 13th International Congress of Nutrition and Metabolism
in Renal Disease, Merida Mexico, Oral presentation

Campbell KL, Ash S, Bauer J, Davies PSW (2006) Estimation of body cell mass in pre-dialysis agreement between two prediction equations and total body potassium 13th International
Congress of Nutrition and Metabolism in Renal Disease, Merida Mexico, Poster presentation

Campbell KL, Ash S, Bauer J, Davies PSW (2006) Rating Methods of Subjective Global
Assessment in Chronic Kidney Disease Nutrition and Dietetics Vol 63 Suppl 1 A28-29

Campbell KL, Ash S, Bauer J, Davies PSW (2005) Assessment of nutrition status - how do
various tools compare for chronic kidney disease patients? New Zealand Dietitians Association
Annual Conference, Auckland New Zealand, Oral presentation

iv

Invited Presentation
Campbell KL, Hulcombe J (2005) Getting Evidence-Based Practice Guidelines Published New
Zealand Dietitians Association Annual Conference, Auckland New Zealand

Workshop Facilitation
Ash S, Campbell KL, Meade A, Fry K (2007). Review of Chronic Kidney Disease guidelines and
th

development of guidelines for transplant, 25 Dietitians Association of Australia National

Conference, Hobart

McCoy E, Campbell KL, Chan M (2004) Implementation of Evidence Based Practice Guidelines
in CKD 22nd Dietitians Association of Australia National Conference, Melbourne

Other conference presentations during candidature

Campbell KL, Gillen L, Smith R, Banks M, Read S (2007). Attention to nutritional status may
th

help optimise rehabilitation after stroke, 25 Dietitians Association of Australia National

Conference, Poster presentation

Juffs P, Gill E, Campbell KL, Gillen L, Ash S, Healy H (2007). Does implementing best practice
th

guidelines improve nutritional status in the dialysis populaiton? 25 Dietitians Association of

Australia National Conference, Poster presentation

Campbell KL, Ash S, Banks M, Healy H (2006) Inadequate protein intake increases all-cause
mortality in haemodialysis patients Renal Week 2006, American Society of Nephrology, San
Diago, United States, Poster presentation

Ash S, Campbell KL, Healy H (2006) Are we measuring the right outcomes for indicators of
improved nutrition status in those receiving renal replacement therapy? Nutrition and Dietetics
Vol 63 Suppl 1 A25

Chalmers C, Neville J, Hulcombe J, Campbell KL (2006) How effective is increased dietetic

intervention for women with gestational diabetes? Nutrition and Dietetics Vol 63 Suppl 1 A30

Smith R, Bell J, Campbell KL (2006) Are multidisciplinary teams effective in maintaining

nutritional status post-stroke? Nutrition and Dietetics Vol 63 Suppl 1 A30

Campbell KL, Ash S, Hulcombe J, Healy H. (2005) Prospective audit of nutrition status in
haemodialysis patients Nephrology Vol 10 Supp 3 A403

Statement of Original Authorship

The work contained in this thesis has not been previously submitted for a degree or diploma at
any other higher education institution. To the best of my knowledge and belief, the thesis
contains no material previously published or written by another person except where due
reference is made.

Signed: ______________________

Date: ____________________

vi

Acknowledgements
Firstly, I would like to extend my sincere thanks to my supervisors, Associate Professor Susan
Ash, Associate Professor Judy Bauer and Associate Professor Peter Davies for their guidance
and support over the past three years. To Sue, you inspired my interest in clinical nutrition
research, provided me with the opportunity to do this research, and (somehow) led me down the
path of a PhD. I cant thank you enough for your enthusiasm and belief in this project. To Judy,
who always provided the right advice at the right time, always positive, encouraging and gave
me confidence in my ability. To Peter, thank you for your timely advice on body composition
assessment, provision of your lab for testing and for your encouragement to keep on time
through the PhD milestones.

To the staff at Royal Brisbane and Womens Hospital, especially Ms Julie Hulcombe and Dr
Helen Healy, without your leadership and support for nutrition research this project would not
have been realized. To all of the dietitians at Royal Brisbane, for your collegiality, and unending
support for my project, I will always look back at my time in the Department with fond memories.
To the medical, nursing and administration staff in the Renal Medicine Department, for providing
a supportive environment in which I could conduct this research and assisting with patient
recruitment and my professional development. Also, to the lab staff at Childrens Nutrition
Research Centre, especially Zoe Lawrie who often went the extra mile for my study participants
and provided me with much comic relief and encouragement.

To my past and present colleagues at Queensland University of Technology, especially those in

A-wing, and throughout the final stages, in IHBI. The ongoing support, friendship, coffee-breaks
and general banter has kept me sane over the past three years. Id like to especially thank Kate
Halton, Kimberley Hinze and Liz Isenring, who, particularly throughout the final stages provided
direction, friendship and plenty of caffeine.

There are a number of people, who despite having limited concept of my research, have
contributed to this journey beyond expectation. A special thanks to my family, Mum, Dad, Sean,

vii

Joanne, Andrew, Chris and Greg for their encouragement in all of my life pursuits. To all of my
wonderful friends for the humour and support through our many coffee/dinner/dessert outings
and keeping my social calendar full.

To my wonderful partner Aaron. Thank you for being the person you are and for taking this
journey with me. Your support gave me the strength to realise my potential. You are a true
inspiration and your motivation in life continues to astound me.

Finally, this study was made possible by the funding support from Queensland University of
Technology, and the Royal Brisbane and Womens Hospital Research Foundation. A special
thanks is extended to all of the participants in this study, their families and carers, without whom
this study would not have been possible.

viii

Table of Contents
Abstract

................................................................................................... i

Publications by the candidate.................................................................... iv

Statement of Original Authorship .............................................................. vi
Acknowledgements.................................................................................... vii
Table of Contents ........................................................................................ ix
List of Tables ............................................................................................. xiii
List of Figures............................................................................................ xvi
List of Acronyms ..................................................................................... xviii
Chapter 1:

Review of the literature ........................................................ 20

1.0

Introduction............................................................................................................................ 20

1.1

Nutrition concerns in chronic kidney disease (CKD)............................................................. 20

1.1.1

Signs and symptoms of CKD......................................................................................... 20

1.1.2

Mechanisms for the development of malnutrition in CKD ........................................... 21

1.1.3

Prevalence and consequence of malnutrition in CKD.................................................. 24

1.1.4

Summary malnutrition in CKD..................................................................................... 27

1.2

Framework for nutritional management in CKD ................................................................... 27

1.3

Assessment of nutritional status in CKD ............................................................................... 29

1.3.1

Biochemistry ................................................................................................................. 30

1.3.2

Body composition definition ......................................................................................... 31

1.3.3

Anthropometry.............................................................................................................. 33

1.3.4

Clinical assessment tools ............................................................................................. 38

1.3.5

Dietary intake assessment............................................................................................. 44

1.4

Quality of life (QOL) assessment in CKD ............................................................................. 46

1.4.1

Renal function and quality of life.................................................................................. 47

1.4.2

Nutritional status and QOL .......................................................................................... 48

1.4.3

Summary QOL .............................................................................................................. 49

1.5

Nutrition intervention in CKD ............................................................................................... 49

1.5.1

Dietary prescription studies in CKD ............................................................................ 50

1.5.2

Dietary implementation studies in CKD ....................................................................... 56

1.6

Summary of the evidence....................................................................................................... 60

ix

Chapter 2

Study Design and Methods .................................................. 62

2.0

Introduction............................................................................................................................ 62

2.1

Aims of the study ................................................................................................................... 62

2.1.1

Objectives and hypotheses............................................................................................ 62

2.2

Framework informing the study design.................................................................................. 64

2.3

Outcome measures ................................................................................................................. 65

2.3.1

Primary Outcome: Nutritional status ........................................................................... 65

2.3.2

Intermediate outcome: Dietary intake ......................................................................... 67

2.3.3

Exploratory outcomes: Clinical variables.................................................................... 68

2.3.4

Secondary outcome: Quality of life .............................................................................. 69

2.4

Research design and methodology......................................................................................... 71

2.4.1

Study setting.................................................................................................................. 72

2.4.2

Target population ......................................................................................................... 72

2.4.3

Sample size considerations ........................................................................................... 72

2.4.4

Study Sample ................................................................................................................ 73

2.4.5

Recruitment methods .................................................................................................... 73

2.4.6

Data collection ............................................................................................................. 74

2.4.7

Phase I: Nutrition assessment methods ........................................................................ 75

2.4.8

Phase II: Randomised-controlled trial of nutrition intervention .................................. 76

2.5

Statistical analysis .................................................................................................................. 82

2.5.1

Phase 1: Nutrition assessment methods........................................................................ 82

2.5.2

Phase II: Randomised-control trial .............................................................................. 83

2.5.3

Generalisability ............................................................................................................ 86

2.6

Ethical considerations ............................................................................................................ 86

Chapter 3:

Comparison of nutrition assessment methods in pre-dialysis

CKD

88

3.0. Introduction............................................................................................................................ 88
3.1

Results: Subjective Global Assessment and Body Cell Mass ................................................ 89

3.1.1

Clinical parameters compared to subjective global assessment................................... 89

3.1.2

Alternative subjective assessment tools as measures of nutritional status ................... 90

3.2

Diagnostic precision of the modified SGA tools.................................................................... 94

3.3

Discussion .............................................................................................................................. 95

3.4

Summary of nutrition assessment measures........................................................................... 99

Chapter 4:

Results of randomised-controlled trial of nutrition

intervention in pre-dialysis CKD on nutritional status.......................... 100

4.1

Characteristics of the study participants............................................................................... 100

4.1.1

Characteristics of participants allocated to each treatment group ............................ 102

4.1.2

Characteristics of participants with missing outcome data........................................ 103

4.2. Primary Outcome: Nutrition status results ........................................................................... 104

4.2.1

Body Cell Mass........................................................................................................... 105

4.2.2

Subjective Global Assessment..................................................................................... 106

4.3

Dietary intake results............................................................................................................ 108

4.3.1

Total energy and protein intake.................................................................................. 108

4.3.2

Proportion of participants meeting energy and protein intake guidelines ................. 110

4.3.3

Macronutrient contribution to energy intake.............................................................. 111

4.4

Change in clinical variables ................................................................................................. 112

4.5

Adjusted estimates of the impact of structured nutrition intervention on nutrition-related outcome

measures ........................................................................................................................................ 115

4.5.1
4.6

Multivariable analysis for nutrition status outcomes ................................................. 115

Gender difference in the impact of structured nutrition intervention ................................... 119

Chapter 5:

Discussion of impact of randomised-controlled trial of

nutrition intervention in pre-dialysis CKD on nutritional status .......... 122

5.0. Introduction.......................................................................................................................... 122
5.1

Nutrition status as a result of intervention............................................................................ 122

5.1.1.

Change in body cell mass ........................................................................................... 122

5.1.2

Change in Subjective Global Assessment ................................................................... 126

5.1.4.

Dietary intake ............................................................................................................. 127

5.2

Change in clinical measures................................................................................................ 132

5.3

Use of telephone intervention .............................................................................................. 134

5.4

Gender differences in response to intervention .................................................................... 134

5.4.1

Nutritional status and body composition .................................................................... 135

5.4.2

The use of self management principles ....................................................................... 136

5.4

Summary of the impact of nutrition intervention in pre-dialysis CKD ................................ 137

Chapter 6:

Quality of Life Results: Relationship with nutrition status and

response to nutrition intervention in pre-dialysis CKD patients.......... 139

6.0

Introduction.......................................................................................................................... 139

6.1

Quality of life and nutritional status..................................................................................... 139

6.1.1

Comparison of pre-dialysis patients to dialysis patients by nutritional status ........... 139

6.1.2

Relationship between Quality of life sub-scales and nutritional status ..................... 140

6.2

Change in Quality of Life following randomised-controlled trial of nutrition intervention. 141

Chapter 7: Quality of Life Discussion: relationship to nutritional status

and impact of nutrition intervention in pre-dialysis CKD patients....... 146
7.1

Quality of life in pre-dialysis compared to dialysis populations .......................................... 146

7.2

Quality of life and nutritional status in pre-dialysis CKD patients ...................................... 147

xi

7.2.1

Kidney-disease specific QOL, relationship to nutritional status and response to nutritional

intervention ............................................................................................................................... 148

7.2.2

Generic Health Related QOL, relationship to nutritional status and response to nutritional

intervention ............................................................................................................................... 150

7.3

Sample considerations and statistical power ........................................................................ 153

7.4

Summary of quality of life in relation to nutritional status and the impact of nutrition

intervention.................................................................................................................................... 154

Chapter 8

Conclusions and recommendations ................................. 156

8.0

Introduction.......................................................................................................................... 156

8.1

Addressing study aims and hypotheses ............................................................................... 156

8.2

Study strengths and limitations ............................................................................................ 160

8.2.1

Phase I ........................................................................................................................ 160

8.2.2

Phase II....................................................................................................................... 160

8.2.3

Considerations of the limitations of this study............................................................ 164

8.3

Recommendations for practice............................................................................................. 166

8.3.1

Criteria for referral to dietitian .................................................................................. 167

8.3.2

Nutrition assessment................................................................................................... 167

8.3.3

Nutrition prescription / intervention........................................................................... 168

8.3.4

Implementation and management............................................................................... 169

8.4

Recommendations for future research.................................................................................. 170

9.0

References........................................................................... 172

10.0

Appendices.......................................................................... 202

Appendix A

Manuscripts and conference presentations related to the thesis

Appendix B

Evidence Based Practice Guidelines for the Nutritional Management of CKD

Appendix C

Development of Body Cell Mass Index

Appendix D

Ethics clearance and patient information package

Appendix E

Data collection and nutrition assessment tools

Appendix F

Intervention support materials

xii

List of Tables
Table 1-1
Kidney Disease Outcomes Quality Initiative (K/DOQI) defined stages of chronic kidney
8
disease (CKD) 21
Table 1-2

Mechanisms for the progression of uraemic malnutrition in CKD ................ 22

Table 1-3
Prevalence of poor nutritional status and outcome from prospective observational
studies in Stage V CKD .......................................................................................................... 25
Table 1-4
51
evidence

National Health and Medical Research Council (NHMRC) classification for levels of
28

Table 1-5
Evidence-based recommendations and corresponding level of evidence for nutritional
management of Stage IV CKD (Adapted from Ash et al 2006).............................................. 28
Table 1-6

Comparison of commonly used body composition methods in CKD ........... 33

Table 1-7
Definition and study design for validity and reliability testing corresponding NHMRC
level of evidence in reference to nutrition assessment tools in CKD ..................................... 39
Table 1-8

Validation studies demonstrating the clinical validity of subjective global assessment

...................................................................................................................... 40

Table 1-9
Description, advantages and disadvantages of SGA-based nutrition assessment tools
utilised in CKD. 41
Table 1-10

Validation studies for SGA-derived tools in CKD, Stage IV and V............... 43

Table 1-11
Common method of participant-reported dietary intake; assessment methods and
considerations 45
Table 1-12
Prospective observational studies investigating the impact of quality of life on clinical
outcome in chronic kidney disease patients........................................................................... 47
Table 1-13
Studies demonstrating the relationship between nutritional status and quality of life in
chronic kidney disease patients ............................................................................................. 49
Table 1-14
Metanalysis investigating outcomes of protein restriction studies in pre-dialysis chronic
kidney disease 51
Table 1-15
Protein restriction studies in pre-dialysis chronic kidney disease investigating change
in nutritional status.................................................................................................................. 53
Table 1-16
Randomised-controlled trials and controlled-clinical trials evaluating the
implementation of nutritional care in chronic kidney disease ................................................. 57
Table 2-1
Steps in MNT cascade and corresponding clinical question for the nutrition
46
management of CKD patients (adapted from Ash et al )...................................................... 65
Table 2-2Extract of the Table of Contents for the Nutrition in Chronic Kidney Disease education
booklets for the intervention and standard care groups ......................................................... 78
Table 3-1
Characteristics of 56 pre-dialysis chronic kidney disease patients based on nutritional
status by Subjective Global Assessment ............................................................................... 89

xiii

Table 3-2
Co-ordinates of the ROC curve showing the sensitivity of BCM-I against 1-specificity
for various cut-off values to determine nutritional status as identified by SGA. ..................... 90
Table 3-3
Index (n=50)

Nutrition assessment tools in pre-dialysis CKD patients against SGA (n=56) and BCM
...................................................................................................................... 92

Table 4-1
Comparison of characteristics of the study population (n=56), non-participants (n=8),
and reference population from Queensland Health Collaborative (n=605) of pre-dialysis CKD patients
.................................................................................................................... 101
Table 4-2
Baseline characteristics (mean SD) of patients with pre-dialysis CKD randomised to
intervention (n=29) or standard care (n=27) treatment ........................................................ 102
Table 4-3
Primary identified cause of CKD for 56 subjects randomised to receive intervention
(n=29) or standard care (n=27) in a sample of pre-dialysis CKD patients ........................... 103
Table 4-4
Baseline characteristics (mean SD) of study participants who completed 12-week
study period in pre-dialysis CKD (n=56)............................................................................... 104
Table 4-5
Mean difference (95% CI) of change in body cell mass (BCM) over the 12-week
treatment period in pre-dialysis CKD patients for intervention (n=21) and standard care (n=20) groups
.................................................................................................................... 105
Table 4-6
Change in Subjective Global Assessment ratings for intervention (n=24) and standard
care (n=26) groups n(%) over the 12-week treatment period, in a sample of pre-dialysis CKD patients
106
Table 4-7
Change in Subjective Global Assessment ratings and relative (%) change in body cell
mass, weight, and change in PG-SGA score, mean (95%CI), over the 12 week treatment period,
independent of treatment group, in a sample of pre-dialysis CKD patients ......................... 107
Table 4-8
Values for PG-SGA score at week 0 and week 12, between the intervention (n=24)
and standard care (n=26) groups, in a sample of pre-dialysis CKD patients....................... 107
Table 4-9
Energy and protein intake as total (kJ or g) and relative to ideal body weight (kJ/kg or
g/kg) for each treatment group at week 0 and week 12 and the mean difference in change between
the intervention (n=24) and standard care (n=24) groups................................................... 108
Table 4-10
Proportion of participants considered to be under-reporting energy intake (EI/BEE
<1.3) at baseline and week 12, between intervention and standard care groups ............... 109
Table 4-11
Ratio between energy intake (kJ) and estimated basal energy expenditure (mean
SD), participants who were weight stable at both baseline and week 12 from intervention (n=15) and
standard care (n=10) ............................................................................................................ 109
Table 4-12
Proportion of subjects in the intervention (n=24) and standard care (n=24) group
46
meeting the recommended level of protein and energy intake at week 0 and week 12.... 110
Table 4-13
Macronutrient contribution to energy intake % (mean SD) for participants in
intervention (n=24) and standard care (n=24) groups.......................................................... 111
46, 47

Table 4-14
Proportion of participants by treatment group compared with guidelines
at
baseline and week 12 between intervention (n=24) and standard care (n=26) treatment groups in a
sample of pre-dialysis CKD patients .................................................................................... 113
Table 4-15
Change in clinical variables by intervention (n=24) and standard care (n=26) groups
during 12-week randomised controlled intervention in pre-dialysis CKD............................. 113

xiv

Table 4-16
Impact of nutrition intervention on nutrition-related outcomes, difference in mean
change between treatment groups (standard care from intervention), both adjusted and unadjusted
values
.................................................................................................................... 117
Table 4.17
Baseline characteristics of participants completing study by gender (n=50) in a sample
of pre-dialysis CKD patients ................................................................................................. 119
Table 4-18
Outcome variables of the intervention study split for gender for week 0 and week 12
(mean (SE)) and mean difference in change (mean (95%CI)) in intervention (n=24) and standard care
(n=26) groups 120
Table 6-1
Kidney disease-specific and short-form 36 quality of life components, compared with
SGA rating and PG-SGA score (n=53) ................................................................................ 141
Table 6-2
Quality of life components for Week 0 and 12 and mean difference in change
(standard care from intervention) ......................................................................................... 143
Table 6-3
Mean difference in change (week 12 week 0; (mean (95%CI))) in quality of life subscales by treatment group split for gender for week 0 and week 12 .................................... 145

xv

List of Figures
Figure 1-1
(CKD)

Mechanisms for the development of uraemic malnutrition in chronic kidney disease

...................................................................................................................... 23

Figure 2-1
Cascade of events leading to evidence of effectiveness of Medical Nutrition Therapy
220
(MNT) (Splett, 1996) ........................................................................................................... 64
Figure 2-2
CKD

Flowchart of participant progression in RCT of nutrition intervention in pre-dialysis

...................................................................................................................... 77

Figure 2-3
Summarised protocol used in this study for the intervention of pre-dialysis CKD
patients adapted from ADA intervention protocol................................................................... 79
Figure 2-4
patients

Example goals, objects and strategies used in the intervention of pre-dialysis CKD
...................................................................................................................... 80

Figure 3-1
Receiver operating characteristic (ROC) curve showing the sensitivity of BCM-I
against 1-specificity for various cut-off values to determine nutritional status as identified by SGA. 90
Figure 3-2
Distribution of well-nourished (SGA A) and malnourished (SGA B), within the ratings
of the 7-point SGA in 56 pre-dialysis CKD patients ............................................................... 91
Figure 3-3

Mean rating of 7-point SGA against BCM Index in CKD patients, n=50. ..... 92
3.5

Figure 3-4
Scatter plot of the BCM index (kg/m ) between the Malnutrition Inflammation Score
(left) and the Patient generated SGA (right), regression line representing slope of the correlation coefficient
...................................................................................................................... 93
Figure 3-5
Scatter plot of the mean difference between the PG-SGA and the MIS scores, against
the average between the scored assessments with limits of agreement of 2 SD from the mean
difference
...................................................................................................................... 93
Figure 3-6
ROC curve showing the sensitivity against 1-specificity for ratings of 7-point SGA
3.5
(inverted), scored PG-SGA and MIS to determine BCM-I <5.25 kg/m . .............................. 94
Figure 4-1
Proportion of change in BCM over the 12-week treatment period for pre-dialysis
chronic kidney disease patients receiving either nutrition intervention (n=20) or standard care (n=21)
.................................................................................................................... 106
Figure 4-2
Bland-Altman plot comparing methods of change in body weight, body cell mass
change compared with weight change over 12 week treatment period ............................... 115
Figure 4-3
Change in energy intake and body cell mass over the intervention period, by
treatment group split for gender ........................................................................................... 121
Figure 6-1a Mean score for general health (SF-36) sub-scales between the US DOPPS population
151
and the current study; US DOPPS n= 2,885 Well-nourished n=43 ; Malnourished n=10
140
Figure 6-1b Mean score for kidney disease specific sub-scales between the US DOPPS study and
151
the current study; US DOPPS n= 2,885 Well-nourished n=43 ; Malnourished n=10 140
Figure 6-2
Change in kidney disease specific and general (SF-36) quality of life subscales as a
result of receiving structured intervention (n=23) ; or standard care treatment (n=22).... 144
Figure 8-1

Cascade of events leading to evidence of effectiveness of Medical Nutrition Therapy

.................................................................................................................... 162

xvi

Figure 8-2
Evidence based practice guideline framework and the clinical questions related to the
nutrition care process for the nutritional management of chronic kidney disease ............... 166

xvii

List of Acronyms
APPR

Acute-Phase Protein Response

BCM

Body Cell Mass

BF

Body Fat

BIA

Bio-Impendence Analysis

BMI

Body Mass Index

CARI

Caring for Australians with Renal Impairment

CKD

Chronic Kidney Disease

CRP

C-Reactive Protein

DXA

Duel X-ray Absorptometery

ECW

Extra-Cellular Water

FFM

Fat-Free Mass

FM

Fat Mass

GFR

Glomerular Filtration Rate

HD

Haemodialysis

HR

Hazard Ratio

IBW

Ideal Body Weight

ICW

Intra-Cellular Water

K/DOQI

Kidney Disease Outcomes Quality Initiative

40

Isotope Potassium 40

KDQOL

Kidney Disease Quality of Life

kg

Kilogram

kJ

Kilojoules

Litre

LBM

Lean Body Mass

MAMC

Mid-Arm Muscle Circumference

MCS

Mental Health Component Score (Quality of Life)

MDC

Minimal Detectable Change

MDRD

Modified Diet in Renal Disease

MIS

Malnutrition Inflammation Score

NI

Nitrogen Index

xviii

n-PCR

Normalised Protein Catabolic Rate

OR

Odds Ratio

PCS

Physical Health Component Score (Quality of Life)

PD

Peritoneal Dialysis

PG-SGA

Patient-Generated Subjective Global Assessment

QOL

Quality of Life

QUT

Queensland University of Technology

RBWH

Royal Brisbane and Womens Hospital

RCT

Randomised Controlled Trial

REE

Resting Energy Expenditure

RR

Relative Risk

SGA

Subjective Global Assessment

TBK

Total Body Potassium

TBN

Total Body Nitrogen

TBW

Total Body Water

TSF

Tri-cep skin fold

xix

Chapter 1:
1.0

Review of the literature

Introduction

Chronic Kidney Disease (CKD) is a significant and increasingly prevalent clinical condition, considered
an epidemic in developed nations, including Australia and the United States

5, 6

. Compromised

nutritional status as a consequence of the disease is considered to be one of the major treatment
priorities for patients with established CKD. However, the aetiology and subsequent treatment of poor
nutritional status in CKD is complex.

Of particular concern is how best to treat patients with

decreasing kidney function prior to dialysis. Current treatment strategies and practice guidelines are
based on either evidence from dialysis populations or from inconclusive studies in pre-dialysis. The
following chapter will consider the literature regarding the nutritional management in chronic kidney
disease, to identify significant gaps in the evidence and to inform the subsequent study design
featured in this thesis.

1.1

Nutrition concerns in chronic kidney disease (CKD)

1.1.1 Signs and symptoms of CKD

The basic functions of the kidney are to regulate fluid, electrolyte and hormone balance, and to
7

facilitate waste product excretion from protein metabolism . With progressive decline in kidney
function, hallmarked by a sustained decrease in glomerular filtration rate (GFR), the kidneys ability to
perform these functions becomes progressively impaired. In particular as GFR decreases, solutes
7

usually excreted by the kidney accumulate in the body and blood (plasma) concentrations increase .
These solutes include urea and creatinine from protein metabolism. As the basic functions of the
kidney are progressively impaired, the incidence of renal-related conditions such as uraemic toxicity
increases with decreasing GFR. With advanced deterioration in kidney function, symptoms of uraemic
toxicity are primary indicators to commence renal replacement therapy.

The diagnosis and management of CKD is defined in five stages by the National Kidney Foundation
8

Kidney Disease Outcomes Quality Initiative (NKF-K/DOQI) clinical practice guidelines (Table 1-1). In
Australia, CKD is a highly prevalent disease with one in three Australians at risk of developing CKD
9

and one in seven over the age of 25 years have at least one clinical sign of existing CKD . However,
20

until pre-dialysis CKD is reached (GFR < 30ml/min or Stage 4 CKD), the presence of CKD is often not
9

associated with significant symptoms and can go unrecognised in 80-90% of cases .

Table 1-1 Kidney Disease Outcomes Quality Initiative (K/DOQI) defined stages of chronic
kidney disease (CKD)

This table is not available online.

Please consult the hardcopy thesis
available from the QUT Library

The stages defined as pre-dialysis CKD are associated with a 3.2- and 5.9-fold increased risk of
10

death associated with Stage 4 (GFR 15-29ml/min) and Stage 5 (<15 ml/min) respectively . A patient
progressing to Stage 4 CKD has a kidney function less than 30% of normal, and is expected to require
renal replacement therapy (RRT, in the form of transplant, or more commonly, dialysis) within the next
11

6 to 18 months . At this stage, the focus of the medical treatment shifts from slowing the decline of
12

renal failure to managing the metabolic disturbances and preparing the patients for RRT . Ideally,
this requires the provision of multidisciplinary care to prevent complications (such as anaemia,
malnutrition and acidosis), treat co-morbidities (including cardiovascular disease and diabetes) and
manage symptoms (including nausea, hypertension and fluid balance)

11, 12

1.1.2 Mechanisms for the development of malnutrition in CKD

Nutritional and metabolic derangements are common in CKD and play a major role in affecting clinical
13

outcome in this patient population . The development of impaired nutritional status, in Stage 4 and 5
CKD, commonly termed uraemic malnutrition, is multi-factorial. The term uraemic malnutrition is used
to encapsulate the dietary and metabolic changes which lead to a state of protein catabolism and loss
of lean body mass. Malnutrition is typically hallmarked by a progressive decline in body weight and
14

body protein . It is the loss of metabolically active component of body mass which creates the
negative effects seen with malnutrition

15, 16

Figure 1-1 provides a schematic diagram highlighting the complex interrelationships of factors leading
to uraemic malnutrition in CKD. Table 1-2 provides an overview of the metabolic consequences of a
number of conditions which may contribute to the development of uraemic malnutrition in CKD.

21

Table 1-2 Mechanisms for the progression of uraemic malnutrition in CKD

Cause
Reduced dietary
intake
Restrictive
dietary intake
Inflammation

Mechanism
Poor appetite from uraemic toxicity, inflammation, reduction in anabolic
17
18
hormones . Spontaneous decline in energy and protein intake
Unmonitored restriction of protein and other nutrients
From whole-body protein catabolism, suppressed albumin synthesis,
3
anorexia, disrupted growth hormone and IGF-1 (decreased anabolism) ,
19
increased resting energy expenditure .
20
Delayed gastric emptying, impaired motility, alterations in bacterial flora

Gastrointestinal
function
Metabolic
acidosis

From stimulation of ubiquitine-mediated proteolysis, resulting in a

negative nitrogen balance, catabolism of branched-chain amino-acids,
21
muscle protein and suppress albumin synthesis
Altered carbohydrate and lipid metabolism, insulin resistance and
22
metabolic acidosis

Abnormal
hormone
response

As evident in Figure 1-1, the development of anorexia, resulting in a reduction of dietary intake is a
14

central factor to compromising nutritional status in CKD . Poor appetite, or anorexia may be induced
by a range of factors, particularly a high plasma concentration of uraemic toxins, metabolic acidosis,
hormone abnormalities and presence of pro-inflammatory cytokines

23

(Figure 1-1).

unmonitored dietary restrictions may contribute to overall decrease in intake.

In addition,

In the absence of

nutritional intervention, there is good evidence that patient with CKD have a spontaneous reduction in
18, 24, 25

energy and protein intake with decreasing renal function

. Particularly compromised is energy

24

intake when renal function deteriorates below 25% (GFR <25 ml/min) . Changes in energy and
protein intake in pre-dialysis CKD have been correlated with renal function and with reductions in
anthropometric nutritional parameters in cross-sectional studies

24, 25

In addition to the decrease in substrate availability from decreased dietary intake, nutritional status
may be further compromised by other mechanisms of metabolic changes as a result of decreased
renal function. The accumulation of hydrogen ions (leading to metabolic acidosis) marked by low
serum bicarbonate and the accumulation of pro-inflammatory cytokines, marked by an increase in
production of C-reactive protein (CRP), both act directly to stimulate protein catabolism and suppress
26

protein synthesis . Abnormal metabolic response includes a reduction in the activity of anabolic
hormones, such as insulin and growth hormone, and an increase in cortisol and glucagon also
22

contribute to a reduced capacity for protein synthesis .

22

Reduced physical
function

Stimulation of protein
degradation
(ubiqutin-protesome
pathway)

Reduced
dietary
intake

Inflammation
Cytokines (TNF- , IL-1)
Resulting in acute phase response
(CRP, albumin)

23

Figure 1-1 Mechanisms for the development of uraemic malnutrition in chronic kidney disease (CKD)

Metabolic acidosis,
uraemic toxins

Anorexia

Nausea, taste changes,

fluid/electrolyte overload,
restrictive dietary intake

Appetite

Abnormal hormone response:

Insulin, IGH-1, growth hormone,
PTH, leptin, glucagon, cortisol

Abnormal lipid and

carbohydrate
metabolism

Uraemic
Malnutrition

urine
concentrating
capacity and
sodium balance

Therefore, due to the metabolic and dietary intake changes resulting from progressive decline in renal
function, nutritional status becomes compromised in pre-dialysis CKD, prior to commencement of
renal replacement therapy and appears to improve with long-term dialysis treatment.

1.1.3 Prevalence and consequence of malnutrition in CKD

There is a significant body of evidence evaluating the prevalence and consequence of malnutrition in
CKD. Multiple epidemiologic studies have confirmed the close association between nutritional status
by commonly available nutritional markers and clinical outcomes are confined to stage 5 CKD, either
commencing dialysis or maintenance dialysis patients. Table 1-3 provides prevalence estimates from
a variety of tools for prospective observational studies in commencing or maintenance dialysis
patients. From this table it is clear that not only is malnutrition significantly prevalent, but also related
to poor clinical outcome.

Prevalence and impact of poor nutritional status varies between studies and particularly varies
depending on the method of assessing nutritional status. The most significant results of poor outcome
with malnutrition are evident from studies using albumin as the assessment standard. However,
basing studies on albumin levels may be representing overall health status as opposed to a direct
measure of nutritional status in CKD. The limitations of albumin as a nutrition status measure will be
detailed further in Section 1.3 Assessment of nutritional status in CKD.

The prevalence of malnutrition as assessed using Subjective Global Assessment (SGA) close to the
start of RRT is 39 48%

1, 2

. Studies assessing nutritional status in stage IV CKD are limited. What is

evident is that when using SGA to assess nutritional status, the prevalence is lower at higher GFR
rates, at around 20-28% for mean GFR 30-20ml/L (respectively), and around 40% in patients with a
38, 39

GFR less than 15 mL/min

- 24 -

28

33

31

194 HD

1854 HD

331 HD

206 commencing
dialysis
15 HD

7719 HD

Albumin (quartiles)

4-point SGA (Rating

2-4)
TBK (Depleted=
observed<predicted)
Appetite (4-point
scale)
Appetite (Appetite &
Diet Assessment)

SGA (B & C)

SGA, MIS

42 months

42%

SGA (C)

83 HD

12 months

55%

7 point SGA (<5)

SGA (B & C)

12 months

144 patients (78

HD, 76 PD)
680 commencing
PD
309 Acute renal
failure in-patients
50 Pre-dialysis

12 months

38% fair/poor

N/A

36 months

8% Poor/very poor 7 years

24% Fair

7 years

36 months

40%

39% (2-4)

6 months

12 months

Not specified
19%

12 months

28%

2-3 years

Albumin (35-40g/L or N/A

30-35g/L)
TBN <80% predicted 25%

Mean followup duration

1200 HD pts

Prevalence of
malnutrition

Assessment tool
(malnutrition)

Population

RR Mortality, 1.5 (95%CI 1.2-2.0)*

RR Hospitalisation 1.4 (95% 1.1-1.2)
RR Mortality, 6.3 times 39-42 g/L; 8.1 times 3239g/L; 26 times 32-39 g/L (Referent >42g/l)**

Risk 12-month Mortality HR 4.7 (95% CI 1.9to 12.2)

Risk Mortality 78% (HR 1.8(95% CI 1.3-2.5))

(Females HR 2.6 (1.5-4.9); Males HR 1.7 (1.2-2.6))
Median survival 55 months vs 100 months normal

Risk Mortality OR 2.7 (1.53.9)

Hospitalisation OR 4.3 (2.65.8)
1-unit SGA Mortality 3.9 (1.53.9); 1.8 (1.2-2.8)
10-unit MIS Mortality 10.4 (2.3-47.6); 3.8 (1.9-7.9)
Risk Mortality 5% SGA B and 33% SGA C

RR Mortality, 2 times for 35-40g/L; 5 times for

30-35 g/L (40-45g/L = referent)
RR Mortality 4.1 times for NI <80%, (NI=>80%
referent).
RR Mortality 0.75 (0.66-0.85)* with 1 unit increase
(improvement)
Risk In-hospital Mortality OR 7.2 (4.1-12.7)

Result of malnutrition*

- 25 -

HD = Haemodialysis, RR = relative risk, PD = Peritoneal Dialysis, TBN = Total body Nitrogen, SGA = Subjective Global Assessment, OR = Odds Ratio, MIS = Malnutrition Inflammation Score; HR =
Hazard ratio, *95% CI **adjusted for CRP

Pupim, USA: 2004

37

Kalantar-Zadeh, USA:
35
2004
36
Burrowes, USA: 2005

Dolson, USA 2003

34

Stenvinkel, Sweden: 2002

Pifer, USA; 2002

Kalantar-Zadeh, USA:
32
2001

Lawson, Australia: 2001

Churchill, Canada, USA

29
1996
30
Fiaccadori 1999 US,

Pollock, Australia 1995

Author, country
Year
27
Lowrie, USA 1990

Table 1-3 Prevalence of poor nutritional status and outcome from prospective observational studies in Stage V CKD

Body composition studies in dialysis patients have verified depleted stores of body protein or cell mass
on initiation of dialysis.

Measurement of body cell mass (BCM) by total-body potassium (TBK)

counting has revealed a 10% reduction in predicted BCM in patients entering maintenance
40

haemodialysis . Also, in a cross-sectional evaluation of 134 Stage V CKD at commencement of renal

dialysis, Cooper et al (2003) established that patients commencing dialysis therapy with
GFR<10ml/min had significantly lower Nitrogen Index (observed/predicted TBN) of 88% verses 106%
p<0.0001. In this study, there was also an independent and significant correlation between renal
41

function (GFR) and TBN (r=0.21 p<0.0001) .

These data all indicate the development of malnutrition predates dialysis commencement. Nutritional
status clearly declines at critical stages prior to dialysis commencement and this is associated with
declines in body composition. This highlights the critical need for effective nutritional management
strategies in the stage preceding dialysis commencement to treat and/or prevent the progression of
malnutrition.

Malnutrition is also a primary indicator for dialysis commencement. In an Australian study of predialysis patients (n=50), after 12 months of standard medical care, 50% of the malnourished and 11%
of the well-nourished patients had reached the endpoint of death or dialysis and likelihood of acute
hospitalisation was increased three-fold in the malnourished group independent of baseline renal
31

function . There is potential that poor nutritional status may be a clinical indicator for increased
likelihood of early dialysis commencement.

Finally, malnutrition can also further exacerbate declining renal function. Firstly, due to a reduction in
the renal blood (plasma) flow, GFR and urine concentrating capacity of the residual function is further
42

reduced as a consequence of malnutrition as indicated in Figure 1-1 . In addition, acid/base

22

regulation and sodium balance may be further impaired .

- 26 -

1.1.4 Summary malnutrition in CKD

Uraemic malnutrition, marked by progressive loss of body protein may be induced by a number of
metabolic mechanisms, which affect body protein homeostasis (synthesis and degradation) and
39, 43

dietary intake

. Evidence of the effect of these symptoms in reducing measures of nutritional

24

status pre-dialysis CKD is well-documented . Malnutrition is common at the onset of RRT (dialysis)
2, 33, 44, 45

and is associated with poor outcome

. A number of studies indicate that malnutrition it is

1, 2

present in up to 48% of patients at the time of dialysis initiation

, indicating deterioration of nutritional

status preceding the onset of renal replacement therapy. In the absence of intervention, patients with
declining renal function appear to be on a trajectory of declining nutritional status leading up to dialysis
18

commencement .

1.2

Framework for nutritional management in CKD

Clinical practice for the nutrition management in CKD is guided by published guidelines

8, 46-50

. Clinical

51

practice guidelines should be based on synthesis of the best available scientific evidence . There
have been a number of published practice guidelines informing the management of CKD patients.
Guidelines including nutritional management include the Caring for Australians with Renal Impairment
47

49

(CARI) , American Kidney Dialysis Outcomes Quality Initiative (K/DOQI) , American Dietetic
Association (ADA) Medical Nutrition Therapy Evidence-Based Guides for Practice: Chronic Kidney
50

52

Disease (non-dialysis) , ADA Guidelines for Nutritional Care of Renal Patients (3rd ed) , European
Dialysis and Transplant Nurses Association and European Renal Care Association (EDTNA/ERCA)
Guidelines for the Nutritional Care of Adult Renal Patients

53

Most recently, Evidence-based practice guidelines for the nutritional management of CKD have been
46

published , endorsed by the Dietitians Association of Australia, by Australia and New Zealand Renal
Guidelines Taskforce, and of which the candidate is an author (Appendix B).

These guidelines

represent a summary of the best available evidence provided by the aforementioned guidelines

47, 50, 52-

54

. The best available evidence was established by evaluating against the classification system by

National Health and Medical Research Council (NHMRC)

51

. Table 1-4 provides the definition for the

critical appraisal of study design, to assign a level of evidence to intervention studies.

- 27 -

Table 1-4 National Health and Medical Research Council (NHMRC) classification for levels of
evidence

51

This table is not available online.

Please consult the hardcopy thesis
available from the QUT Library

Table 1-5 provides an excerpt from the published nutritional management guidelines in CKD,
highlighting the evidence statements relevant for Stage IV CKD. The highest level of evidence relating
to each clinical question is provided. It is acknowledged that where no evidence exists, published
guidelines stating consensus opinion from expert practitioners were included.
Table 1-5 Evidence-based recommendations and corresponding level of evidence for
nutritional management of Stage IV CKD (Adapted from Ash et al 2006)
Evidence Statement
Level of Evidence
Criteria for referral to Dietitian
At what level of GFR should patients be referred to the dietitian in order to maximise nutrition
intervention opportunities?
8
Protein and energy malnutrition increases with deteriorating kidney function Level III-2
and is associated with adverse outcomes
8
Low protein and energy intake is an important cause of poor nutritional status Level III-3
Nutrition Assessment
Which measures best reflects nutritional status or change in nutritional status in CKD?
47
Maintained percent (%) dry body weight reflects optimal nutritional status
Level II
49
Body Mass Index (BMI) = 18.5 25, reflects optimal nutritional status.
Level IV
49
Subjective Global Assessment (SGA) and % ideal body weight (IBW)
Level IV
47
Total body nitrogen, duel x-ray absorptometry or bio-electrical impedance
Level IV
Nutrition Prescription / Intervention
What are the prescriptions for appropriate nutritional intervention(s) to optimise nutritional
status in CKD and prevent malnutrition?
47
Energy intake of at least 146 kJ/kg IBW/day, patients < 60 years
Level II
49
for patients >60 years, an energy intake of 125 kJ/kg IBW/day
Level III-2
47
Protein intake 0.75-1.0g/kg IBW/day. At least 50% of high biological value
Level II
Implementation and Management
What are effective methods of implementation to achieve positive outcomes in CKD?
47
Pre-end stage kidney disease education forms an important part of Level II
management strategy to slow the progression of renal disease.
47
Nutrition counselling to encompass appropriate protein and energy intake
Level II
8
and include fluid, sodium, potassium intake and weight management
Opinion
Monitoring and evaluation: Recommended times for initial consultation are
50
45-60 mins and review 20-30 mins, for all patients
Opinion

- 28 -

From the evidence provided in these guidelines it is clear that the evidence informing practice in this
area have limitations, particularly with guidelines of implementation, monitoring and evaluation. The
literature review to follow will focus on the clinical questions relating to those stated in italics in Table
1-5. In particular, this review will aim to explore further evidence and gaps in the literature regarding
nutrition

assessment

techniques

and

nutrition

intervention,

both

dietary

prescription

and

implementation for use in Stage IV CKD.

Evidence informing nutrition prescription represents the most widely investigated area of CKD
research to date, and will be discussed in detail in Section 1.5.1 Dietary prescription studies in CKD.
46, 47, 49

Methods for nutrition assessment are based on recommendations in dialysis population

. Apart

from maintaining dry weight (level II), there is little evidence informing the other nutrition assessment
methods for this stage of CKD. This will be further explored in the following Section 1.3 Assessment
of nutritional status in CKD. There is only consensus opinion as to the most effective methods of
monitoring and evaluation of interventions to implement adequate dietary prescription and optimise
nutritional status. The evidence from studies in dialysis populations and other chronic conditions will
be further evaluated in section 1.5.2 Dietary implementation studies in CKD.

1.3

Assessment of nutritional status in CKD

Appropriate assessment of nutritional status is essential in order to implement effective and timely
nutrition intervention. It is more difficult to improve the nutritional status of a patient once wasting has
55

begun than to prevent malnutrition . Therefore, strategies to monitor nutritional status of patients predialysis are imperative to ensure appropriate nutrition support is implemented in a timely manner, to
prevent or minimise the development of malnutrition prior to RRT.

An ideal parameter to identify malnutrition should be unaffected by non-nutritional factors and become
normalised when a patients nutritional status is rectified with adequate nutrition support
clinical parameter consistently meets these criteria.
management in CKD,

56

. No single

Therefore, key guidelines for the nutritional

46, 47, 49

, state that assessment of nutritional status in ongoing patient care should

be based on several parameters. Assessment of nutritional status is undertaken using biochemistry,

- 29 -

anthropometry, body composition, subjective clinical assessment tools and assessment of adequacy
of dietary intake.

1.3.1 Biochemistry
Traditionally, serum proteins (particularly albumin) have been utilised as key biochemical parameters
for the assessment of nutritional status, due to their relationship to body protein turnover in clinically
stable conditions. However, it is important to note that abnormalities in serum protein synthesis may
45

not be directly related to changes in nutritional status .

As mentioned previously, synthesis of serum proteins may respond significantly to acute changes in
inflammation and co-morbidity status. Acute phase protein response (APPR), commonly associated
with elevated C-reactive protein (CRP) and low serum bicarbonate (reflecting metabolic acidosis)
results in a reduction of albumin synthesis and is thought to explain much of the high prevalence of
57, 58

hypoalbuminemia in chronic kidney disease

. In addition, serum albumin levels have been shown

to decrease in conditions of fluid volume overload, with the dilution of serum albumin concentration
59

due to an increase of extra-cellular fluid, which is prevalent in CKD patients . Even in the absence of
inflammation, serum albumin does not discriminate between well-nourished and malnourished
60

haemodialysis patients . Rapid, short-term decreases in albumin synthesis can occur with reduction
61

in dietary intake , however, long-term compensatory mechanisms aid to balance serum albumin
concentration, therefore the impact of long-term reduction of dietary intake on serum albumin is
subtle.

Hypoalbuminemia is shown to be highly predictive of mortality risk when present at the time of
initiation of chronic dialysis as well as during maintenance dialysis

62-65

, potentially reflecting conditions

of inflammation and acidosis, which may also be identified by abnormal biochemical markers.
Therefore, albumin is thought to be more closely related to overall health status from co-morbidity and
2

inflammation, rather than nutritional status, particularly in patients close to the start of RRT .

Other serum proteins (e.g., transferrin and pre-albumin) have also been utilised for assessment of
nutritional status in CKD. Again, these serum proteins may be insensitive to changes in nutritional
status and can be influenced by non-nutritional factors, for example inflammation and acidosis

- 30 -

64

. In

particular, transferrin can be influenced by iron status and pre-albumin, which is usually excreted by
the kidneys and may change prospectively in response to a change in kidney function, not necessarily
nutritional status.

Utilising serum proteins is not desirable to measure change in nutritional status, as acute changes are
likely due to an acute phase response or haemodilution, rather than reflecting long-term change in
3

body protein stores . Correction of conditions causing decreased albumin synthesis in-line with the
medical management of the patient can may assist in promoting a positive nitrogen balance and
66

minimise protein degradation , thereby reducing the risk of malnutrition progression.

While APPR and acidosis reflected in changes in albumin may be linked to nutritional status and poor
outcome, serum albumin is not a direct measure of body composition change, which leads to the next
section. However, to date, serum albumin has been the most extensively examined nutritional marker
22

in CKD, due to its availability and association with poor clinical outcome .

This is just one of the

limitations of nutrition management studies in CKD, and will be considered further in Section 1.5.2
Dietary implementation studies in CKD.

1.3.2 Body composition definition

Body composition assessment aims to compartmentalise the bodys contents. A small number of
cadaver analyses has provided reference data for the development of body composition models,
67

dividing body weight into two or more compartments . The two compartment model consists of fat
mass (FM) and fat-free mass (FFM). FFM is quite heterogeneous and groups all non-fat components
together, i.e. muscle, bone, water, electrolytes etc. These components vary independently, especially
in disease states such as pre-dialysis CKD. Therefore, the two-compartment model has its limitations
68, 69

in this population

As represented in Figure 1-2, the 2-compareemnt model can be further

compartmentalised. Two examples are provided in this figure, representing the mineral and cellular
compartment models.

- 31 -

2-C Model

4-C Mineral

4-C Cellular

Fat

Fat

Fat

Water
(TBW)

Body Cell
Mass

Fat-free
Mass
(FFM)

Extra-cellular
fluid
Protein
Mineral

Extra-cellular
solids

Figure 1-2 Body composition compartment models (Adapted from Wang et al (1992)
71
Heymsfield et al (1996) )

70

and

It is recommended that emphasis be placed on assessing FFM in CKD as its compartments of extra72

cellular water (ECW) and body cell mass (BCM) . Extra-cellular mass is made up of extra-cellular
fluids and solids, defined as the component of fat free mass which exists outside the cells, consisting
of fluid (e.g. extra cellular fluids, plasma volume) and solid (e.g. skeleton, cartridge etc). ECM is not
metabolically active. Skeletal muscle protein (termed somatic protein) and visceral protein together
comprise the metabolically active protein known as the body cell mass. Body cell mass (BCM)
reflects the cellular components of metabolically active tissue and is physiologically and chemically
67

more homogeneous than FFM . Malnutrition results in a reduced BCM and concomitant expansion of
extra cellular mass. Hence, weight changes may represent changes in body fat and/or fluid and may
not account for change in BCM compartment.

Therefore, isolating BCM is the ideal method for

monitoring the nutritional status of these patients as BCM reflects the functional status of the patient
and is a better indicator of survival from wasting diseases than body weight

67

Another method provided by the 4-compartment mineral model is to isolate the assessment of body
protein. This will be detailed further in Section 1.3.4. for the assessment of total body nitrogen.
Clinical research into interventions to modify nutritional status, should utilise methods which detect
small yet significant changes in body composition from changes in metabolically active tissues and/or
body protein, which is a direct measure of functional status

- 32 -

67

Table 1-6 Comparison of commonly used body composition methods in CKD

Measure

Method
(providing
xcompartment)
Scales.(1-C)

Assumptions

Advantages

Disadvantages

Dry body weight

BMI

Body size weight

(kg) over height
(m)2 (1-C)

Assumes equal
risk for mortality at
population level

Does not provide

information on
relative change in
body components.
Requires height
measure, useful at
population level

Skin fold
thickness

In-direct estimate
body fat from
estimate of
subcutaneous fat
(2-C)
Mid-upper arm
circumference and
tricep skinfold
indirect estimate of
total body muscle
mass (2-C)

Simple, quick. Can

obtain
retrospective
information
Simple, quick. Can
obtain
retrospective
information
Bed-side
technique, simple,
quick

Only requires 2
measures. Good
precision with arm
circumference

1-dimenstional
measurement.
Insensitive to small
changes in muscle
area.

Simple to apply
can look like
standard scales

Fluid retention or
dehydration will
alter

High precision.
Adequate measure
of fat mass in fluid
abnormality

Estimates
impacted by fluid
status

Direct
measure
nitrogen,

Delivers a
radiation dose,
expense,
availability, MDC
7%
Availability and
expense. May be
limited use in
obese due to size
restrictions

Weight

Bioimpedance

Passing a weak
electrical current
measure
Impendence to
electrical flow (3C)

Dual x-ray
absorptometry

Attenuation of dual
energy x-ray beam
to assess bone
mineral content (3C)

Total body
nitrogen

Protein by in-vivo
neutron capture
analysis (4-C
mineral)

Thickness of site
skin fold reflects a
constant
proportion of body
fat
Assumes
cylindrical limb,
static
compressibility of
fat, hydration and
humerous bone
circumference
Body weight =
Resistance +
TBW. Fluid status
73.2% Use of
healthy population
equations to
calculate FFM
Results in estimate
of soft tissue
mass (incl muscle
mass and water).
To derive FFM,
assumes fluid
status 73.2%
6.26g protein/g N
measurement of N
reflects
body
protein

Total body
potassium

Whole body K-40

by counter (4-C
cellular)

K40- content of
TBK, TBK content
of BCM

Mid-upperarm muscle
area

Direct measure of
intra-cellular
K,
MDC 4%

Requires training.
Inter- and intraobserver reliability
limited.

MDC = Minimal detectable change for an individual (based on reference man model)67; N = Nitrogen; FFM = Fat-free mass; K40
= Isotope potassium-40; TBK= Total body potassium; BCM = Body cell mass

1.3.3 Anthropometry
73

Anthropometric assessment may aim to assess body size and/or body composition . Anthropometric
measures are easy to administer, inexpensive and are able to be monitored over time.

Simple

anthropometric measures such as height, weight, % usual body weight (%UBW) and body mass index
49

(BMI) are regularly measured in patients with chronic kidney disease (CKD) . These measures

- 33 -

provide objective information on body size, or the 1 compartment, or whole-body, therefore do not
provide information on the body compartments or changes within. Monitoring changes in weight over
time may be a useful indicator of changes in nutritional status in individuals

54

, although changes in fat

free mass and fat mass are indistinguishable. In disease conditions with variable fluid status over time
(e.g. periods of fluid retention or dehydration), relative increase in total body water may mask actual
74

weight loss that results from losses of fat or skeletal muscle tissue . As mentioned previously this is
also a limitation of the 2-comparetment model.

