Effects of vegetable diets on nitrogen metabolism in cirrhotic subjects

GASTKOENTEROLOGY

1985:89:538-U

Effects of Vegetable Diets on Nitrogen Metabolism in Cirrhotic Subjects FREDRICK L. WEBER, Jr., DEBRAH MINCO, KATHLEEN M. FRESARD, and JOHN G. BANWELL Veterans Administration Medical Center and Department of Medicine, Division of Gastroenterology and Nutrition, Case Western Reserve University School of Medicine. Cleveland, Ohs

This study compared the effect of a vegetable diet with an animal protein diet on various aspects of nitrogen metabolism to identify what components of the vegetable diet might be causing beneficial therapeutic effects in hepatic encephalopathy. Vegetable diets contained 4.5-fold greater amounts of fiber [56 + 3 g/day) and reduced amounts of methionine, tyrosine, and tryptophan. In 6 stable cirrhotic subjects without encephalopathy, vegetable diets caused a significant reduction in the urea produc. tionratefrom106t5to89+5mg~kg-‘*24h-’oj urea nitrogen. This was mainly accounted for by a fall in urinary urea output. Vegetable diets also caused a fall in total urinary nitrogen, which was accounted for by the fall in urea nitrogen, and a in fecal nitrogen from 12 2 2 to comparable increase 28 2 5 mg * kg-l - 24 h-‘. The fecal bacterial fraction contained 63% of the increase in stool nitrogen. Most plasma amino acids, including methionine, tyrosine, phenylalanine, as well as total and free tryptophan, were unchanged. The effect of vegetable diets on nitrogen metabolism can be mainly accounted for by the increased intake of dietary fiber and increased incorporation and elimination of nitrogen in fecal bacteria. Three recent reports indicated that vegetable diets are beneficial in the treatment of chronic hepatic encephalopathy when compared with diets containing an equivalent amount of animal protein (l-3): one study, however, failed to reach the same conclu-

sion (4). Why vegetable diets might be more effective than conventional diets is unclear. Animal and vegetable protein diets differ in respect to a large number of individual nutrients; most interest, however, has focused on the different amino acid composition of the two types of diets. It has been proposed that the advantageous therapeutic effect of vegetable diets might be explained by either the reduced dietary load of potentially toxic amino acids, i.e., methionine and aromatic amino acids (phenylalanine, tyrosine, and tryptophan) or differences in the potential of the composite amino acids to produce ammonia (11. Addition of vegetable fiber to normal diets, however, may cause an increased rate of colonic transit and increased fecal bulk (5). Complex polysaccharides contained in dietary fiber can increase the colonic bacterial mass and its incorporation of nitrogen (6). Hence, the increased fiber content of vegetable diets has the potential to alter total body nitrogen metabolism by enhancing bacterial metabolism and thereby increasing incorporation of nitrogen into fecal bacteria. In the present study we have compared the effects of vegetable diets with isocaloric, isonitrogenous animal protein diets on the urea synthesis rate, which serves as an index of nitrogen entry into the host, and determined the relationship between this parameter and increased nitrogen output in fiber, bacterial, and soluble fractions of feces.

Materials Received

Address University 44106.

October

3, 1984. Accepted

March

7, 1985. requests for reprints to: Fredrick L. Weber, Jr., M.D., Hospitals, 2074 Abington Road, Cleveland, Ohio

This study was supported by Veterans Administration Medical Research Service and National Institutes of Health grant MO1 RR00080-22. 0 1985 by the American Gastroenterological 0016-5085/85/$3.30

Association

and Methods

Patients Six stable patients with biopsy-proven alcoholic cirrhosis participated in the study. The patients ranged in age from 46 to 69 yr. At the time of the study all patients had abstained from alcohol for at least 3 mo, and none had ascites, gastrointestinal bleeding, or clinical hepatic encephalopathy. They were not further assessed for the

