点击显示 收起
1 From the Department of Medicine, Division of Nephrology (BB, PE, and YV) and the Laboratory of Digestion and Absorption (KV), University Hospital Gasthuisberg, Leuven, Belgium
2 Presented in part at the American Society of Nephrology Annual Meeting, November 2003. 3 Supported by grant no. 1127602N from the Fonds voor Wetenschappelijk Onderzoek. 4 Address reprint requests to P Evenepoel, Department of Medicine, Division of Nephrology, University Hospital Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium. E-mail: pieter.evenepoel{at}uz.kuleuven.ac.be.
ABSTRACT
Background: Protein malnutrition is a common finding in renal disease. Recently, we showed that impaired protein assimilation (digestion and absorption) may contribute to protein malnutrition in nondiabetic patients with chronic renal failure.
Objective: The aim of the present study was to evaluate whether these findings can be extended to the dialysis population. Moreover, relations with indexes of the malnutrition-inflammation-atherosclerosis (MIA) syndrome were investigated.
Design: Protein assimilation was evaluated in 24 healthy control subjects and in 40 patients with end-stage renal disease (ESRD; 14 treated with peritoneal dialysis and 26 with hemodialysis) by means of a [13C]protein breath test, quantification of the generation rate of p-cresol, or both methods. Both approaches provide reliable information on the efficiency of protein assimilation. Breath test results were expressed as the maximum percentage recovery per hour of the administered dose of 13C (%max) and as the cumulative percentage recovery at the end of the test (%cumend). Several indexes of nutritional status, inflammation, and atherosclerosis were assessed.
Results: Compared with the control subjects, ESRD patients had significantly lower breath-test derived indexes of protein assimilation [%max = 3.75 ± 0.30 compared with 4.90 ± 0.25, P = 0.006; %cumend = 12.41 (5.74–23.22) compared with 16.87 (9.42–22.99), P = 0.020] and a higher 24-h p-cresol generation rate corrected for dietary protein intake [3.89 (0.48–11.60) compared with 2.81 (0.21–11.20) mg p-cresol/g urea nitrogen; P = 0.028]. The presence of impaired protein assimilation was associated with indexes of the MIA syndrome.
Conclusion: Our study provides evidence that protein assimilation is impaired in ESRD patients. Moreover, this disorder is associated with the severity of the MIA syndrome.
Key Words: Malnutrition-inflammation-atherosclerosis syndrome • protein assimilation • p-cresol • breath test • end-stage renal disease
INTRODUCTION
Protein malnutrition, defined as low amounts of visceral and somatic proteins and a loss of lean body mass, is a common finding in patients with end-stage renal disease (ESRD). Depending on the variable measured, the prevalence of protein malnutrition ranges from 10% to 75% in hemodialysis (HD) patients and from 18 to 51% in peritoneal dialysis (PD) patients (1–6). Besides an inadequate dietary protein intake and increased protein losses, several other mechanisms have been shown to contribute to the impairment of nutritional status in ESRD. To date, most research has focused on the effect of disturbances in protein metabolism (breakdown, synthesis, and oxidation). The presence of metabolic acidosis has been shown to stimulate amino acid and protein catabolism by increasing the activity of the branched-chain ketoacid dehydrogenase, by impairing albumin synthesis, and by stimulating muscle protein breakdown through the ATP-ubiquitin-proteasome pathway (7–11). The loss of muscle protein in uremia is also caused by impaired action of anabolic hormones such as insulin and insulin-like growth factor I (11). Numerous studies highlight the high concentrations of inflammatory cytokines in ESRD patients (12–14) and consistently describe a relation between these indicators of inflammation on the one hand and impairment of nutritional status and accelerated atherosclerosis on the other (13–23). Although the exact mechanisms responsible for the interplay between the entities of the so-called malnutrition-inflammation-atherosclerosis (MIA) syndrome (15) are still being discovered, some in vitro and animal data suggest the involvement of the ATP-ubiquitin-proteasome pathway as well (10, 24, 25). Recently, we showed another mechanism that possibly contributes to the impaired nutritional status in uremic patients, notably the impairment of small intestinal protein assimilation (digestion and absorption) (26). In a group of nondiabetic, nonacidotic patients with severe renal insufficiency (glomerular filtration rate <30 mL · min–1 · 1.73 m–2), indicators of protein assimilation were significantly lower than in subjects with a glomerular filtration rate 60 mL · min–1 · 1.73 m–2. The question arising from these findings in nondialyzed patients is whether they can be extended to patients with ESRD treated with PD or HD. Moreover, in view of the MIA syndrome (15), it is worthwhile to investigate possible relations between indexes of protein assimilation and indexes of malnutrition, inflammation, and atherosclerosis. It was the aim of the present study to answer these questions by studying protein assimilation by means of breath-test technology in healthy control subjects and in ESRD patients treated with PD or HD. Furthermore, the amount of protein escaping digestion and absorption in the small intestine was evaluated indirectly by estimating the generation of the specific bacterial fermentation metabolite p-cresol from urinary and dialytic removal (26–28).