BMI (weight(kg)/height(m) ) falling outside the reference range for adults may reflect over- or undernutrition at a gross level, and can provide information for classifying a population at a basic level.
However, as above, BMI does not distinguish between weight associated with fat or fat-free mass.
High BMI may be associated with excessive adiposity, muscularity or oedema from excessive fluid
retention

73

Anthropometry measure can also be used to estimate 2compartment model of body composition by
indirect assessment of fat and fat-free mass (FFM). The measure of skin fold thickness may be used
to provide an estimate of percent body fat, and by subtraction of total body weight, may provide a
42, 75

surrogate estimate of FFM

. Tricep skin-fold together with mid-arm circumference is a technique

commonly applied to dialysis patients to estimate body composition (an estimate of mid-arm muscle
42

circumference) .

Interpretation of this technique may be limited by inter-observer variability, which

has been shown to be up to 4.7% for arm circumference and 22.6% in tricep skin-fold when measured
76

by 3 observers . In addition, considering the potential variability of fluid status in pre-dialysis CKD,
42

intra-observer variability has been reported to be as much as 10% in HD patients .

It is also

important to consider that skin-fold assessment is essentially measuring fat mass, which will have low
42

sensitivity for detecting subtle change in body composition as it is relatively stable in the short-term .
This has been demonstrated in a study of dialysis and pre-dialysis CKD patients, where skin fold
77

thickness failed to detect significant changes in FFM compared to TBK assessment . Overall, there
are problems with use of skin-fold assessment for nutrition assessment as this technique is prone to
inter-observer variability, is essentially an estimate of fat mass, and assumes static fluid status.
Therefore, skin-fold assessment has limitations in clinical and research setting.

- 34 -

In summary, anthropometric measures are influenced by fluid levels and do not discriminate between
fat mass, lean body mass and body water. Care must be taken when evaluating prospective weight
and BMI measurements, considering the influence of variable fluid.

In addition, results of

anthropometric assessments should be interpreted with caution and not used in isolation.

In

summary, anthropometry is most useful to detect long term change in euvolemic patients and is best
54

undertaken with the same observer .

1.3.4 Total body nitrogen and total body potassium

Total body nitrogen (TBN, to measure body protein) and total body potassium (TBK, to measure
BCM)

78

provide direct measures of isolated components of the 4-C mineral and cellular models,

respectively.

The measurement of total body nitrogen (TBN) occurs via in-vivo neutron capture analysis to directly
79

assess body protein . Determination of TBN requires a neutron activation machine that delivers a
radiation dose. It relies on the principle that when a subject is irradiated with neutrons, a gamma ray
28

which is specific for the element (in this case nitrogen) is emitted . This separates nitrogen into the
actively metabolising component which is measured by the detectors

67, 79

. TBN reported in HD and

PD has indicated body protein stores significantly lower than controls

28, 80

. Also, Pollack et al and

Arora et al suggest a nitrogen index (observed nitrogen verses predicted nitrogen (from equations
28, 80

based on height and weight)) of 0.8 or less is predictive of mortality

and therefore used as a

81

reference point to identify malnutrition in further studies .

Total body potassium (TBK) is measured with a whole-body counter of 1.46 MeV gamma-ray
emissions from naturally occurring

40

82

K radioactive isotope . As potassium is the major intracellular

cation, approximately 98% of body potassium is present within the cellular compartment. Therefore,
the measurement of body potassium can determine BCM with precision. TBK has been used to
68

measure BCM in patients with pre-end stage kidney disease

haemodialysis

34, 83

and peritoneal dialysis

84-86

- 35 -

and end-stage disease undergoing

Abnormal potassium metabolism resulting in serum potassium fluctuations has caused concern about
the validity of this technique in CKD potentially reducing the sensitivity of TBK as a measure of BCM.
Woodrow and colleagues (2001) conducted a validation study in PD patients with matched healthy
controls, using gold standard techniques to establish the three components of FFM (deuterium oxide
69

(TBW); bromide dilution (ECW) and BCM (TBK)) . Concentrations of intracellular potassium from
gold-standard dilution measures and TBK models were equivalent in the CKD group and the control
69

group . This demonstrated that TBK is a valid technique for use in CKD patients, regardless of
abnormalities of potassium metabolism which may occur in CKD.

Limitations of both TBN and TBK techniques as detailed in table 1-6 are their expense and availability.
In addition, TBN assessment delivers a radiation dose at around 0.20 mSv, which is equivalent to the
average chest x-ray. Therefore, they are deemed most appropriate as research tools rather than for
everyday clinical use. However, a distinct advantage of utilising these tools is that they may be able to
detect small changes in muscle mass occurring in patients at risk of malnutrition who are yet to
develop significant muscle wasting. It is important to note, however, that the minimal detectable
difference of the TBN and TBK are 7 and 4% respectively

67

. Therefore, these tools are considered to

87, 88

be valuable research tool for monitoring progress following nutrition intervention

. This was varied

by some early work by Cohn et al (1983), indicating TBK is a more sensitive and reliable indicator of
changes in body cell mass, compared with TBN. Therefore, measurement of TBK appears to be the
most appropriate gold-standard body composition assessment to measure small changes over time
(as little a 4%).

1.3.5 Other body composition techniques

Other techniques commonly used in CKD patients to estimate FFM are dual x-ray absorptometry
(DXA) and bio-impedance analysis (BIA). DXA measures FM and bone tissue by the different x-ray
42

attenuation properties of body tissues at different energy levels . The remaining mass of body weight
is defined as soft-tissue mass, containing all non-bone components of FFM. The assessment of FFM
from this method assumes that 73.2% of FFM is TBW, therefore, standard a level of hydration. This is
89

then factored in to a standardised equation to estimate BCM .

BIA measures the body tissue

90

impedance to an electrical flow, which is directly proportional to the amount of electrolyte-rich fluid .
These readings (resistance and reactance) are then used to calculate TBW or FFM by standardised

- 36 -

equations. Therefore, the accuracy of BIA depends on the appropriateness of the prediction formula
for the individuals measured and current fluid status. The assumption of static fluid status is a
88

significant limitation for the use of DXA and BIA in pre-dialysis CKD .

Prospective observational studies in pre-dialysis using BIA and DXA, report pre-dialysis patients have
91, 92

relatively good BCM stores which are generally maintained over the course of CKD progression

However, as evident above, these methods rely on equations that assume constant hydration of FFM.
Over the progression, BCM component of FFM may be overestimated as there is no allowance for
intra- or extra cellular fluid expansion within FFM. Certainly these results are in contrast to direct
assessment of TBK and TBN which indicate depleted stores upon commencement of dialysis

40, 41

DXA and BIA are therefore considered to be acceptable amongst the dialysis population when
93

patients are euvolmic . However, the potential presence of an unstable fluid status, deems them
inappropriate to measure lean body tissue and hence, nutritional status in pre-dialysis patients

77, 93

1.3.6 Summary body composition in CKD

Determination of body compartments is essential in the evaluation of patients with wasting diseases in
88

order to optimise their nutritional support . Loss of body protein stores creates the issues relating to
malnutrition.

Therefore, nutrition status assessment tools which aim to identify physical signs of
94

protein-energy malnutrition, ideally monitor changes in body protein status .

Body composition

techniques are important tools to diagnose presence of protein depletion and to monitor efficacy of
42

nutrition therapy . Traditionally, measures such as body weight and body mass index (BMI) have
been used for nutrition assessment and as outcome measures in dietetic practice. However, these
measures do not provide information about which components of the body are affected during disease
progression. Consideration of co morbidities and fluid shifts between body compartments in the CKD
population is essential in the selection and interpretation of body composition techniques to assess
nutritional status.

As a result of the above, it is clear that a direct measure of metabolically active tissues is required for
a valid assessment of nutrition status considering the metabolic anomalies in pre-dialysis CKD.
Therefore, TBN and TBK techniques, however rare, represent the gold-standard for research studies

- 37 -

in pre-dialysis CKD.

Specifically, when assessing change over time, TBK may detect accurately

changes as small as 4%, providing an advantage over TBN assessment in short-term prospective
studies.

1.3.4 Clinical assessment tools 1

A number of tools exist in clinical and research practice in chronic kidney disease (CKD) to measure
95

nutritional status . Many of the tools in use are based on the parameters from the Subjective Global
Assessment (SGA)

95, 96

. Clinical tools are potentially valuable in the assessment of malnutrition to

identify the need for intervention and evaluating change in nutrition status, without influence from nonnutrition factors.

SGA provides a comprehensive appraisal of nutrition status considering a medical history (weight
change, dietary intake change, gastrointestinal symptoms, and changes in functional capacity) and a
96

physical examination (assessment of subcutaneous fat and muscle mass stores) . SGA classifies
nutrition status as well-nourished (A), mild to moderately malnourished (B), or severely malnourished
96

(C) .

96, 97

SGA was originally validated in surgical inpatients in 1987

widespread use in CKD

and has since gained

49, 95, 98, 99

. Limitations in its ability to measure varying degrees of malnutrition

and identify small yet clinically significant changes in nutrition status has resulted in a number of
modified and/or scored SGA-based assessment tools appearing in the CKD literature, creating
uncertainty about which is the most appropriate to use in conjunction with or to replace the original
95

SGA .

Given there are a number of SGA-based tools available for use in CKD, it is critical to evaluate their
validity and reliability for use in CKD, to build the evidence base and promote consistency in practice.
A number of types of validity and reliability outlined in table 1-7 will be used to evaluate the evidence
for SGA and SGA-based tools in CKD literature. The original SGA demonstrated predictive validity as
detailed previously in table 1-3 in prospective observational studies on the impact of malnutrition on
clinical outcome. This tool and the modified-SGAs will be further evaluated for each type of validity in
the following section.

- 38 -

Table 1-7 Definition and study design for validity and reliability testing corresponding NHMRC
level of evidence in reference to nutrition assessment tools in CKD
Definition

37

Study design and NHMRC level of

23
evidence
Cross-sectional comparison with
reference that is not a gold standard,
and/or is not blinded (Level III-2,
Diagnosis)

Clinical
validity

Explores the relationship that exists in

known parameters associated with nutrition
status, but not used in the tool. Validity is
then only established against the parameter
within that investigation

Predictive
Validity

Explores the correlation with another

measure assessed in the future (e.g.
morbidity and mortality)

Prospective, cohort
(Level II, Prognosis, applies only to
studies of diagnostic accuracy)

Criterion
Validity

Evaluates agreement and performance

(sensitivity/specificity) of the tool, against a
valid, gold-standard reference measure.

Inter-rater
Reliability

Tool can provide good agreement between

users

Cross-sectional blinded comparison

in consecutive or non-consecutive
patients
(Level II or III-1, Diagnosis)
Cross-sectional

Intra-rater
Reliability

Tool is reproducible after test-retest

assessment

Test-Retest

Internal
Reliability

An assessment of the correlation across

items within the assessment tool

Cross-sectional

1.3.4.1 SGA
SGA is well studied in CKD. This is evident in table 1-8 summarising studies demonstrating clinical
validity of SGA against a range of biochemical, anthropometrical and body composition measures.
SGA was first investigated in a validation study in haemodialysis (HD) and peritoneal dialysis (PD)
100

patients in 1993

, where SGA displayed clinical validity with albumin, bioimpedance (BIA) phase

angle, mid arm muscle circumference (MAMC), percent body fat (by skin-fold assessment) and
normalised protein catabolic rate (nPCR) with multiple regression analysis confirming the strong
relationship between SGA and the combined objective measures (multiple r = 0.77). Lawson et al
(1999), in a 12-month follow-up study further investigated the predictive validity of the SGA in HD
patients, determining SGA B or C rating results in significantly increased mortality (OR = 2.7 (95% CI
1.5-3.9), likelihood of acute hospitalisation, and depleted FFM (as measured by BIA)

101

In an

investigation by Cooper et al (2000), Nitrogen Index (measured total body nitrogen compared to sexage- height matched general population), was significantly lower in patients rated SGA B and C
compared to A

102

Components of the following section published in Nutr Diet and J Renal Nutr (Appendix A)
- 39 -

Table 1-8 Validation studies demonstrating the clinical validity of subjective global
assessment
Author, country
Year

Population

Prevalence
malnutrition

Outcome of nutritional status

SGA Classification
A
B
C

Fenton, Canada:
103
1987
Young, Europe
104
and US 1991
Enia, Italy
100
1993
Jones, England
64
1997
Qureshi, Sweden
105
1998

118 PD

42% (<3 months)

18% (>3 months)
41%
B 33% C 8%
31%
B 24% C 7%
47%
B 42%; C 5%
64% (2-4)
2 51%; 3/4 15%

% Coronary artery disease

26% 59%

Male residual GFR (ml/min)

Female Albumin
PD Albumin (g/l)
HD BIA ph angle
Male BMI (kg/m2)
Female Strength (kgf)
Albumin (g/l)
I-GF1 (ng/ml)

1.8
37.2
34.3
6.76
25.3
26.3
35
221

44% (2-4)

carotid plaques
LBM (kg) DXA

60% 90%*
51.5 44.5**

Stenvinkel,
38
Sweden 1999

224 PD
36 HD; 23PD
76 PD
128 HD
commencing
dialysis
119
commencing
dialysis
32 HD

Julien, France,
106
2001
Cooper, Australia 52 HD; 24 PD
81
2002
Cupisti, Italy 2004 70
107
Pre-dialysis
Jones, England
72 HD
108
2004
Desbrow, Australia 60 HD
109
2005

76%
B63% C13%
42%
B34%;C8%
29% (B only)

Pre-albumin (mg/l)
Albumin (g/l)
NI %predicted
Urea mg/dL
Bicarbonate mol/L
Composite score A,B,C
Composite score 7-pt
PG-SGA median
PG-SGA range

31% (B only)
20% (B only)

541
37.0
106

0.02*
27.0*

1.6
34.1
28.3*
4.76*
21.8*
16.9*
32
160

30
131

482
31.0
95

381*
19.0*
82**

80
107**
25.8 22.1**
2.1
5.0**
r=-0.472*
2
16*
0-13 7-26

*p<0.001, ** p< 0.01 p<0.05

SGA = Subjective Global Assessment; A=Well-nourished, B=Moderately malnourished, C=Severely Malnourished; PD =
Peritoneal Dialysis; GFR = Glomerular Filtration Rate; HD Haemodialysis; BIA = Bio-impendence Analysis; BMI = Body Mass
Index; LBM = Lean Body Mass; DXA = Dual X-ray Absorptometry; NI = Nitrogen Index; PG-SGA Patient-Generated Subjective
Global Assessment.

110

According to the draft National Health and Medical Research Council criteria for levels of evidence

SGA has level II and III-1 evidence for prognosis and diagnosis (respectively). Merging the SGA
groups representing malnourished (B & C) is adequate for clinical and predictive validity, with
reasonable reliability (Table 1-10). The prognostic validity of SGA was demonstrated in a number of
studies earlier in Table 1-3. SGA is recommended for use in CKD as a clinical tool to determine who
to target for nutrition support, as it represents quick, cost-effective, multi-disciplinary assessment, not
influenced by the metabolic anomalies of CKD
tools

111

95

and reliable, compared with other assessment

1.3.4.2 Modifications to SGA

Modifications of the SGA present in the CKD literature include tools that expand the scale of the SGA
38

29

112

to a larger number of categories (4-point , 7-point ), allow retrospective assessment

109, 113

continuous score for components of the SGA

- 40 -

and incorporate objective measures

114

, provide a
(detailed in

Table 1-9).

A potential benefit of such tools is to improve the ability to measure the degree of

malnutrition and to identify small, yet significant changes in nutrition status

108

To the authors

knowledge, there has been no previously published data using these tools in a pre-dialysis CKD
population. Table 1-9 provides an overview of the modified-SGA tools. Following this, Table 1-10
provides a summary of the SGA and modified-SGA tools validity, reliability, design comments and
NHMRC levels of evidence for the tools featured in this literature review.
Table 1-9 Description, advantages and disadvantages of SGA-based nutrition assessment
tools utilised in CKD.
Tool

Method

Modification from SGA

Advantages

Disadvantages

Retrospective
mSGA

Rating
A,B,C

Retrospective self rating on

A, B, C scale

Conducted as a
survey (self-report)

Rely on self-report
and carers physical
assessment

4-point SGA

Rating
1 to 4

May delineate poor

nutrition status

Difficult to note
changes over time

7-point SGA

Rating
7 to 1

May delineate levels

of nutrition status

May increase interobserver variation

Dialysis
Malnutrition
Score (DMS)
Malnutrition
Inflammation
Score (MIS)

Scored
7 to 35

Expands the B category to

two. Ratings >2 represent
malnutrition
Expands the 3 categories of
the original SGA, to 7 on a
Likert-type scale
Scores 7 components of the
SGA as 1 (normal) to 5 (very
severe)
10 components, the DMS with
BMI, serum albumin and totaliron binding capacity, scored
according to severity 0
(normal) to 3 (very severe)

Scored so less
subjectivity

Allocation of scores
not based on
evidence
Requires
biochemistry
(albumin and iron
studies), and
measures for BMI

Patient-Generate
Subjective Global
Assessment
(PG-SGA)

Scored
0 to 50
and
A,B,C

Scored
0 to 30

Provides a numerical score,

dependant on the impact of
each SGA component on
nutrition status

Includes objective
categories (less
reliance on
subjectivity)
Patient completes
medical history,
scored so less
subjective

May require more

patient input

SGA = Subjective Global Assessment; BMI = Body Mass Index

The 7-point SGA is the method of rating recommended by the National Kidney Foundation K/DOQI
guidelines, as part of a panel of other measures for the regular nutrition assessment of CKD patients

49

(Table 1-10). Jones et al (2004) investigated the criterion validity of 7-point SGA against a composite
108

nutrition score, containing a variety of measures, including the original SGA

. Although statistically

there appeared to be agreement, the composite score distribution between the 7 categories had a
large overlap, and non-systematic distribution of scores, particularly over ratings 4 to 6 (representing
the spectrum of moderate malnutrition). The prognostic ability of 7-point SGA was confirmed in a
large prospective study of PD patients (n=680), an increase in 1 rating on the 7 point scale indicated a
29

25% increased risk of death in the subsequent 12 months . In all studies, the 7-point SGA, has been
29, 108, 115, 116

treated as an interval or continuous score

- 41 -

(Table 1-10). Therefore, relative validity and

reliability results are based on the assumption that the 7-point scale of the SGA is directly linear
(interval).

The tools providing a continuous score, such as the PG-SGA, DMS and MIS have added merit for use
in the clinical setting to monitor small changes in nutrition status over time, all appear to have clinical
109, 113

validity to key nutrition-related variables and reach Level III-2 evidence

. The MIS shows promise

as a potential predictor of outcome and indicate the degree of Malnutrition Inflammation Cachexia
117

Syndrome

Similarly, small changes in PG-SGA score has previously reflected important clinical
118

changes, such as quality of life in oncology

patients

and length of hospitalisation for general medical

119

. The recent study in HD patients indicates the scored PG-SGA may have similar merit for

use in CKD as it has for oncology patients

109

. Tools that monitor nutrition status on a continuous scale

are also beneficial in research, with less reliance on subjectivity and provide greater statistical power
than a categorical rating of nutrition status.

1.3.4.3 Summary of clinical assessment tools

SGA along with modified versions have shown clinical and/or predictive validity in CKD dialysis
populations

2, 29, 101, 114, 116, 117

. Merging the SGA groups representing malnourished (B & C, <5 on 7-

point or >2 on 4 point scales) is common practice in the literature and SGA (A vs. B & C) is a valid
method for assessing the presence of malnutrition for diagnostic and prognostic purposes. Scored
assessment tools, such as PG-SGA and MIS may improve the clinicians ability to assess small
changes in nutrition status, as an ideal nutritional marker should reliably measure change in nutrition
status, not just predict clinically important outcomes.

At the time of this review, to the authors knowledge, there were no studies to date showing
longitudinal changes in SGA rating with nutritional intervention. Given the number of SGA-based tools
available for use in CKD, it is critical to evaluate their applicability to the non-dialysis CKD population,
particularly in a prospective study evaluating the tools response to intervention.

- 42 -

Enia 1993 Clinical validity (36HD; 23PD)

Lawson 2001 Clinical and predictive
validity 50 CKD (Pre-dialysis)
Cooper 2002 Criterion validity and
reliability (24PD; 52HD)

Subjective Global
Assessment
(SGA)

Kalantar-Zadah 1999 Clinical validity and Albumin, Cr, BUN,

Reliability (41 HD)
Iron, Cholesterol,
MAMC (p<0.001)
Albumin, CRP, Cr,
Hct, IL-6

Albumin

Kalantar-Zadah 2001 Clinical and

Predictive validity (83 HD)
Kalantar-Zadah 2004 Clinical and
Predictive validity (378 HD)

Desbrow 2005 Clinical and Criterion

validity and Reliability (60 HD)

Dialysis Malnutrition
Score

Malnutrition
Inflammation Score

Patient-Generated
SGA

Morbidity/Mortality
(12 months)

Mortality/Morbidity
(2-3 years)

NHMRC II (Prognosis) though

interpreted 7-point SGA as interval
NHMRC III-2 (Diagnosis) Not blinded
criterion, a large overlap scores
between groups, and lack of linearity.
Reliability on small sample (n=16)
NHMRC III-2 (Diagnosis)
Not blinded, no gold-standard or
report of number who refused study
Reliability on small sample (n=13)
NHMRC III-2 (Diagnosis), potentially
prognostic, MIS >8 = high risk
mortality/ morbidity, advantageous to
predict outcome over SGA and DMS

Inter-observer
Good, ICC=0.72
Intra-observer
Good, ICC=0.88

Inter-observer
Good, k=0.83

NHMRC III-2 (Diagnosis)

Non-consecutive sample and nature
of the gold-standard did not allow
blinding of the SGA to PG-SGA

No level assigned: question face

validity with death risk for B vs. A

>9 SGA
Internal
PG score 9 good Good (-coefficient
reliability to SGAB: 0.73)
83% sensitivity, 92%
specificity

Composite nutrition
score (incl SGA,
BMI, TSF, MAMC,
Alb)

Mortality (6 months) -

- 43 -

HD = Haemodialysis; PD = Peritoneal Dialysis, BIA (PA) = Bioimpedence Phase angle; MAMC = Mid-arm muscle circumference; BF = Body Fat; nPCR = normalised Protein Catabolic Rate; TSF =
Tricep Skin-fold, Cr = Creatinine; FFM = Fat Free Mass; ICC = Inter-class correlation, k = Kappa-statistic; Hct = Haematocrit; IL-6 = Interleukin 6

Rating <5: Albumin

Churchill 1996 Clinical and Predictive
Rating 1-7: BMI,
validity (680 starting PD)
Visser 1999 Clinical validity and Reliability MAMC, %BF
22 Dx (13HD; 9PD)
Jones 2004 Criterion validity (72 HD)

Not established
Inter-observer
NHMRC II (Prognosis)
(Nitrogen Index, NI Moderate (weighted NHMRC III-1 (Diagnosis) blinded to
<85%= malnutrition) k=0.60)
reference, question gold-standard
cut-off for criterion validity

Albumin, BIA (PA), Mortality/Morbidity

(12 months)
(p<0.001)
MAMC, % BF,
nPCR (p<0.05);
Total body nitrogen

Highest NHMRC level and

design comments

Criterion Validity Reliability

Clinical Validity Predictive

Validity

7-point SGA scale

Retrospective mSGA Pifer 2002 attempt Predictive (7719 HD)

Studies included
Author, year, study type
(sample)

Tool

Table 1-10 Validation studies for SGA-derived tools in CKD, Stage IV and V

1.3.5 Dietary intake assessment

Inadequate dietary intake of energy and protein is recognised as a leading cause of the development
14

of malnutrition in CKD . Observational studies suggest that a reduction in dietary energy intake is the
most important contributor to the development of malnutrition in pre-dialysis CKD patients

18, 24, 120

Estimation of energy and protein intake by different methods also can be used as a surrogate marker
of nutrition status in CKD patients.

1.3.5.1 Methods of dietary intake assessment

Assessment of dietary protein intake may be undertaken though direct biological methods.
Estimations of dietary protein intake in CKD may occur through 24-hour urine urea nitrogen excretion
rate for patients not on dialysis

121

, or protein equivalent of total nitrogen appearance (PNA) in dialysis

22

patients . These measures of dietary protein intake are valid only in steady state (non-catabolic)
states.

In catabolic states, these methods may overestimate actual intake due to the high urea
22

appearance from increased protein breakdown . In addition, 24-hour urine, which is the method
suited for patients in the pre-dialysis phase, represents significant participant burden. Eight collections
122

are required to obtain a precise measure of urinary nitrogen excretion

and then once received

123

validation of the collection is required to verify complete collection

. In previous studies in healthy

populations, up to 25% of collections are rejected on the basis of omission or incompletion rates

124

Therefore, although the 24-hour urine collection may be the most valid measure for use in stable predialysis patients to estimate protein intake, its participant burden and poor validity in catabolic states
renders it not ideal as an outcome measure in nutrition intervention studies for pre-dialysis patients. In
addition, as it only looks at protein, it provides no information on overall energy intake and other
nutrients which are also considered to be important.

Assessment of dietary intake by participant-reported methods are less cumbersome than direct
biological methods, however, rely on participant self-report. The four main methods of dietary
assessment in the literature appear in Table 1-11.

- 44 -

Table 1-11 Common method of participant-reported dietary intake; assessment methods and
considerations
Method

Details

Diet History

An open ended interview recalling usual Useful in clinical practice, can be a time
intake (frequency and quantity) over a effective method to assess habitual
125
specified period
intake when implemented correctly .
Results depend on the skill of the
73
interviewer
A food list incorporating specific food items Useful for epidemiology studies of the
and food groups and their frequency of relation between nutrient intake and
consumption (daily, weekly, seasonally disease, not useful for individual
126
etc), over a given time period
intakes
An interview asking subjects to recall their Useful for assessing average intake of
73
dietary intake during the 24 hours prior to a large population . Does not account
the interview or on the previous day
for attenuation due to day-to-day
127, 128
variation in intake
.

Food-frequency
questionnaire

24-hour recall

Food record or
diary

Considerations for use

A diary for a specified period (commonly 1

to 7 days) in which a subject records all
food and drink consumed, quantified by
weight or portion size

126

Valid assessment of current intake ,

accurate but time consuming. Longer
time frames result in higher respondent
73
burden and lower cooperation .

Level of respondent burden and the purpose of the dietary intake data are necessary considerations
when establishing methodology for use in clinical studies. As dietary assessment methods rely on
participant report, they can therefore lack accuracy. In a large-scale investigation of the validation of
food frequency, 24-hour recall and food record, Bingham (1995) revealed that weighed food record,
followed by estimated food record over 7 days to be the most valid methods for estimating nutrient
127

intake

Further investigations have shown little difference between food records over 3 days

compared with 7-day records

128

129

, particularly if a weekend day is included in the 3 day period

Repeating measures of intake are recommended in recording methods of dietary assessment in order
to reduce the issue of impact from daily variation

130, 131

1.3.5.2 Dietary intake assessment in CKD

Estimation of dietary intake in CKD patients is not without difficulty. The quantification of dietary
protein and energy intake is of importance in pre-dialysis CKD. There is no ideal method to provide a
valid and reliable estimate of both energy and protein intake. Collection of 3 of 7-day dietary records
25

and utilising the method of 24-hour urine collection (to establish protein intake)

18, 25

24,

are the most

common methods used in CKD.

Self-reported energy intakes can be validated by comparing intake with measured energy expenditure
in weight stable participants

73, 132

. Studies using a the gold-standard marker for energy expenditure,

doubly-labelled water, had highlighted systematic bias toward underestimation of energy intake, using
- 45 -

each different method of dietary intake assessment, among both males and females and of all age
groups

124

. Using this approach, under-reporting of energy intake is particularly prevalent in individuals

133, 134

with a high BMI

. This is a consistent finding in the CKD

135

, HD

136

and healthy populations

124, 127

1.3.5.3 Summary of dietary intake methodologies

Dietary intake methodology which can quantify the nutrients of interest adequately with minimal
respondent burden is ideal for use in clinical studies.

Care must be taken when attempting to

establish absolute intake of energy with participant-reported methods, particularly considering the
135, 136

evidence of significant underreporting in patients, particularly those with high BMI

. For the

assessment of changes in dietary intake, 3-day food records, repeated at time-points of interest are
131

considered valid for use in prospective assessment

1.4

Quality of life (QOL) assessment in CKD

The concept of quality of life (QOL) has come to the forefront as a primary outcome measure for the
medical treatment that patients receive.

Health related QOL attempts to measure a patients

functioning, wellbeing and general health perception in physical, psychological and social domains

137

In chronic disease, a close relationship exists between QOL, morbidity and mortality, as demonstrated
138

in Table 1-12.

In CKD specifically, QOL appears to be a strong predictor of survival

139-141

. It has

been established that CKD patients experience a significantly lower QOL compared with healthy
controls, which is more pronounced in the pre-dialysis phase (Stage 4 and 5), especially in the elderly,
and in females on dialysis

142-146

. Patients may move between treatment modality a number of times,

including progressive renal failure (Stages 1 to 5), haemodialysis, peritoneal dialysis and transplant,
therefore over the course of a lifetime, QOL will be affected in a number of ways related to their kidney
disease, and associated morbidities

147-149

- 46 -

Table 1-12 Prospective observational studies investigating the impact of quality of life on
clinical outcome in chronic kidney disease patients
Study

Sample

QOL
tool
SF-36
(PCS,
MCS)

Outcome

Main results

DeOreo,
139
1997

1000 HD

Mortality
Hospitalisation

65 HD

SF-36
(total
score,
PCS
and
MCS)

Mortality
Hospitalisation
BMI
Haemoglobin
Albumin

Lowrie,
150
2003

13,952
HD

SF-36

Mortality
Hospitalisation

Mapes,
151
2004

7378 HD
DOPPS

Mortality
Hospitalisation

Kalantar
-Zadeh,
35
2004

331 HD

KDQOL
(PCS,
MCS,
KD)
SF36
(PCS,
MCS)

5-units = survival advantage

PCS (10%); MCS (2-6%)
PCS (<34 (median value)):
2x death, 1.5 x hospitalisation
Significantly
PCS compared with general
Independent predictor mortality:
10-units SF-36 total: OR 2.1 (1.1-3.9)
SGA category: OR 8.7 (1.9-40.6)
1g/L albumin: OR 7.9 (2.0-30.7)
10-units QOL, predictors mortality:
Bodily Pain: OR 1.8 (1.1-2.9)
Mental Health: OR 2.5 (1.3-4.8)
Social Function: OR 1.5 (1.1-2.1)
Role Emotional: OR 1.3 (1.0-1.7)
2% risk morality AND hospitalisation with 1point increase QOL:
PCS OR 0.98 (0.97-0.98)
MCS OR 0.98 (0.98-0.99) adjust clinical factors
RR death per 10-point QOL:
1.13 MCS, 1.25 PCS, and 1.11 KDCS.
RR Hospitalisation 10-point QOL:
1.06 MCS, 1.15 PCS, and 1.07 KDCS.
38% Appetite (fair-poor) associated with:
HR death 4.74 (95% CI) (1.85-12.16)
Hospitalisation HR 1.43 (1.23-1.66)
PCS; 10-units OR 1.29 (1.14-1.45)
MCS; 10-units OR 1.32 (1.16-1.50)
SF-36; 10-units OR 1.30 (1.15-1.46)
MIS; 5-units OR 2.36 (1.61-3.48)

Kalantar
-Zadeh,
141
2001

Appetite (1-4)
MIS
Mortality
Hospitalisation

QOL = Quality of Life; HD = Haemodialysis; SF-36 = Short-form 36; PCS = Physical Component Summary; MCS = Mental
Component Summary; SGA = Subjective Global Assessment; OR = Odds Ratio; KDCS = Kidney Disease Component Score;
HR = Hazard Ratio; MIS = Malnutrition Inflammation Score

The reliability, validity, and sensitivity of a number of quality of life tools has been demonstrated in
patients with chronic kidney disease, and there is growing experience using this questionnaire to
collect information from renal patients

150

. As evident from Table 1-12, the 2 component summary

scores of the SF-36, physical component summary (PCS) and mental component summary (MCS), is
significantly associated with mortality and morbidity rates among dialysis patients

139, 141, 150

1.4.1 Renal function and quality of life

Shidler et al (1998) found that a negative perception of illness increased with decreasing GFR, and
152

was associated with higher depression scores and lower QOL

Prospective studies confirm a

progressive reduction in QOL as renal function declines and dialysis commencement signals an
improvement in QOL in most individuals

149, 153

. An exception to this is subjects with inadequate pre-

- 47 -

dialysis care and late referral for dialysis preparation (i.e. 1-4 months prior to treatment) experience
137, 154

significantly lower health-related quality of life on dialysis than those with adequate preparation

A recent study using the Health Utilities Index (a multi-attribute assessment of capacity, scale 0 to 1),
subjects with CKD indicated a significant reduction in patients with GFR <30 ml/min (mean 0.54),
compared to the healthy adult population (0.90 to 0.94)

155

. This indicator (addressing the attributes of

vision, hearing, speech, ambulation, dexterity, emotion, cognition and pain) revealed that only persons
with a Stroke and Alzheimers disease reported the same or lower utility as subjects with
144

GFR<30ml/min

. This demonstrates the significant impact CKD has at an individual level, impacting

on day-to-day morbidity and decreasing QOL is a significant predictor of increasing morality.

1.4.2 Nutritional status and QOL

156, 157

In addition to declining GFR, nutritional status has been shown to impact on QOL

. Table 1-13

demonstrates this with lower QOL being associated with lower nutritional status, identified by various
nutrition assessment parameters

156, 157

. In the DOPPS study, Dwyer et al (2002) showed that a

decreased appetite was independently associated with higher MCS and PCS scores. This study
concluded nutritional status assessed by a number of markers significantly impacts QOL, controlling
for co-morbidity and demographics

158

Several studies in other populations have observed poorer QOL outcomes in malnourished patients
159-164

when compared to well-nourished

. Such intervention studies have reported improved nutrition

status correlating with improved QOL for cancer

disease

159, 160

163, 164

- 48 -

, liver disease

161

, type 2 diabetes

162

and heart

Table 1-13 Studies demonstrating the relationship between nutritional status and quality of life
in chronic kidney disease patients
Study
Sample
Ohri289 HD
Vachaspati
165
, 1999

Design
Crosssectional

QOL tool
KDQOL

Laws,
166
2000

64 HD

Crosssectional

QOLI
KI
OLS
IGWB

Bakewell,
146
2002

88 PD

Prospective
(2-year)
observational
6mth r/vs

KDQOL
(PCS,
MCS, KD;
all subscales)

SGA
Albumin
Pre-albumin

Dwyer,
158
2002

1387 HD

Crosssectional

SF-36
(PCS,
MCS)

Appetite
Energy intake
NPCR
Cr, Alb, BMI,
Calf
circumference

KDQOL
Depressio
n screen
3-item

Biochem,
clinical,
demographic

HEMO
study

Walters,
167
2002

422 at
Crossbeginning sectional
HD
case-control
Commencing
From 151 dialysis
centres

Outcome
Albumin
PCR
Demographics

Main results
Albumin, independent associated
with: Physical function, Social
function, Burden of kidney disease
PCR independent associated with:
Physical function; retirement /disability
Significant decreases with SGA B and/or
SGA
Anthropometry, C:
Biochem
Psychological sub-scale; Satisfaction
Hospitalisation (OLS) and Well-being (IGWB)
Function (KI)
Length of stay
Significant, independent QOL with
SGA: (magnitude not available)
Work, social functioning, fatigue, pain,
sleep and emotional well-being
QOL over 2 years, greatest :
General health, burden of KD, emotional
wellbeing and satisfaction
Unadjusted
PCS Appetite, Energy intake; nPCR;
Albumin; BMI; Calf circumference; Age
MCS Appetite; BMI; Calf circumference;
Dialysis vintage
Adjusted for co-morbidity, demographics
PCS Appetite; Energy; BMI; Age
MCS Appetite; Age
56% Hb <100g/L ; 52% Alb <35 g/L
Kidney disease scales, PCS & MCS
in commencing dialysis patients
compared with prevalent HD cohort
(>1,000 pts) and non-renal CHD pts
(n=61)

HD = haemodialysis, KDQOL = Kidney disease quality of life; PCR = protein catabolic rate, QOLI = Quality of Life Index; PCS =
Physical Component Score; MCS = Mental Health Component Score; IGWB = Well Being Scale; KD = Kidney Disease;

1.4.3 Summary QOL

When pre-dialysis CKD (Stage 4 and 5 CKD) is reached, the impact of the illness on functional state
137

and health-related QOL as perceived by the patient is considerable

. The evidence suggests that

adequate pre-dialysis care and implementing interventions to improve the quality of life if CKD patients
should be encouraged throughout treatment.

1.5

Nutrition intervention in CKD

It is well established that deterioration in nutritional status often predates the onset of renal
63,

replacement therapy

168

Subjects with advanced CKD who are not receiving nutritional

management demonstrate the greatest deterioration in nutritional status before dialysis therapy

- 49 -

18

initiation . Considering the prevalence of malnutrition and the strong association between malnutrition
and mortality rate in pre-dialysis CKD, as demonstrated in 1.1.3 Prevalence and consequence of
49

malnutrition in CKD, there is a need for careful nutritional monitoring and treatment at this time .

Nutrition support for the period preceding dialysis initiation may minimise the impact of uraemic
symptoms associated with poor nutritional status and potentially reduce the rate of decline in renal
function in pre-dialysis CKD patients.

High intake of foods rich in protein may accelerate the

accumulation of uraemic toxins, leading to metabolic acidosis, excess sodium can aggravate
52

hypertension and potassium may lead to hyperkalemia .

2, 169, 170

Malnutrition has the potential for reversal following the commencement of dialysis therapy

This is reflected in prevalence estimates provided from the prospective observational studies in table
1-3, identifying malnutrition by SGA was greater in patients commencing dialysis (39-55%), than those
undergoing maintenance dialysis (19-25%). The improvement that does take place is most likely due
to the reduction in uraemic toxicity, thereby improving dietary intake afforded by increasing appetite.
Nutrition intervention aimed at controlling the symptoms of uraemia, reducing further renal injury,
managing the metabolic/electrolyte disturbances, and improving quality of life may therefore assist in
the medical management of CKD patients.

However, conclusive evidence about the mode of

171

effectiveness of intervention in improving nutritional status has not been thoroughly examined

There are two aspects to dietetic management, nutrition prescription and the implementation of
dietetic care.

The success or otherwise depends equally on appropriate prescription and

172

implementation

. The evidence for both components of nutrition care will be assessed below.

1.5.1 Dietary prescription studies in CKD

1.5.1.1 Protein Prescription
The majority of nutrition intervention studies to date in pre-dialysis focus on the effect of protein
restriction on patient outcomes. Dietary protein restriction is a time-honoured intervention method with

- 50 -

progressive CKD. The aim of this intervention is to reduce the production of urea, thereby provide
renal protection and better metabolic control, and potentially delay the progression to dialysis initiation.

A significant number of studies, including 4 meta-analysis

173-177

have investigated the relationship

between dietary protein restriction and progression of renal failure, as evident in Table 1-10. Despite
the strong interest, with over 50 studies in diabetic and non-diabetic patients, the efficacy of a protein
restricted diet, and an optimal level of protein intake in pre-dialysis CKD has yet to be resolved.

Very low protein intakes (<0.5g/kg/day) consist primarily of a vegetarian diet and provide insufficient
amino acids to promote nitrogen balance (protein synthesis). It is for this reason that these diets are
coupled with supplementation with essential amino acids. It is also important to note that dietary
prescription without amino acids should be at least 50% of the protein being of high biological value.
This ensures sufficient availability of essential amino acids for protein synthesis and maintenance of
nitrogen balance with sufficient energy intake.

Table 1-14 Metanalysis investigating outcomes of protein restriction studies in pre-dialysis

chronic kidney disease
Study

Inclusion

Outcome

Results
Compared with
unrestricted intake

Conclusion

Kasiske
176
1998

13 RCTs,
n=1919,
follow-up

Renal
function (rate
of GFR fall)

Rate of decline
0.53 (0.08 to
0.98 mL/min/y)

Save 0.5 mL/min/year loss,

greater for diabetic patients

Waugh
178
1999

5 Trials;
n=60 IDDM,
follow-up
4.5-24
months
8 RCTs,
n=1494,
non-diabetic
follow-up 1224 months

Renal function
(rate of GFR
fall); Creatinine

Rate decline
decreased
0.7ml/min

Controlling protein intake below

1.0g/kg or 0.8g/kg where possible
to reduce loss of renal function in
IDDM

Renal Death
(progression to
dialysis, death
or transplant)

RR death 0.76
(0.54 to 1.05) for
0.6 g/kg
RR death 0.65
(0.49 to 0.86) 0.30.6 g/kg

Reducing from free intake to 0.30.6g/kg may decrease the

occurrence of renal death by 31%

Fouque
173
2006 *

RCT = Randomised controlled trial, RR = Relative Risk of Death *Update of Fouque 1994, 1992 and Pedrini 1996

The largest randomised-controlled (RCT) study to date investigating protein restriction is the Modified
Diet in Renal Disease study (MDRD). Patients were assigned to two Study groups. Study A included
patients with GFR 25-55mL/min (n=585) randomised to a usual protein intake (1.3g/kg/day) or low
protein (0.6g/kg/day), and Study B 13-24mL/min (n=255) randomised to low-protein (0.58g/kg/day) or
very low protein (0.28g/kg/day).

- 51 -

In the most recent Cochrane review, eight eligible randomised-controlled studies were reviewed
comparing two levels of protein intake in patients with moderate to severe renal failure, on the
outcome of renal death (composite of death or commencement of dialysis during the follow-up of
between 12 and 24 months)

173

. Of the eight trials in non-diabetic patients, there were 1494 patients,

753 with restricted protein (moderate-severe 0.3g/kg/day to 0.6g/kg/day), 741 higher/unrestricted

protein intakes (>0.8g/kg/day). The review concluded that there was a 31% relative risk reduction in
renal death (dialysis commencement or death) (RR 0.69, 95% CI = 0.56 to 0.86 p<0.001) for
restricting protein intake to low/very-low levels 0.3 to 0.6g/kg, compared with no restriction. However, it
was acknowledged this meta-analysis was performed with a number of conflicting studies, with only
one significant positive study promoting protein restriction. This positive study was an RCT by Ihle et
al (1989) which demonstrated a significant reduction in renal death with severe protein restriction (OR
=0.34 (0.12 to 0.95)), however the statistical analysis of this trial is likely to be bias towards a positive
effect of protein restriction as it did not proceed as an intention-to-treat and withdrew non-compliant
participants from the analysis.

In another meta-analysis of 13 RCTs (n=1,919) investigating rate of GFR decline as the outcome
measure, Kasiske et al concluded that a low / very low protein restriction (0.3 with amino acids to
0.6g/kg/day) has a small effect on retarding the progression of renal failure

176

. The dietary protein

restriction reduced the rate of decline in estimated GFR by only 0.5 mL/min/yr (95% CI, 0.08 to 0.98
mL/min/year). This is significantly less than the reduction in the rate of decline demonstrated in metaanalysis for the use of lipid-lowering agents 1.9 (0.3 to 3.4) mL/min/year or angiotensin converting
enzyme inhibitors (ACE-I) alone 1.6 (1.2 to 1.9) mL/min/year.

There are significant limitations when considering the evidence from these studies. Firstly, regarding
the issue of compliance, and secondly, there is little focus on the impact of protein restriction on
nutritional status and patient-centred outcomes, for example, quality of life. These issues are explored
further below.

- 52 -

A common theme in the prescription of protein restricted diets is poor compliance.

review of protein restriction studies indicated compliance is less than 50%

A systematic

179

. The dietary prescription

in the above studies between a high/free protein intake and restricted intake prescribed a difference of
up to 0.7g/kg/day (e.g. MDRD study design prescribed a low protein intake of 0.53g/kg/day vs high
protein of 1.3g/kg/day)

180

. Whereas the actual difference in protein intake between the groups (as
179

determined using 24-hour urine) was only 0.2g/kg/day to 0.35g/kg/day

It is this gradient of

difference which represents the true therapeutic effect of protein restricted diets, not necessarily the
initial prescription provided.

Table 1-15 Protein restriction studies in pre-dialysis chronic kidney disease investigating
change in nutritional status
Study

Study details

Williams
181
1991

RCT
n=95,
follow-up
19
months

Damico
182
1994
Kopple
13
1997

RCT N=134

Cupisti,
91
2004

RCT (MDRD)
GFR
25-55
ml/min N=585;
GFR
<25
mL/min) n=255
CCT, n=28 nondiabetic
GFR
<15 mL/min; 28
healthy controls

Prescription
per
kg/day
(energy and
protein)
1) 0.6g/kg or 2)
>0.8g/kg ; Both
energy
>30cKal/kg
1) 0.6g/kg, 2) 1
g/kg
Study 1
0.6g; vs 1.3g
Study 2
0.28g vs 0.6

Outcome
measures

Results

Anthropometry
Dietary compliance

No change in nutrition
status

Anthropometry
Dietary compliance
Albumin
Weight
Arm muscle area,
%body fat

0.6g vs 0.3g +
amino acids

Biochemistry
Anthropometry
Bioimpedance
Skeletal
muscle
oxidation
and
strength

No change in nutrition
status. Poor compliance
Low and very-low protein,
rise in albumin, decrease
weight, body fat %, arm
muscle area. Related to
decline in energy intake
Lower protein decreased
strength and oxidation
skeletal muscles

RCT = Randomised-controlled trial; MDRD = Modified Diet in Renal Disease; GFR = Glomerular Filtration Rate

It is important to consider the effect of this diet on overall nutritional status. Few studies, as indicated
in table 1-10 in protein restriction have considered the impact on nutritional status

183

. When this is

conducted, the measures used such as weight, arm circumference and albumin have limited validity
for use as outcome measures in CKD, as detailed in Section 1.3 Assessment of nutritional status in
CKD. Overall, these studies are inconclusive regarding the effects of low protein intake and nutritional
status. The MDRD study indicated small yet significant declines in nutritional status (weight, body fat
and arm muscle area) in low protein and very low protein diets, compared with the usual protein diet.
Further to this, declines in ability of skeletal muscle to maintain oxidative metabolism and strength has

- 53 -

91

also been indicated with a low-protein diet compared with controls . Other studies have indicated no
change in anthropometry

181, 182

and conflicting changes in serum proteins

107, 184

In these studies, decrements in nutritional status occur in conjunction with recording of low energy
intakes. As demonstrated in the MDRD study, low (~0.6g/kg/day) protein and very low (~0.3g/kg/day)
protein intakes were associated with low energy intakes at 23 kcal/kg/day, 3-4 kcal/kg/day lower than
the usual protein intake group, despite the same energy prescription (of 35 kcal/kg/day). This is
despite regular (minimum monthly) contact with participants

185

Moreover studies which are

conducted over a prolonged period (several years) report subtle changes in the muscle mass or body
composition which occur at intakes of 0.6 g/kg IBW/day or less

184, 186

. Finally, despite the strong

interest and investment in protein restriction studies, there is a distinct lack of patient-centred
outcomes and gold-standard nutritional measures evaluating the effectiveness of this dietary
intervention.

Implications for effectiveness in terms of quality of life years has not to date been

investigated.

Overall, the evidence for low-protein diets retarding renal failure progression is conflicting and severely
179, 183

restrictive diets have poor compliance

. The beneficial effects of severe protein restriction are

small when compared with other reno-protective interventions, such as ACE-I or statins

187

. Therefore,

despite this being a highly researched topic, it appears that there is no greater benefit of restricting
protein below what is recommended for healthy adults

188

. This supports the most recent CARI

guidelines stating patients with progressive chronic kidney disease (CKD), who receive a proteinrestricted diet, the protein content should not be lower than 0.75 g per kilogram ideal body weight
(IBW) per day. The protein should be of at least 50% high biological value. An energy intake of at least
35 kcal/kg IBW/day to promote positive nitrogen balance must accompany a low protein diet. (Level II
evidence)

46, 48

1.5.1.2 Energy Prescription

Adequate energy intake is imperative to the maintenance of nitrogen balance. There is no evidence
19, 48

that energy (kJ) requirements of the CKD population are any different to the general population

Although it has been demonstrated that resting energy expenditure is not impacted by renal function

- 54 -

change, it has been indicated that sub clinical inflammation may impact on energy expenditure
19

therefore energy requirements .

In a study by Avesani et al (2004) in pre-dialysis patients the

difference in resting energy expenditure between the lowest and highest CRP tertiles equated to 123.7
kcal/d (P = 0.003; 95% confidence interval, 42.9 to 204.5) or less than 2 kcal /kg for an average 70kg
19

person .

In a retrospective study, by Kopple et al (1986) the role of energy intake in maintaining nutritional
status was demonstrated in a 24-day study of 16 stable pre-dialysis patients. In this study, each
participant was provided with a 0.6 g protein/kg/day diet with different energy prescriptions, 45, 35, 25
and 15kcal/kg/day.

Decline in nutritional parameters was observed only when energy intake

189

decreased to 25 kcal/day and below

. Therefore an intake of 35kcal/kg/day is recommended to

optimise nutritional status. A level of 3035kcal/kg/day is recommended by K/DOQI guidelines for

46, 48, 49

individuals 60 years and over, and by CARI guidelines for individuals with a sedentary lifestyle

46

Therefore, this equates to a goal intake of 125-145 kJ/kg/day for patients with pre-dialysis CKD.

1.5.1.3 Other dietary strategies in CKD

Aside from extensive studies in protein prescription, few other studies have been conducted regarding
nutrition support in the pre-dialysis CKD area; therefore, one must look to the literature on the dialysis
population, which has been extensively studied.

Appetite stimulants, such as megestrol acetate (a synthetic orally active derivative of the hormone
190

progesterone) have been investigated to stimulate appetite, and therefore improve energy intake

Few studies have investigated its use in HD, with varying results. In a case study, Burrowes et al
reported an increase in appetite and intake, affording an unfavourable body composition change
191

(increase in fat-mass, and decrease in fat-free mass) over the 24 week study

. These findings are

consistent with studies assessing body composition change after administration of megestrol acetate
192, 193

in oncology

headaches and

and HIV

194

dizziness,

. Significant side-effects such as diahorrea, confusion, hyperglycaemia,

reported when long term administration of megestrol acetate was

investigated by Boccanfuso et al

195

. Other side-effects have not been well studied but include

- 55 -

hypogonadism, impotence and increased risk of thromboembolism

22

Therefore, at this stage

megestrol acetate is not recommended for use in clinical practice in CKD.

The use of anabolic hormones and steroids has been investigated to manage malnutrition in dialysis.
In particular, use of growth hormone (GH) has shown positive outcomes in patients with GH
deficiency. In the studies to date, GH in HD improves nitrogen balance and IGF-1 levels, however,
sustainable benefits to nutritional status (muscle mass and function) remain unclear

196

. Studies into

the use of anabolic steroids (nandrolone decanoate) in HD are limited, however, have promising
results. Two studies, including one randomised double-blind trial, determined increase in body weight,
muscle mass and function in malnourished patients over 6 months

197, 198

. The efficacy, side-effects

and ideal dosage of this treatment in CKD patients however, are yet to be determined. Therefore,
both anabolic treatments require further long-term investigation before implementing into clinical
practice.

1.5.2 Dietary implementation studies in CKD

As opposed to the dietary prescription, implementation and delivery of nutrition intervention refers to
the type, duration content and location of intervention.

There are few studies investigating the

method of delivery of nutrition prescription to optimise nutritional status in the literature for CKD, as
outlined in table 1-16. As mentioned previously, most intervention studies are focused on nutrition
support in dialysis or protein prescription in pre-dialysis CKD, therefore, there is little evidence
informing pre-dialysis care.

As is clear from table 1-16 the majority of studies are focused on the dialysis population. In addition,
as per the limitations of protein prescription studies, there is little focus on valid nutrition assessment
measures or patient-centred outcomes. The most common outcome measure used is albumin, with
most studies reporting a significant improvement in albumin by providing intervention. The limitations
of albumin as an outcome measure, particularly in CKD are detailed in Section 1.3. Improvements in
weight and rating of SGA also features in a number of studies detailed in table 1-16.

- 56 -

n=41 HD pts randomly assigned to:

Supplement group n=26
Non-supp group n=14 group

Akpele,
204
2004

203

n=117 HD
High risk (HD-PNI 0.8) n=26
Low risk (HD-PNI 0.8) n=91
HD 3 times/wk; 18 years old;
functioning GIT

Steiber
AL 2003

Eligible:
HD for > 1 mo; BMI <20; Alb
<4.0g/dL

n = 40 MHD non-diabetic pts

- 57 -

HD

RCT 14
months

CCT (3
months)

RCT
1 month

HD

Sharma
MJ et al
202
2002

HD

Case-control
series, (3
months, 3
month
control)

HD

Caglar K
et al
201
2002

Eligible: low alb, av <37 for 3mth

pre 6 mo dialysis
Intervention n=52
Control n=31
n=85 HD pts
Eligible:
HD > 6 months; >18 years old
Alb 3.7g/dL or Prealb <30mg/dL

RCT
(6 months)
pilot

HD

HD n=83

Leon JB
et al
200
2001

Case series
(6 months)

Study
design

CKD

Stage

11 CRF pre-dial
CrCl <25

Sample

Cliffe M
et al
199
2001

Study

Albumin
24hr recall
Weight
Appetite
Karnofsky
index

HD-PNI
24hr recall
SGA

Albumin

Control n=14 Counselling

Treatment 1 n=16 Counselling
+ home-made supplement
Treatment 2 Counselling +
clinical supplement

High risk
oral supplement daily

Treatment
1-2 cans of Nepro/day +
normal diet

Low risk
usual care

Albumin,
Pre-albumin
SGA
Weight

SGA, Weight
Intake
MAMC, GS
Albumin,
I-GF
Albumin
Weight,
Dietary intake,

Outcome
measures

T0-T1 Received conventional

nutrition counselling
T1-T3 Oral supplements over
6 mo

Control Usual care

Treatment Education
addressing barriers intake

Aim energy intake at BMI 2025, protein at 0.8-1.0g/kg/d

phone calls every 3 months

Treatment

Rate of change in serum alb > in nonsupplement pts (dietary counselling alone)
than supplement pts.

