Review
Copyright ©The Author(s) 2016.
World J Nephrol. May 6, 2016; 5(3): 233-257
Published online May 6, 2016. doi: 10.5527/wjn.v5.i3.233
Table 1 Single arm intervention studies of soy protein and kidney function
Ref.StudydesignKidney functionSubjects/groupAmount of soy protein usedControl/comparator proteinDuration of interventionOutcomesNotes
Cupisti et al[93]Single arm dietary intervention studyRenal transplant patients with moderate HC13 subjects completed study (7M, 6F)Goal was to replace 25 g/d animal protein with soy protein (dietary counseling only)Animal protein (baseline)5 wk on soy dietSignificant decrease in urinary creatinine after 5 wk on soy protein compared to baseline (P < 0.05)
Soy protein resulted in significant decrease in TC (P < 0.05) and LDL-C (P < 0.01) after 5 wk compared to baseline; no change in HDL-C
Cupisti et al[94]Single arm dietary interventionRenal transplant patients and and age, sex-matched healthy controls (latter for vascular measure comparisons only)20 per group (12M, 8F)Goal was to replace 25 g/d animal protein with soy protein (dietary counseling only)Animal protein (baseline and WO)5 wk on soy diet followed by 5 wk WORenal transplant patients had significantly reduced FMD compared to age- and sex-matched control subject (P < 0.001) with no differences between groups in non-endothelium-mediated vasodilationFirst study to show improvement in endothelial function in brachial arteries of renal transplant patients when animal protein substituted with soy protein
Soy diet did not change total dietary protein intake, BW, renal function, urinary protein excretion, serum Ca or P
Soy diet reduced TC and LDL-C and LOOH (P < 0.01) compared to baseline diet
Soy diet resulted in improvement in FMD (P = 0.003) compared to baseline while reactive hyperemia and endothelium-independent vasodilation was unchanged; FMD returned to baseline after WO
Increase in FMD correlated to increase in L-arg/ADMA ratio (P < 0.05) with soy diet
D’Amico et al[95,96]Single arm dietary interventionNephrotic patients with proteinuria > 1-5 g/24 h over 25 mo and HL20 subjects (13M, 7F)0.7-0.8 g/kg per day mostly from soy protein in test diet; test diet also contained vegetable oils and no cholesterol0.7–0.8 g/kg per day animal protein (baseline and WO)8 wk baseline diet followed by 8 wk soy diet and then 8 wk WOTC, LDL-C, HDL-C, apoAI and apoB decreased on soy diet compared to baseline diet (P < 0.001); no change in TG; lipids tended to revert to baseline during WOFibre, type of fat and no cholesterol were also other components of the soy protein arm that were different from the control diet; there was a modest but significant decrease in BW on the soy protein diet (no change in BMI)
Urinary protein, urea, Na and P excretion were reduced significantly from baseline during the soy diet (P < 0.001)
Soy diet results in significant decrease in CrCl with no change in serum creatinine; this persisted during WO
BP did not change
Table 2 Chronic controlled intervention studies of soy protein and kidney function
Ref.Study designKidney functionSubjects/groupAmount of soy protein usedControl/comparator proteinDuration of interventionOutcomesNotes
Liu et al[128]RCPre-hypertensive PM women90 subjects/group (85, 87 and 81 completed study in the soy, daidzein and placebo groups, respectively)40 g soy flour/d, 12.8 g soy protein/d40 g lowfat milk powder (placebo) or 40 g lowfat milk powder with 63 mg/d daidzein6 moNo significant changes in most renal parameters were observed between groupsAll subjects were equol producers
Soy flour intake resulted in less decrease in eGFRCockcroft (P = 0.044) and % change in eGFR (P = 0.031) after 6 mo compared to the milk placebo group (P = 0.044)
Effect of soy flour consumption to increase eGFR was greater in women with higher initial plasma cystatin C concentrations (Cys-C > 1.14 mg/L) (P = 0.001 for eGFRCockcroft) compared to milk placebo
Ahmed et al[131]RCGlomerulo-pathy with proteinuria (non-diabetic)9 subjects/group,0.8 g/kg per day soy protein0.8 g/kg per day an animal protein or 0.8 g/kg per day a soy protein + fiber8 wkNo significant changes in anthropometric measures, serum lipids or proteinuria between diet groupsSignificant decreases from baseline in overall energy and protein intake in all groups confounds end of study comparisons