SUBJECTS AND METHODS
Study design
Protein assimilation was evaluated in a cross-sectional observational design by means of breath-test technology and an estimation of daily p-cresol generation. Furthermore, several indexes of the MIA syndrome were measured. The study was approved by the ethical committee of the University Hospital Gasthuisberg, and informed consent was obtained from all subjects.
[13C]Protein breath test
Fifty subjects were included in this part of the study: 24 control subjects and 26 ESRD patients (14 treated with PD and 12 treated with HD). Demographic data of the study participants are shown in Table 1. None of the subjects had a history of gastrointestinal disease or surgery, hepatic disease, or diabetes mellitus; used drugs known to influence gastrointestinal motility (prokinetics or antiemetics) or protein assimilation (prebiotics, probiotics, antibiotics, or inhibitors of gastric acid secretion) during the preceding 3 mo; or had metabolic acidosis as defined by a serum bicarbonate concentration of 20 mEq/L. The control subjects comprised medical students (n = 12) and members of an association of retired persons (n = 12). They had no renal, metabolic, cardiovascular, gastrointestinal, or respiratory disease. The causes of renal disease in the patients included polycystic kidney disease (n = 6), glomerular disease (n = 13), tubulointerstitial disease (n = 4), and unknown etiologies (n = 3). PD patients were treated with standard lactate-buffered glucose solutions (Dianeal; Baxter, Lessines, Belgium) via continuous ambulatory PD (n = 6) or automated PD (n = 8) for 3 mo. In 4 PD patients, the polyglucose icodextrin (Extraneal; Baxter) was used during one long dwell per day. No amino acid-containing or bicarbonate-buffered solutions were applied. Residual renal function (RF) in the PD patients was 4.7 (0–9.1) mL · min–1 · 1.73 m–2. HD patients were treated 3 times, 4 h/wk, with maintenance high-flux HD (n = 7) or postdilution hemodiafiltration (n = 5) for 3 mo. Their mean residual RF was 0.2 (0–1.0) mL · min–1 · 1.73 m–2.
View this table:
TABLE 1. Subject characteristics1
Protein assimilation was evaluated after the subjects fasted overnight for 12 h by means of a [13C]protein breath test as described earlier (26, 28). The solid test meal (being a pancake) consisted of 7.5 g lyophilized [13C]leucine-labeled egg white, 17.2 g lyophilized [13C]leucine-labeled egg yolk, 3.75 g lyophilized unlabeled egg white, 3 g milk powder, 7 g sugar, 17 g flour, 130 mL water, and 5 g butter. The caloric content of the meal amounted to 327.4 kcal (18.67 g protein, 6.15 g fat, and 36.8 g carbohydrates). Subjects were asked to consume the meal together with 200 mL water within 15 min. Breath samples for the detection of 13CO2 were collected twice immediately before ingestion of the test meal and at 15-min intervals for a period of 6 h thereafter.