Dry weight G1= 1.7kg **; G2= 1.8 kg*;

G3=NS
Albumin G1=NS; G2= 6.0g/L*;
G3=5.0g/L *
Karnofsy G1=NS, G2= 0.4*; G3= 0.4*
Intake (food-based): G1 236 kcal or 4
kcal/kg; G2&3 no change
Appetite: G1 improved on non-dialysis day
G2 & G3 no change
Change in HD-PNI scores: 0.50 vs.
0.21
Dietetic intervention with supplements pts
HD-PNI and pts risk for hospitalisation.

Albumin (T0-T3): 3.2 g/L *

Pre-albumin(T0-T3): 4.6 g/L **
SGA (T0-T3): 14% (median 4 to 5)
Dry weight (T0-T3): 3kg

Changes in albumin (P<0.001)

Significant alb with intervention group
Outcome measures: Alb, CRP
67% patients completed the trial

Maintain with 2 of 3 SGA B improving

Weight 1.9kg 4.5; ( 2.6%)
No change in biochemistry

Results
Change in Treatment vs. Control

Table 1-16 Randomised-controlled trials and controlled-clinical trials evaluating the implementation of nutritional care in chronic kidney disease

HD

HD,
PD,
CKD

HD

HD N = 163
Reporting on hypoalbumin groups
only
Treatment n=21(<3.8g/dL)
Control n=20(<3.8g/dL)

14 studies (2 RCTs, 5 CCTs, 6

historical control, 1 cohort)

Dialysis n=180
Eligible Alb <0.37g/dL (> 3 months)
or Alb <0.34g/dL
Treatment n=86
Control n=94

KalantarZadeh K
et al
206
2005

Stratton
207
2005

Leon J et
al
208
2006

RCT
(12 months)

Systematic
review

CCT
4 week pilot

RCT (6
months)

Study
design

Control Usual care

Treatment Education
addressing barriers intake

Oral supplements vs. routine

care

Control
Normal diet + intensive dietary
counselling
Treatment Routine review +
20-30mins monthly of diet
targeting phosphorus
Control
Routine review
Treatment One can Nepro,
One can oxepa on dialysis
Control = no supplements

Treatment

Albumin
Intake
Weight
Not available
for all studies
Albumin
Weight,
Dietary intake,

Albumin

Knowledge
Phosphorus
Calcium

Outcome
measures

Albumin 2.1 vs. 0.6 g/L**

Energy intake 4.1 vs. 0.6 kcal/d/kg*
Protein intake 0.13 vs. 0.06 g/d/kg*

Serum albumin 2.5g/L

Protein intake 0.1 g/kg
Weight 1.7kg (less weight loss)
Oral supplements were associated with
albumin, no change control
2.3 (0.4 - 4.2) g/L; OR 0.4 (0.1 - 0.8)
Energy and protein intake 20-50%
Weight ~3% in 3-4 months

Knowledge test scores 7%

Phosphorus G1< G2= 1.13mg/dL **
Calcium/phosphorus product 11.74**
PTH 26 g/dL

Albumin/month: 0.06 vs 0.04 g/L

Results
Change in Treatment vs. Control

- 58 -

*p<0.001; **p<0.001; P<0.05

Alb = Albumin; HD = Haemodialysis; RCT = Randomised Controlled Trial; PD = Peritoneal dialysis; mo = month; G = Group; PNI = Prognostic Nutrition Index; CrCl = creatinine clearance; CCT =
controlled clinical trial; SGA = Subjective Global Assessment; OR = Odds Ratio; MAMC = Mid-arm muscle circumference; GS = Grip strength; NS = Not significant

HD

Stage

N=63
Intervention n=32; Control n=31
High mean serum phosphorus level
>6.0mg/dL over 3 mo

>18yrs; dialysis 3 times/wk;

Albumin 3.5g/dL

Sample

Ford JC
et al
205
2004

Study

Studies of nutrition support via oral supplementation and enteral feeding are feature most prominently
in the dialysis intervention literature.

Disease-specific formulae with high energy density (up to

2kcal/mL); low potassium, phosphate and sodium are often used for oral supplements or enteral
feeding in dialysis. In a recent systematic review Stratton et al (2005)

207

comparing randomised and

controlled clinical trials revealed enteral nutrition support increased serum albumin of 2.27g/L (95% CI
0.37 to 4.18g/L), with generally little change in electrolyte concentrations, compared with routine care.
In addition, this review highlighted studies suggesting nutrition support can typically increase total
209-211

energy and protein intake by 20 to 50% when used to supplement an oral diet

. Only one study in

haemodialysis outpatients with patients with gastrostomy feeding showed an significant improvement
in nutritional status in the form of muscle mass

212

. Overall, comparison of the studies in nutrition

support is difficult due to the collection of inconsistent outcome measures. There is insufficient data to
suggest that oral/enteral nutrition supplementation improves clinical outcomes (QOL and nutritional
status) over routine care at this stage.

The implementation studies by Leon et al 2002 and 2006 in HD support patient-centred, structured
nutrition intervention. In both of the studies, nutrition intervention was tailored to identify and manage
patient-specific

barriers

(such

as

poor

nutritional

knowledge,

poor

appetite,

help

with

shopping/cooking, low fluid intake, inadequate dialysis, depression, difficulty chewing and swallowing,
GIT symptoms and acidosis).

Both the pilot (n=83) and full study (n=180) resulted in modest

improvements in albumin levels regardless of levels of inflammatory markers (Table 1-16)

204, 208

Although albumin is not an ideal outcome measure of nutritional status, it is important to note that the
improvement in albumin status has significant benefits in terms of its relationship to morbidity and
mortality. The outcome of the HD studies on nutrition investigations highlight overcoming patientspecific nutritional barriers and providing individual dietetic care has the potential to enhance survival
and decrease both hospitalisations and inpatient expenditures

200, 204

Evidence from the large multi-centre trials in protein restriction indicate frequent nutrition counselling
results in compliance with the intervention and improved outcomes

- 59 -

213

. The outcomes from the MDRD

study are that the most effective nutrition interventions involve patients in self-management skills and
frequent, ongoing feedback, and interventions with the nutrition team

185, 213, 214

This review of the literature and the contents of Table 1-16 demonstrate a clear lack of evidence in the
delivery of nutrition intervention studies aimed at optimising nutritional status prior to dialysis. One
small cohort study conducted by Cliffe et al 2001 had eleven patients with GFR < 25mL/min undergo
nutrition intervention. Each patient was followed up 3-monthly by phone following initial face-to-face
intervention. The intervention aimed to provide adequate energy intake to maintain or achieve a BMI
20-25, and protein intake of 0.8-1.0g/kg/d and counselling for six months
patients were mildly to moderately malnourished (SGA rating of B).

199

. At baseline, 3 of the

At six months, mean GFR

decreased by 3.5 +/- 2.3 mL/min, yet only 1 of the patients remained mildly to moderately
199

malnourished and none were severely malnourished

. The decline in renal function was not

199

significantly correlated to any nutritional parameter measured

. Whilst the number of patients in this

study was small, it supports the proposition that dietetic intervention with regular follow-up may
prevent declining nutritional status in patients with CKD. An Australian prospective observational study
in a cohort of pre-dialysis patients has shown that structured nutritional monitoring may maintain
215

nutritional parameters with deteriorating renal function

The recently published practice guidelines introduced in Section 1.2 recommend a blanket referral
46

system for every patient with GFR <30 ml/min with a review every 6 months . A recent survey of
dietetic practice in New South Wales revealed that review of pre-dialysis patients every 6 months was
being met by only 12.5% of facilities
onset of dialysis treatment

217

216

, with >70% of patients not being seen by a dietitian prior to the

Therefore, it appears that the implementation of recommended

guidelines is hindered by the lack of good evidence to support their implementation.

1.6

Summary of the evidence

In chronic kidney disease patients, the high prevalence of malnutrition and the strong association
between measures of malnutrition and mortality rate suggests the need for careful nutritional
monitoring and treatment of these individuals

218

. Malnutrition at the onset of dialysis indicates a

progressive deterioration in nutritional status along with deteriorating kidney function

41

. In particular,

poor nutritional management in these patients has demonstrated inadequate dietary intake,

- 60 -

specifically energy intake, which is identified as a key factor in the development of malnutrition

24

However, significant gaps still also exist in the body of knowledge relating to how best to assess
nutritional status and what types of interventions are likely to prevent or reverse malnutrition in
individuals with pre-dialysis CKD

219

. Therefore there is a need for well designed studies using gold-

standard reference measures to provide evidence for nutrition assessment methods and the
effectiveness of nutrition intervention specifically for pre-dialysis CKD patients.

- 61 -

Chapter 2
2.0

Study Design and Methods

Introduction

From a review of the literature, there is a need to improve the quality of evidence informing practice to
optimise nutritional status in chronic kidney disease patients prior to dialysis treatment. This chapter
will provide the research hypotheses, design and methods guiding this investigation.

2.1

Aims of the study

The aim of this study is to 1) provide a validation of clinical tools for the assessment of malnutrition in
pre-dialysis CKD (Phase I); and 2) to evaluate the impact of an individualised dietetic counselling
intervention in pre-dialysis patients on nutrition status (Primary, clinical outcome, body cell mass
(BCM) and subjective global assessment (SGA)), dietary intake (Intermediate outcome) and quality of
life (Secondary, patient-centred outcome) compared with standard nutrition care (Phase II).

2.1.1 Objectives and hypotheses

The following objectives and hypotheses are to be tested:

To establish the relationship between body cell mass (BCM) (derived from total body potassium
counting (TBK)) and subjective global assessment (SGA), for the assessment of malnutrition in predialysis CKD.

H1:

Patients assessed as malnourished by SGA (B moderately malnourished or C severely

malnourished) will have lower body cell mass (indexed for height) than those assessed as well
nourished (SGA A).

H2:

Body cell mass (indexed for height) provides an accurate assessment of nutritional status

compared with SGA

To investigate the modified versions of the SGA, as cross-sectional measures of nutritional status
compared to BCM in pre-dialysis CKD patients (criterion validity).

- 62 -

H3:

There is a relationship between BCM (indexed for height) and a) 7-point SGA categories, b)

Patient-Generated SGA scores, and c) Malnutrition-Inflammation Scores

2.1.1.2 Nutrition intervention methods

Primary outcome: Nutrition Status

To determine the impact of individualised dietetic counselling with structured follow-up on nutritional
status compared to standard nutrition care.

H3:

Patients who receive nutrition intervention will have maintained or improved nutritional status

as assessed by change in BCM, than those receiving standard care.

H4:

A greater proportion of patients receiving nutrition intervention will have maintained or

improved their nutritional status as assessed by change in SGA, compared to standard care.

Intermediate outcome: Dietary intake

H5:

A greater proportion of patients who receive the structured nutrition intervention will have

maintained or improved appropriate (0.75g/kg-1.0 g/kg) protein and energy intake (at least 125
kJ/kg)

46

compared to standard care.

Secondary outcome: Quality of Life

To determine the impact of individualised dietetic counselling with structured follow-up on quality of life
compared to standard nutrition care.

H6:

Patients who receive structured nutrition intervention will have maintained or improved rating

for health-related and kidney disease specific quality of life measured by KDQOL-SF
receiving standard care.

- 63 -

TM

than those

The above hypotheses are expressed as uni-directional statements, however, this study is designed to
investigate changes in both directions, therefore, the statistical analysis related to the above
statements will be 2-tailed (refer to 2.5 Statistical Analysis, for more detail).

2.2

Framework informing the study design

This figure is not available online.

Please consult the hardcopy thesis
available from the QUT Library

Figure 2-1 Cascade of events leading to evidence of effectiveness of Medical Nutrition Therapy
(MNT) (Splett, 1996)

220

The framework for the provision of nutrition management by Splett (Figure 2-1)

220

underpinned the

design of this series of investigations. Spletts model of nutrition intervention depicts a cascade of
events necessary to achieve positive patient and clinical outcomes

220

. Medical nutrition therapy (MNT)

defines a specific type of nutrition care, one that involves in-depth, comprehensive assessment and
individualised care

221

. Not only does this model, in the first two steps highlight the necessary process

and evidence required to address clinical questions as per Table 2-1, this model also provides a
framework for focusing on outcomes to determine the effectiveness of therapy. These outcomes
relate not just to morbidity or mortality, but also expected intermediate or surrogate outcomes (such as
dietary intake and biochemistry), leading to a variety of long-term outcomes (including clinical, such as
nutritional status, and patient-relevant outcomes such as quality of life).

- 64 -

Table 2-1 Steps in MNT cascade and corresponding clinical question for the nutrition
46

management of CKD patients (adapted from Ash et al )

Step in MNT Cascade

Appropriate
Access to
Necessary
Care

Quality
Nutrition Care

Stage
Criteria
Referral

Related Clinical Question

for

Nutrition
Assessment

Nutrition
Prescription
Intervention

Implementation
and
management

At what level of GFR should patients be

referred to the dietitian in order to maximise
nutritional intervention opportunities?
Which specific measures best reflects
nutritional status or change in nutritional
status?
What is
(are) appropriate
nutritional
intervention(s) to optimise nutritional status in
Chronic Kidney Disease and prevent
malnutrition?
What
is
the
optimal
method
of
implementation and follow-up to ensure
nutritional status is maintained or
improved?

This medical nutrition therapy model was selected for use as the framework to this study design due to
its focus on a variety of patient outcomes. The original cascade by Splett (1996)

220

has been adapted

above to depict this particular study, with both the research questions and outcomes of interest to be
assessed. The clinical questions of interest in this investigation are highlighted in bold in table 2-1
relating to nutrition assessment and implementation and management of nutrition intervention.

2.3

Outcome measures

2.3.1 Primary Outcome: Nutritional status

Nutritional status was determined using total body potassium (TBK) to measure BCM and subjective
global assessment (SGA).

2.3.1.2 Body cell mass

Nutritional status is represented using BCM, the mass of the metabolically active cells of the body.
TBK, a gold-standard measure of BCM, was performed by a shadow shield whole-body counter of the
naturally occurring isotope

40

K (Accuscan; Canberra Industries, Meriden, CT). This isotope represents

a fixed proportion of naturally occurring potassium detected by the 1.46-MeV gamma ray emitted by
40

K. Potassium is 98% intracellular cation, therefore, whole-body counts of

40

K reflects BCM with high

67

precision . Two 1100-s scans were performed for each subject. All personal metallic objects having
been removed, including jewellery and belts; the subject was required to lie supine on a scanning bed
which moved under the detectors (approx 30 cms deep). The measurements of TBK were conducted

- 65 -

at the Childrens Nutrition Research Centre, Body Composition Laboratory, situated at the Royal
Childrens Hospital, Brisbane. The assessments were conducted by a technical assistant, with the
candidate in attendance, blinded to the TBK results. Subjects were instructed that the procedure can
be stopped at any time if they became uncomfortable. Only subjects who could comfortably fit under
the sensor completed the TBK assessment. Background and sensitivity checks were completed daily
and considered in each measurement.

Body cell mass was calculated directly from TBK by the equation by Cohn et al (1985) (BCM (kg) =
0.0092 x TBK (mmol))

222

. This conversion of TBK to BCM was undertaken for this thesis for ease of

interpretation and applicability of the results.

The author acknowledges that by creating a direct

conversion of TBK to BCM assumes a consistent concentration of

(herein referred to as BCM).

40

K in metabolically active tissues

Recent investigations by Wang et al (2004) into the cellular body

composition support this assumption and render the equation BCM (kg) = 0.0092 x TBK (mmol) valid
for the conversion of

40

K to BCM across a range of age groups and for both males and females

223

In the first phase of the study investigating nutrition assessment methods, body cell mass (kg) was
3.5

adjusted for height (m ). This was undertaken to provide an assessment of the quantity of BCM,
224

independent of height, eliminating variation due to body size

. The analysis to establish the body cell

3.5

mass index (BCM-I) equation (kg/m ) was undertaken using a log-log regression of BCM (kg) and
Height (cm) to ensure all variation in BCM attributable to difference in height was accounted for. The
development for this equation is provided in Appendix C Development of Body Cell Mass Index.

2.3.1.3 Subjective global assessment

96

Nutrition status was also assessed simultaneously using subjective global assessment (SGA) , to
provide the diagnostic standard for nutrition status assessment. SGA, in its original form, classifies
patients into category A (well nourished), B (moderately or suspected of being malnourished) or C
(severely malnourished) and is a well validated tool for diagnosis and prognosis of nutrition status in
CKD (Appendix E Data collection and nutrition assessment tools). The assessment is based on
components of a medical history (changes in weight, dietary intake, gastrointestinal symptoms with
nutritional impact and functional capacity) and a physical examination (seeks to identify loss of
subcutaneous fat and muscle wasting).

- 66 -

As was clear form the previous chapter, a number of modified SGA tools, based on the components of
the original SGA assessment, exist in the CKD literature. These tools are commonly used in CKD, but
not sufficiently validated, nor previously investigated in pre-dialysis CKD against criterion, or reference
methods

225

29

. Therefore, in Phase 1, 7-point SGA , and scored tools, MIS

114

226

and PG-SGA

were also

collected for investigation against the reference method, body cell mass. The 7-point SGA is an
expansion of the rating categories for nutrition status of the original SGA, from 3 (A, B and C) to 7
29

(rating 1 severely malnourished to 7 well nourished) . The MIS scores each of the parameters of the
SGA (0-3), and in addition, BMI, albumin and total iron-binding capacity, providing an additive score of
114

0 to 30

. Higher score indicates greater nutritional risk. The rating of the MIS was modified in this

study from the initial protocol by removing dialysis vintage within the co-morbidity scoring component
with permission granted from the author (personal communication). The PG-SGA assesses a broad
range of parameters based on the SGA, focusing on acute changes, scoring each component (0-4), to
give an additive score from 0 up to 50, along with the original global rating (A, B or C)

226

. Again, a

higher score indicated greater risk of malnutrition.

These modified SGA tools aim to better distinguish change in nutrition status. Each of the tools are
supplied in Appendix E Data Collection and Nutrition Assessment tools. The outcome of the Phase 1
will inform the outcome measures to be used in conjunction with SGA for the assessment of nutrition
status in Phase 2.

The assessments were performed simultaneously by the candidate only (KC), a dietitan trained and
experienced in using the SGA-based tools, according to the protocol

96, 114, 226, 227

2.3.2 Intermediate outcome: Dietary intake

Dietary intake was assessed using a three-day food record. Food records were to be completed in the
week prior to both baseline (week 0) and follow-up (Week 12) appointments. The records were to
include two weekdays and one weekend day, which were not required to be consecutive. Subjects
were requested to estimate or measure all food and fluids consumed for those three days.
Instructions in the booklet and from the dietitian covered estimating portion sizes, detailing cooking

- 67 -

methods and providing food labels where possible. If a recipe was prepared, the subjects were to
provide the ingredients list and serve size. Upon collection, all food records were verified by the
dietitan with visual food models and household measures to ensure accurate records at baseline and
follow-up assessment appointments.

The food records were then entered into nutrient analysis software FoodWorks (Professional Version
3.02 Xyris Software, Brisbane, Australia) using computerised food composition data, which are based
on the Australian Nutrient Database (AusNut, Department of Human Services and Health, Canberra).
The food records were coded by selecting the appropriate food and portion size consumed. The
software produced a daily average nutrient consumption by calculating the mean intake over the three
days of recording. Mean protein and energy intake and the proportion of macronutrient contribution to
energy intake were transferred by hand to the statistical software for analysis.

The Goldberg cut-off was used to determine the proportion of under-reporting of energy intake
amongst participants

124

. This method is simple, less expensive and more practical than using the
73

gold-standard doubly labelled water technique . This method was used to compare the daily mean
reported energy intake with 1.3 x BEE estimated by Harris-Benedict equation, on weight stable
participants only. It determines the participants who are low energy reporters.

2.3.3 Exploratory outcomes: Clinical variables

Variables that were clinically available and considered to in-directly influence the development of
uraemic malnutrition and/or identified as parameters of interest for nutritional intervention were
collected where available for participants in this study. Blood samples were taken after an overnight
fast as part of routine care.

The clinical measures of inflammation (CRP), metabolic acidosis

(Bicarbonate) and anaemia (Transferrin saturation) were considered as exploratory variables, in

addition to several routine biochemical measures (detailed for collection by practice guidelines
(Appendix B). Biochemical analysis including albumin by BromoCresol Purple (normal range = 35
48 g/L), C-reactive protein (CRP, high sensitive, only available when ordered by treating nephrologist;
normal range 0 5 g/L), potassium (K+), bicarbonate (CO2), phosphate (PO4), urea, creatinine and
iron studies (including % Transferrin Saturation = serum iron x 100 / total iron biding capacity) was

- 68 -

undertaken at a central laboratory. Estimated GFR was calculated using the abbreviated (4-variable)
228

MDRD equation

with calibrated serum creatinine concentrations.

Biochemical variables were

reported according to the mean value and also proportion meeting normal range, as specified by
guidelines CKD practice guidelines

8, 46, 48

2.3.4 Secondary outcome: Quality of life

Two basic approaches to QOL measures are to use generic instruments that provide a summary of
health-related quality of life, and/or the use of specific instruments that focus on problems associated
with single disease states. Below is a summary of the most commonly used QOL tools, both generic
and kidney disease specific, that have been validated for use in CKD.

2.3.4.1 Generic quality of life assessment

Karnofsky Performance Index (KPI) was first published in 1949 and is still used in recent studies as an
indicator of functional capacity

229

As this is a provider-completed, objective tool of functional

performance it is not a true measure of global QOL. Therefore, KPI is often used in combination with
other generic and disease specific tools to measure QOL. KPI has demonstrated ability to predict
survival in dialysis patients

230-233

. Another generic tool, which complements the KPI by measuring

global QOL through behaviour-based indices, is the Sickness Impact Profile (SIP). SIP is another
provider administered questionnaire of 136 items to evaluate dysfunctional behaviour in 12 activity
categories. SIP has shown reliability, validity and responsiveness in dialysis and transplant patients
234

The final generic tool that is well-researched in CKD is the Short Form 36 (SF-36) which was
235

developed from the Medical Outcomes Survey

This tool includes 36 items assessing eight

dimensions of functioning and well-being. In each dimension, the respondent receives a score from 0
to 100 with the higher the score, the better the health. The reliability, validity, and sensitivity of the test
have been established in patients with chronic kidney disease, and there is growing experience using
this questionnaire to collect information from renal patients

150

. The 2 component summary scores of

the SF-36, physical component summary (PCS) and mental component summary (MCS), can predict
mortality and morbidity rates among dialysis patients
- 69 -

139, 141, 150

. The breadth of information provided

and comparatively short completion time (5-10 minutes, verses 25 minutes for the SIP) of the SF-36,
gives it a distinct advantage for use in large scale studies and for clinical purposes.

2.3.4.2 Kidney disease specific quality of life instruments

The Kidney Disease Questionnaire (KDQ), consists of 26 questions in five dimensions (physical
symptoms, fatigue, depression, relationships with others, and frustration)

236

. The KDQ showed

reliability, construct validity and responsiveness in a randomised-controlled double-blind EPO trial for
anaemic haemodialysis patients

236

. The major limitation of this instrument is it is specific for patients

on haemodialysis, and it does not contain an overall rating of health rating, which would provide a
global assessment of QOL.

Like the KDQ, the KDQOL-SF

TM

was developed for CKD patients undergoing dialysis. KDQOL-SF

TM

combines the SF-36, with a kidney disease-specific module . The disease-specific part includes 43
items directed at the kidney disease (symptoms/problems, effects of kidney disease on daily life,
burden of kidney disease, cognitive function, work status, sexual function, quality of social interaction,
sleep). Also included are multi-item measures of social support, dialysis staff encouragement, and
patient satisfaction, as well as an overall rating of health (a response scale of 0 to 10 ranging from
"worst possible" to "perfect health"). Each question is precoded numerically, and then transformed into
a scale of 0 to 100; the highest values reflect better QOL. The reliability and validity of the KDQOLSF

TM

was proven at in its development have been supported by various studies

4, 237, 238

Use of a combination of a generic and a disease-targeted questionnaire, as contained in the KDQOLSF

TM

234

is likely to be more relevant and responsive to clinically important changes in QOL

this reason that KDQOL-SF

TM

. It is for

version 1.3 is utilised in the largest QOL in CKD study to date, the

Dialysis Outcomes and Practice Study (DOPPS). This is a large international prospective QOL study
of haemodialysis patients in the US and Europe. Results from this study have confirmed its capacity
239

as a predictor for hospitalization and mortality.

TM

. KDQOL-1.3

surveys are completed at baseline

and annually thereafter. Data from the DOPPS study indicated that a higher physical component
score (PCS), mental component score (MCS), and kidney disease score (KDCS) is associated with a
5% to 8% reduction in the risk for hospitalization and a 9% to 23% reduction in mortality rate

- 70 -

239

These results are supported by a large number of previous studies in a range of dialysis populations,
detailed previously in 1.4 QOL assessment in CKD

139-141, 150

Quality of life was measured by Kidney Disease Quality of Life Short Form version 1.3 (KDQOL-SF

TM

v1.3 RAND University) for the investigation featured in this thesis. This tool required minor
modification for use in pre-dialysis patients. Specifically, changing the wording for satisfaction with
care from "kidney dialysis" to kidney disease (item 23), and omitting the questions about dialysis staff
encouragement

The KDQOL-SF

and

TM

support

(items

24A

and

24B)

240

v1.3 was provided to each subject prior to the baseline and follow-up assessments,

together with the food record booklets.

As this was developed as a self-administered survey,

wherever possible, subjects were instructed to complete alone. If they were unable to complete the
survey themselves, subjects were instructed to bring the survey to their assessment appointment. At
this time, the candidate would administer the survey according to the interviewer-administered
survey methods from the KDQOL-SF

TM

241

v1.3 user manual

The scoring spreadsheet for KDQOL-SF

TM

v1.3 was downloaded to Microsoft Office Excel 2003 from

the KDQOL webpage (http://gim.med.ucla.edu/kdqol/downloads/download.html, accessed February

20, 2005). Data from individual surveys were input into this spreadsheet. The direct answered were
then converted to summary scores for each component on a separate spreadsheet.

Following this,

the established QOL summary scores for each sub-scale were manually input into the main SPSS
database. For comparison, the KDQOL-SF

TM

results from the US dialysis population in the DOPPS

151

study were used for comparison to the study population at baseline

2.4

Research design and methodology

The first phase of this study was an investigation into nutrition assessment methods, a cross-sectional
design from the baseline (week 0) data collection. The second phase was to investigate nutrition
intervention methods. This was undertaken as a 12-week, prospective randomised-controlled trial of
individualised nutrition counselling with regular follow-up compared with standard nutrition care.

- 71 -

2.4.1 Study setting

This study was set in the pre-dialysis outpatient clinic of the Department of Renal Medicine, Royal
Brisbane and Womens Hospital (RBWH), located in Brisbane, Queensland, Australia. The RBWH is
the largest, tertiary-referral teaching hospital in Queensland. The Department of Renal Medicine at
RBWH conducted a weekly pre-dialysis clinic, for the medical management and planning for patients
in Stage IV and V CKD, in preparation for future dialysis treatment.

The usual care provided by this clinic involved a registered nurse weighing the patient, preparing
pathology reports taken the week prior and estimating renal function using MDRD equation (using
creatinine, age and gender

228

). This is followed by a medical review by the patients nephrologist. At

the commencement of the study there was no established treatment pathway for the nutrition care of
patients until the patient reached dialysis therapy. Dietitian appointments were provided on a referralonly basis if deemed necessary by the nephrologist or if requested by the patient. Therefore, nutrition
support was provided on an ad-hoc and limited basis.

2.4.2 Target population

The target population for this study was non-dialysing Stage IV CKD patients (as defined by a GFR
228

(estimated by MDRD equation

) < 30ml/min, and >15ml/min for > 3 months duration), undergoing

pre-dialysis treatment. The following exclusion criteria applied:

a) < 18 years of age
b) Patients with previous individualised dietetic intervention for Stage IV CKD
c) Expected to require dialysis within the next 3 to 6 months
d) Patients with emotional, cognitive or language limitations impairing their ability to provide
informed consent or comprehend dietary education, without a carer or interpreter available at
home (determined from a nephrologists assessment).
e) Patients who have established malnutrition, where the aetiology is other than renal disease,
for example, cancer or cardiac cachexia.

2.4.3 Sample size considerations

The estimated number of participants required was calculated based on data from a study by
Castaneda et al (2001) quantifying the reduction in TBK (precursor for BCM) of Stage IV CKD patients

- 72 -

with usual care (n=12), over 12 weeks. This study reported an absolute change (-5g ( 5)) in TBK,
with a mean TBK value of 101.2g TBK, therefore, a change in 5g of TBK represents approximately a
242

5% change

. Based on the premise that the minimum benefit that nutrition intervention will achieve

is weight/muscle mass maintenance, with a 5% TBK difference between the two groups, the sample
size was calculated by comparing the two means (TBK maintenance vs. 5% loss). Therefore, for a
significance of 0.05, power of 80%, two-tailed, detectable difference of 5% and a standard deviation of
5, 20 participants were required for each group. To account for an estimated 30% attrition and an
additional 15% to allow for discrepancies in the randomisation, 30 participants are required for each
group, 60 in total.

2.4.4 Study Sample

The patients were sampled consecutively from the target population of the Department of Renal
Medicine pre-dialysis outpatient clinic. This was conducted over a 13-month period, with an additional
5 months to follow-up the last recruits for baseline (week 0) and follow-up (week 12) assessment
appointments.

2.4.5 Recruitment methods

Prior to their outpatient appointments, patients medical charts were screened by the candidate (KC) to
identify eligibility prior to the patients initial pre-dialysis clinic appointment and were subsequently
identified to the treating nephrologists as a potential participant. The nephrologists then confirmed
eligibility, informed the patient of the study and requested the patient meet with the candidate (KC).
Eligible patients were then provided with a patient information package and a verbal explanation of the
purpose and methods of the study and also an outline of any risks or inconvenience that might occur
when partaking in the study by the candidate (See Appendix D). Informed consent was sought at this
time.

Throughout the recruitment period August 2004, to September 2005, 146 attendees of the RBWH predialysis clinic were screened for eligibility using the criteria described in 2.4.2 Target population. Of
these, 75 (51%) did not meet the inclusion criteria. Of those excluded, the majority (54.5%) did not
meet the level of kidney function under investigation, either too high (GFR>30ml/min) or too low (GFR
<15ml/min); 34% had previously been seen, or had been referred to a dietitian for Stage 4 CKD,

- 73 -

11.5% were excluded due to communication difficulties in the home, therefore, impeding the ability to
undertake telephone follow-up and 2 were excluded due to malnutrition of aetiology other than
symptoms of CKD. A further 5 attendees were subsequently transferred to other units following their
initial appointment, and no longer under the care of RBWH. Of the 66 patients eligible, 4 refused to
participate in the study, resulting in 62 providing consent, a response rate of 94% of eligible patients.

2.4.6 Data collection

Data collection was conducted solely by the candidate (KC), removing the need for training staff or for
establishing inter-rater reliability.

The candidate was also the dietitian providing the nutrition

intervention. Tools used to measure specific outcomes were previously discussed in 2.3 Outcome
measures.

Participants initial data was collected at the time of providing informed consent.

Demographic

characteristics (gender, date of birth), a brief medical and social history (to be detailed below), height
(stadiometer (Holtain), to the nearest 0.5cm), weight (scales (Tanita Wedderburn BWB-600) the
nearest 0.1kg) and a nutrition assessment using SGA were collected at this time. Weight history for
the previous 6 months was self-reported and where possible verified from the medical record.
Participants were then scheduled for the baseline assessment appointment.

A fortnight prior to baseline and again prior to follow-up (Week 12) appointments, participants were
provided with a 3-day food record, KDQOL-SF

TM

survey and appointment reminders.

When

requested, all effort was made to streamline patients baseline and follow-up study appointment with
their nephrologist appointment, to minimise burden on the participant.

As a result, a delay was

experienced between consent and baseline assessment, but the majority were seen within 6 weeks
following consent (median 5.0 weeks (range 0-15)).

Baseline and follow-up assessments involved the collection of outcome measures TBK, SGA,
TM

verification of 3-day food records and collection of KDQOL-SF . In addition routine biochemistry
ordered for nephrology assessment was taken within a week of both week 0 and week 12
assessments.

Further demographic information was collected on participants.

- 74 -

Age (years) was

determined as the participants age on the day of consent.

Socio-demographic information on

participants living arrangements (alone, with partner, family/friends or institution), employment status,
and responsibility of cooking and shopping (self or others) was sourced directly from participants
and/or carers. In addition, the number, duration and primary reason for hospitalisations up to 6 months
prior to baseline assessment was recorded for each participant from medical records admissions data
from the Royal Brisbane and Womens Hospital.

The presence of co-morbidities, smoking history, presence of diabetes and cause of CKD were
5

classified as per ANZDATA . Co-morbidities were also assessed as an index. Charlson index is a
243

weighted index for co morbidity and age class, derived from an estimated risk of 1-year mortality

From 40 years onwards, an increase in age results in one additional point per decade. The weights for
the different co morbid conditions range from one to six points. The weights of all co morbidities that
are present and the concurrent age score are summed to obtain a final score.

This index has been

widely used to adjust for prognostic differences in clinical studies, and validated for use in CKD
245

The participants co morbidities were established from an audit of recording of ICD-10 coding

244

from

previous hospital admissions to Royal Brisbane and Womens Hospital, and verified with the patients
medical chart.

2.4.7 Phase I: Nutrition assessment methods

Phase I of the study investigating nutrition assessment methods proceeded from the cross-sectional
data collection for week 0.

This assessment to answer Hypothesis 1 and 2, regarding nutrition

assessment methods

The first aim for the Phase I study pertains to the relationship between SGA and BCM.

In this

investigation, SGA was the diagnostic standard for the assessment of nutritional status. The validity of
SGA for nutritional status diagnosis (well-nourished (SGA A) versus malnourished (SGA B and C)) is
well established in CKD, as noted in Section 1.3.4 Clinical assessment tools, and detailed in a recent
publication from this thesis (Appendix A Manuscripts and conference presentations related to the
thesis)

225

3.5

. In order to undertake this assessment, body cell mass (kg) was adjusted for height (m ).

This enables an assessment of the quality of fat-free mass, therefore identifying nutritional status

- 75 -

224

independent of height, eliminating variation due to body size

. The analysis to establish the body cell

3.5

mass index (BCM-I) equation (kg/m ) is provided in Appendix C.

The accuracy of BCM-I as a diagnostic tool against SGA was established by assessment of the area
246

under a Receiver Operating Characteristic (ROC) curve

. This ROC curve was also used to provide

a cut-off point of BCM-I to diagnose well-nourished and malnourished patients. This graphical method
provided a spectrum of cut-off points of BCM-I and their relative sensitivity and specificity to assess
nutrition status (by SGA). The cut-off point selected to aim for the highest sensitivity and reasonable
specificity. It was established that the consequences of a higher number of false positives (high
sensitivity) was more important than a large number of false negatives (high specificity)

246

The second component of the Phase I investigation was to establish the validity of use of modified
SGA tools for the assessment of nutritional status. In order to establish this, each of the tools (7-point
SGA, PG-SGA score and MIS) were compared against BCM-I as a continuous variable. Where a
relationship existed, the tools were further investigated for diagnostic capacity by ROC analysis using
the BCM-I cut-off established in the previous investigation.

2.4.8 Phase II: Randomised-controlled trial of nutrition intervention

The intervention phase of this study used a randomised-controlled trial approach, comparing best
practice nutrition intervention with standard care treatment. The length of the intervention was 12
weeks following baseline assessment. This time period was selected on the basis that it is of
sufficient duration to address the study hypotheses and to ensure the patients are captured in the
pre-dialysis phase prior to initiation of dialysis.

2.4.8.1 Randomisation
Participants were randomly allocated to receive either intervention or standard care via a computergenerated number sequence (Figure 2-2 Flowchart of study progression). The sequence was under
the care of the project supervisor (SA, PhD supervisor) and concealed to the candidate (KC) recruiting
the participants. Once a participant consented to the study and was provided a study ID number
(consecutive), the project supervisor (SA) advised the candidate (KC) of the corresponding allocation
for that participant. The participant was not informed of their allocation until following the baseline

- 76 -

data collection, when the participant was either provided with the standard care or intervention
treatment.

Figure 2-2 Flowchart of participant progression in RCT of nutrition intervention in pre-dialysis

CKD
Assessed for eligibility (n=146)
Excluded (n= 84)
Not meeting inclusion criteria (n= 75)
Refused to participate (n= 4)
Transferred from facilitys care (n= 5)
Randomized
n= 62

Allocated to Standard care (n= 30)

Allocated to Intervention (n= 32)

Received allocated intervention (n= 29)
Did not receive allocated intervention
(n= 3)
Reasons:
Voluntary dropout prior to Week 0 = 2
Confirmed non-eligible after allocating = 1

Received
Allocation

n=56

Received allocated intervention (n= 27)

Did not receive allocated intervention
(n= 3)
Reasons:
Dialysis/transplant prior to Week 0 n=2
Confirmed non-eligible after allocating = 1

Completed follow-up (n=24)

Completed

Completed follow-up (n=26)

Lost to follow-up (n=5)

Follow-Up

Lost to follow-up (n= 1)

Reasons:
Deceased during intervention n=4
Dialysis during intervention n=1

n=50

Completed study (n= 26)

Completed study (n= 24)

Incomplete analysis (n= 6)
TBK (n=4)
QoL (n=2)

Reasons:
Dialysis during intervention n=1

Analysis

- 77 -

Incomplete analysis (n=8)

TBK (n=5)
QoL (n=3)
Food record (n=2)

2.4.8.2 Standard care treatment

Those randomised to the standard care group were provided with a booklet developed by the
candidate (KC) specifically for this intervention. This booklet contains overview of nutrition advice for
chronic kidney disease and co-morbidity management. Table 2-2 provides an outline of with Table of
Contents (See Appendix F for full booklet). Given the limited evidence base for dietetic intervention in
Stage IV CKD, at the time of this study, there was a lack of consistency of what constituted standard
care. In Australian practice, individualised education is not provided as standard

216, 217

. At the

institution where this study was carried out, referral to a dietitian was provided on an ad hoc basis, or
at best, written material on various CKD issues, including basic nutrition was provided by a nurse
educator. Considering the ad hoc nature of standard care, the proposal for the standard care arm was
the provision of generic but consistent, education material, based on principles from evidence-based
practice guidelines (Table 2-2 and Appendix F). Therefore, participants in the standard care group
received generic information only and did not receive any individualised advice.
Table 2-2Extract of the Table of Contents for the Nutrition in Chronic Kidney Disease
education booklets for the intervention and standard care groups
Topic
Introduction
Controlling your protein
Maintaining weight
Other nutrients
Fluid
Sodium
Potassium
Phosphate
Your questions answered
Sample menu
Your Nutrition Plan
Meat and Alternatives
Milk and other dairy products
Breads and Cereals
Vegetables
Fruit
What about Exercise?
Appointment Schedule

Intervention

Standard Care

X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X

X
X
X
X
X
X
X
X
X
X
-

Booklets developed by the candidate (KC) specifically for this study. Full versions available in Appendix F

2.4.8.3 Intervention treatment

Participants receiving intervention were provided with an initial individual dietitian consultation (by the
candidate) to meet individual protein and energy requirements, and manage electrolytes, fluid and comorbidities as per the Evidence-Based Practice Guidelines for Nutrition Management in Chronic
- 78 -

46

Kidney Disease .

The background information and tools for meal planning was provided in the

Intervention Booklet (Appendix F).

Content of intervention
The content of the initial and review interventions were based on that of the American Dietetic
50, 247

Association treatment protocol

. This protocol is summarised in Figure 2-3 and presented in more

detail in Appendix F. The focus of the intervention was on self-management strategies to improve
dietary compliance with the above-mentioned guidelines.

Strategies included goal setting, self-

monitoring techniques, addressing potential barriers to change and stimulus control (e.g., managing a
poor appetite, for example). An overview of the common goals, recommended strategies and
expected outcomes provided in Appendix F, with some example in Figure 2-4. In addition to dietary
intervention, encouragement to increase opportunities for incidental activity and recommendations
from the National Physical Activity Guidelines for Australians (Department of Health and Aging,
Commonwealth of Australia, May 1999) were provided to supplement the dietary changes prescribed.

Clinical Data

Interview

Initial

Medical history

Treatment plan

Anthropometry

Biochemistry
Follow-up

Changes in medical
treatment, medications
New Biochemistry

Initial

SGA & Appetite rating

Evaluate food record

Functional ability/exercise

Psychosocial and financial

issues

Readiness to change
Follow-up

24-hour recall
Recall of changes made

Expected outcomes:
Meets set goals
Demonstrates
appropriate food choice
Maintains body weight,
muscle and fat stores
Biochemistry within range

Self-management training:

Educate on identifying
protein, energy, other
nutrients were necessary

Recipe modification

Label reading

Determine treatment
plan:

Discuss role and

effect diet, etc on
renal disease

Nutrition prescription
Goals and target

strategies (Figure x-x

for examples)

Figure 2-3 Summarised protocol used in this study for the intervention of pre-dialysis CKD
patients adapted from ADA intervention protocol

- 79 -

The content of the review intervention sessions involved targeting the self-management skills, reassessing participants goals including problem solving for meal planning and preparation, food
shopping and label reading.
Example Goals / Objective:

Example Strategies:

Control protein intake (0.75-1.0g/kg/day):

Improve knowledge of protein containing foods
Identify appropriate intake of protein containing
foods

Optimise
nutritional
status

Attain optimal intake to maintain weight:

Combat poor appetite and intake
Consume appropriate high energy foods
where necessary

Keep protein-rich portion as a side-dish and bulk

up meals with vegetable and carbohydrate
Have small amounts of protien-containing foods
through day.

Keep a regular eating pattern, try small frequent

meals eg try to eat something every 2 hours
Have pre-prepared foods available (EG, cook in large
quantities and store in freezer, etc)
Avoid fluids at meal times;

To follow the national physical activity

guidelines:
To motivate increasing incidental activity.
Increase the amount of movement each day

Go for a 10-15min walk after dinner/before breakfast

Wear comfortable shoes, increase incidental activity
eg Walk the extra distance from bus/car
Do some work in your backyard or garden

Maintain serum potassium in safe range:

Intake meets guidelines for potassium
Identify high potassium containing foods
Normalises restrictions

Investigate medications (ACE-I, resonium etc) and

monitor changes
Only restrict if K+ >5.5 over time or >6.0 once
Count K+ points for each day up to the previous day
Allow 10 points for Fruit and Vegetables each,
concentrate intake on the lowest potassium foods

Figure 2-4 Example goals, objects and strategies used in the intervention of pre-dialysis CKD
patients

Monitoring and evaluation strategies

The intervention group received 4 counselling sessions via telephone for 10 weeks following the
baseline assessment and consultation, using the structured protocol summarised in Figure 2-3.
Telephone consultations were scheduled one, three, six and nine weeks following the initial
consultation. Additional calls were made where necessary, as determined by the dietitian if pressing
issues needed addressing (e.g. referral to other health professionals). Patient-initiated contact was
also facilitated by providing the patient with the candidates phone details at RBWH. If participants
had scheduled appointments in the Renal Medicine Department during this period; effort was made to
schedule a face-to-face consultation in place of the scheduled telephone session.

- 80 -

Using the telephone to deliver the follow-up intervention was chosen as it minimised the logistical
barriers associated with this level of contact between the dietitian and the patient. The telephone
consultations were structured to the protocol provided in Appendix F.

Firstly, participants were

requested to re-iterate goals for change and identify specific strategies implemented from the previous
session. Following this, a 24-hour recall, using the United States Department of Agriculture multiple248

pass method

was conducted. This provided a form of monitoring to inform and track the patients

dietary intake, and enabled verification of changes implemented to meet dietary goals. The use of
self-management techniques was modelled from a previous nutrition intervention study in this
213

population, the Modification of Diet and Renal Disease Study

. Either the patients underwent the

telephone intervention exclusively, and/or their primary caregiver, depending on the individual
situation.

If intervention participants were having difficulty meeting their nutrition prescription, the dietitian would
provide the participant and/or their carer with details of CKD-specific oral nutrition supplement, which
the participants were to purchase themselves. These nutrition supplements were recommended only
to those in the intervention group who were experiencing difficulties meeting their requirements with
regular foods or losing weight. A typical recommendation was to consume one to two 125ml tetra
packs per day (providing equivalent of 1000 to 2000 kJ and 4 to 8g protein per day).

Education materials for both the intervention and standard care groups were developed specifically for
use in this study. These were developed by the candidate due to lack of appropriate education
material upon initiation of this project. The materials were designed to adhere to the readability
principles guided by McLaughlin (1969)

249

. The content of both booklets are provided in Appendix F

and G. Table 2-2 provides an extract of the table of contents for each publication.

Contact details for the candidate (a dietitian) and project supervisor were provided to all participants
so that if any participants (intervention or standard care) had additional concerns then contact could
be patient-initiated. If nutrition-related problems not related to malnutrition (protein/energy intake)
became apparent during the course of the trial, as identified by the participants nephrologist, the
participants were supplied with the relevant standard dietary information sheets from RBWH.

- 81 -

2.4.8.4 Following intervention

On the follow-up assessment (week 12), the dietitian discussed with all participants the importance of
maintaining adequate oral intake prior to commencement of renal replacement therapy. Following the
study, patients were encouraged to request further dietitian appointments from the nephrologists if
they began to feel symptoms inhibiting sufficient oral intake or experiencing unintentional weight loss.

2.5

Statistical analysis

Statistical analysis was carried out using SPSS Version 13 (SPSS Inc, Chicago, IL, USA).

All

variables were measured at baseline and again at week 12. Frequency distributions were constructed
to check for outliers and ratifications were made with original data sheet where necessary.

All

continuous variables were tested for normality. When normally distributed, mean (sd) and parametric
statistics were utilised.

Where a not-normal or skewed distribution existed, median (range) were

reported descriptively and non-parametric statistics were used.

Statistical significance was set at a conventional 95% level (2-tailed). However, interpretation to the
data is focused on effect size (clinical significance) as much as statistical significance. The data
analysis occurred in two phases, firstly for baseline data collection utilised in Phase I investigation of
nutrition assessment tools and secondly, Phase II, the randomised-controlled trial of nutrition
intervention.

2.5.1 Phase 1: Nutrition assessment methods

Analysis for Phase I was based on data collected at baseline. Characteristics of patients based on
nutritional status by SGA (categorical) were compared with clinical variables including; age, weight,
body mass index (BMI), albumin, C-reactive protein, renal function (GFR), weight change over
previous 6 months by independent sample t-test (for continuous) or chi-square (for categorical). The
assessment for the accuracy of BCM-I as a diagnostic tool against SGA progressed as a ROC
analysis.

- 82 -

The modified SGA tools were evaluated for relationship to BCM-I and also for their diagnostic ability.
A comparison between the assessment measures at baseline between BCM-I (continuous) and the
alternate SGA-based assessment tools, 7-point SGA (categorical), PG-SGA and MIS (continuous) by
bivariate statistics (analysis of variance (ANOVA) and spearman correlation (non-parametric).

further assessment of BCM-I to the modified SGA was conducted in a ROC analysis. The tools which
had a linear relationship to BCM-I were investigated against the BCM-I cut-off for nutritional status
previously established.

Area under the curve was used to compare the diagnostic accuracy between measures included in the
ROC analysis. This gives a measure of the ability of BCM-I to discriminate between those that are
malnourished (SGA B/C) and those well-nourished (SGA A). The ROC analysis provides an area
from zero to one. The literature states, regarding nutritional assessment, an area equivalent to 0.7-0.8
250

provides acceptable discrimination, 0.8-0.9 excellent and >0.9 outstanding discrimination

Therefore, the guidelines of 0.5 - 0.7 = fair; 0.7 to 0.8 = good; 0.8 to 0.9 = excellent and >0.9
outstanding were used in this investigation.

2.5.2 Phase II: Randomised-control trial

2.5.2.1 Determining effective randomisation and group differences at baseline
Baseline characteristics were compared to check for systematic differences amongst groups for
participants in each treatment group, to check for correct randomisation. Differences in the baseline
variables were determined using either the t-test for continuous or the chi-square test for categorical
variables.

Associations were analysed using Pearsons correlation coefficient.

Descriptive

characteristics were also compared for participants completing the study with data on all outcome
measures compared to participants with some missing outcome data, and participants lost to followup.

2.5.2.2 Determining outcome variables

The analysis of the effect of the intervention proceeded on an intention-to-treat basis, irrespective of
how compliant participants were with the intervention protocol.

The independent variable being

treatment group was dichotomous (intervention or standard care).

- 83 -

Dependant variables were

measured at baseline (week 0) and follow-up (week 12) and were analysed as a change over time
score (week 12 minus week 0).

Change variables were created for the outcome measures (BCM, SGA, dietary intake (energy and
protein) and KDQOL components) by taking the difference of the week 12 measure from week 0.
Percent change in BCM was calculated using the equation: % BCM change = ((Week 12 Week 0) /
Week 0) x 100.

To consider the effect of intervention, histograms and scatter plots provided a visual representation of
the frequency distribution and the presence of linearity.

Primary outcome: Nutritional status

Analysis of covariance (ANCOVA) was used to determine whether mean change in BCM, both
absolute (kg) and relative (%) differed between the standard care and intervention groups, adjusting
for the baseline measurement.

Potential confounding variables for the relationship between treatment group and BCM, were identified
and collected at baseline including co-morbidity (as Charlson index), age, gender, GFR, BMI, living
situation and dialysis treatment plan. Bivariate analysis between the potential confounding variables
at baseline and nutrition-related outcome variables (change in BCM, dietary intake and PG-SGA) was
conducted. Baseline variables that had a significant relationship to the continuous outcome variables
were considered in a further ANCOVA analysis to provide adjusted estimates of intervention effects.

The measures of change in nutritional status, SGA and BCM, were categorised into improve, maintain
or decrease. For BCM, this equated to categories of participants change
1% loss, respectively.

1% gain; within 1% or

The categorical outcome variables (SGA change and categorised BCM

change) were compared to treatment group by chi-square analysis.

Categorical outcome variables (SGA change and categorised BCM change) were assessed for effects
of treatment allocation by multinomial logistic regression if relevant assumptions were met.

- 84 -

For each of these models, both the main effects and interactions of each of the factors and covariates
were assessed and withdrawn from each model if not significant (i.e. p>0.05). Both the unadjusted
and adjusted results are presented.

The agreement between weight change and BCM change was compared using a Bland-Altman
251

plot

251

. As described by Bland and Altman

the average change by weight and BCM (x-axis), was

plotted against the difference between percent weight and BCM change (y-axis).

The limits of

agreement were set at 2 standard deviations (SD) of the difference above and below the mean.

Intermediate outcome: Dietary intake

Dietary intake of protein and energy was translated into units per kg of ideal body weight (IBW) and
46

compared with evidence based guidelines . IBW for participants was calculated as current weight if
2

their BMI (kg/m ) was within the current healthy weight range as stated by recently published
2 46

guidelines in Stage IV CKD (BMI between 20 and 25kg/m ) . If a participant was under this BMI
2

range then the weight equivalent to a BMI of 20kg/m was calculated for IBW.
2

Similarly, if a

participants BMI was over 25kg/m then their weight equivalent to a BMI of 25kg/m was determined
as IBW.

In addition, change in dietary intake was assessed for confounding as per the methods provided
above in the analysis for BCM. In addition, a MANCOVA model was built to include both change in
energy and protein intake together with the established confounding variables, to test for the
independent effects of intervention change in energy and protein intake.

Both main effects and

interactions of each of the factors and covariates were assessed and withdrawn from each model if
not significant (i.e. p>0.05).

Both the unadjusted and adjusted mean differences of change in dietary intake are presented.

Secondary outcome: Quality of life

Analysis of covariance (ANCOVA) was used to determine whether mean change in quality of life
differed between the standard care and intervention groups, adjusting for the baseline QOL scores. In

- 85 -

addition, paired t-test was conducted for each treatment group separately to determine the
significance in change within each group over the treatment period.

2.5.3 Generalisability
The baseline characteristics of this study population were compared with the characteristics of a
reference population. Permission was granted from the Queensland Health Renal Collaborative to
access this data for this purpose (Hawley, Personal Communication, 03-01-2007). This reference
population consisted of incident pre-dialysis patients (GFR <30 mL/min) with data collected, initiating
nephrological care in Queensland Health facilities. The data presented here was collected between
April 2005 and December 2006. A number of limitations for use of this data have been identified.
Firstly, there is no way of verifying that every incident patient at every facility undergoes data collection
on their first visit. In addition, there may be significant areas of missing data. Therefore, for the
purposes of comparison, variables with >20% missing data will be highlighted for potential limitations
when assessed against this study population.

2.6

Ethical considerations

This study complied with the ethics guidelines recommended by the National Health and Medical
Research Council (NH&MRC). This study was granted ethics approval by Queensland University of
Technology Human Research Ethics Committee (Level 3) on 19 July 2004, and RBWH Human
Research Ethics Committee on 17 August 2004 (Appendix D). Subsequent minor amendments were
made to the information package and were approved by both committees.

Informed consent was obtained from each subject prior to entry into the study. Eligible patients were
provided with an information package (Appendix D) containing the details of the principal investigators
and the contact details of QUT and RBWH ethics committees in case there were any queries.
Potential participants were informed of: voluntary participation into the study; the purpose and any
possible disadvantage of taking part in the study; and participants were free to withdraw from the
study at any time without jeopardising their medical treatment. Patients willing to take part in the study
were asked to complete a consent form.

