evidence for impaired assimilation of protein in chronic renal failure

Kidney International, Vol. 64 (2003), pp. 2196–2203

Evidence for impaired assimilation of protein in chronicrenal failure

BERT BAMMENS, KRISTIN VERBEKE, YVES VANRENTERGHEM, and PIETER EVENEPOEL

Department of Medicine, Division of Nephrology and Laboratory of Radiopharmaceutical Chemistry,University Hospital Gasthuisberg, Leuven, Belgium

Evidence for impaired assimilation of protein in chronic renalfailure.

Background. Protein malnutrition is a common finding inchronic renal failure (CRF) and is associated with poor out-come. We hypothesized that besides inadequate dietary pro-tein intake and alterations in metabolism, deficient protein as-similation (digestion and absorption) might contribute to thepathogenesis of protein malnutrition in uremia.

Methods. Protein assimilation was evaluated in 64 healthyvolunteers and 119 CRF patients by means of a 13C proteinbreath test and/or quantification of p-cresol in a 24-hour urinecollection. Both approaches provide reliable information onthe efficiency of protein assimilation. Breath test results wereexpressed as maximum percentage of administered dose of13C (%max) and cumulative percentage at the end of the test(%cumend). Data were stratified according to renal function.

Results. As compared to subjects with glomerular filtrationrate (GFR) ≥ 60 mL/min/1.73 m2, subjects with GFR < 30 mL/min/1.73 m2 demonstrated significantly lower breath test–derived parameters of protein assimilation (%max 3.97 ± 0.23 vs.5.20 ± 0.23, P = 0.0017; %cumend 13.91 ± 0.86 vs. 17.40 ± 0.80,P = 0.013) and significantly higher urinary output of p-cresol(54.88 mg/24 hours vs. 28.65 mg/24 hours, P = 0.0005). %max(r = 0.399, P < 0.0001), %cumend (r = 0.347, P = 0.0007), andurinary p-cresol (r = −0.229, P = 0.007) correlated significantlywith GFR. Serum albumin correlated significantly with %max(r = 0.399, P = 0.0002), %cumend (r = 0.408, P = 0.0001), andurinary p-cresol output (r = −0.186, P = 0.035).

Conclusion. Our data provide evidence that protein assimi-lation is impaired in CRF. This impairment might contribute toprotein malnutrition in CRF.

Protein malnutrition is a common finding in patientswith chronic renal failure (CRF). Prevalences up to 65%have been reported in patients with end-stage renal dis-ease (ESRD) [1–6]. Considerable evidence has accumu-lated over the last decade suggesting that malnutrition in

Key words: malnutrition, breath test, p-cresol, protein assimilation.

Received for publication December 2, 2002and in revised form March 26, 2003, and May 15, 2003Accepted for publication July 17, 2003

C© 2003 by the International Society of Nephrology

uremic patients is associated with poor outcome [7–9].A strong association has been demonstrated betweenmalnutrition, inflammation, and atherosclerosis, which isa major cause of morbidity and mortality in advancedCRF [3, 6].

Theoretically, protein malnutrition may result from (1)inadequate dietary protein intake, (2) impaired small in-testinal protein assimilation (digestion and absorption),(3) increased protein losses, and/or (4) alterations in pro-tein metabolism. To date, most researchers have focusedon the impact of disturbances of metabolism. Metabolicacidosis, a frequent complication of uremia, has beenshown to induce net protein catabolism by several mech-anisms [10–12]. High circulating levels of inflammatorycytokines, as seen in CRF patients [13–15], have beendemonstrated to stimulate muscle protein breakdown[16]. Finally, loss of muscle protein in uremia may berelated to impaired action of anabolic hormones such asinsulin.

Small intestinal assimilation of dietary protein, pro-viding essential and nonessential amino acids for proteinsynthesis, plays a pivotal role in the maintenance of to-tal body nitrogen balance. Information on the efficiencyof protein assimilation in uremia is limited, most prob-ably due to the lack of an easy and reliable measuringtechnique [17, 18]. Recently, the availability of protein,highly enriched with 13C, has enabled breath test tech-nology to be adapted for the study of protein assimila-tion [19]. Breath tests are simple, easily repeatable, andnoninvasive. Hence, they are very suitable for performingstudies on a large scale, both in healthy volunteers andin patients. Finally, they are very interesting as a func-tional test because the obtained information representsa dynamic evaluation rather than a static observation.

Protein escaping digestion and absorption in the smallintestine is intensely metabolized by the microbial floraof the colon. Quantitatively, p-cresol is one of the mostimportant phenolic compounds recovered in feces andurine [20]. It is a specific bacterial fermentation metabo-lite of the amino acid tyrosine. Phenolic compounds thatare not used for bacterial protein synthesis are largely

2196

Bammens et al: Protein assimilation in chronic renal failure 2197

absorbed from the colonic lumen, detoxified, and urinar-ily excreted [20, 21]. Urinary output of p-cresol has beendemonstrated to reflect the amount of protein escapingsmall intestinal digestion and absorption [21] and thusmay represent an indirect parameter of the efficiency ofprotein assimilation.

The aim of the present study was to evaluate whetherprotein assimilation is impaired in patients with renal fail-ure. We therefore studied a large cohort of patients andhealthy volunteers with varying glomerular filtration rate(GFR) by means of breath test technology and quantifi-cation of urinary p-cresol output.

