Effect of Low-Protein, Very-Low-Phosphorus Diet on Diabetic Renal Insufficiency With Proteinuria

Effect of Low-Protein, Very-Low-Phosphorus Diet on Diabetic Renal Insufficiency With Proteinuria Masayoshi Shichiri, MD, Yasuhide Nishio, MD, Mitsuo Ogura, MD, and Fumiaki Marumo, MD • We describe the clinical outcome of 13 patients with non-insulin-dependent diabetes mellitus (NIDDM), renal insufficiency, and proteinuria, treated for 12.2 ± 12.9 months (mean ± SO) with a low-protein, very-low-phosphorus diet (LPVLP) containing 30 g protein and 11.3 mmol (350 mg) phosphorus. After a control period of 18.2 ± 20.4 months, LPVLP therapy was initiated and serum urea nitrogen, uric aCid, and phosphate, as well as urinary excretion of protein, creatinine, urea nitrogen, uric acid, and phosphate, decreased significantly. There was no change in mean blood pressure, hemoglobin, blood pH, and HC03 -, as well as in serum creatinine, protein, albumin, calcium, magneSium, cholesterol, triglyceride, tJ-lipoprotein, and high-density lipoprotein (HDL)-cholesterol. Nitrogen balances were measured over 5 weeks in nine patients. Nitrogen balance increased significantly from a negative balance of -0.795 ± 1.367 gl d in the first week, to almost neutral in the fourth week, and later, was neutral or positive. Neither uremic symptoms nor signs of malnutrition appeared during the LPVLP period. These results suggest that negative nitrogen balance during the initial few weeks does not predict future nutritional status of patients with diabetic renal failure. © 1991 by the National Kidney Foundation, Inc. INDEX WORDS: Non-insulin-dependent diabetes mellitus; chronic renal failure; low-protein, very-low-phosphorus diet; nitrogen balance; proteinuria. .

D

IABETIC NEPHROPATHY characteristically progresses toward end-stage renal insufficiency, following the onset of proteinuria. I End-stage renal failure due to diabetic nephropathy constitutes an increasing proportion of patients accepted for renal replacement therapy. The frequent association of defects in water and sodium excretion, infection, and vascular diseases often prevent diabetic renal failure patients from initiation of dietary treatment. Low-protein diets, both supplemented2-4 by amino acids and ketoanalogues and nonsupplemented,5-7 have been used not only for symptomatic relief of uremia, but for retarding the progression of renal failure. Although protein and phosphorus restriction has been used in experimental diabetes in animals 8- 10 and in insulin-dependent diabetic patients with subnormal renal function 11,12 with favorable results, there is a conflicting report suggesting no correlation between dietary protein intake and rate of decline in renal function. 13 When patients with significant proteinuria receive a moderate- to high-protein diet to prevent negFrom the Second Department o/Internal Medicine. Tokyo Medical and Dental University. Tokyo. Japan; the Department o/Internal Medicine, Sashima Kyodo Hospital, Ibaraki. Japan; and the Tokyo Metropolitan Tama Geriatric Medical Center. Tokyo. Japan. Address reprint requests to Masayoshi Shichiri. MD. Second Department o/Internal Medicine. Tokyo Medical and Dental University. 5-45. Yushima I-chome, Bunkyo-ku. Tokyo 113. Japan. © 1991 by the National Kidney Foundation. Inc. 0272-6386/91/1801-0004$3.00/0 26