- 86 -

Each subject was coded with a study number to maintain anonymity. The study numbers were cross
referenced to the patient names, and stored in a locked cabinet in a separate location. The project
supervisor retained a list of participant ID numbers and treatment allocation in a separate locked
cabinet throughout the recruitment phase of the trial.

- 87 -

Chapter 3:

Comparison of nutrition assessment

methods in pre-dialysis CKD

3.0. Introduction
This chapter will address Phase I of this investigation and address the study aims relating to methods
of nutrition assessment. The first section will evaluate the relationship between subjective global
assessment (SGA) and body cell mass index (BCM-I) (derived from total body potassium counting
3.5

(TBK) and indexed for height (BCM-I kg/m )) for the assessment of malnutrition in a sample of predialysis CKD patients. Also in this section, a comparison of the characteristics of patients, according
to nutrition status rating by SGA is provided. The second section provides data to inform the
investigation of the modified versions of the SGA.

This aims to assess each tool (7-point SGA

categories, Patient-Generated SGA scores (PG-SGA), and Malnutrition-Inflammation Scores (MIS)) as

cross-sectional measures of nutritional status, against BCM-I, and to establish their capacity as
diagnostic tools of nutrition status. Finally, the relevance of these results to clinical practice and
Phase II of this investigation are discussed.

Phase I of the study investigating nutrition assessment methods was conducted from cross-sectional
data collection at week 0. Of the 60 eligible participants recruited and consented, 4 participants did
not receive baseline assessment following initial SGA assessment (reached end-point (dialysis or
transplantation n=2), or voluntary withdrawal (n=2) prior to further assessment).

Of the final 56

participants, 6 were withdrawn from BCM assessment, 2 due to voluntary withdrawal from TBK, and 4
were too large to comfortably undertake the measure due to size limitations of the TBK sensor.
Therefore, the number of participants for the SGA-BCM analysis was 50.

- 88 -

3.1

Results: Subjective Global Assessment and Body Cell Mass

3.1.1 Clinical parameters compared to subjective global assessment

Table 3-1 provides the characteristics of participants in this sample of pre-dialysis CKD patients based
on their nutritional status at week 0. There were no severely malnourished, or SGA C patients in this
sample. Moderately malnourished (SGA B) patients had significantly lower BCM, lower body weight,
BMI and greater weight loss over 6 months compared with well-nourished (SGA A, Table 3-2). In this
investigation, there was no relationship between SGA ratings of nutritional status, and the clinical
variables of albumin, CRP, renal function (GFR) or age (Table 3-2).

Table 3-1 Characteristics of 56 pre-dialysis chronic kidney disease patients based on

nutritional status by Subjective Global Assessment
Variable
Gender (M:F)
Age (y)
Weight (kg)
Weight change (6-mo %)
BCM (kg)***
3.5
BCM-I (kg/m )***
2
BMI (kg/m )
Albumin (g/L)

CRP (mg/L)
SD
GFR (ml/min) X

Well-nourished
n = 45
29:16
68.8 11.7**
76.1 14.7
-0.08 2.4
33.4 7.3
5.6 0.8
27.6 4.5
38.9 4.1

3.3 (1.4-16.0)
21.8 7.5

Malnourished
n = 11
4:7
74.0 11.1
64.8 20.9
-6.24 4.3
26.3 6.3
4.8 0.8
23.7 4.6
37.7 6.4
2.10 (1.0-4.8)
22.5 4.3

P-value*
0.090
0.192
0.042
0.004
0.007
0.013
0.015
0.431

0.331
0.755

M = Male; F = Female; BCM = Body Cell Mass; BCM-I = Body Cell Mass Index; BMI = Body Mass Index; CRP = C-reactive
protein; GFR = Glomerular Filtration Rate
* t-test; ** Mean standard deviation; ***n=50; n=33; median (inter-quartile range); Mann-Witney U; statistically significant
BOLD

Figure 3-1 provides the Receiver Operating Characteristic (ROC) curve plotting the sensitivity and 1specificity of BCM-I against identification of malnutrition by SGA. The area under the curve of 76.2%
(95% CI 60.0 to 93.0%) indicating BCM-I estimates nutritional status with good accuracy, based on the
scales previously determined (Section 2.5.1). The co-ordinate providing the highest sensitivity (73.2%)
and specificity (66.7%) for identifying a well-nourished state or SGA A, was a BCM Index of >5.25
kg/m3.5, selected from the coordinates provided in table 3-2.

- 89 -

1
0.9
0.8

Sensitivity

0.7
0.6
BCM-I

0.5

Reference*

0.4
0.3
0.2
0.1
0
0

0.2

0.4

0.6

0.8

1-Specificity

Figure 3-1

Receiver operating characteristic (ROC) curve showing the sensitivity of BCM-I

against 1-specificity for various cut-off values to determine nutritional status as identified by
SGA.
*The broken diagonal line represents a hypothetical ROC curve of a test that yields no diagnostic information.

Table 3-2 Co-ordinates of the ROC curve showing the sensitivity of BCM-I against 1-specificity
for various cut-off values to determine nutritional status as identified by SGA.

BCM-I (kg/m3.5)

Sensitivity

2.69
3.73
3.89
4.41
4.62
4.91
5.25
5.33
5.53
5.93
6.08
6.31
7.01
8.23

1-Specificity
1.000
.976
.951
.927
.902
.805
.732
.659
.561
.317
.195
.122
.024
.000

1.000
1.000
.889
.667
.556
.444
.333
.222
.111
.111
.111
.000
.000
.000

BCM-I = Body Cell Mass Index

3.1.2 Alternative subjective assessment tools as measures of nutritional

status
Modified versions of the SGA (MIS, PG-SGA score, 7-point SGA) were investigated for their
comparison to BCM-I and descriptive difference in ratings against the SGA.

- 90 -

3.1.2.1 Seven-point SGA

The distribution of the 7-point SGA ratings across the original SGA indicated that 5 represents the
overlap in the ratings between SGA A and B (Figure 3-2). It is also important to note there were no
ratings below 3 on the 7-point scale.

This is likely to reflect the fact there were no severely

malnourished (SGA C) patients in this sample, which are related to the lower ratings of 1 or 2 on the
252

7-point scale

25

Number of subjects

20

15
SGA A
SGA B
10

0
7

7-point SGA rating

Figure 3-2

Distribution of well-nourished (SGA A) and malnourished (SGA B), within the

ratings of the 7-point SGA in 56 pre-dialysis CKD patients

The relationship for BCM-I to 7-point SGA is shown in Figure 3-3. Statistical comparison of the mean
BCM-I between the 5 ratings of nutrition status on the 7-point scale (3 to 7) was not significant (F(4) =
2.24, p = 0.080). However, as evident in Figure 3-3, there was a steep increase in BCM-I between
rating 3 at 4, which is equivalent to BCM-I of 5.25kg/m3.5 with a steady incline at the higher ratings,
representing increasing BCM-I. Statistically, this reflects a significant linear trend, with an increasing
mean BCM-I for each category from rating of 3 to 7 on the 7-point SGA (F(1) = 6.71 p=0.013).

- 91 -

Body Cell Mass Index (kg/m3.5)

5.5

4.5

3.5

3
3

7-point SGA rating

Figure 3-3

Mean rating of 7-point SGA against BCM Index in CKD patients, n=50.

3.1.2.2 Scored nutrition assessment tools

Distribution of the overall scoring for the MIS (median (range) 4 (0-12)) and PG-SGA (3 (0-15)) were
very similar. When comparing the scores between the SGA categories, there was a similar average
and spread of scores for SGA A and B (Table 3-3). Scatter plots and corresponding correlations for
the relationship of BCM-I indicate a moderate negative correlation, which was comparable for both
tools. (Table 3-3 and Figure 3-4). Therefore, the scored assessment tools appeared to be assessing
nutritional status to a similar degree when compared with SGA and BCM.

Table 3-3 Nutrition assessment tools in pre-dialysis CKD patients against SGA (n=56) and
BCM Index (n=50)
Assessment tool

PG-SGA Median (IQR)

MIS
Median (IQR)

WellMalnourished Relationship to BCM-I (kg/m3.5) *

nourished
n = 11
n = 45
3.0 (1-4)
9.0 (8-11)
-0.347 p=0.014**
4.0 (1-4)
10.0 (6-12)
-0.361 p=0.011**

CKD = Chronic Kidney Disease; SGA = Subjective Global Assessment; BCM = Body Cell Mass; PG-SGA = Patient-Generated
SGA; MIS = Malnutrition Inflammation Score
*n=50 **spearman correlation; median (inter-quartile range)

- 92 -

3.5

Figure 3-4

Scatter plot of the BCM index (kg/m ) between the Malnutrition Inflammation

Score (left) and the Patient Generated SGA (right), regression line representing slope of the
correlation co-efficient

Using the Bland-Altman plot (Figure 3-5) there was good agreement between the average of PG-SGA
and MIS and mean difference with a minimal mean bias (meanSD -0.091.89). All but 2 of the
251

extreme values were within the 2 SD confidence limits as recommended by Bland-Alman

. The

correlation between the difference and mean of assessment scores was not significant, indicating
there was no systematic bias between the scoring methods (r=-0.181, p=0.183).

6.00

Difference (PG-SGA minus MIS)

4.00

2.00

0.00

-2.00

-4.00

-6.00

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

Average scored estimates from PG-SGA and MIS

Figure 3-5

Scatter plot of the mean difference between the PG-SGA and the MIS scores,

against the average between the scored assessments with limits of agreement of 2 SD from the
mean difference

- 93 -

3.2

Diagnostic precision of the modified SGA tools

Each of the modified tools had a statistically linear relationship with BCM-I. Therefore, they were
compared against the diagnostic cut-off if BCMI of 5.25 kg/m

3.5

for nutritional status, as determined in

Section 3.1.1.

Figure 3-6 provides the ROC curve plotting the sensitivity and 1-specificity of 7-point SGA (inverted),
3.5

MIS and PG-SGA to identify malnutrition with BCMI <5.25 kg/m . The area under the curve for each
tool was between 64 and 67%, indicating fair accuracy. The PG-SGA score, had area under the curve
( SE) of 67.39.0% (95% CI 50.0 to 85.0) represented a diagnostic capacity marginally better than
the other tools. The MIS was similar at 65.19.5% (95% CI 46.6 to 83.7). Finally, the 7-point SGA,
was inverted to present increasing rating with decreasing nutrition status (for ease of comparison to
the scored assessment tools), had a similar average diagnostic capacity at 64.1 9.0% (47.3 to
80.8%).
1
0.9
0.8

Sensitivity

0.7
0.6

7-point SGA

0.5

PG-SGA score
MIS

0.4

Reference*
0.3
0.2
0.1
0
0

0.2

0.4

0.6

0.8

1-Specificity

Figure 3-6

ROC curve showing the sensitivity against 1-specificity for ratings of 7-point
3.5

SGA (inverted), scored PG-SGA and MIS to determine BCM-I <5.25 kg/m .
*The broken diagonal line represents a hypothetical ROC curve of a test that yields no diagnostic information.

- 94 -

3.3

Discussion

A diagnostic gold standard pertains to the best available method for establishing the presence or
253

absence of a condition of interest

. With regard to nutritional assessment in CKD, the primary

purpose is to identify the need for nutritional intervention, establish response to treatment and to
adequately predict outcome. SGA has previously been identified as a diagnostic reference standard
in CKD, capable of meeting these requirements with good inter-rater reliability (to establish well
nourished (SGA A) and malnourished (SGA B and C) (Table 1-3 and 1-10). When establishing
nutritional status by use of SGA, clinicians rely on a set of characteristic signs and symptoms
(decreasing and inadequate dietary intake, etc) and also biological changes, including % recent weight
change.

3.3.1

SGA and BCM

Body cell mass represents a continuous, objective, gold-standard measure of metabolically active
tissue. When indexed for height (metres to power of 3.5) it represents a cross-sectional measure of
the cellular component of body mass independent of height, therefore, a useful biological marker for
nutritional status. To the authors knowledge, BCM, as measured by TBK has not been compared for
diagnostic accuracy to SGA prior to this investigation. In addition, this investigation is unique in that it
investigates the association between SGA and body composition in CKD as it features SGA as the
diagnostic reference standard.

When these two variables were compared there was a clear clinical and statistical association with
greater mean BCM and BCM-I for well nourished, compared with malnourished patients (33.4 7.3 vs.
26.3 6.3 kg p=0.007 and 5.6 0.8 vs. 4.8 0.8 kg/m

3.5

p=0.018, respectively) (Table 3-2). This is in

agreement with an investigation by Cooper et al (2000) in haemodialysis (n=76) patients, where total
102

body nitrogen (TBN) was also significantly lower in patients rated SGA B and C compared to A

Thus SGA, relating to gold-standard biological measures of body composition, is a robust tool for use
in CKD.

The diagnostic accuracy was affirmed by demonstrating an area under the ROC curve equivalent to
250

over 75% against SGA. This demonstrates acceptable discrimination of malnutrition by the BCM-I

- 95 -

From the ROC plot, the cut-off point for BCM-I of <5.25kg/m

3.5

to identify malnutrition was selected.

This was determined by selecting the cut-off value that produces acceptable sensitivity and specificity.
Based on these figures it is acknowledged that use of this cut-off will result in some level of
misclassification, however, the misclassification is likely to produce more false positives, rather than
misclassify malnourished patients.

In the literature, the validation of SGA in CKD has not been without its controversy. One study
criticising the validity of the SGA to assess nutritional status was conducted by Cooper et al (2002).
This investigation used Nitrogen Index (measured by total body nitrogen, TBN) of <80% as the
reference standard for nutritional status. This means if the ratio between measured TBN of a HD
patient was less than 80% of that predicted (by sex- age- height-matched general population), then
102

they were considered to be clinically malnourished

There is an issue with this interpretation.

Malnutrition is fundamentally a result of progressive decrease in body weight/protein stores, indicated

by insufficient intake for requirements

254

, which is difficult to determine by a cross-sectional cut-off,

based on healthy population data. A patient may not register as having body protein stores <80% of
predicted, due to their sheer body size, whereas, significant wasting by an overweight patient (BMI >
2

25kg/m ), and/or moderate weight loss with significant uraemic symptoms can contribute to a
legitimate SGA B rating. This effect is lost when using a dichotomised outcome based on predicted
body composition.

Prior to the investigation featured in this thesis, the identification of malnutrition in CKD utilising TBK
had been performed by comparing measured TBK to TBK from prediction equations

72, 87, 255, 256

or

72

healthy controls . As mentioned above, there is a risk with diagnosing poor nutritional status solely
based on a cross-sectional assessment lower than predicted body mass (protein or cellular) status.
This is particularly relevant considering the limitations with prediction equations themselves, as
reported in Appendix A. To date, validity testing of alternate methods such as bio-impedance analysis
(BIA) and dual-energy X-ray absorptometry (DXA) to TBK to estimate BCM has been restricted to
healthy populations

257, 258

and other clinical conditions

259

, however, these methods are contraindicated

67, 72

in CKD, particularly where fluid abnormalities exist

- 96 -

3.3.2

Relationship between SGA and clinical variables

In this study, there appeared to be no relationship between nutrition status and albumin or CRP (Table
3-2). The prognostic significance of serum albumin and C - reactive protein (CRP) is well-recognised
in Stage V CKD

29, 260

. As mentioned previously, acute phase protein reaction, which is commonly

associated with elevated CRP results in a reduction of albumin synthesis and is thought to explain
much of the high prevalence of hypoalbuminemia in chronic kidney disease

57, 58

. Although the

presence of low albumin and/or a raised CRP strongly influences morbidity and mortality of CKD
patients, it is likely this is independent of nutrition status. In this case, these biochemical parameters
2

were not reliable for the direct assessment of nutrition status in a cross-sectional assessment.

Patients who were malnourished (SGA B) had experienced significantly greater weight loss in the
previous 6 months (results -6.2 4.3 kg, compared with -0.1 2.4 kg p=0.04) and also had
significantly lower BMI (23.7 4.5 kg/m compared with 27.6 4.5 kg/m p=0.02) (Table 3-2). It is
2

also important to note from this data that malnourished patients (SGA B) had an average BMI of 23.7
2

kg/m which is toward the higher end of the healthy weight range for pre-dialysis (20-25 kg/m ). This
supports other studies which indicate patients may be malnourished, yet still be classified as being
261, 262

healthy, if not overweight according to BMI

These results highlight the limitations with

clinically-available objective markers of nutritional status in CKD, and further support the need for a
panel of assessment measures including SGA.

3.3.3 Performance of modified SGA tools

The next aim of this study was to investigate the performance of alternate SGA-based tools with SGA
and BCM-I. For the 7-point SGA, There was a progressive increase in mean BCM-I between the
lowest rating (3) and the highest rating (7), with a steep jump from mean at 3 (3.88 0.16 kg/m ) to 4
3.5

(5.25 0.20 kg/m ), and only a small increase from categories 4 to 7 (5.62 0.83 kg/m ). There
3.5

3.5

was a significant linear trend (F(1) = 6.71 p=0.013), however, this increase was a 2-stage pattern,
where only the rating of 3 was significantly lower from a clinical perspective to the mean BCM-I from
the higher other ratings (Figure 3-3). This two-stage effect agrees with a recent SGA validation study
in HD patients, where the mean of the objective parameters, BMI and albumin were stable from SGA 5
263

to 7, and only significantly lower values were seen for objective markers in categories 3 and 4

- 97 -

Commonly, the ratings of 7-point SGA is dichotomised as well-nourished (rating 6 and 7), and
225

malnourished (<5), as 5 appeared to be the cross-over between A and B ratings in this study

While the published literature applies to the HD population, the findings from this study would show an
overestimation of malnutrition if a score of 6 or 5 and below was used and the 7-point SGA does not
distinguish nutrition status better than the original SGA rating.

Both the MIS and the PG-SGA score increased to the same extent with decreasing nutrition status
showing an identical, yet weak relationship with BCM and a similar distribution of scores with the SGA
categories. They therefore appear to be measuring nutritional status to a similar degree. The BlandAltman plot (Figure 3-1) confirmed this to be the case as there was no systematic difference between
the scoring of the tools. The mean bias was negligible, and limits of agreement (at 2SD) were less
than 4 which could be considered to be a clinically acceptable range.

The additional variables of the MIS aim to improve the assessment of malnutrition-inflammation
114

complex syndrome

. In this investigation, the MIS rated identically with PG-SGA, indicating scoring

of albumin, BMI and total-iron binding capacity did not add any further value in determining nutritional
status. Considering dialysis vintage was not scored as part of the MIS assessment, it is possible to
see a higher scoring MIS in dialysis populations.

However, collection and rating of objective

parameters in this pre-dialysis sample in the MIS provided no further benefit than PG-SGA
assessment score.

In HD, the MIS tool predicts poor clinical outcome (mortality and hospitalisation)

35, 114

Previous

investigations indicate the PG-SGA score has a high inter-rater reliability (oncology), sensitivity and
specificity when compared with SGA classification for both oncology and HD populations

3.3.4

109, 261, 264

Diagnostic ability of modified SGA tools

Each of the tools displayed moderate diagnostic capacity. The 95% confidence intervals for each tool
crossed or went close to crossing 50%. This indicates a risk in their use as diagnostic tools, as for any
place of the rating spectrum, there is an equal chance of being above or below the nutrition status cut3.5

off of BCM-I = 5.25 kg/m .

- 98 -

From a statistical perspective, there are some limitations in establishing a cut-off for BCM-I from SGA
and then using this for rating the modified SGA tools, as they are not entirely independent of this SGA.
Given the clinical limitations of conducting a more rigorous validation study, this assessment is based
on the best available variables, and provides a base for future investigations for assessment methods
in this area. Further limitations included the small sample size, and a low prevalence of malnutrition
(only 20%). Despite these limitations this study provides new information on the cross-sectional rating
ability of these new assessment tools in pre-dialysis CKD.

The modified SGA tools may have greater merits for the prospective assessment of nutritional status
to track smaller changes over time for each has limitations as diagnostic tools. The scored PG-SGA is
considered superior to the other tools as it provides the original SGA rating, as well as a score, which
relates to BCM-I, yet does not require collection of biochemistry or anthropometric measures. The
limitation of 7-point SGA is that it requires clinicians to provide a subjective classification to one of 7
categories, which is difficult for both an inter-rater reliability as evidenced by Steiber et al (2006), and
263

diagnostic ability, as evidenced in this investigation

. Overall, the original SGA is best for the cross-

sectional assessment of nutritional status and PG-SGA needs to be investigated further for this ability
to track small, significant changes over time.

3.4

Summary of nutrition assessment measures

Since initial validation of SGA in haemodialysis (HD) patients

standard of nutritional status in pre-dialysis
1, 64, 105, 109, 112

31, 107

100

, SGA has been a classification

, commencing dialysis

1, 2, 38

and maintenance dialysis

populations. This investigation supports the diagnostic ability of the SGA rating scale to

determine nutrition status and verifies agreement between BCM-I and SGA.

This investigation provided criterion validation based on gold-standard body composition measure to
assess against other clinical tools, scored PG-SGA, MIS and 7-point SGA. From this, it was clear that
each of the modified SGA tools related to BCM-I as a continuous variable and all had similar
3.5

limitations for the diagnostic assessment of nutrition status against cut-off of BCM-I 5.25 kg/m . The
PG-SGA warrants further investigation to assess prospective change in nutritional status.

- 99 -

Chapter 4:

Results of randomised-controlled trial of

nutrition intervention in pre-dialysis CKD on nutritional

status

4.0 Introduction
The results presented in this section represent Phase II of this investigation, a randomised-controlled
trial of nutrition intervention based on Stage IV and V CKD patients selected from a specialist predialysis clinic.

This chapter will address the hypotheses to determine the impact of providing

structured nutrition intervention on the primary outcome of nutritional status and intermediate outcome,
dietary intake compared to standard nutrition care, in pre-dialysis CKD patients. The first section of
this chapter will give an overview of the characteristics of the participants, compare participants
allocated to either treatment group and also participants with complete versus missing outcome data,
to evaluate for any systematic differences. The second section will directly address each of the
hypotheses relating to change in nutritional status (body cell mass (BCM), subjective global
assessment (SGA) and dietary intake, particular energy and protein) to determine the effectiveness of
the intervention at the bivariate level. Finally, the nutrition status and dietary intake outcomes will be
further assessed by multivariable models to evaluate the adjusted effect of the intervention.

4.1

Characteristics of the study participants

Sixty-six patients were eligible for the study during the recruitment period and 62 agreed to participate
(94% response rate). The CONSORT flowchart of the progression of participants was provided in
Figure 2-2. Following the consent of 62 participants, six were excluded from baseline assessment
(two were found to be ineligible, two participants voluntarily withdrew and two transferred care to
dialysis (1) and transplant (1)). A comparison of the characteristics of the participants (n=56) who
underwent baseline assessment, non-participants (n=8) and a reference population in Table 4-1. The
reference population data is from the Queensland Health Renal Collaborative database and consists
of incident CKD patients (GFR <30 ml/min) referred to any Queensland Health facility for nephrological
care between 2005 and 2006.

- 100 -

Table 4-1 Comparison of characteristics of the study population (n=56), non-participants

(n=8), and reference population from Queensland Health Collaborative (n=605) of
pre-dialysis CKD patients
Study participants

Non-participants

Reference
Population
(CKD
Collaborative)

56

Refused or withdrew
before assessment
8

Queensland
Health
Renal Collaborative
605

Gender (% Male)

60.7

25.0

50.0

SD
Age (years) X
GFR (ml/min)

70.2 11.6

66.1 13.6

67.0 13.6

22.0 6.8

16.2 5.9

21.4 8.1

BMI (kg/m )

27.2 4.9

23.1 3.2

29.8 7.2#

SGA (% B or C)

17.9

25.0*

N/A

GFR = Glomerular Filtration Rate; BMI = Body Mass Index

*SGA conducted on non-participants who originally consented (n=4); # 35% missing data

Age, GFR, albumin in the study population was within 5% of the reference population, and a similar
proportion with diabetes mellitus (Table 4-1). There was an equal proportion of males and females in
the reference population, however, slightly greater representation of males in the study population.
The reference population has slightly larger mean BMI; however, the reference data for BMI has
limitations considering the significant amount missing (35%). Overall, the study group appears to be
similar to the reference population. In the eligible patients that did not participate in the study (n=8,
either because they refused the study (n=4), or withdrew from participating (n=4)), there was a greater
proportion of females, with lower renal function than the final study participants and the reference
population. These differences are considered minor, particularly due to the low numbers of nonparticipants.

The study sample had significant co-morbidities, 81% (47/56) with a history of hypertension, 73%
(43/56) with cardiovascular disease, and 42% (25/56) with diabetes mellitus, resulting in a mean
Charlson co-morbidity index

243

of 6.5 1.9. The majority of the participants reported stable weight

(28/56) or weight gain (11/56) over the past six months. Twelve participants reported some level of
unintentional weight loss, and five had intentionally lost weight, collectively resulting in a negligible
mean weight loss of 1.08 3.30% for the sample over the previous six months.

- 101 -

The majority of participants lived with their partner (62%; 35/56) or at home alone (23%; 13/56). Over
sixty percent of the participants were dependent on others (family member, spouse or carer) for food
purchasing and preparation (65 and 68% respectively). Fewer than half of the participants (47%) had
previous dietary intervention, 35% for co-morbidity management (e.g. diabetes, cardiovascular
disease) and 12% specific for electrolyte and fluid management (to control the intake of fluid an/or
electrolytes, potassium and sodium). No participants had previously received nutrition support for predialysis CKD specifically.

4.1.1 Characteristics of participants allocated to each treatment group

There were no statistically significant differences between the characteristics of subjects receiving
intervention or standard care treatments at baseline according to Table 4-2. In addition, there was no
difference between groups regarding co morbidities, living situation, level of care/independence, or
previous attendance to pre-dialysis education.

Table 4-2 Baseline characteristics (mean

SD) of patients with pre-dialysis CKD randomised

to intervention (n=29) or standard care (n=27) treatment

Baseline Characteristics
Age
Gender, Male n (%)
GFR (ml/min)
BMI
% weight change in 6 months
Charlson co-morbidity index
BCM (kg)
Nutritional status n (%)
SGA A Well nourished
SGA B Moderately malnourished
Physical Health Component QOL
Mental Health Component QOL
Mean daily protein intake (g/kg)
Mean daily energy intake (kJ/kg)

Total (n=56)

Intervention Standard Care p-value*

(n=29)
(n=27)
69.5 11.7
17 (58.6)
23.1 7.2
26.8 4.7
-1.8 3.6
6.5 1.9
31.6 7.3

70.9 11.6
17 (63.0)
21.6 6.1
27.6 5.2
-0.3 2.8
6.7 1.9
32.3 8.1

46 (82.1)
22 (75.9)
10 (17.9)
7 (24.1)
34.4 9.8
35.0 9.8
46.4 12.2 44.6 12.7
1.12 0.34 1.07 0.32
105.0 24.0 101.8 23.0

24 (88.9)
3 (11.1)
33.7 10.0
48.5 11.6
1.17 0.35
108.5 25.2

70.2 11.6
34 (60.7)
22.0 6.8
27.2 4.9
-1.1 3.3
6.6 1.9
31.9 7.6

0.645
0.740
0.414
0.525
0.080
0.669
0.768
0.166

0.619
0.254
0.257
0.305

*Independent samples t-test, chi-square test for independence.

GFR = Glomerular Filtration Rate, BMI = Body Mass Index, BCM = Body Cell Mass, SGA = Subjective Global Assessment, QOL
= Quality of Life

Participants in each of the two treatment groups had sightly different weight change over previous 6
months (intervention -1.83.6 % vs standard care -0.32.8 % p=0.080), which approached statistical
significance, although this difference is not considered clinically significant. The distribution of primary

- 102 -

cause of CKD is also similar across the treatment groups (Table 4-3). Therefore, the randomisation
appears to have been successful.

Table 4-3 Primary identified cause of CKD for 56 subjects randomised to receive intervention
(n=29) or standard care (n=27) in a sample of pre-dialysis CKD patients
Analgesic Nephropathy
Diabetes Mellitus
Polycystic Kidney Disease
Hypertension
Glomerulonephritis
Miscellaneous / Unknown causes

Intervention (n=29)
1
9
2
5
6
9

Standard care (n=27)

2
6
4
3
5
10

4.1.2 Characteristics of participants with missing outcome data

Table 4-4 outlines the baseline characteristics of participants grouped for completion of outcome
measures at both week 0 and week 12 of the study. Complete outcome data refers to participants
with complete records for all outcome measures at both week 0 and 12. Of the 50 participants who
completed week 12 assessments, participants with missing data at either week 0 or 12, were grouped
in Table 4-4 as having partial outcome data, and accounted for in Figure 2-2. Incomplete data
occurred in 6 participants from the intervention group (4 with incomplete TBK; 2 with incomplete
KDQOL survey), and 8 from the standard care group (5 with incomplete TBK, 3 with incomplete
KDQOL and 2 with incomplete food records). Also, the characteristics of participants who were lost to
follow up, who had either died (n = 4) or commenced chronic haemodialysis (n = 2) between week 0
and 12 are represented in Table 4-4. There were a greater number of participants allocated to the
intervention group who were lost to follow-up (n=5, 83%), 4 of which were due to death unrelated to
the study intervention.

- 103 -

Table 4-4

Baseline characteristics (mean

SD) of study participants who completed 12-week

study period in pre-dialysis CKD (n=56)

Treatment group (number in Intervention)

Age (mean SD)
Gender % M
GFR (ml/min)
BMI kg/m (mean SD)
% weight change in previous 6 months
2

Charlson co-morbidity index (mean SD)

BCM (kg) (mean SD)
Nutritional status, %
SGA A Well nourished
SGA B Moderately malnourished
Home setting, %
Lives with partner
Lives alone
Lives with family or carers

Complete
outcome data
(n=36)

Partial
outcome data
(n=14)

Lost to followup (n=6)

18
70.8 12.0
64.9
22.6 6.1
26.6 4.2
-0.92 3.2
6.2 1.8
32.4 8.1

6
66.7 11.3
53.8
22.5 9.5
29.4 6.7
-0.60 3.6
7.0 1.9
29.9 5.3

5
71.6 10.2
50.0
20.6 2.5
25.6 3.5
-3.60 3.1
8.5 1.5
30.9 7.1

31
5

11
3

4
2

21
11
4

12
1
1

4
1
1

GFR = Glomerular Filtration Rate, BMI = Body Mass Index, BCM = Body Cell Mass, SGA = Subjective Global Assessment, QOL
= Quality of Life

There was no systematic bias between patients based on the full or partial completion outcome
measures. The most notable discrepancy between the groups exists for BMI, as it is higher in the
participants with partial outcome data (29.4 6.7 kg/m ) compared with complete data (26.6 4.2
2

kg/m ). This can be explained by the participants with a large waist circumference who were excluded
from the TBK analysis due to size restrictions (as per Methods 2.3.1).

There was an acceptable

distribution of participants from the intervention and standard care group with missing TBK data
(Figure 2-2). Considering the observed group difference in characteristics between the complete and
partial outcome date are only minor, it appears that participants with incomplete data are at random,
and there is no systematic difference between the groups in Table 4-4.

4.2. Primary Outcome: Nutrition status results

The following results reflect the final 50 participants who completed the study (intervention n=24, and
control n=26). Each of the outcome measures were assessed as continuous change scores and were
adjusted for the baseline measurement using analysis of covariance (ANCOVA) models as per 2.5
Statistical Analysis.

- 104 -

4.2.1 Body Cell Mass

Body Cell Mass (BCM) change was assessed in 82% (n=41) of the final 50 participants. Table 4-5
shows the mean difference for absolute (kg) and relative (%) change in BCM between the 2 treatment
groups. The effect of intervention on change in BCM showed a trend for an increased BCM for the
intervention group and a reduction in the standard care group, resulting in an overall difference in
change between the groups which was not statistically significant (Table 4-5).

Table 4-5

Mean difference (95% CI) of change in body cell mass (BCM) over the 12-week
treatment period in pre-dialysis CKD patients for intervention (n=21) and standard
care (n=20) groups

Variable

Intervention (n=21)

Standard care (n=20)

Mean difference
between groups

p-val

BCM (%)
BCM (kg)

2.0 (-1.9 to 5.9)

0.5 (-0.7 to 1.8)

-1.5 (-5.5 to 2.5)

-0.5 (-1.8 to 0.7)

3.5% (-2.1 to 9.1)

1.1kg (-0.7 to 2.8)

0.209
b
0.238

F (1,38) 1.63, adjusted r2 = 0.005; F (1,38) 1.44, adjusted r2 = 0.10; BCM = Body Cell Mass

4.2.1.1 Categorised change in BCM

Figure 4-1 shows the BCM data divided into participants who maintained ( 1%), improved or
decreased (>1%) in BCM by treatment group over the 12-week period. More than twice as many
participants in the intervention group showed an increase (61.9% vs 30.0%), and less than half the
proportion decreased in BCM (23.8% vs 50.0%), compared to the subjects receiving standard care (x
(2) = 4.37 p=0.118).

- 105 -

14

Number of Participants

12
10
8

Intervention
Standard Care

6
4
2
0
Improve >1%

Maintain

Decrease >1%

Body cell mass change

Figure 4-1

Proportion of change in BCM over the 12-week treatment period for pre-dialysis

chronic kidney disease patients receiving either nutrition intervention (n=20) or standard care
(n=21)

4.2.2 Subjective Global Assessment

Most of the participants (n=40, 80%) maintained their baseline nutrition status according to SGA over
the 12-week treatment period (Table 4-6). Of those who had a change in SGA status, all of the
malnourished subjects in the intervention group improved their nutritional status (from SGA B to A),
resulting in no malnutrition at week 12. This is in contrast to the participants receiving standard care
treatment, which resulted in a rise in the proportion malnourished from 11% at week 0, to 22%
(including 1 severely malnourished, SGA C) for the standard care group at week 12. This difference in
2

change in SGA between the 2 groups was statistically significant X (2) = 12.76 (P=0.002).

Table 4-6 Change in Subjective Global Assessment ratings for intervention (n=24) and
standard care (n=26) groups n(%) over the 12-week treatment period, in a sample of
pre-dialysis CKD patients
Change in SGA

Intervention (n=24)

Standard care (n=26)

0 (0)
19 (79.2)
5 (20.8)

4 (14.8)
22 (84.5)
0 (0)

Deteriorated
No change
Improved
SGA = Subjective Global Assessment

The categories of change in SGA, independent of treatment group were assessed against change in
body cell mass, weight and PG-SGA score. A clear trend for decreasing body cell mass was noted for
participants who deteriorated in SGA status.

This was further verified by BCM maintenance in

- 106 -

participants maintaining SGA status and BCM increase in participants who improved in SGA rating.
This was further paralleled by weight change (Table 5-7).

Table 4-7 Change in Subjective Global Assessment ratings and relative (%) change in body
cell mass, weight, and change in PG-SGA score, mean (95%CI), over the 12 week
treatment period, independent of treatment group, in a sample of pre-dialysis CKD
patients

Change in SGA

Deteriorated n=4
No change n=41
Improved n=5

Change Body Cell

Mass (%)*

Change Weight (%)*

Change in PG-SGA*

-7.1 (-13.5 to -0.6)

0.2 (-2.8 to 3.2)
4.8 (-11.4 to 21.0)

-5.0 (-16.7 to 6.7)

0.6 (-0.3 to 1.5)
2.7 (0.1 to 3.7)

6.5 (2.7 to 10.3)

-0.4 (-1.0 to 0.2)
-5.2 (-6.8 to -3.6)

*mean (95% CI); SGA = Subjective Global Assessment, PG-SGA=Patient Generated Subjective Global Assessment

The Patient-Generated SGA (PG-SGA), a tool scoring the assessment parameters of the SGA, and
validated for use in this sample in Chapter 3 has merit for measuring nutrition status prospectively.
The change in PG-SGA score between SGA category changes was clinically significant evidenced in
table 4-7, with over a 5 point difference between those who improved, maintained or decreased.

Table 4-8 Values for PG-SGA score at week 0 and week 12, between the intervention (n=24)
and standard care (n=26) groups, in a sample of pre-dialysis CKD patients

Week 0
Week 12

Intervention
median (IQ-range)

Standard care
median (IQ-range)

3.0 (1.3-5.0)
2.0 (2.0-3.0)*

2.5 (1.0-6.0)
2.0 (1.8-6.3)

IQR = Inter-quartile range * p<0.05 paired t-test, baseline and week 12 values.

Table 4-8 provides the median (inter-quartile range) of PG-SGA score between the treatment groups.
The mean difference in change of PG-SGA score between the groups was statistically significant (F
(1,48) = 4.74 p=0.034; adjusted r2 = 0.071), representing a magnitude of -1.82 (-3.58 to 0.07), for
change in intervention minus standard care, therefore representing a trend for increase in score for
control and decrease in intervention groups (see Table 4-8). Although statistically significant, given
the small magnitude of change in PG-SGA score between groups, this is not considered to be
clinically significant.

- 107 -

4.3

Dietary intake results

In this section the results of change in the intermediate outcome measure, dietary intake is explored.
Firstly, , the difference in change in energy and protein intake between each treatment group, as
pertaining to the study hypotheses is shown. This is followed by a brief assessment to consider the
validity of reported energy intakes by use of the Goldberg method described in Section 2.3.2. An
evaluation of the proportion of participants meeting energy and protein intake guidelines by treatment
group follows. Finally, the change in macronutrient distribution contributing to total energy intake
between the treatment groups is considered as a qualitative assessment in dietary intake change.

4.3.1 Total energy and protein intake

Energy intake, as detailed in Table 4-9, significantly increased over the treatment period for the
intervention group and decreased in the standard care group. Conversely, mean protein intake did not
change significantly in either group, except for a decrease in protein intake relative to body weight in
the standard care group. Overall, there was little difference between the mean change in protein
intake between the treatment groups for absolute intake or relative to body weight.

Table 4-9 Energy and protein intake as total (kJ or g) and relative to ideal body weight (kJ/kg
or g/kg) for each treatment group at week 0 and week 12 and the mean difference in
change between the intervention (n=24) and standard care (n=24) groups
Variable

Time

Intervention (n=24)

Standard care (n= 24)

Energy Intake (total kJ)

Baseline
Week 12
Mean change

6707 1567
7463 1393
756 (302 to 1211)**

7247 1614
6785 1558
-485 (-966 to -3)*

Energy Intake (kJ/kg IBW)

Baseline
Week 12
Mean change

102.6 24.8
109.3 25.2
114.5 25.6
102.7 22.2
11.36 (4.68 to 18.04)* -6.33 (-13.00 to 0.35)

Protein intake (total g)

Baseline
Week 12
Mean change

71.4 22.1
69.2 12.4
-2.2 (-10.4 to 6.0)

78.6 22.4
70.7 20.3
-8.0 (-15.3 to -0.7)*

Protein Intake (kJ/kg IBW)

Baseline
Week 12
Mean change

1.09 0.34
1.06 0.18
-0.07 (-0.15 to 0.02)

1.19 0.35
1.06 0.27
-0.11 (-0.19 to -0.03)*

*p<0.05 **p<0.01

- 108 -

4.3.1.1 Under-reporting of energy intake

Participants were identified as being low energy reporters if their energy intake (EI) ratio to basal
energy expenditure (BEE, calculated from the Harris Benedict equation) equated to less than 1.3.
Only patients considered weight stable, defined by an arbitrary cut-off of less than 2.5% change in
past 6 months underwent assessment for under-reporting.

Table 4-10 Proportion of participants considered to be under-reporting energy intake (EI/BEE

<1.3) at baseline and week 12, between intervention and standard care groups .
Overall

Intervention

Control

Baseline n=

38

19

19

n (%) under-report
Follow-up n=
n (%) under-report

25 (65.8)
31

16 (84.2)
17

9 (47.4)
14

18 (58.1)

10 (58.8)

8 (57.1)

EI:BEE = Ratio of energy intake to basal energy expenditure

When assessing the proportion of weight stable patients with EI/BEE <1.3 at either baseline or week
12, it indicates that overall there is around 60% under-reporting of energy intake at both baseline and
week 12. The presence of under-reporting has implications for the interpretation of the dietary intake
results shown in Table 4-9. In particular, as shown in Table 4-10, it appears those participants in the
intervention group who were weight stable at baseline were more likely to under-report (16 of 19) than
those weight-stable at week 12 (10 of 17). Therefore, the magnitude of increase in energy intake
shown in Table 4-9, may be slightly enhanced by the presence of more accurate reporting by the
intervention group. Similarly, declines in standard care intakes may be affected by the presence of
greater proportion of participants considered to be under-reporting at week 12 (8 of 14) compared with
baseline 9 of 19).

Table 4-11 Ratio between energy intake (kJ) and estimated basal energy expenditure (mean
SD), participants who were weight stable at both baseline and week 12 from
intervention (n=15) and standard care (n=10)
n
EI:BEE baseline
EI:BEE follow-up

Overall

Intervention

Control

25
1.24 0.30
1.18 0.25

15
1.11 0.27
1.16 0.27

10
1.42 0.23
1.21 0.25**

**p<0.05 paired t-test; EI:BEE = Ratio of energy intake to basal energy expenditure

- 109 -

Table 4-11 indicates the mean ratio of EI/BEE for participants identified as weight stable at baseline
and week 12 separated by treatment group. There was no statistically significant difference between
the EI/BEE between groups at baseline. Over the 12 week treatment period, mean ratio of EI to BEE
decreased for weight stable participants in the standard care group (t (9) = 3.84 p=0.004), but did not
change in the intervention group.

This indicates a mean increase in under-reporting in this cohort of

standard care participants between the two data collection times. This is therefore consistent with
Table 4-10, and relates to declines in intakes recorded for the standard care group in Table 4-9. The
decrements in energy and protein intake demonstrated by the standard care group between baseline
and week 12 may have been impacted by increased rates of under-reporting. This will be discussed
further in section 5.1.4

4.3.2 Proportion of participants meeting energy and protein intake guidelines

Dietary intake was also categorised according to guideline recommendations from absolute values at
week 0 and week 12.

Evidence-based guideline recommendations were utilised to guide the

categories in Table 4-12. Protein intake should not be less than 0.75g/kg, with an optimum level 0.7546

46

1.0g/kg . Energy intake for patients >60years an intake of 125kJ/kg is recommended . Therefore,
considering the mean age of the study population 70.2 11.6 years, >125 kJ/kg was set as the goal
intake, and <100kJ/kg was considered inadequate.

Table 4-12 Proportion of subjects in the intervention (n=24) and standard care (n=24) group
46

meeting the recommended level of protein and energy intake at week 0 and week
12
Intervention
%(n)
Protein intake
Meeting guideline 0.75-1.0 g/kg
Above guideline >1.0g/kg
Under guideline <0.75g/kg
Energy intake
Adequate >125kJ/kg
Moderate 100-125kJ/kg
Inadequate <100kJ/kg

Standard care
%(n)

Week 0

Week 12

Week 0

Week 12

41.7 (10)
54.2 (13)
4.2 (1)

45.8 (11)
54.2 (13)
0.0 (0)

32.0 (8)
64.0 (16)
3.7 (1)

25.0 (6)
58.3 (14)
16.7 (4)

16.7 (4)
25.0 (6)
58.3 (14)

29.2 (7)
37.5 (9)
33.3 (8)

40.0 (10)
24.0 (6)
36.0 (9)

20.8 (5)
33.3 (8)
45.8 (11)

- 110 -

Table 4-12 details the proportion of subjects in each group meeting the recommended guidelines for
dietary intake of protein and energy as set above. The trends in proportion of participants meeting the
guidelines reflected the changes in intake reported in section 4.3.1 Total energy and protein intake.

There was relative maintenance in the proportion of subjects in the intervention group who were
meeting the guideline for protein intake, with no subjects consuming under 0.75g/kg protein at week
12. However, in the standard care group, the proportion of subjects below the guideline of 0.75g/kg/d
increased, with a relative decrease in those meeting the guideline intake (Table 4-12).

A greater proportion of intervention group were consuming the recommended energy intake of
>125kJ/kg following intervention (17% to 29%), however, the standard group decreased the proportion
with adequate intake (40% to 21%) and increased in the proportion consuming inadequate kJ
(<100kJ/kg; 36% to 46%). Both groups had a similar proportion of subjects consuming 100-125kJ/kg
(38% and 33% respectively).

As mentioned above in Section 4.3.1.1 Under-reporting of energy

intake, these results must be considered in context of potential changes in the accuracy of participants
reporting of energy intake. This will be considered further in Section 5.1.4.

4.3.3 Macronutrient contribution to energy intake

Qualitative change in dietary intake was assessed from the contribution of energy intake, from each
macronutrient, protein, fat, carbohydrate and alcohol, at both baseline and week 12 for each group, as
per Table 4-13. Statistically significant changes were noted in a reduction in the energy contribution
from protein in the intervention group, and increase in contribution of energy from carbohydrate in the
standard care group.

Table 4-13 Macronutrient contribution to energy intake % (mean

SD) for participants in

intervention (n=24) and standard care (n=24) groups.

Macronutrient

Time

Protein

Baseline
Week 12
Baseline
Week 12
Baseline

Fat
Carbohydrate

Intervention
n=24
18.5
16.4
33.0
33.8
45.9
- 111 -

Standard care
n=24*
18.6
18.1
34.5
43.2

Alcohol

Week 12
Baseline
Week 12

44.9
2.5
4.7

46.0
3.4
2.3

*2/26 missing intake data; **p<0.05

4.4

Change in clinical variables

A number of exploratory clinical variables collected were assessed for change over the treatment
period. Whilst these variables do not form part of the study hypotheses, they are indicators of interest
in context of with the nutrition status results presented. Descriptive results and bivariate analysis will
be presented for each variable by treatment group at baseline and week 12. Within group difference
at each time point by paired t-test, and a difference in change over time by independent samples ttest.

Biochemistry collected at each time-point were electrolytes and serum protein (albumin) as standard.
Iron studies (including transferrin saturation) and C-reactive protein were collected less routinely, and
were available for around 80% and 60% of patients respectively (evident in Table 4-14).

- 112 -

46, 47

Table 4-14 Proportion of participants by treatment group compared with guidelines

at

baseline and week 12 between intervention (n=24) and standard care (n=26)
treatment groups in a sample of pre-dialysis CKD patients
Intervention
%(n)
Week 0

Standard care
%(n)
Week 12

Week 0

Week 12

Albumin status
>40 g/L
35-39.9 g/L
<35 g/L

45.8 (11)
45.8 (11)
8.3 (2)

45.8 (11)
45.8 (11)
8.2 (2)

53.8 (14)
34.6 (9)
11.5 (3)

30.8 (8)
46.1 (12)
23.1 (6)

No evidence
Evidence (HCO3 <22 mmol/L)
Inflammation
No inflammation (CRP <5 mg/L)
Mild inflammation (CRP 5 10 mg/L)
Inflammation (CRP >10 mg/L)
Missing
Transferrin Saturation
Good (> greater 20%)
Inadequate (<20%)
Missing
Appetite
Fair/Poor
Good
Very Good

75.0 (18)
25.0 (6)

70.8 (17)
29.2 (7)

65.4 (17)
34.6 (9)

69.2 (18)
30.8 (8)

62.5 (10)
12.5 (2)
25.0 (4)
8

37.5 (6)
43.8 (7)
18.1 (3)
8

53.3 (8)
26.7 (4)
20.0 (3)
11

50.0 (8)
25.0 (4)
25.0 (4)
10

60.0 (12)
40.0 (8)
4

66.7 (14)
33.3 (7)
3

61.9 (13)
38.1(8)
5

45.0 (9)
55.0 (11)
6

33.4 (8)
50.0 (12)
16.7 (4)

20.8 (5)
62.5 (15)
16.7 (4)

34.6 (9)
34.6 (9)
30.8 (8)

38.5 (10)
46.2 (12)
15.4 (4)

Metabolic Acidosis

HCO3 = Bicarbonate; CRP = C-reactive protein

At baseline, the proportion of patients in the intervention compared to the standard care groups were
similar with low albumin, metabolic acidosis, inflammation, inadequate transferrin saturation and
appetite rating of fair or poor (Table 4-14). At week 12 there was a greater proportion of the standard
care group with inadequate transferrin saturation (11/20 versus 8/21), low albumin status (6/26 versus
2/24) and poor appetite (10/26 versus 4/24) compared with the intervention group.

Table 4-15 Change in clinical variables by intervention (n=24) and standard care (n=26) groups
during 12-week randomised controlled intervention in pre-dialysis CKD.

- 113 -

Clinical variable
(ideal value)

Weight (kg)
GFR (ml/min)
Albumin (>35 g/L)
Bicarbonate (22.029.0 mmol/L)
C-reactive
protein
(<5mg/L)
Transferrin
Saturation (>22%)
Serum Potassium (<
5.5 mmol/L)
Serum
Phosphate
(0.8 - 1.6 mmol/L)

Intervention n=24
(mean sd)

Standard Care n=26

(mean sd)

Mean difference in
change# (95% CI)

Week 0
73.516.1
23.47.9
39.24.8
24.24.7

Week 12
73.815.7
22.96.8
39.54.9
23.55.3

Week 0
76.918.0
21.76.2
39.24.4
23.54.4

Week 12
77.420.1
21.47.2
37.14.6**
25.75.0

0.14 (-1.3 to 1.6)

0.3 (-1.75 to 2.34)
-2.26 g/L (-4.0 to -0.5)*
3.46 (1.42 to 5.50)**

2.4 (0-25)

5.8 (0-13)

3.3 (0-53)

4.9(0-135)

-11.16 (-13.02 to 35.35)

23.610.3

26.311.2

25.311.6

21.811.0

-6.25 (-12.2 to -0.4)*

4.70.6

4.70.6

4.70.7

4.60.5

-0.13 (-0.41 to 0.16)

1.30.3

1.20.2

1.40.3

1.30.2

0.06 (-0.13 to 0.13)

*p<0.05; **p<0.01 t-test # change in standard care - intervention

Table 4-15 provides the clinical variables as continuous, and evaluates statistical differences in
change over the treatment period. There was negligible difference in change between groups for
CRP, GFR, potassium and phosphate. Albumin, bicarbonate and transferrin saturation had a mean
difference in change which reached statistical significance. The significant change in albumin status is
reflected in the increased proportion of standard care participants with low albumin status at week 12
compared with the intervention group (23 versus 8% <35 g/L).

Mean weight appeared to be stable, and if not, to increase slightly in standard care group (Table 415), which is in contrast to the changes evident with BCM (Section 4.2.1 Body Cell Mass). The
correlation between weight change and body cell mass change was not significant (r=0.178, p=0.266).
The relationship was further assessed for agreement using the Bland and Altman method (Figure 4-2).
There was a negligible mean difference in weight change from BCM change of -0.21%, however the
limits of agreement ( 2SD) were significantly wide as change from -17.9% to 17.6% which is
considered unacceptable for clinical purposes.
The difference in weight change from body cell mass change showed systematic bias, represented by
a significant correlation between the mean of BCM and weight change and difference (BCM change
weight change) r=0.46 p=0.003 (equation y=1.5x-1.1 SEE=5.3 (y=0, x=0.7)), indicated on the diagonal
line below. Therefore, for values below 0%, change tended to be signified by a greater reduction in
BCM than weight, and above 0%, by a greater increase in BCM change compared with weight
change.

- 114 -

Difference (% Body Cell Mass Change - % Weight Change)

25.00

0.00

-25.00

-10.00

-5.00

0.00

5.00

10.00

Average (% Body Cell Mass and % Weight Change)

Figure 4-2

Bland-Altman plot comparing methods of change in body weight, body cell

mass change compared with weight change over 12 week treatment period

4.5

Adjusted estimates of the impact of structured nutrition

intervention on nutrition-related outcome measures

Each of the nutrition status and dietary intake outcome variables were further investigated in a
multivariable analysis as described in 2.5 Statistical analysis.

4.5.1 Multivariable analysis for nutrition status outcomes

Baseline variables identified to potentially impact change in outcome variables between the treatment
groups (including co-morbidity (as Charlson index), age, gender, GFR, BMI, recent hospitalisations or
dialysis access placement, living situation and dialysis treatment plan) were assessed for confounding
by bivariate statistics. The few variables identified as confounders were added to each of the original
ANCOVA analysis. In an exploratory analysis, gender was also assessed as an additional covariate
to each of the ANCOVA models. Main effects and interactions were assessed. The final model only
included significant covariates.

- 115 -

Table 4-16 is a summary of the change in nutrition related outcomes as a result of the treatment group
allocation. Both unadjusted and adjusted values are provided, also with the statistical significance and
2

adjusted r , to give an estimation of the proportion of variation explained for each of the outcome
variables by treatment group allocation, fitted with the variables placed in the footnote. Adjusted
estimates did not change the outcome from treatment group allocation significantly, however,
strengthened the evidence against the null.

4.5.1.1 Adjusted change in BCM

When change in BCM was assessed for the confounding variables, the main effect of gender was
significant (p=0.01) was therefore left in (interaction with treatment group was not significant
(gender*treatment group p=0.115)). For both absolute (g) and relative change in BCM, after adjusting
for the main effect of gender on relative change in BCM (%), the predictive ability of the model
improved. The mean difference in change between the treatment groups increased in magnitude from
the previous bivariate analysis presented in Section 4.2.1. Although this did not reach statistical
significance, this difference was largely contributed by a significant reduction in BCM in the standard
care group.