total 27 subjects (4M, 23F)
Soroka et al[129]RCNon-diabetic, non-nephrotic CRF patients (urinary protein excretion < 3 g/d)9 completed study (5M, 4F)0.71 g/kg BW protein, mostly soy protein with egg (VPD)0.85 g/kg per day APD (1:1, animal sources:grains)6 moNo difference in renal function between groups seen; both groups saw reduction in rate of GFR declineHigh dropout and small number of subjects
BUN, Urinary N excretion, PCR, 24 h urinary creatinine and phosphate were lower in VPD groupDifferences in total energy and protein intake in VPD and APD
D’Amico et al[102], Gentile et al[103]RCNon-diabetic, nephrotic patients with proteinuria > 2.5 g/24 h for a mean of 24 mo and HC20 subjects (9M, 11F)Protein intake at end of study was calculated from urinary urea excretion to be 1.16 ± 0.04 g/kg per day (98% of this estimated to be soy protein)Soy protein used in both experimental arms of study; baseline diet was comparator8 wk for each arm (baseline diet, soy ± 5 g/d fish oil in random order) followed by WO for 3 mo on baseline diet)Soy diet significantly reduced TC, LDL-C, HDL-C, apoB (P < 0.0001) and apoAI (P < 0.01) compared to baseline; TGs were unaffected; lipids tended towards baseline values after WOBoth diet interventions resulted in modest decrease in BW and BMI (-4%) which was significantly different from baseline; both values tended towards baseline during WO
Addition of 5 g/d fish oil to soy diet resulted in significant elevation of TC and apoB compared to soy diet alone (P < 0.01)
Urinary protein, urea, P, Na and creatinine excretion was significantly decreased by both diet interventions (P < 0.01); measures tended towards baseline after WO
Blood glucose was significantly reduced by both diet interventions (P < 0.01) however soy diet alone reduced blood glucose more than soy diet and fish oil (P < 0.01)
Anderson et al[97]RCT2D with proteinuria, obese, and HTN8 M1.0 g/kg per day protein, 50% soy protein in soy test diet1.0 g/kg per day protein, 50% ground beef in animal test diet8 wk, 4 wk WOTC and TG decreased by soy diet (P < 0.05) vs animal protein dietLow number of subjects
SUN sig decreased by soy protein (P < 0.05)
Change in GFR similar with both diets
Urine protein excretion increased by soy vs animal protein diet (P = 0.028)
Azadbakht et al[98]RPT2D subjects with nephropathy, proteinuria, HTN41 subjects: 18 M, 23 F0.8 g/kg per day protein, 35% soy protein (textured soy protein), 35% animal protein, 30% vegetable protein0.8 g/kg per day protein, 70% animal and 30% vegetable protein4 yrDecreased FPG in soy group (P = 0.03
Soy protein group decreased TC (P < 0.01), LDL-C (P = 0.01) and TG (P = 0.01
Serum CRP decreased (P = 0.02) on soy protein diet
Soy protein diet reduced proteinuria (P = 0.001) and urinary creatinine (P = 0.01)
Miraghajani et al[108,113]RCT2D subjects with nephropathy25 subjects completed the study (10 M, 15 F)2.5 g soy protein (240 mL soymilk/d)3.3 g cow milk protein (240 mL milk/d)4 wk interventions with 2 wk WOSoy protein consumption resulted in a significant difference in % change of fibrin D-dimer concentrations compared to milk protein (P = 0.04); there were no differences in % changes in TNFα, IL-6, CRP, MDA or fibrinogen concentrations between groupsAmount of soy protein used in diet intervention was low
Soy protein consumption resulted in significant decrease in systolic BP compared to cow milk protein (-4.50% vs + 5.89%, P = 0.02)
Teixeira et al[106]RCT2D subjects with nephropathy14 male subjects0.5 g/kg per day soy protein (Approximately equal 50% of total daily intake)0.5 g/kg per day casein8 wk interventions with 4 wk WOUrinary albumin-creatinine ratio was significantly reduced by ISP (P < 0.0001) and increased by casein (P = 0.002)
Change in urinary albumin-creatinine ratio correlated inversely with plasma isoflavone levels (P = 0.012)
CrCl did not change (GFR) with either diet
HDL-C was increased after ISP (P = 0.0041) while it tended to be lower after casein (P = 0.0847)
TC and LDL-C not changed by either diet
Total and glycated hemoglobin did not change in either group
No differences in BP between groups; however soy diet resulted in higher plasma arg/lys ratios (P = 0.0097) which persisted after fasting
