Review: Effects of Dietary Protein and Phosphorus

Restriction on the Progression

of Chronic Renal Failure
BY KATHLEEN R. ZELLER, MD

ABSTRACT: In recent years, evidence has accu- will amount to a direct cost in excess of $2 billion for
mulated suggesting that early dietary interven- 1986. Projections for 1990 are as high as $3.5 billion,
tion in the form -of protein restriction can dra- and this does not include the cost oflost productivity
matically alter the natural history of chronic in the work place. In addition, over 10,000 Ameri-
renal insufficiency. This article reviews the lit- cans are expected to die in the next year as a
erature in this area and summarizes the ques- consequence of renal failure or related complications
tions that remain for future investigators to stemming from therapy.l Thus, any therapeutic mo-
answer. dalities that slow the progression of end-stage renal
Interest in the role of protein intake in renal disease would have a major impact on morbidity,
disease dates back to the early 1900s, when sev- mortality, and national health-care costs.
eral investigators found that the progression of In recent years, evidence has accumulated to sug-
renal failure was accelerated in rats and rabbits gest that early dietary intervention, in the form of
fed high-protein diets. Subsequent work in ani- protein restriction, can dramatically alter the natu-
mal models has demonstrated that nephron ral history of chronic renal insufficiency. This article
loss, resulting from a variety of disease states, is reviews the literature in this area and summarizes
-a ssociated with "elevated"intraglomerular pres- the questions that remain for future investigators to
sures and flows. Several investigators now have answer.
postulated that these abnormal hemodynamics, Interest in the role of protein intake in renal dis-
resulting from intrarenal vasodilation and hy- ease dates back to the early 1900s.2 Newburgh and
perperfusion, damage the glomerulus and may Clarkson first reported in 1919 that high dietary
produce further nephron loss independent of protein resulted in significant renal structural dam-
the initial renal insult. Dietary protein restric- age. 3 ,4 Over the next 30 years, many other in-
tion appears to reduce these pressures and vestigators reported that high dietary protein intake
flows towards normal in animals, although the had a similar detrimental effect on the course of both
operative mechanism has yet to be identified. experimental nephritis and renal failure following
Limited studies of dietary protein restriction in subtotal nephrectomy. In 1928, Moise and Smith
human renal disease have suggested a bene- noted that rats subjected to uninephrectomy rapidly
ficial effect, but carefully controlled prospec- developed renal damage when fed a high-protein
tive studies will be necessary to establish clear diet. 5 Chanutin and Ludewig demonstrated that in-
therapeutic efficacy. KEY INDEXING TERMS: creasing the protein content of the diet from 10% to
Protein Restriction; Therapy of Chronic Renal 80% in rats subjected to partial renal ablation re-
Failure. [Am J Med Sci 1987; 294(5): 328-340.] sulted in a progressive increase in proteinuria, renal
failure, and mortality.6 Similarly, Farr and Smadel
(1939) reported that reducing dietary protein to 5%
A Pproximately 80,000 Americans currently re-
quire chronic dialysis therapy, and an addi-
tional 30,000 can be expected to require such
in rats with nephrotoxic serum nephritis resolved
clinical and morphologic evidence of the disease. 7,s
Control rats fed a 40% protein diet died of renal
treatment during the coming year. Adding an failure within one year. 7,s
anticipated 8,000 renal transplants this year, this In 1948, Addis suggested that protein intake be
restricted in patients with early renal insufficiency.
From the Department of Internal Medicine, University of Texas He argued that the excretion of urea required ther-
Health Science Center at Dallas, Dallas, Texas.
Reprint requests: Kathleen Zeller, MD, James W. Aston Center, modynamic work by the proximal tubule, and, there-
U4.318, University of Texas Health Science Center at Dallas, 5303 fore, catabolism of large amounts of protein put a
Harry Hines Boulevard, Dallas, TX 75235. strain on this segment of the nephron. This conclu-

328 November 1987 Volume 294 Number 5

 Zeller

sion was supported by the early observation that sumably reflecting "adaptive" hyperfiltration in
proximal nephron mass in rats increased substan- nephron units least damaged by the original disease
tially with high dietary protein intake. By reducing process.
protein intake, Addis hoped to decrease the strain on Likewise, it has become clear that patients with
surviving nephrons in diseased kidneys, thus in- renal insufficiency secondary to etiologic factors
creasing their longevity.9 Subsequent studies would such as post-streptococcal glomerulonephritis, corti-
largely invalidate the notions that secretion from cal necrosis, vesicoureteral reflux, or renal agenesis,
the proximal tubule is the major mode of urea excre- can progress to end-stage renal disease, even though
tion, or that this requires large amounts of renal the initial insult is inactive or corrected. Uncon-
work. With the advent of hemodialysis and trans- trolled hypertension or recurrent infection may con-
plantation, interest in the more general concept of tribute to this progression, but not in all cases. Pa-
dietary intervention in renal disease gradually tients with renal insufficiency solely secondary to
eroded. hypertension can develop progressive renal failure
despite effective blood pressure control. Similarly,
Brenner Hypothesis children with corrected vesicoureteral reflux fre-
Recently, there has been a renewal of interest in quently progress to end-stage renal disease despite
the possibility that protein restriction slows the correction of obstruction and infection. 13 Virtually
progression of chronic renal insufficiency. This all nephrologists would agree that most patients
largely stems from a unique hypothesis resulting with GFR's below 25/mllmin eventually will require
from a series of observations based on recent and dialysis or renal transplant, regardless of the origi-
remote literature. nal cause of reduced function.14
Simply stated, the hypothesis generated by Dr. In recent years, researchers have found that GFR
Barry Brenner in the late 1970s proposes that an does indeed vary in normal humans, depending on
initial insult damages the kidney, resulting in dietary protein intake. Pullman demonstrated a rise
compensatory changes by remaining "functioning" in basal GFR (estimated by inulin clearance) of 22
tissue. These changes include a decrease in intra- ml/min in 20 normal adults followed for a period of
renal vascular resistance, promoting arteriolar vaso·· 14 days, first on a very low-protein diet (0.3 g/kgl
dilation, increased single-nephron blood flow, and day), and then on a high-protein diet (2.6 g/kg/day).15
therefore, increased single-nephron glomerular fil- Similar results were obtained by Bosch (using Cr Cl)
tration rate (GFR). Brenner hypothesizes that these on five members of the Nephrology Division at
changes occur in all types of renal disease, but be- Mount Sinai. 16 Bergstrom found substantially
come increasingly significant as fewer functioning smaller changes in inulin clearance (13 ml/min) be-
nephrons remain. 10, 11 tween a 0.3 and 2 g/kg protein diet after six days.
He and others propose that the rise in intra- Both Bosch and Bergstrom went on to demonstrate a
glomerular pressure, resulting from arteriolar di- further acute increase in GFR in the first three
lation, damages the glomerulus directly by altering hours following a high-protein meal, Bosch finding a
its permselective properties, and indirectly by aug- 50/mllmin increment in inulin clearance with an
menting the transcapillary convective flux of plasma 80-g protein meal, and Bergstrom a 15 mllmin in-
proteins. The result is enhanced protein filtration crease in response to 60 g of protein. Bergstrom also
with accumulation in the mesangium. This then reported that the response to an acute protein load
serves as a stimulus to proliferation of mesangial was independent of previous dietary intake. 17
cells and matrix, culminating in glomerulosclerosis. These observations, together with anecdotal
As functioning glomeruli are destroyed, less severly reports by several nephrologists of a beneficial effect
afflicted gomeruli undergo compensatory hyper- of dietary protein restriction on the course of human
filtration, establishing a positive feedback loop re- renal disease,tB have culminated in a massive re-
sulting in progressive glomerular injury and search effort, first in animals and then in humans.
eventual loss of renal function. 12 Protein restriction
should normalize glomerular pressures and flows, Animal Models of Renal Disease
thus breaking this cycle (Figure 1).10 Animal research in this field has primarily used a
variety of rat models of renal disease, although more
Observations recent investigators have reported results in dogs.
It is now well established that removal of one Studies that bear on the possible role of dietary pro-
human kidney will result in "compensatory" hyper- tein restriction fall into one of five major categories.
trophy of the other. Glomerular plasma flow and 1. Nephron Loss Results In Progressive Renal
filtration rate increase by about 40%.12 further- Insufficiency. The most common technique used to
more, pathologic studies in a variety of human renal produce nephron depletion is subtotal nephrectomy
diseases reveal marked hypertrophic changes, pre- or infarction. In 1932, Chanutin and Ferris first

