HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: CJN.00060505v1 1/1/52    most recent Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gennari, F. J. Articles by Levey, A. S. Search for Related Content PubMed PubMed Citation Articles by Gennari, F. J. Articles by Levey, A. S.

Effect of Dietary Protein Intake on Serum Total CO₂ Concentration in Chronic Kidney Disease: Modification of Diet in Renal Disease Study Findings

F. John Gennari^*, Virginia L. Hood^*, Tom Greene, Xulei Wang, and Andrew S. Levey

^* University of Vermont College of Medicine, Burlington, Vermont; Cleveland Clinic Foundation, Cleveland, Ohio; and New England Medical Center, Boston, Massachusetts

Address correspondence to: Dr. F. John Gennari, Rehab 2319 UHC Campus, Fletcher Allen Health Care, Burlington, VT 05401. Phone: 802-847-2534; Fax: 802-847-8736; E-mail: f-john.gennari{at}uvm.edu

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Metabolic acidosis is a feature of chronic kidney disease (CKD), but whether serum bicarbonate concentration is influenced by variations in dietary protein intake is unknown. For assessing the effect of diet, data that were collected in the Modification of Diet in Renal Disease study were used. In this study, patients with CKD were enrolled into a baseline period, then randomly assigned to follow either a low- or a usual-protein diet (study A, entry GFR 25 to 55 ml/min) or a low- or very low–protein diet, the latter supplemented with ketoanalogs of amino acids (study B, entry GFR 13 to 24 ml/min). Serum [total CO₂] and estimated protein intake (EPI) were assessed at entry (n = 1676) and again at 1 yr after randomization, controlling for changes in GFR and other key covariates (n = 723). At entry, serum [total CO₂] was inversely related to EPI (1.0 mEq/L lower mean serum [total CO₂]/g per kg body wt increase in protein intake/d; P = 0.009). In an intention-to-treat analysis, the reduction in mean EPI in the low-protein group as compared with the usual-protein group (0.41 g/kg body wt per d) was independently associated with a 0.9-mEq/L increase in serum [total CO₂], after adjustment for covariates (P < 0.001). No such effect was evident in study B, in which the very low–protein diet group received dietary supplements. Serum [total CO₂] is inversely correlated with dietary protein intake in patients with CKD. A reduction in protein intake results in an increase in serum [total CO₂].

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Serum total CO₂ concentration ([total CO₂]) falls as GFR decreases in chronic kidney disease (CKD) (1,2), but the values vary widely among patients with similar levels of kidney function. Part of this difference may be due to diet-induced differences in endogenous acid production. In individuals with normal kidney function, the effect of diet on serum [total CO₂] is small and not detectable unless the influence of Paco₂ is removed (3). Given the impairment in acid excretion that occurs in CKD (4–6), one might anticipate that differences in diet-induced acid production would have a greater influence on serum [total CO₂] than in people with normal kidney function. The major source of endogenous acid production comes from metabolism of dietary protein (7); therefore, protein restriction should result in an increase in serum [total CO₂] if this hypothesis is correct. Despite extensive studies on the effects of protein restriction on metabolic parameters and kidney disease progression (8), little attention has been directed to the effect of this intervention on serum [total CO₂] in CKD. In one study in humans with severe CKD, serum [HCO₃^–] was notably higher when the patients were compliant with a very low–protein diet (0.3 g/kg body weight) supplemented with ketoanalogs of amino acids (9). In two other studies in humans, serum [total CO₂] did not increase significantly after a 50% reduction in protein, but both studies contained very few subjects (10,11). The best data supporting an effect of protein restriction on body alkali stores in CKD come from a partial nephrectomy rat model (12). In this model, steady-state serum [HCO₃^–] was significantly higher (by 2 mEq/L) in animals that ingested a low-protein diet (6% of total calories/d) as compared with animals that ingested a normal-protein diet (24% of total calories/d).

To determine whether systematic changes in dietary protein intake affect serum [total CO₂] in a large cohort of patients with CKD, we evaluated the data collected in patients who participated in the Modification of Diet in Renal Disease (MDRD) study, focusing on GFR, protein intake, and serum [total CO₂] (13). Our analysis demonstrates that [total CO₂] is inversely related to dietary protein intake in this group of patients and that a deliberate reduction in dietary protein intake increases serum [total CO₂].