September

1985

possible presence of subclinical hepatic encephalopathy. On entrance to the study conventional liver tests were either normal or mildly abnormal and were stable throughout the study period: bilirubin 1.3 * 0.8 mg/dl, mean & 1 SD (normal 0.2-1.0); albumin 4.3 +- 0.3 g/d1 (normal 3.55.0); aspartate aminotransferase 30 + 7 IUiL (normal lo40); and alkaline phosphatase 95 + 34 IUiL (normal 30110). Serum creatine and creatinine clearance were 1.3 + 0.3 mgidl and 84 + 24 mlimin, respectively, and were unchanged throughout the study period. Serum electrolytes were normal and remained unchanged. All patients received a multivitamin capsule and 1 mg of folic acid daily. Except for 3 subjects who were receiving spironolactone as outpatients, no patients received any other medications. The study protocol was approved by the institutional review board on February 2, 1982.

Diets Each patient had individual diets formulated based on their food preferences. For each patient diets were isocaloric and had the same amount of protein. Food came from constant sources throughout a study, and during each diet period, identical food was served on alternate days. Thirty percent of the protein in each patient’s diet was constant throughout the study and was primarily derived from highly refined grains used in breads, macaroni, spaghetti, and cereals; however, potatoes and rice were also included. This was done so as to increase diet palatability. The remainder of the protein in the diets (70%) was derived entirely from either animal sources (including meat, dairy products, and fish) or vegetable sources. Prior experience with these patients gave us a clear understanding of the amount of food that the patients could be expected to eat reliably, and the bulky nature of the vegetable diets limited the amount of protein that could be provided. Hence, different patients were on varying protein intakes; 1 patient received 1 g protein/kg body wt, whereas the other 5 patients received relatively less protein (0.87, 0.70, 0.68, 0.65, and 0.63 g/kg body wt). All patients ate their meals under observation of the ward staff and any uneaten food was reweighed. Analysis of nutrient intake was accomplished by use of a computer program listing 50 individual nutrients for each type of food. Nutrient analysis of the different foods was compiled from standard sources (7-13). Values for glutamic and aspartic acid were not included because conventional methodology for protein hydrolysis converts asparaginine to aspartic acid and glutamine to glutamic acid. Values given in standard sources were also used to calculate protein content of the different diets. These figures are derived from nitrogen analysis of foods, and conversion to protein is based on the finding that most protein contains 16% nitrogen. This figure, however, varies for certain foods because of the presence of nonprotein nitrogen and variations in the amino acid composition of different proteins (14). These different conversion factors were used to calculate nitrogen intake during different diets (7). Additionally, we directly analyzed the nitrogen content of most foods used in this study and found close agreement with published values.

VEGETABLE DIETS AND NITROGEN METABOLISM

539

Protocol All patients were studied under metabolic ward conditions. Four of the 6 patients were studied during three diet periods; two animal diet periods separated by a vegetable period. In these patients the results from the two animal diet periods were averaged. The other 2 patients were studied during two diet periods; 1 patient first received the animal and then the vegetable diet, whereas the other patient received diets in the reverse order. Each diet period lasted 9 days. The first 3 days were considered an equilibration period; during the following 6 days samples of blood, urine, and stool were collected when plasma urea and urinary urea excretion were constant. Twentyfour-hour urine samples were collected in refrigerated containers and then frozen. Stool samples were collected individually in plastic bags and immediately frozen. Subsequently stools collected over 3-day periods were pooled, thawed, and thoroughly mixed while still cold. Duplicate samples were promptly lyophilized for subsequent analysis. On day 6 or 7 of each diet period urea synthesis rates were determined by the method of Walser and Bodenlos (15). After an overnight fast, patients were injected intravenously at 8:30 AM with 5 &i of [‘*C]urea (New England Nuclear, Boston, Mass.) that was sterile and pyrogen free. The patients then ate their protocol diet throughout the day. Blood and urine collections were taken for determination of [14C]urea as previously described (16). Fasting plasma samples were obtained twice during the latter part of each diet period for determination of amino acids and prealbumin.

