Clinical Science and Molecular Medicine (1977) 52,485498.
Protein metabolism in human neonates: nitrogen-balance studies, estimated obligatory losses of nitrogen and whole-body turnover of nitrogen P. B. PENCHARZ,* W. P. STEFFEE,? W. COCHRAN, N . S. SCRIMSHAW, w. M. R A N D A N D V. R. YOUNG Department of Nutrition and Food Science, Massachusetts Institute of Technology, Cambridge, Massachusetts, U.S.A., and Boston Hospital for Women, Boston, Massachusetts, U.S.A.
(Received 3 March 1976; accepted 30 December 1976)
SummarY 1. Aspects of nitrogen metabolism in the human neonate were assessed in one full-term infant and six premature infants by means of nitrogen-balance measurements, estimates of obligatory nitrogen losses and determinations of whole-body nitrogen turnover. 2. Our data indicate that the mean protein requirement for maintenance is 1.1 g of protein day-' kg-' and that 3.8 g of protein day-' kg-' should be sufficient for adequate growth in healthy premature babies. 3. The mean obligatory urinary, faecal and total nitrogen losses were estimated to be 24, 106 and 145 mg day-' kg-' respectively. These figures are compared with published values for older infants, and the possible metabolic basis for changes in nitrogen losses during growth and development is discussed. 4. Mean values for whole-body protein synthesis and breakdown were 26.3 k 7.0 and 23.8 +7-4 g of protein day-' kg-I respectively. Dietary nitrogen intake accounted for 6 1 8 % of the nitrogen flux through the metabolic pool; urea excretion accounted for 2% of the nitrogem flux.
* Present address: The Montreal Children's Hospital, Department of Metabolism, Montreal, Que. H3H IP3, Canada. t Present address: Boston University Medical Center, 75 East Newton Street, Boston, Mass. 02118, U S A . Correspondence: PublicationsUnit, 16-334A, Department of Nutrition and Food Science, Massachusetts Institute of Technology, Cambridge, Mass. 02139, USA.
5. The net protein gain, estimated from nitrogen-balance data, accounted for 9.6% of total daily protein synthesis. 6. These results are discussed in relation to published estimates of whole-body protein synthesis and breakdown at various ages. Their possible significance in the assessment of a 'maintenance' requirement for protein and amino acids during the period of rapid growth and development is also considered. Key words: amino acids, growth, neonates, nitrogen balance, protein. Introduction Relatively few studies have examined the human neonate's requirement for dietary protein and the characteristics of its protein metabolism (Davidson, Levine, Gauer & Dann, 1967; Babson & Bramhall, 1969; Nicholson, 1970; Irwin & Hegsted, 1971; Raiha, 1974). Although the protein needs of premature babies may be higher than those for full-term infants, this greater requirement is neither well characterized nor explained metabolically. For these reasons, we have carried out a series of studies in premature infants within 1-45 days after their births, in order to characterize this group's nitrogen metabolism by measurements of nitrogen balance and estimates of obligatory nitrogen losses. Body protein metabolism is more intense in species of small mammals than in larger ones,
485
P. B. Pencharz et al.
486
including man (Munro, 1969). Furthermore, it decreases with normal growth and development (Waterlow & Stephen, 1967; Picou & TaylorRoberts, 1969; Steffee, Goldsmith, Pencharz, Scrimshaw & Young, 1976). As part of the present study, we estimated the rate of body nitrogen turnover in premature neonates, using the approach proposed by Picou & TaylorRoberts (1969). We have previously applied this method in our studies on young adults (Steffee et al., 1976) and on elderly subjects (Winterer, Steffee, Perera, Uauy, Scrimshaw & Young, 1976). Our findings show that the intensity of nitrogen metabolism, per unit of body weight, is approximately eight times higher in neonates than in young adult subjects. Methods
Subjects and diet One full-term (C.C.) and six premature infants were studied; Table 1 presents the anthropometric and gestational characteristics of these seven infants. One baby (S.H.) was studied on three different occasions and two (L.C. and B.M.) were each studied twice (L.C.l, L.C.2; B.M.l, B.M.,). We have excluded the nitrogen-balance data from the first study with infant B.M. because this child received an exchange transfusion 11 h after the start of the balance period. The birth weights ranged from 1120 to 1758 g for the premature babies, and their gestational ages were estimated to be 30-36 weeks.
Infant J.B. was small for gestational age (Usher, McLean & Scott, 1966). Infant C.C., who was less than 2 days old, lost weight during the study; this weight loss was probably caused by a change in body water content (Smith, 1946), since the baby was in positive nitrogen balance at the time of the study. All the babies except M.N. consumed a commercial infant milk formula (Similac) ; infant M.N. was fed with expressed human breast milk. In addition, during the first two studies with infant S.H., 0.5 ml of mediumchain triglyceride was included in each feeding. All babies were fed every 3 h throughout the study period, except S.H. who was fed at 2 h intervals during the first study. As shown in Table 2, the babies' total protein intakes varied from 1.5 to 4.4 g day-' kg-l and their energy intakes varied from 255 to 732 kJ day-' kg-l, equal to approximately 170 kJ/g of consumed protein (except for infant S.H., who received the medium-chain triglyceride supplement ; his energy intakes were 230 and 194 kJ/g of protein during the first and second nitrogen-balance studies respectively).
Procedures The experiments received the administrative approval of the MIT Committee on the Use of Humans as Experimental Subjects, the Executive and Policy Committees of the MIT Clinical Research Center and the Research Advisory Committee of the Boston Hospital for Women. Written informed consent was obtained from
TABLE 1. Anthroponietric and gestational data on secen neonates studied for characteristics of body nitrogen metabolism Surface area was calculated from Wt.0'425x Ht.0.725x 71.84, where Wt. = weight in kg and Ht. = height in cm (surface area cm2). Subject and study no.
Birth wt. (B)
w.c.1
1729 1503 1129
J.B.1
M.N.1 L.C.1 L.C.2 B.M.l B.M.2
c.c.1 S.H., S.H.2 S.H.S
1758 1758 1673 1673 3473 I120 I I20 1120
wt. during study (g) 2218 1998 I836 1772 2055 1612 1807 3345 1091 1495 1935
Age during study
Birth length (cm)
Surface area (m2)
24 days 24 days 14 days 44 hours 14 days 29 hours 13 days 45 hours 16 days 31 days 45 days
44.4 45.5 45.7 42.0 42.0 41.9 41.9 48.5 38.1 38.1 38.1
0157 0154 0.148 0.138 0.146 0.132 0140 0200 0.104
0.128 0.147
Gestation (weeks) 30 34-35 32-33 35-36 35-36 34 34 42 31 31
31
Wt. gain during study (g day-' kg-') 19.2 26.1 12.6 15.9 17.3 -11.7 15.6 - 2.7 10.9 19.1 22.0
TABLE 2. Protein intake, energy intake and nitrogen-balance data for sewn neonates studied for characteristics of body nitrogen metabolism
S.H.' S.H.2 S.H.3
c.c.1
J.B.1 M.N.l L.C.1 L.C.2 B.M.I B.M.2
w.c.1
Subject and study no.