Stephens et al[104]RCT1D subjects with hyper-filtration GFR > 120 mL/min/1.73 m212 subjects completed study (6 M, 6 F 0)45-55 g/d soy protein to substitute for animal protein in control diet45-55 g/d animal protein8 wk interventions; no WOGFR sig lower in soy group vs control group (P = 0.02)No washout between interventions
Excretion of urinary creatinine, urea and Na not diff between groups
Microalbuminuria within normal ranges and unaffected by diet
TC and LDL-C significantly reduced in soy group (P < 0.02, 0.05, respectively) whereas TG and HDL-C not diff between groups
Serum glucose was not affected by soy protein diet but was significantly increased on the control diet (P < 0.05) compared to baseline
Serum albumin did not change but total serum protein decreased in soy group (P < 0.05)
Chen et al[101]RPNondiabetic hemodialysis patientsSoy group: 10 HL (7 F, 3 M) and 8 NL (6 F, 2 M)30 g/d soy protein30 g/d milk protein12 wkNo significant differences between groups in serum nutritional parameters or hemodialysis adequacyTest proteins consumed on top of usual hemodialysis diet
TC and TG decreased in HL subjects consuming soy vs milk protein over time (P < 0.05 at 12 wk)
Control group: 9 HL (7 F, 2 M) and 10 NL (7 F, 3 M)Non-HDL-C, apoB, TC/HDL-C ratio and insulin decreased in HL subjects consuming soy vs milk protein at 12 wk (P < 0.05)
Non-significant differences between protein groups in NL subjects
Soy protein resulted in significant decrease in fasting insulin in NL group at 12 wk compared to values at baseline (P < 0.05)
Chen et al[100]RPNon-diabetic hemodialysis patients with HCSoy group: 13 (9 M, 4 F)30 g/d soy protein30 g/d milk protein12 wkNo significant differences between groups in serum nutritional parameters or hemodialysis adequacyNot clear if some of the subject are the same as reported in Chen et al[101] as study protocols are the same
Milk group: 13 (10 M, 3 F)TC, non-HDL-C, apoB, TC/HDL-C and LDL-C/HDL-C ratios decreased in subjects consuming soy vs milk protein at 12 wk (P < 0.05)
No differences in TG between soy and milk groups
Soy protein resulted in significant decrease in fasting insulin at 12 wk compared to milk protein group (P < 0.05)
Imani et al[110]RPPD patients18 subjects14 g soy protein at dinner each dayMeat instead of soy protein at dinner8 wkSoy protein diet resulted in significant 17% reduction in plasma coagulation factor IX activity compared to control group (P < 0.05)Study was not blinded
Soy group (9 M, 9 F),No significant changes in oxLDL, P, fibrinogen or activities of coagulation factors VII and X between groupsMean energy and protein intakes were less than recommended amounts (30 kcal/kg per day and 1.2 g/kg per day, respectively) which is common among PD patients
Control group (9 M, 9 F)
Fanti et al[112]RPESRD patients on chronic HD with elevated CRP (> 10 mg/L)Soy group = 15; control milk group = 1025 g/d25 g/d milk protein8 wk5 to 10-fold increase in mean serum IF concentration in soy group at end of study (P < 0.001)Small number of subjects
No significant change in CRP between groups, however, significant inverse correlation of CRP with IF concentrationTest proteins provided as beverages, a cereal-type product and as snack bar
Significant positive correlation of serum IF concentration and serum albumin and IGF-1
Siefker et al[109]RPHD patients17 subjects25 g soy protein (4 times per week)Whey protein (exact amount not specified); provided 4 times per week4 wkNo difference between groups on serum markers of renal function except creatinine; whey protein showed a significant decrease in creatinine from baseline (P < 0.05) whereas there was no change in the soy protein group from baselineSmall number of subjects
8 subjects on soy protein diet; 9 on whey proteinoxLDL was significantly decreased after soy protein consumption (P < 0.05) compared to baseline; the % change in oxLDL compared to the whey group was significantly different (P < 0.05)
No differences in plasma concentrations of 8-iso-PGF, TNFα, or CRP between diet groups