METHODS

Subjects

Sixty-four volunteers and 119 patients with CRF wereincluded. The volunteers comprised members of the nurs-ing and laboratory staff, medical students, and mem-bers of an association of retired persons. All volunteersclaimed to be “healthy,” which was confirmed by a medi-cal history taken by a physician (B.B.). None of them wasknown to have renal disease or impaired renal function.No metabolic, cardiovascular, gastrointestinal, or respi-ratory disease was present. The patients were recruitedfrom the outpatient nephrology clinic. The causes of re-nal failure included polycystic kidney disease (N = 16),glomerular disease (N = 56), tubulointerstitial disease(N = 20), vascular disease (N = 3), and unknown etiolo-gies (N = 24). The study was approved by the EthicalCommittee of the University Hospital Gasthuisberg andinformed consent was obtained from all subjects.

Study design

Protein assimilation was evaluated in a cross-sectionalobservational design by means of breath test technologyand/or quantification of urinary p-cresol output.

13C breath test studies

Eighty-four subjects (54 healthy volunteers and 30CRF patients) were included in this part of the study.None of them (1) had a history of gastrointestinal diseaseor surgery, hepatic disease or diabetes mellitus, (2) useddrugs known to influence gastrointestinal motility (proki-netics or antiemetics) or protein assimilation (prebiotics,probiotics, antibiotics, or inhibitors of gastric acid secre-tion) during the preceding 3 months, (3) or had metabolicacidosis as defined by a serum bicarbonate of 20 mEq/Lor less.

13C protein breath test. Protein assimilation was evalu-ated after an overnight fast of at least 12 hours by means ofa 13C protein breath test. The solid test meal (being a pan-cake) consisted of 7.5 g of lyophilized 13C-leucine–labeledegg white, 17.2 g of lyophilized 13C-leucine–labeled egg

yolk, 3.75 g of lyophilized unlabelled egg white, 3 g ofmilk powder, 7 g of sugar, 17 g of flour, 130 mL ofwater, and 5 g of butter. The methodology for obtain-ing large amounts of egg proteins intrinsically labeledwith 13C has been described elsewhere [22]. The caloriccontent of the meal amounted to 327.4 kcal (18.67 gof protein, 6.15 g of fat, and 26.8 g of carbohydrates).Subjects were asked to consume the meal together with200 mL of water within 15 minutes. Breath samples forthe detection of 13CO2 were collected twice immediatelybefore the test meal and at 15-minute intervals for aperiod of 6 hours thereafter.

13C-leucine breath test. The absorption and the oxida-tive metabolism of leucine were evaluated in an unse-lected subgroup of 17 subjects (seven healthy volunteersand ten CRF patients). After an overnight fast of at least12 hours, a single-lumen duodenal tube was positionedin the second part of the duodenum under gastroscopiccontrol. Two hundred milligrams of 13C-leucine (99 MP)dissolved in 20 mL of water was administered as a bolusthrough the tube. Immediately after instillation, the tubewas removed. Breath samples for the detection of 13CO2

were collected twice immediately before the administra-tion of the label and thereafter every 5 minutes duringthe first hour and every 15 minutes for the following3 hours.

Sampling, analytic procedures, and calculations.Breath was collected by blowing air through a strawin a small tube, called an exetainer (Europa Scientific,Crewe, UK). The 13C content of the breath was de-termined by on-line gas chromatographic purification-isotope ratio mass spectometry (IRMS) (ABCA)(Europa Scientific, Crewe, UK). The d values given bythe IRMS were converted to percentage 13C recovery perhour (%dose/hour) of the initial amount administered ac-cording to calculations described in detail by Ghoos et al[23]. Cumulative percentages of label recovery (%cum)were calculated by means of the trapezoidal rule. Fromthese data, the following parameters of protein assimi-lation were derived: the maximum percentage of admin-istered dose of 13C excreted per hour (%max) and thecumulative percentage of 13C administered doserecovered in breath at the end of the test (%cumend).

Urinary output of p-cresol

One hundred and twenty-two subjects (34 healthy vol-unteers and 88 CRF patients) performed a 24-hour urinecollection. The collection bottles were refrigerated (0 to4◦C) at the participant’s home to prevent bacterial growthand brought to the laboratory within hours of termina-tion of the collection. After measurement of the volume,samples were taken and stored at –80◦C until analysis.

p-Cresol was analyzed by gas chromatography-massspectrometry (GC-MS) technology. The pH of 950 lL

2198 Bammens et al: Protein assimilation in chronic renal failure

of urine sample was adjusted to pH 1 with concentratedH2SO4 and the solution was heated to 90◦C for 30 min-utes. After a cooling down period to ambient tempera-ture, 50 lL 2,6-dimethylphenol solution (20 mg/100 mL)was added as internal standard. One milliliter ethyl ac-etate was added for the extraction of p-cresol. The solu-tion was well mixed during 30 seconds and centrifugedat 3300 rpm for 20 minutes. Five hundred microliters ofthe supernatant was dried over anhydrous sodium sul-fate. One hundred microliters of the resultant sample wastransferred to the GC-MS (Trace GC-MS) (Thermofinni-gan, San Jose, CA, USA) for automatic splitless injec-tion of 0.5 lL. The analytical column used was a 30 m ×0.32 mm internal diameter, film thickness 1 lm AT5-MS(Alltech, Deerfield, IL, USA). Helium GC grade wasused as a carrier gas with a constant flow of 1.3 mL/min.The oven was programmed from 75◦C (isotherm for5 minutes) to 280◦C with 15◦C per minute. After sepa-ration, p-cresol was identified by MS (Electron Impactfull scan mode from m/z 59 to m/z 590 at 2 scan/second).Quantitative results were obtained by the internal stan-dard method. Results were calculated as concentrations(mg/L) and data were expressed as amounts excreted per24 hours. To normalize for differences in dietary proteinintake, results were also expressed as the ratio of urinaryp-cresol to urinary urea nitrogen according to the follow-ing equation:

Urinary p-cresol/urea nitrogen ratio

= urinary p-cresol (mg/24 h)urinary urea nitrogen (g/24 h)

Assuming no changes in serum urea nitrogen and bodyweight over a 24-hour period in stable nondialyzed CRFpatients, urinary urea nitrogen can be used as a parameterfor protein intake [24].

Anthropometrics and biochemical analyses

Body mass index (BMI) was calculated from measuredheight and body weight (kg/m2). Serum albumin, serumbicarbonate, serum creatinine, and serum and urinaryurea nitrogen concentrations were measured using stan-dard laboratory techniques. GFR was estimated by theCockcroft and Gault formula. Values were normalized forbody surface area (BSA) by multiplying with 1.73/BSA(m2) and were expressed as mL/min/1.73 m2. BSA wascalculated from the formula of Haycock, Schwartz, andWisotsky [25]:

BSA(m2) = 0.024265 × weight (kg)0.5378

× height (cm)0.3964

For comparative purposes subjects were categorized intothree groups according to GFR: group A (GFR ≥ 60 mL/min/1.73 m2), group B (GFR 59 to 30 mL/min/1.73 m2),and group C (GFR < 30 mL/min/1.73 m2).

Statistical analysis

Data are expressed as mean ± SEM if normallydistributed or as median and range otherwise. Nor-mality of distributions was assessed by Kolmogorov-Smirnov and Cramer-von Mises test. In the figures, 95%confidence intervals of the data are indicated whereappropriate.

Differences in measured and calculated data betweenmore than two subject groups were evaluated using para-metric one-way analysis of variance (ANOVA) or theKruskal-Wallis test, followed by Bonferroni correctedpost hoc analysis for pair-wise comparisons. In two sam-ple analyses unpaired Student t tests or Mann-WhitneyU tests were used. Pearson correlation coefficients werecalculated to assess possible associations between 13Cprotein breath test results, urinary output of p-cresoland urea nitrogen, age, serum albumin, BMI, and re-nal function. P values less than 0.05 were consideredsignificant. The SAS version 8.02 (SAS Institute, Cary,NC, USA) software program was used for the statisticalanalysis.

RESULTS

Breath test studies

Demographics and nutrition. Subjects with GFR ≥60 mL/min/1.73 m2 (group A) were significantly youngerthan subjects with GFR < 60 mL/min/1.73 m2 (groupsB and C). Genders were equally distributed among thedifferent groups. A lower serum albumin was seen in sub-jects with GFR < 30 mL/min/1.73 m2. Statistical signifi-cance, however, was not reached (Table 1).

Breath test data. The mean 13CO2 excretion curves,expressed as %dose/hour and %cum of the administereddose obtained after ingestion of the 13C protein test mealare depicted in Figure 1. The curves in the subjects withGFR < 30 mL/min/1.73 m2 (group C) are flattened ascompared to those observed in subjects with GFR ≥60 mL/min/1.73 m2 (group A). Breath test derived pa-rameters of protein assimilation were significantly lowerin group C as compared to group A (%max 3.97 ± 0.23 vs.5.20 ± 0.23, P = 0.0017; %cumend 13.91 ± 0.86 vs. 17.40 ±0.80, P = 0.013) (Table 1). The differences remained sig-nificant after stratification for age (%max 4.02 ± 0.25 vs.5.71 ± 0.70, P = 0.014; %cumend 14.01 ± 0.93 vs. 19.94 ±2.52, P = 0.017).

The mean 13CO2 excretion curves, expressed as%dose/hour and %cum of the administered dose ob-tained after intraduodenal instillation of 13C-leucine ingroups A and C are depicted in Figure 2. The curves arealmost identical in both groups. There were no differencesin breath test derived parameters (%max 7.80 ± 0.84 vs.7.49 ± 0.52, P = 0.78; %cumend 14.69 ± 1.27 vs. 13.44 ±1.17, P = 0.50) (Table 1).


