Adaptive Response to a Low-Protein Diet in Predialysis Chronic Renal Failure Patients

Many studies have addressed the safety of a moderate restriction in protein intake on the nutritional status during chronic renal failure (CRF). Whereas it is well established that in healthy individuals, low-protein diets result in a reduction of whole-body protein turnover and amino acid oxidation (1,2), there have been some discrepancies about this adaptive response in CRF (3). In addition, the metabolic effects of the ketoanalogs of amino acids have been evaluated only in CRF patients during more restricted protein diets (4,5). To clarify these adaptations during a short period, we designed a randomized, controlled study and administered a moderately lowprotein diet supplemented or not with ketoanalogs of amino acids during 3 mo in 12 patients who presented with mild CRF.

Materials and Methods

Results

As shown in Table 1, patients were not malnourished. The degree of renal failure was mild, corresponding to a GFR of approximately 30 ml/min per 1.73 m². Renal failure was not progressing rapidly, as evidenced by the fact that patients' serum creatinine did not increase by more than 50 μmol/L within the last 3 mo. None of them experienced nephrotic syndrome or massive proteinuria (Table 1). Patients did not present metabolic acidosis or advanced renal osteodystrophy.

Compliance to the diet was obtained after training with our specialized renal dietitian. Patients were asked not to reduce their energy intake; this was accomplished by increasing slightly the marmalade in breakfast and by increasing to satisfactory levels the daily oil intake mainly in salad dressing. As shown in Table 2 and based on dietary interviews, the energy intake of patients, although not excessive, was kept within acceptable values, e.g., >30 kcal/kg per d. Protein intake was assessed by two different methods. We determined protein nitrogen appearance (PNA) from urinary urea output (Table 2), according to Maroni's formula (8). We also monitored protein intake from the 3-d home dietary records obtained monthly (Table 2). The protein equivalent of the nitrogen content in the ketoanalogs was not included in the estimated dietary protein intake (DPI) and averaged 0.08 ± 0.004 g of protein/kg per d, partly explaining a greater PNA than the DPI (Table 2). Overall, patients gradually reduced their daily protein intake by approximately 40% (diet records), 44% (24-h urinary urea), and 50% (Maroni's estimation; Table 2). Patients did not present body weight or body mass index changes over 3 mo (Table 1).

Discussion

The present study shows that compliance with a low-protein diet can be obtained by intensive dietitian interviews and that a moderately low-protein diet can be prescribed safely with an adequate energy intake. In addition, we provide further data that show clearly an adequate metabolic response when patients with mild CRF consume a moderately low-protein diet for an extended period.

Compliance with any diet is the key to success. Renal diets often are complicated by superimposed diseases, chronic medications, spontaneous anorexia, or inappropriate counseling. Furthermore, there is a regular spontaneous reduction in protein and energy intakes after renal function deterioration (9). In the present study, we assessed energy intake by food records and protein intake using a combined food record and urinary nitrogen measurement (8). Table 2 shows the decrease in daily protein intake as estimated by two independent means, i.e., food records and urinary urea nitrogen. The baseline protein intake was 1.13 g/kg per d, slightly lower than the mean protein intake in the French adult population (approximately 1.3 g/kg per d). After the protein restriction was initiated, both protein intake from records and urinary urea decreased to attain statistical significance from baseline at month 3 (Table 2). It is interesting to note that the difference between these two techniques in evaluating actual protein intake is approximately 0.20 to 0.30 g/kg per d during the first two diet interviews but drops to 0.15 g/kg per d after 3 mo of patient training and diet adjustment by the dietitian. These differences likely are explained by the fact that we did not recorded as “protein” the amount of nitrogen in the ketoanalog supplement (approximately 6 g of protein equivalent/d for 6 of the 12 patients), thus explaining a greater nitrogen output than recorded on the diet report (Table 2). In addition, a PNA higher than DPI could be explained by an underestimation of intakes by home diet records because this difference seemed to diminish over time after the number of dietitian interviews was increased (0.19 g of protein/kg per d at month 1, 0.29 at month 2, and 0.15 at month 3; Table 2), whereas patients did not change their ketoanalogs intake during the study. Although none of these differences was statistically significant, we believe that the quality of diet reports can be improved after repeated interviews between patient (and spouse if available) and the specialized renal dietitian. Other reasons for observing a PNA greater than the DPI include acute inflammatory or septic episodes, steroid treatments, or chronic acidosis. None of these conditions occurred during the study, and mean serum bicarbonate was 22.3 and 23.7 mmol/L at baseline and at the end of study, respectively (Table 1).

To avoid a well-described energy limitation associated with the reduction in protein intake (4,10,11), energy intake was monitoring carefully during diet interviews and kept constant during the study. There was no change in actual energy intake of patients during the 3-mo period (Table 2), and this level of energy intake can be considered adequate (12). Generally, energy intake and energy need are less reported than protein intake because it is more difficult to monitor energy needs than to collect urine for urea or nitrogen output measurement. Indeed, sophisticated research techniques, such as direct calorimetry performed in a metabolic chamber or double-labeled water, should be used to assess energy metabolism reliably. A study of 29 women reported that a 7-d diet record underestimated by approximately 20% the true energy needs as estimated by double-labeled water (13). If this were true for patients with CRF, then it may explain partly why such low energy intakes, i.e., 20 to 22 kcal/kg per d, are sometimes reported without concomitant severe malnutrition and/or impaired outcome (14). Another study comparing diet records and weighting food trays at the same time concluded that a diet record underestimated energy intake by <2% (15). In any event, the present energy intake of patients could be considered to be adequate, thus allowing a valid interpretation of the metabolic study (12).