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES 329

Protein Restriction In Renal Failure

LONGSTANDING
REDUCTION IN ~---------------------
RENAL MASS

~
CHRONIC RENAL
VASODILATION

INCREASED
GLOMERULAR /
./ ~ELEVATED GLOMERULAR
BLOOD FLOW PRESSURES
I I
I
~
GLOMERULAR
1
DIMINISHED DIRECT
HYPERFILTRATION PERMSELECTIVITY CELLULAR
INJURY

\ / INCREASED
PROTEIN FLUX
CELLULAR
PROLIFERATION

I~
PROTEINURIA MESANGIAL
/\ GLOMERULAR _____ _
CELL INJURY SCLEROSIS

\/
INCREASING
MESANGIAL
MATRIX
Figure 1. Possible long-term effects of compensatory Increases In glomerular pressures and flows. (Adapted from
Brenner, 1986.1Q )

demonstrated in the rat that infarction of greater mesangial matrix were evident, and endothelial and
than five-sixths of the total renal mass was associ- epithelial cells disappeared from these areas. The
ated with proteinuria, hypertension, and progres- increased matrix eventually occluded capillary lu-
sive renal dysfunction. 4 Shimamura and Morrison mens and Bowman's space, leading to the formation
carefully documented the development of glomerular of obsolescent giomerulL 19
pathologic features in the same mode1. 19 They were Progressive injury to remnant glomeruli in the
able to demonstrate, as early as the 10th week, an renal ablation model occurs at a rate proportional to
increase in glomerular size and hypertrophy of the the amount of tissue ablated surgically or by in-
visceral glomerular epithelial cells. Significant farction. This injury also is reflected in increasing
glomerular hyalinization started on the 25th week, proteinuria. Olsen demonstrated that there was a
and gradually became more extensive. On electron four-fold increase in the protein excretion of rats
microscopic examination, the hypertrophic glomeru- following 90% renal ablation. When corrected for the
lar epithelial cells showed many osmophilic bodies reduced nephron mass, this amounted to a 20-fold
in their cytoplasm, and fusion of foot processes. Be- increase in protein excretion per nephron. Studies
ginning on the 30th week, areas of increasing using macromolecular tracers found defects in both

330 November 1987 Volume 294 Number 5

 Zeller

the charge and size-selective properties of the glo- proportional to the increase in systemic blood pres-
merular basement membrane to account for this. 20 sure. In contrast, Azar found that once nephron loss
2. Hemodynamic Abnormalities Associated with was established, renal autoregulation failed, and
Nephron Depletion. Kaufman examined the func- glomerular blood flow increased several-fold second-
tional and hemodynamic abnormalities that occur in ary to marked intrarenal vasodilation. 24
the renal ablation model. 21 After removal of 50% of Glassock 25 found similar hemodynamic abnormal-
the renal mass, mean nephron GFR increased by ities in rats with nephrotoxic serum nephritis, as did
60%, and after ablation of 75% of the tissue, it in- Hostetter in rats made diabetic with streptozotocin
creased to 150%. These changes were paralleled by and then treated with insulin to maintain moderate
increases in renal growth. In comparison, mean glo- hyperglycemia. 26
merular blood flow rose 90% and 240%, after 50% 3. Increased Renal Blood Flow Correlates with Pro-
and 75% nephrectomy, respectively. Using labeled gressive Renal Damage. Substantial evidence
microspheres to examine intrarenal blood flow dis- suggests that the hemodynamic adaptations to renal
tribution, Kaufman found a disproportionate rise in damage that occur in response to a disease state such
blood flow to the inner cortex, following partial as diabetes, are, in and of themselves, causative fac-
nephrectomy, and proposed that this was secondary tors in progressive renal deterioration. Steffes found
to marked intrarenal vasodilation. 21 that unilateral nephrectomy in the diabetic rat
At approximately the same time, Brenner charac- greatly hastened the development of typical diabetic
terized the physiologic determinants of GFR as the lesions. 27 This effect was independent of systemic
following: blood pressure. In 1978, Mauer studied the effect of
1. Initial glomerular plasma flow, Goldblatt hypertension on glomerular pathology in
2. Systemic oncotic pressure, diabetic rats. 28 After 4 months of diabetes, the glo-
3. Glomerular transcapillary hydraulic pressure meruli of the unclipped kidney of hypertensive di-
gradient, and abetic rats had markedly increased diabettc.changes,
4. Glomerular capillary ultrafiltration coefficient including mesangial matrix thickening and me-
(K r), which represents the product of glomeru- sangial immunoglobin and complement localization,
lar capillary hydraulic permeability and the when compared with glomeruli of the contralateral
total surface area available for filtration. 22 clipped kidney. More importantly, glomeruli of the
Deen studied glomerular hemodynamics in rats 3 clipped kidney in hypertensive diabetic rats had less
weeks after unilateral nephrectomy.23 Using micro- mesangial thickening than glomeruli of normoten-
puncture techniques, he demonstrated that the sive diabetic rats. The authors concluded that alter-
elevated GFR in the residual kidney was associated ations in nephron hemodynamics combine with the
with an 80% increase in single-nephron GFR diabetic state to influence the rate of development of
(SNGFR). This was the result of two factors. First, diabetic glomerulopathy in rats. 28 Similarly, during
the plasma flow rate in single glomeruli was in- an autopsy on a patient with diabetes and unilateral
creased proportionately to the GFR because of pro- renal artery stenosis, Berkman and Rifkin observed
nounced intrarenal vasodilation of both afferent and that Kimmelstiel-Wilson lesions were present in the
efferent arterioles. Secondly, the glomerular trans- kidney with the patent renal artery and absent on
capillary hydraulic pressure gradient was increased the contralateral side with the tight renal artery
and also contributed to the hyperfiltration. Kr and stenosis. 29
systemic oncotic pressure did not vary. Therefore, Since the articles by Azar and Steffes, numerous
the increased GFR and SNGFR found in the renal investigators have reported deleterious effects of
ablation model were a consequence of the increased unilateral nephrectomy on the progressive renal
glomerular pressures and flows resulting from renal failure associated with a variety of diseases in the
vasodilation. 23 rat, suggesting that the hemodynamic abnormal-
Since that time, luJ.merous other investigators ities produced by nephron loss have broad implica-
have demonstrated similar hemodynamic abnormal- tions.
ities in other rat models of renal disease. Azar 4. Protein Restriction Modifies Hemodynamic Ab-
showed that intrarenal hemodynamics varied normalities. Hostetter explored the effect of reducing
greatly between desoxycorticosterone (DOC) salt hy- dietary protein intake on the hemodynamic abnor-
pertensive rats with normal renal function, and malities found in the subtotal nephrectomy model.
those in which mild dysfunction was produced by By putting rats with a 1 5/6 nephrectomy on a low
unilateral nephrectomy.24 In established hyper- protein (6%) diet for 14 days, he was able to maintain
tension in animals with intact kidneys, there is no SNGFR at a level not significantly different from
increase in glomerular pressure, and glomerular that of control rats without a nephrectomy fed a nor-
plasma flow tends to be decreased. This primarily is mal protein diet. This was associated with propor-
the result of an increase in intravascular resistance tional reductions in the glomerular transcapillary