- 116 -

Adjusted

Unadjusted

Adjusted

Unadjusted

Adjusted

Unadjusted

Adjusted

Unadjusted

Adjusted

Unadjusted

0.98 (-0.15 to 2.11)

0.39 (-0.81 to 1.58)

-1.44 (-2.72 to -0.16)

-1.41 (-2.58 to -0.24)

-0.13 (-0.21 to -0.05)

- 117 -

-2.39 (-4.02 to -0.77)

-1.82 (-3.58 to 0.07)

-0.08 (-0.03 to 0.20)

-0.04 (-0.73 to 0.16)

-0.11 (-0.19 to -0.03)

-0.05 (-0.13 to -0.03)

-0.07 (-0.15 to 0.02)

22.11 (12.77 to 31.45)

17.69 (8.20 to 27.18)

1.20 (-0.30 to 2.70)

1.07 (-0.74 to 2.88)

3.85 (-1.00 to 8.70)

3.51 (-2.05 to 9.06)

mean (95%CI)

Mean difference

-7.92 (-14.29 to -1.55)

-6.33 (-13.00 to 0.35)

11.36 (4.68 to 18.04)

14.19 (7.62 to 20.75)

-1.15 (-2.28 to -0.03)

-0.54 (-1.83 to 0.76)

-3.25 (-6.89 to 0.38)

-1.52 (-5.49 to 2.46)

0.05 (-1.05 to 1.14)

0.53 (-0.73 to 1.80)

0.60 (-2.94 to 4.13)

1.99 (-1.88 to 5.87)

Standard care

BCM = Body Cell Mass; PG-SGA = Patient-Generated Subjective Global Assessment

a
adjusted for baseline BCM and gender
b
adjusted for baseline kJ and protein intake and gender (note treatment group*gender significant)
c
adjusted for gender (note treatment group*gender significant)
BOLD = statistically significant difference
Italics = models identifying significant interaction effect with gender

PG-SGA

Protein (g/kg)

Energy (kJ/kg)

BCM (kg)

BCM (%)

Intervention

from intervention), both adjusted and unadjusted values

8.79

4.41

2.30

0.56

22.85

14.11

2.63

1.44

2.59

1.63

F-stat

0.005

0.042

0.137

0.458

<0.001

<0.001

0.114

0.238

0.116

0.209

p-value

0.21

0.19

0.44

0.56

0.63

0.34

0.32

0.01

0.24

0.01

Adj r

Table 4-16 Impact of nutrition intervention on nutrition-related outcomes, difference in mean change between treatment groups (standard care

Proportion of change in BCM was assessed by multinomial logistic regression (Figure 4-1). Although
the direct relationship between category change in BCM and treatment group did not reach statistical
2

significance(X (2) = 4.46 p=0.108; Naglekerke r = 0.118), relative to a decrease in BCM, participants
who received intervention were 4.3 times (95% CI 1.02-18.38) more likely to have increased in BCM
2

(p = 0.047). When adjusted for gender, the relationship to treatment approached significance (X (2) =
2

4.90 p=0.086; Naglekerke r = 0.349) and relative to decrease in BCM, the odds of participants from
the intervention group to increase in BCM was to 5.2 times (95% CI 1.08-25.05), compared standard
care (p=0.040).

4.5.1.2 Adjusted change in PG-SGA score

Change in PG-SGA score had a significant relationship with both gender and Charlson co-morbidity
index (p<0.05).

Therefore placed together into the ANCOVA model this analysis provided an

increased adjusted mean difference in PG-SGA score as a result of treatment allocation (Table 4-16).
In this model, however, the treatment and gender interaction was statistically significant (p=0.009 for
PG-SGA score), when the main effect of gender was not (PG-SGA*Gender p=0.132). This issue will
be addressed later in 4.7 Gender difference in the impact of structured nutrition intervention. As a
result of this finding (significant gender and treatment group interaction), future discussion for PG-SGA
will be based on the bivariate (unadjusted) results.

4.5.1.3 Adjusted change in dietary intake

Change in protein (g/kg) and energy intake (kJ/kg) were included together in a MANCOVA model to
determine the independent relationship between each dietary intake variable and treatment group
allocation. Both dietary intake variables had a significant relationship with gender (p<0.05). Therefore
the MANOVA model, evaluated change in energy and protein intake, included baseline intakes (per
kg) as covariates, and gender as a factor (both as a main effect (p=0.242) and interaction (p=0.006)
2

(Wilks Lamda 0.626 (2,41) 11.27 p<0.001 n = 0.355).

Independent of baseline intakes and gender, change in kJ/kg intake remained significant (F (1,42)
2

21.57 p<0.001, adjusted r = 0.432) and change in protein intake increased, however, remained non2

significant (F (1,42) 2.26 p=0.141, adjusted r = 0.638). It is important to re-emphasis that in this
model, the gender and treatment interaction was significant, and the main effect of gender was not

- 118 -

significant. Again, as a result of this finding (significant gender and treatment group interaction), future
discussion on dietary intake will be based on the bivariate (unadjusted) results.

4.6

Gender difference in the impact of structured nutrition

intervention
Throughout the multivariable nutrition outcomes analysis, a significant treatment and gender
interaction was apparent in a number of the models. This suggests an effect modification by gender
influencing the treatment outcomes. As this was an unexpected result, sub-group statistical analysis
was not performed. However, a descriptive comparison of gender difference at baseline and changes
due to treatment allocation are presented.

Table 4.17 Baseline characteristics of participants completing study by gender (n=50) in a

sample of pre-dialysis CKD patients

Male (n=31)
Intervention n (%)
Age (mean SD)
GFR (ml/min)
Charlson co-morbidity index
2
BMI (kg/m )
BCM (kg)
% weight change in previous 6 months
Nutritional status
SGA A Well nourished % (n)
SGA B Moderately malnourished % (n)
Home setting
Lives with partner % (n)
Lives alone % (n)
Lives with family or carers % (n)
Shopping (% primary) % (n)
Cooking (% primary) % (n)
Attended pre-dialysis education % (n)
Previous individual dietetic intervention % (n)

15 (50)
69.9 11.3
22.6 6.7
6.6 1.9
27.6 4.6
36.5 5.4
-0.3 2.5

Female
(n=19)
9 (43)
69.4 13.4
22.5 7.7
6.1 1.8
26.9 5.8
24.3 4.0
-1.6 4.2

90.3 (28)
9.7 (3)

73.7 (14)
26.3 (5)

71.0 (22)
16.1 (5)
12.9 (4)
22.2 (8)
22.2 (8)
55.9 (19)
33.3 (12)

56.2 (10)
36.8 (7)
10.6 (2)
54.2 (13)
45.8 (11)
59.1 (13)
37.5 (9)

BMI = Body Mass Index; BCM = Body Cell Mass; SGA = Subjective Global Assessment

Baseline characteristics by gender are provided in Table 4-17. There appeared to be no systematic
differences in clinical variables between the groups (age, GFR, co-morbidities and BMI). There were
a greater proportion of females malnourished at baseline (5/19 versus 3/31 for males) and also a
greater proportion of females living alone (7/19 versus 5/31).

Females were more likely to be the

primary food providers, in terms of food purchasing (13/19 versus 8/31) and preparation (11/19 versus
8/31). Both genders had a similar amount of dietary intervention at baseline.

- 119 -

A descriptive assessment of the influence of gender on study outcomes was performed in Table 4-18
and Figure 4-3.

There was a significant improvement in study outcomes for the female gender,

however, negligible change for most outcomes for the male gender.

Overall, the response to

intervention appeared to have a different outcome depending on gender. The potential influences of
this effect will be discussed further in 5.4 Gender Differences.

Table 4-18 Outcome variables of the intervention study split for gender for week 0 and week 12
(mean (SE)) and mean difference in change (mean (95%CI)) in intervention (n=24)
and standard care (n=26) groups

M
F
Both

Intervention
ControlMean difference in change
M (n=15); F (n=9) M (n=16); F (n=10)
mean differencese mean differencese
1.46 1.94
1.35 2.61
0.11 (-6.86 to 6.65)
9.85 (0.10 to 19.6)
2.93 8.70
-6.92 8.70
1.99 (-1.88 to 5.87) -1.52 (-5.49 to 2.46)
3.51 (-2.05 to 9.06)

M
F
Both

0.55 0.70
0.18 1.02
0.54 0.70
-1.90 0.85
0.53 (-0.73 to 1.80) -0.54 (-1.83 to 0.76)

0.38 (-2.2 to 2.9)

2.43 (0.0 to 4.8)
1.07 (-0.74 to 2.88)

Energy kJ/kgM
F
Both

6.53 16.17
-1.69 16.5
21.8 14.08
-15.01 18.11
11.36 (4.68 to 18.04)-6.33 (-13.00 to 0.35)

8.22 (-4.22 to 20.68)

36.85 (21.2 to 52.5)
17.7 (8.2 to 27.2)

Gender

BCM (%)*

BCM (kg)

Protein g/kg M
F
Both
PG-SGA

M
F
Both

-0.28 0.06
0.10 0.19

-0.11 0.09
-0.30 0.30

-0.09 (-0.30 to 0.13)

0.40 (0.02 to 0.64)

-0.07 (-0.15 to 0.02) -0.11 (-0.19 to -0.03)

0.04 (-0.73 to 0.16)

-0.81 0.61
-1.89 2.80

-0.93 0.60
2.70 3.56

-0.1 (-1.8 to 1.63)

-4.59 (-7.7 to 1.5)

-1.44 (-2.72 to -0.16) 0.39 (-0.81 to 1.58)

-1.82 (-3.58 to 0.07)

*n=13 (M), n=8 (F); significant difference in BOLD, M = Males, F = Females, BCM = Body Cell Mass; PG-SGA = Patient
Generated Subjective Global Assessment

- 120 -

Figure 4-3

Change in energy intake and body cell mass over the intervention period, by

treatment group split for gender, intervention

standard care
MALES

FEMALES
25

20
15
10
5
0
-5

Week 0

Week 12

-10
-15

Change in energy intake (kJ/kg/day)

Change in enregy intake (kJ/kg/day)

25

20
15
10
5
0
-5

Week 0

-15
-20

-20

MALES

FEMALES

0.5

0
Week 0

Week 12

Change in BCM (kg)

Change in BCM (kg)

0.5

-0.5

Week 12

-10

0
Week 0

Week 12

-0.5

-1

-1

-1.5

-1.5

The graphs in figure 4-3 represent a similarly patterned magnitude of change between change in
reported energy intake and BCM change in both females and males. The females clearly demonstrate
a greater magnitude and difference in change for both variables between treatment groups compared
with males.

- 121 -

Chapter 5:

Discussion of impact of randomised-

controlled trial of nutrition intervention in pre-dialysis CKD

on nutritional status

5.0. Introduction
The aim of this phase of the study was to investigate the impact of providing individualised structured
nutrition intervention on nutrition status outcomes, including body composition, nutritional status and
dietary intake, in patients with pre-dialysis chronic kidney disease (CKD). The first section of this
chapter addresses the results of the randomised-controlled trial (RCT) in relation to the study
hypotheses. This discussion will compare the results to other studies in the literature and consider the
clinical implications of the findings.

The next section addresses effect modification due to the

significant gender differences displayed in 4.7 Gender Difference in the impact of structured nutrition
intervention. This section aims to put the results of this study into context with gender considerations
in the literature and also to direct recommendations for future research in this area.

5.1

Nutrition status as a result of intervention

5.1.1. Change in body cell mass

The data demonstrate that overall, there was an adjusted mean difference (95% CI) in BCM change
between intervention and standard care groups of 1.2kg (-0.3 to 2.7 kg), or 3.9% (-1.0 to 8.7%). This
was contributed by a considerable decrease in BCM in the standard care group (adjusted mean
change -1.15kg (-2.3 to -0.0 kg) or -3.3% (-6.9 to 0.4%)) and a maintenance in the intervention group
(adjusted mean change 0.05kg (-1.1 to 1.1 kg) or 0.6% (-2.9 to 4.1%)). It is known that having change
265

of less than 0.7% in BCM over three months is equivalent to weight stabilisation

, as displayed in the

intervention group. Therefore, the mean 3.3% loss of BCM in the standard care group is clinically
significant and appears to capture a trajectory of body cell mass loss at this stage of CKD, which was
attenuated in the intervention group

- 122 -

The benefit of the intervention was further highlighted when the change in BCM was categorised into
improved (>1%), maintained (1%) or decreased (<1%).

More than twice the proportion of the

intervention group had improved (change >1%) BCM compared with standard care. Relative to a
decrease in BCM, the odds of participants from the intervention group increasing in BCM was 5.2
times (95% CI 1.08-25.05), greater than those receiving standard care treatment (p=0.04). This again
highlights the clinical significance of this intervention.

5.1.1.1 Clinical relevance of BCM change

BCM is the functional body compartment determining energy expenditure, protein needs, and the
metabolic response to physiologic stress. BCM declines are linked with loss of strength, immune
266, 267

function, pulmonary function, increased disability and mortality

. This highlights the prognostic

significance of the regression seen in the standard care group. Not surprisingly, individuals with
limited reserves of muscle mass respond poorly to metabolic stressors

268

. This is notable in the study

by Dolson et al (2003) where there was a significantly reduced survival of patients initiating dialysis (a
87

stressor) with depleted BCM .

BCM represents the metabolically active tissue component of fat-free mass (FFM). FFM is a more
commonly defined unit to describe changes in body composition related to nutritional status.
Therefore, converting the original TBK values from this investigation to FFM can assist with further
comparison of the results from this study to the literature. From the development of a cellular model,
Wang et al (2001) established the mean TBK/FFM ratio to be 60.4 mmol/kg, in agreement with original
studies by Burkinshaw and Morgan (1985)

89 269

. However, Wang et al (2001) also noted a wide

variation range of possible TBK/FFM ratio from 48.3 to 70.7 mmol/L due to the influence of variable
269

fluid distribution

. Using the formula TBK/FFM = 60.4 mmol/kg (range 48.3 to 70.7 mmol/kg), the

mean difference in change in FFM from this investigation is approximately 2.4kg (range 1.9 to 2.7kg).
Therefore, regardless of the inaccuracies for the formula, it is clear this change far exceeds what is
55

considered to be a clinically significant change in FFM (0.5 1.0kg) .

5.1.1.2 Body protein change in response to nutrition support

To the authors knowledge, the only literature to date reporting change in BCM by TBK in pre-dialysis
patients are either observational or in exercise training interventions. Observational studies using TBK

- 123 -

270

over the treatment period report a progressive reduction in BCM significant for females only

. This

difference in gender is in agreement with the results of this study, to be considered in 5.4 Gender
difference in response to intervention. Other studies have used bioelectrical impedance (BIA) for the
assessment of FFM and BCM in pre-dialysis. Using this method, it appears that FFM can be relatively
well preserved in pre-dialysis, even with progressive decline in renal function when patients receive
91, 92, 271

regular medical and dietitian contact for a low-protein diet

. However, this does not account for

the potential expansion of the fluid component of FFM, particularly intra-cellular fluid, with declining
renal function. Therefore, declines in cellular mass, as demonstrated in the standard care group of
this sample could be potentially masked by increases in fluid balance.

Similar to the findings in this study, a control group arm in an exercise intervention study has also
indicated BCM decline (measured by TBK) in pre-dialysis CKD

242

. A decline of around 5% over 12-

weeks was measured in patients prescribed a low protein diet (0.6g/kg/day) without follow up

242

. This

re-emphasises that in the absence of regular individualised dietetic consultation, pre-dialysis patients
are likely to decrease BCM. It is important to note, provision of a restrictive diet prescription resulted in
a greater loss of TBK ~5% compared with 3.3% in the standard care group featured in this thesis.

The intervention group with patients in pre-dialysis (median GFR 24.8ml/min) in the abovementioned
exercise study, demonstrated a significant improvement in nitrogen balance including increase in
242

TBK

, following 12 weeks of resistance training.

The anabolic effects of resistance training (mean

difference in change TBK approx 6%) were observed despite subjects uraemia and self-reported low
energy intakes (<90 kJ/kg). This indicates resistance training is an effective counter-measure to the
negative effects of uraemia, protein utilisation and nutritional status. Other studies based on low
242

intensity resistance and aerobic training have shown significant improvements in the physical
272

psychosocial capacity

and

of pre-dialysis patients. However, limitations of the exercise interventions

include the generalisability and feasibility in practice. Each of the interventions have experienced poor
response rates and the requirement for specialty exercise physiologists for each session to supervise,
monitor progress and adjust programs are possibly unsustainable in clinical practice.

- 124 -

The clinical relevance of the results of this thesis that is the attenuation of functional body cell mass
with nutrition intervention compared to significantly decreased functional body mass (BCM) with
standard care (1.2 kg BCM; approximately 2.4kg FFM), holds up when compared with results of
interventions from other clinical areas. Nutrition interventions in oncology and HIV have shown similar
response in BCM and/or FFM to nutrition support. Provision of a combination of resistance training
and protein supplementation over 14-weeks resulted in 0.74kg gain in BCM in a group of HIV-infected
273

men and women, compared with maintenance during the lead up control period

. Recently, the use

of testosterone (oxandrolone) in HIV-infected males patients resulted in a comparable magnitude of

gain in BCM between the placebo and highest dosage group (1.3kg BCM)

274

. In an investigation in

the change in body composition in 36 cancer outpatients receiving radiotherapy to the head and neck
area, FFM loss (measured by BIA) was significantly higher in the standard care group with a mean
275

loss of 2.2 kg versus 0.3 kg in the intervention group

5.1.1.3 Gender difference in BCM change

When exploring the effect of gender on the relationship between treatment group allocation and
change in BCM, it was possible to account for significantly more variance in the ANCOVA models (%
2

BCM, R 0.01 to 0.24). It appeared it was not only the physiological difference in BCM at baseline by
gender which contributed to this, but also the change in BCM results by treatment group when split for
gender. Between treatment groups the mean difference in BCM change (95% CI) in females was
large at 9.9% (0.1 to 19.6%), or 2.4kg (-0.1 to 4.8kg), whereas for males, no real difference in change
between the groups 0.1% (-6.9 to 6.6%) and 0.4kg (-2.2 to 2.9kg) was evident (Table 4-18). This
highlights the significant magnitude of change in female participants, contributed mostly by a reduction
in the standard care group (6.9 8.7%), while men displayed a mean maintenance of BCM,
regardless of treatment allocation.

This result supports another observational investigation in CKD, where BCM measured by TBK
deteriorated in females, yet was maintained in males in the stage prior to dialysis

270

. This distinction

may be due to inherent biological differences between genders in response to uraemic malnutrition.
This will be explored further in 5.4 Gender difference in response to intervention. Overall, this study
confirmed that by providing a regular, nutrition intervention, based on the principles of selfmanagement, that muscle wasting in uraemia may be attenuated or reversed.
- 125 -

This impact of

intervention was significantly greater for females, who did have a higher rate of malnutrition at
baseline, and a poor response to the standard care treatment.

5.1.2 Change in Subjective Global Assessment

Subjective global assessment rating for patients in the intervention group maintained in all well
nourished (79%) and improved in all malnourished (21%) patients (Table 4-6). In the standard care
group, most patients maintained a well-nourished nutritional status, however, a number of wellnourished patients regressed in nutritional status, resulting in the rate of malnutrition increasing from
11% at baseline to 22% (including 1 severely malnourished, SGA C) for this group. In addition, the
change in SGA over the treatment period was verified by change in BCM. Participants who regressed
an SGA category significantly decreased in BCM (-7.1% (95% CI -13.5 to -0.6%)) and improvement in
SGA category saw an increase in BCM of 4.8% (-11.4 to 21.0) (Table 4-7).

As with BCM, change in SGA is not a commonly reported outcome measure in the CKD literature.
SGA is most commonly used to determine the prevalence and prognostic significance of
99

malnutrition . In Stage IV and V pre-dialysis, SGA B or C are significant independent predictors of

poor outcome in CKD and related to higher mortality and morbidity

29, 101

. Also, malnutrition in pre-

dialysis appears to be a predictor of shorter time to dialysis initiation. A prospective observational

study over 12-months in the pre-dialysis phase of 50 patients, provided a strong association between
malnutrition (SGA B or C) and progression to dialysis, increased rate of acute hospitalisations and
31

death, independent of GFR at baseline .

Only one other intervention study with SGA as an outcome measure was identified in pre-dialysis
patients.

This small intervention study (n=11) reported improvements in nutritional status when

patients received three-monthly nutrition intervention.

At 6 months, all well-nourished patients

maintained nutritional status, and 2 of the 3 malnourished patients were reclassified as well199

nourished

This intervention was not as intensive as the investigation featured in this thesis,

however, highlights the importance of intervention and follow-up for at least maintenance of nutritional
status in pre-dialysis.

5.1.3.1 Change in scored nutrition assessment

- 126 -

Improvement or maintenance in category of nutrition status with intervention was reflected by change
in PG-SGA score, with a significant mean difference in PG-SGA score (1.8 (0.1 to 3.6, p=0.04)). This
was attributable to a reduction in score in the intervention relating to a decreased risk of malnutrition.
While the change in PG-SGA by treatment group was statistically significant, the magnitude of change
recorded of 1.8 is not considered to be clinically significant.

Table 4-7 showed a change in PG-SGA score of 5 or 6 was evident in patients who improved or
deteriorated according to SGA. In an investigation of radiation-oncology outpatients, the change in
55

PG-SGA score of 9 was required to move one SGA category . A previous investigation into PG-SGA
109

score in dialysis patients indicated a median score of 16 for SGA B

, which is well above the median

score of 9 in this study (Table 3-3). The MIS contains scoring components for dialysis vintage, which
for the purposes of this investigation were omitted.

Therefore, it is possible that in pre-dialysis,

compared with dialysis that the scoring of both the MIS and PG-SGA would be of a smaller magnitude.
Further investigation into the breakdown of scoring components and inter-rater reliability in both predialysis and non-dialysis patient groups is warranted to further develop the evidence base for
preferential use of either tool in CKD.

5.1.4. Dietary intake

Dietary intake represents the intermediate outcome measure, which is integral in the cascade of
nutrition intervention to facilitate the flow-on effect of positive long-term outcome measures. The aim
of the intervention was to maintain or improve energy and protein intake of the intervention
46

participants, to within the recommendations of the evidence-based guidelines .

5.1.4.1 Dietary energy intake

Dietary energy intake significantly improved as a result of the intervention and decreased in standard
care, resulting in a final mean difference of 17.7kJ/kg (8.2 to 27.2 kJ/kg) or 1202kJ (307 to 2095 kJ/kg)
(Table 4-9). Improving energy intake is an important intermediate outcome driving improvements in
158, 189

nutritional status. This is in agreement with the literature

. In a short-term observational study,

an increase in energy intake of pre-dialysis CKD patients resulted in an improvement in various

nutritional parameters

189

- 127 -

However, the overall energy intake results must be considered with reference to the varying levels of
under-reporting identified in Table 4-10 and 4-11.

Overall there appeared to be underreporting in

around 60% of the weight stable patients in the sample at each time point (Table 4-10). This level of
under-reporting is equivalent to levels reported in the CKD literature

135, 136

. The proportion of subjects

in the intervention group who were identified to be under-reporting decreased from baseline to week
12. It is possible that the regular intervention and dietary recall may have improved the intervention
groups reporting of daily intake in 3-day record book at week 12 compared with their baseline record.
This has implications to the overall results as it may represent a bias supporting an inflated
improvement in dietary energy intake in the intervention group.

This is particularly relevant

considering the significant increase in reported energy intake overall (mean increase 11.4 kJ/kg (95%
CI 4.7 to 18.0)).

Moreover, the standard care group reported a decrease in EI:BEE ratio in

participants who were weight stable at baseline and week 12, representing an increase in underreporting by week 12. This may account for some of the overall reduction in energy intake noted in
this group (mean decrease of 6.3 kJ/kg (-13.0 to 0.4)). However, despite the differences in underreporting and how that may have impacted the overall results, the real point is the significant
magnitude of difference between the treatment groups, and in particular, how the change in dietary
intake influenced the primary outcomes of change in nutritional status and body cell mass. The
relationship between change in energy intake and body cell mass is highlighted in the results (Table 416). This is particularly reflected in the magnitude demonstrated by the gender sub-analysis looking at
change in energy intake and BCM in the female participants (Table 4-18 and Figure 4-3).

Whilst it is agreed that underestimation of energy intake is a common phenomena, it is difficult to

276

identify accurately, particularly depending on individual variation

associated with estimates such as the Goldberg cut-off

124

and the many assumptions

method used in this situation. This method

73

is simple, less expensive and more practical than using the gold-standard, doubly labelled water .
This method compares the daily mean reported energy intake with an expected minimum intake
assuming a sedentary lifestyle and using a generic equation to calculate basal energy expenditure (or
BEE)) for all individuals. Therefore, this method is based on a number of key assumptions. Firstly, the
use of this equation does not account for individual variation in energy expenditure or requirements.
In addition, use of the equation is appropriate only for individuals in energy balance. Energy balance

- 128 -

is difficult to determine, particularly in this sample, where only weight stable participants, those who
had within 2.5% weight change in the previous 6-months were included. Considering the issues of
weight change in pre-dialysis CKD patients being significantly impacted by fluid shifts (see Section 4.4
Change in clinical variables), this fundamental assumption of energy balance for use of the Goldberg
cut-off cannot be guaranteed in this sample. Therefore, use of the Goldberg cut-off estimate expected
energy intake for weight maintenance is considered a very crude estimate of the validity of reported
intake, and the results of the levels of under-reporting must be considered together with these
limitations. Therefore, the affect of under-reporting and in particular change in the level of underreporting is considered relatively minor in the overall message of this study.

This intervention provided mostly food-based strategies to improve energy intake (Appendix F)
affording improvements in nutritional status. During the course of this study, 3 of the intervention
patients were recommended to purchase oral supplements (1-2 per day) designed for pre-dialysis
patients (high energy, low protein and electrolyte beverages), providing 1050kJ and 6g protein each
serve, to assist them when strategies to increase food intake were insufficient to meet requirements.
All of these patients were female, 2 were malnourished at baseline, and all were rated as wellnourished at week 12. All these patients were also provided with strategies to improve energy intake
from food-based sources and were encouraged to use the supplements as an addition to their usual
daily intake. At week 12, only 1 of the participants continued with purchasing supplements, which
indicates the change in mean intake was mostly a result of increased intake from food sources.
Overall the intake of these participants significantly increased by a mean of 34kJ/kg/day or 1990kJ/day
from baseline to week 12, which had a greater magnitude of increase compared with those not
recommended supplements in the intervention group (n=21, mean increase of 9.1kJ/kg/day or
820kJ/day).

In a Cochrane review examining the evidence for dietary strategies in the treatment of illness-related
malnutrition, the provision of oral supplements (e.g. high energy protein beverages) was
278

advantageous for promoting improved energy intake and weight gain over dietary advice alone

Provision of dietary advice appeared to make no difference to energy intake or weight compared with
no advice.

This is in contrast to the results of the study reported in this thesis and recent studies in

- 129 -

159, 279

oncology

. In particular, Ravasco et al reported individualised dietary counselling advantageous

to improve nutritional intake, status and QOL in radiation-oncology patients over provision of dietary
279

supplements or no advice

. A criticism of the studies included in the Cochrane review is the narrow

outcome measures used to assess the benefit of intervention. The results of more recent studies in
oncology and of this current intervention study in CKD would add to the Cochrane review highlighting
that nutrition counselling with or without supplements improves energy intake and a broad range of
159, 279

other patient outcomes, compared with providing no advice

Further to this, in the case of pre-dialysis CKD, the provision of dietary advice to improve energy
intake may be practically advantageous over the provision of oral supplements only. Dietary advice
providing skill development for managing uraemic symptoms and optimising nutritional status with
food ( supplements as required), the strategy adopted in the study featured in this thesis, is
potentially a most sustainable model.

In a study of 111 colorectal oncology patients undergoing

radiation treatment, Ravasco et al demonstrated that providing nutritional counselling based on regular
foods, as opposed to nutritional supplements resulted in lower incidence and/or severity of symptoms
279

compromising intake

. Considering CKD is a progressive and life-long disease, providing

interventions that aim to up-skill patients to manage issues of uraemia-induced anorexia, is thought to
be valuable through the progression of their disease and dialysis treatment.

5.1.4.2 Dietary protein intake

Sufficient energy and quality protein (>50% high biological value) intake is essential for the
171

maintenance of nitrogen balance

In this study protein intake remained relatively stable, with

negligible difference in change between the intervention and standard care groups. At week 12, 46%
of the intervention group, and only 25% of the standard care group were consuming the recommended
protein intake.

In reference to the under-reporting discussed above, evidence in the literature

indicates under reporters preferentially record lower intakes of fat and higher intakes of protein and
133, 277

carbohydrate as a percent of total energy

. Considering this, there is likely to be little impact of

under-reporting on recorded protein intakes.

- 130 -

46

The goal of the intervention was to maintain protein intake between 0.75g/kg and 1.0g/kg .
Controlling protein intake around 0.8g/kg level was encouraged for patients with significant uraemic
symptoms. In the absence of symptoms, controlling protein below 1.0g/kg was encouraged to not
46

compromise total energy intake .

Anecdotally from this study, participants who were non-symptomatic, and had good appetite had the
most trouble adhering to the recommendation of controlling protein intake (i.e. generally consuming
>1.0g/kg/day). In the case of the intervention group, a clinically significant result was achieved by
managing a slight reduction in protein intake (change of -0.07g/kg (-0.15 to 0.02) with an increasing
energy intake 11.36kJ/kg (4.68 to 18.04) (Table 4-9). This result was supported by a statistically
significant shift of decreasing contribution of energy intake from protein in the intervention group in the
macronutrient assessment (Table 4-13). In the environment of increasing overall energy intake this is
of interest, particularly noting increases in energy intake can be achieved with regular foods with a
lower contribution of energy from protein.

Although there remains a school of thought and some evidence that protein intake >0.8g/kg may
280

induce a faster rate of decline in GFR compared with low-protein diet

, the clear risks with provision

of a restricted-protein diet (<0.6g/kg/day) which may include simultaneous reduction in energy intake
and nutrition status parameters (TBK, weight, albumin, etc)

242, 281

. In most studies reporting outcomes

from low- or very-low protein diets, generally only young patients (<60 years) and uncomplicated (no
co-morbid disease, including diabetes) with established compliance to treatment are included. Even
so, to successfully achieve the goals of protein-restricted regimens, intensive nutritional counselling
and medical monitoring (at intervals at least monthly, mostly weekly) are essential components of
91, 256, 271, 282

care

. A prospective observational study of 239 patients confirmed that if patients in pre-

dialysis maintain sufficient energy intake (prescription of 35 kcal/kg or 146kJ/kg) and are medically
283

stable, BCM will remain stable on a low-protein diet

. However, in this particular study, BCM in 1 in 3

of the participants decreased significantly (mean decrease BCM 13%) due to an increase in co-morbid
illness and subsequent decrease in energy intake, highlighting the key to maintaining BCM is close
clinical monitoring, with a focus on sufficient energy intake

- 131 -

283

It is interesting to note that not one of the outcomes of protein restriction studies focused on patientcentred outcomes, such as quality of life, and very few examined valid indicators of nutritional
status

173, 179, 284, 285

. It can be hypothesised that by providing interventions that promote sufficient

energy intake, whilst controlling protein intake rather than restricting, potential gains in functional
status, nutritional status and quality of life may mimic benefits in delaying dialysis initiation seen in
protein restriction studies.

Overall the results of change in dietary intake reflect the goals of the nutrition prescription in the
intervention group and the decrements associated with providing standard care only. The changes
noted in each group indicate the importance of the intermediate outcome measure for influencing
change in the primary outcome of nutritional status.

5.2

Change in clinical measures

Change in clinical markers was assessed through the treatment period. Changes in clinical markers
for conditions such as inflammation, acidosis and anaemia and other exploratory variables are
reported so to be considered along with the investigation in change between the two treatment groups.
In part, this discussion will focus on change in serum albumin and weight as these measures are
historically the primary markers of body protein status for clinical interventions in dialysis patients
204, 206, 208

5.2.1

199-202,

and therefore will be useful for comparison to the literature.

Serum Albumin

In this intervention study, albumin significantly decreased in the standard care group (39.2g/L to
37.1g/L) and was maintained in the intervention group, resulting in a significant mean difference in
albumin status between the two groups at week 12 (-2.3 (-4.0 to -0.6) g/L p=0.012) (Table 4-14). This
magnitude of difference reflects the results of other nutrition support interventions in CKD.

systematic review of controlled trials providing nutrition supplementation in dialysis indicated a mean
difference in change in albumin status with oral supplements versus standard care of 2.4g/L (0.4 to 4.2
g/L)

207

. A similar magnitude of difference has been also reported in a community-based intervention

addressing barriers to food intake, with an increase in 2.1g/L in the intervention and 0.6 g/L in the
200, 208

usual care group

27

. Hypoalbuminemia has long-been considered an indicator of poor outcome .

- 132 -

In a recent investigation of the association between change-over-time and mortality predictability of

albumin, a significantly greater survival was associated with patients with increasing or stabilised
286

serum albumin

The worst outcome was associated with continually decreasing albumin,

286

particularly below 38 g/L

For the purposes of this thesis albumin status was not considered a primary outcome set by the
original hypotheses, due to the fact that albumin is considered to be a poor direct measure of
nutritional status. In addition, considering the weight increase and BCM decrease in the standard care
group, it suggestive of fluid overload, indicating the decrease in albumin is likely to be contributed by
haemodilution, rather than an overall decrease in synthesis. Also considering the proportion of the
standard care group with inadequate transferrin saturation and increased acidosis, it is difficult to
determine if this change in albumin is a true reflection of altered protein status during the investigation
period. Further research in this area is encouraged to establish the independent effect of nutrition
support on serum protein. To do so, the study must be designed to consider and adjust for not only
inflammatory status, but metabolic acidosis, hyperparathyroidism and other co-morbidities affecting
acute albumin synthesis.

5.2.2

Weight change

As indicated by the analysis in 5.4 Change in clinical variables, BCM changed to a different degree to
weight change, indicating a systematic bias in change between the two measures. It is acknowledged
that body cell mass may be lost even in the presence of stable or increasing weight. This investigation
indicates weight change is a poor indicator of change in body cell mass over a 12-week period in predialysis patients and highlights the difficulty in using weight to monitor changes in nutritional status in
the clinical setting for patients with CKD, particularly in the presence of variable fluid status.

It is for

this reason that assessment tools, such PG-SGA may be considered most useful as surrogate
measures to be included to the current panel of markers for change in nutrition status. Weight change
only reflects 10% of the total scoring system in these tools and other factors such as the impact of
recent gastrointestinal symptoms and dietary intake changes influence the assessment, amongst
other items.

- 133 -

There is an argument that if BCM changes at a different rate to weight then it is likely due to the action
287

of cytokines as opposed to inadequate dietary intake

. This is reported in a range of conditions

287

including cardiac cachexia, rheumatoid arthritis, burn trauma, AIDS and oncology

. However, the

limited data available from this study indicates no significant change in C-reactive protein over the 12
weeks between the groups; therefore one may deduce there was limited change in the action of
cytokines (see Table 4-5).

In addition, in this study BCM stabilised with an increased kJ intake

(intervention) and decreased, on average, with a decreased kJ intake (as seen in the standard care
group). Therefore, it is likely that the wasting noted is associated with poor dietary intake.

It may be concluded that the lack of agreement between BCM change and weight change is likely to
be due to fluid shifts, however, fluid was not a variable measured by this study, and therefore this can
only be hypothesised.

5.3

Use of telephone intervention

Telephone-delivered interventions have emerged as an increasingly popular means of delivering

288, 289

health promotion and behaviour change intervention for patients with chronic disease

. In a

recent evidence-based review, telephone counselling was identified to promote significantly greater
289

improvements in fruit/vegetable and dietary fat intake

. Similarly, a recent large scale intervention in

chronic heart failure identified significant lowering of all-cause mortality in patients receiving regular
290

telephone intervention, compared to those receiving written information only

. The study featured in

this thesis adds to, and supports this body of evidence for the use of telephone counselling to provide
a practical way to follow-up patients and ensure the translation of nutrition support messages to the
home environment.

5.4

Gender differences in response to intervention

In this study, there appeared to be significant effect modification by gender for each of the nutrition
related outcomes (Section 4.7). This is reflected in the similarities in outcomes for males receiving
either the intervention or standard care treatments; however, the females had a distinct and significant
difference in change depending on treatment allocation. The significant difference evident in females

- 134 -

by treatment group drove the overall results discussed above. It must be noted that this study was not
designed to detect differences according to gender therefore only inferences can be made from the
data. This finding of a gender difference has significant implications for practice and future research
studies.

5.4.1 Gender difference in nutritional status and body composition

Stenvinkel et al (2002) in a prospective observational study at the onset of dialysis noted significant
differences in the comparability of nutritional makers by gender. Reduced hand-grip strength, serum
2

albumin and lean body mass (assessed by DXA) was related to poor outcomes in males only . In
addition, elevated CRP was a strong and independent risk factor in males, elevated CRP was not
2

associated with poor outcome in females at all .

It was theorised this is perhaps related to the

atherogenic potential of inflammation, and protective nature of female sex hormones. Regardless,
there are clear disparities between nutrition-related measures to predict outcome depending on
gender.

Disparities have also been noted in an investigation to determine sex-specific associations between
HIV disease characteristics and body composition, prior to anti-retroviral therapy. Body fat differed
significantly by gender, with women with prior AIDS-defining illness having less fat than women
291

without such illnesses, whereas both BCM and fat was affected in men

. This is supported by

additional studies in HIV which have shown that women lose more fat mass than lean body tissue, but
292

men lose about equal quantities of both

. These distinctions appear to be due to inherent biological

differences between the sexes.

Longitudinal and cross-sectional studies report a greater magnitude of change in FFM in females with
293, 294

aging

. It is possible to therefore consider that the final phase of CKD decline prior to dialysis
295

mimics a fast-track of the natural trajectory of FFM decline

In addition to body composition

differences, in the MDRD pilot study a gender difference for dietary intake and response to
progressive renal failure was apparent.

In women, GFR

correlated with dietary energy intake,

whereas there was no relationship between renal function and energy intake in males

296

This

supports the trajectory of decreasing intake and subsequent BCM loss displayed by females in the
standard care group of this thesis.

- 135 -

Over the 12 week study period, the symptoms of uraemia and outcome measures did not change
significantly in males for either group. Therefore, it is important to consider the stage of CKD for this
study, with a mean GFR of 22.5 ml/min and inclusion criteria of GFR 15 30 ml/min. Adequate
nutritional status is common in cohorts of pre-dialysis patients without symptoms
GFR >10 ml/min.

297

, particularly with

Therefore, the underlying mechanisms resulting in the need for nutrition

intervention, for example, uraemic toxicity, may have less of an effect on males at this stage than
females, however, at this stage this can only be theorised.

5.4.2 Gender difference in response to the use of self management principles

Aside from the differences noted in the physiology between genders, it is possible to also consider the
difference in the style of treatment offered by the intervention. The primary focus of the intervention
was to facilitate self-management skills, which have been shown to improve clinical outcomes
298

compared with information only patient education.

Self-management interventions rely on patients

to take responsibility for their own care and are considered an integral part of medical nutrition therapy
298

in chronic disease management

The results of this study are similar to a study following dietary intervention for sodium-restriction in
heart failure patients.

In this investigation, despite men and women reporting similar perceived

barriers to compliance with the diet, women were shown to have greater knowledge and subsequently
299

better adherence to the dietary prescription

Participants in the intervention group required some level of behaviour change to meet the treatment
goals. Behavioural theories that have been related to dietary compliance in weight loss studies were
drawn on to assist in the explanation of these findings. Behavioural response to chronic illness can be
further examined by Health Belief Model theories

300

In the nutrition setting, the behaviour and

compliance with recommendations depends on the value that the patient places on improving their
nutritional status and the belief that receiving intervention will improve their nutritional status. In the
above example of the sodium-restricted diet, men and women identified similar barriers to the diet;
however it was females who actually complied best with the dietary prescription.

In the cohort

featured in this thesis, over twice the proportion of females had the primary responsibility of food

- 136 -

purchasing and preparation. Men were more likely to have a wife or caregiver and thus did not have
the added responsibility of shopping/cooking. Perhaps this fact may have contributed to males being
less responsive to the intervention.

Further investigation is necessary to identify whether being

responsible for the shopping and meal preparation is in fact a key factor to dietary adherence.

However, having the belief that treatment will be beneficial is not necessarily adequate to sustain
behaviour change.

According to social learning theory, behaviour change is a function of both

301

expectancy beliefs and the importance placed on the outcome

. Key concepts from this model

investigated in nutrition interventions are self-efficacy and locus of control. Self-efficacy, established
by Bandura (1986) is the self-belief that one is capable of achieving a desired outcome. In the case of
the pre-dialysis intervention, implementation of this theory would involve the perceived efficacy to
manage sufficient/appropriate dietary intake, despite issues with appetite and taste changes; and to
develop skills in food choice and preparation to support dietary changes.

Studies have indicated that in males, the locus of control is dependant on age, with older males relying
302

more heavily on chance and other factors, rather than themselves for change

. Whereas females

reported a strong orientation toward self-management of health-care needs and were more likely to
utilise information seeking and threat-minimisation strategies to instigate behaviour change

302

In a

recent study into compliance for haemodialysis patients, in males, self-efficacy and the internal health
locus of control were negatively correlated with all compliance measures. Whereas in females, all of
303

the correlations were positive, except for between self-efficacy and interdialytic weight gain

5.4

Summary of the impact of nutrition intervention in pre-dialysis

CKD
Structured nutrition counselling with telephone follow-up provides beneficial outcomes including
limiting the deterioration in nutritional status, improving dietary intake and improving quality of life in
patients with CKD in the stage prior to dialysis treatment. When considering the evidence, the cause
of this gender difference in response to intervention is largely unknown, with a need for this to be
considered in future intervention studies in pre-dialysis. However, overall, there is a clear benefit for

- 137 -

the provision of well-structured dietetic intervention to optimise nutritional status and improve
intermediate and patient-centred outcomes prior to dialysis treatment.

- 138 -

Chapter 6:

Quality of Life Results: Relationship with

nutrition status and response to nutrition intervention in

pre-dialysis CKD patients
6.0

Introduction

This chapter considers the relationship between nutritional status and quality of life in this pre-dialysis
sample at baseline and also evaluates the impact of providing nutrition intervention on quality of life
compared to standard care treatment. The same sample of patients described in 4-1 Characteristics
of study participants.

6.1

Quality of life and nutritional status

6.1.1 Comparison of pre-dialysis patients to dialysis patients by nutritional

status
Figure 6-1a and b display the mean values for KDQOL for well-nourished and malnourished patients
(n=53), compared with dialysis reference population established from the United States cohort of the
151

Dialysis Outcomes and Practice Patterns study (US-DOPPS) populaiton

The SF-36, generic

quality of life components, the well-nourished patients had similar mean QOL ratings to that of the
dialysis population. A trend for lower mean QOL rating in malnourished patients is clear for physical
and social functioning, and the role limitations due to physical and emotional problems, compared to
both the dialysis population and well nourished pre-dialysis patients. The sub-scales of effects and
burden of CKD had a trend for a lower rating in dialysis patients, compared with the pre-dialysis
figures from this study.

- 139 -

80
70

Mean score

60
50
40
30
20
10
0
Physical
functioning

Role
Physical

Bodily Pain

General
Vitality
Social
Role
Health
Function Emotional
Generic Quality of Life Components

Mental
Health

PHC

MHC

Figure 6-1a Mean score for general health (SF-36) sub-scales between the US DOPPS
151

population and the current study; US DOPPS

n= 2,885

Well-nourished n=43

Malnourished n=10
90
80

Mean score

70
60
50
40
30
20
10
0
Symptoms
of KD

Effects of
KD

Burden of
KD

Work Status

Cognitive
function

Social
interactions

Sleep

Social
support

Overall
Health

Satisfaction

Kidney-disease specific components

Figure 6-1b

Mean score for kidney disease specific sub-scales between the US DOPPS study
151

and the current study; US DOPPS

n= 2,885

Well-nourished n=43

; Malnourished n=10

PHC = Physical Health Component; MHC = Mental Health Component; KD = Kidney Disease; US-DOPPS = United States
Dialysis Outcomes and Practice Patterns cohort; SF = Short Form

6.1.2

Relationship between Quality of life sub-scales and nutritional status

Nutritional status, as determined by SGA was compared with ratings of the components of the KDQOL
tool (Table 6-1), with higher score equating to better quality of life.

In the majority of the QOL

components well-nourished patients had a higher mean rating than the malnourished patients and this
was particularly evident in the components of symptoms of kidney disease, physical functioning, social
functioning and role emotional (Table 6-1). Clinically significant differences (>10 point difference
between ratings of well-nourished and malnourished) are indicated in italics in Table 6-1. The PGSGA score demonstrated a significant negative correlation with each of the short-form (SF) 36

- 140 -

components, relating to physical and mental health and the ratings for symptoms and effects for
kidney disease and overall health from the kidney disease sub-scales.

Table 6-1

Kidney disease-specific and short-form 36 quality of life components, compared

with SGA rating and PG-SGA score (n=53)
Well-nourished
SGA A, n = 43
Mean SD

Malnourished P-value*
SGA B, n = 10
Mean SD

PG-SGA score
r**; p-value

Kidney-disease components
Symptoms of Kidney Disease
Effects of Kidney Disease
Burden of Kidney Disease
Work Status
Cognitive function
Social interactions
Sleep
Social support
Overall Health
Satisfaction
SF-36 Components
Physical functioning
Role Physical
Bodily Pain
General Health
Vitality
Social Function
Role Emotional
Mental Health
Physical Health Component
(PHC)
Mental Health Component
(MHC)

77.2 16.1
80.2 19.0
64.4 26.2
27.9 33.3
78.9 18.7
75.6 17.9
66.2 19.4
75.8 26.3
62.6 21.2
73.2 22.0

61.0 23.2
68.4 18.3
51.3 31.6
30.0 25.8
74.3 24.9
76.0 18.4
63.3 25.4
81.7 12.3
47.0 20.0
74.1 20.6

0.028
0.094
0.240
0.665
0.769
0.963
0.891
0.699
0.056
0.927

-0.480 p<0.001
-0.307 p=0.030
-0.162 p=0.247
-0.010 p=0.941
-0.251 p=0.072
-0.219 p=0.119
-0.179 p=0.200
0.025 p=0.863
-0.422 p=0.002
-0.073 p=0.615

43.4 27.6
32.6 38.8
66.8 25.9
41.5 18.1
41.6 23.0
71.8 28.5
62.8 39.3
70.8 20.9

20.5 22.7
12.5 31.7
63.5 32.7
33.5 13.6
28.0 25.8
46.3 32.8
26.7 40.9
59.6 23.2

0.015
0.112
0.774
0.331
0.087
0.020
0.018
0.138

-0.458 p=0.001
-0.375 p=0.006
-0.304 p=0.027
-0.313 p=0.022
-0.345 p=0.011
-0.405 p=0.003
-0.485 p<0.001
-0.410 p=0.002

35.9 10.2

29.9 7.9

0.107

-0.348 p=0.011

47.7 11.7

40.0 12.6

0.069

-0.388 p=0.004

SGA= Subjective Global Assessment; PG-SGA= Patient-Generated Subjective Global Assessment; SF=Short-Form
Italics = Clinically significant difference (>10 points); Bold = statistically significant
*determined by independent sample t-test; ** determined by Pearson correlation

6.2

Change in Quality of Life following randomised-controlled trial

of nutrition intervention
Table 6-1 summarises the quality of life ratings for the 10 kidney-disease, and 8 SF-36 subscales at
baseline and week 12 for each of the treatment groups, including an evaluation of the mean difference
in change between groups. There was a statistically significant difference in mean change for scores
of symptoms of kidney disease, cognitive function and vitality in favour of the intervention (Table 6-2).

- 141 -

There was a clear trend for a mean increase in ratings from the intervention group with a clinically
significant mean improvement in 10 of the 18 sub-scales from baseline to week 12, indicated in Table
6-2 in italics.

During the 12-week treatment period, most of the kidney disease-specific components of the KDQOL
had an increasing trend in the intervention group and decreasing trend in the standard care group
(Table 6-2 and Figure 6-2). From week 0 to week 12, there was a consistent increase the SF-36 subscales for the intervention group (except for body pain), whilst they remained relatively stable with the
standard care treatment.

- 142 -

82.032.67*
81.162.43
65.873.96
82.605.93
64.354.39
80.304.33

62.965.22

27.785.55

76.203.99*

75.803.17

65.003.64

77.804.95

58.524.19

74.074.74

Burden of Kidney Disease

Work Status

Cognitive function

Quality of social interactions

Sleep

Social support

Overall Health

Satisfaction

68.485.11

Bodily Pain

45.212.40

MHC

48.582.18

35.972.47

23.085.71

28.856.89

47.292.36

34.342.01

70.313.98

61.538.00

65.386.21

40.584.99

39.103.97

63.9227.75

29.817.07

33.646.14

72.453.72

60.774.29

76.004.62

66.254.35

75.64.53

48.892.70

33.132.10

74.724.71

63.899.18

65.386.91

38.404.29

42.624.30

62.856.44

30.658.13

34.776.62

76.003.98

59.104.39

73.336.53

63.084.12

74.674.09

72.534.90**

54.816.56

60.825.55
80.03.83**

75.504.58

74.503.17

76.694.43

75.353.46

Week 12
Mean SE

0.04 (-6.38 to 6.46)

3.65 (-1.20 to 8.51)

-1.21 (-11.86 to 9.44)

5.65 (-9.10 to 20.40)

2.29 (-25.12 to 29.72)

12.03 (4.56 to 19.50)

-0.81 (-10.10 to 8.48)

1.38 (-12.50 to 15.26)

8.55 (-10.92 to 28.01)

9.15 (-4.65 to 22.94)

3.41 (-6.25 to 13.08)

6.47 (-4.84 to 17.80)

10.78 (-7.47 to 29.03)

8.13 (-0.89 to 17.15)

8.85 (-0.26 to 17.96)

14.55 (5.39 to 23.71)

9.76 (-4.37 to 24.00)

9.04 (-6.95 to 25.04)

2.08 (-6.68 to 10.84)

7.08 (0.11 to 14.05)

Change
(Mean (95%CI))

Mean difference a

0.990

0.136

0.820

0.444

0.867

0.002

0.861

0.843

0.381

0.188

0.479

0.256

0.240

0.076

0.057

0.003

0.171

0.260

0.635

0.047

P-value

- 143 -

change in standard care minus intervention, adjusted for mean baseline score; Italics = Clinically significant difference (>5 points); Bold = statistically significant (*P <0.05 ** p <0.01from paired ttest (week 0 and week 12 values)) PHC = Physical Health Component; MHC = Mental Health Component

35.131.92

PHC

71.133.64

60.6110.44

50.628.22

67.114.51

Social Function

Mental Health

71.575.72

68.525.83

Vitality

Role Emotional

40.173.47
46.743.77**

40.812.84

37.594.37**

General Health

62.206.30

51.305.35
42.059.65*

45.435.86

27.787.82*

Role Physical

61.085.89

Physical functioning

SF-36 Components

30.706.80

79.503.23

Effects of Kidney Disease

81.523.15

73.013.76

Symptoms of Kidney Disease

80.093.53

Week 0
Mean SE

Week 0
Mean SE

Week 12
Mean SE

Standard Care

Intervention

Quality of life components for Week 0 and 12 and mean difference in change (standard care from intervention)

Kidney-disease specific

Table 6-2

Mean change in scores (Week 12 - baseline)

-4.0

Cog. Function
Difference between Intervention and Standard Care*
Mean Changes

-0.8

2.3

-1.2

3.7

0.1

. Mean difference in change (standard care from intervention) greater increase in intervention

- 144 -

KD = Kidney Disease; Cog = Cognitive Functioning; Soc = Social, Phys = Physical, PHC = Physical Health Component; MHC = Mental Health Component

standard care treatment (n=22)

; or
greater in standard care

Change in kidney disease specific and general (SF-36) quality of life subscales as a result of receiving structured intervention (n=23)

-4.0

1.4

General Health

-2.0

0.0

2.0

4.0

5.7

Mental Health

-2.0

Satisfaction
3.4

6.0

8.6

Bodily Pain

0.0

Symptom of KD
8.0

9.2

12.0

Soc. Function

2.0

Effects of KD

2.1

Overall Health
6.5

10.0

Phys. Function

4.0

Sleep
8.1

12.0

14.0

Role Emotional

6.0

Burden of KD

7.1

Soc interaction
8.9

Mean change in scores (Week 12 - baseline)

16.0

Role Physical

8.0

9.8

Work Status

9.0

10.8

Soc support

10.0

14.6

-10

-5

10

15

Vitality

12.0

14.0

16.0

-10

-5

10

15

General (SF-36) Domains

PHC

Figure 6-2

Difference between Intervention and Standard care*

Mean Changes

Kidney disease specific domains

MHC

6.2.1 Gender difference in the change in quality of life in response to nutrition

intervention
As with the results from Chapter 4, the mean difference in quality of life change over the intervention
was split to establish descriptive difference by gender. When evaluating these results, a similar
pattern emerged to the results presented previously, with a greater magnitude of change evident in
females, favouring the intervention group (indicated by a positive mean difference in change, Table 63). A clinically significant difference in change (> 5 points) was evident in majority of the components
of quality of life in females. Social support and body pain sub-scales had opposing results by gender,
of a magnitude of clinical significance.