3.4 3.7 2.8 2.5 3.6 2.3 4.4 1.5 2.9 3.2 3.6
Protein intake (g day-' kg-')
605
575 625 462 420 615 38 1 735 258 672 620
Energy intake (kJ day-' kg-')
400 585 361 70 1 242 464 508 576
444
548 596
N intake (mg day-' kg-')
143 122 166 62 76 113 134
09
156 134 71
Urine N (mg day-' kg-')
42 42
100
123 46 97 I68 49 166 97 95
Faecal N (rng day-' kg-')
422 73 270 335 384
-
253 40 1 260 127 379
N balance (mg day-' kg-')
37 38 41
50
60 63 68
71 66 45 54
total N)
(% urinary
Urea N
N balance is corrected for estimated integumental and miscellaneous N losses, as discussed in the text. N balance results for B.M.' are excluded, as discussed in the Methods section.
s'
488
P. B. Pencharz et al.
the parents, and permission was obtained from the infants’ pediatricians. The technique used to determine nitrogen balance was a modification of that described by Hepner & Lubchenco (1960). Each nitrogenbalance period lasted 30-36 h. Plastic bags for urine were attached to the skin around the baby’s anus in order to collect each stool separately. After a stool had passed, the bag was removed and the faeces were wiped from the infant’s skin with cotton balls, which were included with the stools for analysis. Use of Carmine Red as a faecal marker facilitated the preparation of pooled samples (Fomon, 1974b). The apparent nitrogen balance was calculated as the difference between nitrogen intake and the sum of faecal and urinary nitrogen losses. We corrected this calculation for integumental and miscellaneous losses of nitrogen, as discussed below, in order to estimate the protein requirements for nitrogen equilibrium. Gestational age was assessed in two ways: from an evaluation of (a) the infant’s anthropometric characteristics (length, weight and head circumference) and (b) his external characteristics, as described by Usher et al. (1966). All babies were examined for gestational age by one of us and then independently by the baby’s pediatrician, with essentially the same results. When a discrepancy existed between the clinical assessment and the obstetrical data, the clinical evaluation was used to estimate gestational age. We measured each infant’s body length from crown to heel, ensuring full extension without pelvic tilt. All body weights were recorded daily between 07.30 and 08.00 hours. Each infant’s weight gain was calculated as the weight change over a 3 days period, including the period of nitrogen-balance measurement. Experinrental model To study dynamic aspects of whole-body nitrogen metabolism, we modified slightly the approach described by Picou & TaylorRoberts (1969), which involves the administration of [ 5N]glycine at a constant rate in order to achieve a steady state of 15N enrichment in the metabolic nitrogen pool. This method overcomes some of the difficulties (e.g. Waterlow, 1969) inherent in the single-isotope-dose method of San Pietro & Rittenberg (1953).
The Picou & Taylor-Roberts method allows estimates to be made of the nitrogen flux (Q) through the metabolic pool, i.e. the nitrogen disposal rate (Shipley & Clark, 1972), and the rates of whole-body protein synthesis (S) and breakdown (C). The assumptions inherent in this approach to the study of whole-body protein turnover have been discussed and tested experimentally (Picou & Taylor-Roberts, 1969; Steffee et al., 1976). Isotope administration We obtained [15N]glycine (95 atoms % excess) from the Stohler Chemical Corp. (Waltham, Mass., U.S.A.). The enrichment of the purchased material was confirmed by massspectrographic analysis. In all of the studies, the [15N]glycinewas administered at a known rate to supply about 0.6 mg of 15N 24 h-’ kg - l body weight. Picou & Taylor-Roberts (1969) administered the labelled amino acid by constant intravenous or intragastric infusion. In our studies with healthy young adults given an adequate diet (Steffee et al., 1976), we found that [15N]glycine given orally every 3 h for 60 h allowed a satisfactory concentration of urinary [l 5N]urea to be obtained. Furthermore, in the same study, we noted that estimates of body nitrogen flux were similar whether the subjects were given [ 5N]glycine orally or intravenously. Therefore, in this study with neonates, the [‘5NN]glycinewas mixed with the bottle feeds, which the infants received every 2 or 3 h for a total of 30-36 h. Infant B.M. received an exchange transfusion (2 volumes) with whole blood 11 h after the start of the first study. The exchange was performed for hyperbilirubinaemia, secondary to her being the recipient in a twin-transfusion syndrome. For these reasons the nitrogenbalance data from this study were excluded. However, we determined that the exchange caused a nitrogen loss of 539 mg and that this loss was mainly associated with erythrocytes. In addition, the time-course of urinary [ * 5N]urea enrichment resembled that found with the other infants. We decided therefore to include the B.M.I results on body nitrogen turnover in the present analysis. Analyses Food, urine and faeces were analysed for total nitrogen; in addition, urea and creatinines
Nitrogen metabolism in neonates were measured in the urine (Young, Taylor, Rand & Scrimshaw, 1973). After pretreating the urine with Permutit (Folin & Bell, 1917), we isolated the urinary urea nitrogen by a modification of the Conway diffusion method (Hawk, Oser & Summerson, 1954). The I5N enrichment of urinary urea was determined with the aid of a dual-collector, isotope-ratio mass spectrometer (Vacumetrics model MS11, Waltham, Mass., U.S.A.) (Steffee et al., 1976). The 15N enrichments of the administered isotope and of faeces were determined after digestion of the faecal samples (Munro & Fleck, 1969) and steam-distillation of the resulting ammonia (Rittenberg, 1946). Samples for mass-spectrographic analysis were measured in duplicate. The reproducibility of measurements of excess I5N was within 5 % ; the massspectrographic analysis has been previously described (Steffee et al., 1976). Data analysis We calculated the SNenrichment of urinary urea according to the method described by Rittenberg (1946). Enrichment was corrected for background values that were determined from urine samples collected over a 3 h period preceding each tracer study. The plateau value of 5N enrichment of urea was estimated by two methods. For the analysis of the whole-body nitrogen turnover results described below, plateau enrichment was calculated by averaging the values for the last three data points of each turnover study, provided they did not differ by more than 25%.