Breath was collected by blowing air through a straw in a small tube called an exetainer (Europa Scientific, Crewe, United Kingdom). The 13C content of the breath was determined by online gas chromatographic purification-isotope ratio mass spectrometry (Automated Breath 13C Analyzer-New Technology 20-20 Stable Isotope Analyzer; Europa Scientific). The values obtained by isotope ratio mass spectrometry were converted to percentage 13C recovery per hour (%dose/h) of the initial amount administered according to calculations described in detail by Ghoos et al (29). Cumulative percentages of label recovery (%cum) were calculated by means of the trapezoidal rule. From these data, the following indexes of protein assimilation were derived: the maximum percentage of the administered dose of 13C excreted per hour (%max) and the cumulative percentage of the administered dose of 13C recovered in the breath at the end of the test (%cumend).
p-Cresol generation rate
In a steady-state situation, total daily elimination of a molecule equals its daily generation. p-Cresol is a unique bacterial fermentation metabolite of tyrosine in the colon. Because most of the produced p-cresol is absorbed by the colonic mucosa, urinary output of the molecule reflects its colonic generation in nondialyzed subjects (26, 27, 30, 31). When dealing with ESRD patients, however, the elimination of p-cresol through the dialytic procedure also has to be taken into account. All control subjects (n = 24) and PD patients (n = 14) performed a 24-h urine collection. The collection bottles were refrigerated (0–4°C) at the participant's home to prevent bacterial growth and brought to the laboratory within hours of the time that the collection ended. During the same 24-h period, peritoneal drainage was collected by the PD patients. In 14 HD patients, total dialysate was collected in a 300-L polyethylene vessel during a midweek dialysis session. Although they were not the same patients as those who underwent the breath-test study, they met the same eligibility criteria, and their demographic variables were nearly identical (first set of values is for the 14 HD patients): age, 72 y (range: 42–82 y) compared with 64 y (49–80 y); sex distribution, 7 males compared with 9 males; body weight, 62.1 ± 2.0 compared with 65.5 ± 4.1 kg; body mass index (BMI; in kg/m2), 22.3 (17.6–26.8) compared with 22.0 (17.7–35.0); and residual RF, 0.1 (0–6.4) compared with 0.2 (0–1.0) mL · min–1 · 1.73 m–2 (two-sided P value > 0.3 in all comparisons). They were all treated trice weekly with maintenance high-flux HD. If not anuric (n = 7), the patients also performed an interdialytic (44 h) urine collection. Collected volumes of urine and dialysate were vigorously stirred, weighed, and sampled. All samples were stored at –80°C until analyzed. The total weekly elimination of the solute was calculated on the basis of measured p-cresol concentrations and volumes. The results were then expressed as amounts eliminated per 24 h. As a reflection of dietary protein intake, 24-h elimination of urea nitrogen was calculated as well by using the same approach (32).
Indexes of the MIA syndrome
The anthropometric indicators of nutritional status measured were body weight (kg) and BMI, calculated from height (kg) and body weight (m2). Dietary protein (g · kg–1 · d–1) and energy intake (kcal · kg–1 · d–1) were evaluated by computer calculations based on a 7-d dietary record. Serum C-reactive protein (CRP) concentrations (mg/L) were measured as an index of inflammation. On the basis of a review of the medical records, comorbidity was assessed by using the method described by Davies et al (33). In brief, 7 categories of comorbid conditions were scored 1 (present) or 0 (absent). The total sum of scores was used to determine the grade of comorbidity. Grade 0 (low risk) is a score of 0, grade 1 (medium risk) is a score of 1–2, and grade 2 (high risk) is a cumulative score of 3. Subjects scoring 1 in the categories "ischemic heart disease" and "peripheral vascular disease" were considered to have atherosclerotic disease. Serum albumin (g/L) and the Malnutrition-Inflammation Score (MIS) were determined as indicators of both nutritional status and inflammation. The latter comprehensive scoring system, with a possible range of from 0 to 30, combines components of the Subjective Global Assessment (34) and 3 additional elements (serum albumin, total-iron-binding capacity, and BMI) (23).