ative nitrogen balance, 14 beneficial effects of protein and possibly phosphorus restriction could be negated. Recent studies suggest that patients with proteinuria can have a decrease in protein excretion when placed on protein-restricted diet. 12,15,16 We have applied the low-protein, very-Iowphosphorus diet (LPVLP) to protein uric renal failure patients with non-insulin-dependent diabetes mellitus (NIDDM). During the observation periods, patients were examined repeatedly to reveal any untoward outcome of the therapy. PATIENTS AND METHODS Patients with NIDDM and chronic renal failure who had been maintained on either an unrestricted diet or a diabetic dietary regimen were fully informed about the trial. Chronic renal failure was defined as a persistent elevation of serum creatinine for greater than 6 months. For patients with a pretherapy observation period of less than 6 months, prior medical records at other hospitals were used to satisfy this criterion. Excluded from this trial were patients with malignant hypertension, renal obstruction, and malignancies, and patients requiring immediate dialysis therapy. Patients who consented to participate were admitted to our hospital ward for nutritional therapy of chronic renal failure and placed on a diet with restricted protein (30 to 35 g) and phosphorus (12.9 to 16.1 mmoljd [400 to 500 mgjd]). The initial diet contained the same amount of calories the patients had been eating as estimated by dietary recall unless markedly elevated blood glucose or glycosylated hemoglobin levels were present. During the first week of this dietary adjustment, glycemic control was assessed. The diet was then changed to a more strictly controlled diet containing 30 g of protein and 11.3 mmol (350 mg) of phosphorus per day. The diet also contained the maximum calories with which postprandial blood glucose

American Journal of Kidney Diseases, Vol XVIII, No 1 (July), 1991: pp 26-32

27

PROTEIN AND PHOSPHORUS RESTRICTION IN NIDDM

levels could be controlled at values below 1 ) .)mmol/L (200 mg/dL) in order to maintain nitrogen balance and body mass, and to reduce net urea generation. 17 This special diet was similar to one reported by Barsotti et al 3 with respect to the choice of foods of vegetable and animal origin that contain a particularly low level of phosphorus. Protein and phosphorus contents of foods were also reduced by using whey protein isolates, containing extremely low levels of protein and/or phosphorus. Care was taken to design the diet according to the preferences of each patient, and intensive dietary education was given by three expert dietitians during the first 5-week period of metabolic study, so that each patient could consume a menu consisting of the desired amount of calories, protein, and phosphorus. The intensive dietary education program consisted of repeated interviews, lectures, tests, and easy-to-comprehend literature. Each day, patients were provided with a detailed list of each food served, and they calculated the amount of protein, phosphorus, and calories actually ingested. The dietitians checked the accuracy of the records and weighed uneaten foods and calculated the results. The protein contents of individual portions of food weighed before and after each meal were calculated with reference to the Standard Tables of Food Composition in Japan. 18 These dietary surveys and education were continued from admission for at least 5 weeks and occasionally during long-term follow-up. We also assessed outpatient dietary compliance using records made by the patients themselves. No supplements were supplied. Accuracy of the survey was also assessed from chemistry and electrolytes analyses data for urine collections, and if results of dietary survey showed a significant error, the patient was excluded from the analysis. Examinations performed on admission and on the first, eighth, 15th, 22nd, and 29th morning after the start of LPVLP included complete blood cell count, serum chemistries (total protein, albumin, creatinine, urea nitrogen, uric acid, sodium, potassium, chloride, calcium, phosphate, magnesium, total cholesterol, triglyceride), and blood gases. The following serum chemistry tests were performed on admission and on the first and 29th morning: alkaline phosphatase, leucine aminopeptidase, glutamate oxaloacetate transaminase, glutamate pyruvate transaminase, lactic dehydrogenase, 'Y-g1utamic transpeptidase, total bilirubin, direct bilirubin, creatine phosphokinase, and amylase. Chest roentgenography, electrocardiography, erythrocyte sedimentation rate, and measurements of fj-Iipoprotein and high-density lipoprotein (HDL)-cholesterol were performed on admission and on the 29th morning. Twenty-four-hour urine collections were repeated every day from admission to the end of 35th day, and urinary urea, creatinine, uric acid, sodium, potassium, chloride, calcium, phosphate, magnesium, protein, and glucose were measured. The means of seven straight 24-hour excretion rates during each week were used for calculation. Blood pressure was measured three times a day, at 6:00 AM, 2:00 PM, 7:00 PM. Most of the above-mentioned tests were repeated at least every 1 to 3 months after an initial intensive care period of 5 weeks' duration. If medication was known to affect a result, that test result of the patient was excluded from the analysis. During hospitalization, weekly nitrogen balance (Bn) was calculated from the following equation I7 ,19,20: Bn = In - U - NUN, where In is nitrogen intake calculated by the dietitians, U is the urea nitrogen appearance rate (sum of urinary