Analytic

Methods

[‘*C]Urea was determined in the injection solution, plasma, and urine on a Packard Tricarb 300 scintillation counter (Packard Instrument Co., Downers Grove, Ill.) using an internal standard to correct for quench. Amino acids were determined by ion exchange chromatography on a Beckman 11X amino acid analyzer (Beckman Instrument Co., Palo Alto, Calif.) as previously described (17). Tryptophan was determined spectroflurometrically by the method of Denkla and Dewey (18) as modified by Bloxam and Warren (19). Free tryptophan was also determined by this method in a plasma filtrate obtained by centrifuging plasma at 800 g for 30 min at 25°C through an Amican CF50 Diaflo membrane cone (20). Prealbumin was determined in plasma using a radial immunodiffusion kit (Behring Diagnostics, Somerville, N.J.) (normal lo-40 mgi dl). Other analyses were determined in the clinical chemistry laboratory by standard automated methods. The dry weight of stool was determined after specimens were lyophilized to constant weight. Stool solids were then separated into fiber and bacterial fractions by the method of Stephen and Cummings (21) in which fecal solids are mixed in the presence of a detergent and physical separation of different size particles is accomplished by filtering through nylon mesh (100 pm and 150 pm). Near complete separation of fiber and bacterial fractions was confirmed by scanning electron microscopy.

540

WEBER

GASTROENTEROLOGY

ET AL

Prior chemical and microscopic analysis of these fractions has determined that the bacterial fraction contains only 6%-7% plant material, and a maximum of 2% of the bacterial mass may contaminate the fiber fraction. The soluble fraction was determined by subtracting fiber and bacterial solids from total fecal solids. In the experiment herein described, this method has revealed close (99%) agreement with soluble fractions directly determined by homogenization and filtration of fecal solids in the absence of detergent (21). Total nitrogen was determined in food components, urine, and fecal solid fractions on a Coleman nitrogen analyzer as previously described (16). Nitrogen in the fecal soluble fraction was calculated as the difference between total fecal nitrogen minus nitrogen in the fiber and bacterial fractions.

Urea Synthesis Calculations

and

Degradation

Rate

A detailed explanation of the principles and methods used in determining urea synthesis and degradation rates has been reported previously (16,22,23). In brief, the urea synthesis rate was determined by analysis of the logarithmic decline in plasma [“C]urea specific activity after injection of a known amount of [“Clurea. When the plasma urea concentration is constant, there are two fates of synthesized urea: (a) excretion in the urine and (b) degradation by gut flora. Urinary urea excretion was measured directly and urea degradation was calculated by subtracting urinary excretion from the total urea synthesis rate. The intestinal or extrarenal clearance rate of urea was determined by dividing the urea degradation rate by the plasma urea concentration. The calculated radiation exposure for each subject was 0.3 millirads to the total body and 2.7 millirads to the critical organ (bladder). Data in the text and tables have been presented as the mean ? SEM. Statistical significance was determined by paired, two-tailed t-tests. In most cases the data are corrected for body weight. The mean weight of these subjects was 83 5 4 kg, and use of this figure permits calculation of the uncorrected data.

Results Dietary

Table

1. Dietary Protein

Intake During Periods

Nutrient Protein (g kg ‘. 24 h ‘1 Carbohydrate, total h-‘) (g.kg-‘.z4 Carbohydrate, simple (g. kg-‘. 24 h-‘) Carbohydrate, complex (g. k ’‘24 h ‘) Fat(g.kg ‘.24h ‘) K Calories (kg-’ 24 h-‘) Fiber, total (g’kg-‘,24 h-‘) Fiber, soluble” (mg kg ” 24 h Threonine’ Serine Proline Glycine Alanine Cystine Methionine Valine Isoleucine Leucine Tyrosine Phenylalanine Tryptophan Lysine Histidine Arginine

Animul

Vol. 8~. No. 3

und Vegetable p cc”