Biochemical analyses
Serum albumin, total-iron-binding capacity, CRP, serum bicarbonate, serum creatinine, and serum urea nitrogen concentrations were measured by using standard laboratory techniques. p-Cresol was analyzed by gas chromatography-mass spectrometry (GC-MS) technology. Five hundred microliters of serum was diluted with 450 µL water. The pH of a 950-µL sample (diluted serum, urine, or dialysate) was adjusted to pH 1 with concentrated H2SO4, and the solution was heated to 90°C for 30 min. After a cooling-down period to ambient temperature, 50 µL 2,6-dimethylphenol solution (20 mg/100 mL) was added as internal standard. One milliliter ethyl acetate was added for the extraction of p-cresol. The solution was well mixed for 30 s and centrifuged at 1583 x g (3300 rpm) for 20 min. Then, 500 µL of the supernatant fraction was dried over anhydrous sodium sulfate, and 100 µL of the resultant sample was transferred to the GC-MS apparatus (Trace GC-MS; Thermofinnigan, San José, CA) for automatic splitless injection of 0.5 µL. The analytic column used was 30 m (length) x 0.32 mm (internal diameter) and had a film thickness of 1 µm (AT5-MS; Alltech, Deerfield, IL). Helium GC grade was used as a carrier gas with a constant flow of 1.3 mL/min. The oven was programmed from 75°C (isotherm for 5 min) to 280°C at 15°C/min. After separation, p-cresol was identified by MS (electron impact full scan mode with mass-to-charge ratios of from 59 to 590 at 2 scan/s). Quantitative results were obtained by the internal standard method and calculated as concentrations. The R2 of the calibration line was 0.998. The method has low intra- and interassay variabilities (CVs of 3.33% and 5.30%, respectively). It is able to detect total p-cresol concentrations ranging from 0.15 to 60 mg/L; the extraction efficiency is 91.5%.
(Residual) RF was estimated by calculating the arithmetic mean of renal urea nitrogen and creatinine clearance. Values were normalized for body surface area (BSA) by multiplying by 1.73/BSA(m2) and were expressed as mL · min–1 · 1.73 m–2. BSA was calculated from the formula of Haycock et al (35):
RESULTS
[13C]Protein breath test
The mean 13C excretion curves in control subjects and ESRD patients, expressed as %dose/h and %cum obtained after ingestion of the [13C]protein test meal, are depicted in Figure 1. The curves for the ESRD patients are flattened compared with those observed for the control subjects. Breath test-derived indexes of protein assimilation (%max and %cumend) were significantly lower in the ESRD patients than in the control subjects. (Table 2)
FIGURE 1.. Mean 13C enrichment of breath carbon dioxide expressed as a percentage of the administered dose of [13C]leucine per hour and as a cumulative percentage (%cum) in control subjects (, n = 24) and patients with end-stage renal disease (, n = 26) after ingestion of the [13C]protein meal. Bars represent 95% CIs.
View this table:
TABLE 2. [13C]Protein breath test1
p-Cresol generation rate
As illustrated in Table 3, the difference in 24-h elimination of p-cresol between the ESRD patients and the control subjects was not statistically significant. When the data were expressed per gram of urea nitrogen eliminated, as a correction for dietary protein intake, the difference was significant.
View this table:
TABLE 3. p-Cresol generation rate1
Indexes of the MIA syndrome
Data on nutritional status, inflammation, and atherosclerosis are shown in Table 4. Anthropometric measures did not differ significantly between the 2 groups, although there was a trend for a lower BMI in the ESRD patients. No significant difference in protein intake was seen between the groups. Daily energy intake, however, was significantly lower in the patient group than in the control subjects. Significantly higher serum concentrations of CRP were seen in the ESRD patients than in the control subjects. Comorbidity, evaluated by Davies score and comorbidity grade, was significantly higher in the ESRD patients as was the frequency of atherosclerotic disease. Serum albumin concentrations were significantly lower and MIS was significantly higher in the ESRD patients than in the control subjects.
View this table:
TABLE 4. Indexes of the malnutrition-inflammation-atherosclerosis syndrome1
Correlations
As shown in Table 5, both %max and %cumend were significantly related with serum creatinine, RF, Davies score, Davies grade, and the presence of atherosclerotic disease. In addition, %max was significantly correlated with serum urea nitrogen, serum p-cresol, CRP, and MIS.