urea nitrogen [UVun], urinary protein, and change in urea nitrogen pool), and NUN is non urea nitrogen excretion. Assumptions are that 60% of body weight is the volume of distribution of urea, 19,20 that the nonurea nitrogen excretion averages 0.031 g/kg/d,20 and that protein is 16% nitrogen. Urea nitrogen appearance and urinary protein excretion were weekly averages of seven straight 24-hour urine collection data, and nitrogen intake was an average of seven daily nitrogen intake measurements, Patients showing negative nitrogen balance at the end of the fifth week were asked to remain hospitalized for continuation of the nitrogen balance study. Data were analyzed by analysis of variance and by Friedman rank test. Differences from baseline values and after LPVLP on lipid profiles were analyzed using paired I test or Wilcoxon's test for non normally distributed data. Results are given as mean ± SD. The rate of progression was calculated from simple regression analysis, and differences of the declining rate before and during LPVLP in patients treated for greater than 6 months were assessed by analysis of covariance.

RESULTS

Initially, 14 patients who fulfilled our criteria were asked to participate in the trial. Although all agreed and underwent a 5-week training period, one patient had much larger excretions of solutes than were expected. Since her family and nurses witnessed her taking unprescribed meals surreptitiously, her results were not considered further. The remaining 13 patients (eight men and five women) completed the dietary protocols described above (Table 1). All patients had diabetic retinopathy and/or neuropathy. Patients 3, 5, and 12 required daily insulin injection to control hyperglycemia. They were maintained under good glycemic control during the entire observation period and did not present any major untoward outcome thought to result from the dietary therapy. The majority of the patients who adhered to a diet prescribed for diabetes were initially reluctant to begin LPVLP, but showed good compliance during the intensive education. Body weight decrease observed in nine of the 13 patients coincided with loss of edema. Ten of the 13 patients were hypertensive and two of them (patients 1 and 3) were not controlled well with antihypertensives. Patients 1, 6, and 8 had visual impairment and occasionally failed to collect urine completely. Patients 1 and 4 had incomplete bladder emptying. These four patients were excluded from the urine data analysis. Average daily ingestion of calories gradually increased from 1,646 ± 186 kcaljd in the first week to 1,790 ± 103 kcaljd in the fourth week. After the second week, the weekly average protein

SHICHIRI ET AL

28 Table 1. Clinical Features of 13 NIDDM Patients With Nephropathy Prescribed Dietary Regimen Weight (kg) Patient No.

Age'/Sex

Height (cm)

Pre

Post

Cal (kcal)

Prot (g)

Phos (mg)

1 2 3 4 5 6 7 8 9 10 11 12 13

40/M 52/M 42/F 68/M 53/F 54/M 61/M 63/F 52/M 51/M 72/F 76/F 59/M

167 161 158 160 157 174 167 146 176 164 141 148 168

50.2 53.3 45.5 50.0 62.6 75.1 63.5 38.4 59.3 61 .9 58.1 44.5 70.4

51 .9 51 .7 43.0 48.6 60.9 65.8 60.5 40.0 58.0 62.1 58.5 43.2 69.0

1,900 1,900 1,700 1,900 1,700 1,900 1,700 1,900 1,900 1,900 1,700 1,700 1,900

30 30 30 30 30 30 30 30 30 30 30 30 30

350 450 350 350 350 350 350 450 350 350 350 350 350

Antihypertensives

MOP, Nif, Cap, Atn, Fro MOP Nif, Cap, Atn, Fro Fro Nit, Cap, Pra Nit Olt Nit

Fro Fro, Olt

Abbreviations: MOP, methyldopa; Nit, nitedipine; Cap, captopril; Atn, atenolol; Fro, furosemide; Pra, prazosin hydrochloride; Olt, diltiazem . • Age at start ot LPVLP.