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES 331

Protein Restriction In Renal Failure

hydraulic pressure gradient and glomerular capil- restriction also have been evaluated in other rat
lary blood flow. Afferent and efferent arteriolar re- models of renal disease. Neugarten looked at the
sistance (RA and RE) almost doubled, with no change effects of low- (4.6%), standard (23%), and high-
in K r. Systemic blood pressure was not significantly (57.5%) protein diets on the course of nephrotoxic
different between groups.30 serum nephritis.38 This model of immune nephritis
Subsequently, Wen went on to examine the effect was produced by injecting goat serum with anti-rat
of a similar low protein diet on the elevated GFR and glomerular basement membrane (GBM) antibody
renal plasma flow found in diabetic rats with mod- activity, followed by infusion of rabbit anti-goat
erate hyperglycemia. Once again, a 40% reduction in gamma globulin. Neugarten found that proteinuria,
these parameters occurred on the low protein diet. 31 which exceeded 275 mg/24 hr at the start of dietary
Recently, Dworkin found a marked reduction toward intervention (ie, while on standard chow), remitted
normal of the elevated glomerular transcapillary hy- to normal levels « 10 mg/24 hr) in rats fed the low-
draulic pressure gradient found in uninephrecto- protein diets, wheras heavy proteinuria continued in
mized DOC-salt rats and spontaneously hyperten- nephritic rats fed the 57% and 23% protein diets,
sive (SHR) rats fed a low protein diet for several with excretion rates of 289 and 133 mg/24 hr, re-
weeks following surgery.32 spectively, at 60 days. More importantly, creatinine
5. Protein Restriction Modifies the Outcome of Renal clearance was not different at this time between
Disease. Hostetter initially reported that the early nephritic and control rats both fed the 4% low-
morphologic abnormalities found in rats subjected protein diet. Thus, nephritic rats fed the low-protein
to a 1 5/6 nephrectomy were ameliorated by placing diet demonstrated no loss of renal function. In con-
the animals on a low-protein diet. 30 Likewise, the trast, nephritic rats fed the standard protein (23%)
20-fold increase in protein excretion found in control or high-protein (57%) chow lost approximately 25%
rats was eliminated. Kenner compared the effect of of their baseline renal function, a difference that was
modest protein restriction (14%) to a high protein highly significant (Table 1).38
diet (37%) in this model (5/6 nephrectomy), and Histologic findings at sacrifice confirmed minimal
found a significantly higher rate of protein excretion changes in the low-protein group. Progressive
in rats fed a high protein diet, as well as a four-fold mesangial expansion, focal and segmental prolif-
increase in the rate of progression to renal failure eration, and sclerosis correlated in severity with in-
and a seven-fold increase in mortality.33 Similar creasing dietary protein. Systemic blood pressure
results have been found by Kleinknecht,34 Laouari,35 and dietary caloric and mineral contents were iden-
El-Nahas,36 and Kikuchi. 37 tical in all groups. It should be recognized, however,
The long-term consequences of dietary protein that the 4% low-protein diet used in this study Js

TABLE 1
The Effect of Dietary Protein Content on NSN:
Clinical and Laboratory Features at 60 days
Creatinine
Body Blood Urinary Serum Clearance
Weight Pressure Protein Albumin (ml/24 hr/100 9
(g) (mmHg) (mg/24 hr) (g/dl) body weight)
Control, 4% 315 100 2 2.6 297
Control,20% 371 115 11 * 3.3t 460*
Control,50% 418 108 10t 3.0t 471*
NSN,4% 329 103 7 2.6 280
NSN, 20% 393 100 133 2.6 294
NSN, 50% 448 107 289 2.5 362

•p < 0.05 compared to the corresponding nephritic dietary group.

t p < 0.001 compared to the corresponding nephritic dietary group.
Adapted from Neugarten. 1983. 38

332 November 1987 Volume 294 Number 5

 Zeller

insufficient to support normal growth and develop- hypophosphatemia, marked anorexia, and growth
ment. Thus, renal function was stabilized, but possi- arrest. When animals subjected to subtotal nephrec-
bly at the expense of adequate nutrition.3s tomy were put on a normal or high-phosphorus diet,
Wen demonstrated that proteinuria and early there was little difference in survival, despite a sig-
morphologic abnormalities were prevented by di- nificant increase in mean phophorus intake per day,
etary protein restriction in diabetic rats. 31 More if total food intake was restricted to that of rats fed a
recently, Rennke confirmed that these early obser- very low-phosphorus diet. 42
vations translate into long-term benefits. 39 They In 1983, Kikuchi looked at survival following sub-
reported preliminary results of a morphometric anal- total renal ablation in rats fed one of four diets: a
ysis performed in four groups of animals followed high-protein, normal phosphorus diet; a high-
for 12 months: moderately hyperglycemic rats on a protein, low-phosphorus diet; a low-protein, normal
diet containing either 50% or 12% protein, and non- phosphorus diet; and a low-protein, low-phosphorus
diabetic controls on similar diets. There was no dif- diet. 37 The high- and low-protein diets contained
ference in GBM thickness, glomerular volume, me- 24% and 6% protein, respectively, and the non-
sangial volume, segmental glomerular obsolescence, restricted and phosphorus-restricted diets, 0.5% and
or urinary albuminl24 hr between diabetic rats fed 0.12% phosphorus, respectively. They reported that
a low-protein diet and nondiabetic rats fed a high- . protein restriction alone had some benefit on renal
protein diet. Furthermore, diabetic rats fed a high- function and survival, but that combined phos-
protein diet developed far worse morphometric and phorus and protein restriction had a more dramatic
functional abnormalities than their low-protein effect.37 This study did not demonstrate comparable
counterparts. 39 food intake between groups, nor was renal function
Likewise, Dworkin found that normalizing the ele- found to be different by creatinine clearance.
vated glomerular transcapillary hydraulic pressure Gimenez found that when male rats with 70%
gradient in uninephrectomized SHR rats, by dietary renal ablation were fed a diet containing 2.2% phos-
protein restriction, prevented these rats from devel- phate (very high phosphorus), they had a progressive
oping proteinuria and morphologic evidence of glo- increase in serum creatinine. 43 '!\vo weeks after sur-
merular disease up to 31 weeks after nephrectomy.32 gery, their mean serum creatinine was three times
Uninephrectomized SHR rats fed normal chow devel- that of comparable, partially nephrectomized rats
oped increasing proteinuria and progressive mesan- fed a normal phosphate diet (0.5%). This deteriora-
gial expansion and glomerular sclerosis over time. tion in renal function was prevented by the intra-
peritioneal injection of 3-phosphocitric acid, an in-
Phosphorus Restriction-An Additional Benefit? hibitor of calcium-phosphorus precipitation and
In 1978, TheIs reported that dietary restriction crystallization. Care was taken to standardize the
of phosphate prevented proteinuria, renal calcifica- nutrient content of the diet. Recently, Lumlertgul
tion, histologic changes, and functional deteriora- compared two groups of rats subjected to a subtotal
tion and death from uremia in rats undergoing 1 3/4 nephrectomy, with one group receiving a phosphorus
nephrectomy.4o Animals fed a diet of normal calcium binder (DHAAA).44 Both groups of animals were pair-
and phosphate contents (1.4%, 0.5% respectively) fed identical diets. At 14 weeks, creatinine clearance
showed renal deposition of calcium and phosphorus in the phosphorus-restricted group was 282 f.LI/min
in cortical tubular cells, basement membrane, and versus 46 f.LlImin in the phosphate-replete group.
interstitium. Phosphate restriction (0.04%) mark- Histologic examination of renal tissue from the
edly reduced the amount of calcium and phosphate phosphate-replete rats demonstrated more severe
deposited in the kidney. Likewise, the mean serum glomerular sclerosis, interstitial inflammation with
creatinine in the animals fed the normal diet was fibrosis, and tubular atrophy and dilation, compared
3.75 mg/dL 4 months after surgery, as opposed to to the phosphorus-depleted group. It should be noted,
1.06 mg/dl in the phosphate-restricted animals. In however, that this study produced severe hypophos-
addition, 19 of the 26 animals on the normal diet phatemia in the low-phosphorus group (2.85 vs.
died, compared to only 2 of the 12 on the phosphorus- 6 mg/dl in normal rats), and may have had a variety
restricted diet. 4o Similar results were obtained by of effects, such as blunting normal tissue hyper-
Karlinsky using the same diets in rats with nephro- trophy of the remnant kidney or decreasing cardiac
toxic serum nephritis. 41 output and renal blood flow, that would not have
These results were challenged by Laouari, who been seen if phosphate had been less severely re-
felt that the protective effect attributed to phos- stricted. 2
phorus was, in fact, mediated by a decrease in food
intake. 42 Decreased protein was most likely to be the Summary
major beneficial factor. In this study, the authors There is now substantial evidence from a variety
demonstrated that the severe phosphorus restriction of rat models of renal disease for a clear-cut bene-
imposed by Thel and Karlinsky resulted in severe ficial effect of dietary protein restriction in rats. The
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES 333
Protein Restriction In Renal Failure