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
The details of the entry criteria, design, and results of the MDRD trial have been published previously (13–15). In brief, 1782 men and women were screened for this study. A total of 840 men and women who were aged 18 to 70 yr and had CKD were subsequently randomly assigned to diets that differed in protein content. On the basis of GFR, measured by iothalamate clearance, the participants were enrolled in either study A or study B. Patients in study A (n = 585) had entry GFR of 25 to 55 ml/min per 1.73 m²; patients in study B (n = 255) had entry GFR of 13 to 24 ml/min per 1.73 m². After a baseline period of measurements, the patients in study A were randomly assigned to either a usual-protein diet or a low-protein diet. The usual-protein diet contained 1.3 g/kg body wt per d protein and 16 to 20 mg/kg body wt per d phosphorus. The low-protein diet was designed to provide 0.575 g/kg body wt per d protein (with 65% of protein from high biologic value sources) and 5 to 10 mg/kg body wt per d phosphorus. In study B, the patients were randomly assigned either to the low-protein diet described above or to a very low–protein diet (0.28 g/kg body wt per d) supplemented with a mixed salt preparation made up of basic amino acids (tyrosine and threonine) and ketoacid analogs of other essential amino acids (totaling 0.28 g/kg body wt per d) (16). In this study, total protein and protein precursor intake from food and supplements was identical in the low- and very low–protein diets. Both the low- and very low–protein diets contained 4 to 9 mg/kg body wt per d phosphorus in study B. In both studies, serum [total CO₂] was recorded every 4 mo from local laboratory measurements, and 24-h urine urea nitrogen excretion was measured centrally monthly throughout follow-up (average duration 2.2 yr). Dietary protein intake was estimated from urine urea nitrogen measurements (15). This report focuses on two analyses of these data. The first is a cross-sectional examination of the interrelationships among serum [total CO₂], GFR, and estimated protein intake (EPI) at the initial baseline visit in all of the screened patients for whom these three measurements were available (n = 1676). The second is a longitudinal examination of the effect of dietary protein restriction on serum [total CO₂], using an intention-to-treat analysis. We chose a 1-yr interval for this latter examination to minimize bias as a result of attrition from death, development of ESRD, or other loss to follow-up and to minimize the confounding effects of changes in GFR. For this longitudinal assessment, serum [total CO₂] and EPI were evaluated at the initial baseline and at 1 yr. This analysis was carried out in all of the randomly assigned patients in whom [total CO₂], GFR, and EPI were assessed both at the initial baseline visit and at 1 yr after randomization (n = 723).

Statistical Analyses
The cross-sectional association of [total CO₂] with other factors was evaluated at the initial baseline visit in 1676 patients with values at this visit for [total CO₂], GFR, and EPI. We first related [total CO₂] to GFR and EPI in a multiple regression model that included indicator variables for the 15 clinical centers as covariates. A segmented regression model with separate slopes for the relationship of [total CO₂] with GFR above and below 25 ml/min per 1.73 m² was used to account for a steeper relationship between these factors at lower GFR levels. This regression analysis was repeated after adjustment also for demographic factors and indicators for antihypertensive medications at baseline.