Animal

Vegetable

0.74 k 0.06 3.48 _t 0.13

0.78 + 0.07 3.78 2 0.14

0.02

2.48 + 0.07

2.11 t 0.15

0.05

0.98

1.68 + 0.07

0.001

1.01 t 0.05 25.8 t 0.9 0.15 t 0.01

0.91 2 0.06 26.2 t 1.1 0.67 -t 0.05

0.001

0.05 -+ 0.002

0.25 + 0.03

0.001

31.4 + 2.3 30.4 + 2.7 35.9 t 3.8 29.8 ? 3.1 34.6 k 3.2 10.8 ?z 0.8 18.5 + 1.4 40.2 + 3.0 36.7 + 2.7 59.2 Ii 4.3 27.3 + 2.1 34.2 -c 2.5 9.2 r 0.7 55.9 2 4.7 27.7 + 2.5 41.5 t 3.4

27.5 k 2.6 31.4 + 3.0 27.0 + 3.3 26.8 i 2.8 26.7 k 2.6 10.4 + 0.9 10.5 -t 0.8 36.8 + 3.2 32.9 -' 2.8 54.9 + 4.5 21.6 i- 1.7 38.4 -c 3.1 8.2 -c 0.8 41.8 + 4.3 18.6 rt 1.5 52.1 5 5.6

0.02

f

0.08

‘)

0.05 0.01 0.001 0.02 0.001 0.025 0.001 0.01 0.02 0.001 0.005 0.02

u Significance by paired t-test when p < 0.05.'ISoluble fiber refers to water-soluble dietary fiber such as pectin and gums. ’Amino acids expressed as g kg- ’ 24 h I.

drates rose. The intake of most amino acids differed. On vegetable diets methionine intake was significantly reduced along with the three branched-chain amino acids (valine, isoleucine, and leucine). Among the aromatic amino acids, tyrosine and tryptophan were reduced, whereas phenylalanine was increased on vegetable diets.

Urea Metabolism Intake

The intake of individual nutrients (Table 1) was based on the actual amount of food eaten during the last 6 days of each diet period. Total intake of protein and calories were not different on the two diets, and both diets met the recommended daily allowances for all measured components. There were, however, numerous differences in the intake of individual nutrients on the two diets. Total fiber intake was increased 4.5-fold during vegetable diets, which represented a mean fiber intake of 56 + 3 gl day. Total carbohydrate intake was somewhat greater during vegetable diets, but the intake of simple carbohydrates fell, whereas that of complex carbohy-

Differences in the effects of animal and vegetable diets on urea metabolism and other parameters were consistently seen in all 6 subjects. In the 2 patients studied during two diet periods, changing the order in which diets were administered gave similar results. In 4 patients studied during three diet periods there was no systematic difference in any measured parameter between the two animal diet periods, hence results from these periods were averaged. As seen in Table 2, vegetable diets caused reductions in the urea synthesis rate, urinary urea excretion, total body urea pool, and plasma urea concentration. Mean reductions in these four parameters

September 1985

Table

VEGETABLE

2. Urea Metabolism During Animal Vegetable Protein Diets

Vegetable

p <”

5

89 2 5

0.05

90 + 6

73 + 4

0.005

16 + 4

17 t

Animal

Urea synthesis

106?

',24 h ‘) (mg~k Urinary urea excretion (mg kg ‘. 24 h ‘)

Urea degradation (mg kg ‘. 24 h ‘) Intestinal (ml

Total

kg body

119 ? 28

urea clearance ‘. 24 h ‘) urea

5

133 5 29 64 -c 8

0.02

149 t 21

122 + 17

0.01

530 ? 16

525 t

78 + 10

pool

and

CmgW

Urea concentration (mgkl

Urea distribution

volume

18

(ml/kg) n = 6. Milligram measurements are milligrams of urea ” Significance by paired t-test when p < 0.05.

nitrogen.

were very similar, averaging 170&19°~ below values obtained during animal diets. The reduction in the urea synthesis rate was not explained by any change in the urea degradation rate. Similarly, the intestinal urea clearance rate was unchanged.

Fecal

3. Stool Components Diets Total

Stool stool solids

AND

24 h-‘) was quantitatively both total urea synthesis excretion (-18 ? 4).