Tomayko et al[114]RPMHD patientsSoy group = 1227 g/d soy protein27 g whey protein or noncaloric placebo powder (2 g Crystal Light)6 moA significant time x treatment effect for IL-6 levels (P = 0.036) with both whey and soy protein groups decreasing compared to control groupFirst study to observe improvements in inflammation and physical function after intradialytic nutritional support in MHD patients with serum albumin ≥ 3.9 g/dL (i.e., not malnourished)
Whey group = 11Soy diet resulted in a significant decrease in neutrophil-lymphocyte ratio (systemic inflammation marker) compared to control or whey diet (P = 0.02)
Placebo control = 15Alkaline phosphatase, a marker of bone turnover, was increased in the control diet compared with both protein diet groups (P = 0.04)
A significant time by treatment interaction was seen for gait speed when all 3 groups analyzed (P = 0.048); both soy and whey groups indicated improved gait speed while control diet had a decline
Shuttle walk test time was significantly improved in the whey group (P < 0.05) and when protein groups were combined (P < 0.05) versus the control group; shuttle walk test time was increased in the soy group but was not significant compared to the control group (which had decreased test times)
Table 3 Human studies with soy protein and renal isoflavone metabolism
Ref.Study designKidney functionSubjects/groupAmount of soy protein usedControl/comparator proteinDuration of interventionOutcomesNotes
Fanti et al[115], Franke et al[192]3 separate protocols:ESRD patients on HD and normal healthy subjects23 HD subjects and 10 healthy subjects for baseline IF measures20 g soy proteinBaseline diet is self-selected standard renal dietSingle meal interventions55%-65% of HD patients had undetectable serum IFs and 35%-45% had concentrations > 200 nM on standard renal dietFirst study to report blood levels of genistein and daidzein in ESRD patients
Assessment of baseline serum concentrations of IFs7 HD patients and 8 healthy subjects for meal intervention study (8 h only); 2 healthy subjects and 3 HD subjects had multiple serum and urine timepts collectedSerum concentrations of IFs greater post-soy protein ingestion compared to baseline for both groups (P < 0.001); concentrations in HD subjects after 8 h of soy protein consumption were greater than those in healthy subjects (P < 0.05)Daidzein metabolites equol and O-DMA were not detected in sera of any of the subjects
Post-ingestion concentrations of IFs5 HD patients for pre- and post-dialysis IF measuresHalf-lives of genistein and daidzein averaged 3.5 and 6 h in healthy subjects, respectively but were increased to an average of 47 and 58 h in HD patients
Effects of hemodialysis on IF concentrationsHD did not effectively remove IFs from serum since (due to higher molecular weight of conjugates and large proportion of unconjugated IFs are bound to albumin)
Fanti et al[116]ObservationalRandomly selected HD patients residing in the United States, Japan or ThailandSubjects from:Habitual dietary intake of soy was assessed by questionnaires developed by their renal replacement therapy programme dieticiansStudy aim was to compare habitual dietary intake of soy in 3 countriesN/ASerum IF concentrations significantly higher in HD patients from Japan compared to United States or Thailand (P < 0.0001)
United States = 20Significant correlation between soya intake and genistein (P < 0.0001), daidzein (P < 0.0001), glycitein (P < 0.001) and O-DMA (P < 0.01) in subjects from all 3 countries
Japan = 20ESRD HD patients displayed consistently higher concentrations of daidzein compared to genistein, while the reverse occurs in healthy subjects
Thailand = 17Concentrations of sulphated and unconjugated compounds in HD subjects (Japan only studied) are comparable to those detected in healthy subjects
Locati et al[117]Single arm intervention studyRenal transplant patients16 subjects (11 M, 5 F)25 g soy protein substituted for 25 g animal protein25 g animal protein (as habitual diet)5 wkSerum IFs were measured and 5 different groups were identified on the basis of the IF profiles: (1) 4 subjects had no detectable IFs; (2) only genistein was quantifiable in 7 patients; (3) 3 patients had only detectable genistein and daidzein; (4) 2 subjects only had detectable genistein and equol; and (5) 1 subject had the highest observed genistein and daidzein with detectable dihydrogenistein and equolConcentrations of serum IFs in the renal transplant patients were similar to those observed in healthy subjects