The effects of the diet interventions on the change in [total CO₂] from baseline to 1 yr were tested in 517 study A patients and in 206 study B patients who remained in the trial and who provided [total CO₂] measurements at baseline and at 1 yr. The diet effects were estimated using analysis of covariance, with the model adjusting for baseline [total CO₂]; baseline GFR; and indicator variables for randomized BP group, clinical center, and the randomization stratification factors (inverse serum creatinine slope before the trial and baseline mean arterial BP). In this analysis, patients were analyzed according to their randomized group assignment, irrespective of achieved protein intake during follow-up. The sensitivity of the results to potential confounding by changes in GFR after randomization was evaluated by adding the 12-mo GFR value as an additional covariate in the analysis. Because [total CO₂] was measured in the local laboratories associated with the 15 clinical centers in the MDRD study, we expect there to be systematic center difference in the assay. This was accounted for in the statistical analyses by including clinical center as a covariate and by assessing the effects of the diet intervention on the basis of the change in [total CO₂] over 1 yr, thus using each patient as his or her own control.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Fifteen centers participated in the study, each contributing between 52 and 64 participants who met the baseline characteristics and were randomly assigned. The spectrum of diagnoses included polycystic kidney disease (22%), glomerular diseases (27%), hypertensive nephropathy (17%), tubulointerstitial diseases (7%), and other or unknown (14%). Only 3.5% had diabetic nephropathy, because patients who had diabetes and required insulin therapy were excluded from the MDRD study. The remaining 9.5% included hereditary renal diseases, urinary tract disease, and diseases associated with a single kidney. The results of the cross-sectional analyses at baseline are shown in Tables 1 through 3. Mean baseline serum [total CO₂] varied significantly among the centers, ranging from 20.2 to 25.7 mEq/L (P < 0.0001). The level of [total CO₂] was also significantly related to GFR (Table 1), but the differences among mean [total CO₂] levels persisted when GFR was taken into account. After controlling for clinical center, [total CO₂] correlated directly with GFR and inversely with estimated protein intake (Table 2). The relationship of [total CO₂] to GFR was significantly stronger at lower GFR than at higher GFR values (P < 0.001); thus, two regression coefficients are shown in Table 2. Regression analysis of baseline [total CO₂] against GFR and estimated protein intake, now corrected for all other significant variables (center, age, gender, and drugs) showed the same significant correlations (Table 3). Of note, this analysis shows a significant inverse correlation between baseline serum [total CO₂] and the use of angiotensin-converting enzyme (ACE) inhibitors and a direct correlation with the use of diuretics. Serum [total CO₂] also showed an inverse correlation with age and direct correlations with male gender and other antihypertensive medication use. The latter grouping was too diverse to analyze further. After controlling for clinical center, GFR, protein intake, age, race, antihypertensive medications, and proteinuria, the kidney disease diagnosis was not significantly associated with serum [total CO₂] (P = 0.24).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
This analysis of the participants in the MDRD study, both at the time of entry into the study and at 1 yr after deliberate changes in diet, indicates that dietary protein intake has a significant influence on serum [total CO₂] in patients with CKD. Before randomization, serum [total CO₂] varied inversely with EPI, an effect that persisted when other influences were accounted for. When protein intake was reduced deliberately in the low-protein diet group in study A, moreover, serum [total CO₂] increased notably in comparison with the usual-protein diet group, indicating a significant diet effect at 1 yr (Table 5). On the basis of this latter analysis, a reduction in dietary protein intake of 0.2 g/kg body wt per d in patients with CKD would be predicted to increase serum [total CO₂] by almost 0.5 mEq/L.

In contrast to study A, no significant effect of diet on serum [total CO₂] was seen in study B, but this study was confounded by the diet supplement that was given to the patients in the very low–protein diet group. This supplement contained mixed salts of basic amino acids and ketoanalogs of other amino acids (15,16). Reviewing the components of these salts, one would expect them, if anything, to generate new HCO₃^– when metabolized, likely accounting for the large increase in serum [total CO₂] between baseline and 1 yr in the very low–protein group (Table 5). Because serum [total CO₂] also trended upward in the low-protein group in study B and possibly because of the small numbers of patients, no "diet effect" was seen when the low- and very low–protein diet groups were compared. Regardless of the outcome, however, one cannot draw any conclusions about the effect of diet in study B because of the potential alkali contained in the supplements that were given to the very low–protein diet group.

This study reconfirms the strong relationship between GFR and serum [total CO₂] (1,2,17), and this influence is clearly a confounding factor in assessing the effect of diet in any group of patients with widely varying and changing GFR. Fortuitously, the fall in GFR was essentially the same over 1 yr in the two groups in both study A and study B (Table 4). The power of the present study is that the randomization procedure limits the degree of confounding for this key variable. Furthermore, our sensitivity analyses using multivariable regression showed that the estimated diet group effects are essentially unchanged after adjustment for the small differences in change in GFR that were observed.

Other significant cross-sectional relationships involving serum [total CO₂] are outlined in Table 3. Of interest, the use of ACE inhibitors correlated inversely with serum [total CO₂], and the use of diuretics correlated directly. Although it is difficult to draw any firm conclusions from these correlations, the ACE inhibitor effect could be due to a reduction in aldosterone secretion, thereby impairing acid excretion. The diuretic effect, by contrast, suggests that even damaged kidneys can respond to an increase in distal Na⁺ delivery with an increase in H⁺ secretion. Further studies are needed to explore these effects in greater detail.

Three earlier studies have provided data to suggest that reducing protein intake increases serum [HCO₃^–] in patients with CKD (9–11), but all had fewer subjects and none assessed critically other variables that could have influenced serum [total CO₂]. In the largest of the three studies, capillary blood pH and [HCO₃^–] were measured in a group of patients who had severe CKD (mean creatinine clearance = 10 ml/min) and were put on a very low–protein diet (dietary protein intake 0.3 g/kg body wt per d), supplemented by a different mixture of essential amino acids and ketoanalogs than in the present study (9). In contrast to the present study, all patients were on the same diet and supplements. Patient compliance was assessed by measuring urine urea nitrogen excretion, and acid-base status measurements were obtained at varying intervals after the diet was started (from 5 to 58 mo). Blood [HCO₃^–] and pH both were significantly higher (22 versus 19 mEq/L and 7.39 versus 7.33; P < 0.01 in both cases) in a subset of patients who were compliant with the diet. Although mean GFR was not significantly different at the time of assessment, the rate of decline was greater in the noncompliant patients; thus, a confounding effect of GFR is likely. The two other studies involved very small numbers of patients (six in one and 12 in the other), and both showed a nonsignificant increase in serum [HCO₃^–] (1 to 2 mEq/L) 2 wk to 3 mo after dietary protein intake was reduced by 50% (10,11).