489
Because the urinary I5N urea enrichment, which occurred after the beginning of isotope administration, usually represented a smooth, curvilinear change, we also predicted, by the use of asymptotic regression analysis, that plateau enrichment would occur. The midpoint of each urine collection was used as the time for each sample. The least-squares method was used to fit the individual data sets to an exponential equation: y = C (l-e-kL). Totalbody nitrogen flux was calculated by estimating the asymptote, y (a) = C,as the plateau value. The fitting was carried out by means of a non-linear program of a computer program package (Dixon, 1973). The program uses a modified Gauss-Newton technique (Hartley, 1961; Jennrich & Sampson, 1968) to search until five successive iterations do not change the total error mean square by more than 0.001%. In addition, the program calculates the standard deviation of the estimates. Statistical evaluations were performed with the Student's t-test, regression and correlation analysis (Dixon & Massey, 1969).
Results Nitrogen balance Table 2 summarizes the nitrogen-balance data for the seven infants. These results were based on stool and urine collections made during a 30-36 h period of [lsN]glycine administration. All infants were in positive apparent nitrogen balance, with values ranging from +85 to +438mgday-'kg-'bodyweight
TABLE 3. An estimate of the protein content of weight gain as calculatedfrom nitrogen-balance data in six neonates Subject and study no.
Wt. gain (g day-, kg-')
N balance (mg day-' kg-I)
Protein content of weight gain
(%)
w.c.1 J.B.1 M.N., L.C.1 L.C.2 B.M., S.H.1 S.H.2 S.H.3
Mean +SD
19.2 26.1 12.6 15.9 17.3 15.6 10.9 19.1 22.0
253 40 1 260 127 379 422 270 335 384
8.2 9.6 12.9 8.0 13.6 16.9 15.5 11.0 10.9
17.6+ 4.6
314+95
11*8+3.1
P . B. Pencharz et al.
490
TABLE 4. Results of regression analysis ofprotein intake on nitrogen balance, weight gain and ihe relationship between absorbed nitrogen and retained nitrogen Equation Regression
(Y = a+ b x )
sb
r
~~
(1) (2) (3) (4) (5)
Apparent N balance ( y ) on N intake ( x ) * N balance ( y ) on N intake (x)* Apparent N balance ( y ) on N absorbed (x)* N balance ( y ) on N absorbed (x)* Weight gain (y)? on protein intake (x)j
-
y = 131+0,86x y = -145+0.86~ y = - 24.2+0.78x
0.12 012 0.050
y = -38.9+078~ y = - 8.2+ 7 . 5 ~
0.048 2.22
0.93 093 098 0.99 0.77
* x and y expressed as mg of N day-' kg-'. Apparent N balance = N intake-(urinary N+faecal N). N balance = apparent N balance - integumental N loss. t y expressed as g day-' kg-'. x expressed as g of protein day-' kg-'. 500 C
-- 400 m
O*m-
v
m'
-Es 2000 c
x=
looI
1
I
In order to estimate the protein requirements for nitrogen equilibrium, we needed to include sweat and other miscellaneous losses of nitrogen in our calculations. Because these losses have not been determined for infants, we used data obtained in adults. Based on a review of published data, the FAO/WHO (1973) Committee chose a value of 5 mg of nitrogen day-' kg-' to cover the integumental and other minor routes for nitrogen loss in young men. Assuming a surface area of 1-75 mz for a young adult subject, we can calculate the unmeasured nitrogen losses at about 200 mg/m2. Applying this figure on a surface-area basis, we have estimated the miscellaneous nitrogen losses for the seven infants (Table 2). From these calculations, mean integumental and other unmeasured nitrogen losses are estimated to be 15.2 mg day-' kg-' in premature infants. The regression analysis indicates that a mean intake of 170 mg of nitrogen day-' kg-' is required to replace urinary, faecal and other losses. Hence, a protein intake (N x 6.25) of 1.06 g day-' kg-I would be required, on average, to compensate for the major and minor routes of body loss of nitrogen. The value for the y-intercept, 145 mg day- kg-', provides a prediction of the mean total obligatory nitrogen loss for the group studied (Table 4, eqn. 2). Finally, the slope of the regression line is an estimate of the efficiency of dietary nitrogen utilization, and this value was found to be 86% for the range of protein intake given to the seven infants. We also obtained a significant positive relationship (r = 0.77; P i0.01)between dietary protein intake (g day-' kg-') and body weight gain (g day-' kg-') (Fig. 2). The regression
Nitrogen metaboilism in neonates
T
* /
24
491
with an efficiency of 78%, or only slightly less than the value derived from the regression of nitrogen balance on total intake of nitrogen (Fig. 1).
/ Whole-body nitrogen turnover
FIG.2. Relationship between protein intake and body weight gain in human neonates. The continuous line depicts the regression equation y = -8.2+7.5x, where y = weight gain (g day-' kg-') and x = protein intake (g day-' kg-I).
t
500
0
I
I
I
I
200
400
600
h b e d N (mg day-' kg-')
Fro. 3. Relationship between retained N and absorbed N in human neonates. The continuous line depicts the regression equation y = -38*9+078x, where y = N retention (mg of N day-' kg-') and x = N absorption (mg of N day-' kg-').
equation was y = -8*2+73x, where y = weight gain (g day-' kg-I) and x = protein intake (g day-I kg-I). Hence, according to these calculations, a baby must have a mean intake of 1.09 g of protein day-l kg-' for weight maintenance alone, and this value is comparable to that calculated for body nitrogen maintenance (Table 4, eqn. 5). Finally, we examined the relationship of retained nitrogen to absorbed nitrogen (Fig. 3); these values showed a significant linear correlation (r = 0.99; P < 0.001). The y-intercept provides an estimate of mean obligatory urinary N plus integumental and other minor losses; it was determined to be 38.9 mg of N day kg-' (Table 4, eqn. 4). The slope of the regression line indicates that absorbed nitrogen is utilized D
Table 5 shows the ['5N]urea enrichment during the final phase of the I5N turnover period and also shows the plateau values of enrichment based on methods described earlier. Fig. 4 depicts, for a representative case, the predicted and observed change in I5N enrichment of urinary urea during a 36 h period of I5N administration. By means of regression analysis, we estimated a plateau (asymptote) value for [15N]urea enrichment for each tracer study (Table 5). Only a small fraction of the ingested lSN was excreted in the infants' stools (Table 5). In calculating parameters of body nitrogen metabolism, we have assumed that all of the administered dose entered the metabolic pool. For one infant (L.C.'), the time-course of [15N]urea enrichment appeared to change linearly throughout and, in this case, it was impossible to predict a plateau value. There was a wide range of variability in the precision with which it was possible to predict the plateau value by regression analysis. The estimated standard deviations of the plateau ranged from 5 to 76%. We therefore decided not to use this procedure for calculation of nitrogen flux, but rather to estimate the plateau from the last three data points if they varied by no more than 25%. With this criterion, we were able to utilize data from eight of the eleven turnover studies; the mean difference in the last three data points in these eight subjects was 15% (range 3-24%). In the remaining three subjects, the differences were 95% (M.N.'), 42% (C.C.1) and 81 % (S.H.J. Table 6 summarizes the estimates for body nitrogen flux and the rates of whole-body protein synthesis and breakdown for the seven neonates. The mean flux was equivalent to 181 mg of N h-' kg-', which can be compared with the mean value of 25.7 mg of N h-I kg-' for healthy young adults (Steffee et al., 1976). In the present series of studies we did not obtain a statistically significant change in the nitrogen flux with age, although the data do suggest a trend towards lower values with increasing age. The mean rates of protein (N x 6.25) synthesis
P. B. Pencharz
492
et
al.
