Stable isotopes have been used extensively for measuring whole-body protein turnover in children (Van Goudoever et al. 1995), young adults (Conway et al. 1980) and elderly subjects (Golden and Waterlow 1977). Methods with ¹⁵N-labeled tracers, e.g., ¹⁵N-glycine, have been favored for many years because of their noninvasive character. Protein turnover is calculated by measurement of isotope enrichment in end products of protein breakdown in urine. This method assumes that the whole-body free amino acids are present in one or two metabolically active pools from which urea and ammonia are derived. The ¹⁵N label enters the free amino acid pool and will exchange with other amino acids such that the ¹⁵N label can be regarded as a label for the total free amino acid pool. Urea and ammonia, the urinary end products of N metabolism, are derived solely from the free amino acid pool. Therefore their labeling can be taken as representative of that pool. In the case of ¹⁵N-glycine, the dose is usually given by constant nasogastric infusion, repeated oral doses or in one bolus dose.

Recently, precursor methods have been applied to a wide range of patient groups to derive quantitative information regarding protein metabolism. In this case, the label is infused intravenously until the plasma amino acid enrichment reaches a constant value. By giving a priming dose, the plateau can be achieved more rapidly. The protein turnover rate is calculated from the tracer dilution in the plasma and the known infusion rate. However, in the case of leucine, the tracer is infused in the plasma and samples are taken from the plasma, whereas leucine is metabolized within the cells. Because the [1-¹³C]-leucine plasma enrichment is not equal to the enrichment within the cells, plasma -ketoisocaproate (KIC)³ is usually considered to reflect more closely the enrichment of the precursor pool (Bruce et al. 1990, Matthews et al. 1982). KIC is formed intracellularly from leucine by transamination and is released to the plasma. In many studies, it has been shown that plasma KIC enrichment is a good reflection of the intracellular enrichment of leucine (e.g., Matthews et al. 1982).

Protein turnover measurement techniques are based on several assumptions (Halliday and Rennie 1982, Millward et al. 1991, Waterlow et al. 1978); each technique has its own limitations as discussed elsewhere (e.g., Garlick and Fern 1984, Halliday and Rennie 1982, Waterlow et al. 1978). Therefore absolute values for protein turnover rates tend to depend on the technique employed. However, it has been shown that there is broad agreement in general among values obtained by different approaches. Reported rates of protein synthesis and protein degradation in young adults obtained with different techniques range from 2 to 5 g/(kg·d) (Conway et al. 1980, Hoffer et al. 1985, Waterlow and Jackson 1981). Acceptable methods to measure protein turnover should not only give similar absolute values but should also reveal similar physiological directional changes in protein turnover whenever they occur (e.g., changes induced by trauma, feeding/fasting or diet of various composition). Garlick et al. (1991) and Young et al. (1987) reviewed the response in protein turnover to differences in protein intake. They concluded that dietary protein affects protein turnover at two levels, i.e., an immediate response to the intake of protein in meals (studies done during the fed state, e.g., Bruce et al. 1990, Garlick et al. 1987, Melville et al. 1989, Young et al. 1987) and a longer-term adaptation after a change in protein intake (studies done in the fed or fasted state after subjects had adapted to a certain level of protein intake, e.g., Conway et al. 1988, Pannemans et al. 1995a and 1995b). Only a few studies have reported on the effect of different levels of protein intake on whole-body protein turnover as measured with different tracers (Motil et al. 1981, Pacy et al. 1994, Quevedo et al. 1994, Yang et al. 1986, Zello et al. 1992), and no such studies have been undertaken in elderly subjects. Previous work of our group revealed a lower protein turnover in elderly women compared with elderly men and young women (Pannemans et al. 1995a and 1995b) at the same protein intake. Therefore we decided to investigate whole-body protein turnover in elderly women in more detail with different tracers. Thus it was of interest to compare the response of whole-body protein turnover with variations in dietary protein level as simultaneously measured with two commonly used tracers, ¹⁵N-glycine and [1-¹³C]-leucine, in the same elderly subjects.

SUBJECTS AND METHODS

Table 1. Diet components [View Table]

Fat-free mass (FFM) was also measured at the start and at the end of each dietary period with deuterium dilution (Van Marken Lichtenbelt et al. 1994) to determine whether muscle mass, e.g., FFM, was lost during these periods. No significant changes in FFM were observed during the dietary intervention period ( FFM diet A, +0.3 ± 0.9 kg; diet B, +0.1 ± 1.0 kg in 2 wk). Fat-free mass, measured simultaneously with protein turnover, was used to express data per kilogram FFM.