Table 6-3 Mean difference in change (week 12 week 0; (mean (95%CI))) in quality of life subscales by treatment group split for gender for week 0 and week 12
Mean difference
a
whole sample
Change
(Mean (95%CI))

Males
n=29
Change
(Mean (95%CI))

Females
n=17
Change
(Mean (95%CI))

Symptoms of Kidney Disease

7.1 (0.1 to 14.1)

0.16 (-8.3 to 8.6)

18.1 (4.2 to 31.9)

Effects of Kidney Disease

2.1 (-6.7 to 10.8)

4.3 (-7.5 to 16.2)

-1.7 (-22.1 to 18.7)

Burden of Kidney Disease

9.0 (-7.0 to 25.0)

6.0 (-16.6 to 28.5)

20.1 (-9.6 to 49.8)

Work Status

9.8 (-4.4 to 24.0)

14.2 (-2.9 to 31.3)

5.6 (-27.2 to 38.3)

14.6 (5.4 to 23.7)

9.7 (-1.5 to 20.8)

25.6 (6.9 to 43.1)

Quality of social interactions

8.9 (-0.3 to 18.0)

4.1 (-6.3 to 14.4)

17.8 (-7.6 to 43.1)

Sleep

8.1 (-0.9 to 17.2)

7.4 (5.3 to 20.2)

14.4 (-2.5 to 31.5)

Social support

10.8 (-7.5 to 29.0)

-9.2 (-32.2 to 13.7)

46.8 (4.5 to 89.0)

Overall Health

6.5 (-4.8 to 17.8)

6.1 (-12.0 to 24.3)

13.3 (-2.9 to 29.6)

Satisfaction

3.4 (-6.3 to 13.1)

-3.0 (-17.2 to 11.2)

7.2 (-13.2 to 27.6)

9.2 (-4.7 to 23.0)

-2.3 (-19.7 to 15.0)

14.2 (-13.2 to 41.5)

Role Physical

8.6 (-10.9 to 28.0)

4.3 (-21.9 to 30.5)

15.3 (-17.1 to 48.4)

Bodily Pain

1.4 (-12.5 to 15.3)

-7.9 (-25.3 to 9.4)

15.1 (-11.5 to 41.7)

General Health

-0.8 (-10.1 to 8.5)

-0.6 (-14.4 to 13.1)

1.9 (-11.7 to 15.4)

Kidney-disease specific

Cognitive function

SF-36 Components
Physical functioning

Vitality

12.0 (4.6 to 19.5)

5.3 (-5.2 to 15.8)

27.8 (11.4 to 44.1)

Social Function

2.3 (-25.1 to 29.7)

-3.4 (-16.8 to 10.0)

19.3 (-20.7 to 59.2)

5.7 (-9.1 to 20.4)

13.3 (-21.8 to 48.5)

0.0 (-63.4 to 63.4)

-1.2 (-11.9 to 9.4)

-1.9 (-13.4 to 9.7)

4.9 (-22.3 to 32.1)

PHC

3.7 (-1.2 to 8.5)

2.5 (-4.2 to 9.3)

5.4 (-3.7 to 14.5)

MHC

0.0 (-6.4 to 6.5)

-2.1 (-9.7 to 5.7)

6.8 (-10.2 to 37.9)

Role Emotional
Mental Health

aadjusted for mean baseline score; Italics = Clinically significant difference (>5 points); Bold = statistically significant (95% CI does not cross 0)

- 145 -

Chapter 7: Quality of Life Discussion: relationship to

nutritional status and impact of nutrition intervention in
pre-dialysis CKD patients
This chapter provides a discussion based on the results of Chapter 6. This discussion aims to provide
a patient-centred evaluation of the impact of poor nutritional status and the effect of providing a
nutrition intervention aimed at optimising nutritional status, on various quality of life (QOL)
components, in a sample of pre-dialysis chronic kidney disease (CKD) patients.

7.1

Quality of life in pre-dialysis compared to dialysis populations

To date, previous studies in the CKD literature investigating quality of life have been focused on
haemodialysis (HD) patients. Investigations reporting the significant and independent relationship
between QOL with clinical outcome (morbidity/mortality) and nutrition status are contained to HD
populations. . In the pre-dialysis phase, a decline in GFR and an increase in uraemic symptoms
(including fatigue, weakness, anorexia and muscular cramps) is associated with a reduced functional
39

status and quality of life . A number of studies have supported a direct correlation between GFR and
functional status, with a lower physical function and functional capacity with declining GFR

304-306

In the overall pre-dialysis study sample the global physical component score was significantly lower
than general population norms described in the literature (PCS (meanSD) 34.710.0 versus
45.911.3)

, however the mental health component score (MCS 46.2 12.16) of the pre-dialysis

307

patients was similar to the general population (MCS 51.19.7)

307

. This same pattern is in other CKD

populations, compared with healthy age- and sex-matched controls, the physical health measured by
151, 308-310

PCS is lower, and more variable than mental and emotional health as measured by MCS

The ratings from the kidney-disease specific sub-scales indicate a similar reported level QOL between
the pre-dialysis patient sample and reference dialysis data from the US-DOPPS population (Figure 61). However, burden of CKD is greater for the dialysis reference population compared with both wellnourished and malnourished pre-dialysis patients. This significantly greater burden of kidney disease

- 146 -

on the quality of life of dialysis patients represents a greater perception of frustration and interference
of kidney disease on ones life in the dialysis patients, presumably attributable to the nature of dialysis
treatment. There is a similar, if not greater reported burden of kidney disease from other dialysis
populations, including peritoneal dialysis

7.2

148, 311

Quality of life and nutritional status in pre-dialysis CKD

patients
No study could be sourced that compared all KDQOL subscales with SGA ratings in CKD, or any
nutritional markers against QOL in pre-dialysis CKD. In this investigation, 4 of the 18 QOL sub-scales
were statistically significantly lower in the malnourished pre-dialysis patients, compared with the wellnourished patients (Table 6-1). A further 6 subscales had a mean score greater than 10 points lower
in the malnourished patients, considered to be a clinically significant difference (highlighted in italics in
Table 6-1).

Change in quality of life following the randomised controlled trial (RCT) of 12-week

nutrition intervention improvements in QOL reported by the intervention group spanned components in
the global physical and mental health dimensions (of SF-36) as well as the components of coping with
the impact of kidney disease.

Significant regression in QOL was noted only for disease-specific

components with standard care treatment. The mean difference in change was statistically significant
for 3 sub-scales and a further 9 had clinically significant difference in change (i.e. difference >5
points

312

) favouring the intervention (Table 6-2).

The impact of poor nutritional status in dialysis patients (as measured by poor appetite, low albumin,
creatinine and nPCR) has been most significantly related to reduced physical health components, and
less so for mental health components

158

. In another cross-sectional study of Australian haemodialysis

patients (n=53), severe malnutrition (SGA C) was associated with a significantly lower index of general
well-being, life satisfaction and functional status (measure by Karnofsky index)

166

. In this same study,

using the quality of life index malnutrition (SGA B and C) was significantly lower in the psychological
166

and spiritual well-being domain

- 147 -

7.2.1 Kidney-disease specific QOL, relationship to nutritional status and

response to nutritional intervention
Symptoms, effects and burden of kidney disease were statistically and/or clinically lower in
malnourished patients at baseline, indicating poorer QOL (Figure 6-1 and Table 6-1). There was a
statistically significant, lower rating for the symptoms/problems relating to kidney disease thus
indicating malnourished participants rated a greater extent of difficulty due to soreness in muscles,
pain, headaches, cramps, bruising, itchy skin, shortness of breath, dizziness, lack of appetite,
4

excessive thirst, numbness in hands or feet, trouble with memory, blurred vision and nausea . As
evident in Table 6-1, this variable also had a significant correlation with PG-SGA score, indicating the
sensitivity of this nutrition assessment tool to monitor changes in kidney disease symptoms directly
impacting on nutritional status.

Following the RCT, rating of the symptom/problem sub-scale relating to kidney disease was
significantly reduced in intervention participants over 12 weeks (73.03.8 to 80.1 3.5). The standard
care group did not change in this component resulting in a significant mean difference in change
between the groups (7.1 (0.1 to 14.05) p=0.047, Table 5-14). Similar results have been reported with
improvements in the symptoms and burden of kidney disease 6-months following parathyroidectomy
in HD patients (both increased 10 points)

313

. It is possible these improvements were afforded by

concurrent improvement of managing appetite and uraemic symptoms, as noted in the exploratory
outcomes (Section 4.4, Table 4-14).

The effects of kidney disease, where participants rated the level of burden associated with restrictions
on fluid and dietary intake, impact on carrying out family responsibilities, work, and personal
appearance had greater impact on the malnourished patients, identified by a clinically significant
difference between well-nourished and malnourished patients at baseline.

These issues maybe

related to uraemia, thereby indicative of the development of uraemic malnutrition. The burden of
kidney disease, where participant rated their feelings of frustration regarding the interference of kidney
disease on their life and burden on others, clinically lower in the malnourished group.

- 148 -

The rating of nutritional status by SGA appears sensitive to distinguishing a significant difference in
the concern and burden caused by patients symptoms and feelings of well-being. In addition, PGSGA score had a significant correlation also with the sub-scale ratings for the effect of kidney disease
on daily life. However, neither of these components changed significantly as a result of intervention or
standard care treatment over 12 weeks.

The cognitive function sub-scale was not related to nutritional status at baseline, however, had the
greatest magnitude of change between groups in the RCT, significantly improving in the intervention
and decreasing in the standard care group resulting in a significant mean difference in change
between the two groups (14.6 (5.4 to 23.7) p=0.003, Table 6-2). The items contained in the cognitive
314

function sub-scale are indicators of impaired thinking derived from the sickness impact profile

and

have been validated against the global objective measure of cognitive function, the Modified MiniMental State Exam

315

Advanced stages of CKD are associated with increased risk of cognitive

impairment and more than twice the prevalence in the age-matched population, even after adjusting
for potential clinical mediators to cognitive impairment including anaemia, inflammation and
316

hyperlipidemia

A significant trend in decreasing cognitive function scale demonstrated in the standard care group is
perhaps reflective of a gradual decline that is associated with progressive renal function loss. A
reduced cognitive function score from KDQOL has previously been related to loss of kidney function in
a number of cross-sectional studies

152, 315, 316

. Also, of note is that the standard care group had an

increasing proportion of participants with indications of anaemia (transferrin saturation <20% from 38%
to 55% at Week 12), which is closely linked with cognitive functioning

317

. Therefore, it is possible the

cognitive function decline in this standard care group potentially reflects the metabolic changes and
psychological stressors, a result of this stage of the disease process.

A mechanism for the significant improvement in cognitive function evident in the intervention group is
potentially control of metabolic changes and may also be related to depressive symptoms.
Depression has been associated with lower performance on cognitive function testing, and a higher

- 149 -

318

risk of dementia

. In the elderly, a depressed state is closely linked with symptoms of anxiety and

appear to increase in prevalence with increasing co-morbidity load, higher physical and emotional
dependency319.

It is also associated with decrements in sub-scales measuring quality of social

interactions, social support and ratings of general health, which also significantly improved in the
intervention group and remained relatively stable for ratings in standard care group from baseline to
week 12. Griva et al (2006) recently demonstrated a reversibility of the cognition problems evidenced
in dialysis patients once patients underwent a kidney transplant. This demonstrates improvements in
cognitive function may also be due to improvements in symptoms and coping with the burden kidney
disease, considering cognitive impairment evident in the dialysis patients (attention, concentration,
320

psychomotor ability and memory) which was ameliorated with transplantation

7.2.2 Generic Health Related QOL, relationship to nutritional status and

response to nutritional intervention
Of the SF-36 subscales, physical and social functioning, vitality, mental health and role limitations due
to both physical and emotional problems (Role Physical and Role Emotional) were rated statistically
and/or clinically lower in the malnourished pre-dialysis patients (Table 6-1). In addition, each of the
eight SF-36 components had a moderate, yet significant negative relationship to the PG-SGA score.
Therefore, indicating a direct relationship between scored nutrition assessment and generic quality of
life components. This further demonstrates the strengths of PG-SGA as a sensitive assessment tool
for patient-centred evaluation.
Of the sub-scales under the physical health dimension, malnourished patients rated significantly lower
in physical functioning, therefore indicating decreased capacity to perform the physical requirements
of life, such as attending to personal needs, walking and flexibility. In addition, the problems with daily
activities as a result of physical health (Role Physical) were greater in malnourished patients, rating an
average of 20 points less than the well nourished patients. In a cross-sectional study, increased
dietary energy intake in haemodialysis patients was associated with higher physical components
scores from the SF-36

158

Overall, there were significant improvements in a number of the physical health scales as a result of
patients receiving nutritional intervention.

Similar improvements have been noted in studies

- 150 -

normalising haemotocrit and haemoglobin levels, for measures of physical function subscales of SF321

36

322

in pre-dialysis and functional status by Karnofsky Performance Index and SIP

in

haemodialysis. Resistance training also improved physical functioning sub-scales and physical health
components in HD patients

323

An indirect improvement in physical functioning due to improved

nutrition and encouraged incidental activity, reduced the impact of symptoms related to CKD.

Role limitation due to physical functioning (Role Physical, Table 6-2) improved a mean of 14.3 points
as a result of the intervention. This is of significant clinical value, considering the support provided in
the intervention enabled participants on average to rate a lower degree of problems encountered as a
result of their physical health. In addition, role limitation due to emotional problems also improved by
10 points in the intervention group. This links with the results of improvement in the role limitations
and symptoms subscales, that it appears partaking in the intervention assisted participants to
implement strategies to better deal with their symptoms. It is possible that strategies implemented
during the intervention period assisted participants to better manage issues resulting from emotional
or physical limitations.

Vitality, indicated substantially greater feelings of lethargy in the patients identified as malnourished at
baseline.

The vitality sub-scale, sensitive to both physical and mental health outcomes

324

, also

significantly improved in participants who received the intervention resulting in a significant mean
difference in change. This indicates the intervention group improved ratings for feelings of increased
energy levels as opposed to standard care.

More recent investigations into this scale and the

Nottingham Health Profile has confirmed it is strongly linked to the emotional reactions, suggesting it
is a concept closely allied to psychological scales and motivation, further adding weight to the benefit
of the intervention

325

In terms of scales under the mental health dimension, the consistent impact of poor nutritional status
on QOL at baseline continued. Social functioning, a scale evaluating the impact of the amount of time
that physical or emotional problems interfered with family, friends and other social interactions rated
as significantly more burdensome in the malnourished group. Problems with daily activities as a result
of emotional problems were similarly related to nutritional status, as measured by the Role Emotional

- 151 -

sub-scale.

This was further supported by a lower rating in the mental health sub-scale, indicating a

greater burden of feelings of nervousness and depression.

The intervention group demonstrated clinically significant improvements in the quality of social
interactions, social support and role limitations due to emotional problems, with no significant change
in the standard care group. Recently, an investigation into KDQOL in dialysis patients indicated that
psychosocial variables, trait anxiety and depressive symptoms substantially affected the way in which
309

patients evaluated their HRQOL

. This may explain the improvement noted in the ratings of the

intervention patients with psychosocial components reflected in the subscales of vitality, cognitive
functioning (as above) and socialisation. It is known that CKD patients under high physiological stress
326

in general report poorer health related and kidney-disease specific quality of life

. It is therefore

important to consider the quality of psychological care provided throughout the pre-dialysis phase,
where anxiety related to prognosis and treatment options may be at their peak and multidisciplinary
individual support systems may have the most benefit.

In the literature, the social functioning subscale has been strongly associated with other measures of
309

mental health including anxiety and depression

. Social functioning itself is multidimensional and

325

may be involved in the relationship of health impairment

. Indicated by low score on mental health

sub-scale amongst HD patients is associated with increased risk of mortality. In a study of 137
commencing haemodialysis patients investigating short-term mortality, for each increase in the Mental
Component Summary (MCS) score, the adjusted RR of death diminished by 4% (RR = 0.96, 95% CI
0.940.99; P = 0.006)

327

Therefore is strong clinical relevance of this relationship between

140, 141

malnutrition in pre-dialysis and poor mental state

The synergistic improvement between nutrition and quality of life in CKD patients evident in this study
was been recently reported in other CKD studies.
coaching in pre-dialysis and dialysis cohorts
330

supplementation

272

, resistance-training in HD

and use of icodextrin in PD

markers of nutrition status.

Provision of daily hemodialysis

331

323

and PD

328

329

, exercise

, L-carnitine

all result in improvements in quality of life and

To date, there have been no specific nutrition support/counselling

- 152 -

interventions reporting quality of life change in CKD. However, the significant improvements in quality
of life as a result of direct nutrition support intervention as evident in this study are supported by
studies in other clinical areas. Particularly in the tertiary treatment environments for the areas of
oncology, surgery and HIV, where treatment goals are to optimise nutrition status by limiting the
progression of malnutrition and muscle wasting, the improvements recorded in quality of life were
directly reflected by improvements in nutritional status, as measured by estimates of FFM, clinical
332-336

tools and/or dietary intake

. In a sample of oncology patients receiving radiotherapy, the greatest

magnitude of improvement in intake and QOL as a result of nutrition intervention was in those
identified as higher risk (head/neck and gastrointestinal cancer)

336

weight-stabilisation over an 8 week period with provision of

nutrition support in weight losing

Using the EORTC QLQ-C30

pancreatic cancer patients resulted in significant improvements in QOL (increased 8 points on 100
point scale)

337

Providing regular

nutrition support in

333, 336

radiotherapy

338

, treatment for congestive heart failure

patients

332

receiving chemotherapy

have also reported this benefit.

This self-management focused nutrition intervention produced clinically important improvement as a

result of participants receiving the intervention (Table 6-2). The kidney-disease specific components
had the largest magnitude of difference in change between the treatment groups and significantly
favoured intervention in many components. Change in SF-36 components was less consistent, with
some components changing to a similar degree in both treatment and standard care groups. The
issue of statistical power will be discussed further in 6.3.1 Sample considerations and statistical
power.

The gender difference in response to the intervention, detailed in section 6.3, indicated the

QOL results mirrored those of the nutritional status outcome presented in Chapters 4 and 5. That is,
providing a self-management focused intervention resulted in improvements, and providing standard
care treatment resulted in maintenance and/or decrements in quality of life, to a greater magnitude in
the female gender than males.

7.3

Sample considerations and statistical power

As a five-point difference in the scales of quality of life measured are considered to define differences
312, 324

which are clinically and socially relevant

, this study was under-powered to detect statistically this

difference in a number of sub-scales. The sample size required to detect a statistically significant

- 153 -

difference from a difference considered clinically significant is dependant on the scale in question.
The sample size needed per group to detect a 5 to 10 point difference in mean change is anything
from 34 (10 point difference in change in the mental health sub-scale) to 468 (required for significance
from a 5 point difference in the role-physical scale)

324

. Approximately twice as many subjects are

required to detect a given difference in the relatively coarse role limitations due to physical and
emotional problems as they are based on only 4 components compared with 20 or more with the least
coarse scales, for example, physical functioning

324

. Therefore, this study was underpowered to detect

a statistical difference in many of the subscales of quality of life.

7.4

Summary of quality of life in relation to nutritional status and

the impact of nutrition intervention

In the past, studies investigating nutrition support in pre-dialysis CKD has generally focused on
nutrition perception and the monitoring of compliance with this prescription.

Therefore, previous

studies in CKD have generally overlooked patient-focused outcomes such as quality of life, including
patient satisfaction.

This study confirms pervious observations that malnutrition is related to poorer quality of life. The
associations between nutritional status and quality of life indicated in this investigation, particularly in
the mental health dimension indicate the need for more focussed interventions in an effort to provide
targeted therapy with consideration to psychological limitations which may enhance treatment
outcomes

In addition, providing nutrition intervention resulted in improvement or maintenance of quality of life.

Perhaps due to the supportive nature of the intervention, improvement in nutritional status and
metabolic markers in the intervention group is reflected in the improvement in QOL sub-scales. It is
likely that improvements in QOL were facilitated by providing an individualised, patient-centred
supportive intervention, assisting patients with coping with the symptoms of their reduced kidney
function and reducing anxiety associated with this stage of their disease.

Patients feeling more

positive and hopeful in dealing with their kidney disease, and therefore less depressed and anxious,

- 154 -

provided improvements in most of the kidney disease sub-scales of QOL, particularly cognitive
functioning.

- 155 -

Chapter 8

8.0

Conclusions and recommendations

Introduction

The studies presented in this thesis provide insight into each stage of the cascade of medical nutrition
220

therapy

for the management of pre-dialysis chronic kidney disease (CKD) patients. This chapter

provides a synthesis of the findings stated across the previous chapters in relation to the study
hypotheses.
detailed.

Firstly, evidence gained to support or refute each of the study hypotheses will be

Secondly, a discussion around the study strengths and considerations of the study

limitations, including issues relating to the sample population and power. Finally, recommendations to
direct clinical practice in context with the Framework for Evidence-Based Practice in CKD, and
recommendations for future research will be discussed.

This study confirms pervious observations that malnutrition is related to poorer quality of life. In
addition, providing nutrition intervention resulted in improvement or maintenance of quality of life.

8.1

Addressing study aims and hypotheses

In the following section, the study aims and hypotheses introduced in Section 2.1 are discussed with
the results to support or refute each hypothesis.

H1:

Patients assessed as malnourished by SGA (B moderately malnourished or C severely

malnourished) will have lower body cell mass (indexed for height, BCM-I) than those assessed as well
nourished (SGA A).

BCM and BCM-I was significantly lower in the malnourished, or SGA B group, compared with SGA A.
There was no SGA C or severely malnourished patients. Therefore, hypotheses H1 is accepted; there
is a significant relationship for BCM and SGA in this sample of pre-dialysis CKD patients.

H2:

Body cell mass (indexed for height) provides an accurate assessment of nutritional status

compared with SGA

- 156 -

BCM-I provided an acceptable assessment of nutritional status by discriminating between wellnourished and malnourished pre-dialysis CKD patients. This was demonstrated by producing and
area under the receiver operating characteristic curve of 76.2% (Section 3.11). Therefore, hypotheses
H2 is accepted;

H3:

There is a relationship between BCM (indexed for height) and a) 7-point SGA categories, b)

Patient-Generated SGA scores, and c) Malnutrition-Inflammation Scores

Each of the modified SGA tools had a statistically significant relationship with cross-sectional BCM-I.
Therefore hypotheses H2 is accepted. However, when assessed for their diagnostic ability against
BCM-I cut-off for malnutrition, all tools displayed only fair discrimination. The PG-SGA was the only
tool to demonstrate significant discrimination between malnourished and well-nourished based on
BCM-I cut-off.

Primary outcome: Nutrition Status

H4:

Patients who receive nutrition intervention will have maintained or improved nutritional status

as assessed by change in BCM, than those receiving standard care.

The results in section 4.2.1 demonstrate that BCM was maintained in the intervention group over the
treatment period (0.6% (-2.9 to 4.1%)), compared with those receiving standard care who had a
significant deterioration in BCM (-3.3% (-6.8 to 0.4%)). The mean difference in change between both
groups equates to 3.9% (-1.0 to 8.7%) p=0.114, or approximately 2.4kg of fat-free mass. This is
considered to be a clinically significant magnitude of difference. When the change in BCM was
categorised into improved ( 1%), maintained (1%) or decreased ( 1%), relative to decrease in BCM,
the odds of participants from the intervention group to increase in BCM was 5.2 times (95% CI 1.1 to
25.1 (p=0.04)), compared to those receiving standard care.

Hypothesis H4 is accepted on the basis

of the intervention effectively attenuating a clinically significant deterioration in BCM, as established by

the reduction in the standard care group.

- 157 -

When the data is split for gender, change in BCM for females had a much greater magnitude, resulting
in mean difference in change of 9.9% (0.1 to 19.6%), with improvement in the intervention (increase in
2.9 8.7%) and reduction in the standard care (6.9 8.7%) for females only. In both groups males
had a similar, small change in BCM (mean increase of 1.5 1.9% for intervention; 1.4 2.6% for
control) resulting in a negligible mean difference in change (0.1% (-6.9 to 6.7)). Further research, with
a sample size that is adequately powered, is required to establish if this is a true difference.

H5:

A greater proportion of patients receiving nutrition intervention will have maintained or

improved their nutritional status as assessed by change in SGA, compared to standard care.

All of the participants in the intervention group either maintained (19/24) or improved (5/24) nutritional
status according to SGA over the 12-week treatment period, resulting in no malnutrition at week 12. In
contrast, 4 participants in the standard care group regressed an SGA category (SGA A to B, n=3; SGA
B to C, n=1) and 2 maintained SGA B (malnourished) status, resulting in a rise in the proportion
malnourished from 11% at week 0, to 22% for the standard care group at week 12. This difference in
2

change in SGA between the 2 groups was statistically significant X (2) = 12.76 (P=0.002). Therefore,
H5 is accepted, a significantly greater proportion of the intervention group improved or maintained
SGA compared with the standard care group.

Intermediate outcome: Dietary intake

H6:

A greater proportion of patients who receive the structured nutrition intervention will have

maintained or improved appropriate (0.75g/kg-1.0 g/kg) protein and energy intake (at least 125
kJ/kg)

46

compared to standard care.

To optimise nutritional status in CKD and prevent malnutrition, a protein intake no less than
0/75g/kg/day is recommended. Over the 12-week treatment period, the decrease in mean protein
intake in the control group alone was statistically significant (1.19 0.4 to 1.06 0.3 g/kg p=0.02)
however, overall the mean difference in the change in protein intake between the treatment groups

- 158 -

was not (0.04g/kg (-0.7 to 0.2 g/kg), p = 0.46). The proportion of subjects who were meeting the
guideline for protein intake was maintained in the intervention group (42 to 46%), with a relative
decrease in those meeting the guideline intake (32% to 25%). No subjects in the intervention group
were consuming under 0.75g/kg protein at week 12, however, in the standard care group, the
proportion of subjects below 0.75g/kg/d increased from 4% to 17%.

Energy intake significantly increased over the treatment period for the intervention group from 102.6
kJ/kg to 114.5 kJ/kg (p=0.002) and decreased 109.3 kJ/kg to 102.7 kJ/kg (p=0.06) in standard care
participants, resulting in a significant mean difference in change for energy intake between the
treatment groups 17.7 kJ/kg (8.2 to 27.2 kJ/kg) (p<0.001). A greater proportion of intervention group
were consuming the recommended energy intake of >125kJ/kg following intervention (17% to 29%),
however, the standard group decreased the proportion with adequate intake (40% to 21%) and
increased in the proportion consuming inadequate kJ (<100kJ/kg; 36% to 46%). Overall, the evidence
indicates an improvement in energy intake and greater proportion meeting energy and protein
guidelines at week 12 for those receiving the intervention and provides sufficient evidence to accept
H6.

Secondary outcome: Quality of Life

H7:

Patients who receive structured nutrition intervention will have maintained or improved rating

for health-related and kidney disease specific quality of life measured by KDQOL-SF

TM

than those

receiving standard care.

Throughout the treatment period, there was a clear trend for a mean increase in QOL ratings from the
intervention group with a significant mean improvement in 10 of the 18 sub-scales from baseline to
week 12. Benefits of the intervention was reflected in a significant improvement in cognitive function
(meanSE, 76.24.0 up to 82.02.0 p<0.05) and vitality (37.64.4 up to 46.73.8 p<0.01) and a
clinically significant improvement in the symptoms of kidney disease (73.03.8 up to 80.13.5), quality
of social interactions (75.83.2 up to 81.22.43), social support (77.84.9 up to 82.65.9), reported
limitations due to problems with physical (27.87.8 up to 42.19.6 p=0.058), emotional health
(50.68.2 to 60.610.4) and overall health (58.54.2 up to 64.44.4). For participants receiving the
standard care treatment, a statistically significant reduction in cognitive functioning (80.03.8 down to

- 159 -

72.54.9 p<0.05), in addition to a clinically significant reduction in the ability to cope with the burden of
kidney disease (60.85.5 down to 54.86.56) and work status (28.96.9 down to 23.15.7) was
evident. Patient satisfaction increased in both intervention (74.14.7 up to 80.34.3) and standard
care groups (72.453.72 to 76.003.98, p=0.06). Therefore H7 is accepted for the 10 of 18 subscales, participants receiving nutrition intervention had a greater improvement in QOL than those
receiving standard care.

8.2

Study strengths and limitations

8.2.1 Phase I
To the authors knowledge, this is the first study to compare BCM, measured by total body potassium
(TBK) and SGA in pre-dialysis CKD.

This investigation established that traditional biochemical

measures of protein status (albumin) had no relationship to nutritional status by SGA. This reflects the
interaction of albumin in the acute phase response and other mechanisms which affect albumin status
with CKD.

In addition to this, using a cut-off of BCM-I to determine discriminatory ability of a range of alternate
subjective assessment to measure nutrition status were investigated. This is considered to be the first
step in providing a sufficient evidence base in the validation of such tools in pre-dialysis CKD. One of
the advantages of the modified SGA-based tools is the potential to be more sensitive to small changes
in nutritional status. Phase II of this investigation aimed to assess the ability of the scored tools to
accurately track changes in nutrition status.

This is of particular importance in chronic wasting

conditions when ability to detect small, yet clinically significant changes in nutritional status is
important. PG-SGA is recommended as the assessment tool for pre-dialysis CKD as it provides the
global SGA rating, is simple to use and had the best discrimination for nutritional status based on the
BCM-I cut-off.

8.2.2 Phase II
Overall this study determined that a structured nutrition intervention program results in positive
outcomes for pre-dialysis chronic kidney disease (CKD) patients. This was achieved through a
randomised-controlled trial, providing a structured, individualised program using self-management

- 160 -

strategies with regular telephone counselling to monitor progress, and comparing this to standard care
(provision of written information only). This study aimed at improving intake rather than restriction of
particular nutrients. This is in contrast to previous published studies in pre-dialysis, where nutrition
advice is focused on restricting protein- and/or phosphate-containing foods. In contrast, based on the
conflicting evidence for benefits of restrictive dietary strategies in pre-dialysis CKD, this intervention
was promoting a moderate protein intake (0.8-1.0g/kg), it also provided strategies to ensure sufficient
energy intake (>125 kJ/kg).

These results show that in CKD patients in pre-dialysis, prone to

development of malnutrition and reduced quality of life, supplying individualised, structured nutrition
care during pre-dialysis medical nutrition therapy improves nutritional intake, limits deterioration in
nutritional parameters and improves quality of life.

Malnutrition has significant prevalence and also major implications on initiation of dialysis.
Development of malnutrition is multi-factorial and may be due to mechanisms of pro-inflammatory
cytokines and metabolic acidosis rather than a lack of dietary intake. Although this study did not set
out to determine aetiology, it demonstrated that provision of strategies to improve intake and maintain
weight resulted in improved nutritional status and quality of life even in the presence of mild levels of
inflammation, acidosis and anaemia, which is typical of this stage of CKD. Therefore, regardless of
the mechanism or aetiology for the development of poor nutritional status, strategies to improve
dietary intake are useful in treatment and prevention of malnutrition longer-term.

Absence of

intervention resulted in lower dietary intake and significant decrease in nutritional status and body
protein (BCM), as demonstrated in the standard care group, particularly in female participants.

8.2.2.1 Utilised a range of outcome measures

The methods employed in this study were of a high standard and included both objective and
subjective outcomes measures (providing clinically practical, as well as using gold-standard methods).
This study was designed based on the medical nutrition therapy framework to influence outcomes
(Figure 8-1). By providing nutrition intervention, this resulted in positive outcomes in terms of dietary
intake (intermediate), body composition, nutritional status (clinical) and quality of life (patient-centred).
Quantifying change in BCM measured by the gold-standard TBK as a result of a nutrition intervention
is a first for CKD. In addition, this study validated the PG-SGA for use in pre-dialysis. This quick, easy
and inexpensive tool was able to differentiate against BCM and identify change in nutritional status

- 161 -

over time. It is likely that each of the established improvements will further improve the cost outcomes
listed in Figure 8-1.

This figure is not available online.

Please consult the hardcopy thesis
available from the QUT Library

Figure 8-1

Cascade of events leading to evidence of effectiveness of Medical Nutrition

Therapy (Splett, 1996)

220

8.3.2.2 Utilised translational intervention methods

It is a well-documented fact that there is commonly a significant gap between evidence-based
recommendations and practice in chronic disease management

339

. One reason for this is the types of

clinical trials reported in guidelines have little relevance to clinicians, administrators and policymakers

340

. It is thought that this can be partly attributed to the lack of generalisability of studies due to

stringent inclusion criteria, use of non-clinically relevant outcome measures and the fact that
commonly the interventions are carried out by researchers and experts as opposed to being practicing
clinicians

339

Successful interventions stabilising BCM loss in CKD to date has been limited to exercise-based
treatments. Resistance training methods requiring specialty exercise physiologists on-hand to
supervise each individual session and adjust programs at each session, are questionably sustainable
in clinical practice due to the significant participant burden and cost. In addition, in the Castaneda
(2001) study, only a small recruitment rate was recorded (20%)

- 162 -

242

Although the importance of

exercise training for BCM maintenance cannot be underestimated, the findings of this nutrition
intervention study are comparable and potentially more sustainable in clinical practice.

Nutrition status outcomes in previous prospective observational protein restriction studies have utilised
BCM and FFM by DXA or BIA. These studies generally support that stabilisation of body tissue can
occur with protein restriction, if patients are closely monitored regularly (commonly every 1 to 4 weeks)
in an outpatient setting by both the nephrologist and dietitian. These studies selected only patients
with demonstrated compliance and the restrictive study design limits the applicability to clinical
practice, so therefore limits their generalisabiliy.

In contrast, the intervention featured in this thesis was aimed at providing a translational, sustainable
approach. Using the telephone to deliver the follow-up intervention was chosen as it minimised the
logistical barriers associated with this level of contact between the dietitian and the patient. Structured
programs, involving frequent contact with health professionals demonstrate improved knowledge and
behaviour change in weight management

341

. This study answered the call for well-designed RCTs to

278

access the impact of nutritional support in illness-related malnutrition

8.3.2.3 Gender difference in response to treatment

A gender difference in response to the delivery of intervention was established in this study. Females
were responsive to the intervention aimed at self-management, and regular follow-up and maintained,
or slightly improved, nutritional status. Whereas, those in the standard care group, responded poorly
to the provision of written information only. On the other hand, it seemed that males were able to
maintain their dietary intake and nutrition status and do equally regardless of their treatment allocation.
It is difficult to determine the direct mechanism for this, other than the potential sex-hormone
difference in muscle maintenance or response to individualised intervention may be perhaps related to
gender differences in health-belief system, self-efficacy, or locus of control. The study highlighted the
need to consider the issue of gender difference in future cross-sectional and intervention studies to
explore these issues further.

- 163 -

8.2.3 Considerations of the limitations of this study

Selection of subjects for the trial may have affected the ability to generalise these results to other predialysis clinics. Patients who were referred for pre-dialysis care late (considered to be within 3-6
months of dialysis commencement upon first visit to clinic) are more likely to have poorer outcomes
145, 342, 343

due to inadequate pre-dialysis care

More than a quarter of people starting dialysis in

Australia are not referred to a nephrologist for specialist care until less than three months before
344

initiating their treatment

. As the goal of this study was to complete the intervention prior to dialysis

commencement, these patients were excluded due to the impracticalities of providing a 12 week
treatment when a sufficient time-frame to dialysis could not be confirmed. However, when this study
population were compared with a reference population of incident patients referred to nephrology care
in Queensland with a GFR <30 mL/min, renal function and other characteristics such as age were
comparable.

In addition, although this study had a 96% consent rate of eligible participants, analysis of the
characteristics between the participants and non-participants (did not consent, undergo baseline
assessment, or complete the follow-up) demonstrated that those completing the study were potentially
better nourished than those that did not complete the study (Table 3-1). In addition, patients with
cognitive, intellectual or communication problems that prevented them from completing the study were
excluded. This may have excluded those at greater risk of poor nutritional status. This point is
indicative of bias towards the null, which implies that the reported benefit of providing structured
nutrition treatment in pre-dialysis care may be conservative and an even stronger association might be
expected if patients more at risk were included.

In addition, due to inherent issues associated with conducting an intervention in the pre-dialysis
phase, the duration of this intervention study was 12 weeks, which is equivalent to other intervention
trials in the literature in pre-dialysis

242

. Being of short duration, a limitation of this study design is that

the long-term effectiveness of this dietary intervention cannot be determined. The question remains,
beyond the follow-up data collection point at week 12, did the intervention group sustain the benefits
associated with receiving individualised care in the long term? Although this is beyond the scope of
the hypotheses set out in this thesis, it is a valid concern. Therefore, to address this, a follow-up study

- 164 -

345

was conducted as an extension of this work,

with the published abstract in Appendix A. This paper

demonstrates participants who received standard care treatment were more likely to reach an endpoint of dialysis initiation or death, compared with the intervention group during the follow-up period.
This investigation was conducted as a quality activity, outside of the scope of this thesis, however,
assists in addressing this identified limitation.

An unavoidable limitation of this study is the fact that the randomised treatment allocation was not
blinded. This limitation is inherent in all nutrition studies supplying an intervention. The potential bias
was minimised by the investigator not alerting the standard-care group that they were getting any less
treatment than the intervention group. The standard care group were all supplied with the education
booklet and then provided with a follow-up appointment in 12 weeks time, where they were offered the
opportunity of a nutrition consultation.

A RCT was selected as it is the strongest study design

possible, and the most practical and feasible method to control for potential bias.

In addition, as the standard care treatment prior to this intervention involved referral to a dietitian on an
ad-hoc basis, it is difficult to determine if the standard care treatment provided in this intervention
was any more or less than usual care. In fact, the comprehensive, practical-based education material
may have been considered to be more than standard care as a consistent message was promoted. In
addition, the fact that the trial was being conducted in the pre-dialysis clinic may have heightened the
awareness of the other practitioners about nutrition support, therefore resulting in more than typical
usual care.

One further limitation involved the difficulty with establishing validity of reported intakes, in particular
energy intake. The crude Goldberg cut-off method was used to identify the proportion of low energy
reporters. Through this method, a similar proportion of weight stable individuals were identified as
135, 136

under-reporting energy intake as has been previously reported in the CKD literature.

However,

the inherent assumptions of this method made it difficult to then determine the relative impact of
under-reporting on the overall results of the study. Ideally, use of a gold-standard methodology, such
as doubly-labelled water or urinary nitrogen to assess energy expenditure and protein balance,
respectively, would assist in accurately determining the validity of intakes, the assessment of these

- 165 -

were considered to be beyond the scope of this investigation. However, overall conclusions maybe
drawn from the consistent trends of improved energy intake in the intervention group, associated with
attenuation of reductions in body cell mass apparent in the standard care group, which recorded an
overall reduction in energy intake.

Finally, the investigator collecting the baseline and follow-up data was also the dietitian providing the
intervention. There is potential that this could introduce bias to the outcome measures of the study. A
number of steps were undertaken to minimise this bias. Firstly, nutrition assessment was undertaken
prior to randomisation.

Secondly, data collection tools based on a subjective assessment were

coupled with objective assessment (TBK) that was blinded to the investigator until the completion of
the study.

8.3

Recommendations for practice

Criteria for referral to Dietitian

At what level of GFR should patients be referred to the dietitian in order to
maximise nutritional intervention opportunities?

Nutrition Assessment
Which specific measures best reflects nutritional status or change in nutritional
status in Chronic Kidney Disease?

Nutrition Prescription/Intervention
What is (are) the appropriate nutritional intervention(s) to optimise nutritional status
in Chronic Kidney Disease and prevent malnutrition?

Implementation and Management

What is the optimal method of implementation and follow up to ensure nutritional
status is maintained or improved?

Figure 8-2

Evidence based practice guideline framework and the clinical questions related

to the nutrition care process for the nutritional management of chronic kidney disease
This thesis makes a significant contribution to the evidence base for nutritional management of Stage
IV and V pre-dialysis CKD, in both the nutrition assessment and the implementation and management

- 166 -

stages of the nutrition care process (see Figure 8-2). It addition, it assists to support existing evidence
for the referral and nutrition prescription stages informing the criteria for dietitian referral and the
nutrition prescription appropriate to optimise nutritional status and prevent malnutrition.

8.3.1 Criteria for referral to dietitian

At what level of GFR should patients be referred to the dietitian in order to maximise nutrition
intervention opportunities?
This study supports individualised dietetic input at the beginning of Stage IV CKD. At the baseline
assessment, the rate of malnutrition was 20% in this cohort who had previously not seen a dietitian for
nutrition support in Stage IV CKD. If the trend of decreasing nutritional status continued as it did in
standard care group, after 12 weeks, the rate of malnutrition would have been over 30%. Provision of
the dietetic intervention with structured follow-up not only prevented further regression, but also
improved the nutrition status of those malnourished at baseline. In addition, quality of life variables
associated with mental health, coping and anxiety rated significantly worse in malnourished patients,
therefore, considering the enormity of this stage of CKD, it is important to consider the effect of further
regression of QOL on other aspects of treatment, such as decisions and suitability for certain modes
of dialysis

Therefore, it is recommended that patients are seen at stage IV to focus on optimising nutritional
status and treating or preventing malnutrition.

The significant co-morbidity load apparent in the

baseline data collection (Section 4-1) also highlights the need for stages I to III to focus on comorbidity management, including maximising physical functioning to promote optimal muscle status
prior to further regression in kidney function.

8.3.2 Nutrition assessment

Which specific measures best reflects nutritional status or change in nutritional status in Chronic
Kidney Disease?

The early detection and management of deteriorating nutritional status in the pre-dialysis population is
important and requires the use of tools that are non-invasive, easy to use, not affected by non-

- 167 -

nutritional factors (e.g. fluid overload), and have little subject burden. SGA was confirmed in this study
for its diagnostic abilities, identifying significantly lower BCM.

In addition, in response to the

randomised-control trial, when SGA change was classified as improve, maintain or regress, there was
a significant change in BCM between the groups.

The results of this study also indicate the scored PG-SGA can provide a diagnosis of malnutrition with
fair accuracy and is also significantly related to BCM-I. Also the benefit of the PG-SGA is that it can
allow identification of small, clinically significant changes in nutrition status.

In addition PG-SGA

demonstrated significant change assessed against the categorised of change in SGA (improved,
maintained or regressed). The PG-SGA has the advantage over other modified SGA tools in that it
identifies a more extensive range of nutrition impact symptoms without requiring the use of
biochemical parameters and the clinical history can be completed by the patient. Finally, the PG-SGA
score significantly correlated with all the generic QOL scales (from physical and mental health
domains) and also symptoms and effects of kidney disease therefore indicating it is also an
appropriate measure of patient-centred wellness.

8.3.3 Nutrition prescription / intervention

What is (are) the appropriate nutritional intervention(s) to optimise nutritional status in Chronic Kidney
Disease and prevent malnutrition?

This intervention supports promoting current Australian-based guidelines for energy and protein to
optimise nutrition status management in pre-dialysis CKD. Dietary counselling in pre-dialysis CKD,
taking into account patients individual clinical condition, co-morbidities and symptoms was an
effective intervention, ensuring an improvement in dietary intake and other related outcomes.
Although energy intake in this intervention appeared below the recommended >125 kJ/kg, and protein
intake slightly over the range of 0.75g/kg to 1.0 g/kg, the aim of this intervention was not focused on
absolute intake values, rather the end points of nutritional status through the intermediate outcome of
improved intake.

In addition, this investigation revealed a significant prevalence of low-energy

reporting in 60% of participants at baseline.

The intervention group, in particular in females,

increasing energy and controlling protein intake resulted in positive patient outcomes.

- 168 -

On the

contrary, when this principle was provided as written material only (as in the standard care group),
regression in dietary intake occurred, and subsequent reduction in body composition, nutritional and
quality of life outcomes were evident, particularly represented by females in the standard care group.

It is recommended that with future research into the benefits of optimising nutritional status to the
progression to dialysis. Due to the improved ability to cope with the symptoms of the disease, this
style of intervention has the potential to delay the progression to dialysis simply by management of
uraemia.

There is evidence in a range of conditions that support the hypothesis that enabling provision of the
appropriate nutritional therapy leads to an improvement in functional body weight (BCM), reflecting an
94

improvement in nutritional status .

8.3.4 Implementation and management

What is the optimal method of implementation and follow up to ensure nutritional status is maintained
or improved?

A major goal of this study was to establish if providing individual nutrition counselling with structured,
regular follow-up, to achieve the prescription as discussed above, resulted in beneficial patient
outcomes. Certainly this style of intervention appeared to be an appropriate vehicle for delivering the
nutrition prescription in pre-dialysis CKD.

In addition, this study demonstrated that provision of

intervention, based on self-management principles; primarily over telephone is an appropriate vehicle

for the follow-up intervention.

As CKD is progressive, following initial nutritional care and follow-up advice, regular nutrition
screening and assessment throughout pre-dialysis care is recommended.

Phase I of this study

demonstrated that recent weight change and SGA related to BCM. Phase II identified the assessment
tools SGA and scored PG-SGA most appropriate to track changes in nutritional status. Therefore,
routine assessment to track progress is recommended using these tools prior to dialysis
commencement.

- 169 -

8.4

Recommendations for future research

1. Further investigate the mechanisms for the gender difference demonstrated in this study. This
may include studies to develop an understanding of health behaviours in CKD, particularly in
relation to the theories of health belief model and social learning theory, regarding types of
interventions (e.g. knowledge vs. skill development) to promote compliance. In addition, further
research is warranted to establish potential gender difference in response to uraemic symptoms.
The results of this study demonstrated females were most at risk of malnutrition than males, and
responded poorly to standard care treatment.

2. Cost-benefit analysis for the implementation of individualised counselling with regular, telephonebased follow-up in pre-dialysis patients. One of the limitations in providing this structured nutrition
intervention is the cost involved with providing more dietetic services than previously established.
It is acknowledged that the intervention featured in this thesis is of greater intensity than previous
50

recommendations for clinical practice . If the cost of having a service far outweighs the savings
to the healthcare facility (less outpatient appointments, fewer hospital admissions, commencement
of dialysis with good nutritional status etc), then this would strengthen the argument for
implementing a model of dietetic care as demonstrated in this intervention.

3. Further research studies should determine the ideal frequency and duration of nutrition support
throughout the entirety of the pre-dialysis phase, to deliver the best outcomes. In particular,
malnourished and/or patients with symptoms of uremia may require follow-up at a different
intensity than those who are identified as low-risk, or SGA A. In addition, a clinical pathway to
establish how regular should patients be re-assessed, or followed up extending to dialysis
commencement, for example, 3-monthly, following initial intervention phase.

4. Future research is required to demonstrate the effectiveness of nutrition intervention in patients

referred late for nephrological care (i.e. referred in less than 6 months of anticipated dialysis
commencement).

As the literature indicates these patients are most at risk of increased

complications due to uraemic malnutrition and poor quality of life. It is possible that these patients

- 170 -

with a lower GFR may have a different focus of intervention, for example, a greater need for the
use of oral supplementation to meet energy requirements and sustain positive nitrogen balance.

5. Based on the studies completed to date in pre-dialysis, it would now be useful to investigate
combining nutritional support with exercise intervention (resistance training) on outcomes such as
body cell mass, with the goal of determining the best treatment approach to optimising nutritional
status prior to dialysis.

6. This study demonstrated limitations for the use of weight change as a marker of change in body
protein, when compared with change in BCM, assessed by TBK. It was concluded this may be
related to fluid status. Future studies to investigate BCM change in conjunction with gold standard
fluid status measure (e.g. deuterium dilution), too adequately estimate functional weight change
and provide additional clinical data on fluid management. At this time, assessment in conjunction
with a clinically available tool, for example, bio-impendence analysis or spectroscopy to establish
the validity for use in clinical practice as a surrogate for the gold-standard measures above.

7. Prior to this intervention, nutrition support studies in CKD utilised biochemical parameters and/or
weight to monitor change in nutritional status. It is recommended that future studies incorporate a
range of outcome measures to determine the effectiveness of dietetic intervention, especially
patient-centred measures including quality of life.

8.5 Conclusion
The studies presented in this thesis provide valuable insight into informing medical nutrition therapy for
the effective management of pre-dialysis CKD patients. This thesis provides strong evidence to inform
the monitoring and evaluation of nutritional status in pre-dialysis patients. It addresses the important
issue of quality of life associated with nutritional status in pre-dialysis and the effect of providing
dietetic treatment on this patient-centred outcome. Finally, the benefits of implementing individualised
dietetic counselling, demonstrated by a strong study design, makes a significant contribution to the
CKD literature and justification of dietetic services within pre-dialysis care.

- 171 -

9.0 References
1.

Heimburger O, Qureshi AR, Blaner WS, Berglund L, and Stenvinkel P: Hand-grip muscle
strength, lean body mass, and plasma proteins as markers of nutritional status in patients with
chronic renal failure close to start of dialysis therapy. Am J Kidney Dis 36:1213-25, 2000

2.

Stenvinkel P, Barany P, Chung SH, Lindholm B, and Heimburger O: A comparative analysis of

nutritional parameters as predictors of outcome in male and female ESRD patients. Nephrol
Dial Transplant 17:1266-1274, 2002

3.

Mitch WE and Maroni BJ: Factors causing malnutrition in patients with chronic uremia. Am J
Kidney Dis 33:176-9, 1999

4.

Hays RD, Kallich JD, Mapes DL, Coons SJ, and Carter WB: Development of the kidney
disease quality of life (KDQOL) instrument. Qual Life Res 3:329-338, 1994

5.

ANZDATA Registry Report 2005, in Australia and New Zealand Dialysis and Transplant
Registry. 2006: Adelaide, South Australia.

6.

Saydah S, Eberhardt M, Rios-Burrows N, Williams D, Geiss L, and Dorsey R: Prevalence of

Chronic Kidney Disease and Associated Risk Factors - United States, 1999--2004. MMWR
Morb Mortal Wkly Rep 56:161-165, 2007

7.

Mitch W and Klahr S, Handbook of Nutrition and the Kidney. 2005: Lippincott Williams &
Wilkins.

8.

National Kidney Foundation: K/DOQI clinical practice guidelines for chronic kidney disease:
evaluation, classification, and stratification. Am J Kidney Dis 39:S1-266, 2002

9.

Chadban SJ, Briganti EM, Kerr PG, Dunstan DW, Welborn TA, Zimmet PZ, et al.: Prevalence
of Kidney Damage in Australian Adults: The AusDiab Kidney Study. J Am Soc Nephrol
14:S131-138, 2003

10.

Go AS, Chertow GM, Fan DJ, McCulloch CE, and Hsu CY: Chronic kidney disease and the
risks of death, cardiovascular events, and hospitalization. N Engl J Med 351:1296-1305, 2004

11.

National Chronic Kidney Disease Strategy - Draft January 2006,

http://www.kidney.org.au/assets/documents/NCKD%20Strategy%20Final%20Report%20Jan
%202006.pdf Accessed October 4, Editor.

- 172 -

12.

Bolton WK: Renal physicians association clinical practice guideline: appropriate patient
preparation for renal replacement therapy: guideline number 3. J Am Soc Nephrol 14:1406-10,
2003

13.

Kopple J: Effect of nutrition on morbidity and mortality in maintenance dialysis patients. Am J

Kidney Dis 24:1002-9, 1994

14.

Mehrotra R and Kopple JD: Nutritional Management of Maintenance Dialysis Patients: Why
Aren't We Doing Better? Ann Rev Nutr 21:343-379, 2001

15.

Tchekmedyian N, Zahyna D, Halpert C, and Heber D: Assessment and maintenance of

nutrition in older cancer patients. Oncology (Huntingt) 6:105-11, 1992

16.

Cederholm T, Jagren C, and Hellstrom K: Outcome of protein-energy malnutrition in elderly

medical patients. Am J Med 98:67-74, 1995

17.

Pupim LB and Ikizler TA: Uremic Malnutrition: New Insights Into an Old Problem. Semin Dial
16:224-232, 2003

18.

Ikizler TA, Greene JH, Wingard RL, Parker RA, and Hakim RM: Spontaneous dietary protein
intake during progression of chronic renal failure. J Am Soc Nephrol 6:1386-1391, 1995

19.

Avesani CM, Draibe SA, Kamimura MA, Colugnati FA, and Cuppari L: Resting energy
expenditure of chronic kidney disease patients: influence of renal function and subclinical
inflammation. Am J Kidney Dis 44:1008-16, 2004

20.

Kang DH, Lee R, Lee HY, Han DS, Cho EY, Lee CH, et al.: Metabolic acidosis and composite
nutritional index (CNI) in CAPD patients. Clin Nephrol 53:124-31, 2000

21.

Mitch W: Mechanisms activated by kidney disease and the loss of muscle mass. Am J Kidney
Dis 38:1337-42, 2001

22.

Pupim LB, Cuppari L, and Ikizler TA: Nutrition and Metabolism in Kidney Disease. Semin
Nephrol 26:134-157, 2006

23.

Mitch WE and Maroni BJ: Nutritional considerations in the treatment of patients with chronic
uremia. Miner Electrolyte Metab 24:285-9, 1998

24.

Kopple JD, Greene T, Chumlea WC, Hollinger D, Maroni BJ, Merrill D, et al.: Relationship
between nutritional status and the glomerular filtration rate: Results from the MDRD Study.
Kidney Int 57:1688-1703, 2000

- 173 -

25.

Duenhas MR, Draibe SA, Avensani CM, Sesso R, and Cuppari L: Influence of renal function
on spontaneous dietary intake and on nutritional status of chronic renal insufficiency patients.
Eur J Clin Nutr 57:1473, 2003

26.

Pupim LB and Cuppari L: Malnutrition in end-stage renal disease: beyond inadequate nutrient
intake. Nephrol News Issues 17:66-71, 2003

27.

Lowrie EG and Lew NL: Death risk in haemodialysis patients: the predictive value of
commonly measured variables and an evaluation of death rate differences between facilities.
American Journal of Kidney Diseases 15:458-482, 1990

28.

Pollock CA, Ibels LS, Allen BJ, Ayass W, Caterson RJ, Waugh DA, et al.: Total body nitrogen
as a prognostic marker in maintenance dialysis. J Am Soc Nephrol 6:82-88, 1995

29.