role of moderate phosphate restriction has yet to be Dietary histories for the prior 2 years were ob-
determined. The implications for human renal dis- tained at the start of the study for Group 3 patients
ease are less clear. The rat kidney tends to grow by a trained dietician. Protein intake was estimated
throughout the rat's lifetime, which is typically 24 at approximately 70 g/day, with a 900-mg mean
months. In addition, most normal rats who die of old phosphorus intake. This compares to the study diet,
age die from chronic renal insufficiency. By age 24 which consisted of 0.6 g/kg body weight/day protein
months, 60% to 100% of such animals have severe (60% to 70% high in biologic value) (equivalent to 40
glomerulosclerosis associated with proteinuria and g/day total in a 70-kg man) and 700 mg phosphorus.
hypertension. 45 Thus, a beneficial effect of dietary To put this in perspective, the current Recommended
protein restriction in the rat may not translate into a Daily Allowance (RDA) for protein is 0.8 g/kg/day.
beneficial effect in humans or other species. Most Amet:icans consume an average of 1.5 g/kg/day
Studies of dietary protein restriction in dogs have (or 105 g/d'ay in a 70-kg man).49 In this study, renal
yielded confusing results. In 1983, Polzin evaluated function deterioration was calculated using linear
the effects of low-, medium-, and high-protein diets regression analysis of the relationship between time
on dogs in which chronic renal failure was induced and the reciprocal of serum creatinine for each
by a combination of renal artery ligation and con- group. The follow-up period was 18 to 76 months.
tralateral nephrectomy.46 He reported that dogs fed Data for the control group were obtained from retro-
the high-protein diet had a high mortality rate (50%) spective chart review. Mean blood pressure was com-
resulting from uremia. This occurred in the first parable in all groups at the start of the study. These
several weeks following renal insult, suggesting that results (Table 2)48 demonstrate a 20-fold increase in
something else was happening in these dogs, since the rate of progression of the control group over the
the remaining dogs in that diet group (50% of total) two treatment groups. There was no evidence of pro-
demonstrated stability of renal function over a gressive phosphorus or protein depletion in the two
40-week period. More recently, Robertson (1986) treatment groups, although specific nutritional
reported no association between dietary protein in- assessment techniques were not reported. 48 Limita-
take and proteinuria, morphologic abnormalities, tions of this particular study include the retro-
and renal function decline in dogs subjected to 75% spective controls, the diversity of renal diseases
nephrectomy (50% residual renal function) and fol- analyzed, and the lack of continued blood pressure
lowed for 4 years. 47 Most surviving dogs maintained reporting during the study. In addition, compliance
stable renal function over this period, although dogs data were not presented, and renal function changes
fed a high-protein diet had a higher GFR as a group were reported only as changes in the reciprocal of
before and after surgery. While these results indi- serum creatinine.
cate that the dog is different from the rat in response In 1983, several researchers published their indi-
to 75% renal ablation, the author correctly points vidual experiences with protein-restricted diets in a
out that they do not necessarily rule out a beneficial special edition of Kidney International. The study
effect of protein restriction on progressive renal in- reported by Barsotti in this issue was the first at-
sufficiency. Progressive deterioration of renal func- tempt at a controlled tria1. 50 In this study, 19 con-
tion was not achieved in these dogs. It may well be trols with various renal diseases were followed
that a greater portion of renal mass would need to monthly for 11 months while they were on a diet
be removed to trigger deterioration, or a longer time containing 0.8 g/kg/day protein and 12 mg/kg/day
interval could be required. These experiments have phosphorus. The treatment group consisted of 20
yet to be done.

Human Studies
In 1982, Maschi048 reported the results of a retro- TABLE 2
spectively controlled trial designed to assess the ef- Progression of Renal Failure as a
fects of combined dietary protein and phosphorus
restriction on the progression of human renal dis- Function of Protein Intake
ease. He studied 75 patients with a variety of renal Slope of 1/cr* VS, Time
diseases, ranging from chronic pyelonephritis to
chronic glomerulonephritis. Patients were divided Group 1 -0.0008
into three groups as follows: Group 2 -0.0010
Group 1: 25 patients; creatinine, 1.5-2.7; low- -0.020
Group 3
protein, low-phosphate diet.
Group 2: 20 patients; creatinine, 2.9-5.4; low-
• cr = creat/nlnes
protein, low-phosphate diet.
Group 3: 30 patients; creatinine, 1.6-4.7; retro- Adapted from Maschlo, 1982,
spectively determined control diet.
334 November 1987 Volume 294 Number 5
 Zeller

patients with comparable diseases followed on a diet quality protein) with a ketoanalogue, amino acid
consisting of 0.5 g/kg/day protein and 7 mg/kg/day mixture (11 g) in 24 patients with severe renal fail-
phosphorus. Prior to study, these patients had been ure (creatinine clearance, 2-15 mllmin).51 Looking
following the control diet. The mean creatinine at the 17 patients for whom prior information was
clearance in each group at the start of the study was available (mean pre-study intake: 56 g protein/day),
approximately 30 ml/min, and renal function was he compared their course pre- and post-institution of
followed by serum creatinine and creatinine the diet. The reciprocal of serum creatinine re-
clearance. Blood pressure was comparable over the mained stable in six of seven patients, with an initial
study interval. Compliance with the protein re- serum creatinine!5 8 mg/dl for an average of2 years.
striction was verified by assessment otmean urinary This was verified further by measurement of cre-
urea nitrogen at monthly intervals. Results are atinine (cr) clearance. Four additional patients
given in Table 3. experienced a significantly decreased rate of pro-
Limitations of this study include the lack of gression compared with their previous course, as
randomization and the large category of renal assessed by lIcr versus time. Compliance was as-
diseases reported as "chronic glomerulonephritis" sessed by urinary urea nitrogen measurements and
without further, specific details. Two patients in the dietary history. Whenever nitrogen intake esti-
control group had an unexpectedly rapid deterio- mated by these techniques differed by more than
ration after entering the study (ie, slope changes 20% from prescribed, the diet was reviewed by the
despite no change in diet). Without these patients, research dietician, and appropriate adjustments
the rates of progression would have been slower. were made. Limitations of this study include the
Finally, compliance data were not normalized for absence of a prospective, randomized control group,
body weight, so true compliance could not be as- use of serum creatinine and creatinine clearance as
sessed. Simply put, this means that a 50-kg patient the sole markers of GFR in patients with advanced
eating 2 g/kg/day protein could not be differentiated renal insufficiency, no mention of blood pressure,
from a 100-kg patient eating 1 g/kg/day. Thus, the no reported compliance data, and no anthropometric
data reported could have been achieved despite gross assessment.
noncompliance. The paper by Rosman in 1984 is the only published
In 1984, Mitch reported the effects of supple- prospective, randomized, controlled trial of dietary
menting a very low-protein diet (20-30 g mixed- protein restriction in renal failure. 52 Two hundred
twenty-eight patients with renal insufficiency of di-
verse causes were stratified for age, sex, and renal
function, and then randomized to one of four groups:
TABLE 3 A1: 70 patients; creatinine clearance, 31-60 mll
minl1.73m2; control diet
Progression of Renal Failure as a A2: 40 patients; creatinine clearance, 10-30 mll
Function of Protein Intake minll. 73m2; control diet
B: 66 patients; creatinine clearance, 31-60 mll
A Serum A Creatinine min/l.73m2; low-protein diet
Creatinine Clearance
(mg/dllmonth) (mllmin/month)
C: 52 patients; creatinine clearance, 10-30 mll
minll.73m 2; low-protein diet
Controls The mean protein intake of Group A1 patients was
Pre-study 70 g/day versus 0.6 g/kg/day (40 g/day in a 70-kg
0.8 g/kg/protein. +0.08 ± 0.07 -0.50 ± 0.66 man) in the corresponding treatment group B. The
12 mg/kg phosphorus mean protein intake of Group A2 patients was 55
During study g/day versus 0.4 g/kg/day (30 g/day in a 70-kg man)
0.8 g/kg/protein. +0.08 ± 0.01 -0.44 ± 0.10 in Group C. All patients were treated with alumi-
12 mg/kg phosphorus num hydroxide to maintain normal serum phos-
Treatment group phorus levels. Blood pressure was lowered to com-
Pre-study parable levels in all groups throughout the study,
0.8 g/kg/proteln. +0.073 ± 0.05 -0.59 ± 0.7 and compliance was assessed by urinary urea nitro-
12 mg/kg phosphorus gen, although this was not normalized for individual
During study weight. Renal function deterioration was followed
0.5 g/kg/proteln. -0.03 ± 0.08 +0.1 ± 0.4 over a minimum of 18 months, and was reported as
7 mg/kg phosphorus the change in the reciprocal of serum creatinine with
time (Table 4).52.
Adapted from 80(50ttI1983.50 The authors noted that protein restriction re-
sulted in a significant decrease in proteinuria