intake remained almost unchanged, ranging from 30.17 g to 30.73 g (Table 2). Average nitrogen balance showed a gradual but steady improvement from a negative balance in the first week (-0.795 ± 1.367 g/d) to almost neutral (0.075 ± 0.885 g/d) in the fourth week (Fig 1). Unmeasured sources of nitrogen loss, such as from skin, exfoliation, hair, sweat, respiration, and toothbrushing, are estimated to be approximately 415 mgfd.2 1 The average amount of blood drawn in this study was approximately 2.1 mL/d, which is

equivalent to about 50 mg of nitrogen per day. Subtracting the nitrogen content of ingested medicines, we estimate that average total unmeasured nitrogen losses were approximately 460 mg/d and that six patients (patients 3, 5, 7,9, 10, and 13) still remained negative in the fifth week. Patients 5, 10, and 13 exceeded 0.460 g/d in the sixth week (0.462 to 0.907 gfd), and patient 9 did in the seventh week (0.505 g/d), while patients 3 and 7 remained negative until the eighth week. During the first 5 weeks of therapy, there was

Table 2. Dietary Protein and Calorie Intake of 13 Patients Average Total Protein Intake (g/d) (weeks) Patient No.

1 2 3 4 5 6 7 8 9 10 11 12 13 Mean ±SD

37.90 36.23 29.04 31 .02 32.05 32.59 31.42 30.33 32.37 33.74 33.53 34.67 30.12 32.69 ± 2 5. 1

Average Calories Intake (kcal/d) (weeks)

2

3

4

5

32.56 32.61 28.99 29.06 30.06 30.05 30.83 28.61 31 .16 30.27 27.19 31.93 29.60 30.22 ± 1.59

31 .37 32.34 28.40 28.50 29.91 28.99 30.63 30.26 31 .10 30.19 31 .37 30.86 28.28 30.17 ± 1.29

32.10 32.10 29.90 29.13 29.87 29.40 30.03 30.32 30.02 30.94 31.41 33.87 30.37 30.73 ± 13 .3

34.23 31.74 29.81 28.89 29.90 29.65 32.32 30.35 30.43 30.31 30.09 30.74 28.84 30.56 ± 14 .7

1,817 1,419 1,638 1,657 1,706 1,679 1,646 1,730 1,804 1,861 1,241 1,394 1,802 1,646 ± 186

2

3

4

5

1,887 1,896 1,672 1,687 1,698 1,690 1,631 1,850 1,813 1,840 1,521 1,468 1,772 1,725 ± 134

1,884 1,891 1,614 1,693 1,696 1,688 1,716 1,915 1,922 1,851 1,693 1,560 1,731 1,758 ± 120

1,920 1,920 1,712 1,702 1,694 1,710 1,738 1,875 1,907 1,821 1,731 1,638 1,901 1,790 ± 103

2,035 1,924 1,698 1,721 1,701 1,708 1,724 1,883 1,923 1,855 1,692 1,509 1,731 1,777 ± 139

29

PROTEIN AND PHOSPHORUS RESTRICTION IN NIDDM

g/day

± 10.2 g/L, n = 12), fourth (58.3 ± 9.5 g/L, n = 11), and 12th months (60.0 ± 10.6 g/L, n = 4) were not significantly different from the start of LPVLP (58.5 ± 6.5 g/L, n = 13). Likewise, the hemoglobin level did not show any significant decrease during the follow-up period (102.3 ± 25.3 giL [10.2 ± 2.6 gldL) at the end of the 12th month, n = 4). Six of these patients had defined rate of progression before treatment as assessed by serial determinations of serum creatinine values. In four of them, the rate of progression decreased significantly when they changed from an unrestricted diet to LPVLP (Table 3). Figure 4 shows the changes in reciprocal serum

1

o -1

:t ':IHtttl · ,·L L·L ".