Table 1. 13C breath test studies

Group A Group B Group C(GFR ≥ 60 mL/min/1.73 m2) (GFR 59 to 30 mL/min/1.73 m2) (GFR < 30 mL/min/1.73 m2)

Number 46 8 30Age years 23 (18–70) 71 (59–78)a 55 (33–74)a,b

Male/female 23/23 4/4 17/13Body weight kg 66.0 (47.0–116.0) 76.2 (54.5–87.5) 67.65 (44.1–101.6)Body mass index kg/m2 21.48 (17.96–33.62) 25.83 (21.98–28.23)a 23.77 (15.55–32.55)Bicarbonate mEq/L 23.55 ± 0.28 24.08 ± 0.71 24.71 ± 0.71Serum urea nitrogen mg/dL 25.0 (14.0–42.0) 33.5 (25.0–41.0) 170.0 (65.0–255.0)a,b

Serum creatinine mg/dL 0.92 (0.73–1.19) 1.09 (1 00–1.45) 5.93 (1.78–10.61)a,b

Glomerular filtration rate mL/min/1.73 m2 102.78 (60.36–173.31) 56.98 (46.11–58.74)a 12.85 (6.79–29.78)a

Serum albumin g/L 41.35 (28.80–46.30) 41.75 (37.40–43.90) 40.30 (33.10–51.07)Urinary urea nitrogen g/24 hours 19.9 (9.34–36.05) (N = 26) 14.65 (9.3–21.3) (N = 8) 16.0 (5.1–22.8) (N = 19)13C protein breath test

%max 5.20 ± 0.23 4.73 ± 0.32 3.97 ± 0.23a

%cumend 17.40 ± 0.80 16.72 ± 0.82 13.91 ± 0.86a

13C-leucine breath test%max 7.49 ± 0.52 (N = 7) — 7.80 ± 0.84 (N = 10)%cumend 13.44 ± 1.17 (N = 7) — 14.69 ± 1.27 (N = 10)

Data are expressed as mean ± SEM if normally distributed or as median and range otherwise. Differences between groups were evaluated by one-way analysis ofvariance/unpaired Student t test or Kruskal-Wallis test where appropriate, followed by post hoc pair-wise comparisons.

aP < 0.05 vs. group A; bP < 0.05 vs. group B.

65

4

3

2

00 60 120 180 240 300 360 360300240180120600

1

% d

ose/

h

20

15

10

0

5

% c

um

Fig. 1. 13CO2 excretion curves of the 13C pro-tein breath test expressed as %dose/hour (A)and %cum (B). Mean values and 95% con-fidence intervals are illustrated. Symbols are:(�) group A; (�) group C.

1210

8

6

4

02

% d

ose/

h

0 060 60120 120180 180240 240

15

10

5

0

% c

um

Fig. 2. 13CO2 excretion curves of the 13C-leucine breath test expressed as %dose/hour(A) and %cum (B). Mean values and 95%confidence intervals are illustrated. Symbolsare: (�) group A; (�) group C.

Urinary output of p-cresol

Demographics and nutrition. Subjects with GFR ≥60 mL/min/1.73 m2 (group A) were significantly youngerthan subjects with GFR < 60 mL/min/1.73 m2 (groupsB and C). There were no statistical differences in gen-der distribution among the different groups. Serum al-bumin was significantly lower in subjects with GFR <

30 mL/min/1.73 m2 as compared to subjects with GFR ≥60 mL/min/1.73 m2 (Table 2).

Urinary output of p-cresol. The 24-hour urinary p-cresol output in the three study groups is illustrated inFigure 3A and Table 2. Groups B and C subjects hadsignificantly increased urinary p-cresol output as com-

pared to subjects from group A. After normalization fordietary protein intake, an even more pronounced step-wise increase along with renal function impairment wasobserved (Fig. 3B) (Table 2). After stratification for age,the differences in urinary p-cresol output, and urinaryp-cresol/urea nitrogen ratio remained statistically signif-icant (P = 0.0021, P = 0.0004, respectively).

Correlations

A significant positive correlation was observed be-tween 13C protein breath test–derived parameters andGFR (%max r = 0.399, P < 0.0001; %cumend r = 0.347,P = 0.0007) (Fig. 4). A significant inverse relationship


Table 2. Urinary output of p-cresol and urea nitrogen

Group A Group B Group C(GFR ≥ 60 mL/min/1.73 m2) (GFR 59 to 30 mL/min/1.73 m2) (GFR < 30 mL/min/1.73 m2)

Number 44 29 49Age years 35.5 (20–70) 69 (24–83)a 68 (35–92)a

Male/female 17/27 17/12 27/22Body weight kg 64.5 (47.0–116.0) 69.3 (45.6–108.4) 67.7 (44.1–101.6)Body mass index kg/m2 22.21 (18.03–38.41) 25.66 (18.27–37.07)a 24.71 (15.63–30.75)Serum urea nitrogen mg/dL 30.0 (15.0–75.0) 50.0 (25.0–118.0)a 122.5 (45.0–234.0)a,b

Serum creatinine mg/dL 0.93 (0.73–1.38) 1.50 (1.00–2.13)a 3.20 (1.78–8.91)a,b

Glomerular filtration rate mL/min/1.73 m2 86.35 (60.36–111.18) 46.11 (31.82–58.74)a 19.49 (7.04–29.93)a,b

Serum albumin g/L 41.7 (31.2–50.0) 40.95 (21.8–48.1) 40.6 (24.6–46.0)a

Urinary p-cresol mg/24 hours 28.65 (0.33–381.11) 50.67 (7.32–231.93)a 54.88 (5.37–417.91)a

Urinary urea nitrogen g/24 hours 20.05 (9.34–36.05) 15.7 (5.2–40.1)a 16.0 (3.09–34.3)a

Urinary p-cresol/urea nitrogen ratio 1.38 (0.02–14.60) 3.23 (0.72–14.45)a 4.19 (0.41–30.19)a

Data are expressed as median and range. Differences between groups were evaluated by Kruskal-Wallis test followed by post hoc pair-wise comparisons.aP < 0.05 vs. group A; bP < 0.05 vs. group B.