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES 335

Protein Restriction In Renal Failure

probably should be excluded, at least in initial evalu-

ations.
TABLE 4 One technique commonly used to assess renal
Progression of Renal Failure as a failure progression is to plot the reciprocal or the
Function of Protein Intake logarithm of plasma creatinine against time. Mitch
and Rutherford have shown that more than 80% of
Slope of 1/cr* vs. Time patients with a variety of different diseases will have
CrClt, 10-30 ml/mln crC!. 31-60 ml/mln a linear course when plotted in this fashion. This
implies a constant rate of loss with the reciprocal
Control -0.037 -0.036 method, and a constant fractional loss with the loga-
Treated -0.016 -0.007 rithm method. 53 -55
p<0.01 p<0.05 In a majority of studies, reciprocal creatinine con-
centrations pre- and post-institution of a low-protein
diet have been compared. 53 Unfortunately, serum
• cr = creatinine.
creatinine is not an adequate measure of changes in
t erGi = creatinine clearance.
GFR, particularly when dietary protein is altered.
Adapted from Rosman, 1984. 52 Creatinine is formed from the nonenzymatic irre-
versible dehydration of creatine. About 2% of total
body creatine is converted to creatinine daily. Cre-
(p::5 0.02), but specific data were not reported. There atine is produced by endogenous synthesis from
were no differences in the following parameters at amino acids, but also is ingested preformed in the
9-month intervals: body weight, blood pressure, he- diet. Thus, a low-protein diet entails an immediate
matocrit, serum calcium, and serum albumin. Group reduction in the creatine pool, which can result in a
compliance, assessed at 3-month intervals, was 15% or greater loss in creatinine production per 24
greater than 90%. While this is clearly the most hours. This means that serum creatinine will fall
thorough study to date, it still has limitations. The when a patient begins a low-protein diet despite no
randomization protocol did not produce an even dis- change in GFR. 53 •56 While Mitch and Walser have
tribution of diagnoses between groups, and since the reported that a new steady state of creatinine excre-
numbers were small, this could be important. Like- tion is achieved within 4 months, this will only be
wise, a number of patients were classified simply as true if muscle mass remains constant, a hypothesis
"chronic glomerulonephritis," as in prior studies. to be tested in the long-term evaluation of low-
Protein intake was not normalized for body weight protein diets. 57
in controls, and compliance in treated patients was Direct measurement of GFR would be preferable
not normalized. Creatinine clearance data were not to measurement of serum creatinine. Creatinine
reported, nor was any other direct assessment of clearance may not be the best marker for reasons
GFR. Finally, the nutritional adequacy of the diet noted previously, as well as the fact that GFR usu-
can be questioned. While weight and serum albumin ally is overestimated when there is poor renal func-
did not change, no other nutritional information tion. It is not clear whether tubular handling of
such as anthropometric data was supplied. The con- creatinine also might be affected by protein intake.
tinuous fall in total 24-h urine creatinine output of For these reasons, clearance of a radioactive marker
Group C patients over the study interval suggests such as 1251 iodothalamate, which behaves much like
progressive muscle wasting, which is not surprising, inulin, would be optima1. 53
since the 0.4 g protein/kg/day supplied by this diet is 2. What are the effects of other variables? In a re-
well below minimum recommendations to maintain cent paper, Bergstrom reported that the progression
neutral or positive (+) nitrogen balance in most of renal failure was significantly slower in control
normal individuals. patients after they had entered a prospective study
that called for more frequent follow-up visits. They
Unanswered Questlons53 also found a significant association between the re-
1. How should we assess the progression of chronic tardation of progression and improved blood pres-
renal failure? In order to evaluate the effectiveness sure contro1. 58 It is likely that close monitoring of
of a specific therapy on the course of renal disease, other variables, such as urinary tract infections
we need to deal with a relatively homogeneous popu- and calcium and phosphorus balance has a beneficial
lation at the start. In addition, patients should dem- effect. Future studies will need to control for these
onstrate progressive loss of renal function over time. variables. 53
Disease processes such as membranous glomeru- 3. How should compliance be assessed? As men-
lonephritis that exhibit significant variability in tioned previously, urinary urea nitrogen has been
their course (one-third of patients remain stable) used frequently to assess compliance. Changes in