5. ouiput in faeces, urinary urea enrichmeni and plateau values for urea ' N , obtained by visual inspection and by non-linear regression analysis, in seven neonates given ['5N]glycine during 30-36 h
TABLE
Urine collections were usually made every 3 h ; times given are the midpoint of each collection. Plateau I5N enrichment was calculated by determination of the mean value for the last three collection periods. If the values of the final three collections differed by more than 25%, a plateau estimate was not made. The predictive method is described in the text.
Subject and study
no. W.C.,
J.B.,
M.N.,
L.C.,
L.C.2
B.M.,
U.M.2
C.C.,
S.H.,
S.Hq
S.H.3
Time of sample (h)
0 21.04 23.79 29.84 0 27.95 30.82 34.02 0 16.58 19.75 31.75 0 28.46 31.12 34.21 0 29.25 32.16 34.83 0 28.15 31.79 34.40 0 29.12 32.75 35.12 0 2408 28.26 34.25 0 18.16 22.25 33.74 0 2846 31.62 34.25 0 28.96 32.12 34.61
IOd3xI5N enrichment (atoms % excess)
10-3x Plateau [I5N]urea (atoms % excess)
I5N in faeces*
Inspected
Predicted
(%)
12.3
2 1.3+_2.7
4.0
15.1 15.2 15.5
16.1
42.7+ 31.0
0.4
0.7 5.0 3.9 7.6
-
lZ.O+ 7.2
I .2
0.1
11.8 11.7 13.7
- 0.8
1 .o
16.3 18.3 19.8 1.7 20.7 17.7 22.0 0.5 8.3 8.8 10.3 0.4 16.4 17.6 17.6
17.1
-
18.4
3 1 . 2 2 7.2
1.2
8.6
15.9+ 3.9
2.6
16.8
20.9+ 1.0
0.8
-
17.3+ 13.2
I .o
-
21.3f8.6
4.6
11.4
13.4k2.3
0.3
11.7
27.4+_8.2
0.7
1.5
9.4 9.3 13.2 1.4 6.4 9.0 11.6
- 0.8 11.3 10.0 106 0.5
12.0 11.7 13.0
* Expressed as the percentage of the administered dose that appeared in the faeces passed during the I5N tracer study.
Nitrogen metabolism in neonates
["N]Glycine administroton ( h 1
FIG.4. Time-course of I5N enrichment of urinary urea N after oral doses of [lsN]glycine given every 3 h to subject B.M.2: 0 , predicted; 0 , observed.
and breakdown were 26.3 and 23.8 g day-l kg-' respectively. These two values are approximately eight times the rates that occur in young adults (Steffee et al., 1976) and about four times the values reported by Picou 8t Taylor-Roberts (1969) for infants of about 1 year of age. The entry of nitrogen from food into the metabolic pool accounted for 6-18% (mean 12%) of the total nitrogen flux; urea nitrogen excretion constituted about 2% of the total nitrogen flux through the metabolic pool. The daily protein gain, as calculated from the nitrogen-balance data, amounted to 11.8%
493
of total body-weight gain (Table 3). Thus, from the net body protein gain and the total body protein synthesis rate, we estimate that on average the protein gain accounted for only 9.6% of total protein synthesis per day (Table 7). These data indicate that the greater proportion of protein synthesis and, thus, of amino acid utilization is associated in the neonate with a rapid turnover of existing protein, rather than with net protein synthesis. The results in Table 6 reveal considerable individual variation in estimated rates of wholebody protein synthesis and breakdown. Because dietary protein intakes varied among the infants,cwe examined the relationships between protein and energy intakes and the parameters of whole-body nitrogen metabolism. For this purpose, the parameters were expressed per unit body weight and per unit surface area; Table 8 shows the correlation matrices. No statistically significant correlation existed between protein or energy intakes and the rates of synthesis, breakdown, nitrogen flux, or the fraction of nitrogen flux that was excreted as urea. However, both protein and energy intakes were correlated significantly with synthesis and catabolism when they were expressed as fractions of whole-body nitrogen flux, except for energy intake (kJ day-' kg-') in relation to the term S/Q.
Discussion Nitrogen balance
Nitrogen-balance determinations among any age group are difficult to carry out with pre-
TABLE 6. Whole-body nitrogen flux and rates of whole-body protein synihesis and breakdown in five neonates
Subject and study no.
Whole-body N flux (Q) (g day-' kg-')
Whole-body protein synthesis ( S ) (g day-' kg-')
Whole-body protein breakdown (C) (g day-' kg-')
W.C.' J.B., L.C.1 L.C.2 B.M.1 B.M.2 S.H.2 S.H.3 Mean f s ~
4.18 3.51 3.60 3.19 6.38 3.58 4.99 5.34 435* 1.12
25-2 21.1 21.9 19.0 39.1 21.3 305 32.5 26.3f 1.0
22.1 18.2 20.0 16.3 37.6 18.0 28.0 29.8 23,8+ 7.4
P . B. Pencharz et al.
494
TABLE7. Relationship between body protein gain and whole-body protein synthesis in five neonates
Subject and study no.
Estimated protein gain* (g day-' kg-')
(Protein gainlprotein synthesis) x 100
2.26 3.08 1.88 2.04 1.84 2.25 2.60
9.0 14.6 8.6 10.7 8.6 7.4
2.28+ 0.43
9 . 6 k 2.4
W.C., J.B.1
L.C.' L.C.2 B.M.2 S.H.2 S.H., Mean ~
S
D
8.0
* 11.8'A of weight gain was estimated to be protein (Table 3); protein gain = weight gain x 0.1 18. TABLE 8 . Statistical correlations ( r ) in neonates between dietary protein and energy intake andparariieters of whole-body protein turnover expressedper unit body weight and per unit body surface area Q = N flux;
S = whole-body protein synthesis; C = whole-body protein breakdown. * P
Protein metabolism in human neonates: nitrogen-balance studies, estimated obligatory losses of nitrogen and whole-body turnover of nitrogen P. B. PENCHARZ,* W. P. STEFFEE,? W. COCHRAN, N . S. SCRIMSHAW, w. M. R A N D A N D V. R. YOUNG Department of Nutrition and Food Science, Massachusetts Institute of Technology, Cambridge, Massachusetts, U.S.A., and Boston Hospital for Women, Boston, Massachusetts, U.S.A.