Churchill D, Taylor, WD., Keshaviah, PR.: Adequacy of dialysis and nutrition in continuous
peritoneal dialysis: association with clinical outcomes. Canada-USA (CANUSA) Peritoneal
Dialysis Study Group. J Am Soc Nephrol 7:198-207, 1996

30.

Fiaccadori E, Lombardi M, Leonardi S, Rotelli C, Tortorella G, and Borghetti A: Prevalence

and clinical outcome associated with preexisting malnutrition in acute renal failure: a
prospective cohort study. J Am Soc Nephrol 10:581-93, 1999

31.

Lawson J, Lazarus R, and Kelly J: Prevalence and prognostic significance of malnutrition in

chronic renal insufficiency. Journal of Renal Nutrition 11:16-22, 2001

32.

Kalantar-Zadeh K and Kopple JD: Relative contributions of nutrition and inflammation to

clinical outcome in dialysis patients. Am J Kidney Dis 38:1343-50, 2001

33.

Pifer TB, McCullough KP, Port FK, Goodkin DA, Maroni BJ, Held PJ, et al.: Mortality risk in
hemodialysis patients and changes in nutritional indicators: DOPPS. Kidney Int 62:2238-45,
2002

34.

Dolson GM, Ellis KJ, Johnson ML, and Adrogue HJ: Incidence and consequences of total
body potassium depletion in chronic hemodialysis patients. Frontiers in Bioscience 8:a126132, 2003

35.

Kalantar-Zadeh K, Block G, McAllister CJ, Humphreys MH, and Kopple JD: Appetite and
inflammation, nutrition, anemia, and clinical outcome in hemodialysis patients. Am J Clin Nutr
80:299-307, 2004

- 174 -

36.

Burrowes JD, Larive B, Chertow GM, Cockram DB, Dwyer JT, Greene T, et al.: Self-reported
appetite, hospitalization and death in haemodialysis patients: findings from the Hemodialysis
(HEMO) Study. Nephrol Dial Transplant 20:2765-74, 2005

37.

Pupim LB, Caglar K, Hakim RM, Shyr Y, and Ikizler TA: Uremic malnutrition is a predictor of
death independent of inflammatory status. Kidney Int 66:2054-2060, 2004

38.

Stenvinkel P, Heimburger O, Paultre F, Diczfalusy U, Wang T, Berglund L, et al.: Strong

association between malnutrition, inflammation, and atherosclerosis in chronic renal failure.
Kidney Int 55:1899-911, 1999

39.

Caravaca F, Arrobas M, Pizarro JL, and Sanchez-Casado E: Uraemic symptoms, nutritional

status and renal function in pre-dialysis end-stage renal failure patients. Nephrol Dial
Transplant 16:776-82, 2001

40.

Attman PO: Long-term treatment with low protein diet in uremia. Contrib Nephrol 53:128-36,
1986

41.

Cooper BA, Aslani A, Ryan M, Ibels LS, and Pollock CA: Nutritional state correlates with renal
function at the start of dialysis. Perit Dial Int 23:291-5, 2003

42.

Kamimura MA, Majchrzak KM, Cuppari L, and Pupim LB: Protein and Energy Depletion in
Chronic Hemodialysis Patients: Clinical Applicability of Diagnostic Tools. Nutr Clin Pract
20:162-175, 2005

43.

Toigo G, Aparicio M, Attman PO, Cano N, Cianciaruso B, Engel B, et al.: Expert Working
Group report on nutrition in adult patients with renal insufficiency (part 1 of 2). Clin Nutr
19:197-207, 2000

44.

Chung SH, Lindholm B, and Lee HB: Is malnutrition an independent predictor of mortality in
peritoneal dialysis patients? Nephrol Dial Transplant 18:2134-40, 2003

45.

Cooper BA, Penne EL, Bartlett LH, and Pollock CA: Protein malnutrition and hypoalbuminemia
as predictors of vascular events and mortality in ESRD. Am J Kidney Dis 43:61-6, 2004

46.

Ash S, Campbell KL, MacLaughlin H, McCoy E, Chan M, Anderson K, et al.: Evidence Based
Practice Guidelines for Nutritional Management of Chronic Kidney Disease. Nutr Diet 63:S35S45, 2006

47.

Australian Kidney Foundation, The CARI Guidelines (Caring for Australians with Renal
Impairment). 2001, Australian Kidney Foundation.

- 175 -

48.

Voss D: The CARI Guidelines: Nutriton and Growth in Kidney Disease. Nephrol 10:S177S203, 2005

49.

K/DOQI: Clinical practice guideline for nutrition in chronic renal failure. Am J Kidney Dis 39:s1s140, 2000

50.

American Dietetic Association, Chronic Kidney Disease (non-dialysis) Medical Nutrition

Therapy Protocol, in Medical Nutrition Therapy Evidence-Based Guides for Practice. 2002,
American Dietetic Association.: Chicago.

51.

NHMRC, How to review the evidence: systematic identification and reivew of the scientfic
literature. 2000, National Helath and Medical Research Council: Canberra.

52.

Wiggins K: Guidelines for Nutritional Care of Renal Patients. Renal Dietitians Dietetic Practice
Group, American Dietetic Association. Chicago: American Dietetic Association 2002

53.

EDTNA/ERCA, European Guidelines for the Nutritional Care of Adult Renal Patients. 2003.

54.

K/DOQI: Kidney Dialysis Outcomes Quality Initiative Nutrition in Chronic Renal Failure.
American Journal of Kidney Diseases 35:s17-s104, 2000

55.

Isenring E, The impact of nutrition support in onclology outpatients recieving radiotherapy to

the gastrointestinal or head and neck area, in Centre for Health Research - Public Health.
2003, QUT: Brisbane.

56.

Ferguson M. Nutrition parameters - Are you wasting your time? in Best Practice in Nutrition
and Dietetics Research. 1996. Brisbane, Australia: QUT Publications.

57.

Eustace JA, Astor B, Muntner PM, Ikizler TA, and Coresh J: Prevalence of acidosis and
inflammation and their association with low serum albumin in chronic kidney disease. Kidney
Int 65:1031-40, 2004

58.

Mitch WE: Getting beyond cross-sectional studies of abnormal nutritional indexes in dialysis
patients. Am J Clin Nutr 77:760-761, 2003

59.

Kaysen GA: Serum albumin concentration in dialysis patients: why does it remain resistant to
therapy? Kidney Int Suppl 87:S92-8, 2003

60.

dos Santos NSJ, Draibe SA, Kamimura MA, Canziani MEF, Cendoroglo M, Junior AG, et al.:
Is Serum Albumin a Marker of Nutritional Status in Hemodialysis Patients without Evidence of
Inflammation? Artif Organs 27:681-686, 2003

- 176 -

61.

Fleck A, Raines G, Hawker F, Trotter J, Wallace P, Ledingham I, et al.: Increased vascular

permeability: a major cause of hypoalbuminaemia in disease and injury. Lancet 1:781-4, 1985

62.

Owen WF: Nutritional status and survival in end-stage renal disease patients. Miner
Electrolyte Metab 24:72-81, 1998

63.

Barrett BJ, Parfrey PS, Morgan J, Barre P, Fine A, Goldstein MB, et al.: Prediction of early
death in end-stage renal disease patients starting dialysis. Am J Kidney Dis 29:214-22, 1997

64.

Jones CH, Newstead CG, Will EJ, Smye SW, and Davison AM: Assessment of nutritional
status in CAPD patients: serum albumin is not a useful measure. Nephrol Dial Transplant
12:1406-13, 1997

65.

Avram MM, Mittman N, Bonomini L, Chattopadhyay J, and Fein P: Markers for survival in
dialysis: a seven-year prospective study. Am J Kidney Dis 26:209-19, 1995

66.

Zeier M: Risk of mortality in patients with end-stage renal disease: the role of malnutrition and
possible therapeutic implications. Horm Res 3:30-4, 2002

67.

Ellis KJ: Human Body Composition: In Vivo Methods. Physiol. Rev. 80:649-680, 2000

68.

Woodrow G, Oldroyd B, Turney JH, Davies PS, Day JM, and Smith MA: Four-component
model of body composition in chronic renal failure comprising dual-energy X-ray
absorptiometry and measurement of total body water by deuterium oxide dilution. Clin Sci
(Lond) 91:763-9, 1996

69.

Woodrow G, Oldroyd B, Wright A, Coward WA, Turney JH, Brownjohn AM, et al.: The
measurement of total body potassium in patients on peritoneal dialysis. Perit Dial Int 21:S1637, 2001

70.

Wang Z, Pierson R, Jr, and Heymsfield S: The five-level model: a new approach to organizing
body-composition research. Am J Clin Nutr 56:19-28, 1992

71.

Heymsfield S: Techniques used in the measurement of body composition: an overview with

emphasis on bioelectrical impedance analysis. 1996

72.

Woodrow G, Oldroyd B, Wright A, Coward WA, Turney JH, Brownjohn AM, et al.:
Abnormalities of body composition in peritoneal dialysis patients. Perit Dial Int 24:169-75,
2004

73.

Gibson R: Principles of nutritional assessment (2nd ed). 2005

- 177 -

74.

Heymsfield S, McManus III C, Seitz S, and Nixon D: Anthropometric assessment of adult

protein-energy malnutrition. Nutritional assessment. Boston: Blackwell Scientific
Publications:2782, 1984

75.

Durnin J and Womersley J: Body fat assessed from total body density and its estimation from
skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. Br J
Nutr 32:77-97, 1974

76.

Hall JC, O'Quigley J, Giles GR, Appleton N, and Stocks H: Upper limb anthropometry: the
value of measurement variance studies. Am J Clin Nutr 33:1846-51, 1980

77.

Woodrow G, Oldroyd B, Turney JH, Tompkins L, Brownjohn AM, and Smith MA: Whole body
and regional body composition in patients with chronic renal failure. Nephrol Dial Transplant
11:1613-1618, 1996

78.

Cohn SH, Vartsky D, Yasumura S, Vaswani AN, and Ellis KJ: Indexes of body cell mass:
nitrogen versus potassium. American Journal of Physiology - Endocrinology and Metabolism
244:E305-E310, 1983

79.

Beddoe AH, Zuidmeer H, and Hill GL: A prompt gamma in vivo neutron activation analysis
facility for measurement of total body nitrogen in the critically ill. Phys Med Biol 29:371-83,
1984

80.

Rayner H, Stroud D, Salamon K, Strauss B, Thomson N, Atkins R, et al.: Anthropometry

underestimates body protein depletion in haemodialysis patients. Nephron 59:33-40, 1991

81.

Cooper BA, Bartlett LH, Aslani A, Allen BJ, Ibels LS, and Pollock CA: Validity of subjective
global assessment as a nutritional marker in end-stage renal failure. American Journal of
Kidney Diseases 40:126-132, 2002

82.

Treuth MS, Butte NF, Ellis KJ, Martin LJ, and Comuzzie AG: Familial resemblance of body
composition in prepubertal girls and their biological parents. American Society for Clinical
Nutrition 74:529-533, 2001

83.

Sanders HN, Tyson IB, Bittle PA, and Ramirez G: Effect of potassium concentration in
dialysate on total body potassium. J Ren Nutr 8:64-8, 1998

84.

Woodrow G, Oldroyd B, Wright A, Coward WA, Turney JH, Brownjohn AM, et al.: The
measurement of total body potassium in patients on peritoneal dialysis. Perit Dial Int 21 Suppl
3:S163-7, 2001

- 178 -

85.

Johansson AC: Nutritional status in peritoneal dialysis: studies in body composition,

lipoprotein metabolism and peritoneal function. Scand J Urol Nephrol Suppl:7-31, 2002

86.

Stall S, Devita MV, Ginsberg NS, Frumkin D, Lynn RI, and Michelis MF: Body composition
assessed by neutron activation analysis in dialysis patients. Annals of the New York Academy
of Sciences 904:558-563, 2000

87.

Dolson GM, Ellis KJ, Johnson ML, and Adrogue HJ: Incidence and consequences of total
body potassium depletion in chronic hemodialysis patients. Front Biosci 1:a126-32, 2003

88.

Pichard C and Kyle UG: Body composition measurements during wasting diseases. Curr Opin
Clin Nutr Metab Care 1:357-61, 1998

89.

Burkinshaw L and Morgan D: Mass and composition of the fat-free tissues of patients with
weight-loss. Clin Sci (Lond) 68:455-62, 1985

90.

Pichard C, Kyle UG, Morabia A, Perrier A, Vermeulen B, and Unger P: Nutritional assessment:
lean body mass depletion at hospital admission is associated with an increased length of stay.
Am J Clin Nutr 79:613-618, 2004

91.

Cupisti A, Licitra R, Chisari C, Stampacchia G, D'Alessandro C, Galetta F, et al.: Skeletal

muscle and nutritional assessment in chronic renal failure patients on a protein-restricted diet.
J Intern Med 255:115-24, 2004

92.

Dumler F and Kilates C: Nutritional Status Assessment and Body Composition Analysis in
Pre-End Stage Renal Disease Patients. Miner Electrolyte Metab 25:397-9, 1999

93.

Kooman JP and Leunissen KML: Malnutrition in patients with end-stage renal disease:
diagnosis, pathophysiology and treatment. Neth J Med:120-132, 1997

94.

Stratton RJ GC, Elia M, Chapter 1: Scientific Criteria for Defining Malnutrition, in Diseaserelated malnutrition: an evidence-based approach to treatment. 2003, CABI Publishing:
Wallingford.

95.

Steiber AL, Kalantar-Zadeh K, Secker D, McCarthy M, Sehgal A, and McCann L: Subjective

Global Assessment in chronic kidney disease: A review. J Ren Nutr 14:191-200, 2004

96.

Detsky AS, McLaughlin JR, Baker JP, Johnston N, Whittaker S, Mendelson RA, et al.: What is
subjective global assessment of nutritional status? J Parenter Enteral Nutr 11:8-13, 1987

- 179 -

97.

Detsky AS, Baker JP, Mendelson RA, Wolman SL, Wesson DE, and Jeejeebhoy KN:
Evaluating the accuracy of nutritional assessment techniques applied to hospitalized patients:
methodology and comparisons. J Parenter Enteral Nutr 8:153-9, 1984

98.

McCann L: Using subjective global assessment to identify malnutrition in the ESRD patient.
Nephrol News Issues 13:18-19, 1999

99.

Campbell KL, Ash S, Bauer J, and PSW D: A critical review of the use of nutrition assessment
tools in chronic kidney disease. Nutr Diet In Press2006

100.

Enia G, Sicuso C, Alati G, and Zoccali C: Subjective global assessment of nutrition in dialysis
patients. Nephrol Dial Transplant 8:1094-8, 1993

101.

Lawson JA, Lazarus R, and Kelly JJ: Prevalence and prognostic significance of malnutrition in
chronic renal insufficiency. J Ren Nutr 11:16-22, 2001

102.

Cooper BA, Bartlett LH, Aslani A, Allen BJ, Ibels LS, and Pollock CA: Validity of subjective
global assessment as a nutritional marker in end-stage renal disease. Am J Kidney Dis
40:126-32, 2002

103.

Fenton SS, Johnston N, Delmore T, Detsky AS, Whitewell J, O'Sullivan R, et al.: Nutritional
assessment of continuous ambulatory peritoneal dialysis patients. ASAIO J 33:650-653, 1987

104.

Young GA, Kopple JD, Lindholm B, Vonesh EF, De Vecchi A, Scalamogna A, et al.: Nutritional
assessment of continuous ambulatory peritoneal dialysis patients: an international study. Am J
Kidney Dis 17:462-471, 1991

105.

Qureshi AR, Alvestrand A, Danielsson A, Divino-Filho JC, Gutierrez A, Lindholm B, et al.:

Factors predicting malnutrition in hemodialysis patients: A cross-sectional study. Kidney Int
53:773-782, 1998

106.

Julien JP, Combe C, and Lasseur C: Subjective global assessment of nutrition a useful
diagnostic tool for nurses? Edtna Erca J 27:193-6, 2001

107.

Cupisti A, D'alessandro C, Morelli E, Rizza GM, Galetta F, Franzoni F, et al.: Nutritional status
and dietary manipulation in predialysis chronic renal failure patients. J Ren Nutr 14:127-133,
2004

108.

Jones CH, Wolfenden RC, and Wells LM: Is subjective global assessment a reliable measure
of nutritional status in hemodialysis? J Ren Nutr 14:26-30, 2004

- 180 -

109.

Desbrow B, Bauer J, Blum C, Kandasamy A, McDonald A, and K M: Assessment of nutritional

status in hemodialysis patients using patient-generated subjective global assessment. J Ren
Nutr 15:211-216, 2005

110.

Coleman K, Norris S, Weston A, Grimmer K, Hillier S, Merlin T, et al., NHMRC additional

levels of evidence and grades for recommendations for developers of guidelines - pilot
program 2005-2007, National Health and Medical Research Council: Canberra.

111.

Afsar, Sezer, Arat, Tutal, Ozdemir, and Haberal: Reliability of Mini Nutritional Assessment in
Hemodialysis Compared With Subjective Global Assessment. J Ren Nutr 16:277-282, 2006

112.

Pifer TB, Mccullough KP, Port FK, Goodkin DA, Maroni BJ, Held PJ, et al.: Mortality risk in
hemodialysis patients and changes in nutritional indicators: DOPPS. Kidney Int 62:2238-2245,
2002

113.

Kalantar-Zadeh K, Kleiner M, Dunne E, Lee GH, and Luft FC: A modified quantitative
subjective global assessment of nutrition for dialysis patients. Nephrol Dial Transplant
14:1732-8, 1999

114.

Kalantar-Zadeh K, Kopple JD, Block G, and Humphreys MH: A malnutrition-inflammation

score is correlated with morbidity and mortality in maintenance hemodialysis patients. Am J
Kidney Dis 38:1251-63, 2001

115.

Visser R, Dekker FW, Boeschoten EW, Stevens P, and Krediet RT: Reliability of the 7-point
subjective global assessment scale in assessing nutritional status of dialysis patients. Adv
Perit Dial 15:222-5, 1999

116.

Davies SJ, Phillips L, Griffiths AM, Naish PF, and Russell GI: Analysis of the effects of
increasing delivered dialysis treatment to malnourished peritoneal dialysis patients. Kidney Int
57:1743-1754, 2000

117.

Kalantar-Zadeh K, Kopple JD, Humphreys MH, and Block G: Comparing outcome

predictability of markers of malnutrition-inflammation complex syndrome in haemodialysis
patients. Nephrol Dial Transplant 19:1507-19, 2004

118.

Isenring E, Bauer J, and Capra S: The scored Patient-generated Subjective Global

Assessment (PG-SGA) and its association with quality of life in ambulatory patients receiving
radiotherapy. Eur J Clin Nutr 57:305-9, 2003

- 181 -

119.

Mobley SL, Hogan EK, Gormley SE, Seifried RM, and Wozney A: PG-SGA is the best
nutrition-related predictor for length of hospital stay in hospitalized patients. J Am Diet Assoc
104:15, 2004

120.

Duenhas MR, Draibe SA, Avesani CM, Sesso R, and Cuppari L: Influence of renal function on
spontaneous dietary intake and on nutritional status of chronic renal insufficiency patients. Eur
J Clin Nutr 57:1473-8, 2003

121.

Maroni BJ, Steinman TI, and Mitch WE: A method for estimating nitrogen intake of patients
with chronic renal failure. Kidney Int 27:58-65, 1985

122.

Bingham S: Urine nitrogen as an independent validatory measure of dietary intake: a study of

nitrogen balance in individuals consuming their normal diet. Am J Clin Nutr 1985

123.

Bingham S and Cummings J: The use of 4-aminobenzoic acid as a marker to validate the
completeness of 24 h urine collections in man. Clin Sci 64:629-35, 1983

124.

Livingstone MBE and Black AE: Markers of the Validity of Reported Energy Intake. J. Nutr.
133:895S-920, 2003

125.

Hoidrup S, Andreasen AH, Osler M, Pedersen AN, and al e: Assessment of habitual energy
and macronutrient intake in adults: Comparison of a seven day food record with a dietary
history interview. Eur J Clin Nutr 56:105, 2002

126.

Bingham SA, Gill C, Welch A, Day K, Cassidy A, Khaw KT, et al.: Comparison of dietary
assessment methods in nutritional epidemiology: weighed records v. 24 h recalls, foodfrequency questionnaires and estimated-diet records. Br J Nutr 72:619-43, 1994

127.

Bingham SA, Cassidy A, Cole TJ, Welch A, Runswick SA, Black AE, et al.: Validation of
weighed records and other methods of dietary assessment using the 24 h urine nitrogen
technique and other biological markers. Br J Nutr 73:531-50, 1995

128.

Stuff J, Garza C, Smith E, Nichols B, and Montandon C: A comparison of dietary methods in

nutritional studies. Am J Clin Nutr 37:300-306, 1983

129.

Cellier KM and Hankin ME: Studies of Nutrition in Pregnancy: I. Some Considerations in

Collecting Dietary Information. Am J Clin Nutr 13:55-62, 1963

130.

Durnin J, McKillop F, Grant S, and Fitzgerald G: Is nutritional status endangered by virtually

no extra intake during pregnancy? Lancet 2:823-5, 1985

- 182 -

131.

Bingham S: The dietary assessment of individuals; methods, accuracy, new techniques and
recommendations. Nutr Abstr Rev 57:705-42, 1987

132.

Hill R and Davies P: The validity of self-reported energy intake as determined using the doubly
labelled water technique. Br J Nutr 85:415-30, 2001

133.

Briefel R, Sempos C, McDowell M, Chien S, and Alaimo K: Dietary methods research in the
third National Health and Nutrition Examination Survey: underreporting of energy intake. Am J
Clin Nutr 65:1203S-1209, 1997

134.

Johansson L, Solvoll K, Bjorneboe G, and Drevon C: Under- and overreporting of energy

intake related to weight status and lifestyle in a nationwide sample. Am J Clin Nutr 68:266274, 1998

135.

Avesani CM, Kamimura MA, Draibe SA, and Cuppari L: Is energy intake underestimated in
nondialyzed chronic kidney disease patients? J Ren Nutr 15:159-65, 2005

136.

Kloppenburg WD, de Jong PE, and Huisman RM: The contradiction of stable body mass
despite low reported dietary energy intake in chronic haemodialysis patients. Nephrol Dial
Transplant 17:1628-1633, 2002

137.

Valderrabano F, Jofre R, and Lopez-Gomez JM: Quality of life in end-stage renal disease
patients. Am J Kidney Dis 38:443-64, 2001

138.

Pais-Ribeiro JL: Quality of life is a primary end-point in clinical settings. Clin Nutr 23:121-130,
2004

139.

DeOreo PB: Hemodialysis patient-assessed functional health status predicts continued

survival, hospitalization, and dialysis-attendance compliance. Am J Kidney Dis 30:204-12,
1997

140.

Kimmel PL, Peterson RA, Weihs KL, Simmens SJ, Alleyne S, Cruz I, et al.: Psychosocial
factors, behavioral compliance and survival in urban hemodialysis patients. Kidney Int 54:24554, 1998

141.

Kalantar-Zadeh K, Kopple JD, Block G, and Humphreys MH: Association among SF36 quality
of life measures and nutrition, hospitalization, and mortality in hemodialysis. J Am Soc Nephrol
12:2797-806, 2001

- 183 -

142.

Khan IH, Garratt AM, Kumar A, Cody DJ, Catto GRD, Edward N, et al.: Patients Perception of
Health on Renal Replacement Therapy - Evaluation Using a New Instrument. Nephrol Dial
Transplant 10:684-689, 1995

143.

Neto JFR, Ferraz MB, Cendoroglo M, Draibe S, Yu L, and Sesso R: Quality of life at the
initiation of maintenance dialysis treatment - a comparison between the SF-36 and the KDQ
questionnaires. Qual Life Res 9:101-107, 2000

144.

Gorodetskaya I, Zenios S, McCulloch CE, Bostrom A, Hsu C-y, Bindman AB, et al.: Healthrelated quality of life and estimates of utility in chronic kidney disease. Kidney Int 68:28012808, 2005

145.

Loos C, Briancon S, Frimat L, Hanesse B, and Kessler M: Effect of end-stage renal disease
on the quality of life of older patients. J Am Geriatr Soc 51:229-33, 2003

146.

Bakewell AB, Higgins RM, and Edmunds ME: Quality of life in peritoneal dialysis patients:
decline over time and association with clinical outcomes. Kidney Int 61:239-48, 2002

147.

Locatelli F and Manzoni C: Treatment modalities in comparison: when do clinical differences

emerge? Nephrol Dial Transplant 1:29-35, 2000

148.

Kutner NG, Zhang R, Barnhart H, and Collins AJ: Health status and quality of life reported by
incident patients after 1 year on haemodialysis or peritoneal dialysis. Nephrol Dial Transplant
20:2159-2167, 2005

149.

Wu AW, Fink NE, Marsh-Manzi JVR, Meyer KB, Finkelstein FO, Chapman MM, et al.:
Changes in Quality of Life during Hemodialysis and Peritoneal Dialysis Treatment: Generic
and Disease Specific Measures. J Am Soc Nephrol 15:743-753, 2004

150.

Lowrie EG, Curtin RB, LePain N, and Schatell D: Medical outcomes study short form-36: a
consistent and powerful predictor of morbidity and mortality in dialysis patients. Am J Kidney
Dis 41:1286-1292, 2003

151.

Mapes DL, Bragg-Gresham JL, Bommer J, Fukuhara S, McKevitt P, Wikstrom B, et al.:

Health-related quality of life in the Dialysis Outcomes and Practice Patterns Study (DOPPS).
Am J Kidney Dis 44:54-60, 2004

152.

Shidler NR, Peterson RA, and Kimmel PL: Quality of life and psychosocial relationships in
patients with chronic renal insufficiency. Am J Kidney Dis 32:557-66, 1998

- 184 -

153.

Moreno F, Sanz-Guajardo D, Lopez-Gomez JM, Jofre R, and Valderrabano F: Increasing the

hematocrit has a beneficial effect on quality of life and is safe in selected hemodialysis
patients. Spanish Cooperative Renal Patients Quality of Life Study Group of the Spanish
Society of Nephrology. J Am Soc Nephrol 11:335-42, 2000

154.

Sesso R and Yoshihiro MM: Time of diagnosis of chronic renal failure and assessment of
quality of life in haemodialysis patients. Nephrol Dial Transplant 12:2111-2116, 1997

155.

Furlong WJ, Feeny DH, Torrance GW, and Barr RD: The Health Utilities Index (HUI (R))
system for assessing health-related quality of life in clinical studies. Ann Med 33:375-384,
2001

156.

Kimmel P, Peterson R, Weihs K, Simmens S, Boyle D, Cruz I, et al.: Aspects of quality of life
in hemodialysis patients. J Am Soc Nephrol 6:1418-1426, 1995

157.

Hickson M and Frost G: An investigation into the relationships between quality of life,
nutritional status and physical function. Clin Nutr 23:213-221, 2004

158.

Dwyer J, Larive B, Leung J, Rocco M, Burrowes J, Chumlea W, et al.: Nutritional status affects
quality of life in Hemodialysis (HEMO) Study patients at baseline. J Ren Nutr 12:213-223,
2002

159.

Isenring EA, Capra S, and Bauer JD: Nutrition intervention is beneficial in oncology
outpatients receiving radiotherapy to the gastrointestinal or head and neck area. British
Journal Of Cancer 91:447-452, 2004

160.

Davidson W, Ash S, Capra S, and Bauer J: Weight stabilisation is associated with improved
survival duration and quality of life in unresectable pancreatic cancer. Clinical Nutrition 23:239247, 2004

161.

Hickman IJ, Jonsson JR, Prins JB, Ash S, Purdie DM, Clouston AD, et al.: Modest weight loss
and physical activity in overweight patients with chronic liver disease results in sustained
improvements in alanine aminotransferase, fasting insulin, and quality of life. Gut 53:413-419,
2004

162.

Wolf AM, Conaway MR, Crowther JQ, Hazen KY, L Nadler J, Oneida B, et al.: Translating
Lifestyle Intervention to Practice in Obese Patients With Type 2 Diabetes: Improving Control
with Activity and Nutrition (ICAN) study. Diabetes Care 27:1570-1576, 2004

- 185 -

163.

Kuehneman T, Saulsbury D, Speltt P, and Chapman DB: Demonstrating the Impact of

Nutrition Intervention in a Heart Failure Program. Journal of the American Dietetic Association
102:1790-1794, 2002

164.

Gleason JA, Bourdet KL, Koehn K, Holay S, and Schaefer EJ: Cardiovascular Risk Reduction
and Dietary Compliance with a Home-Delivered Diet and Lifestyle Modification Program.
Journal of the American Dietetic Association 102:1445-1451, 2002

165.

Ohri-Vachaspati P and Sehgal AR: Quality of life implications of inadequate protein nutrition
among hemodialysis patients. J Ren Nutr 9:9-13, 1999

166.

Laws RA, Tapsell LC, and Kelly J: Nutritional status and its relationship to quality of life in a
sample of chronic hemodialysis patients. J Ren Nutr 10:139-47, 2000

167.

Walters BAJ, Hays RD, Spritzer KL, Fridman M, and Carter WB: Health-related quality of life,
depressive symptoms, anemia, and malnutrition at hemodialysis initiation. Am J Kidney Dis
40:1185-1194, 2002

168.

Canada-USA CPDSG: Adequacy of dialysis and nutrition in continuous peritoneal dialysis:

association with clinical outcomes. Canada-USA (CANUSA) Peritoneal Dialysis Study Group.
Journal of the American Society of Nephrology 7:198-207, 1996

169.

Mehrotra R and Nolph KD: Treatment of advanced renal failure: Low-protein diets or timely
initiation of dialysis? Kidney Int 58:1381-1388, 2000

170.

Mehrotra R, Berman N, Alistwani A, and JD K: Improvement of nutritional status after initiation

of maintenance hemodialysis. Am J Kidney Dis 40:205-7, 2002

171.

Kopple JD: Dietary protein and energy requirements in ESRD patients. Am J Kidney Dis
32:S97-104, 1998

172.

Capra S, Bauer J, Davidson W, and Ash S: Nutritional Therapy for Cancer-Induced Weight
Loss*. Nutr Clin Pract 17:210-213, 2002

173.

Fouque D, Laville M, and Boissel JP: Low protein diets for chronic kidney disease in non
diabetic adults (Cochrane Review). Cochrane Database of Systematic Reviews 2006

174.

Fouque D, Wang P, Laville M, and Boissel JP: Low protein diets for chronic renal failure in non
diabetic adults (Cochrane Review). Cochrane Database of Systematic Reviews 2004

- 186 -

175.

Pedrini MT, Levey AS, Lau J, Chalmers TC, and Wang PH: The effect of dietary protein
restriction on the progression of diabetic and nondiabetic renal diseases: A meta-analysis.
Ann Intern Med 124:627-+, 1996

176.

Kasiske B, Lakatua J, Ma J, and Louis T: A meta-analysis of the effects of dietary protein

restriction on the rate of decline in renal function. Am J Kidney Dis 1998:6, 1998

177.

Fouque D, Laville M, Boissel JP, Chifflet R, Labeeuw M, and Zech PY: Controlled Low Protein
Diets in Chronic Renal-Insufficiency - Metaanalysis. Br Med J 304:216-220, 1992

178.

Waugh Nr RAM: Protein restriction for diabetic renal disease. Waugh NR, Robertson AM.
Protein restriction for diabetic renal disease 32004

179.

Aparicio M, Chauveau P, and Combe C: Low protein diets and outcome of renal patients. J
Nephrol 14:433-439, 2001

180.

Klahr S, Levey AS, Beck GJ, Caggiula AW, Hunsicker L, Kusek JW, et al.: The Effects of
Dietary-Protein Restriction and Blood-Pressure Control on the Progression of Chronic RenalDisease. N Engl J Med 330:877-884, 1994

181.

Williams B, Hattersley J, Layward E, and Walls J: Metabolic acidosis and skeletal muscle
adaptation to low protein diets in chronic uremia. Kidney Int 40:779-86, 1991

182.

D'Amico G, Gentile MG, Fellin G, Manna G, and Cofano F: Effect of dietary protein restriction
on the progression of renal failure: a prospective randomized trial. Nephrol. Dial. Transplant.
9:1590-1594, 1994

183.

Johnson DW: Dietary protein restriction as a treatment for slowing chronic kidney disease
progression: The case against. Nephrol 11:58-62, 2006

184.

Kopple JD, Levey AS, Greene T, Chumlea WC, Gassman JJ, Hollinger DL, et al.: Effect of
dietary protein restriction on nutritional status in the Modification of Diet in Renal Disease
Study. Kidney Int 52:778-91, 1997

185.

Doleck TA, Olson MB, Caggiula AW, Dwyer JT, Milas NC, Gillis BP, et al.: Registered Dietitian
Time Requirements in the Modification of Diet in Renal Disease Study. J Am Diet Assoc
95:1307-1312, 1995

186.

Guarnieri G, Toigo G, Situlin R, Faccini L, Coli U, Landini S, et al.: Muscle biopsy studies in
chronically uremic patients: evidence for malnutrition. Kidney Int Suppl 16:S187-93, 1983

- 187 -

187.

Fried LF, Orchard TJ, and Kasiske BL: Effect of lipid reduction on the progression of renal
disease: A meta-analysis. Kidney Int 59:260-269, 2001

188.

Kellet E, Smith A, and Schmerlaib Y, The Australian Guide to Healthy Eating - Background
information for nutrition educators. 1998, Canberra ATC : Commonwealth Department of
Health and Family Services, 1998.

189.

Kopple J, Monteon F, and Shaib J: Effect of energy intake on nitrogen metabolism in

nondialyzed patients with chronic renal failure. Kidney Int 29:734-42, 1986

190.

Bossola M, Muscaritoli M, Tazza L, Giungi S, Tortorelli A, Fanelli F, et al.: Malnutrition in

hemodialysis patients: what therapy? Am J Kidney Dis 46:371-86, 2005

191.

Burrowes J, Bluestone P, Wang J, and Pierson Jr R: The effects of moderate doses of

megestrol acetate on nutritional status and body composition in a hemodialysis patient. J Ren
Nutr 9:89-94, 1999

192.

Simons J, Schols A, Hoefnagels J, Westerterp K, Velde G, and Wouters E: Effects of

medroxyprogesterone acetate on food intake, body composition, and resting energy
expenditure in patients with advanced, nonhormone-sensitive cancer. Cancer 82:553-560,
1998

193.

Oster M, Enders S, Samuels S, Cone L, Hooton T, Browder H, et al.: Megestrol Acetate in

Patients with AIDS and Cachexia. Ann Intern Med 121:400-408, 1994

194.

Strang P: The effect of megestrol acetate on anorexia, weight loss and cachexia in cancer and
AIDS patients (review). Anticancer Res 17:657-62, 1997

195.

Boccanfuso J, Hutton M, and McAllister B: The effects of megestrol acetate on nutritional

parameters in a dialysis population. J Ren Nutr 10:36-43, 2000

196.

Bossola M, Tazza L, Giungi S, and Luciani G: Anorexia in hemodialysis patients: An update.

Kidney Int 70:417-422, 2006

197.

Johansen K, Mulligan K, and Schambelan M: Anabolic Effects of Nandrolone Decanoate in

Patients Receiving Dialysis A Randomized Controlled Trial. JAMA 281:1275-1281, 1999

198.

Gascon A, Belvis JJ, Beser E, Perez A, Rubio V, and Berisa F: Comparison between
erythropoietin (EPO) and nandrolone decanoate (ND) in elderly patients on hemodialysis:
erythropoietic and nutritional effects. Kidney Int 55:366-367, 1999

- 188 -

199.

Cliffe M, Bloodworth LL, and Jibani MM: Can malnutrition in predialysis patients be prevented
by dietetic intervention? J Ren Nutr 11:161-5, 2001

200.

Leon J, Majerle A, Soinski J, Kushner I, Ohri-Vachaspati P, and Sehgal A: Can a nutrition

intervention improve albumin levels among hemodialysis patients? A pilot study. J Ren Nutr
11:9-15, 2001

201.

Caglar K, Fedje L, Dimmitt R, Hakim RM, Shyr Y, and Ikizler TA: Therapeutic effects of oral
nutritional supplementation during hemodialysis. Kidney Int 62:1054-1059, 2002

202.

Sharma RK and Sahu KM: Nutrition in dialysis patients. J Indian Med Assoc 99:206-8, 2001

203.

Steiber A, Handu D, Cataline D, Deighton T, and Weatherspoon L: The impact of nutrition

intervention on a reliable morbidity and mortality indicator: the hemodialysis-prognostic
nutrition index. J Ren Nutr 13:186-90, 2003

204.

Akpele L and Bailey JL: Nutrition counseling impacts serum albumin levels. J Ren Nutr
14:143-148, 2004

205.

Ford JC, Pope JF, Hunt AE, and Gerald B: The effect of diet education on the laboratory
values and knowledge of hemodialysis patients with hyperphosphatemia. J Ren Nutr 14:3644, 2004

206.

Kalantar-Zadeh K, Braglia A, Chow J, Kwon O, Kuwae N, Colman S, et al.: An antiinflammatory and antioxidant nutritional supplement for hypoalbuminemic hemodialysis
patients: a pilot/feasibility study. J Ren Nutr 15:318-31, 2005

207.

Stratton R, Bircher G, Fouque D, Stenvinkel P, de Mutsert R, Engfer M, et al.: Multinutrient

oral supplements and tube feeding in maintenance dialysis: a systematic review and metaanalysis. Am J Kidney Dis 46:387-405, 2005

208.

Leon JB, Albert JM, Gilchrist G, Kushner I, Lerner E, Mach S, et al.: Improving albumin levels
among hemodialysis patients: A community-based randomized controlled trial. Am J Kidney
Dis 48:28-36, 2006

209.

Cuppari L, Medeiros F, Papini H, Neto M, Canziani M, Martini L, et al.: Effectiveness of oral

energy-protein supplementation in severely malnourished hemodialysis patients. J Ren Nutr
4:127-135, 1994

210.

Patel M, Kitchen S, and Miligan P: The effect of dietary supplements on the nPCR in stable
hemodialysis patients. J Ren Nutr 10:69-75, 2000

- 189 -

211.

Patel M and Raftery M: The Use of Dietary Supplements in Continuous Ambulatory Peritoneal
Dialysis Patients. J Ren Nutr 7:129-133, 1997

212.

Sayce H, Rowe P, and McGonigle R: Percutaneous endoscopic gastrostomy feeding in

haemodialysis out-patients. J Hum Nutr Diet 13:333-341, 2000

213.

Gillies BP, Caggiula AW, Chiavacci AT, Coyne T, Doroshenko L, Milas NC, et al.: Nutrition
Intervention Program of the Modification of Diet in Renal Disease Study: A Self-Management
Approach. J Am Diet Assoc 95:1288-1294, 1995

214.

Milas N, Nowalk MP, Akpele L, Castaldo L, Coyne T, Doroshenko L, et al.: Factors Associated
with Adherence to the Dietary Protein Intervention in the Modification of Diet in Renal Disease
Study. J Am Diet Assoc 95:1295-1300, 1995

215.

Chan M and Brown M: Nutrition Management in Pre-Dialysis Assessment Clinic 18 Months

Experience. Nephrol 2004

216.

Patwardhan A, Bartlett L, Chan M, and Ryan C: A Survey of Renal Dietitan Staffing in New
South Wales. Nephrol 9:A37, 2004

217.

Chan M and Brown M. Nutrition Assessment in Pre-Dialysis Assessment Clinic. in 12th

ICNMRD. 2004. Venice, Italy.

218.

K/DOQI.: Clinical practice guideline for nutrition in chronic renal failure. Americn Journal of
Kidney Diseases 39:s1-s140, 2000

219.

Pupim LB and Ikizler TA: Assessment and monitoring of uremic malnutrition. J Ren Nutr 14:619, 2004

220.

Splett P, Cost Outcomes of Nutrition Interventions, Part I Outcomes Research, Mead Johnson
&. Co, Evansvillc, Ind,(1996).

221.

Lacey K and Pritchett E: Nutrition Care Process and Model: ADA adopts road map to quality
care and outcomes management. J Am Diet Assoc 103:1061-1072, 2003

222.

Cohn SH, Vaswani AN, Yasumura S, Yuen K, and Ellis KJ: Assessment of cellular mass and
lean body mass by noninvasive nuclear techniques. J Lab Clin Med 105:305-11, 1985

223.

Wang Z, St-Onge MP, Lecumberri B, Pi-Sunyer FX, Heshka S, Wang J, et al.: Body cell mass:
model development and validation at the cellular level of body composition. Am J Physiol
Endocrinol Metab 286:123-128, 2004

- 190 -

224.

Talluri A, Liedtke R, Mohamed EI, Maiolo C, Martinoli R, and De Lorenzo A: The application of
body cell mass index for studying muscle mass changes in health and disease conditions.
Acta Diabetol 40:S286-9, 2003

225.

Campbell KL, Ash S, Bauer J, and Davies PSW: Critical review of nutrition assessment tools
to measure malnutrition in chronic kidney disease. Nutr Diet 64:23-30, 2007

226.

Ottery F, Patient-Generated Subjective Global Assessment, in The Clinical Guide to Oncology

Nutrition, P. McCallum, Polisena, C.,, Editor. 2000, American Dietetic Association: Chicago. p.
11-23.

227.

Kopple JD: The National Kidney Foundation K/DOQI clinical practice guidelines for dietary
protein intake for chronic dialysis patients. Am J Kidney Dis 38:S68-73, 2001

228.

Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, and Roth D: A more accurate method to
estimate glomerular filtration rate from serum creatinine: a new prediction equation.
Modification of Diet in Renal Disease Study Group. Ann Intern Med 130:461-470, 1999

229.

Karnofsky D and Burchenal J, The clinical evaluation of chemotherapeutic agents in cancer, in

Evaluation of Chemotherapeutic Agents. 1949, Columbia University Press: New York, NY. p.
191-205.

230.

Allen KL, Miskulin D, Yan G, Dwyer JT, Frydrych A, Leung J, et al.: Association of nutritional
markers with physical and mental health status in prevalent hemodialysis patients from the
HEMO study. J Ren Nutr 12:160-9, 2002

231.

Burrowes J, Cockram D, Dwyer J, Larive B, Paranandi L, Bergen C, et al.: Cross-sectional

relationship between dietary protein and energy intake, nutritional status, functional status,
and comorbidity in older versus younger hemodialysis patients. J Ren Nutr 12:87-95, 2002

232.

Chung SH, Lindholm B, and Lee HB: Influence of initial nutritional status on continuous
ambulatory peritoneal dialysis patient survival. Perit Dial Int 20:19-26, 2000

233.

Sietsema KE, Amato A, Adler SG, and Brass EP: Exercise capacity as a predictor of survival
among ambulatory patients with end-stage renal disease. Kidney Int 65:719-724, 2004

234.

Edgell ET, Coons SJ, Carter WB, Kallich JD, Mapes D, Damush TM, et al.: A review of healthrelated quality-of-life measures used in end-stage renal disease. Clin Ther 18:887-938, 1996

235.

Ware JE, Jr and Sherbourne CD: The MOS 36-item short-form health survey (SF-36). I.
Conceptual framework and item selection. Medical Care 30:473-483, 1992

- 191 -

236.

Lapaucis A, Muirhead N, Keown P, and Wong C: A disease-specific questionnaire for

assessing quality of life in patients on hemodialysis. Nephron 60:302-306, 1992

237.

Korevaar JC, Merkus MP, Jansen MAM, Dekker FW, Boeschoten EW, Krediet RT, et al.:
Validation of the KDQOL-SF: a dialysis-targeted health measure. Qual Life Res 11:437-447,
2002

238.

Rao S, Carter WB, Mapes DL, Kallich JD, Kamberg CJ, Spritzer KL, et al.: Development of
subscales from the symptoms/problems and effects of kidney disease scales of the kidney
disease quality of life instrument. Clin Ther 22:1099-1111, 2000

239.

Mapes DL, Lopes AA, Satayathum S, McCullough KP, Goodkin DA, Locatelli F, et al.: Healthrelated quality of life as a predictor of mortality and hospitalization: the Dialysis Outcomes and
Practice Patterns Study (DOPPS). Kidney Int 64:339-349, 2003

240.

Hays RD, The KDQOL Working Group - Frequently Asked Questions. 2004.

241.

KDQOL-SF(TM) Users Manual Version 13. p. 15-16.

242.

Castaneda C, Gordon PL, Uhlin KL, Levey AS, Kehayias JJ, Dwyer JT, et al.: Resistance
training to counteract the catabolism of a low-protein diet in patients with chronic renal
insufficiency. A randomized, controlled trial. Ann Intern Med 135:965-76, 2001

243.

Charlson ME, Pompei P, Ales KL, and MacKenzie CR: A new method of classifying prognostic
comorbidity in longitudinal studies: development and validation. J Chronic Dis 40:373-83, 1987

244.

van Manen JG, Korevaar JC, Dekker FW, Boeschoten EW, Bossuyt PMM, and Krediet RT:
Adjustment for Comorbidity in Studies on Health Status in ESRD Patients: Which Comorbidity
Index to Use? J Am Soc Nephrol 14:478-485, 2003

245.

Sundararajan V, Henderson T, Perry C, Muggivan A, Quan H, and Ghali WA: New ICD-10
version of the Charlson comorbidity index predicted in-hospital mortality. J Clin Epidemiol
57:1288-94, 2004

246.

Metz C: Basic principles of ROC analysis. Semin Nucl Med 8:283-98, 1978

247.

Wiggins K and Harvey K: A review of guidelines for nutrition care of renal patients. J Ren Nutr
12:190-196, 2002

248.

Conway JM, Ingwersen LA, Vinyard BT, and Moshfegh AJ: Effectiveness of the US
Department of Agriculture 5-step multiple-pass method in assessing food intake in obese and
nonobese women. Am J Clin Nutr 77:1171-1178, 2003

- 192 -

249.

McLaughlin G: SMOG Grading- A New Readability Formula. J Reading 1969

250.

Hosmer D and Lemeshow S, Applied Logistic Regression. 2000. 2000, New York, NY: John
Wiley & Sons.

251.

Bland JM and Altman DG: Statistical methods for assessing agreement between two methods
of clinical measurement. Lancet 1:307-10, 1986

252.

McCann L, Use of the 7-point SGA. 2004. p. Personal Communication.

253.

Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al.: Towards
Complete and Accurate Reporting of Studies of Diagnostic Accuracy: The STARD Initiative.
Clin Chem 49:1-6, 2003

254.

Jeejeebhoy KN: Nutritional assessment. Nutrition 16:585-590, 2000

255.

Johansson A, Samuelsson O, Haraldsson B, Bosaeus I, and Attman P: Body composition in

patients treated with peritoneal dialysis. Nephrol Dial Transplant 13:1511-1517, 1998

256.

Attman P, Ewald J, and Isaksson B: Body composition during long-term treatment of uremia
with amino acid supplemented low-protein diet. Am J Clin Nutr 33:801-810, 1980

257.

Lukaski H, Johnson P, Bolonchuk W, and Lykken G: Assessment of fat-free mass using

bioelectrical impedance measurements of the human body. Am J Clin Nutr 41:810-817, 1985

258.

Manuela Dittmar HR: Validation of different bioimpedance analyzers for predicting cell mass
against whole-body counting of potassium as a reference method. Am J Hum Biol 16:697-703,
2004

259.

Corcoran C, Anderson EJ, Burrows B, Stanley T, Walsh M, Poulos AM, et al.: Comparison of
total body potassium with other techniques for measuring lean body mass in men and women
with AIDS wasting. Am J Clin Nutr 72:1053-1058, 2000

260.

Avram MM, Bonomini LV, Sreedhara R, and Mittman N: Predictive value of nutritional markers
(albumin, creatinine, cholesterol, and hematocrit) for patients on dialysis for up to 30 years.
Am J Kidney Dis 28:910-7, 1996

261.

Bauer J, Capra S, and Ferguson M: Use of the scored Patient-Generated Subjective Global
Assessment (PG-SGA) as a nutrition assessment tool in patients with cancer. Eur J Clin Nutr
56:779-85, 2002

262.

Ottery FD: Definition of standardized nutritional assessment and interventional pathways in

oncology. Nutrition 12:S15-S19, 1996

- 193 -

263.

Steiber AL, Leon J, Sehgal A, Secker D, McCarthy M, Kalantar-Zadeh K, et al.: A multicenter

study of subjective global assessment (SGA) validity and reliability in the hemodialysis (HD)
population. J Ren Nutr 16:172, 2006

264.

Persson C, Sjoden PO, and Glimelius B: The Swedish version of the patient-generated
subjective global assessment of nutritional status: gastrointestinal vs urological cancers. Clin
Nutr 18:71-7, 1999

265.

Rosenbaum K, Wang J, Pierson R, and Kotler D: Time-dependent variation in weight and

body composition in healthy adults. J Parenter Enteral Nutr 24:52-55, 2000

266.

Castaneda C, Charnley J, Evans W, and Crim M: Elderly women accommodate to a lowprotein diet with losses of body cell mass, muscle function, and immune response. Am J Clin
Nutr 62:30-39, 1995

267.

Arora N and Rochester D: Respiratory muscle strength and maximal voluntary ventilation in
undernourished patients. Am Rev Respir Dis 126:5-8, 1982

268.

Wolfe RR: The underappreciated role of muscle in health and disease. Am J Clin Nutr 84:475482, 2006

269.

Wang Z, Pi-Sunyer FX, Kotler DP, Wang J, Pierson RN, Jr., and Heymsfield SB: Magnitude
and variation of ratio of total body potassium to fat-free mass: a cellular level modeling study.
Am J Physiol Endocrinol Metab 281:E1-7, 2001

270.

Brennan BL, Letteri JM, Cohn SH, and Ellis KJ: Serial measurements of body composition and
total body mineral content in dialysis and nondialysis patients with renal failure. Miner
Electrolyte Metab 13:451-61, 1987

271.

Chauveau P, Vendrely B, Haggan W, Barthe N, Rigalleau V, Combe C, et al.: Body

composition of patients on a very low-protein diet: A two-year survey with DEXA. J Ren Nutr
13:282-287, 2003

272.

Fitts SS, Guthrie MR, and Blagg CR: Exercise coaching and rehabilitation counseling improve
quality of life for predialysis and dialysis patients. Nephron 82:115-21, 1999

273.

Agin D, Gallagher D, Wang J, Heymsfield SB, Pierson RN, Jr, and Kotler DP: Effects of whey
protein and resistance exercise on body cell mass, muscle strength, and quality of life in
women with HIV. AIDS (London, England) 15:2431-2440, 2001

- 194 -

274.

Grunfeld C, Kotler DP, Dobs A, Glesby M, and Bhasin S: Oxandrolone in the treatment of HIVassociated weight loss in men: a randomized, double-blind, placebo-controlled study. J Acquir
Immune Defic Syndr 41:304-314, 2006

275.

Isenring E, Capra S, Bauer J, and Davies PS: The impact of nutrition support on body
composition in cancer outpatients receiving radiotherapy. Acta Diabetol 40:S162-4, 2003

276.

Rutishauser I: Dietary intake measurements. Pub Health Nutr 8:1100-1107, 2005

277.

Voss S, Kroke A, Klipstein-Grobusch K, and Boeing H: Is macronutrient composition of dietary

intake data affected by underreporting? Results from the EPIC-Potsdam study. European
Journal of Clinical Nutrition 52:119-126, 1998

278.

Baldwin C and Parsons TJ: Dietary advice and nutritional supplements in the management of
illness-related malnutrition: systematic review. Clin Nutr 23:1267-1279, 2004

279.

Ravasco P, Monteiro-Grillo I, Vidal PM, and Camilo ME: Dietary Counseling Improves Patient
Outcomes: A Prospective, Randomized, Controlled Trial in Colorectal Cancer Patients
Undergoing Radiotherapy. J Clin Oncol 23:1431-1438, 2005

280.

Mandayam S and Mitch WE: Dietary protein restriction benefits patients with chronic kidney
disease. Nephrol 11:53-57, 2006

281.

Meloni C, Morosetti M, Suraci C, Pennafina MG, Tozzo C, Taccone GM, et al.: Severe dietary
protein restriction in overt diabetic nephropathy: benefits or risks? J Ren Nutr 12:96-101, 2002

282.

Castaneda C: Protein Restriction in Renal Disease. Nutr Clinical Care 4:103-112, 2001

283.

Aparicio M, Chauveau P, De Precigout V, Bouchet JL, Lasseur C, and Combe C: Nutrition and
outcome on renal replacement therapy of patients with chronic renal failure treated by a
supplemented very low protein diet. J Am Soc Nephrol 11:708-716, 2000

284.

Meloni C, Tatangelo P, Cipriani S, Rossi V, Suraci C, Tozzo C, et al.: Adequate protein dietary
restriction in diabetic and nondiabetic patients with chronic renal failure. J Ren Nutr 14:208213, 2004

285.

Zarazaga A, Garcia-de-Lorenzo L, Garcia-Luna PP, Garcia-Peris P, Lopez-Martinez J,

Lorenzo V, et al.: Nutritional support in chronic renal failure: systematic review. Clin Nutr
20:291-299, 2001

286.

Kalantar-Zadeh K, Kilpatrick RD, Kuwae N, McAllister CJ, Alcorn Jr H, Kopple JD, et al.:
Revisiting mortality predictability of serum albumin in the dialysis population: time

- 195 -

dependency, longitudinal changes and population-attributable fraction. Nephrol Dial

Transplant:gfh941, 2005
287.

Roubenoff R, Heymsfield S, Kehayias J, Cannon J, and Rosenberg I: Standardization of

nomenclature of body composition in weight loss. Am J Clin Nutr 66:192-196, 1997

288.