336 November 1987 Volume 294 Number 5

 Zeller

dietary protein intake are reflected primarily in

urea nitrogen excretion. Non-urea losses tend to
remain relatively fixed, and can be estimated by a TABLE 5
factor proportional to weight. Thus, urea nitrogen Progression of Renal Failure as a
appearance (UNA) is an objective way to assess Function of Protein and
compliance accurately, provided that (1) complete Phosphorus Intake
urine specimens are obtained, (2) allowance is made
for changes in body weight and total body water, and Creatinine Clearance Changes
(3) results are normalized for body weight and take (ml/min/mo)
into account differences in proteinuria. This, in com-
bination with frequent dietary histories taken by a Free Diet Controlled Diet
trained dietician, should provide a reasonable as- ClND* -0.79 -0.53
sessment of protein intake. 57 lPlNDt -0.90 -0.07
4. How should nutritional status be followed? If
low-protein diets are to be regarded as safe, it is • CLND = diet with a moderate phosphorus res/ric/Ion.
necessary to show that malnutrition does not occur.
A number of different parameters have been evalu- t LPLND = die/ with a severe phosphorus res/ric/Ion.
ated to assess nutritional status. These include body Adap/ed from Barsotti, 1984, ~9
weight, anthropometry, serum albumin and trans-
ferrin, and nitrogen balance. Substantial muscle
loss without a change in serum albumin and trans- and other nutrients. The RDA for protein for normal
ferrin has been observed by several investigators in adults is 0.8 g/kg/day.49 The RDA is based on the
patients on some low-protein diets, indicating the average protein (N) lost by individuals on a
need for careful anthropometric follow-up. Formal protein-free diet (0.45 g/kg/day) plus two standard
nitrogen balance studies can be done only at infre- deviations (0.15 g/kg/day), to cover 97% of the popu-
quent intervals and may not reflect intermittent lation, plus an increment (0.2 g/kg/day) to cover the
periods of catabolism. Combined serial measure- less efficient use of dietary protein as intake is
ments over time offer the most complete nutritional increased from a point of negative ( - ) nitrogen bal-
information. 53 ance to neutral or ( + ) nitrogen balance.6o Most, but
5. Is there an additional benefit of phosphorus not all, investigators have found that protein intake
restriction? Until recently, all studies of dietary pro- can be reduced safely to 0.55 to 0.6 g/kg/day, pro-
tein restriction combined this with moderate phos- vided that 60% to 70% of the protein is high in bio-
phorus restriction. Barsotti attempted to clarify this logic value (high content of essential amino acids).
issue by comparing the decline in creatinine clear- Such diets also may require a higher than usual
ance between two groups of patients fed low-protein total energy intake (35-50 kcallkg/day) to maintain
diets: one with a moderate phosphorus restriction neutral N balance. 61- 63 It is unclear that this level of
(CLND) and the other severely restricted in phos- intake will provide ( + ) nitrogen balance in all pa-
phorus intake (LPLND).59 Creatinine clearance data tients with chronic renal insufficiency over extended
were compared for each patient pre- and post- periods of time. This is particularly questionable in
institution of the diet (Table 5).59 While the authors patients with other medical problems. Thus, there is
conclude that these results demonstrate an addi- a need for careful, short-term balance studies and
tional beneficial effect of phosphorus restriction, this more long-term nutritional evaluation.
study suffers from many of the limitations described Ketoanalogues of essential amino acids may allow
previously. The control data appear retrospective, patients to achieve nitrogen balance, despite a lower
and the patients were not randomized. There was no daily nitrogen intake. Several studies have sug-
assessment of blood pressure during the control or gested that uremic patients actually may prefer to
treatment periods, and information concerning ingest 0.3 g/kg/day protein of mixed quality with
etiologic factors of renal insufficiency is scanty. supplements of essential amino acids or keto-
Thus, it remains to be determined whether or not analogues in preference to a 0.6 g/kg/day (60-70%
phosphorus plays an independent role in the pro- HBV) diet. 64.65 Whether this will be true in otherwise
gression of chronic renal failure. Future studies healthy patients with moderate renal insufficiency
should address this issue, as well as that of dele- is not known. It is also unclear whether the further
terious side effects. nitrogen reduction offered by ketoanalogues will be
6. What is the "best" low-protein diet? If low- beneficial, or whether these medications will have
protein diets are found to be effective in slowing the deleterious effects. Thus, each diet will need to be
progression of renal insufficiency, it will be neces- evaluated separately in a variety of diseases before
sary to determine the optimal daily intake of protein specific dietary recommendations can be made.

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES 337

Protein Restriction In Renal Failure

Dietary Protein and Altered Renal However, prostaglandins frequently function as

Hemodynamics-Possible Mechanisms modulators of the action of other vasoactive sub-
The mechanism by which ingested protein en- stances such as angiotensin II. Thus, the reported
hances GFR has not been established, although ab- increases in local prostaglandin production simply
sorption of amino acids from the intestine into the may represent a response to an as yet unidentified
circulation appears to be required. O'Connor and primary effector system.
Summerill have clearly shown that the rise in GFR Finally, in a series of elegant experiments, Seney
following a meat meal in dogs cannot be reproduced demonstrated that the tubuloglomerular feedback
by feeding urea, sulfate, or acid in quantities equiv- (TGF) system in rats is rendered less responsive by
alent to that produced by catabolism of the mea1. 66.67 a high-protein diet. 78 This system adjusts glomeru-
Moreover, increases in GFR in dogs and humans are lar vascular resistance and the glomerular ultra-
induced by intravenous infusion or oral intake of filtration coefficient (two determinants of GFR
mixed or individual amino acids. Free amino acids known to be altered by protein intake) in response to
probably do not influence the GFR directly, since variations in some correlate of the early distal tu-
unilateral renal artery infusion of amino acids in the bule flow rate. (As distal tubular flow increases,
dog produces bilateral increases in GFR. 68 Similarly, GFR normally falls.) Thus, a reduction in function of
studies in the isolated perfused kidney of the rat the TGF system by high dietary protein intake re-
demonstrate little direct effect of infused amino duces the normal suppression of GFR effected by TG
acids on filtration. 69 For this reason, it has been sug- feedback, causing the filtration rate to rise. 78 Other
gested that a circulating hormone is responsible. investigators have found that single-nephron GFR
Both growth hormone (GH) and glucagon have been remains elevated even when the effects of TGF are
suggested as possible effectors, since each is released eliminated, suggesting that additional factors are
in response to amino acid infusion or oral protein also operative. 79
ingestion, and each has been shown to produce a rise
in GFR when infused. 68 Moreover, somatostatin, studies In Progress
which inhibits the endogenous release of both hor- The National Institutes of Health is currently
mones, prevents protein-induced renal hyperemia. 71 sponsoring a multicenter trial entitled "The
Recently, Hirschberg found that normal and GH- Modification of Diet in Renal Disease (MDRD).,,80
deficient adults demonstrated a comparable rise in The purpose of this study is to evaluate the effects of
GFR and renal blood flow (RBF) during arginine two low-protein diets on the course of mild, mod-
infusion, thus negating the importance of GH as an erate, and severe renal insufficiency in a variety of
effeCtor in this response. 72 Premen looked at the ef- diseases. Nine centers are involved, and a total of
fect of glucagon infusion at a level comparable to 540 patients will be enrolled over the next 3 to 4
that found postprandially in dogs. Neither RBF nor years. Patients are evaluated by retrospective chart
GFR was altered significantly from its respective review to determine that their disease is indeed pro-
fasting control value. A plasma glucagon concentra- gressive. Active collagen vascular disease, systemic
tion more than ten times greater than that mea- infection, uncontrolled hypertension, and insulin-
sured postprandially was required to achieve the rise requiring diabetes mellitus are criteria for exclu-
in GFR demonstrated after the meat meals. Thus, sion. Patients aged 18 to 40 years with GFRs 25-80
glucagon does not appear to be the primary mediator mllmin/1.73m 2and patients aged 41 to 75 years with
of protein-induced renal hyperemia. 71 GFRs 25-(120-age) mllmin/1.73 m 2 will be random-
Other investigators have suggested that angio- ized to a low-protein diet containing 0.55 to 0.6 g/kg
tensin II, which appears to be generated locally in protein and 5 to 10 mg/kg phosphorus; a very low-
the kidney as well as in the systemic circulation, and protein diet containing 0.28 g/kg protein and 4 to 9
renal prostanoids may represent a common effector mg/kg phosphorus (supplemented with amino acids
system by which many hormones influence the GFR. and ketoanalogues); or a normal protein diet con-
Acute and chronic protein loading in normal and taining 1 to 1.4 g/kg protein and 16 to 20 mg/kg
subtotal nephrectomized rats is associated with a phosphorus. Patients with GFRs between 10 and 24
significant rise in the glomerular production of the mllmin/1. 73 m 2 will be randomized to either the 0.55
vasodilatory prostaglandins PGE 2 and PGI2.73-75 to 0.6 g/kg protein diet or one of the very low-protein
Similar results have been found in moderately hy- diets described previously. Patients will be followed
perglycemic diabetic rats, which also exhibit renal monthly, and compliance will be monitored by uri-
hyperperfusion and hyperfiltration. 73 .77 Administra- nary urea nitrogen and dietary history. Decline in
tion of the nonspecific prostanoid synthesis inhibi- renal function will be followed by iodothalamate
tor, indomethacin, significantly reduces prostaglan- clearance at regular intervals, and nutritional status
din production and prevents the rise in GFR assoc- assessed by repetitive anthropometric evaluation
iated with protein ingestion or the diabetic state. 73 •76 and serum protein determinations. Total follow-up