Although not measured, we presume that the diet effect observed in the MDRD study is caused by a reduction in endogenous acid production, engendered by reduction of dietary protein intake. In individuals who are eating a Western diet, a major component of endogenous acid production is due to the catabolism of sulfur-containing amino acids contained in animal proteins, and endogenous acid production correlates directly with protein intake (3). In study A, dietary phosphate intake was also reduced in the low-protein diet group. Because of the study design, we could not separate potential effects of phosphate reduction from those of protein reduction, but dietary phosphate is normally a net contributor to endogenous acid production (3). Thus, the reduction in phosphate intake, if it has any effect, would contribute to the diet-induced reduction in endogenous acid production. In individuals with normal GFR or with age-related reductions in GFR, endogenous acid production shows an inverse correlation with serum [HCO₃^–], but only after the influences of Pco₂ and GFR are removed (3,17). This study suggests that when CKD is present, diet-induced reductions in endogenous acid production have an effect on serum [total CO₂] that is demonstrable without controlling for variations in Pco₂.

When the total variability in serum [total CO₂] in the present study was evaluated at baseline, only 35% was accounted for by all of the variables tested (Table 3), indicating that other, unmeasured factors contribute to this variability. The MDRD study was not designed to study the hypothesis that we proposed, and it is not surprising, therefore, that many undefined influences are present. One can only speculate about these, but one is almost certainly Pco₂. Variations in Pco₂ are well recognized to have a potent effect on steady-state serum [total CO₂] (3,17,18), most likely through a direct effect of CO₂ on HCO₃^– reabsorption and acid excretion by the kidney. Nonetheless, a systematic bias in Pco₂ is unlikely to be present in this study. All contributing centers were at or near sea level, so the distribution of variations in Pco₂ was likely random in the participants. Another likely factor is variations in renal H⁺ secretion in patients who had the same level of GFR. It is widely recognized, for example, that a fraction of patients with CKD waste HCO₃^– (5,6), and in these individuals, serum [total CO₂] would not be expected to increase with a reduction in endogenous acid production. A third possibility is variation as a result of differences in kidney disease diagnoses. The effect of a given diagnosis on serum [total CO₂] has not been examined critically, but one large retrospective analysis that included a broad spectrum of diagnoses did not note any specific effect of kidney disease diagnosis on serum [total CO₂] (2). Our cross-sectional analysis did not suggest that kidney disease diagnosis influenced acid-base status.

A limitation of our study is that we cannot exclude with certainty other, as-yet-undefined factors that may introduce a bias in our findings. We have adjusted for the key factors that were measured, however, particularly the influence of changes in GFR. The paired observations add strength to our conclusion by minimizing population variation, and the diet effect is robust in study A, which is the most straightforward study of the effect of protein restriction on serum [total CO₂] in patients with CKD.

In summary, analysis of the MDRD data indicates that dietary protein intake is inversely associated with serum [total CO₂]. In addition, a deliberate reduction in protein intake results in a significant increase in serum [total CO₂]. Given the now well-documented effects of a low serum [total CO₂] to promote muscle catabolism and bone calcium loss in patients with CKD (19,20), further exploration of dietary approaches to ameliorating the reduction in [total CO₂] that accompanies CKD is warranted.

	Footnotes

 
Published online ahead of print. Publication date available at www.cjasn.org.