Whole-body protein turnover was measured with ¹⁵N-glycine given orally (200 mg) and with [1-¹³C]-leucine (99 atom %; Cambridge Isotope Laboratories, Woburn, MA) administered by continuous infusion (see below) on the last day of each diet period.

Rates of protein breakdown and synthesis with the ¹⁵N-glycine method were estimated from the urinary excretion of ¹⁵N in ammonia and urea during the 12 h after tracer administration (Pannemans et al. 1995a). Subjects were not allowed to eat or drink (except water) for at least 6 h before administration of the ¹⁵N-glycine (~1200 h) and during the following 12 h of the study period. Determination of urinary ammonia and urea concentrations, of ¹⁵N enrichment of urinary ammonia and urea and of plasma urea were performed as described earlier (Pannemans et al. 1995a, Read et al. 1982) Calculations were done as described by Fern et al. (1981). Rates of whole body protein synthesis and breakdown were derived from the expression Q = E + S = I + B, where E is the rate of excretion of total nitrogen in urine, S is the rate of whole-body protein synthesis, I is the rate of intake of nitrogen from the diet and B is the rate of whole-body protein breakdown. Because these studies were conducted in the fasting state, protein breakdown is equal to the calculated flux. A factor of 6.25 was used to convert grams of nitrogen into grams of protein. Whole-body protein turnover was calculated as the harmonic average of the estimates of flux based on urea and ammonia (Fern et al. 1984b).

Whole-body protein turnover was also measured employing a primed continuous infusion of [1-¹³C]-leucine during the last 4 h of the glycine study (Fig. 1). Subjects reported to the human physiology laboratory at 0700 h, and two venous catheters were inserted with aseptic technique into an anticubital vein of each arm. One catheter was used for withdrawing blood samples, the other for the isotope infusion. A priming dose of NaH¹³CO₃ (0.085 mg/kg) and [1-¹³C]-leucine (3.8 µmol/kg) was administered into the other catheter at 0800 h and immediately followed by the infusion of [1-¹³C]-leucine [3.6 µmol/(kg·h)] at a constant rate (50 mL/h) for 4 h using a calibrated pump (model 561, IVAC, San Diego, CA). Blood and expired air were obtained just before the administration of the isotope and at 30-min intervals thereafter during the first 2 h. During the 3rd and the 4th h, blood and expired air samples were collected at 15-min intervals. Each blood sample was collected with a heparinized syringe and transferred immediately to a tube that was kept on ice. Blood samples were centrifuged at 600 g for 10 min and plasma was stored at 20°C until analysis. After derivatization (Wolfe 1992), plasma ¹³C-KIC isotope enrichment was measured on a gas chromatography mass spectrometer (Finnigan Incos XL MS with a Varian 3400 GC with 1075 split/splitless injector, Bremen, Germany). Expired air samples were obtained by having the subject breathe normally for 3 min into a 6.75-L mixing chamber. After 3 min, a 20-mL vacutainer was filled with a sample of the mixed expired air. ¹³CO₂ enrichment in the expired air was measured with a GC continuous flow-IRMS (Finnigan MAT-252). At 3 and 4 h after the start of the infusion, the subject's CO₂ production rate was determined by means of a computerized open-circuit ventilated hood system.

Calculations of the rate of protein turnover were performed as described by Matthews et al. (1980). In short, leucine turnover is measured from the dilution of [1-¹³C]-leucine infusion in plasma KIC at isotopic steady state Q = i[(E_i/E_p)  1] where i is the [1-¹³C]-leucine infusion rate [µmol/(kg·h)]; E_i is the enrichment of the [1-¹³C]-leucine infused (MPE); E_p is the [1-¹³C]-KIC enrichment in the plasma at steady state (MPE). The rate of leucine oxidation is O = F¹³CO₂ [(1/E_p)  (1/E_i)] × 100, where F¹³CO₂ is the rate of ¹³CO₂ released by leucine tracer oxidation [µmol ¹³C/(kg·h)]. From these calculations, the rate of leucine incorporation into protein (nonoxidative leucine disappearance) is calculated, as described above, where Q = S + O = B + I. The leucine parameters were converted to corresponding estimates of whole-body protein turnover by multiplying the leucine values by the constant (24 h/d)/(590 µmol leucine/g protein) to give values of g protein/d (Matthews et al. 1980).

RESULTS

Table 2. Nitrogen balance data in six elderly women after consuming diets differing in protein content1 [View Table]

Table 3. Protein turnover as measured with 15N-glycine and [1-13C]-leucine in six elderly fasting women1 [View Table]

As measured with ¹⁵N-glycine, protein turnover (breakdown and synthesis) was significantly greater when subjects consumed 20% of energy (en%) as protein in the diet rather than 10 en% (P < 0.05), whereas no differences in protein turnover between the diets were measured with [1-¹³C]-leucine. Protein oxidation, measured with [1-¹³C]-leucine, was significantly higher when subjects consumed diet B compared with diet A.