Table 4 Human studies with soy protein on renal calcium metabolism
Ref.Study designKidney functionSubjects/groupAmount of soy protein usedControl/comparator proteinDuration of interventionOutcomesNotes
Breslau et al[119]RCNormal15 subjects completed animal and ovo-vegetarian diet phases; 10 completed all 3 phases (including vegetarian)Soy protein accounted for most of the 75 g protein/d in vegetarian phase; accounted for an unspecified but lower amount in ovo-vegetarian phaseAnimal protein accounted for most of the 75 g per day in the animal protein phase; consisted of dairy, beef, chicken and fish12 d No WOSerum uric acid concentrations were significantly lower with the vegetarian and ovo-vegetarian diets compared to the animal protein diet (P < 0.01); urinary uric acid excretion was significantly lower in ovo-vegetarian diet vs animal diet only (P < 0.02)Diets were constant for Ca, P, Na and total protein
Urinary Ca and P were significantly lower in vegetarian diet compared to beef diet (P < 0.02); urinary oxalate was significantly higher in vegetarian vs beef diet (P < 0.02)
Animal protein diet resulted in lower PTH level vs vegetarian diet (P < 0.05)
Serum 1,25-(OH)2D was higher in the vegetarian vs animal protein diet(P < 0.01)
Roughead et al[120]RCNormal PM women13 female subjects25 g soy protein substituted for 25 g meat protein25 g meat protein in control diet7 wkCa retention was not affected by substituting soy protein for meat protein
Urinary pH was higher on the soy diet compared to the control diet (P < 0.0001); renal acid excretion was lower during soy diet (P = 0.0001) however urinary Ca excretion was similar between soy and meat diets
Substitution of soy protein for meat protein did not affect bone metabolism as indicated by no differences between diets in a number of specific bone biomarkers
No differences between soy and meat protein diets in plasma lipid or hemostatic measures
Table 5 Single meal intervention studies of soy protein and kidney function
Ref.Study designKidney FunctionSubjects/groupAmount of soy protein usedControl/comparator proteinDuration of interventionOutcomesNotes
Bilo et al[122]Single meal intervention study (crossover)Normal healthy subjects6 normal subjects; 5 M, 1 FStudies in normal subjects only: 80 g soy protein in single oral administrationStudies in normal subjects only: 80 g lactoprotein or beef protein or 36 g amino acidsNormal subjects: 8 individual renal function tests run on separate daysSoy protein ingestion induced significantly lower rises in GFR and ERPF compared to beef protein but not compared to lactoprotein or 36 g amino acid ingestionSubjects with chronic renal insufficiency (PKD, NS, or MGP) were studied in a separate series of experiments in this publication, but were not used to evaluate soy protein
Buzio et al[127]Single meal intervention study (crossover)Normal healthy subjects7 (gender not specified)80 g (0.9-1.3 g.kg BW)80 g red meat or 80 g dairy (cheese)Single meal interventions conducted 1 wk apartCrCl and urinary protein were not different between protein loadsPublication describes 2 separate experiments; soy protein effects on renal function only assessed in second experimental protocol
UAp was significantly lower after soy protein meal versus red meat or cheese meals (P < 0.01) (samples taken 4 h post-meal)
Water excretion rate was higher after soy protein load versus meat (P < 0.05) or cheese (P < 0.01)
Serum total protein was lower after soy protein load compared to meat (P < 0.01) or cheese (P < 0.01) loads
Deibert et al[126]Single meal intervention study (crossover)Normal healthy and metabolic syndrome subjects; all with normal kidney function10 subjects per group (All males)1st intervention: 1 g/kg/BW soy protein: Milk protein (83% soy protein);N/ASingle meal intervention in normal healthy subjects; 2 meal interventions in subjects with metabolic syndrome (1 wk apart)Patients with metabolic syndrome had significantly elevated baseline GFR and ERPF compared to healthy subjects (P = 0.02)0.3 g/kg/BW is amount of protein used in meal replacement therapy