Nitrogen

NITKOGEK

METABOLISM

541

similar to reductions in (-17 2 6) and urinary urea

Balance

During vegetable protein diets, the increase in fecal nitrogen excretion was accompanied by a fall in total urinary nitrogen excretion (-21 * 4 mg kg-l * 24 h-‘) [Table 4). Hence, there was no significant change in nitrogen balance. The tendency toward a more positive mean nitrogen balance during vegetable diets was accounted for by a somewhat greater nitrogen intake, although it was not significantly different from the animal diets. Two of the patients on lower protein intakes (0.63 and 0.68 g protein/kg) were in significantly negative nitrogen balance throughout the study (- 15.2 ? 2.2 and -12.1 ‘_ 2.3 mg nitrogen * kg-’ - 24 h I, respectively). Other patients were either in nitrogen equilibrium or positive nitrogen balance. In contrast to the findings of nitrogen balance studies, plasma concentrations of the rapid turnover protein prealbumin were ll”/, lower during vegetable diets (19.0 ? 2.3 mgidl) than during animal diets (21.3 * 2.6, p < 0.05).

Excretion

Both total fecal weight and solid weight rose by 160% on vegetable diets when compared with animal diets (Table 3). The increase in fecal solids was caused by increases in all major fecal solid components: fiber, bacterial, and soluble. Total fecal nitrogen excretion also more than doubled on vegetable diets. Sixty-three percent of the increase was contained in the bacterial fraction. Smaller increments were seen in the soluble and fiber fraction, which accounted for 35% and 3%, respectively, of the increase in nitrogen excretion. The increase in total fecal nitrogen excretion (+ 16 t 3 mg . kg-’ *

Table

DIETS

weight

on Animal

Animal diet

Vegetable diet

87 5 29

231 2 53”

17.4

-+ 1.8

Fiber fraction

0.9 + 0.2

Bacterial fraction

10.2 + 0.9

Solubln fraction

6.2 + 1.5

Significantly

different

Nitrogen kg ‘. 24 h

Animal diet

‘)

Vegetable diet

k 6.1”

11.5

+ 1.7

27.6

+ 4.7”

3.9 I 0.6”

0.22

t 0.05

0.67

2 0.13”

45.2

26.8

k 3.3”

5.7 f

0.7

15.8

?

3.1”

L 2.1”

by paired

t-test:

5.6 f

1.2

” p < 0.005:

11.1

* 1.5”

” P < 0.05

Acids

Discussion The results of this vegetable diets increased Table

study the

clearly indicate fecal elimination

4. Nitrogen

Intake and Excretion and Vegetable Protein Diets” Animal

Nitrogen intake Urinary nitrogen Fecal nitrogen Nitrogen

14.5

Amino

The two diets caused few significant changes in plasma amino acids (Table 5). In general, the differences in amino acid intake on the two diets caused corresponding, modest alterations in plasma aminograms, but only with threonine and lysine were reduced intakes associated with significant reductions in their plasma concentrations during vegetable diets. Two nonprotein amino acids showed the greatest changes during vegetable diets; a-amino-N-butyric acid fell by 30% and a-amino adipic acid rose by 277%.

and Vegetable

(mg

(g/24 h)

Plasma

balance’

During

that of

Animal Vegetable

118.2

t 9.0

124.8

106.0

+ 6.5

85.3

-+ 4.1”

2 1.7

27.6

-t 4.7”

11.5 -3.4

t 4.1

t

11.5

6.4 2 6.5

n = 6. ” Data expressed as milligrams per kilogram per 24 h. ” p i 0.005. ’Calculated by subtracting urinary and fecal raxcretion from intake minus 5 mgikg for unmeasurt~d losses.

542

Table

WEBER

ET AL.