-2

TP

1

2

3

4

5

weeks

Hb

:!N+tH

°012345

-3

Alb

oIL

°012345

"

HI

:1H+H1

°012345

°012345

~~'~Il~ ~L-' ~~1~-:~:~ °012345

°012345

012345

012345

weeks

Fig 1. Effects of LPVLP on nitrogen balance. Results (mean ± SO) of nine patients for whom 5-week data were complete.

a significant decrease in serum urea nitrogen (P < 0.01), serum phosphate (P < 0.05), serum uric acid (P < 0.01), and daily urinary excretions of protein (P < 0.01), creatinine (P < 0.01), urea nitrogen (P < 0.01), uric acid (P < 0.01), and phosphate (P < 0.01). There was no significant change in mean blood pressure, serum creatinine, total protein, albumin, calcium, total cholesterol, triglyceride, ~-lipoprotein, HDL-cholesterol, hemoglobin, and hematocrit, as shown in Figs 2 and 3. Endogenous creatinine clearance did not change significantly after 5 weeks (0.53 ± 0.41 v 0.47 ± 0.35 mL/s/1.73 m 2 [31.6 ± 24.3 v 28.2 ± 21.1 mL/min/ 1.73 m 2]). Other laboratory data showed no significant changes after LPVLP. Serum total protein levels at the end of each of the second (57.6 ± 9.3 g/L, n = 12), third (58.8

°

0 1 2 3 -45

012345

0012345

012345

weeks

iLIl~~ '~llrhll ~rHffi1

~til~::~ ~L °

pre

post

pre

post

012345

012345

weeks

Fig 2. Changes of serum protein (TP), albumin (Alb), hemoglobin (Hb), hematocrit (Ht), urea nitrogen (BUN), creatinine (Cr), uric acid (UA), calcium (Ca), phosphate (P), magnesium (Mg), total cholesterol (T.Chol), triglyceride (TG), p-lipoprotein (P-lipo), and HOL-cholesterol (HOL) on the mornings of the admission (e), and after first (1), second (2), third (3), fourth (4), and fifth (5) week since the start of LPVLP. Values are mean ± SO. *P < 0.05, **P < 0.01 for differences with time against (e); *P < 0.05, **P < 0.01 for differences against time (1); and ... P < 0.05 against (2).

SHICHIRI ET AL

30 UVprot gram

UVe,

UVUN

Patient 9

Patient 5

UVUA

mlllol

mmol

mmol

-

.~ 1.0

.;

.. .

5

~ 0.'

UVea

....

UVp

-.

UV...

:~t~:t: °12345

1234

5

°'

2

1'"

- 2

- 1

-3

-2

-I

0

T_(~)

Fig 4. Reciprocal serum creatinine values and calculated regression lines for patients 5 and 9 before and after LPVLP.

DISCUSSION

345 wMks

Fig 3. Urinary excretions of protein (UVprot), urea nitrogen (UVUN ), creatinine (UVer), uric acid (UVUA ), calcium (UVea), phosphate (UVp ), and magnesium (UVMg). Values are mean ± SO. • p < 0.05, •• p < 0.01 for differences with time against (e); .p < 0.05, •• p < 0.01 against (1); and A P < 0.05 against (2).

creatinine of patients 5 and 9 before and after LPVLP; progression was virtually arrested after beginning LPVLP. Patients 1, 3, 6, and 13 showed marked water and sodium retention and edema, and had to begin hemodialysis or extracorporeal ultrafiltration treatment even though no uremic symptoms required therapy. Patients 1 and 12 died of cerebrovascular accidents.