338 November 1987 Volume 294 Number 5

 Zeller

time will range from 24 to 45 months. Thus, final 19. Shimamura T, Morrison AB: A progressive glomerulo-
results will not be available before 1991. sclerosis occurring in partial five-sixths nephrectomized rats.
Am J Pathol79(1):95-106, 1975.
At the University of Texas Health Science Center, 20. Olsen JL, Hostetter TH, Rennke HG, Brenner BM,
we now have a similar study ongoing in Type I di- Venkatachalam MA: Altered charge-and-size selective
abetic patients with moderate to severe renal insuf- properties of the glomerular wall: A response to reduced
ficiency. Patients are monitored by the same tech- renal mass. Kidney lilt 16:857, 1979.
21. Kaufman JM, Siegel NJ, Hayslett JP: Functional and hemo-
niques as in the MDRD study, and are randomized to dynamic adaptation to pl'ogressive renal ablation. Circ Res
a low-protein diet with or without phosphorus re- 36:286-293, 1975.
striction, or a normal protein diet. Results of both of 22. Brenner BM, Humes HD: Mechanics of glomerular ultra-
these studies may have a significant impact on the filtration. N ElIgl J Med 297:148-154, 1977 .
clinical management of progressive renal disease. 23 . Deen WM, Maddox DA, Robertson CR, Brenner BM: Dynam-
ics of glomerular ultrafiltration in the rat. VII. Response to
reduced renal mass. Am J Physiol227:556-562, 1974.
24. Azar S, Johnson MA, Hertel B, Tobian L: Single-nephron
References pressures, flows and resistances in hypertensive kidneys with
1. Private Communication, Department of Health and Human nephrosclerosis. Kidlley lilt 12:28-40, 1977.
Services, Health Care Financing Administration, Office of 25. Glassock R, Bennett C, Kayser B, Martin C, Marsh D: Glo-
Research and Demonstrations, August 1986. merular hemodynamics in nephrotoxic serum nephritis. Kid-
2. Klahr S, Buerkert J, Purkerson ML: Role of dietary factors in lIey lilt 14:725, 1978.
the progression of chronic renal disease. Kidney lilt 24: 26. Hostetter TH, Troy JL, Brenner BM: Glomerular hemo-
579-587,1983. dynamics in experimental diabetes. Kidlley lilt 19:410-415,
3. Newburgh LH: Production of Bright's Disease through feed- 1981.
ing high protein diets. Arch Illtem Med 24:359-377,1919. 27. Steffes MW, Brown DM, Mauer SM: Diabetic glomerulopathy
4. Newburgh LH, Clarkson S: Renal injury produced in rabbits following unilateral nephrectomy in the rat. Diabetes
by diets containing meat. Arch Illtem Med 32:850-869,1923. 27:35-41, 1978.
5. Moise TS, Smith AH: Effect of high protein diet on the kid- 28. Mauer SM, Steffes MW, Azar S, Sandberg SK, Brown DM:
neys. Experimental study. Arch Pathol4:530-542, 1927. The effects of Goldblatt hypertension on development of the
6. Chanutin A, Ludewig S: Experimental renal insufficiency glomerular lesions of diabetes mellitus in the rat. Diabetes
produced by partial nephrectomy. V. Diets containing whole 27:738-744,1978.
dried meat. Arch Illtern Med 58:60-80, 1936. 29. Berkman J, Rifkin H: Unilateral nodular diabetic glomeru-
7. Farr LE, Smadel JE: The effect of dietary protein on the losclerosis (Kimmelsteil-Wilson): Report of a case. Metabo-
course of nephrotoxic nephritis in rats. J Exp Med lism 22:715-722,1973.
70:615-627,1939. 30. Hostetter TH, Olson JL, Rennke HG, Venkatachalam MA,
8. Smadel JE, Farr LE: The effect of diet on the pathologic Brenner BM: Hyperfiltration in remnant nephrons: A poten-
changes in rats .with neprhotoxic serum nephritis. Am J tially adverse respnose to renal ablation. Am J Physiol
Pathol15:199-216,1939. 241:F85-F93, 1981.
9. Addis T: Glomerular Nephritis: Diagllosis and Treatmellt. 31. Wen SF, Huang TP, Moorthy AV: Effects of low-protein diet
New York, Macmillan, 1948. on experimental diabetic nephropathy in the rat. J Lab Clill
10. Brenner BM, Meyer TW, Hostetter TH: Dietary protein in- Med 106:589-597,1985.
take and the progressive nature of kidney disease. The role of 32. Dworkin LD, Feiner HD: Glomerular injury in uninephrec-
hemodynamically mediated glomerular injury in the patho- tomized spontaneously hypertensive rats. A consequence of
genesis of progressive glomerular sclerosis in aging, renal glomerular capillary hypertension. J Clill Illvest 77:797-809,
ablation, and intrinsic renal disease. N Engl J Med 1986.
307:652-659, 1982. 33. Kenner CH, Evan AP, Blomgren P, Aronoff GR, Luft FC:
11. Meyer TW, Lawrence WE, Brenner BM: Dietary protein and Effect of protein intake on renal function and structure in
the progression of renal disease. Kidlley lilt 24(S16): partially nephrectomized rats. Kidlley lilt 27:739-750, 1985.
S243-S247, 1983. 34. Kleinknecht C, Salusky I, Broyer M, Gubler MC: Effect of
12. Hostetter TH, Rennke HG, Brenner BM: The case for intra- various protein diets on growth, renal function, and survival
renal hypertension in the initiation and progression of di- of uremic rats. Kidlley lilt 15:534-541, 1979.
abetic and other glomerulopathies. Am J Med 72:375-380, 35. Laouari D, Kleinknecht C, Gubler MC, Broyer M: Adverse
1982. effect on proteins on remnant kidney: Dissociation from that
13. Mitch WE, Steinman TI: Can the course of chronic renal fail- of other nutrients. Kidlley lilt 24 (SuppI16):S248-S253 , 1983.
ure be altered by diet? Kidney 16:31-35, 1983. 36. El-Nahas AM, Paraskevakou H, Zoob S, Rees AJ, Evans DJ:
14. Blachley JD: The role of dietary protein in the progression Effect of dietary protein restriction on the development of
and symptomatology of chronic renal failure . Am J Med Sci renal failure after subtotal nephrectomy in rats. Clill Sci
288:228-34, 1984. 65:399-406, 1983.
15. Pullman TN, Alving AS, Dern RJ, Landowne M: The influ- 37. Kikuchi H, Matsushita T, Hirata K: Improved dietary treat-
ence of dietary protein intake on specific renal functions in ment with low protein and phosphorus restriction in uremic
normal man. J Lab Clill Med 44:320-332, 1954. rats. Kidlley lilt 24 (SuppI16): S254-258, 1983.
16. Bosch JP, Saccaggi A, Lauer A, Ronco C, Belledonne M, 38. Neugarten J, Feiner HD, Schacht RG, Baldwin DS: Ameliora-
Glabman S: Renal functional reserve in humans-effect of tion of experimental glomerulonephritis by dietary protein
protein intake on glomerular filtration rate. Am J Med restriction. Kidlley lilt 24:595-601,1983.
75:943-950, 1983. 39. Rennke HG, Sandstrom D, Zatz R, Meyer TW, Cowan RS,
17. Bergstrom J, Ahlberg M, Alvestrand A: Influence of protein Brenner BM: The role of dietary protein in the development of
intake on renal hemodynamics and plasma hormone concen- glomerular structural alterations in long term experimental
trations in normal subjects. Acta Med Scalld 217:189-196 , diabetes mellitus. Kidlley lilt 29:289,1986.
1985. 40. Ibels LS, Alfrey AC, Haut L, Huffer WE: Preservation offunc-
18. Walser M: Does dietary therapy have a role in the predialysis tion in experimental renal disease by dietary restriction of
patient? Am J Clill Nutr 33:1629-1637, 1980. phosphate. N ElIgl J Med 298:122-126, 1978.