Protein breakdown values measured with ¹⁵N-glycine were not different than those obtained with [1-¹³C]-leucine when subjects consumed both diets. However, protein synthesis rates were significantly lower when measured with ¹⁵N-glycine than with [1-¹³C]-leucine (diets A and B). Table 4 shows the protein balance data as derived by both tracers during both diet periods, as calculated from protein synthesis minus protein breakdown. Because subjects were studied in the postabsorptive state, breakdown was higher than synthesis, resulting in negative net protein balances. Both tracers showed that the absolute difference between breakdown and synthesis was higher during consumption of diet B (P < 0.05) compared with diet A, and the absolute difference between breakdown and synthesis was also higher when measured with ¹⁵N-glycine than with [1-¹³C]-leucine. However the relative increase in net balance was similar for both tracers (¹⁵N-glycine, 25 ± 27%; [1-¹³C]-leucine, 36 ± 4%).

Table 4. Protein balance data as measured with 15N-glycine and [1-13C]-leucine in six fasting elderly women1 [View Table]

DISCUSSION

In the past, protein requirements of the elderly have been studied extensively. Cheng et al. (1978) determined protein requirements in 15 elderly subjects and recommended a protein intake level of 0.8 g/(kg·d). Uauy et al. (1978) reported the same recommendation for elderly women [0.83 g/(kg·d)], whereas the protein intake for elderly men was established at 0.70-0.85 g/(kg·d). In another study, it was estimated that the protein requirements for elderly men were 0.72 g/(kg·d) (Zanni et al. 1979). Gersovitz et al. (1982) studied the long-term effects of a protein intake of 0.8 g/(kg·d) (as recommended by the FAO) on nitrogen balance in elderly men and women. Because most of the subjects were in negative nitrogen balance, it was concluded that the current recommended protein intake was too low for the majority of subjects 65 y. In the present study, a protein intake of 1.3 g/kg (diet B) resulted in a mean nitrogen balance that did not differ significantly from zero; at this protein intake, the elderly women were in nitrogen balance. It is therefore concluded on the basis of the current study and on retrospective nitrogen balance data that a safe protein intake for elderly adults would probably be higher than 0.8 g/(kg·d) because this amount may not always be sufficient to reach nitrogen balance, especially not in the frail elderly who have a low energy intake, low energy expenditure, low body weight and low physical activity levels.

Comparison of the results with previous studies is complicated by the differences in the experimental protocols, e.g., differences in tracers, subjects and the feeding status of the subjects. In summary, it can be concluded that protein turnover rates measured with leucine and glycine are either comparable or turnover rates measured with leucine are somewhat higher.

In the past, a few other investigators have reported on the effect of dietary protein levels on protein turnover as measured with stable isotope tracers in humans. The present ¹⁵N-glycine data are in agreement with data previously published from our laboratory concerning the effect of protein intake on protein kinetics in both young and elderly fasting subjects (Pannemans et al. 1995a and 1995b). As in the present study, we found significantly higher protein turnover rates when subjects consumed the high protein diet in both groups. Using the same tracer, Meredith et al. (1989) and Gersovitz et al. (1980) described the effects of a variable protein intake on protein turnover measured during 60 h of feeding and fasting. Both groups found increased overall fluxes with increasing protein intake. Previous studies measuring the effect of protein intake on protein metabolism as measured with [1-¹³C]-leucine are also more or less in agreement with the present data, i.e., increasing the protein intake has no effect on protein turnover of fasting subjects as measured with [1-¹³C]-leucine (Pacy et al. 1994, Yang et al. 1986, Zello et al. 1992). Quevedo et al. (1994) described the effect of a reduced protein intake and concluded that changing the protein intake from high to moderate intake levels had no effect on rates of protein synthesis and breakdown in fasting subjects, whereas oxidation rates decreased. The latter result was also described by Pacy et al. (1994), whereas others found no effect of protein intake on oxidation rates in fasting subjects when changing from marginal to surfeit protein intake (Motil et al. 1981). In contrast with the present study, Motil et al. (1981) reported increased protein breakdown and synthesis rates with increasing protein intake, whereas Yang (1986) described decreased protein fluxes as intake increased from 0 to 1.5 g/(kg·d). Recently, Casteneda et al. (1995) reported increased protein fluxes with increasing protein intake in fed elderly women, whereas no effects were observed on protein synthesis and breakdown. The effects of protein intake on leucine flux probably is less pronounced in fasting than in fed subjects (Motil et al. 1981, Pacy et al. 1994, Quevedo et al. 1994). In summary it can be concluded that the results of the present study are in agreement with data previously obtained under similar experimental conditions for both the ¹⁵N-glycine and [1-¹³C]-leucine tracers.