In the present study, patients with diabetic renal failure were treated with a protein- and phosphorus-restricted diet and their in.take, as well as 24-hour urine values, was examined every day. There was a marked reduction in serum urea nitrogen and prevention of uremic symptoms. Even in patients whose renal function deteriorated and serum creatinine level increased during therapy, the increase in serum urea nitrogen was halted to a considerable degree. Serum phosphate decreased to the normal range in all patients. Serum uric acid levels also decreased significantly. These changes were associated with marked reduction in urinary excretions of urea, uric acid, and phosphate, suggesting decreased production of urea and uric acid, and a reduction in ingestion

Table 3. Clinical Outcome of 13 Patients With Dietary Therapy Rate of Progression Observation Period (mo)

(Xl0- 2 dL/mg/mo) p

Patient No.

Pre

Post

1 2 3 4 5 6 7 8 9 10 11 12 13

2 5 4 0 33 0 48 53 42 0 40 0 10

5 5 8 0 44 19 5 16 33 5 8 4 6

Mean ± SO

18.2 ± 20.4

12.2 ± 12.9

Abbreviation: HD, hemodialysis.

Pre

Post

Value

-13.16

-2.98

<0 .01

- 2.08

-0.34

<0.01

-0.74 -0.53

-1.87 +0.16

<0.05

-1 .69

- 1.80

NS

- 1.38

-0.80

<0.05

NS

Outcome

HD HD HD Died Diet continued HD HD HD Diet continued HD HD Died HD

31

PROTEIN AND PHOSPHORUS RESTRICTION IN NIDDM

of phosphorus. During the initialS weeks, serum total protein did not change significantly, nor did hemoglobin or hematocrit. These parameters remained unchanged during subsequent periods, suggesting that LPVLP did not cause hypoproteinemia or anemia. Urinary excretions of calcium and magnesium markedly decreased, while serum levels remained unchanged, suggesting significant renal conservation of these minerals despite decreased intake. Plasma lipid profiles did not change significantly in the present study. A reduction in meat intake following LPVLP can lower creatinine appearance, and it takes a few months for a new steady-state level to be achieved. 22 Hence, any change in the rate of decline in the slope occurring immediately after protein restriction cannot be attributed to a change in the progression of renal disease. Therefore, we compared the slopes of the reciprocal serum creatinine plot before and during LPVLP in patients treated for greater than 6 months. Of the six patients (patients 3, 5, 8, 9, 11, and 13), four showed a significant halt in the decline in reciprocal serum creatinine during LPVLP (Table 3). Patients 5 and 9 started LPVLP when their serum creatinine level was 133 JLmolfL (1.5 mg! dL) and 221 JLmolfL (2.5 mg/dL), and adhered to LPVLP for 44 and 34 months, respectively. They showed slowing of progression (Fig 4), and no decrease in body weight during long-term follow-up. They are still enjoying full-time jobs and exercise without any complaints except for some postural edema.

Little information is available as to whether diabetic renal failure patients with massive proteinuria maintained on LPVLP may develop protein malnutrition. The average nitrogen balance of our patients was negative for the first 3 weeks after initiation of LPVLP, but later increased. Six patients with slightly negative balance in the last week shortly thereafter showed positive balance. In nutritional therapy of chronic renal failure, dietary nitrogen should be reduced so as to minimize the production of protein-derived waste product, and at the same time sufficient dietary protein and energy must be provided to prevent catabolism of endogenous body proteins. Nitrogen balance of patients on a restricted protein intake is reported to increase as energy intake increases. 17 Our results suggest that negative nitrogen balance during the initial few weeks following LPVLP does not predict the .future nutritional status of patients with diabetic renal failure. However, such patients should be carefully monitored to avoid the development of inadvertent protein malnourishment, since the overall long-term safety of the regimen has not been established. ACKNOWLEDGMENT We are grateful to Professor William E. Mitch, MD for helpful advice and critical review of the manuscript. We are also greatly indebted to Hiro Kanakawa, Toshiko Honjohya, and Ema Nishi for providing skillful dietary programs and their efforts in patients care.

