- Email: [email protected]
Milk protein quantity and quality in low-birth-weight infants
March 1977
356
TheJournalofPEDIATRICS
L
Milk protein quantity and quality in low-birth-weight infants IV. Effects on tyrosine and phenylalanine in plasma and urine
Well, appropriate-for-gestational age, low-birth-weight infants were divided into three gestational age groups and assigned randomly within each age group to one of five feeding regimens: pooled human milk (BM); formula 1 (F1) = 1.5 gm/dl protein, 60 parts bovine whey proteins: 40 parts bovine caseins," F~ = 3.0 gm/dl, 60:40; F3 = 1.5 gm/dl, 18:82; F4 = 3.0 gm/dl, 18:82. Plasma and urine concentrations of tyrosine and phenylalanine were far higher in the infants fed F1 to F~ especially F~ and Fo than in the infants fed BM. These findings offer further evidence for the limited capacity of the low-birth-weight infant to catabolize tyrosine. Infants fed F3 had significantly higher plasma tyrosine concentrations than infants fed F~ and those fed F4 had higher concentrations than those fed F~. Thus, increased plasma tyrosine concentrations in low-birth-weight infants are related directly both to the quantity and to the quality of the protein in their diets.
David K. Rassin, Ph.D.,* Gerald E. Gaull, M.D., Neils C,R. R~iih~i, M.D., and Kirsti Heinonen, M.D., New York, N. Y., and HelsinkL Finland
IN THIS PAPER, we report on the metabolism of phenylalanine and tyrosine in the same group of preterm infants reported on in the preceding paper ~ and in two papers published elsewhereY :~ The results give further evidence that incomplete biochemical development of the system for the metabolism of these aromatic amino acids 4 results in a limited ability to metabolize adequately protein loads that would be handled more readily in the adult. METHODS
AND
MATERIALS
T h e e x p e r i m e n t a l design, m e t h o d s , a n d c o m p o s i t i o n o f
the feeding regimens are as previously reported? -~ (See From the Department of Pediatric Research, New York State Institute for Basic Research in Mental Retardation, and Mount Sinai School of Medicine of the City University of New York, and the Children's Hospital, University of Hetsinki. Supported in part by Wyeth Laboratories, the Juselius Foundation, and the New York State Department of Mental Hygiene. *Reprint address: Department of Pediatric Research', N.Y.S. Institute for Basic Research in Mental Retardation, 1050 Forest Hill Road, Staten .Island, NY 10314. Vol. 90, No. 3, pp. 356-360
preceding paper for description of feeding regimens; human milk and four cow milk-based formulas.) RESULTS Phenylalanine in plasma and urine. The infants fed the 3.0 gm/dl protein formulas, especially the casein-predomSee related articles, pp. 348, 504, and 507.
Abbreviations used F: formula BM: pooled human milk
inant formula, had plasma phenylalanine concentrations two- to fourfold higher than those fed human milk (Fig. 1). T h e i n f a n t s f e d t h e 1.5 g m / d l p r o t e i n f o r m u l a s h a d i n t e r m e d i a t e values. T h e c o n c e n t r a t i o n s o f p h e n y l a l a n i n e in the u r i n e reflected t h o s e in t h e p l a s m a . * In a g e n e r a l *Complete tables with means _+ standard error of the mean for urine and plasma are available to interested readers directly from the authors. The variance of the mean concentrations of amino acids in the plasma and urine has not been indicated in the figures because they would become unreadable.
Volume 90
Milk protein." Phenylalanine and tyrosine
357
Number 3
PLASMA
PHENYLALANINE
PLASMA
TYROS/NE
90 o
18
80
0
16
70 14 60
o
.<
\-o
I0
50 40
8
0 6
..~s= ..-u---u "~
"'0 ....
9 ~
^.--'~o'
"'0. O"
o. ...e."
~ ""-"- O - -
". ""
~ "''" ~ ~
-0 . . . . v--.^
~~
4
.
~ " 1.5 ~ r
160:401
1.5%
30
(18:82)
9v . '~- 1.5e/= (18:82) 20
~"0 BM
I0
I
i
|
i
~
3
4 5 WEEKS
i
i
i
i
6
7
8
Fig. l. Effect of dietary regimen on the mean plasma concentration of phenylalaninc. The open circles (o) indicate p < 0.05 for the infants fed the formulas when tested for significance by Student's t test against the infants fed the pooled, banked, human milk (BM), The whey-predominant protein formulas (F,, F2) are indicated by the symbols 1.5% (60:40) and 3.0% (60:40), respectively, and the casein-predominant protein formulas (F3, F4) by the symbols 1.5%(18:82) and 3.0% (18:82). See text for details of feeding regimens. way, the concentrations of phenylalanine in the plasma and urine reflected the intake of phenylalanine (Table I), although the plasma phenylalanine concentrations of infants fed the high-protein, casein-predominant formula (F4) were disproportionately high compared to those of infants fed the high-protein, whey-predominant formula (F2). The plasma differences were noted by the first week of life. Only 2% of all plasma phenylalanine determinations of infants fed pooled human milk fell into the high-normal range (8 to 20 ~moles/dl) and not a single plasma determination fell into moderately (20 to 100 ~moles/dl) or greatly ( > 100 #moles/dl) increased ranges. Only infants fed the high-protein, casein-predominant formula had phenylalanine concentrations that were in the moderately or greatly increased range. Tyrosine in plasma and urine. The differences in plasma and urine tyrosine concentrations between infants fed pooled human milk and those fed the high-protein formulas, especially the casein-predominant formula, were even more striking than for phenylalanine (Fig. 2). These differences are compatible with the possibility that there is a limiting catabolic step distal to tyrosine, presumably at the p-hydroxyphenylpyruvic acid oxidase step? The average plasma tyrosine concentration of infants fed pooled human milk never rose above 8 /~moles/dl. In
i
|
i
I
2
3
i
i
i
i
4 5 WEEKS
i
6
7
8
Fig. 2. Effect of dietary regimen on the mean plasma concentration of tyrosine. Symbols as in Fig. 1. Table !. Tyrosine and phenylalanine intake from protein in five feeding regimens*
Pooled l F-.5% F, F.~ F, human 3.(~7~ 1.5% 3.0% milk (60.'40) (60.'40) (18:82) (18:82) Tyrosine 554t Phenylalanine 522 Total 1.076
628 681 1.309
1.258 1,353 2,611
762 753 1,515
1,522 1,516 3,038
*See preceding paper for composition of formulas, F, to F 4. tpmoles intake/kg/day.
