Recruitment: 

• 89% enrolled via one nephrologist (100% of practice)
• 11% enrolled via 4 other nephrologists (small percentage of each practice)

"Analyses showed no difference in results on the basis of treating nephrologist."

No mention of informed consent. 

Design:

1. Cross-sectional study:  Baseline data collected for analyses.

2. Nonrandomized subtrial:  Subgroup of subjects reassessed within 3 months of beginning dialysis and again 6-9 months from beginning dialysis.

Blinding used:  none mentioned - lab tests

Intervention:

1. Cross-sectional study: no intervention provided as part of this study, but all subjects were assessed to determine previous exposure to low protein diet instruction.

• Prestudy diet intervention included physicians often assessing protein intake by dietary recall and providing advise regarding high- and low- protein foods which included materials developed by the Australian Kidney Foundation.  Dietary compliance and reinforcement was given at 3 to 6 month intervals.  Patients were referred to a dietitian if dietary compliance was unsatisfactory.
• Dietitian intervention included measurement of BMI, assessment of protein and calorie intake (protein intakes 0.6-0.8 g/kg IBW/d and 35 kcal/kg IBW/d), nutrition counseling & follow-up in 4 wk to evaluate compliance.  Follow-up as recommended by patient, physician, dietitian.

2. Dialysis subgroup: received uniform diet counseling regarding adequate protein intake (1.2 g/kg/day for HD; 1.5 g/kg/day for CAPD) and kcal (35-45 kcal/day, with emphasis on PUFA and complex carbohydrates) at commencement of dialysis.  Diets were reviewed within one week of commencement of diaysis, again within one month, and then "as needed" or at regular 6 month intervals.  All dialysis diet counseling was received by a dietitian.

Statistical analysis: 

• Statview II (Abacus Concepts, Inc., Berkeley CA)
• Pearson correlation coefficient for correlation testing
• Univariate and multivariate analyses to determine independent effects of study variables
• Unpaired or paired t tests for comparisons between groups, or repeated comparisons, respectively
• ANOVA  to detect differences between multiple groups
• P<0.05 considered significant

Timing of measurements:

Baseline (total group):

• Biochemical data: a) blood: urea, creatinine, albumin, glucose, cholesterol, and trigylcerides.  b) 24-hour urine collection: urea, creatinine, protein; data used to determine DPI. 
• Anthropometrics: see below 
• Diet: subjects were assessed as to whether or not they had been advised about a low-protein diet and whether the instruction had been given by a dietitian.

Dialysis subgroup (at commencement of dialysis, thereafter "as needed" or regular 6 month follow up):

• food intake (total calories and protein requirements based on IBW, determined in this study to be BMI=25)
• biochemistry, including albumin
• anthropometrics
• total body nitrogen (to reflect protein stores) 

Dependent variables: 

• Estimated dietary protein intake (DPI).

• Methods to assess daily protein intake (DPI):

  1. predialysis (method of Meroni et al, 1985): g/d = 6.25 x [UUN (g/d) + ideal body weight (kg) x 0.031] + urinary protein losses
  2. post-HD (method of Borah et al, 1978): not quoted in article
  3. post-CAPD (method of Bergstrom et al, 1993): not quoted in article

Independent variables: 

• Serum creatinine
• Dietary modification advice

Control variables:

Initial N:

1. Baseline: N=766

2. Dialysis subgroup: N=52

Attrition (Final N):

• 3 subjects from dialysis subgroup died within 12 wk commencement of dialysis
• 49 dialysis subjects were reassessed at 6-11 wk commencement of treatment
• 39 dialysis subjects were reassessed at 6-9 months from commencement of treatment

Age:  54.4±0.6 (7-88) years

Ethnicity:  not mentioned

Other relevant demographics:

Anthropometrics:  

Location:  New South Wales, Australia

Baseline data was grouped by serum creatinine for comparison (<0.12 mmol/L and >0.12 mmol/L):

• Subjects with normal serum creatinine (<0.12mmol/L) without protein restriction had significantly higher DPI than those who had been advised to restrict protein (1.08±0.01 vs. 0.96±0.02 g/kg/d; P<0.01).
• Subjects with elevated serum creatinine advised to follow a low protein diet had a lower DPI than those that had not received diet advise (0.87±0.02 vs 0.93±0.01 g/kg/d; P<0.05).
• Subjects with higher serum creatinine were older, had lower BMI, and higher levels of plasma glucose and triglycerides.  Albumin was lower, but WNL.

For all subjects at baseline, DPI correlated most strongly with CrCl, independent of prestudy diet advise (r=0.50, P<0.001), BMI (r=0.23, P<0.0001), and serum albumin (r=0.14, P<0.0001).

Dialysis subgroup:  52 subjects commenced dialysis within 3 months of baseline assessment.  47 (or 49) were reassessed 6-11 weeks after dialysis.  Despite advice on increasing dietary protein after initiation of dialysis, there was not a significant change (0.79±0.04 vs. 0.82±0.03 g/kg/d; P=0.64); however, there was a significant increase by 6 to 9 months (1.04±0.04 g/kg/d vs. both prior measurements; P<0.005).

A low dietary protein intake in renal impairment occurs independently of dietary advice, but compliance with such advice is evident because patients advised to consume a low protein diet had significantly lower protein intake than patients receiving no dietary advice. Adaptation to a high protein intake after instigation of dialysis is unsuccessful in the short-term, irrespective of whether advice is given regarding a low protein diet before dialysis is initiated. However, 6 to 9 months after the commencement of dialysis, a significant increase in protein intake occurs which in the hemodialysis population correlates with dialysis delivery.