5

4

3

2

1

0

B

Urin

ary

p-cr

esol

, mg/

24h/

urin

ary

ure

a ni

trog

en, g

/24h

GroupA

GroupB

GroupC

**

55504540353025201510

50

A

Urin

ary

p-cr

esol

, mg/

24h

GroupA

GroupB

GroupC

**

Fig. 3. Urinary output of p-cresol. (A) The median values of urinaryp-cresol output (mg/24 hours) in the three study groups. (B) The me-dian values of the urinary p-cresol (mg/24 hours)/urinary urea nitrogen(g/24 hours) ratio. ∗P < 0.05 vs. group A.

was found between urinary p-cresol output (r = −0.229,P = 0.007) and urinary p-cresol/urea nitrogen ratio (r =−0.310, P = 0.0003) on the one hand and GFR onthe other. Urinary p-cresol output was positively corre-lated with urinary urea nitrogen output (r = 0.193, P =0.028). Similar relationships were found, when the analy-sis was confined to the CRF patients only: %max vs. GFR(r = 0.470, P = 0.007), %cumend vs. GFR (r = 0.374,

30

20

10

00 50 100 150

% c

umen

d

GFR, mL/min/1.73 m2

r = 0.347P = 0.0007

Fig. 4. %cumend of the 13C protein breath test vs. glomerular fil-tration rate (GFR) (mL/min/1.73 m2). r is Pearson correlationcoefficient.

P = 0.035), and urinary p-cresol/urea nitrogen ratio vs.GFR (r = −0.233, P = 0.023). Serum albumin was posi-tively correlated with parameters of protein assimilation(%max r = 0.399, P = 0.0002; %cumend r = 0.408, P =0.0001) and inversely with urinary output of p-cresol (r =−0.186, P = 0.034). There was a positive correlation ofalbumin with GFR (r = 0.239, P = 0.002) and BMI (r =0.423, P < 0.0001). Urinary urea nitrogen output waspositively correlated with GFR (r = 0.301, P = 0.0005)(Table 3). No correlations were found between age and13C protein breath test–derived parameters (%max r =−0.080, P = 0.448; %cumend r = −0.009, P = 0.931), norbetween age and urinary p-cresol output (r = 0.159, P =0.060).

DISCUSSION

Information on the efficiency of protein assimilation inCRF patients is scarce, most probably due to the lack ofan easy and reliable measuring technique. In the present


Table 3. Correlations

Urinary Urinary urea p-cresol/urea Serum%cumend %max p-cresol nitrogen nitrogen ratio albumin BMI GFR

%cumend 1 0.941c 0.160 −0.147 0.110 0.408c 0.203 0.347c

%max 1 0.095 −0.137 0.060 0.399c 0.164 0.399c

Urinary p-cresol 1 0.193a 0.837c −0.186a −0.107 −0.299b

Urinary urea nitrogen 1 −0.171 0.109 −0.008 0.301c

p-cresol/urea nitrogen ratio 1 −0.159 −0.087 −0.310c

Serum albumin 1 0.423c 0.239c

BMI 1 −0.073GFR 1

Abbreviations are: BMI, body mass index; GFR, glomerular filtration rate. Pearson correlation coefficients (r) for 13C protein breath test parameters, urinary outputof p-cresol, and urea nitrogen, serum albumin, BMI, and renal function.

aP < 0.005; bP < 0.01; cP < 0.005.

study, protein assimilation was evaluated by means of arecently developed and validated 13C protein breath test[19, 21, 26, 27] and by quantification of the urinary out-put of p-cresol [19, 21]. Our breath test data provide evi-dence that protein assimilation is impaired in uremia. Themean 13CO2 excretion curve, expressed as %dose/hour,was flattened in subjects with an estimated GFR <

30 mL/min as compared to subjects with GFR ≥ 60 mL/min and mathematically derived parameters of proteinassimilation (%cumend, %max) were significantly lower inthe former group (Fig. 1) (Table 1). Moreover, a positiverelationship between these parameters and renal functionwas demonstrated (Fig. 4) (Table 3). We recognize thatthe appearance of 13CO2 in breath is not only influencedby determinants of the protein assimilation process, butalso by factors related to the postabsorptive metabolism(oxidation rate, bicarbonate pool retention). In order toascertain the validity of the 13C breath test as a tool for theevaluation of protein assimilation in uremia, it was aimedto minimize metabolic interference by excluding patientswith diabetes mellitus, hepatic failure, and/or metabolicacidosis. These conditions have previously been shownto interfere with protein metabolism [10–12, 28–33]. Inthe absence of overt metabolic acidosis, uremic patientsshow normal adaptive responses to a low-protein diet [11,34–36]. One of these responses consists of a reductionin irreversible oxidation of essential amino acids, whichwould result in a lower 13CO2 excretion derived fromadministered 13C-leucine. However, major differences inprotein intake as assessed by urinary urea nitrogen out-put [24] were not found between the different groups inthis part of the study. Furthermore, mean 13CO2 excre-tion curves, obtained after intraduodenal instillation ofa standard 13C-leucine solution, did not differ betweengroups, suggesting that the observed differences in 13Cprotein breath test results merely reflect impaired proteindigestion in uremia rather than differences in absorptionand/or postabsorptive metabolism.