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES 339

Protein Restriction In Renal Failure

41. Karlinsky ML, Haut L, Buddington B, Schrier NA, Alfrey 61. Ford J, Phillips ME, Toyee FE, Luch VA, Wardener HE:
AC: Preservation of renal function in experimental Nitrogen balance in patients with chronic renal failure on
glomerulonephritis. Kidney lnt 17:293-302, 1980. diets containing varying quantities of protein. Br Med J 1:
42. Laouari D, Kleinknecht C, Cournot-Witmer G, Habib R, Mou- 735-740,1969.
nier F, Broyer M: Beneficial effect of low phosphorus diet in 62. Bergstrom J: Discovery and rediscovery of low protein diet.
uraemic rats: A reappraisal. Clin Sci 63:539-548, 1982. Clill Nephrol 21:29-35, 1984.
43. Gimenez L, Walker WG, Tew WP, Hermann JA: Prevention of 63. Kopple JD, Sorensen MK, Coburn JW, Gordan S, Rubini ME:
phosphate-induced progression of uremia in rats by Controlled comparison of 20-g and 40-g protein diets in the
3-phosphocitric acid . Kidlley lilt 22:36-41, 1982. treatment of chronic uremia. Am J Clin NuIr 21:553, 1968.
44. Lumlertgul D, Burke TJ, Gillum DM, Alfrey AC, Harris DC, 64. Kampf D, Fischer H, Kessel M: Efficacy of an unselected
Hammond WS, Schrier RW: Phosphate depletion arrests protein diet (25 g) with minor oral supply of essential amino
progression of chronic renal failure independent of protein acids and keto analogues compared with a selective protein
intake. Kidney lnt 29:658-666,1986. diet (40 g) in chronic renal failure. Am J Clill Nulr 33:
45. Blachley JD: The role of dietary protein in the progression 1673-1677,1980.
and symptomatology of chronic renal failure. Alii J Med Sci 65. Lee HA, Talbot S, Rowlands A, Jackson MA: Dietary manage-
228:228-234, 1984. ment of chronic renal failure with oral amino acids. Nutr
46. Polzin DJ, Osborne CA, Hayden DW, Stevens JB: Influence of Metab 24:50-63, 1980.
reduced protein diets on morbidity, mortality, and renal func- 66. O'Connor WJ, Summerill RA: The excretion of urea by dogs
tion in dogs with induced chronic renal failure. Am J Vet Res following a meal of meat. J PhysioI256:93-102, 1976.
45:506-516,1984. 67. O'Connor WJ, Summeri11 RA: Sulphate excretion by dogs fol-
47. Robertson JL, Goldschmidt M, Kronfeld DS, Tomaszewski JE, lowing ingestion of ammonium sulphate 01' meat. J Physiol
Hill GS, Bovee KC: Long-term renal responses to high dietary 260:597-607,1976.
protein in dogs with 75% nephrectomy. Kidney lilt 68. Bergstrom J, Ahlberg M, Alvestrand A: Influence of protein
29:511-519, 1986. intake on renal hemodynamics and plasma hormone concen-
48. Maschio G, Oldrizzi L, Tessitore N, D'Angelo A, Valvo E, Lupo trations in normal subjects. Acta Med Scalld 217:189-196,
A, Loschiavo C, Fabris A, Gammaro L, Rugiu C, Panzetta G: 1985.
Effects of dietary protein and phosphorus restriction on the 69. Maack T, Johnson V, Tate SS, Meister A: Effects of amino-
progression of early renal failure. Kidney lnt 22:371-376, acids on the function of the isolated perfused rat kidney. Fed
1982. Proc 33:305, 1974.
49. National Research Council, Food and Nutrition Board, Rec- 70. Johannesen J, Lie M, KiiI F: Effect of glycine and glucagon on
ommended Dietary Allowances, 9th ed. Washington, DC, Na- glomerular filtration and renal metabolic rates. Am J Physiol
tional Academy of Science, 1980. 233:F61-F66,1977.
50. Barsotti G, Morelli E, Giannoni A, Guiducci A, Lupetti S, 71. Premen AJ, Hall JE, Smith MJ: Postprandial regulation of
Giovannetti S: Restricted phosphorus and nitrogen intake to renal hemodynamics: Role of pancreatic glucagon. Alii J
slow the progression of chronic renal failure: A controlled PhysioI248:F656-F662, 1985.
trial. Kidney lnt 24(Suppl 16):S278-S284, 1983. 72. Hirschberg R, Kleinman K, Giassock R, Kopple JD: Effect of
51. Mitch WE, Walser M, Steinman TI, Hill S, Zeger S, Tung- arginine infusion on renal plasma flow and GFR in normal
sanga K: The effect of a keto acid-amino acid supplement to a and growth hormone deficient man. Kidney lilt 29:190, 1986.
restricted diet on the progression of chronic renal failure. N 73. Stahl RA, Kudelka S, Schollmeyer P: Glomerular pros-
Engl J Med 311:623-629, 1984. taglandins modulate GFR in rats with reduced renal mass on
52. Rosman JB, terWee PM, Meijer S, Piers-Becht TP, Sluiter high protein intake. Kidney Int 27:267,1986.
WJ, Donker AJ: Prospective randomized trial of early dietary 74. Schambelan M, Hutchison FN, Kayser GA, Blake S: Dietary
protein restriction in chronic renal failure. Lancet protein intake modulates glomerular production of eicosa-
2:1291-1296, 1984. noids in rats with Heymann nephritis. Kidney Inl 29:344,
53. EI-Nahas AM, Coles GA: Dietary treatment of chronic renal 1986.
failure: Ten unanswered questions. Lancet 1:597-600, 1986. 75. Levine MA, Kirschenbaum MA, Chaudhari A, Wong MW,
54. Mitch WE, Buffington GA, Lemann J, Walker M: A simple Bricker NS: Effect of a protein bolus on GFR and glomerular
method of estimating progression of chronic renal failure. prostanoid synthesis in normal and uremic rats. Kidlley Int
Lancet 2:1326-1328,1976. 29:339, 1986.
55. Rutherford WE, Blondin J, Miller JP, Greenwalt AS, Varva 76. Schambelan M, Blake S, Sraer J, Bens M, Nivez MP, Wahbe
JD: Chronic progressive renal disease: Rate of change of F: Increased prostaglandin production by glomeruli isolated
serum creatinine concentration. Kidney Inl 11:62-70, 1977. from rats with streptozotocin-induced diabetes mellitus. J
56. Blachley JD: The role of dietary protein in the progression Clin blUest 75:404-415, 1985.
and symptomatology of chronic renal failure. Am J Med Sci 77. Craven PA, DeRubertis FR: Glomerular prostaglandins in
288:228-234, 1984. the hyperfiltration of early diabetes. Kidney lnt 29:332, 1986.
57. Mitch WE: Nutritional therapy in renal failure. A,I1lU Rev 78. Seney FD, Wright FS: Dietary protein suppresses feedback
Med 35:249-264, 1984. control of glomerular filtration in rats. J Clin Invest
58. Bergstrom J, Alvestrand A, Bucht H, Gutierrez A: 75:558-568, 1985.
Progression of chronic renal failure in man is retarded with 79. Schnermann J, Gokel M, Weber PC, Schubert G, Briggs JP:
more frequent clinical follow-ups and better blood pressure Tubuloglomerular feedback and glomerular morphology in
control. Clin NephroI25:1-6, 1986. Goldblatt hypertensive rats on varying protein diets. Kidney
59. Barsotti G, Giannoni A, Morelli E, Lazzeri M, Vlamis I, Baldi lilt 29:520-529, 1986.
R, Giovannetti S: The decline of renal function slowed by very 80. National Institute of Diabetes, and Digestive and Kidney Dis-
low phosphorus intake in chronic renal patients following a eases: Modification of diet in renal disease study. Referring
low nitrogen diet. Clill Nephrol21:54-59, 1984. Physician's Information Handbooll, 1986.
60. Holliday MA: Nutrition therapy in renal disease, Kidlley 1111
30:S3-S6, 1986.

340 November 1987 Volume 294 Number 5