Received May 3, 2005. Accepted July 28, 2005.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Widmer B, Gerhardt RE, Harrington JT, Cohen JJ: Serum electrolyte and acid base composition. The influence of graded degrees of chronic renal failure. Arch Intern Med 139: 1099–1102, 1979[Abstract]
2. Hakim RM, Lazarus JM: Biochemical parameters in chronic renal failure. Am J Kidney Dis 11: 238–247, 1988
3. Kurtz I, Maher T, Hulter HN, Schambelan M, Sebastian A: Effect of diet on plasma acid-base composition in normal humans. Kidney Int 24: 670–680, 1983[Medline]
4. Tizianello A, De Ferrari G, Garibotto G, Gurreri G, Robaudo C: Renal metabolism of amino acids and ammonia in subjects with normal renal function and in patients with chronic renal insufficiency. J Clin Invest 65: 1162–1173, 1980[Medline]
5. Schwartz WB, Hall PW 3rd, Hays RM, Relman AS: On the mechanism of acidosis in chronic renal disease. J Clin Invest 38: 39–52, 1959[Medline]
6. Goodman AD, Lemann J Jr, Lennon EJ, Relman AS: Production, excretion, and net balance of fixed acid in patients with renal acidosis. J Clin Invest 44: 495–506, 1965[Medline]
7. Harrington JT, Lemann J Jr: The metabolic production and disposal of acid and alkali. Med Clin North Am 54: 1543–1554, 1970[Medline]
8. Pedrini MT, Levey AS, Lau J, Chalmers TC, Wang PH: The effect of dietary protein restriction on the progression of diabetic and nondiabetic renal diseases: A meta-analysis. Ann Intern Med 124: 627–632, 1996[Abstract/Free Full Text]
9. Barsotti G, Cupisti A, Ciardella F, Morelli E, Niosi F, Giovannetti S: Compliance with protein restriction: Effects on metabolic acidosis and progression of renal failure in chronic uremics on supplemented diet. Contrib Nephrol 81: 42–49, 1990[Medline]
10. Williams B, Hattersley J, Layward E, Walls J: Metabolic acidosis and skeletal muscle adaptation to low protein diets in chronic uremia. Kidney Int 40: 779–786, 1991[Medline]
11. Bernhard J, Beaufrere B, Laville M, Fouque D: Adaptive response to a low-protein diet in predialysis chronic renal failure patients. J Am Soc Nephrol 12: 1249–1254, 2001[Abstract/Free Full Text]
12. Maddox DA, Horn JF, Famiano FC, Gennari FJ: Load dependence of proximal tubular fluid and bicarbonate reabsorption in the remnant kidney of the Munich-Wistar rat. J Clin Invest 77: 1639–1649, 1986[Medline]
13. Klahr S, Levey AS, Beck GJ, Caggiula AW, Hunsicker L, Kusek JW, Striker G: The effects of dietary protein restriction and blood-pressure control on the progression of chronic renal disease. Modification of Diet in Renal Disease Study Group. N Engl J Med 330: 877–884, 1994[Abstract/Free Full Text]
14. Greene T, Bourgoignie JJ, Habwe V, Kusek JW, Snetselaar LG, Soucie JM, Yamamoto ME: Baseline characteristics in the Modification of Diet in Renal Disease Study. J Am Soc Nephrol 4: 1221–1236, 1993[Abstract]
15. The Modification of Diet in Renal Disease Study: Design, methods, and results from the feasibility study. Am J Kidney Dis 20: 18–33, 1992
16. Mitch WE, Abras E, Walser M: Long-term effects of a new ketoacid-amino acid supplement in patients with chronic renal failure. Kidney Int 22: 48–53, 1982[CrossRef][Medline]
17. Frassetto LA, Morris RC Jr, Sebastian A: Effect of age on blood acid-base composition in adult humans: Role of age-related renal functional decline. Am J Physiol 271: F1114–F1122, 1996[Medline]
18. Madias NE, Adrogue HJ, Horowitz GL, Cohen JJ, Schwartz WB: A redefinition of normal acid-base equilibrium in man: Carbon dioxide tension as a key determinant of normal plasma bicarbonate concentration. Kidney Int 16: 612–618, 1979[Medline]
19. Garibotto G, Russo R, Sofia A, Sala MR, Robaudo C, Moscatelli P, Deferrari G, Tizianello A: Skeletal muscle protein synthesis and degradation in patients with chronic renal failure. Kidney Int 45: 1432–1439, 1994[Medline]
20. Bushinsky DA: The contribution of acidosis to renal osteodystrophy. Kidney Int 47: 1816–1832, 1995[Medline]

This article has been cited by other articles:

				O. Moranne, M. Froissart, J. Rossert, C. Gauci, J.-J. Boffa, J. P. Haymann, M. B. M'rad, C. Jacquot, P. Houillier, B. Stengel, et al. Timing of Onset of CKD-Related Metabolic Complications J. Am. Soc. Nephrol., January 1, 2009; 20(1): 164 - 171. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: CJN.00060505v1 1/1/52    most recent Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gennari, F. J. Articles by Levey, A. S. Search for Related Content PubMed PubMed Citation Articles by Gennari, F. J. Articles by Levey, A. S.