(Received 3 March 1976; accepted 30 December 1976)
SummarY 1. Aspects of nitrogen metabolism in the human neonate were assessed in one full-term infant and six premature infants by means of nitrogen-balance measurements, estimates of obligatory nitrogen losses and determinations of whole-body nitrogen turnover. 2. Our data indicate that the mean protein requirement for maintenance is 1.1 g of protein day-' kg-' and that 3.8 g of protein day-' kg-' should be sufficient for adequate growth in healthy premature babies. 3. The mean obligatory urinary, faecal and total nitrogen losses were estimated to be 24, 106 and 145 mg day-' kg-' respectively. These figures are compared with published values for older infants, and the possible metabolic basis for changes in nitrogen losses during growth and development is discussed. 4. Mean values for whole-body protein synthesis and breakdown were 26.3 k 7.0 and 23.8 +7-4 g of protein day-' kg-I respectively. Dietary nitrogen intake accounted for 6 1 8 % of the nitrogen flux through the metabolic pool; urea excretion accounted for 2% of the nitrogem flux.
* Present address: The Montreal Children's Hospital, Department of Metabolism, Montreal, Que. H3H IP3, Canada. t Present address: Boston University Medical Center, 75 East Newton Street, Boston, Mass. 02118, U S A . Correspondence: PublicationsUnit, 16-334A, Department of Nutrition and Food Science, Massachusetts Institute of Technology, Cambridge, Mass. 02139, USA.
5. The net protein gain, estimated from nitrogen-balance data, accounted for 9.6% of total daily protein synthesis. 6. These results are discussed in relation to published estimates of whole-body protein synthesis and breakdown at various ages. Their possible significance in the assessment of a 'maintenance' requirement for protein and amino acids during the period of rapid growth and development is also considered. Key words: amino acids, growth, neonates, nitrogen balance, protein. Introduction Relatively few studies have examined the human neonate's requirement for dietary protein and the characteristics of its protein metabolism (Davidson, Levine, Gauer & Dann, 1967; Babson & Bramhall, 1969; Nicholson, 1970; Irwin & Hegsted, 1971; Raiha, 1974). Although the protein needs of premature babies may be higher than those for full-term infants, this greater requirement is neither well characterized nor explained metabolically. For these reasons, we have carried out a series of studies in premature infants within 1-45 days after their births, in order to characterize this group's nitrogen metabolism by measurements of nitrogen balance and estimates of obligatory nitrogen losses. Body protein metabolism is more intense in species of small mammals than in larger ones,
485
P. B. Pencharz et al.
486
including man (Munro, 1969). Furthermore, it decreases with normal growth and development (Waterlow & Stephen, 1967; Picou & TaylorRoberts, 1969; Steffee, Goldsmith, Pencharz, Scrimshaw & Young, 1976). As part of the present study, we estimated the rate of body nitrogen turnover in premature neonates, using the approach proposed by Picou & TaylorRoberts (1969). We have previously applied this method in our studies on young adults (Steffee et al., 1976) and on elderly subjects (Winterer, Steffee, Perera, Uauy, Scrimshaw & Young, 1976). Our findings show that the intensity of nitrogen metabolism, per unit of body weight, is approximately eight times higher in neonates than in young adult subjects. Methods
Subjects and diet One full-term (C.C.) and six premature infants were studied; Table 1 presents the anthropometric and gestational characteristics of these seven infants. One baby (S.H.) was studied on three different occasions and two (L.C. and B.M.) were each studied twice (L.C.l, L.C.2; B.M.l, B.M.,). We have excluded the nitrogen-balance data from the first study with infant B.M. because this child received an exchange transfusion 11 h after the start of the balance period. The birth weights ranged from 1120 to 1758 g for the premature babies, and their gestational ages were estimated to be 30-36 weeks.
Infant J.B. was small for gestational age (Usher, McLean & Scott, 1966). Infant C.C., who was less than 2 days old, lost weight during the study; this weight loss was probably caused by a change in body water content (Smith, 1946), since the baby was in positive nitrogen balance at the time of the study. All the babies except M.N. consumed a commercial infant milk formula (Similac) ; infant M.N. was fed with expressed human breast milk. In addition, during the first two studies with infant S.H., 0.5 ml of mediumchain triglyceride was included in each feeding. All babies were fed every 3 h throughout the study period, except S.H. who was fed at 2 h intervals during the first study. As shown in Table 2, the babies' total protein intakes varied from 1.5 to 4.4 g day-' kg-l and their energy intakes varied from 255 to 732 kJ day-' kg-l, equal to approximately 170 kJ/g of consumed protein (except for infant S.H., who received the medium-chain triglyceride supplement ; his energy intakes were 230 and 194 kJ/g of protein during the first and second nitrogen-balance studies respectively).
Procedures The experiments received the administrative approval of the MIT Committee on the Use of Humans as Experimental Subjects, the Executive and Policy Committees of the MIT Clinical Research Center and the Research Advisory Committee of the Boston Hospital for Women. Written informed consent was obtained from
TABLE 1. Anthroponietric and gestational data on secen neonates studied for characteristics of body nitrogen metabolism Surface area was calculated from Wt.0'425x Ht.0.725x 71.84, where Wt. = weight in kg and Ht. = height in cm (surface area cm2). Subject and study no.
Birth wt. (B)
w.c.1
1729 1503 1129
J.B.1
M.N.1 L.C.1 L.C.2 B.M.l B.M.2
c.c.1 S.H., S.H.2 S.H.S
1758 1758 1673 1673 3473 I120 I I20 1120
wt. during study (g) 2218 1998 I836 1772 2055 1612 1807 3345 1091 1495 1935
Age during study
Birth length (cm)
Surface area (m2)
24 days 24 days 14 days 44 hours 14 days 29 hours 13 days 45 hours 16 days 31 days 45 days
44.4 45.5 45.7 42.0 42.0 41.9 41.9 48.5 38.1 38.1 38.1
0157 0154 0.148 0.138 0.146 0.132 0140 0200 0.104
0.128 0.147
Gestation (weeks) 30 34-35 32-33 35-36 35-36 34 34 42 31 31
31
Wt. gain during study (g day-' kg-') 19.2 26.1 12.6 15.9 17.3 -11.7 15.6 - 2.7 10.9 19.1 22.0
TABLE 2. Protein intake, energy intake and nitrogen-balance data for sewn neonates studied for characteristics of body nitrogen metabolism
S.H.' S.H.2 S.H.3
c.c.1
J.B.1 M.N.l L.C.1 L.C.2 B.M.I B.M.2
w.c.1
Subject and study no.