View this table:
TABLE 5. Spearman correlation coefficients for indexes of protein assimilation1
Data on both the p-cresol generation rate and the other variables were available in only 38 subjects (ie, 24 control subjects and 14 ESRD patients). Within this group, the 24-h elimination of p-cresol as well as the elimination of p-cresol/urea nitrogen correlated significantly with serum urea nitrogen, serum creatinine, serum p-cresol, RF, CRP, and serum albumin. (Table 5)
Because there was a significant age difference between the groups (Table 1), we evaluated the relations between age and all the other measured variables. Significant correlations with age were found only for serum urea nitrogen, serum creatinine, RF, BMI, CRP, Davies score, Davies grade, serum albumin, and MIS (Table 6) No correlations were found between age and [13C]protein breath test-derived indexes or between age and the p-cresol generation rate. (Spearman correlation coefficients shown in last row of Table 5, all P values > 0.150.)
View this table:
TABLE 6. Spearman correlation coefficients for indexes of the malnutrition-inflammation-atherosclerosis syndrome1
As illustrated in Table 6, significant relations were found between RF indexes (serum urea nitrogen, serum creatinine, serum p-cresol, and RF) and indexes of the MIA syndrome. In view of the correlations found between indexes of protein assimilation and indexes of the MIA syndrome (Table 5), a comparison was made between ESRD patients with a MIS below and above the median value of 6. As denoted in Table 7, the latter were older and had a higher Davies score and comorbidity grade than did those with a MIS <6. Their serum concentrations of urea nitrogen, creatinine, and CRP were higher as well, whereas RF was lower. Moreover, the ESRD patients with a MIS >6 had lower breath test-derived indexes of protein assimilation than did their counterparts with a MIS <6 (%max = 3.24 ± 0.37 compared with 4.23 ± 0.43, one-sided P = 0.045; %cumend = 12.11 (5.74–19.84) compared with 12.88 (8.91–23.22), one-sided P = 0.085). The high-MIS group consisted predominantly of patients treated with HD (n = 10, P = 0.0016).
View this table:
TABLE 7. Comparison of end-stage renal disease patients with a high or low Malnutrition-Inflammation Score (MIS)1
DISCUSSION
The findings of this cross-sectional observational study indicate that small intestinal protein assimilation (digestion and absorption) is impaired in ESRD patients. Moreover, indexes of protein assimilation were related to indexes of the MIA syndrome. Small intestinal protein assimilation was studied by 2 different approaches. First, the validated [13C]protein breath test was used (26, 28); background bias due to the dialysis procedure was taken into account where appropriate. The mean 13CO2 excretion curve, expressed as %dose/h, was flattened in the ESRD patients as compared with that in the control subjects and mathematically derived indexes of protein assimilation (%max and %cumend) were significantly lower in the former group (Figure 1 and Table 2). Second, daily colonic generation of the bacterial fermentation metabolite p-cresol, corrected for dietary protein intake, was higher in the ESRD patients than in the control subjects (Table 3). This is a reflection of the higher amount of protein escaping small intestinal digestion and absorption in the former (27). We measured the total daily elimination of p-cresol to estimate its generation rate. An essential assumption for this approach is the steady state situation. We believe that this assumption is applicable to our study population, because subjects in whom recent changes in the colonic flora could be expected were excluded from the study. Moreover, the participants were all consuming their usual diets, which minimized the possible influence of dietary changes on p-cresol generation.
In a recent study, using the same methodology, we found evidence for impaired protein assimilation in nondialyzed uremic patients (26). The present data extend these findings to the dialysis population. Several gastrointestinal abnormalities have been reported in dialyzed as well as in nondialyzed uremic patients, including gastrointestinal motility disorders (38, 39), small-bowel bacterial overgrowth (40, 41), gastric hypochlorhydria (30), and pancreatic abnormalities (42, 43). Each of them could theoretically disturb protein digestion or absorption. The clinical relevance of these abnormalities and their pathogenetic role in the impairment of protein assimilation, however, remain to be elucidated.