McBride CM and Rimer BK: Using the telephone to improve health behavior and health
service delivery. Patient Educ Couns 37:3-18, 1999

289.

VanWormer JJ, Boucher JL, and Pronk NP: Telephone-Based Counseling Improves Dietary
Fat, Fruit, and Vegetable Consumption: A Best-Evidence Synthesis. J Am Diet Assoc
106:1434-1444, 2006

290.

GESICA Investigators: Randomised trial of telephone intervention in chronic heart failure:

DIAL trial. Br Med J 331:425, 2005

291.

Visnegarwala F, Raghavan SS, Mullin CM, Bartsch G, Wang J, Kotler D, et al.: Sex
differences in the associations of HIV disease characteristics and body composition in
antiretroviral-naive persons. Am J Clin Nutr 82:850-856, 2005

292.

Kotler DP, Thea DM, Heo M, Allison DB, Engelson ES, Wang J, et al.: Relative influences of
sex, race, environment, and HIV infection on body composition in adults. Am J Clin Nutr
69:432-439, 1999

293.

Roubenoff R, Hughes VA, Dallal GE, Nelson ME, Morganti C, Kehayias JJ, et al.: The Effect
of Gender and Body Composition Method on the Apparent Decline in Lean Mass-Adjusted
Resting Metabolic Rate With Age. J Gerontol A Biol Sci Med Sci 55:M757-760, 2000

294.

Newman AB, Haggerty CL, Goodpaster B, Harris T, Kritchevsky S, Nevitt M, et al.: Strength
and Muscle Quality in a Well-Functioning Cohort of Older Adults: The Health, Aging and Body
Composition Study. J Am Geriatr Soc 51:323-330, 2003

295.

Kyle U, Genton L, Hans D, Karsegard L, Slosman D, and Pichard C: Age-related differences

in fat-free mass, skeletal muscle, body cell mass and fat mass between 18 and 94 years. Eur
J Clin Nutr 55:663-672, 2001

296.

Kopple J, Berg R, Houser H, Steinman T, and Teschan P: Nutritional status of patients with
different levels of chronic renal insufficiency. Modification of Diet in Renal Disease (MDRD)
Study Group. Kidney Int Suppl 27:S184-94, 1989

- 196 -

297.

Carvalho KT, Silva MIB, and Bregman R: Nutritional profile of patients with chronic renal
failure. J Ren Nutr 14:97-100, 2004

298.

Bodenheimer T, Lorig K, Holman H, and Grumbach K: Patient self-management of chronic

disease in primary care. JAMA 288:2469, 2002

299.

Chung ML, Moser DK, Lennie TA, Worrall-Carter L, Bentley B, Trupp R, et al.: Gender
differences in adherence to the sodium-restricted diet in patients with heart failure. J Card Fail
12:628-634, 2006

300.

Becker M, The Health Belief Model and Personal Health Behavior. 1974: Charles B. Slack.

301.

Miller Ck ELKGSL: Nutrition education improves metabolic outcomes among older adults with
diabetes mellitus: results from a randomized controlled trial. Prev Med 2002

302.

Buckelew S, Shutty Jr M, Hewett J, Landon T, Morrow K, and Frank R: Health locus of control,
gender differences and adjustment to persistent pain. Pain 42:287-94, 1990

303.

Takaki J and Yano E: Possible gender differences in the relationships of self-efficacy and the
internal locus of control with compliance in hemodialysis patients. Behav Med 32:5-11, 2006

304.

Kusek JW, Greene P, Wang SR, Beck G, West D, Jamerson K, et al.: Cross-sectional study of
health-related quality of life in African Americans with chronic renal insufficiency: The African
American Study of Kidney Disease and Hypertension Trial. Am J Kidney Dis 39:513-524, 2002

305.

Rocco MV, Gassman JJ, Wang SR, and Kaplan RM: Cross-sectional study of quality of life
and symptoms in chronic renal disease patients: The modification of diet in renal disease
study. Am J Kidney Dis 29:888-896, 1997

306.

Chow FYF, Briganti EM, Kerr PG, Chadban SJ, Zimmet PZ, and Atkins RC: Health-related
quality of life in Australian adults with renal insufficiency: A population-based study. Am J
Kidney Dis 41:596-604, 2003

307.

Ware J, Kosinski M, and Keller S: SF-36 Physical and Mental Health Summary Scales: A
Users Manual. Health Institute, Boston 8:23-28, 1994

308.

Carmichael P, Popoola J, John I, Stevens PE, and Carmichael AR: Assessment of quality of
life in a single centre dialysis population using the KDQOL-SFTm questionnaire. Qual Life Res
V9:195-205, 2000

- 197 -

309.

Vazquez I, Valderrabano F, Fort J, Jofrac R, Lopez-Gomez JM, Moreno F, et al.: Psychosocial

Factors and Health-Related Quality of Life in Hemodialysis Patients. Qual Life Res V14:179190, 2005

310.

Martinez-Castelao A, Gorriz JL, Garcia-Lopez F, Lopez-Revuelta K, De Alvaro F, and Cruzado

JM: Perceived health-related quality of life and comorbidity in diabetic patients starting dialysis
(CALVIDIA study). J Nephrol 17:544-51, 2004

311.

Fukuhara S, Lopes AA, Bragg-Gresham JL, Kurokawa K, Mapes DL, Akizawa T, et al.:
Health-related quality of life among dialysis patients on three continents: the Dialysis
Outcomes and Practice Patterns Study. Kidney Int 64:1903-1910, 2003

312.

Hays R and Woolley J: The Concept of Clinically Meaningful Difference in Health-Related

Quality-of-Life Research. Pharmacoeconomics 18:419-423, 2000

313.

Chow K, Szeto C, Kum L, and Kwan B: Improved health-related quality of life and left
ventricular hypertrophy among dialysis patients treated with parathyroidectomy. J Nephrol
16:878-885, 2003

314.

Bergner M, Bobbitt R, Carter W, and Gilson B: The Sickness Impact Profile: development and
final revision of a health status measure. Med Care 19:787-805, 1981

315.

Kurella M, Luan J, Yaffe K, and Chertow G: Validation of the Kidney Disease Quality of Life
(KDQOL) Cognitive Function subscale. Kidney Int 66:2361, 2004

316.

Kurella M, Chertow GM, Fried LF, Cummings SR, Harris T, Simonsick E, et al.: Chronic
Kidney Disease and Cognitive Impairment in the Elderly: The Health, Aging, and Body
Composition Study. J Am Soc Nephrol 16:2127-2133, 2005

317.

Kurella M, Chertow GM, Luan J, and Yaffe K: Cognitive Impairment in Chronic Kidney
Disease. J Am Geriatr Soc 52:1863-1869, 2004

318.

Wilson R, Barnes L, Mendes de Leon C, Aggarwal N, Schneider J, Bach J, et al.: Depressive

symptoms, cognitive decline, and risk of AD in older persons. Neurology 59:364-370, 2002

319.

Mehta KM, Simonsick EM, Penninx BWJH, Schulz R, Rubin SM, Satterfield S, et al.:
Prevalence and Correlates of Anxiety Symptoms in Well-Functioning Older Adults: Findings
from the Health Aging and Body Composition Study. J Am Geriatr Soc 51:499-504, 2003

- 198 -

320.

Griva K, Thompson D, Jayasena D, Davenport A, Harrison M, and Newman SP: Cognitive

functioning pre- to post-kidney transplantation--a prospective study. Nephrol Dial Transplant
21:3275-3282, 2006

321.

Rossert J, Levin A, Roger SD, Horl WH, Fouqueray B, Gassmann-Mayer C, et al.: Effect of
early correction of anemia on the progression of CKD. American Journal Of Kidney Diseases:
The Official Journal Of The National Kidney Foundation 47:738-750, 2006

322.

Moreno F, Sanz-Guajardo D, Lopez-Gomez JM, Jofre R, and Valderrabano F: Increasing the

hematocrit has a beneficial effect on quality of life and is safe in selected hemodialysis
patients. Spanish Cooperative Renal Patients Quality of Life Study Group of the Spanish
Society of Nephrology. Journal Of The American Society Of Nephrology: JASN 11:335-342,
2000

323.

Painter P, Carlson L, Carey S, Paul SM, and Myll J: Physical functioning and health-related
quality-of-life changes with exercise training in hemodialysis patients. Am J Kidney Dis
35:482-492, 2000

324.

Ware J, Snow K, Kosinski M, and Gandek B, SF-36 Health Survey: Manual and Interpretation
Guide. Health Institute, Boston. Vol. 2. 2000.

325.

Stansfeld SA, Roberts R, and Foot SP: Assessing the validity of the SF-36 General Health
Survey. Qual Life Res V6:0, 1997

326.

Franke GH, Reimer J, Philipp T, and Heemann U: Aspects of quality of life through end-stage
renal disease. Qual Life Res V12:103-115, 2003

327.

Valdes C, Garcia-Mendoza M, Rebollo P, Ortega T, and Ortega F: Mental health at the third
month of haemodialysis as a predictor of short-term survival. Nephrol Dial Transplant 21:32233230, 2006

328.

Williams AW, Chebrolu SB, Ing TS, Ting G, Blagg CR, Twardowski ZJ, et al.: Early clinical,
quality-of-life, and biochemical changes of "daily hemodialysis" (6 dialyses per week). Am J
Kidney Dis 43:90-102, 2004

329.

Lo C, Li L, Lo W, Chan M, So E, Tang S, et al.: Benefits of exercise training in patients on

continuous ambulatory peritoneal dialysis. Am J Kidney Dis 32:1011-8, 1998

- 199 -

330.

Sloan RS, Kastan B, Rice SI, Sallee CW, Yuenger NJ, Smith B, et al.: Quality of life during
and between hemodialysis treatments: role of L-carnitine supplementation. Am J Kidney Dis
32:265-72, 1998

331.

Guo A, Wolfson M, and Holt R: Early quality of life benefits of icodextrin in peritoneal dialysis.
Kidney Int 62:S72-S79, 2002

332.

Bauer JD and Capra S: Nutrition intervention improves outcomes in patients with cancer
cachexia receiving chemotherapy--a pilot study. Supportive Care In Cancer: Official Journal Of
The Multinational Association Of Supportive Care In Cancer 13:270-274, 2005

333.

Isenring EA, Capra S, and Bauer J: Improved outcomes with nutrition support in ambulatory
oncology patients receiving radiotherapy. Clin Nutr 222003

334.

Earthman CP, Reid PM, Harper IT, Ravussin E, and Howell WH: Body cell mass repletion and
improved quality of life in HIV-infected individuals receiving oxandrolone. JPEN. Journal Of
Parenteral And Enteral Nutrition 26:357-365, 2002

335.

Beattie AH, Prach AT, Baxter JP, and Pennington CR: A randomised controlled trial evaluating
the use of enteral nutritional supplements postoperatively in malnourished surgical patients.
Gut 46:813-818, 2000

336.

Ravasco P, Monteiro-Grillo I, and Camilo ME: Does nutrition influence quality of life in cancer
patients undergoing radiotherapy? Radiotherapy And Oncology: Journal Of The European
Society For Therapeutic Radiology And Oncology 67:213-220, 2003

337.

Davidson W, Ash S, Capra S, and Bauer J: Weight stabilisation is associated with improved
survival duration and quality of life in unresectable pancreatic cancer. Clin Nutr 23:239-247,
2004

338.

Kuehneman T, Saulsbury D, Splett P, and Chapman DB: Demonstrating the impact of nutrition
intervention in a heart failure program. Journal Of The American Dietetic Association
102:1790-1794, 2002

339.

Glasgow R, Magid D, Beck A, Ritzwoller D, and Estabrooks P: Practical clinical trials for
translating research to practice: design and measurement recommendations. Med Care
43:551-7, 2005

340.

Tunis SR, Stryer DB, and Clancy CM: Practical Clinical Trials: Increasing the Value of Clinical
Research for Decision Making in Clinical and Health Policy. JAMA 290:1624-1632, 2003

- 200 -

341.

References S, Jeffery R, and French S: Preventing weight gain in adults: the pound of
prevention study. Am J Public Health 89:747-751, 1999

342.

Schwenger V, Morath C, Hofmann A, Hoffmann O, Zeier M, and Ritz E: Late referral - a major
cause of poor outcome in the very elderly dialysis patient. Nephrol Dial Transplant 21:962967, 2006

343.

Jansen M, Korevaar J, Dekker F, Jager K, Boeschoten E, and RT K: Renal function and

nutritional status at the start of chronic dialysis treatment. J Am Soc Nephrol 12:157-63, 2001

344.

ANZDATA, Chapter 2 - New patients commencing treatment in 2004, in Australia and New
Zealand Dialysis and Transplant Registry 28th Annual Report, E.L. McDonald S., Shtangey V.,
Editor. 2005.

345.

Hayes-St Clair S CK, Ash S, Bauer J: Long-term outcomes following a randomised-controlled

trial of nutrition intervention in chronic kidney diesase. Nutr Diet 64:S19, 2007

- 201 -

10.0 Appendices

Appendix A Manuscripts and conference presentations related to the thesis

Appendix B Evidence Based Practice Guidelines for the Nutritional Management of
CKD
Appendix C Development of Body Cell Mass Index
Appendix D Ethics clearance and patient information package
Appendix E Data collection and nutrition assessment tools
Appendix F Intervention support materials

- 202 -

Manuscript One: A critical review of nutrition assessment

tools to measure malnutrition in chronic kidney disease
Exploring validity and reliability testing of subjective global assessment and related
tools for the stages of chronic kidney disease

Citation:

Campbell KL, Ash S, Bauer J, Davies PSW, A critical review of the use

of nutrition assessment tools in chronic kidney disease. Nutr Diet 64:23-30, 2007

Date Submitted:

January 2006

Date Accepted:

August 2006

Corresponding Author: Katrina L Campbell, Institute of Health and Biomedical

Innovation, Queensland University of Technology, GPO Box 2434, Brisbane, QLD
4001, Australia
Contribution of Authors:
KLC was the main author of the manuscript, initiated the study, collected and analysed the literature.
SA, JB and PSWD initiated the study, assisted in interpretation of the literature and writing of the
manuscript.

Appendix A: Manuscripts and conference abstracts relating to thesis

Manuscript Two:

Evaluation of nutrition assessment tools

compared with body cell mass for the assessment of

malnutrition in Stage IV chronic kidney disease

Citation: Campbell KL, Ash S, Bauer J, Davies PSW, Evaluation of nutrition

assessment tools compared with body cell mass for the assessment of
malnutrition in Stage IV chronic kidney disease. J Ren Nutr 17:189-195, 2007

Date Submitted:

March 2006

Date Accepted:

October 2006

Corresponding Author: Katrina L Campbell, Institute of Health and Biomedical

Innovation, Queensland University of Technology, GPO Box 2434, Brisbane, QLD
4001, Australia

Contribution of Authors:
KLC was the main author of the manuscript, initiated the study, collected and
analysed the literature.

SA, JB and PSWD iniated the study, assisted in

interpretation of the literature and writing of the manuscript.

Appendix A: Manuscripts and conference abstracts relating to thesis

The impact of nutrition support on body composition: a randomized controlled

trial in pre-dialysis chronic kidney disease
Accpeted for presentation European Renal Association Conference, Barcelona, June
2007
Campbell KL1,2, Ash S1,2, Bauer J3, Davies PSW4
1 Queensland University of Technology, Australia
2 Royal Brisbane and Womens Hospital, Australia
3 The Wesley Research Institute, Australia
4 University of Queensland, Australia
INTRODUCTION AND AIMS: Patients with chronic kidney disease (CKD) are at high risk of body
protein loss due to mechanisms related to uremic malnutrition. With progressive reduction in renal
function, symptoms such as anorexia and subsequent reduction in dietary intake contribute to an
increased risk of malnutrition. Despite this, there is little evidence informing the impact of nutrition
intervention on body composition of pre-dialysis CKD (Stage 4 and 5). In particular, common methods
to assess body composition at this stage of CKD have limitations due to the assumption of static body
fluid status. Therefore, this study investigated the change in body composition, as measured by total
body potassium (TBK), to determine body cell mass (BCM), a direct measure of body protein stores in
41 pre-dialysis patients (Males 62%, Age (mean(SD)) 49.5 (12.3) years, GFR 22.3 (5.8)).

METHODS: Subjects were randomized to receive either standard care (n=20), provided by group
education and written material, or intervention (n=21), consisting of individualized nutrition counseling
with regular follow-up by a dietitian. There were no significant differences in baseline characteristics
between the groups. BCM was calculated via a measurement of TBK at baseline and three months
later.

RESULTS: The standard care group lost significantly more BCM (-3.3% (-6.9 to 0.4)) than the
intervention group (0.6% (-2.9 to 4.1)) over the treatment period. The overall mean difference in
change between the treatment groups was 3.9% (-1.0 to 8.7) or 1.2kg (-0.3 to 2.7) of BCM. Converted
to fat-free mass, this change is equivalent to 2.4kg (range 1.9 to 2.7kg). Interestingly, the mean
difference in BCM change between treatment groups was the most significant for females at 9.9% (0.1
to 19.6%), or 2.4kg (-0.1 to 4.8kg).

CONCLUSIONS: Individualized dietetic intervention with regular follow-up provides beneficial

outcomes in terms of minimizing loss of functional, metabolically active tissue (BCM), when compared
with standard care treatment for patients in pre-dialysis

Appendix A: Manuscripts and conference abstracts relating to thesis

Randomised-controlled trial of structured nutrition intervention in chronic

kidney disease
Accepted for presentation at 25th National Conference for the Dietitians Association
of Australia, May 2007

Campbell KL1,2, Ash S1,2, Bauer J3, Davies PSW4

1 Queensland University of Technology, Australia
2 Royal Brisbane and Womens Hospital, Australia
3 The Wesley Research Institute, Australia
4 University of Queensland, Australia

There is a high prevalence of malnutrition in patients at the onset of renal dialysis. Despite this, there
is little evidence informing the delivery of nutrition intervention in pre-dialysis chronic kidney disease
(CKD, stage 4 and 5). To determine the impact of intervention in pre-dialysis CKD, 56 patients (Male
62%; age mean (SD) 70.7 (14.0) years) were randomly allocated to receive either individual
counselling with fortnightly telephone follow-up (n=29), or standard care (written material only, n=27).
Intervention aimed to attain evidence-based dietary prescription (energy >125kJ/kg and protein 0.81.0g/kg) whilst managing symptoms of reduced kidney function. Outcome measures included change
in body cell mass (by gold-standard total body potassium counting), dietary intake (3-day food record)
and quality of life (KDQOLTM). At 12 weeks, there was a clinically significant improvement in all
outcome measures in the intervention group. Between treatment groups, the adjusted mean
difference in change in body cell mass was 3.9% (95%CI -1.0 to 8.7%, p=0.114) and energy intake
22.1 kJ/kg (12.8 to 31.4kJ/kg, p<0.001). Quality of life improved significantly in 10 of the 18 subscales in the intervention group. There was a statistically significant mean difference in change
between groups for: coping with the symptoms of kidney disease (7.1 (0.1 to 14.1) p=0.047); cognitive
functioning (14.6 (5.4 to 23.7) p=0.003); and feelings of energy/vitality (12.0 (4.6 to 19.5) p=0.002)
favouring intervention treatment. Structured nutrition intervention provides beneficial patient outcomes
including limiting the deterioration in nutritional status, improving dietary intake and improving quality
of life in patients with pre-dialysis CKD.

Appendix A: Manuscripts and conference abstracts relating to thesis

Performance of assessment tools to measure nutrition status in pre-dialysis

Presented at XIII International Congress on Metabolism and Nutrition in Renal
Disease, Mrida, Yucatn, Mxico February 28 March 4, 2006
Campbell KL1,2, Ash S1,2, Bauer J3, Davies PSW4
1 Queensland University of Technology, Australia
2 Royal Brisbane and Womens Hospital, Australia
3 The Wesley Research Institute, Australia
4 University of Queensland, Australia
There is a growing variety of nutrition assessment tools in chronic kidney disease (CKD). This study
will investigate a number of available nutrition assessment tools for their performance in assessing the
nutrition status of Stage IV CKD patients.
Fifty consecutive consenting CKD patients (Male 61%; (mean (SD)) age 70.7 (11.9) years;
GFRMDRD 22.4 (6.5)mls/min) attending a pre-dialysis outpatient clinic were recruited. Crosssectional assessment of total body potassium (TBK), a gold-standard measure of body cell mass
(BCM) was undertaken. Blinded to the TBK results, the dietitian assessment included the original
Subjective Global Assessment (SGA); 7-point SGA; Patient Generated-SGA (PG-SGA, a scored
assessment tool) and Malnutrition Inflammation Score (MIS, a scored tool including objective
measures, adjusted for pre-dialysis).
TBK was significantly related with the original SGA (92.520.3 vs. 74.618.0 p<0.02), and approached
statistical significance with 7-point SGA (ANOVA f(1)=3.54 (p=0.067)), MIS (Pearson r=-0.26
(p=0.077)) and PG-SGA (Spearman r=-0.24 (p=0.093)). Clinical and statical significance for
distribution of scores for SGA A and B ratings was noted for all SGA-based tools as shown below.
Table 1: Comparison of SGA-based tools
SGA-derived tools
SGA A (80%) n = 40

SGA B (20%) n = 10

MIS

Median (range)*

4.0 (0-7)

9.0 (5-12)

PG-SGA Median (range)*

3.0 (0-8)

9.0 (4-15)

7-pt SGA Score 7-1 (%)**

(7)32%;(6)45%;(5) 4%
* p<0.0001 by t-test; **p<0.0001 by chi-square

(5)11%;(4) 5%;(3) 4%

In this sample, the original SGA best detected depleted BCM. The MIS and PG-SGA appear to be
measuring nutritional status (by SGA and TBK) to a similar degree. Further prospective investigation
of the new assessment tools is warranted as potential measures to indicate change in nutrition status
and/or predict outcome in pre-dialysis patients.

Appendix A: Manuscripts and conference abstracts relating to thesis

Estimation of body cell mass in pre-dialysis - agreement between two

prediction equations and total body potassium
Presented at XIII International Congress on Metabolism and Nutrition in Renal
Disease, Mrida, Yucatn, Mxico February 28 March 4, 2006
Campbell KL1,2, Ash S1,2, Bauer J3, Davies PSW4
1 Queensland University of Technology, Australia
2 Royal Brisbane and Womens Hospital, Australia
3 The Wesley Research Institute, Australia
4 University of Queensland, Australia
Optimising body cell mass is an important outcome of nutrition therapy. This investigation aims to
establish the feasibility of using healthy population-based equations to predict body cell mass (BCM)
in CKD. The gold-standard measure of BCM (total body potassium, BCMTBK) will be compared to
BCM derived from 2 TBK prediction equations (BCMELLIS (1) and BCMLARRSON (2)) in Stage IV
CKD patients.
Fifty consecutive patients attending a pre-dialysis clinic were recruited (Male 61%; age
70.0(11.9)years; GFRMDRD 22.6(6.6)mls/min). T-test was used to assess the differences between
mean BCMTBK and prediction equations. Bland-Altman analysis used to assess the agreement
between prediction equations relative to BCMTBK.
Mean(SD) BCM results were 28.9(6.8)kg (BCMTBK), 30.9(7.1)kg (BCMELLIS), 24.4 (9.6)kg
(BCMLARRSON). There was a significant difference between mean BCMTBK and BCMELLIS
(t49=3.659 p<0.001) and BCMLARRSON (t49= -3.264 p<0.002).
Table 1: Bias and limits of agreement of 2 methods predicting
by TBK in Stage IV CKD
Equation
Bias (kg)
Limits of agreement
(2 SD) kg
BCMELLIS
3.98
8.60
BCMLARRSON
- 4.53
10.94

BCM in comparison to measured BCM

Correlation of
mean difference
-0.238 (p=0.097)
-0.531 (p<0.001)

Both prediction equations had clinically significant mean bias and wide limits of agreement. The mean
difference for both equations was inversely correlated and reached statistical significance for
BCMLARRSON, meaning the bias was inconsistant across the range of BCM measured.
At both a population and individual level, these data indicate the use of BCM prediction equations are
not reliable for the assessment of BCM in Stage IV CKD.
(1) Ellis K.J., Shulka K.K., et al. J Lab Clin Med 83:716, 1974
(2) Larrson I, Lindroos A.K., et al AJP Endo 284:416, 2003

Appendix A: Manuscripts and conference abstracts relating to thesis

Rating methods of subjective global assessment in chronic kidney disease

Presented at 24th National Conference for the Dietitians Association of Australia,
Sydney, May 2006

Katrina L Campbell1,2, Susan Ash1,2, Judith Bauer3, Peter SW Davies4

1 Queensland University of Technology, Brisbane QLD 4006 Australia
2 Royal Brisbane and Womens Hospital, Brisbane QLD 4059 Australia
3 The Wesley Research Institute, Brisbane QLD 4066 Australia
4 University of Queensland, Brisbane QLD 4029 Australia
Guideline recommendations and supporting evidence for the rating of Subjective Global Assessment
(SGA) is inconsistent in chronic kidney disease (CKD) literature. This investigation aims to determine
the validity of the original ABC tool and a modified 7-point scale for rating the SGA in pre-dialysis CKD
patients.
Forty-six consecutive consenting patients (Male 61.5%; age 69.5 (11.7) years; glomerular filtration
rate (GFR) 21.4 (6.6) ml/min) attending a specialist pre-dialysis outpatient clinic were recruited.
Assessment included the original SGA, 7-point SGA, anthropometry and total body potassium (TBK),
a gold-standard measure of body cell mass, used to derive fat-free mass (FFM). Statistical
association for both tools compared to the objective measures were undertaken using SPSS (Version
13).
The original SGA rated 17.3% (n=8) moderately malnourished (SGA B; no C ratings), with significant
differences between A and B categories and body-mass index (meanstandard deviation;
27.74.6kg/m2 to 24.34.8kg/m2), weight change (6 months; -0.22.3% to -5.44.6%) (p<0.05), and
strong association for FFM (47.810.5kg to 36.37.4kg) (p<0.0001). The 7-point SGA by one-way
weighted ANOVA was related to weight change (6 months; p<0.002) only. The relationship between
7-point SGA and FFM was not significant (p = 0.125), indicating FFM was not related to the seven
categories in a systematic way.
These data support the validity of the original SGA to detect malnutrition in CKD. The 7-point SGA did
not accurately delineate nutrition status of individuals on the basis of FFM. This investigation does not
support the validity of a 7-point rating system to determine nutrition status in CKD.

Appendix A: Manuscripts and conference abstracts relating to thesis

Long-term outcomes following a randomised-controlled trial of nutrition

intervention in chronic kidney disease
Presented at 25th National Conference for the Dietitians Association of Australia,
Hobart, May 2007
Samantha Hayes-St Clair1, Katrina L Campbell1,2, Susan Ash1,2, Judith Bauer3
1 Queensland University of Technology, Brisbane QLD 4006 Australia
2 Royal Brisbane and Womens Hospital, Brisbane QLD 4059 Australia
3 The Wesley Research Institute, Brisbane QLD 4066 Australia
Poor nutritional status on commencement of renal dialysis is associated with increased morbidity and
mortality. This study aims to investigate long-term outcomes following completion of a 12-week
randomised-control trial of nutrition intervention in pre-dialysis chronic kidney disease. Fifty-six
participants (Male 62%; age meanSD 70.714.0 years; glomerular filtration rate (GFR) 22.16.9
ml/min) were allocated to either intervention (individual counselling with fortnightly follow-up, n=29) or
control (written material only, n=27) groups. Six participants were lost to follow-up during the treatment
period and were therefore not considered for this investigation. Of the participants who completed the
treatment, 41.6% (10/24) of the control group and 19.2% (5/26) of the intervention group commenced
dialysis or died during the follow-up period of 13.44.1 months (range 9 to 22 months) (p=0.14). There
was no difference in hospitalisations (bed days per hundred, intervention 1.33.4, control 1.62.5 days
p=0.74) between the groups during the follow-up period. From values collected at the end of the
intervention period, lower serum albumin (35.66.0 vs 39.04.3 g/L p=0.04) and GFR (16.56.8 vs
23.66.2 ml/min p<0.01) was associated with dialysis commencement and lower dietary energy
(85.912.7 vs 108.923.2 kJ/kg/day p=0.06) and protein (0.840.29 vs 1.070.21 g/kg/day p<0.05)
intake was evident in participants who died during the follow-up period. These results indicate those
who completed the nutrition intervention were less likely to have progressed to dialysis or die in the
follow-up period, compared with the control group. Poor dietary intake may be a surrogate factor
associated with illness relating to death in pre-dialysis patients.

Appendix A: Manuscripts and conference abstracts relating to thesis

Randomised-controlled trial of structured nutrition intervention in chronic

kidney disease
Presented at 25th National Conference for the Dietitians Association of Australia,
Hobart, May 2007

Katrina L Campbell1,2, Susan Ash1,2, Judith Bauer3, Peter SW Davies4

1 Queensland University of Technology, Brisbane QLD 4006 Australia
2 Royal Brisbane and Womens Hospital, Brisbane QLD 4059 Australia
3 The Wesley Research Institute, Brisbane QLD 4066 Australia
4 University of Queensland, Brisbane QLD 4029 Australia
There is a high prevalence of malnutrition in patients at the onset of renal dialysis. Despite this, there
is little evidence informing the delivery of nutrition intervention in pre-dialysis chronic kidney disease
(CKD, stage 4 and 5). To determine the impact of intervention in pre-dialysis CKD, 56 patients (Male
62%; age mean (SD) 70.7 (14.0) years) were randomly allocated to receive either individual
counselling with fortnightly telephone follow-up (n=29), or standard care (written material only, n=27).
Intervention aimed to attain evidence-based dietary prescription (energy >125kJ/kg and protein 0.81.0g/kg) whilst managing symptoms of reduced kidney function. Outcome measures included change
in body cell mass (by gold-standard total body potassium counting), dietary intake (3-day food record)
and quality of life (KDQOLTM). At 12 weeks, there was a clinically significant improvement in all
outcome measures in the intervention group. Between treatment groups, the adjusted mean
difference in change in body cell mass was 3.9% (95%CI -1.0 to 8.7%, p=0.114) and energy intake
22.1 kJ/kg (12.8 to 31.4kJ/kg, p<0.001). Quality of life improved significantly in 10 of the 18 subscales in the intervention group. There was a statistically significant mean difference in change
between groups for: coping with the symptoms of kidney disease (7.1 (0.1 to 14.1) p=0.047); cognitive
functioning (14.6 (5.4 to 23.7) p=0.003); and feelings of energy/vitality (12.0 (4.6 to 19.5) p=0.002)
favouring intervention treatment. Structured nutrition intervention provides beneficial patient outcomes
including limiting the deterioration in nutritional status, improving dietary intake and improving quality
of life in patients with pre-dialysis CKD.

Appendix A: Manuscripts and conference abstracts relating to thesis

Appendix B Evidence Based Practice Guidelines for the Nutritional Management of

CKD

SEE MANUSCRIPT:
Ash S, Campbell KL, MacLaughlin H, McCoy E, Chan M, Anderson K, et al.: Evidence Based Practice
Guidelines for Nutritional Management of Chronic Kidney Disease. Nutr Diet 63:S35-S45, 2006

Appendix A: Manuscripts and conference abstracts relating to thesis

APPENDIX C: Development of Body Cell Mass indexed for height

In the literature, body cell mass index (BCM-I) has been recommended for use over absolute body cell
mass (BCM) as it provides an assessment of metabolically active tissue independent of body
size/height. This occurs for the same reason that body mass is commonly indexed for height (BMI
kg/m2) is used in place of weight (kg) when defining the cross-sectional weight status of populations.
From the baseline data of the pre-dialysis sample in this thesis, the regression equation and graph in
Figure 1 indicate a strong and significant relationship between BCM and height (m) (r=0.775 p<0.001).
This justifies the need to index BCM for height to provide an independent assessment of BCM.

Linear Regres sion

Baseline BCM (kg)

5 0.0 0

Baseline BCM (kg) = -82.77 + 69.35 * HGHTB_M

R-Square = 0.60
4 0.0 0

3 0.0 0

2 0.0 0

1 .50

1 .60

1 .70

1 .80

Height (m)

Figure 1: Body cell mass (kg) by height (m)

A number of studies to date utilising BCM-I as a variable have used the simple equation of BCM-I =
2
1-4
BCM (kg) / Height (m) to provide the assessment of body cell mass independent of height . The
2
relationship between BCM-I as kg/m and height (m) was further evaluated from the samples
baseline data, with the graph and equation in Figure 2. This relationship was also strong and
significant (r=0.503 p<0.001), indicating that by indexing BCM (kg) to height to the power of 2 does not
provide a BCM-I which is independent of height.

Linear Regres sion

BCM_Index_B

1 6.0 0

1 4.0 0

BCM_Index_B = -8.06 + 11.86 * HGHTB_M

R-Square = 0.25
1 2.0 0

1 0.0 0

8 .00

1 .50

1 .60

1 .70

1 .80

Height (m)

Appendix C: Development of Body Cell Mass Index

Figure 2: Body cell mass index as kg per m2 by height (m)

Therefore, this calculation is taken further, similar to the method adopted by White et al (2005),
5
undertaking a Log-Log regression of BCM and Height . This regression provides the following
equation: Ln(BCM) = 1.54 + 3.78 * Ln(Height).

4 .00

Linear Regre s sion

Ln_BCMB = 1.54 + 3.78 * Ln_HGHT_m
R-Square = 0.61

Ln_BCMB

3 .75

3 .50

3 .25

3 .00

0 .40

0 .45

0 .50

0 .55

0 .60

Ln_HGHT_m

Figure 3: Regression of Log Body Cell Mass by Log Height

The value of interest is the co-efficient of 3.78 (standard error = 0.43). This is then used to modify the
previous BCM-I equation to equal BCM (kg) / height raised to the power of 3.78. The relationship
3.78
between this BCM-I (kg/m ) is provided in Figure 4.

Linear Reg re s sion

BCM_HGT_EXP1

6 .00

5 .00

BCM_HGT_EXP1 = 4.72 + -0.01 * HGHTB_M

R-Square = 0.00
4 .00

1 .50

1 .60

1 .70

1 .80

Height (m)

Figure 4: Relationship between BCM-I (kg/m

3.78

) and height (m)

3.78

In figure 4 it is clear that there is no relationship between BCM-I (kg/m ) and height (r=-0.001
p=0.993). Therefore, use of this equation for BCM-I enables the successful elimination of any
influence that body size/height has on the estimate of BCM.
However, for convenience the equation was modified to raise height to the power of 3.5, instead of the
original co-efficient from the Log-log regression 3.78. The value of 3.5 provides a numerically
convenient and statistically valid alternative (within 2 standard errors of original co-efficient). When

Appendix C: Development of Body Cell Mass Index

3.5

assessing this final BCM-I (kg/m ) against height, the values remain unrelated r=0.01, p=0.89, see
3.5
Figure 5. Therefore, for the purposes of this thesis, BCM-I were calculated as BCM/Height

Linear Regres sion

BCM_HGHT_PW R3.5

7 .00

6 .00

BCM_HGHT_PWR3.5 = 3.98 + 0.87 * HGHTB_M

R-Square = 0.01
5 .00

4 .00

1 .50

1 .60

1 .70

1 .80

Height (m)

Figure 5: Body cell mass index (kg/m3.5) by height (m)

References:
1.
Talluri, A., R. Liedtke, E.I. Mohamed, et al., The application of body cell mass index for
studying muscle mass changes in health and disease conditions. Acta Diabetol, 2003. 40(1):
p. S286-9.
2.
Talluri, T., Qualitative human body composition analysis assessed with bioelectrical
impedance. Coll Antropol, 1998. 22(2): p. 427-32.
3.
Langkamp-Henken, B., J. Hudgens, J. Stechmiller, et al., Mini nutritional assessment and
screening scores are associated with nutritional indicators in elderly people with pressure
ulcers. J Am Diet Assoc, 2005. 105(10): p. 1590-6.
4.
Bergman, P. and G. Hauser, Biosocial and nutritional effects on body composition in young
adults from Wroclaw, Poland. Journal of Biosocial Science, 2005. 38(06): p. 721-734.
5.
White, M., A.J. Murphy, Y. Hastings, et al., Nutritional status and energy expenditure in
children pre-bone-marrow-transplant. Bone Marrow Transplantation, 2005. 35(8): p. 775-779.

Appendix C: Development of Body Cell Mass Index

Appendix D Ethics clearance and patient information package

Patient Information Sheet

Intensive Nutrition Intervention in Chronic Kidney Disease
You are invited to help with this research study because your kidney disease may put
your nutrition at risk.
This sheet explains what is involved. Please read it carefully and take your time
before deciding whether to take part. Please discuss anything that you dont
understand or that is not clear to you with your doctor.
Why is the research being done?
Many patients who have chronic kidney disease are undernourished. We know that
being undernourished can worsen your condition and you are then more likely to
have other health problems compared to those who are well nourished.
This study will help us to develop a quick, convenient and reliable way of checking
nutritional status for the future. It will also help to decide what the best way is to
ensure that your nutrition status is at its best before dialysis starts.

What does the study involve?

This study will be conducted at the Royal Brisbane and Womens Hospital (RBWH). It
involves nutritional assessments and consultations with the research officer.
Firstly, if you consent to the study, you will be asked to complete a survey about how
you are feeling and coping with daily activities. You will then be randomised into
either a group that gets a fortnightly follow-up with a dietitian (by phone and in person
at the RBWH Renal Department) for three months, OR a group that receives
standard care treatment, which is follow-up in three months. The dietitian will provide
you with a booklet and give you instructions to complete a record of the food you eat
for three days.
An appointment will then be made for a number of tests to be conducted at the Royal
Childrens Hospital, by the research officer. This appointment will last about 1
hours. You are required to bring the completed food record with you to your
appointment.

Appendix D: Ethics information

At this appointment, you will be asked questions relating to your nutrition, including
questions about your weight, dietary intake, medical symptoms and activities of daily
living. The assessment also involves a physical examination of fat and muscle
stores.
Then you will also be asked to undergo a test to screen your muscle stores. This
involves a machine where you lie still for 40 mins while it passes over your body. This
machine is only available for use at the body composition laboratory at the Royal
Childrens Hospital, therefore, all of your tests will occur here.
Three months later:
The assessment process is repeated again, you will be provided with an appointment
to the Royal Childrens Hospital, and:

complete another record of your food intake for three days

have a repeat nutritional assessment
complete a survey about how you are feeling and coping with daily activities.
be tested with a machine that can count your active body mass (muscle stores).
Again, you will need to lie still on a bed for 40 minutes and this will be a total of a
one-hour visit.

How will taking part in the study help me?

You will have been asked to take part in the study because you have chronic kidney
disease and may be at risk of being undernourished.
By taking part in the study, you will receive nutrition information to manage your
kidney disease and have the opportunity to receive a full nutrition assessment. You
will be monitored over three months to check your nutritional status does not become
worsened.
In addition to the benefits you personally may receive by taking part, you may be
helping all other patients with kidney disease in the future.
Are there any risks or discomforts if I decide to take part?
The tests and measurements are not harmful in any way.
The physical examination of muscle and fat stores involves the research officer
observing certain sites on your body and may require you to remove any bulky items
of clothing (i.e. pullover/jumper) and uncovering your legs. It does not involve
invasive procedures and is commonly performed by the dietitians on all renal
patients.
The machine that measures you bodys muscle mass is simple and painless, but
involves lying down still for about 40 minutes while detectors move from over the
bed. You will not feel anything. The only discomfort may be the length of time
required to lie still.

Appendix D: Ethics information

Any new information that comes to light that may affect your decision to continue in
the study will be made known to you.
Do I have to take part in the trial?
Participating in this trial is voluntary. If you chose not to participate in this study, your
usual care will not be affected. You will be treated under the regular care of the
dietitian at your pre-dialysis clinic, which involves a minimum of 6 monthly follow-up
appointments.
This research is conducted as part of a PhD study. Therefore, the results of this are
expected to be published. If you consent, you may request to be provided with a copy
of the groups results. If you are interested, please tick the box provided on the
consent form.
Confidentiality
Your identity will always be treated as confidential and will not be disclosed to the
public. All information will be assessed without your name or address on them.
The doctors looking after your care in the hospital will be told you are taking part in
the study. Your hospital medical record will state that you are in the study. This is
done to make sure that all people involved in your care are aware of the treatment
you are receiving. It will not change your care in any way if you chose not to take
part.
Before starting in the study, you will be asked to sign a consent form. This is so that
other people know that we have explained the study to you and that you are doing it
of your own free will.
By signing the form you are not waiving any of your legal rights.
If at any time you feel you do not wish to continue, you may withdraw from the study.
No questions will be asked. This will not affect any future care from your doctor or the
hospital in any way.
Your doctor may withdraw you from the study if it is felt that it is not right for you to
carry on for some reason. This would be done with your permission.
You may contact Dr Susan Ash (Phone 3636 7997 & Page 40509), Katrina Campbell
(Phone 3864 9619) or a member of the ethics committee (ph 3636 5490) about
anything that might concern you in the study.
Thank you for your assistance
Appendix D: Ethics information

Dr Susan Ash
Research Co-ordinator

Katrina Campbell
Research Dietitian

Department of Nutrition and Dietetics

Royal Brisbane and Womens Hospital

This study has been reviewed and approved by the Royal Brisbane Hospital and Health Service
District Human Research Ethics Committee. Should you wish to discuss this study with someone not
directly involved, particularly in relation to matters concerning policies, information about the conduct
of the study and your rights as a participant, or should you wish to make an independent complaint,
you can contact the Coordinator or Chairperson, Human Research Ethics Committee, Royal Brisbane
Hospital, Herston, Qld or telephone (07) 3636 5490, and/or the Research Ethics Officer, Queensland
University of Technology on 3864 2340 or ethicscontact@qut.edu.au.
This study is being partially funded by the Royal Brisbane and Womans Hospital Foundation

Appendix D: Ethics information

Appendix E Data collection and nutrition assessment tools

Subjective Global Assessment
SUMMARY SCORECARD
Weight

Height

Weight change
7-point SGA
(6 month)
7 5%
6 5-7% improving
5 5-7% stable
4 5-10% improving
3 7-10%
2 > 10% stablising
1 >10% and continuing

BMI

Weight 2/52
/-/

MIS
(3-6 month)
0, none - <0.5kg
1, minor 0.5-1kg
2, >1kg but <5%
3, >5%

Weight 1/12

Weight 6/12
%

PG-SGA
(6 mth; 1mth) + 1 one if
dec in last 2wk)
0, 0-1.9%
1, 2-5.9%; 2-2.9%
2, 6-9.9%, 3-4.9%
3, 10-19.9%; 5-9.9%
4, >20%; >10%

Appetite
%

Regular SGA

A 0-5%
B 5-10%
C >10%

Dietary intake
MIS
0
Good appetite, no deterioration

PG-SGA
0
Unchanged / more than usual

1
Somewhat sub-opt solid intake

1
Less than normal

2
Moderate decrease up to full liquid

1+1
normal food, less than normal amount

3
Hypo-caloric liq to total starvation

1+2
Little solid food
1+3
only liquids or supps
1+4
very little of anything

Regular SGA
A
No change/adequate; intake
borderline/increasing
B
No change and inadequate; Change
and suboptimal/full liquid, borderline
no change, poor and no change/INC
C
Change to hypocaloric/starvation,
poor and decreasing

Gastrointestinal symptoms (anorexia, nausea, vomiting, and diarrhoea).

7-point SGA
7
no symptoms, or they are
uncommon
6
if few symptoms but
uncommon
5
one or more symptom, but
not every day
4
>1 symptom, almost daily
3
most symptoms are present,
almost daily
2
if all symptoms are present,
almost daily
1
if all symptoms are present,
daily

MIS
0
No symptoms with good
appetite
1
Mild Symptoms, poor
appetite or nausea
occasionally
2
Moderate GI symptoms or
occasional vomiting
3
Frequent diahorrea or
vomiting or anorexia

PG-SGA (additive)
0
No problems eating
1
Nausea
Constipation
Smells bother me
Dry mouth
Feel full quickly
Taste changes
2
Mouth sores
Swallowing problems

3
No appetite
just dont feel like eating
Vomiting
Diahorrea
Pain (where)

Appendix E: Data collection and nutrition assessment tools

Regular SGA
A
No symptoms or
intermittent
B
Some of the symptoms,
daily for >2weeks
C
All of the symptoms
daily for >2 weeks

Functional Capacity
MIS
0
Normal to improved functional capacity
feeling fine
1
Occational difficulty in baseline
ambulation, or tired frequently
2
Difficulty with otherwise independent
activities (ADLs)
3
Bed/chair ridden, or little to no physical
activity

PG-SGA (past month)

0
Normal, no limitations
1
Not normal self, but able to be up to
normal activities
2
not feeling up to most things, in bed
and chair less than day
3
able to do little activity most of day in
bed/chair
4
bedridden, rarely out of bed

Regular SGA
A
No dysfunction or recent
improvement
B
Difficulty with ambulation/normal
activities, no change in 2 wk
C
Bed/chair ridden or function
regressed in 2 wk

Co morbidity (MIS only)

Major co morbid conditions (MCC) = CHF class 3 or 4, full blown AIDS, severe CAD, moderate to sever COPD,
major neurological, metastatic malignancies or Chemo
0 No major co morbid conditions
1 Mild co-morbidity (no MCC)
2 Moderate co-morbidity (including one MCC)
3 Multiple co-morbidity (including 2 or more MCC)
Physical examination Subcutaneous fat Below the eyes, skin above the triceps and biceps.
7-point SGA
MIS
PG-SGA
Regular SGA
7
0
0
A
little or no loss, all areas
No change
No deficit
No sign
6
1
1
B
small loss, some areas
mild
mild deficit
Mild to moderate
5
2
2
C
small loss, most areas
moderate
moderate deficit
Severe
4
3
3
small loss, all areas
severe
severe deficit
3
moderate loss, most areas
2
moderate loss, all areas
1
severe loss, most or all areas
Physical examination Muscle mass and wasting Temporalis muscle, clavicles, shoulders (rounded VS
squared), Visibility of Scapula, ribs, Interosseous and quadriceps muscle mass.
7-point SGA
7
little or no loss, all areas
6
small loss, some areas
5
small loss, most areas
4
small loss, all areas
3
moderate loss, most areas
2
moderate loss, all areas
1
severe loss, most or all areas

MIS
0
No change
1
mild
2
moderate
3
severe

PG-SGA
0
No deficit
1
mild deficit
2
moderate deficit
3
severe deficit

7-point

MIS

PG score

Appendix E: Data collection and nutrition assessment tools

Regular SGA
A
No sign
B
Mild to moderate
C
Severe

SGA rating

kJ (per Kg)

GFR (CG-MDRD)

PG score (0-35)
7-point (1-7)
MIS (0-30)
Pro (per Kg)
kJ (per Kg)

Height (0.5cm)

BMI

Weight Hx (1+6/12)

SGA

GFR (CG-MDRD)

Appendix E: Data collection and nutrition assessment tools

Appetite (1-4)

Weight (0.1kg)

Subjective Ax

Pro (per Kg)

SGA

IBW:

MIS (0-30)

Weight Hx (1+6/12)

Follow-up Data

7-point (1-7)

BMI

Diabetes
T1DM
T2-ins
T2-no ins
No

PG score (0-35)

Subjective Ax

Smoking
Current
Former
Unsure
Never

Height (0.5cm)

Appetite (1-4)

IBW:

Gender M
F

Age

Demographics

TBK2

TBK1

TBW

% BF

Body Comp

Objective Ax

TBK2

TBK1

TBW

% BF

Body Comp

Objective Ax

Cause CKD
Analgesic
T1DM
T2DM ins
T2DM no ins
Nephritis

Stable
intentional
unintentional

Wght Hx (6mth)

CRP

Alb

PO4

Crea

Urea

K+

CRP

Alb

PO4

Crea

Urea

K+

Misc
HTn
Polycystic
Reflux
Uncertain

Living
With Partner
With Others
Alone
Institution

SUMMARY ASSESSMENT FORM

Weight (0.1kg)

Baseline Data

Follow-up date

Treatment Arm

CVD
CAD
PVD
CLD

Baseline date

UR #

Medical History
PMHx

Consent date

ID #

Intensive Nutrition Intervention in CKD

ID#
Self
Others
Self
Others
Yes
No

PTH

Calcuim

TGs

Homocys

Cholest

HbA1c

PTH

Calcuim

TGs

Homocys

Cholest

HbA1c

Placed Bked Not yet

HD
PD
Not planned
Not for Dialysis

Dx Planning HD/PD

Employment

Shopping

Cooking

Total Pro

Fe Tx

Ferritin

Fe Sat %

FE

TIBC

Total Pro

Fe Tx

Ferritin

Fe Sat %

FE

TIBC

Medications
ACE-I
Other HT
Diuretic
Insulin
CA/PO4

Lip
S
__
O

(attac

Previous Diet Inter

CKD K+ Othe
Dietitan
Doctor
Other

Pt

Appendix E: Data collection and nutrition assessment tools

Intensive Nutrition Intervention in CKD

ID#
SUMMARY ASSESSMENT FORM

Pt

Intervention Support Materials

Department of Nutrition and Dietetics

Ph: (07) 3636-7997

Your Guide to

Appendix F Intervention support materials

Intervention Support Materials

&,
&,
&-

2
+
3

Intervention Support Materials

&*

/&

&0

&/

&)

&&

$
(

$
$

"

'

&

"

! "

"

Intervention Support Materials

!!

!$

"

"

& ,&

Intervention Support Materials

!
+

"

'
*

&

"

&

"

!%

!-

"

&

"

"

"

Intervention Support Materials

"

"

As your body breaks down foods containing protein,

it produces a waste called UREA. Urea can build up
in your blood and may cause:

'

1
$"

"

"

Intervention Support Materials

TIP:

"

"

"
"

"!

5,

%
3
2

42
43

&

"

"

'

1
9

1
"
"
"
1
1 8 "8 "

"

&

"

"

"

!6

Intervention Support Materials

7
7

'
-5

"

"

"

*4

"

"

"

"

"

"

"
"

"

"

.) )

"8
"
"
"
"

"8 "

"

-/0
%

"
"

"
"

"

)%10

034%

%2-3 %

.) )

Intervention Support Materials

. 4

"

%2-3 %

,
-

"

"

"
"

5 ;

Intervention Support Materials

"

"

"
!

'

'

"

"

"

"

"

* "

"

"

Intervention Support Materials

(
/

"
!

5%

"

"

"

"

"6+

!'

"

"

"

"

"

!'

"
"

!
!#

:
!

Intervention Support Materials

"

Low Potassium Cooking Hints:

Fruits and vegetables are some of the richest

sources of Potassium

'

"

"

Intervention Support Materials

"

"
"

"

"

"8

"

"
"
"

"0

8
8
"
"
"

&

"

'

"1

"
"

"

"

"
"

"

10

"
"

"
"

"
"

"
"

"

8
8

"

"

"

"

"

!
" &

!<
!

"

!<

"

"

!'

11 Intervention Support Materials

If necessary, ask your healthcare provider for

more information on Sodium, Fluid, Potassium
or Phosphate.

8
9
:

8
9
:

Intervention Support Materials

.
%

12

13 Intervention Support Materials

DINNER

Afternoon Tea

LUNCH

Morning Tea

BREAKFAST

!"#$% # "%$%"

>

Intervention Support Materials

_________________________

_________________________

_________________________

_________________________

! &'( "

_________________________

. "

(')$

Other Information:

&''
"*

14

. %
"

"

"

"

"

,%1

15 Intervention Support Materials

2
?

: "

34 3.) 3(4

"

"

.34,

'

,%1

&

"

Intervention Support Materials

2?
3

2?

. %

,(; 3.) 4
)2 4%

"
4

&

)3, /

16

"

, D

1 >

.
>

"

,
%1

" !,

"

2
4
B
+

3(%

17 Intervention Support Materials

3.

2
4
C

4
B

3)% <

. %
2

'
:

"+

"

. %

. %

"
8

"

,
%1

"

"

EEEEEEEEEEEE

,
%1

Intervention Support Materials

21

21

++
EEEEEEEEEEEE

0 2,
4

- 43'( %
+
+0

18

:
:
'

<

"

!0

19 Intervention Support Materials

6.5F'

'

'

'

1
1

"

=
1

"

G
H
H
G

.1

"

"

Intervention Support Materials

$"

,"

"

!1

20

!5

= 6

= 6

= 7

++

21 Intervention Support Materials

Adapted from the National Heart Foundation Making Every Step Count

! , 2
0

'

" "

'

"

)
D

EEEEEEEEEEEEEEEEEEE

Intervention Support Materials

&

EEEEEEEEEEEEEEEEEE

22

'

"

'

&

7
<

0
7

'

Intervention Support Materials

>

&

!7

"
&

"

"

"

"

"

!. "

"
"

"

"-

"

"

&
!

"

&

"

'

4 +

Intervention Support Materials

<
+

5
1

#
G
#

&

"

! <

1
2 &2
?

&

"

"

,"
M
*K

'
=

"

"

"

1
.G"

"

"

"

%
G

"

Intervention Support Materials

0""

"

'

. 8
'

"

N"

"
/

&

%<J 1
%0G
+

1
"

(
%

"

"

"

&?"

"

>

2
?

"

B1

"
B"

2
&2
!?D
-$<1 <
5
!

LO
*

N
"

%LH

"

;A

"

"

Intervention Support Materials

"

"

&

'-J

'
P2
!
C
'

>

-J
-J

+
"

"

LO
"

"LOP?!
?
'
<./"

'
LO

"

"

LO
(

" &
3
" PN!

)
0

LO

"

"-J

%.&'
"

>

>

>

-J