GASTROENTEROLOGY

5. Effect of Animal and Vegetable Protein on Fasting Plasma Amino Acids Amino acid

Animal diet (NM]

Vegetable diet (PM)

Taurine Aspartic acid Threonine Serine Glutamate Proline a-Amino-adipic acid Glycine Alanine Citrulline a-Amino-N-butyric acid Valine Cystine Methionine Isoleucine Leucine Tyrosine Phenylalanine Ornithine Lysine Histidine Arginine Tryptophan Free tryptophan

48 2 4 9+-l 137 k 6 103 2 9 37 + 6 211 k 32 9kl 218 2 9 404 + 28 49 + 6 23 + 2 207 t 10 105 k 5 31 + 3 64 t 3 119 t 5 85 t 11 70 2 5 66 t 4 172 k 14 95 lr 16 113 k 7 60 + 6 6.5 k 0.3

47 + 10 + 125 f 108 k 35 + 214 + 34 t 225 + 384 % 49 t 16 k 190 t 104 + 29 + 61 t 111 ? 84 k 74 k 71 + 150 + 90 k 124 k 58 t 7.1 +

n = 6. " Significance

5 2 6 8 5 32 2 12 28 7 2 12 5 2 3 6 11 6 5 14 9 4 5 1.0

Diets

p <”

0.05

0.001

0.001

0.05 0.005

by paired t-test is given when p < 0.05.

urea precursor nitrogen. The fall in the urea synthesis rate could be quantitatively accounted for by an increase in fecal nitrogen excretion. Most of the increase in fecal nitrogen was contained in the fecal bacterial compartment, and the results suggest that the increased incorporation of nitrogen into bacteria was caused by the large amount of dietary fiber provided in the vegetable diets. Only 7% of dietary fiber was excreted in stool while patients were on vegetable diets. The remainder of the fiber was probably metabolized by the colonic microflora, causing an increase in fecal bacterial mass. Dietary fiber that can be metabolized by fecal bacteria has been shown to increase fecal bacterial mass and nitrogen content in human studies (6). The 16.6 g increase in bacterial mass on vegetable diets followed a 43.1 g net increase in fiber intake and metabolism. This equated to a 0.41 g increase in bacterial mass per gram of fiber metabolized, which corresponded to previous observations (24). Nitrogen contained in the fecal soluble fraction also increased. Although this may result from reduced transit time, increased nitrogen in this fraction may also be caused by catabolite inhibition or repression whereby carbohydrates cause inhibition of synthesis or activity, respectively, of enzymes mediating bacterial catabolism of amino acids (25). Other studies have noted the laxative effects of vegetable diets (l-3). DeBruign et al. (3) measured

Vol. 89, No. 3

fecal and urinary output in their patients and found a significant decrease in urinary nitrogen and a trend, which was not significant, toward increased fecal nitrogen excretion. These patients, however, were receiving a fiber source (psyllium mucilloid) throughout the study that may have blunted the increase in fecal nitrogen found with vegetable diets. The study of Shaw et al. (4) revealed a trend toward increased fecal nitrogen output on vegetable diets, but no change in urinary nitrogen and no improvement in encephalopathy. However, 4 of their 5 patients were receiving therapeutic doses of lactulose throughout the study, which may have obscured treatment effects as well as changes in nitrogen excretion as vegetable diets and lactulose appear to influence these parameters in a similar manner (26). Lactulose and digestible plant fiber may have additive effects as different bacterial species may be capable of metabolizing one substance but not the other. Greenberger et al. (1) studied 1 patient in whom encephalopathy was improved when lactulose was added to a vegetable diet regimen, but in a larger series of patients, vegetable diets had no discernable effect on number connection tests when (4). The given to patients already receiving lactulose question of whether lactulose and dietary fiber may act additively to alter nitrogen metabolism in cirrhotic patients deserves further study. Although there were many differences in the intake of individual amino acids between the two diets, the corresponding changes in plasma aminograms were not likely to be of any practical significance. In particular, plasma-free tryptophan, which has been correlated with the course of encephalopathy (27,28), did not change. Plasma-free tryptophan has not previously been measured on these different types of diets. Ammonia production might be expected to be lower on vegetable diets because of differences in the amino acid composition (1). Individual free amino acids have been shown by Rudman et al. (29) to cause varying rates of ammonia production after oral ingestion. Based on their data and the dietary intake of protein amino acids (exclusive of glutamine, glutamate, aspartic acid, and asparagine), we calculated that ammonia production on vegetable diets would be only 69% of that expected on animal diets. However, there may be a fallacy in trying to predict ammonia production by this method due to uncertainties regarding (a) the relative amounts of asparganine versus aspartic acid, and glutamine versus glutamic acid in the diets, (b) the amounts of nonprotein nitrogen contained in the diets (14,30), and (c) whether amino acids that are primarily absorbed as oligopeptides, as may be expected after a protein meal (31), cause ammonia production in a similar

