[This article originally appeared in Issue #8 of the Schwanz Nutrition Journal.]
Protein restriction has been recommended for individuals with chronic kidney disease for decades. The ideal amount of protein is still debated, but most agree that some degree of protein restriction is beneficial to slow the progression of kidney disease. More recently there has been an increased focus on the source of protein, with some researchers suggesting that plant protein may be less likely to cause kidney damage than animal protein.
Research on this topic mostly consists of short-term and poor quality studies; however, many reviews have been published on this topic that misrepresent studies and only cover a small portion of the evidence base, so a thorough review is needed.
Multiple trials have been completed that tested the response to a single meal that varied by protein source. One showed increased GFR, renal plasma flow, and fractional albumin clearance after beef consumption, with no change in these variables after soy consumption (1). Another trial showed increased GFR and creatinine levels after tuna consumption, with no change after tofu, egg white, or cheese (2). Another showed that tuna consumption led to increased GFR in healthy subjects, with a stable or decreased GFR in diabetics with micro- or macroalbuminuria (3). No changes were seen with tofu consumption. The primary limitation of these studies is their extreme short-term nature. It is essentially unknown if these single-meal responses will lead to kidney damage over a long period of time. Caution is needed when interpreting studies such as this, as a few measures of kidney function or blood tests will not capture the overall impact on human physiology. A prime example is exercise; if we measure blood pressure during or after exercise it will likely show significant increases compared to baseline. If we use this single measure alone, we may conclude that exercise is harmful; but if we look at the long-term impact on overall health, we will most certainly find that exercise is beneficial.
A cross-sectional study was completed that showed lower levels of albuminuria were associated with increased low-fat dairy intake and that a better albumin-to-creatinine ratio was associated with low-fat dairy and fruit intake (4). A case-control study was completed that showed a higher intake of fish protein was associated with lower levels of albuminuria (5). Both of these studies suffer from the same problem: a very weak study design. Both studies will be highly prone to confounding and error as there is no randomization and dietary intake is estimated from a single food frequency questionnaire. In regards to protein intake, both studies showed that animal protein was associated with more favorable measures of renal function, with no associations found for plant proteins. These studies have minimal if any influence on the final conclusions on this topic, but both lean in favor of animal protein.
Prospective cohort studies provide a stronger study design, but are still highly prone to confounding and error with no randomization. The Nurses’ Health Study (6) is one popular cohort study that was used to assess dietary intake and renal function. Subjects completed food frequency questionnaires and had GFR estimated based on serum creatinine levels. Results showed that nondairy animal protein was associated with a decrease in GFR over time, with no association found for dairy or vegetable protein. The Nurses’ Health Study is infamous for its “healthy user bias” with many healthy behaviors such as increased exercise, less smoking, and a higher intake of fruits and vegetables seen together, and also with unhealthy lifestyle behaviors seen together. Thus, if negative associations are found between meat intake and renal function, it is very difficult to determine if the negative effects are due to the meat intake or the unhealthy lifestyle factors that are associated with meat intake. Also, in this particular study, the final food frequency questionnaire was completed in 1994, and the last blood draw to estimate the GFR was completed in 2000 – a significant amount of time for dietary intake to change. Another observational study showed decreased total mortality with increasing plant protein intake (7). Again the significance of this trial is severely limited by its observational design, its one-time intake analysis, one-time estimation of GFR with serum creatinine, and a notable healthy user bias. Subjects with less plant protein intake had higher levels of smoking and alcohol intake along with a higher BMI. Observational studies function as hypothesis-generating or supporting studies rather than hypothesis-testing studies. Both of these observational studies were supportive of the idea that plant protein is beneficial for kidney function, but similar to the cross-sectional and case-control studies, they carry minimal weight on the overall conclusions.
Randomized, controlled trials (RCT) are the most informative studies and will be the most important to consider when drawing final conclusions. An eight-week RCT was completed that compared diets based on animal protein, plant protein, or plant protein with added fiber; all diets provided 0.8 g/kg/day of protein. This study showed no significant difference among the three groups on all outcomes which included proteinuria, creatinine clearance, and serum creatinine. Numerous randomized, crossover trials were completed that compared animal protein to plant protein. The actual interventions varied among the studies and included soy powder, casein powder, whole soy, and general animal and vegetable proteins. These studies were short-term (range of 3-8 weeks) and had small sample sizes (range of 8-17 subjects). Results of these studies were mixed, but overall were more supportive of plant protein. Some studies (1,9,10) but not all (11) showed decreased proteinuria with plant proteins; one showed mixed results (12) and one showed increased proteinuria with the soy intervention compared to the animal protein14. These studies were mixed in terms of the effect on GFR, with some showing decreased GFR (1,9) with plant proteins and some showing no difference in GFR (11,13). These studies had mostly comparable designs, though in one study all foods were prepared by the researchers and provided to subjects, which allowed the researchers to design the diets to be equal in all nutrients and differed only by the protein source. This study included the largest sample of subjects (n=17) and lasted for six weeks. This study showed no difference in GFR or albumin excretion between the two groups (11).