2nd intervention same protein source at 0.3 g/kg BWAfter ingestion of 1 g/kg/BW protein, GFR and ERPF increased in both groups however the subjects with metabolic syndrome had significantly higher increases in GFR (P < 0.002) and ERPF (P < 0.02) compared to normal subjects; no significant effect of ingestion of 0.3 g/kg per BW protein on renal parameters in subjects with metabolic syndrome
Howe et al[118]Single meal intervention study (Latin square crossover)Healthy PM women8 F subjects45 g soy protein0 g protein, 45 g beef or dairy protein (cottage cheese)Single meal intervention; 6 meal interventions (1 wk apart)Urinary Ca excretion was significantly greater after 45 g protein meal for all proteins compared to basal (0 g protein) meal (P < 0.05)
% Ca resorbed by the kidney was significantly reduced after the dairy and soy protein meals (P < 0.05)
Serum ionized Ca was unaffected, however, serum P was significantly lowered by all protein meals (P < 0.05) compared to 0 g protein meal
Soy protein meal significantly reduced calcitonin versus baseline (P < 0.05) however, all protein means tended to lower calcitonin compared to baseline
Dairy protein significantly increased PTH (P < 0.05) compared to baseline, however all protein meals tended to elevate PTH compared to baseline
Serum insulin was significantly increased by all protein meals (over time) compared 0 g protein meal (P < 0.05)
Kontessis et al[70]Single meal intervention study (crossover)Normal healthy subjects7 M subjects80 g soy protein80 g lean beef2 separate single meal interventionsGFR and ERPF increased significantly after acute beef protein load (P < 0.005 compared to baseline) but did not increase with soy protein loadAmount of soy protein in vegetable protein diet in the reported chronic study was not specified so is therefore not summarized
Renal vascular resistance fell significantly after beef load (P < 0.05) but was unchanged after soy protein load; plasma 6-keto-PGF1α rose significantly after meat load (P < 0.05) but not after soy protein load
Fractional albumin and IgG clearance rose after beef load (P < 0.05 and P < 0.001, respectively) but did not change significantly after soy protein load; plasma protein concentrations were not different between different protein loads; UAp was not different between groups
Plasma glucagon increase was higher after meat load (P < 0.05) compared to soy protein load; no differences were seen between proteins on plasma insulin or growth hormone
Nakamura et al[123]Single meal intervention study (crossover)Healthy and T2D subjects (T2D divided into 3 groups based on AER: Group A ≤ 20 μg/min (Normal); B = 20-200 μg/min; C ≥ 200 μg/min11 healthy subjects (8M, 3F); 20 T2D patients (10 M, 10 F)1g/kg soy protein (as bean curd)1 g/kg tuna fish proteinMeals fed on separate daysIn healthy subjects, eGFR increased (P < 0.01) after tuna meal but no significant difference after soybean curd meal
In Grp A, eGFR increased with tuna meal (P < 0.01) but not after soybean curd
In Group B there was no difference in GFR with either protein
In group C, GFR sig decreased after tuna meal (P < 0.05) but not with soy protein
No changes in AER with any protein in any group
Nakamura et al[124]Single meal intervention study (crossover)Healthy and T2D subjects10 healthy subjects and 6 T2D subjects0.7 g/kg soy protein (as bean curd)0.7 g/kg tuna fish protein or egg white protein or dairy protein (cheese)Meals fed on separate dayseGFR was only significantly increased after ingestion of tuna fish protein (P < 0.001) and not after consumption of soy, egg white or dairy proteins
Orita et al[125]Single meal intervention study (crossover)Healthy subjects6 male subjects86.9 g soy protein86.9 g beef protein or fasting (0 g protein)Meals fed 1 wk apartInulin clearance (GFR) was significantly increased over baseline at 2 h post beef or soy protein compared to fasting (P < 0.005 and P < 0.05, respectively)First study to show an increase in GFR after a soy protein load in healthy subjects
Creatinine clearance (GFR) was significantly increased by both beef and soy proteins at 2 and 3 h post-ingestion compared to fasting (P < 0.01)
Plasma glucagon was significantly increased at 1 to 3 h post-ingestion by both beef and soy protein compared to fasting (P < 0.01)