p-Cresol is a unique bacterial fermentation metabo-lite of tyrosine in the colon. Since most of the p-cresol

produced in the colon is urinarily excreted, it can beassumed that in a steady-state situation, urinary outputreflects colonic generation [19–21]. Significantly higherurinary p-cresol excretion was seen in subjects withGFR < 60 mL/min/1.73 m2 than in subjects with GFR ≥60 mL/min/1.73 m2 (Fig. 3A) (Table 2). In agreement witha previous study, the colonic production of p-cresol wasshown to correlate with the dietary protein intake [21](Table 4). However, after normalization for dietary pro-tein intake, the stepwise increase of urinary p-cresol out-put along with renal function impairment was even morestriking (Fig. 3B) (Table 2). These observations supportthe hypothesis of impaired protein assimilation in ure-mia. We also analyzed the data of the CRF patients sep-arately. The results of this analysis were almost identicalto those obtained in the entire study population. Over-all, the correlation data indicate that the efficiency ofprotein assimilation declines along with renal functionimpairment. Whether protein assimilation is impaired inpatients with incipient nephropathy remains an unan-swered question. Although our study was not designed toresolve this issue, a reanalysis of the urinary data for thispurpose at least does not support this hypothesis (data notshown).

Our findings are not that surprising considering thebroad range of uremia-related gastrointestinal abnor-malities reported in literature, including gastrointestinalmotility disorders [37–41], small bowel bacterial over-growth [42, 43], gastric hypochlorhydria [26, 37, 44], anduremic pancreopathy [44–52]. Each of them could theo-retically disturb protein digestion and/or absorption. Thelack of a correlation between the 13C protein breath testresults and the urinary output of p-cresol (Table 3), inour opinion, does not question our conclusions but ratherreflects the fact that urinary output of p-cresol is deter-mined not only by protein assimilation efficiency. Urinaryoutput of p-cresol is known to be influenced by dietaryprotein intake. Both quantity and quality of the proteinmay be important in this regard [21, 53]. We normal-ized our data for the quantity of dietary protein. Since


subjects were on a free diet, however, we were not ableto evaluate the possible impact of qualitative differences.Other factors involved are characteristics of the bacte-rial flora and colonic transit time [54, 55]. In contrast, thebreath test results are not influenced by these factors andhence only depend on the efficiency of the assimilationprocess.

As in other studies [1–6], we found a positive associ-ation between the nutritional parameter serum albuminand GFR. Of interest were the correlations found be-tween serum albumin and protein assimilation param-eters (%cumend, %max, and urinary output of p-cresol).These correlations were independent of C-reactive pro-tein levels (multivariate regression analysis, data notshown). Although there was a trend for an associationwith BMI, statistical significance was not reached, prob-ably due to small power. The findings strengthen thehypothesis of impaired small intestinal protein assimi-lation as a contributing factor to malnutrition in uremia.Although it was shown that protein requirements of non-dialyzed CRF patients do not differ from those of nor-mal subjects [11, 36], impaired protein assimilation willat least make them more vulnerable in case of lower di-etary protein intake due to intercurrent illnesses. More-over, it will probably aggravate the detrimental effects ofinflammation and metabolic acidosis.

Further studies are required to confirm our data andto elucidate the underlying pathogenetic mechanisms.

CONCLUSION

Our data provide for the first time evidence that pro-tein assimilation is impaired in uremia. Impaired proteinassimilation might, besides other well-known factors aslower dietary protein intake, metabolic acidosis, inflam-mation, and resistance to anabolic hormones, contributeto protein malnutrition in uremia. Furthermore, our find-ings underscore the role of the colon in the uremic syn-drome, since bacterial fermentation metabolites, such asp-cresol, are increasingly recognized as important ure-mic toxins [56]. This study demonstrates that besides adecline in renal clearance an increased colonic produc-tion may contribute to the accumulation of these toxins inuremia.

ACKNOWLEDGMENTS

T. Coopmans, M. Dekens, V. De Preter, N. Gorris, A. Luypaerts,S. Rutten, L. Swinnen are acknowledged for their excellent technicalassistance. This work was supported by the FWO grant no. 1127602N.

Part of this work was presented as a poster at the American Societyof Nephrology Annual Meeting, November 2002.

Reprint requests to Pieter Evenepoel, M.D., Ph.D., Department ofMedicine, Division of Nephrology, University Hospital Gasthuisberg,Herestraat 49, B-3000 Leuven, Belgium.E-mail: [email protected]

REFERENCES

1. QURESHI AR, ALVESTRAND A, DANIELSSON A, et al: Factors predict-ing malnutrition in hemodialysis patients: A cross-sectional study.Kidney Int 53:773–782, 1998

2. QURESHI AR, ALVESTRAND A, DIVINO-FILHO JC, et al: Inflamma-tion, malnutrition and cardiac disease as predictors of mortality inhemodialysis patients. J Am Soc Nephrol 13(Suppl 1):S28–S36, 2002

3. STENVINKEL P, HEIMBURGER O, PAULTRE F, et al: Strong associationbetween malnutrition, inflammation and atherosclerosis in chronicrenal failure. Kidney Int 55:1899–1911, 1999

4. BURROWES JD, COCKRAM DB, DWYER JT, et al: Cross-sectional re-lationship between dietary protein and energy intake, nutritionalstatus, functional status, and comorbidity in older versus youngerhemodialysis patients. J Ren Nutr 12:87–95, 2002