3.4 3.7 2.8 2.5 3.6 2.3 4.4 1.5 2.9 3.2 3.6
Protein intake (g day-' kg-')
605
575 625 462 420 615 38 1 735 258 672 620
Energy intake (kJ day-' kg-')
400 585 361 70 1 242 464 508 576
444
548 596
N intake (mg day-' kg-')
143 122 166 62 76 113 134
09
156 134 71
Urine N (mg day-' kg-')
42 42
100
123 46 97 I68 49 166 97 95
Faecal N (rng day-' kg-')
422 73 270 335 384
-
253 40 1 260 127 379
N balance (mg day-' kg-')
37 38 41
50
60 63 68
71 66 45 54
total N)
(% urinary
Urea N
N balance is corrected for estimated integumental and miscellaneous N losses, as discussed in the text. N balance results for B.M.' are excluded, as discussed in the Methods section.
s'
488
P. B. Pencharz et al.
the parents, and permission was obtained from the infants’ pediatricians. The technique used to determine nitrogen balance was a modification of that described by Hepner & Lubchenco (1960). Each nitrogenbalance period lasted 30-36 h. Plastic bags for urine were attached to the skin around the baby’s anus in order to collect each stool separately. After a stool had passed, the bag was removed and the faeces were wiped from the infant’s skin with cotton balls, which were included with the stools for analysis. Use of Carmine Red as a faecal marker facilitated the preparation of pooled samples (Fomon, 1974b). The apparent nitrogen balance was calculated as the difference between nitrogen intake and the sum of faecal and urinary nitrogen losses. We corrected this calculation for integumental and miscellaneous losses of nitrogen, as discussed below, in order to estimate the protein requirements for nitrogen equilibrium. Gestational age was assessed in two ways: from an evaluation of (a) the infant’s anthropometric characteristics (length, weight and head circumference) and (b) his external characteristics, as described by Usher et al. (1966). All babies were examined for gestational age by one of us and then independently by the baby’s pediatrician, with essentially the same results. When a discrepancy existed between the clinical assessment and the obstetrical data, the clinical evaluation was used to estimate gestational age. We measured each infant’s body length from crown to heel, ensuring full extension without pelvic tilt. All body weights were recorded daily between 07.30 and 08.00 hours. Each infant’s weight gain was calculated as the weight change over a 3 days period, including the period of nitrogen-balance measurement. Experinrental model To study dynamic aspects of whole-body nitrogen metabolism, we modified slightly the approach described by Picou & TaylorRoberts (1969), which involves the administration of [ 5N]glycine at a constant rate in order to achieve a steady state of 15N enrichment in the metabolic nitrogen pool. This method overcomes some of the difficulties (e.g. Waterlow, 1969) inherent in the single-isotope-dose method of San Pietro & Rittenberg (1953).
The Picou & Taylor-Roberts method allows estimates to be made of the nitrogen flux (Q) through the metabolic pool, i.e. the nitrogen disposal rate (Shipley & Clark, 1972), and the rates of whole-body protein synthesis (S) and breakdown (C). The assumptions inherent in this approach to the study of whole-body protein turnover have been discussed and tested experimentally (Picou & Taylor-Roberts, 1969; Steffee et al., 1976). Isotope administration We obtained [15N]glycine (95 atoms % excess) from the Stohler Chemical Corp. (Waltham, Mass., U.S.A.). The enrichment of the purchased material was confirmed by massspectrographic analysis. In all of the studies, the [15N]glycinewas administered at a known rate to supply about 0.6 mg of 15N 24 h-’ kg - l body weight. Picou & Taylor-Roberts (1969) administered the labelled amino acid by constant intravenous or intragastric infusion. In our studies with healthy young adults given an adequate diet (Steffee et al., 1976), we found that [15N]glycine given orally every 3 h for 60 h allowed a satisfactory concentration of urinary [l 5N]urea to be obtained. Furthermore, in the same study, we noted that estimates of body nitrogen flux were similar whether the subjects were given [ 5N]glycine orally or intravenously. Therefore, in this study with neonates, the [‘5NN]glycinewas mixed with the bottle feeds, which the infants received every 2 or 3 h for a total of 30-36 h. Infant B.M. received an exchange transfusion (2 volumes) with whole blood 11 h after the start of the first study. The exchange was performed for hyperbilirubinaemia, secondary to her being the recipient in a twin-transfusion syndrome. For these reasons the nitrogenbalance data from this study were excluded. However, we determined that the exchange caused a nitrogen loss of 539 mg and that this loss was mainly associated with erythrocytes. In addition, the time-course of urinary [ * 5N]urea enrichment resembled that found with the other infants. We decided therefore to include the B.M.I results on body nitrogen turnover in the present analysis. Analyses Food, urine and faeces were analysed for total nitrogen; in addition, urea and creatinines
Nitrogen metabolism in neonates were measured in the urine (Young, Taylor, Rand & Scrimshaw, 1973). After pretreating the urine with Permutit (Folin & Bell, 1917), we isolated the urinary urea nitrogen by a modification of the Conway diffusion method (Hawk, Oser & Summerson, 1954). The I5N enrichment of urinary urea was determined with the aid of a dual-collector, isotope-ratio mass spectrometer (Vacumetrics model MS11, Waltham, Mass., U.S.A.) (Steffee et al., 1976). The 15N enrichments of the administered isotope and of faeces were determined after digestion of the faecal samples (Munro & Fleck, 1969) and steam-distillation of the resulting ammonia (Rittenberg, 1946). Samples for mass-spectrographic analysis were measured in duplicate. The reproducibility of measurements of excess I5N was within 5 % ; the massspectrographic analysis has been previously described (Steffee et al., 1976). Data analysis We calculated the SNenrichment of urinary urea according to the method described by Rittenberg (1946). Enrichment was corrected for background values that were determined from urine samples collected over a 3 h period preceding each tracer study. The plateau value of 5N enrichment of urea was estimated by two methods. For the analysis of the whole-body nitrogen turnover results described below, plateau enrichment was calculated by averaging the values for the last three data points of each turnover study, provided they did not differ by more than 25%.