A common assumption found in most peer-reviewed studies and review articles is that all animal proteins have the same impact on renal function; the same is assumed about vegetable proteins. However, evidence actually shows that this assumption is wrong. Two similar randomized crossover trials were completed where subjects followed a typical diet high in red meat, a diet that substitutes either chicken and fish (14) or chicken alone (15) for red meat, and a reduced protein vegetarian diet. Subjects followed the diets for 3-4 weeks and results showed that the diet that replaced red meat with chicken or fish showed the same effect as the low protein vegetarian diet. The first study showed decreases in the GFR in the low protein vegetarian diet and the chicken and fish diet compared to the red meat diet, with no differences in albumin excretion. The second study showed no differences in GFR, but showed higher levels of albumin excretion on the red meat diet than the two other diets. Again, no differences were found between the chicken diet and the low protein vegetarian diet. Additionally, in one of the single-meal trials mentioned previously, multiple animal proteins were compared. This study (2) compared tuna, tofu, egg white, and cheese and showed increased GFR and creatinine with tuna consumption, while no differences were seen between the tofu, egg white, and cheese.
The primary limitation to the RCTs presented so far is their short-term duration. These studies are able to show short-term effects on GFR, albumin excretion, and blood tests, but it is unclear what the long-term effects of these changes are. It’s possible that these changes could be a normal physiological response to ingestion of different types of foods. Also, it’s important to not lose sight of the bigger picture. Human physiology is incredibly complex and there can be many different changes after eating. Measuring a few of these changes while ignoring others may not give an accurate representation of the overall effect on health. Another good example of this is sodium restriction in subjects with CHF. Sodium restriction may be beneficial to prevent fluid retention and exacerbation of CHF, but too much of a restriction may lead to increases in the renin-angiotensin-aldosterone system and have a negative impact on overall health.
The primary goal of protein manipulation with chronic kidney disease is to prevent the decline in renal function that can lead to the initiation of dialysis. This can be a slow, gradual process, so short-term studies that last for only a few weeks will typically be unable to show differences between the interventions. Three RCTs were found that included six months of each intervention, which is much better than the previously reported studies but still shorter than ideal to show progression to end-stage kidney disease. The first was published only in abstract form, and compared meat and soy-based diets in 23 subjects with CKD (16). After six months of intervention, no significant difference was found between the groups in eGFR or other measures. A second trial compared subjects taking supplements of either soy powder, low-fat milk powder, or low-fat milk powder plus a soy isolate (17). A total of 265 women completed the six month trial. Results showed that the soy powder and the low-fat milk powder plus soy isolate had nearly identical results, which were superior to the low-fat milk group in measures of serum creatinine and GFR (81.9 mL/min/1.73m2 at baseline to 82.8 in the soy powder group; 86.9 mL/min/1.73m2 at baseline to 82.6 in the low-fat milk powder group). These results indicate that it is most likely not a negative effect from the milk powder but a beneficial effect from the soy. A third trial had a crossover design with nine subjects following both an animal-protein diet and a plant-protein diet for six months each (18). No differences were found in GFR, creatinine clearance, or proteinuria between the diets. All subjects in this study were being followed at a nephrology clinic for at least one year prior to this study, so researchers were able to compare the changes in kidney function during the study period to the year prior to starting the study. They found the rate of GFR decline decreased by an average of 73% after starting the study, with no difference between the groups.
Six-month trials give a more moderate-term view of the impact of dietary changes on renal function, but none of the six-month studies reported any subjects starting on dialysis. Most only reported on minor declines in GFR. One last study goes far beyond the six-month mark and does report on new dialysis cases. Unfortunately the study had a fairly small sample size of 41 subjects, but the study was carried out for four years (19). Subjects were randomized to follow a standard diet high in animal protein or a diet high in soy protein. This study measured the same markers as previous studies and found similar results: no differences were seen between the groups on GFR, but the soy diet did result in lower levels of proteinuria. So did the higher levels of proteinuria in the animal protein group result in higher rates of end stage kidney disease that required dialysis after four years? At the end of the trial, there were three subjects in the soy group that had started dialysis, and two in the animal protein group. The small sample size of this study along with the small amount of subjects that started dialysis make it unlikely that the study would be able to show a statistically significant difference in this endpoint, even if it did exist. One other study that was previously mentioned also reported on new dialysis patients during a seven-week crossover trial (10), with one subject in each group starting dialysis.