September

VEGETABLE

1985

manner as their constituent free amino acids. The predicted large reduction in ammonia production on vegetable diets would be expected to cause a quantitatively similar reduction in the urea synthesis rate. Although the urea synthesis rate did fall, it was entirely accounted for by an increase in fecal nitrogen, primarily caused by the effect of dietary fiber in augmenting bacterial mass. Feeding purified fiber should permit one to separate the effects of alterations in dietary fiber content from the different amino acid composition of the two diets in reducing urea production. The practicality and utility of vegetable protein diets for the treatment of hepatic encephalopathy requires discussion. These diets are not well tolerated due to their bulk. Our patients were stable clinically, without evidence of encephalopathy, and would not be considered anorectic. Most of them, however, were unable to ingest sufficient amounts of vegetables to bring their protein intake to I g/kg body wt even though they selected vegetables that they considered most palatable. Two of the patients on lower protein intakes were in negative nitrogen balance throughout the study. Although they could easily have ingested more animal protein, they did not feel that they could ingest any more vegetables than they were given. Patients with more advanced liver disease and hepatic encephalopathy might have a greater likelihood of being in negative nitrogen balance on long-term therapy because of anorexia and a reduced protein intake. These results do not necessarily imply that increased fecal nitrogen excretion caused by vegetable diets will lead to a net negative nitrogen balance as increased fecal nitrogen excretion was compensated for by a decrease in urinary nitrogen. Patients who receive less protein than our subjects, however, might not be able to compensate in a similar manner for increased fecal nitrogen losses and could undergo protein wasting on vegetable diets. Encephalopathic patients with more impaired liver function or more extensive portal-systemic shunting, or both, may have a different, and possibly a relatively greater, metabolic response to vegetable diets than the well-compensated, nonencephalopathic patients studied herein. References 1. Greenberger NJ, Carley J, Schenker S. Bettinger I, Stamnes C, Beyer P. Effect of vegetable and animal protein diets in chronic hepatic encephalopathy. Dig Dis Sci 1977;22:845-5.5. 2. Uribe M, Marquez MA, Ramos GG, Ramos-Uribe MH, Vargas F. Villalobos A. Ramos C. Treatment of chronic portalsystemic encephalopathy with vegetable and animal protein diets: a controlled crossover study. Dig Dis Sci 1982:27:110916. 3. DeBruijn KM, Blendis LM, Zilm DH. Carlen PL. Anderson

4.

5.

6. 7.

8. 9.

10. 11.

12.

13. 14.

15. 16.

17.

18. 19.

20.

21. 22.

23.

24.

25.

26.