5. ROCCO MV, PARANANDI L, BURROWES JD, et al: Nutritional status inthe HEMO study cohort at baseline. Am J Kidney Dis 39:245–256,2002

6. FUNG F, SHERRARD DJ, GILLEN DL, et al: Increased risk for car-diovascular mortality among malnourished end-stage renal diseasepatients. Am J Kidney Dis 40:307–314, 2002

7. MITTMAN N, AVRAM MM, OO KK, CHATTOPADHYAY J: Serum pre-albumin predicts survival in hemodialysis and peritoneal dialysis:10 years of prospective observation. Am J Kidney Dis 38:1358–1364,2001

8. KALANTAR-ZADEH K, KOPPLE JD, BLOCK G, HUMPHREYS MH: Amalnutrition-inflammation score is correlated with morbidity andmortality in maintenance hemodialysis patients. Am J Kidney Dis38:1251–1263, 2001

9. STENVINKEL P, BARANY P, CHUNG SH, et al: A comparative analysis ofnutritional parameters as predictors of outcome in male and femaleESRD patients. Nephrol Dial Transplant 17:1266–1274, 2002

10. PRICE SR, MITCH WE: Metabolic acidosis and uremic toxicity: Pro-tein and amino acid metabolism. Semin Nephrol 14:232–237, 1994

11. MITCH WE, PRICE SR, MAY RC, et al: Metabolic consequences ofuremia: Extending the concept of adaptive responses to proteinmetabolism. Am J Kidney Dis 23:224–228, 1994

12. PICKERING WP, PRICE SR, BIRCHER G, et al: Nutrition in CAPD:Serum bicarbonate and the ubiquitin-proteasome system in muscle.Kidney Int 61:1286–1292, 2002

13. PANICHI V, MIGLIORI M, DE PIETRO S, et al: C-reactive proteinand interleukin-6 levels are related to renal function in predialyticchronic renal failure. Nephron 91:594–600, 2002

14. PECOITS-FILHO R, BARANY P, LINDHOLM B, et al: Interleukin-6 isan independent predictor of mortality in patients starting dialysistreatment. Nephrol Dial Transplant 17:1684–1688, 2002

15. YEUN JY, LEVINE RA, MANTADILOK V, KAYSEN GA: C-reactive pro-tein predicts all-cause and cardiovascular mortality in hemodialysispatients. Am J Kidney Dis 35:469–476, 2000

16. MITCH WE: Insights into the abnormalities of chronic renal diseaseattributed to malnutrition. J Am Soc Nephrol 13(Suppl 1):S22–S27,2002

17. MOSENTHIN R, SAUER WC, AHRENS F: Dietary pectin’s effect on ilealand fecal amino acid digestibility and exocrine pancreatic secretionsin growing pigs. J Nutr 124:1222–1229, 1994

18. READ NW, AL JANABI MN, BATES TE, BARBER DC: Effect of gas-trointestinal intubation on the passage of a solid meal through thestomach and small intestine in humans. Gastroenterology 84:1568–1572, 1983

19. EVENEPOEL P, CLAUS D, GEYPENS B, et al: Evidence for impairedassimilation and increased colonic fermentation of protein, relatedto gastric acid suppression therapy. Aliment Pharmac Ther 12:1011–1019, 1998

20. BONE E, TAMM A, HILL M: The production of urinary phenols by thegut bacteria and their possible role in the causation of large bowelcancer. Am J Clin Nutr 29:1148–1454, 1976

21. EVENEPOEL P, CLAUS D, GEYPENS B, et al: Amount and fate of eggprotein escaping assimilation in the small intestine of humans. AmJ Physiol 277:G935–G943, 1999

22. EVENEPOEL P, HIELE M, LUYPAERTS A, et al: Production of egg pro-teins, enriched with L-leucine-13C1, for the study of protein assimila-tion in humans using the breath test technique. J Nutr 127:327–331,1997


23. GHOOS Y, GEYPENS B, MAES B, et al: Breath tests in gastric emp-tying and transit studies: Technical aspects of 13CO2-breath tests,in Progress in Understanding and Management of Gastro-intestinalMotility Disorders, edited by Janssens J, Belgium, KU Leuven, 1993,pp 169–180

24. MARONI BJ, STEINMAN TI, MITCH WE: A method for estimatingnitrogen intake of patients with chronic renal failure. Kidney Int27:58–65, 1985

25. HAYCOCK G, SCHWARTZ G, WISOTSKY D: Geometric method for mea-suring body surface area: A height-weight formula validated in in-fants, children and adults. J Pediatr 93:62–66, 1978

26. EVENEPOEL P: Alteration in digestion and absorption of nutrientsduring profound acid suppression. Best Pract Res Clin Gastroenterol15:539–551, 2001

27. EVENEPOEL P, HIELE M, GEYPENS B, et al: 13C-egg white breath test:A non-invasive test of pancreatic trypsin activity in the small intes-tine. Gut 46:52–57, 2000

28. HALVATSIOTIS PG, TURK D, ALZAID A, et al: Insulin effects on leucinekinetics in type 2 diabetes mellitus. Diabetes Nutr Metab 15:136–142,2002

29. HALVATSIOTIS P, SHORT KR, BIGELOW M, NAIR KS: Synthesis rate ofmuscle proteins, muscle functions, and amino acid kinetics in type2 diabetes. Diabetes 51:2395–2404, 2002