REFERENCES I. McCray RF, Pitts TO, Puschett JB: Diabetic nephropathy: Natural course, survivorship and therapy. Am J Nephrol 1:206-218,1981 2. Mitch WE, Walser MW, Steinman TI, et a1: The effect of a keto-amino acid supplement to a restricted diet on the progression of chronic renal failure. N Eng! J Med 311 :623629, 1984 3. Barsotti G, Guiducci A, Ciardella F, et al: Effects on renal function of a low-nitrogen diet supplemented with essential amino acids and ketoanalogues and of hemodialysis and free protein supply in patients with chronic renal failure. Nephron 27:113-117, 1981 4. Mitch WE, Abras E, Walser M: Long-term effects of a new ketoacid-amino acid supplement in patients with chronic renal failure. Kidney Int 22:48-53, 1982 5. Lucas PA, Meadows JH, Roberts DE, et al: The risks and benefits of a low protein-essential amino acid-keto acid diet. Kidney Int 29:995-1003, 1986 6. Maschio G, Oldrizzi L, Tessitore N, et a1: Effects of di-

etary protein and phosphorus restriction on the progression of early renal failure. Kidney Int 22:371-376, 1982 7. Rosman JB, TerWee PM, Meijer S, et al: Prospective randomized trial of early dietary protein restriction in chronic renal failure. Lancet 2:1291-1295,1984 8. Hostetter TH, Olson JL, Rennke HG, et al: Hyperfiltration in remnant nephrons: Apotentially adverse response to renal ablation. Am J Physiol 241 :F85-F93, 1981 9. Smadel JE, FaIT LE: The effect of diet on the pathologic changes in rats with nephrotoxic serum nephritis. Am J Pathol 15:199-216,1939 10. Kleinknecht C, Salusky I, Broyer M, et al: Effect of various protein diets on growth, renal function and survival of uremic rats. Kidney Int 15:534-541, 1979 II. Barsotti G, Giannoni A, Morelli E, et al: The decline of renal function slowed by very low phosphorus intake in chronic renal failure patients following a low nitrogen diet. Clin Nephrol 21 :54-59, 1984 12. Evanoff G, Thompson C, Brown J, et a1: Prolonged

32 dietary protein restriction in diabetic nephropathy. Arch Intern Med 149:1129-1133,1989 13. Nyberg G, Norden G, Attman PO, et al: Diabetic nephropathy: Is dietary protein harmful? J Diabetic Complications 1:37-40, 1987 14. Blainey JD. High protein diets in the treatment of the nephrotic syndrome. Clin Sci 13:567-581 , 1954 15. Kaysen GA, Gambertogiio J, Jimenez I, et al: Effects of dietary protein intake on albumin homeostasis in nephrotic patients. Kidney Int 29:572-577, 1986 16. EI Nahas AM, Masters Thomas A, Brady SA, et al: Selective effect oflow protein diets in renal diseases. Br J Med 289:1337-1341,1984 17. Kopple JD, Monteou FJ, Shaib JK: Effect of energy intake on nitrogen metabolism in non-dialyzed patients

SHICHIRI ET AL

with chronic renal failure. Kidney Int 29:734-742, 1986 18. Science and Technology Agency: Standard Tables of Food Composition in Japan (ed 4). Prime Minister's Office of Japan, Tokyo, Japan, 1982 19. Mitch WE: Nutritional therapy and the progression of renal insufficiency, in Mitch WE, Klar S (eds): Nutrition and the Kidney. Boston, MA, Little Brown, 1989, pp 165-170 20. Maroni BJ, Steinmann TI, Mitch WE: A method for estimating nitrogen intake of patients with chronic renal failure. Kidney Int 27:58-65, 1985 21. Calloway DH, Odell ACF, Margen S: Sweat and miscellaneous nitrogen losses in human balance studies. J Nutr 101:775-786, 1970 22. Mitch WE: The influence of the diet on the progression of renal insufficiency. Annu Rev Med 35:249-264, 1984