489
Because the urinary I5N urea enrichment, which occurred after the beginning of isotope administration, usually represented a smooth, curvilinear change, we also predicted, by the use of asymptotic regression analysis, that plateau enrichment would occur. The midpoint of each urine collection was used as the time for each sample. The least-squares method was used to fit the individual data sets to an exponential equation: y = C (l-e-kL). Totalbody nitrogen flux was calculated by estimating the asymptote, y (a) = C,as the plateau value. The fitting was carried out by means of a non-linear program of a computer program package (Dixon, 1973). The program uses a modified Gauss-Newton technique (Hartley, 1961; Jennrich & Sampson, 1968) to search until five successive iterations do not change the total error mean square by more than 0.001%. In addition, the program calculates the standard deviation of the estimates. Statistical evaluations were performed with the Student's t-test, regression and correlation analysis (Dixon & Massey, 1969).
Results Nitrogen balance Table 2 summarizes the nitrogen-balance data for the seven infants. These results were based on stool and urine collections made during a 30-36 h period of [lsN]glycine administration. All infants were in positive apparent nitrogen balance, with values ranging from +85 to +438mgday-'kg-'bodyweight
TABLE 3. An estimate of the protein content of weight gain as calculatedfrom nitrogen-balance data in six neonates Subject and study no.
Wt. gain (g day-, kg-')
N balance (mg day-' kg-I)
Protein content of weight gain
(%)
w.c.1 J.B.1 M.N., L.C.1 L.C.2 B.M., S.H.1 S.H.2 S.H.3
Mean +SD
19.2 26.1 12.6 15.9 17.3 15.6 10.9 19.1 22.0
253 40 1 260 127 379 422 270 335 384
8.2 9.6 12.9 8.0 13.6 16.9 15.5 11.0 10.9
17.6+ 4.6
314+95
11*8+3.1
P . B. Pencharz et al.
490
TABLE 4. Results of regression analysis ofprotein intake on nitrogen balance, weight gain and ihe relationship between absorbed nitrogen and retained nitrogen Equation Regression
(Y = a+ b x )
sb
r
~~
(1) (2) (3) (4) (5)
Apparent N balance ( y ) on N intake ( x ) * N balance ( y ) on N intake (x)* Apparent N balance ( y ) on N absorbed (x)* N balance ( y ) on N absorbed (x)* Weight gain (y)? on protein intake (x)j
-
y = 131+0,86x y = -145+0.86~ y = - 24.2+0.78x
0.12 012 0.050
y = -38.9+078~ y = - 8.2+ 7 . 5 ~
0.048 2.22
0.93 093 098 0.99 0.77
* x and y expressed as mg of N day-' kg-'. Apparent N balance = N intake-(urinary N+faecal N). N balance = apparent N balance - integumental N loss. t y expressed as g day-' kg-'. x expressed as g of protein day-' kg-'. 500 C
-- 400 m
O*m-
v
m'
-Es 2000 c
x=
looI
1
I
In order to estimate the protein requirements for nitrogen equilibrium, we needed to include sweat and other miscellaneous losses of nitrogen in our calculations. Because these losses have not been determined for infants, we used data obtained in adults. Based on a review of published data, the FAO/WHO (1973) Committee chose a value of 5 mg of nitrogen day-' kg-' to cover the integumental and other minor routes for nitrogen loss in young men. Assuming a surface area of 1-75 mz for a young adult subject, we can calculate the unmeasured nitrogen losses at about 200 mg/m2. Applying this figure on a surface-area basis, we have estimated the miscellaneous nitrogen losses for the seven infants (Table 2). From these calculations, mean integumental and other unmeasured nitrogen losses are estimated to be 15.2 mg day-' kg-' in premature infants. The regression analysis indicates that a mean intake of 170 mg of nitrogen day-' kg-' is required to replace urinary, faecal and other losses. Hence, a protein intake (N x 6.25) of 1.06 g day-' kg-I would be required, on average, to compensate for the major and minor routes of body loss of nitrogen. The value for the y-intercept, 145 mg day- kg-', provides a prediction of the mean total obligatory nitrogen loss for the group studied (Table 4, eqn. 2). Finally, the slope of the regression line is an estimate of the efficiency of dietary nitrogen utilization, and this value was found to be 86% for the range of protein intake given to the seven infants. We also obtained a significant positive relationship (r = 0.77; P i0.01)between dietary protein intake (g day-' kg-') and body weight gain (g day-' kg-') (Fig. 2). The regression
Nitrogen metaboilism in neonates
T
* /
24
491
with an efficiency of 78%, or only slightly less than the value derived from the regression of nitrogen balance on total intake of nitrogen (Fig. 1).
/ Whole-body nitrogen turnover
FIG.2. Relationship between protein intake and body weight gain in human neonates. The continuous line depicts the regression equation y = -8.2+7.5x, where y = weight gain (g day-' kg-') and x = protein intake (g day-' kg-I).
t
500
0
I
I
I
I
200
400
600
h b e d N (mg day-' kg-')
Fro. 3. Relationship between retained N and absorbed N in human neonates. The continuous line depicts the regression equation y = -38*9+078x, where y = N retention (mg of N day-' kg-') and x = N absorption (mg of N day-' kg-').
equation was y = -8*2+73x, where y = weight gain (g day-' kg-I) and x = protein intake (g day-I kg-I). Hence, according to these calculations, a baby must have a mean intake of 1.09 g of protein day-l kg-' for weight maintenance alone, and this value is comparable to that calculated for body nitrogen maintenance (Table 4, eqn. 5). Finally, we examined the relationship of retained nitrogen to absorbed nitrogen (Fig. 3); these values showed a significant linear correlation (r = 0.99; P < 0.001). The y-intercept provides an estimate of mean obligatory urinary N plus integumental and other minor losses; it was determined to be 38.9 mg of N day kg-' (Table 4, eqn. 4). The slope of the regression line indicates that absorbed nitrogen is utilized D
Table 5 shows the ['5N]urea enrichment during the final phase of the I5N turnover period and also shows the plateau values of enrichment based on methods described earlier. Fig. 4 depicts, for a representative case, the predicted and observed change in I5N enrichment of urinary urea during a 36 h period of I5N administration. By means of regression analysis, we estimated a plateau (asymptote) value for [15N]urea enrichment for each tracer study (Table 5). Only a small fraction of the ingested lSN was excreted in the infants' stools (Table 5). In calculating parameters of body nitrogen metabolism, we have assumed that all of the administered dose entered the metabolic pool. For one infant (L.C.'), the time-course of [15N]urea enrichment appeared to change linearly throughout and, in this case, it was impossible to predict a plateau value. There was a wide range of variability in the precision with which it was possible to predict the plateau value by regression analysis. The estimated standard deviations of the plateau ranged from 5 to 76%. We therefore decided not to use this procedure for calculation of nitrogen flux, but rather to estimate the plateau from the last three data points if they varied by no more than 25%. With this criterion, we were able to utilize data from eight of the eleven turnover studies; the mean difference in the last three data points in these eight subjects was 15% (range 3-24%). In the remaining three subjects, the differences were 95% (M.N.'), 42% (C.C.1) and 81 % (S.H.J. Table 6 summarizes the estimates for body nitrogen flux and the rates of whole-body protein synthesis and breakdown for the seven neonates. The mean flux was equivalent to 181 mg of N h-' kg-', which can be compared with the mean value of 25.7 mg of N h-I kg-' for healthy young adults (Steffee et al., 1976). In the present series of studies we did not obtain a statistically significant change in the nitrogen flux with age, although the data do suggest a trend towards lower values with increasing age. The mean rates of protein (N x 6.25) synthesis
P. B. Pencharz
492
et
al.