DIETS

AND

NITKOGEN

METABOLISM

543

GH. Effect of dietary protein manipulations in subclinical portal systemic encephalopathy. Gut 1983:24:53-60. Shaw S, Worner TM. Lieber CS. Comparison of animal and vegetable protein sources in the dietary management of hepatic encephalopathy. Am J Clin Nutr 1983;38:59-63. Cummings JH, Branch W, Jenkins DJA. Southgate DAT, Houston H. James WPT. Colonic response to dietary fiber from carrot, cabbage, apple, bran, and guar gum. Lancet 1978;i:5-9. Stephen AM, Cummings JH. Mechanism of action of dietary fiber in the human colon. Nature 1980;284:283-4. Watt BK, Merrill AL. Composition of foods. Agriculture handbook No. 8. Washington, D.C.: llnited States Department of Agriculture, 1963. Orr ML, Watt BK. Amino acid content of foods. Washington, D.C.: United States Department of Agriculture, 1968. Paul AA, Southgate DAT. McCance and Widdowson’s the composition of foods. London: ElsevieriNorth-Holland Biomedical Press, 1978. Church CF, Church HN. Food values of portions commonly used. 11th ed. Philadelphia: JB Lippincott, 1970. Anderson JM, Chen W. Ward K. Composition of foods commonly used in diets for persons with diabetes. Diabetes Care 1978:1:293-302. Harvey D. Tables of amino acids in foods and feeding stuffs. Commonwealth Agricultural Bureaux Technical Communication, No. 19. Farnham Royal, 1985. Krause MV, Mahan LK. Food, nutrition and diet therapy. 6th ed. Philadelphia: WB Saunders. 1979. Munro HN, Crim MC. The proteins and amino acids. In: Goodhart RS, Shils ME, eds. Modern nutrition in health and disease. 6th ed. Philadelphia: Lea & Febiger, 1980:51-98. Walser M, Bodenlos LJ. Urea metabolism in man. J Clin Invest 1959;38:1617-26. Weber FL Jr. The effect of lactulose on urea metabolism and nitrogen excretion in cirrhotic patients. Gastroenterology 1979;77:518-23. Weber FL Jr, Snodgrass PJ, Powell DE, Rao P. Huffman SL, Brady PG. Abnormalities of hepatic mitochondrial urea-cycle enzyme activities and hepatic ultrastructure in acute fatty liver of pregnancy. J Lab Clin Med 1979;94:27-41. Denkla WD, Dewey HK. The determination of tryptophan in plasma, liver. and urine. J Lab Clin Med 1967;69:160-9. Bloxam DL. Warren WH. Error in the determination of tryptophan by the method of Denkla and Dewey. A revised procedure. Analyt Biochem 1974;60:621-5. Curzon G. Kantamaneni BD, Winch J, Rojas-Bueno A, MurryLyon IM, Williams R. Plasma and brain tryptophan changes in experimental acute hepatic failure. 1 Neurochem 1973; 21:137-45. Stephen AM, Cummings JH. The microbial contribution to human fecal mass. J Med Microbial 1980;13:45-56. Weber FL Jr, Fresard KM. Comparative effects of lactulose and magnesium sulfate on urea metabolism and nitrogen balance in cirrhotic subjects. Gastroenterology 1981;80:994-8. Weber FL Jr, Fresard KM, Lally BR. Effects of lactulose and neomycin on urea metabolism in cirrhotic subjects. Gastroenterology 1982;82:213-7. Cummings JH. Dietary fibre and the intestinal microflora. In: Hallgren B, ed. Nutrition and the intestinal flora: proceedings of a symposium jointly sponsored by the Swedish Diaries Association and the Swedish Nutrition Foundation. Stockholm, Sweden: Almqvist & Wiksell International. 1983:7786. Paigen K, Williams B. Catabolite repression and other control mechanisms in carbohydrate utilization. Adv Microbial Physiol 1970:4:251-324. Weber FL Jr. Banwell JG, Fresard KF. Cummings JH. Effect of

544

WEBER ET AL.

lactulose and lactulose plus neomycin on the compartmentalization of stool nitrogen (abstr]. Gastroenterology 1982;82: 1208. 27. Cascino A, Cangiano C, Calcaterra V, Rossi-Fanelli F, Capocaccia L. Plasma amino acid imbalance in patients with liver disease. Dig Dis Sci 1978;23:591-8. 28. Hutson DC, Ono J, Dombro RS, Levi JU, Livingstone A, Zeppa R. A longitudinal study of tryptophan involvement in hepatic coma. Am J Surg 1979;137:235-9. 29. Rudman D, Galambos JT, Smith RB III, Salam A, Warren WD.

GASTROENTEROLOGY

Vol. 89. No. 3

Comparison of the effect of various amino acids upon the blood ammonia concentration of patients with liver disease. Am J Clin Nutr 1973;26:916-25. 30. Rudman D, Smith RB, Salam A, Warren WD, Galambos JT, Wegner J. Ammonia content of food. Am J Clin Nutr 1973;26: 487-90. 31. Adibi SA, Kim YS. Peptide absorption and hydrolysis. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1981:1073-95.