An association between symptoms of malnutrition, inflammation, and atherosclerosis in uremic patients was described recently (15). The existence of this so-called MIA syndrome was confirmed in several publications (5, 19, 20, 23, 44). Because digestion and subsequent absorption of dietary protein play a pivotal role in the maintenance of a neutral nitrogen balance, disturbances in these processes may contribute to the impaired nutritional status seen in uremic patients. Hence, as a secondary endpoint of our study, we investigated the relations between indexes of protein assimilation and indexes of the abovementioned MIA syndrome. In a correlation analysis, breath test-derived indexes of protein assimilation were associated with comorbidity, atherosclerosis, and inflammation. p-Cresol generation rate correlated significantly with CRP and serum albumin (Table 5). Furthermore, although not uniformly significant, lower breath test-derived indexes of protein assimilation were observed in the ESRD patients with a higher MIS, most of whom were treated with HD, than in those with a lower MIS. The latter finding suggests that the above correlations are not simply a statistical consequence of covarying differences between control subjects and ESRD patients in protein assimilation (Tables 2 and 3) and indexes of the MIA syndrome (Table 4). Although our observations do not prove causality, they certainly suggest that impairment of protein assimilation might be one of the factors playing a role in the MIA syndrome (15). Although the sample size of our study does not allow for an adequate multivariate analysis to address this issue, (residual) RF seems to play a key role in the interplay between the MIA indexes and protein assimilation. Both indexes of protein assimilation and indexes of the MIA syndrome were strongly related to (residual) RF. Also, when the analysis was confined to the ESRD patients, a significantly lower residual RF was noted in the patients with a higher MIS and lower breath test-derived indexes of protein assimilation (Table 7). This may not be a surprise considering the evidence of lower residual RF as a cardiovascular risk factor and the relation between RF, inflammatory indexes, and nutritional indexes described by others (45–48).
Some methodologic issues have to be addressed. First, to minimize metabolic interference with the breath tests performed in this study, we excluded patients with diabetes mellitus, hepatic failure, or metabolic acidosis (7–9, 49, 50). Hence, on the basis of our findings, no conclusions can be drawn with regard to uremic patients belonging to these subpopulations. Second, although daily generation of p-cresol reflects the amount of protein escaping digestion and absorption in the small intestine (26, 27), this is not the only determining factor. As stated earlier, daily p-cresol generation is known to be influenced by dietary protein intake. Both the quantity and quality of the protein may be important in this regard. We normalized our data for the quantity of dietary protein. Because subjects were consuming a free diet, however, we were not able to evaluate the possible effect of qualitative differences. Other factors involved are the amount of (fermentable) dietary fiber, characteristics of the bacterial flora, and colonic transit time. Nevertheless, taking into account the breath-test data, which are not influenced by these factors, we consider the p-cresol generation rate as a valid, although rather crude, estimate of the amount of protein escaping small intestinal assimilation.
Finally, because of the inclusion and exclusion criteria and the fact that breath-test performance requires full cooperation of the tested subjects, most participants in the present study were in good clinical condition. This was reflected by the relatively mild indexes of the MIA syndrome. As compared with the mean values in our center's total dialysis population, the ESRD patients in the present study were significantly younger, had significantly lower Davies scores and grades, had a lower prevalence of atherosclerotic disease, had higher serum albumin concentrations, had lower CRP concentrations, and had higher daily protein intakes (data not shown; two-sided P value < 0.05 in all comparisons). From our correlation data, it can be speculated that the findings of impaired protein assimilation would have been more pronounced in sicker patients.
In conclusion, our study provides evidence that protein assimilation is impaired in ESRD patients. Although the exact pathogenesis of the disorder has to be clarified by further investigation, its presence seems to be associated with the severity of the MIA syndrome.
ACKNOWLEDGMENTS
We acknowledge M Dekens, C Dewit, N Gorris, S Rutten, R Servaes, and L Swinnen for providing excellent technical assistance and D Kuypers and B Maes for providing scientific comments and reviewing the manuscript.
BB and PE developed the idea for the study and were involved in the planning and design of the study. BB, PE, KV, and YV were involved in the interpretation of the results and writing of the manuscript. BB collected the data and was responsible for the data analysis, interpretation of the data, writing of the first draft, and integration of the comments from the coauthors. None of the authors had a conflict of interest.
REFERENCES