5. ouiput in faeces, urinary urea enrichmeni and plateau values for urea ' N , obtained by visual inspection and by non-linear regression analysis, in seven neonates given ['5N]glycine during 30-36 h
TABLE
Urine collections were usually made every 3 h ; times given are the midpoint of each collection. Plateau I5N enrichment was calculated by determination of the mean value for the last three collection periods. If the values of the final three collections differed by more than 25%, a plateau estimate was not made. The predictive method is described in the text.
Subject and study
no. W.C.,
J.B.,
M.N.,
L.C.,
L.C.2
B.M.,
U.M.2
C.C.,
S.H.,
S.Hq
S.H.3
Time of sample (h)
0 21.04 23.79 29.84 0 27.95 30.82 34.02 0 16.58 19.75 31.75 0 28.46 31.12 34.21 0 29.25 32.16 34.83 0 28.15 31.79 34.40 0 29.12 32.75 35.12 0 2408 28.26 34.25 0 18.16 22.25 33.74 0 2846 31.62 34.25 0 28.96 32.12 34.61
IOd3xI5N enrichment (atoms % excess)
10-3x Plateau [I5N]urea (atoms % excess)
I5N in faeces*
Inspected
Predicted
(%)
12.3
2 1.3+_2.7
4.0
15.1 15.2 15.5
16.1
42.7+ 31.0
0.4
0.7 5.0 3.9 7.6
-
lZ.O+ 7.2
I .2
0.1
11.8 11.7 13.7
- 0.8
1 .o
16.3 18.3 19.8 1.7 20.7 17.7 22.0 0.5 8.3 8.8 10.3 0.4 16.4 17.6 17.6
17.1
-
18.4
3 1 . 2 2 7.2
1.2
8.6
15.9+ 3.9
2.6
16.8
20.9+ 1.0
0.8
-
17.3+ 13.2
I .o
-
21.3f8.6
4.6
11.4
13.4k2.3
0.3
11.7
27.4+_8.2
0.7
1.5
9.4 9.3 13.2 1.4 6.4 9.0 11.6
- 0.8 11.3 10.0 106 0.5
12.0 11.7 13.0
* Expressed as the percentage of the administered dose that appeared in the faeces passed during the I5N tracer study.
Nitrogen metabolism in neonates
["N]Glycine administroton ( h 1
FIG.4. Time-course of I5N enrichment of urinary urea N after oral doses of [lsN]glycine given every 3 h to subject B.M.2: 0 , predicted; 0 , observed.
and breakdown were 26.3 and 23.8 g day-l kg-' respectively. These two values are approximately eight times the rates that occur in young adults (Steffee et al., 1976) and about four times the values reported by Picou 8t Taylor-Roberts (1969) for infants of about 1 year of age. The entry of nitrogen from food into the metabolic pool accounted for 6-18% (mean 12%) of the total nitrogen flux; urea nitrogen excretion constituted about 2% of the total nitrogen flux through the metabolic pool. The daily protein gain, as calculated from the nitrogen-balance data, amounted to 11.8%
493
of total body-weight gain (Table 3). Thus, from the net body protein gain and the total body protein synthesis rate, we estimate that on average the protein gain accounted for only 9.6% of total protein synthesis per day (Table 7). These data indicate that the greater proportion of protein synthesis and, thus, of amino acid utilization is associated in the neonate with a rapid turnover of existing protein, rather than with net protein synthesis. The results in Table 6 reveal considerable individual variation in estimated rates of wholebody protein synthesis and breakdown. Because dietary protein intakes varied among the infants,cwe examined the relationships between protein and energy intakes and the parameters of whole-body nitrogen metabolism. For this purpose, the parameters were expressed per unit body weight and per unit surface area; Table 8 shows the correlation matrices. No statistically significant correlation existed between protein or energy intakes and the rates of synthesis, breakdown, nitrogen flux, or the fraction of nitrogen flux that was excreted as urea. However, both protein and energy intakes were correlated significantly with synthesis and catabolism when they were expressed as fractions of whole-body nitrogen flux, except for energy intake (kJ day-' kg-') in relation to the term S/Q.
Discussion Nitrogen balance
Nitrogen-balance determinations among any age group are difficult to carry out with pre-
TABLE 6. Whole-body nitrogen flux and rates of whole-body protein synihesis and breakdown in five neonates
Subject and study no.
Whole-body N flux (Q) (g day-' kg-')
Whole-body protein synthesis ( S ) (g day-' kg-')
Whole-body protein breakdown (C) (g day-' kg-')
W.C.' J.B., L.C.1 L.C.2 B.M.1 B.M.2 S.H.2 S.H.3 Mean f s ~
4.18 3.51 3.60 3.19 6.38 3.58 4.99 5.34 435* 1.12
25-2 21.1 21.9 19.0 39.1 21.3 305 32.5 26.3f 1.0
22.1 18.2 20.0 16.3 37.6 18.0 28.0 29.8 23,8+ 7.4
P . B. Pencharz et al.
494
TABLE7. Relationship between body protein gain and whole-body protein synthesis in five neonates
Subject and study no.
Estimated protein gain* (g day-' kg-')
(Protein gainlprotein synthesis) x 100
2.26 3.08 1.88 2.04 1.84 2.25 2.60
9.0 14.6 8.6 10.7 8.6 7.4
2.28+ 0.43
9 . 6 k 2.4
W.C., J.B.1
L.C.' L.C.2 B.M.2 S.H.2 S.H., Mean ~
S
D
8.0
* 11.8'A of weight gain was estimated to be protein (Table 3); protein gain = weight gain x 0.1 18. TABLE 8 . Statistical correlations ( r ) in neonates between dietary protein and energy intake andparariieters of whole-body protein turnover expressedper unit body weight and per unit body surface area Q = N flux;
S = whole-body protein synthesis; C = whole-body protein breakdown. * P