Protein metabolism was assessed by the leucine turnover measurement. This technique has been used extensively in healthy humans. Adaptation to a reduction in protein intake is associated with a decrease in amino acid oxidation during fasting and postprandial states in healthy volunteers (16,17,18,19) and in patients with renal disease and/or renal failure (3,4,5,20,21,22). Table 3 indicates the values obtained before and after 3 mo on a low-protein diet. The fasting leucine rate of appearance (Table 3) and fasting leucine oxidation (Table 3 and Figure 1) decreased by 8% and 18%, respectively (P < 0.05) after reducing protein intake. These facts indicate that a proteinsparing mechanism occurred in response to the reduction of protein intake. The metabolic adaptation, hence, is present in patients with mild CRF, and the magnitude of the amino acid oxidation decrease we observed after 3 mo of protein reduction (- 18%, Table 3 and Figure 1) is suggestive of an important and sustained protein-sparing mechanism.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

The individual leucine oxidation value before and after a 3-mo low-protein diet. The low-protein diet was supplemented with ketoanalogs of amino acids (○) or not ([UNK]); ^*, P < 0.05 from baseline for all patients (paired t test); no difference between groups at baseline or at the end of study.

Other studies have reported similar findings in CRF patients who were undergoing different types of diet intervention but mostly during shorter studies. Goodship et al. (3) studied six patients with moderate CRF during a short-term (1 wk) regular (1 g of protein/kg per d) or reduced protein intake (0.6 g of protein/kg per d) and energy intakes of 32.5 kcal/kg per d. Fasting leucine oxidation did not change significantly during the lower protein intake, whereas postprandial leucine oxidation decreased by approximately 25% (P < 0.05). Extending a comparable diet up to 3 mo in the present study induced a significant and sustained reduction in fasting leucine oxidation (Table 3). Altogether, these data show that patients with moderate CRF can adapt their protein metabolism during acute and chronic protein intake reductions by reducing amino acid oxidation during both postprandial and fasting states.

More profound protein restrictions reduce amino acid oxidation to a greater magnitude. Masud et al. (5) showed in six predialysis patients that a diet that provided 0.35 g of protein/kg per d supplemented with either ketoacids or essential amino acids for 25 d allowed for maintaining neutral nitrogen balances and body composition. These diets were associated with very low leucine oxidation values, which were not different whether patients were supplemented with ketoanalogs or essential amino acids (5). In a long-term follow-up of these patients (16 mo), the fasting leucine oxidation remained at a low level of 10.0 ± 2.2 μmol/kg per h (4). As compared with the present study (baseline, 16.8 ± 4.2; low-protein diet, 13.8 ± 2.4 μmol/kg per h; Table 3), these amino acid oxidation values seem to be even lower in response to a lower protein intake, thus suggesting a potential “functional reserve” for protein sparing in CRF patients.

The addition of ketoacid supplements to the low-protein diet did not modify the protein metabolism as assessed by leucine turnover measurement (Table 3). Although there was a trend to a greater decrease in turnover values in the nonsupplemented group, none of these changes was significant. It could be argued that the protein restriction that we studied here was not restricted enough to observe the nitrogen-sparing effect of ketoacid administration. Indeed, in other diet intervention studies in uremia, the protein intake was reduced to a greater extent and generally averaged 0.3 to 0.5 g of protein/kg per d (4,5,23).

Amino acids were measured before and after the protein intake reduction (Table 4). Overall, there was almost no change in plasma essential, nonessential, and total amino acids. Branched chain amino acids did not vary between periods and within groups, whether patients received ketoacids or not, as reported by Masud et al. (5). In the ketoanalog-supplemented group, plasma tyrosine decreased by 15% (P < 0.05; Table 4) although tyrosine was included in the ketoanalog supplement (153 mg/pill, corresponding to an intake of 1.2 to 2 g/d). In the low-protein group that was not supplemented with ketoanalogs, arginine decreased by 34% (P < 0.05). More important, the essential/nonessential amino acid ratio (EAA/NEAA) decreased by 23% from 0.47 ± 0.07 to 0.36 ± 0.04 (P < 0.05) as a result of an increase in nonessential amino acids. These values are in agreement with those published by Kopple et al. (11,14). The absence of a comparable decrease of EAA/NEAA in the ketoanalog-supplemented group may be compatible with a better nutritional response. However, the absence of significant essential amino acid decrease in both groups during the study also is an important fact to consider (Table 4).

Finally, another indirect index for an adequate protein metabolism response is that serum insulin-like growth factor-1, a sensitive marker of body protein status (24,25), did not decrease during the 3-mo diet in each group (all patients: baseline, 278 ± 20 μg/L; end, 257 ± 20 μg/L) (26), thus indicating a well-preserved body protein compartment.

In summary, this is the first report of a 3-mo moderate reduction in protein intake in mild CRF patients that shows an adequate metabolic and body composition response. This suggests that under an energy intake >31 kcal/kg per d, a protein intake of 0.7 g/kg per d is metabolically and nutritionally safe. According to the current evidence for prescribing a low-protein diet to patients with mild CRF (27,28,29), this study confirms that such a diet therapy is worth being proposed to patients. Whether patients eventually will accept it certainly may rely more on physicians' beliefs and enthusiasm.