When reviewing the totality of the evidence on this topic, it is very clear that it is weak, mixed, and inconclusive. It is notable that in a number of different studies, plant proteins resulted in less proteinuria and less increase in GFR in the short-term. This is a fairly weak observation given several other trials that show no difference in these measures. However, it seems reasonable to conclude that certain animal proteins and plant proteins have different impacts on renal function. One clear conclusion is that it is inappropriate to assume that all protein sources of animal origin have the same impact on renal function. Less research has been done that compared different plant proteins, but it is likely that different effects could also be seen. No differences were seen in any of the RCTs on the two most important outcomes: mortality and progression to dialysis; however, the studies did not have the statistical power to really test these outcomes. Therefore the conclusion on the effect of plant proteins vs animal proteins on these outcomes is inconclusive, but should be tested in long-term randomized controlled trials with an adequate number of subjects and a focus on individual protein sources rather than broad categories of protein types.
Current evidence does not support the idea that plant proteins can delay dialysis or reduce mortality when substituted for animal proteins in individuals with chronic kidney disease. Red meats seem to have the largest impact on renal function, showing the most consistent elevation of proteinuria and GFR in short-term studies. Soy products may have a beneficial impact on kidney function. Therefore, even though evidence is inconclusive, it may be prudent to recommend a reduction in red meat intake to be replaced with plant proteins, dairy proteins, fish, eggs, or white meats as a possible preventative measure. These interventions will be the most important in subjects with hypertension and/or hyperglycemia (20).
1. Kontessis, P. et al. Renal, metabolic, and hormonal responses to ingestion of animal and vegetable proteins. Kidney International 1990. 38:136-144.
2. Nakamura, H. et al. Glomerular filtration response to acute loading with protein from different sources in healthy volunteers and diabetic patients. Tohoku Journal of Experimental Medicine 1990. 162:269-278.
3. Nakamura, H. et al. Effects of acute protein loads of different sources on renal function of patients with diabetic nephropathy. Tohoku J Exp Med 1989. 159:153-162.
4. Nettleton, J. et al. Associations between microalbuminuria and animal foods, plant foods, and dietary patterns in the Multiethnic Study of Atherosclerosis. Am J Clin Nutr 2008. 87:1825-1836.
5. Mollsten, A. et al. Higher intakes of fish protein are related to a lower risk of microalbuminuria in young Swedish type 1 diabetic patients. Diabetes Care 2001. 24:805-810.
6. Knight, E. et al. The impact of protein intake on renal function decline in women with normal renal function or mild renal insufficiency. Ann Intern Med 2003. 138:460-467.
7. Chen, X. et al. The associations of plant protein intake with all-cause mortality in CKD. Am J Kidney Dis 2016. 67(3):423-430.
8. Ahmed, M. et al. Short-term effects of soy protein diet in patients with proteinuric glomerulopathies. J Bras Nefrol 2011. 33(2):150-159
9. Kontessis, P. et al. Renal, metabolic, and hormonal responses to proteins of different origin in normotensive, nonproteinuric type 1 diabetic patients. Diabetes Care 1995. 18(9):1233-1240.
10. Azadbakht, L. et al. Soy protein consumption and kidney-related biomarkers among type 2 diabetics: a crossover, randomized clinical trial. Journal of Renal Nutrition 2009. 19(6):479-486
11. Wheeler, M. et al. Animal versus plant protein meals in individuals with type 2 diabetes and microalbuminuria. Diabetes Care 2002. 25:1277-1282
12. Teixeira, S. et al. Isolated soy protein consumption reduces urinary albumin excretion and improves the serum lipid profile in men with type 2 diabetes mellitus and nephropathy. J Nutr 2004. 134:1874-1880.
13. Anderson, J. et al. Effects of soy protein on renal function and proteinuria in patients with type 2 diabetes. Am J Clin Nutr 1998. 68(suppl):1347S-1353S.
14. Pecis, M. et al. Chicken and fish diet reduces glomerular hyperfiltration in IDDM patients. Diabetes Care 1994. 17(7):665-672.
15. de Mello, V. et al. Withdrawal of red meat from the usual diet reduces albuminuria and improves serum fatty acid profile in type 2 diabetes patients with macroalbuminuria. Am J Clin Nutr 2006. 83:1032-1038
16. Tze-Wah Kao et al. Effects of soy protein and nutrition education on patients with chronic kidney disease. Kidney Research and Clinical Practice 2012. 31(2):A89.
17. Liu, Z. et al. Effect of whole soy and purified isoflavone daidzein on renal function - a 6-month randomized controlled trial in equol-producing postmenopausal women with prehypertension. Clinical Biochemistry 2014. 47:1250-1256.
18. Soroka, N. et al. Comparison of a vegetable-based (soya) and an animal-based low-protein diet in predialysis chronic renal failure patients. Nephron 1998. 79:173-180.
19. Azadbakht, L. et al. Soy protein intake, cardiorenal indices, and c-reactive protein in type 2 diabetes with nephropathy. Diabetes Care 2008. 31:648-654
20. Toeller, M. et al. Protein intake and urinary albumin excretion rates in the EURODIAB IDDM Complications Study. Diabetologia 1997. 40:1219-1226.