30. NAIR KS, FORD GC, EKBERG K, et al: Protein dynamics in wholebody and in splanchnic and leg tissues in type I diabetic patients. JClin Invest 95:3926–3937, 1995

31. LUZI L, CASTELLINO P, SIMONSON DC, et al: Leucine metabolism inIDDM. Role of insulin and substrate availability. Diabetes 39:38–48,1990

32. STATEN MA, MATTHEWS DE, BIER DM: Leucine metabolism in typeII diabetes mellitus. Diabetes 35:1249–1253, 1986

33. MCCULLOUGH AJ, MULLEN KD, TAVILL AS, KALHAN AD: In vivodifferences between the turnover rates of leucine and leucine’s ke-toacid in stable cirrhosis. Gastroenterology 103:571–578, 1992

34. MITCH WE: Uremia and the control of protein metabolism. Nephron49:89–93, 1988

35. LIM VS, YARASHESKI KE, FLANIGAN AMJ: The effect of uraemia, aci-dosis, and dialysis treatment on protein metabolism: A longitudinalleucine kinetic study. Nephrol Dial Transplant 13:1723–1730, 1998

36. GOODSHIP THJ, MITCH WE, HOERR RA, et al: Adaptation to low-protein diets in renal failure: Leucine turnover and nitrogen balance.J Am Soc Nephrol 1:66–75, 1990

37. BABAK E: Gastrointestinal complications of renal failure. Gastroen-terol Clin North Am 27:875–892, 1998

38. VAN VLEM B, SCHOONJANS R, VANHOLDER R, et al: Dyspepsia and gas-tric emptying in chronic renal failure patients. Clin Nephrol 56:302–307, 2001

39. EL-LAKANY S, EAGON PK, GAVALER JS, et al: Gastrointestinal func-tion, morphology, and immune status in uremia. Nutrition 6:461–468, 1990

40. SCHIFFL H, MUCKE C, SORODOC J, BRENDEL C: Gastrointestinal tran-sit of an indigestible solid in patients on CAPD. Perit Dial Int 20:787–789, 2000

41. BROWN-CARTWRIGHT D, SMITH HJ, FELDMAN M: Gastric emptyingof an indigestible solid in patients with end-stage renal disease oncontinuous ambulatory peritoneal dialysis. Gastroenterology 95:49–51, 1988

42. VANTRAPPEN G, JANSSENS J, HELLEMANS J, GHOOS Y: The interdi-gestive motor complex of normal subjects and patients with bacte-rial overgrowth of the small intestine. J Clin Invest 59:1158–1166,1977

43. SIMENHOFF ML, SAUKKONEN JJ, BURKE JF, et al: Bacterial popula-tions of the small bowel in uremia. Nephron 22:63–68, 1978

44. MALYSZKO J, SOSNOWSKI S, MAZERSKA M, et al: Gastric and pancre-atic functions in haemodialyzed patients. Int Urol Nephrol 27:471–478, 1995

45. ABU-ALFA A, IVANOVICH P, MUJAIS SK: Uremic exocrine pancre-opathy. Nephron 48:94–100, 1988

46. MASOERO G, BRUNO M, GALLO L, et al: Increased serum pancreaticenzymes in uremia: Relation with treatment modality and pancre-atic involvement. Pancreas 13:350–355, 1996

47. OWYANG C, MILLER LJ, DIMANGO EP, et al: Pancreatic exocrinefunction in severe human chronic renal failure. Gut 23:357–361,1982

48. VENTRUCCI M, CAMPIERI C, DI STEFANO M, et al: Alterations ofexocrine pancreas in end-stage renal disease. Do they reflect aclinically relevant uremic pancreopathy. Dig Dis Sci 40:2576–2581,1995

49. VENTRUCCI M, CIPOLLA A, MIDDONNO M, et al: Impaired fecal elas-tase excretion in uremic pancreopathy. Dig Dis Sci 45:2265–2269,2000

50. SCHURR D, LEVY E, GOLDSTEIN R, et al: Intestinal fat malabsorptionin the uremic rat. Int J Pediatr Nephrol 8:129–134, 1987

51. DRUKKER A, LEVY E, BRONZA N, et al: Impaired intestinal fat ab-sorption in chronic renal failure. Nephron 30:154–160, 1982

52. ABU-ALFA A, MUJAIS SK: Uremic pancreopathy: Impaired secre-tory function in vitro. Am J Kidney Dis 20:585–588, 1992

53. GEYPENS B, CLAUS D, EVENEPOEL P, et al: Influence of dietary proteinsupplements on the formation of bacterial metabolites in the colon.Gut 41:70–76, 1997

54. CUMMINGS JH, HILL MJ, BONE ES, et al: The effect of meat proteinand dietary fibre on colonic function and metabolism. II. Bacte-rial metabolites in feces and urine. Am J Clin Nutr 32:2094–2101,1979

55. BIRKETT A, MUIR J, PHILLIPS J, et al: Resistant starch lowers fecalconcentrations of ammonia and phenols in humans. Am J Clin Nutr63:766–772, 1996

56. VANHOLDER R, DE SMET R, LESAFFER G: p-Cresol: A toxin revealingmany neglected but relevant aspects of uraemic toxicity. NephrolDial Transplant 14:2813–2815, 1999

evidence for impaired assimilation of protein in chronic renal failure

Documents