It remains unclear why protein turnover increased in subjects consuming the high priotein diet with ¹⁵N-glycine as tracer, whereas measurements done with [1-¹³C]-leucine did not detect any difference. To detect changes in protein turnover, the reproducibility of the technique used is important. However, the coefficients of variation of the leucine tracer technique and the glycine tracer technique are similar [6.8 and 6.7%, respectively (Bier et al. 1981, Fern et al. 1984a)]. It is not known why these differences were observed. Because both techniques rely on several assumptions, the differences may be attributable to the differences in assumptions made for both techniques. Promotion of the concept that a single given methodology represents the gold-standard approach to the estimation of whole-body protein turnover, applicable in both health and disease and to all age groups, would be difficult to defend. Direct comparison of results in studies in which ¹⁵N-glycine has been employed is often difficult in that so many variations on the central methodological theme have been reported (length of the study, oral or intravenous administration of the label, amount of tracer or choice of the end product). Given that the ammonia pool of the body is small and rapidly turning over, any pathophysiological disturbance of this pool could markedly affect the measured ¹⁵N enrichment in urinary ammonia. It has been reported in some neonatal studies that no ¹⁵N enrichment was measurable in urinary urea following ¹⁵N-glycine administration, thus precluding the harmonic mean calculation (Van Goudoever et al. 1995). There is only one published study reporting the effect of dietary protein intake on whole-body protein turnover data as obtained with ¹⁵N-glycine and [1-¹³C]-leucine, measured simultaneously in the same subjects, i.e., the study of Pacy et al. (1994) who used both ¹⁵N-glycine and [1-¹³C]-leucine to measure whole-body protein synthesis and degradation in healthy subjects given diets with increasing protein intakes [0.36, 0.77 and 1.59 g/(kg·d)]. They concluded that the ¹⁵N-results generally supported the leucine infusion data. However, measurements were done only in the fed state because the relatively long period required to clear all ¹⁵N from the metabolic pool precluded measurements in both fed and fasted subjects.

To make results of the present study directly comparable to previous studies (Pannemans et al. 1995a and 1995b), we choose to use the ¹⁵N-glycine method. Protein turnover was measured in the postabsorptive state with an orally administered dose (200 mg) of ¹⁵N-glycine. Protein turnover was calculated on the basis of excreted end products, urea and ammonia, during 12 h, according to an established procedure (Fern et al. 1981). This method was used because of its simplicity and its convenience for use outside the laboratory, especially when elderly people are involved. To investigate whole-body protein turnover in more detail and to allow a comparison between the two most commonly used stable isotopes, it was decided to employ the ¹³C-leucine method as well. Furthermore, the ¹⁵N-glycine method gives no information about the protein oxidation levels; thus the ¹³C-leucine method was used to directly measure this variable. However, this method requires a continuous infusion of the tracer, which means that subjects have to be in the laboratory during the 4-h measurement. For practical reasons, this period was chosen at the end of the overnight fasting period and not in the middle of the night. The differences in time needed to collect the necessary data (12 h in the case of ¹⁵N-glycine and 4 h in the case of ¹³C-leucine) resulted in the protocol as described. This might be another possible explanation for the differences between the two tracers, i.e., the fact that glycine measurements were done over 12 h after a 6-h fast, whereas leucine measurements were done over 4 h after a 14-h fast.

In summary, protein breakdown as measured with ¹⁵N-glycine is comparable to results from [1-¹³C]-leucine, whereas protein synthesis is significantly lower. Furthermore, measurements with ¹⁵N-glycine as tracer revealed higher protein turnover rates when subjects consumed the high protein diet, whereas no significant differences in protein turnover were found during consumption of the normal and high protein diet with the [1-¹³C]-leucine method. Leucine oxidation rates were higher though when subjects consumed the high protein diet. Both tracers showed that the absolute difference between breakdown and synthesis (net protein breakdown) was higher during consumption of the high protein diet (P < 0.05). Despite the fact that the absolute difference was higher as measured with the ¹⁵N-glycine method compared with the [1-¹³C]-leucine method, the relative increase in net protein breakdown during consumption of the high protein diet is comparable for both tracers. These data indicate that care is needed with the choice of the tracer used when the components of protein turnover are measured in elderly women with the aim of understanding the physiological basis behind the adequacy of the level of protein intake.

ACKNOWLEDGMENTS