Authors' objectives

To examine the effects of restricting dietary protein on the rate of decline in renal function.

Searching

MEDLINE was searchd from 1980 to 1996, and the bibliographies of published reviews were examined.

Study selection

Assessment of study quality

Validity was assessed on the basis of the following criteria: publication as a peer-reviewed report; random allocation to treatment and control groups; use of a parallel control group design, rather than a sequential or other study design; assessment of dietary protein intake by measuring urea nitrogen accrual, rather than assuming adherence to prescribed diets or using diet recall methods; analysis of results using an intention to treat protocol. Each study was reviewed independently by two of the authors.

Data extraction

Each study was reviewed independently by two of the authors. In the case of duplicate publications, data were extracted from the most recent and/or the most complete report. The following data were extracted: the mean age; the proportion of males; and the proportion of patients with renal disease caused by glomerulonephritis, tubulointerstitial disease, autosomal dominant polycystic kidney disease, diabetes, hypertensive nephrosclerosis, or other causes. Baseline data (renal function, blood-pressure, level of urine protein excretion) and outcome data (rate of decline in renal function, protein intake, phosphorus intake) were also extracted.

For each randomised controlled trial (RCT), the difference in the rate of decline in GFR between treatment and control groups was calculated using, preferentially, (1) urine excretion or plasma decay of an indicator compound; (2) creatinine clearance, multiplied by 0.85 to correct for systematic overestimation; or (3) Cockcroft-Gault estimated GFR (see Other Publications of Related Interest). The variance of the difference in the rate of change in GFR between treatment and control groups for each study was based on the sum of the sample size divided by the variance for each group. In the five trials with insufficient data to calculate the variance, the value was randomly attributed across the range of values from the studies in which variance could be calculated. This was performed five times, yielding five different data sets.

Methods of synthesis

Results of the review

Overall, 24 studies (2,248 participants) were included: 13 RCTs (n=1,919), 1 non-randomised controlled trial (n=30), and 10 before-and-after studies (n=299).

The pooled effect size from 13 RCTs (n=1,919) suggested that protein restriction reduced the rate of decline in renal function by 0.53 mL/minute/year (95% CI: 0.08, 0.98). The magnitude and variability of the treatment effect appeared to be inversely proportional to study size. In a funnel plot of treatment effect versus study sample size, the smaller studies tended to report positive results, indicating possible publication bias in favour of low-protein diets.

Regression analysis, based on all 24 studies (n=,2248), suggested that the benefit of low-protein diet was significantly less in RCTs, compared with other study designs (regression coefficient -5.2 mL/minute/year, 95% CI: -7.8, -2.5, p<0.05). However, the benefit was relatively greater among diabetic versus non-diabetic patients (regression coefficient 5.4 mL/minute/year, 95% CI: 0.3, 10.5, p<0.05), and there was a trend toward a greater effect with each additional year of follow-up (regression coefficient 2.1 mL/minute/year, 95% CI: -0.05, +4.2, non significant). However, the number of diabetic patients was small (n=129) and the duration of follow-up was short in most trials. None of the other patient or study design characteristics examined were independently associated with differences in the effect of a reduced protein diet on GFR.

Authors' conclusions

Dietary protein restriction retarded the rate of renal function decline. However, the relatively weak magnitude of this effect suggests that better therapies are needed to slow the rate of renal disease progression.

CRD commentary

The research question, and the methods used for the data synthesis, were presented clearly. The search strategy was limited, and the search terms were not provided. If other relevant sources had been accessed, it is possible that further eligible material may have been identified. However, the authors did estimate the impact of publication bias using a funnel plot, which was useful. Detailed selection criteria were presented for study design and outcomes, but there was less information on the participants and interventions. In addition, it was unclear what regimens were received by the control groups. An assessment of study validity was reported, but the results of this were not presented, and there were very few other details of the individual studies. Few details of the study selection process were provided, i.e. how many reviewers were involved, whether the decisions were made independently, and how any disagreements were resolved.

The authors' conclusions appeared to supported by the results of the review. However, they have focused on clinical aspects, i.e. the view that the relatively small estimated effect size may not be useful in clinical practice. It would also have been appropriate for them to have mentioned the methodological limitations of this review, particularly given the limited literature search.

Implications of the review for practice and research

The authors state that additional studies should be completed before concluding that a low-protein diet is uniquely beneficial among diabetic patients. Evidence was found for publication bias that induced a tendency for the published literature to overestimate the benefits of a low-protein diet. This analysis suggests that better therapies are needed to retard the rate of GFR decline in patients with renal disease. It is possible that studies of longer duration could show a greater effect of a low-protein diet.

Bibliographic details

Kasiske B L, Lakatua J D, Ma J Z, Louis T A. A meta-analysis of the effects of dietary protein restriction on the rate of decline in renal function. American Journal of Kidney Diseases 1998; 31(6): 954-961. [PubMed: 9631839]

Other publications of related interest

1. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31-41. 2. Martinez-Maldonado M, Sattin RD. Rate of progression of renal disease and low-protein diet [editorial]. Am J Kidney Dis 1998;31:1048-9.

Indexing Status

Subject indexing assigned by NLM

MeSH

Diabetic Nephropathies /diet therapy /physiopathology; Dietary Proteins /administration & dosage; Disease Progression; Glomerular Filtration Rate; Humans; Kidney Failure, Chronic /diet therapy /physiopathology; Randomized Controlled Trials as Topic

AccessionNumber

11998001020

Database entry date

31/10/1999

Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.