striking contrast, the plasma tyrosine concentrations of infants fed the 3.0 gm/dl casein-predominant formula rose to a maximum four- to tenfold higher than the mean for infants fed human milk. Plasma tyrosine concentrations of infants fed the 1.5 gm/dl, whey-predominant formula were closest to those of infants fed human milk. The concentrations of tyrosine in the urine* reflected those in the plasma. It is instructive to compare the plasma tyrosine concentrations of the infants fed the casein-predominant formulas with those of the infants fed the whey-predominant formulas in order to distinguish differences due to protein quality from those due to protein quantity alone. Clearly the mean plasma tyrosine concentrations of the infants fed the casein-predominant formulas are disproportionately greater than those of the infants fed the whey-predominant formulas at the same protein concentration (compare Figs. 3, A and B, with Table I), The difference tends to disappear at about the fifth to
358
Rassin et aL
The Journal o f Pediatrics March 1977
90
80
PLASMA
40
~
PLASMATYROSINE
70
TYROSINE
1.5% (18:82)
3O Oi,**#'~ j W -1
o
2O
It,.,.,I O~
60
'-.. I0
50
~
" y " 1.5% (60z40)
o
-O
In tlJ 0 :E
I 2
40
I 4
I 6
I II
! I0
WEEKS
Fig. 3B. Effect of dietary regimen on the mean plasma concentration of tyrosine. The 1.5%(18:82) formula tested against the 1.5% (60:40) formula. The open circles (o) indicate p < 0.05.
30
20
DISTRIBUTION OF ALL TYROSINE VALUES
\
I0
3.0% 160:40) l
I
I
I
I
,!
2
4
6
8
I0
12
WEEKS
Fig. 3A. Effect of dietary regimen on the mean plasma concentration of tyrosine. The 3.0% (18:82) formula tested against the 3.0% (60:40) formula. The open circles (o) indicate p < 0.05.
60
'k
40
20 0
sixth postnatal week, which is consistent with further maturation of the p-hydroxyphenylpyruvic acid oxidase system. Differences in the frequency of high plasma tyrosine concentrations between feeding groups were even more striking for tyrosine than for phenylalanine (Fig. 4). Not a single plasma tyrosine determination in any infant fed human milk showed a concentration above our normal values and only 10% were in the high-normal range. In contrast, 22% of all plasma tyrosine concentrations in infants fed the high-protein, casein-predominant formula (F,) were in the greatly increased range ( > 100 t~moles/ dl), 23% were in the moderately increased range (20 to 100 /~moles/dl), and fewer than 10% were below 8 #moles/dl (Fig. 4). The distribution of the plasma tyrosine concentrations in the infants who received the other formulas were intermediate and directly reflected the intake of tyrosine. DISCUSSION For the entire period of study, the mean plasma and urine concentrations of phenylalanine and tyrosine were
3.0%(60:40)
1.5%(60:40)
80
80
BM
1.5%( 1 8 : 8 2 }
:5.0%(t8:82)
60
~
20
0
....9 ~ ~ t ~ , 1 0 0
/k.l/i ,10 ~ - ~ 2 [ N l ~ d l ~
,~
~ MOLES %
Fig. 4. Effect of dietary regimen on the distribution of all the individual determinations of concentration of plasma tyrosine. The determinations within each range are plotted as a percent of the total determinations for each particular feeding group: n - - 172, BM; n = 130, 1.5% (60:40); n = 127, 1.5% (18:82); n = 131, 3.0% (60:40): n = 123, 3.0% (18:82).
considerably greater in the infants fed formulas providing 4.5 gm protein/kg/day than in the infants fed pooled human milk (approximately 1.7 gm protein/kg/day). In infants fed formulas providing 2.25 protein/kg/day these concentrations were intermediate. In this regard, phenylalanine and tyrosine resembled most of the aliphatic:' and
Volume 90 Number 3
the sulfur-containing' amino acids (with the notable exception of taurine). The mean concentrations of tyrosine were far greater than those of phenylalanine and persisted for four to six weeks after birth. It is striking that the mean plasma tyrosine concentrations of infants who received the high-protein, caseinpredominant formula (F~) were considerably higher than those of the infants who received the high-protein, wheypredominant formula (FA (Fig. 3, A). Thus there is a metabolic difference which can be attributed to the quality of the protein, rather than to the quantity of protein alone. Furthermore, the difference in mean plasma tyrosine concentration is far higher than the 20% greater intake of tyrosine in the casein-predominant formula (Table I). A less striking, albeit highly significant, difference in mean plasma tyrosine concentrations attributable to protein quality was found when the two lowprotein formulas were compared to one another (Fig. 3, B). Both the severity and the duration of these metabolic sequelae in preterm infants fed high-protein, caseinpredominant formulas have been noted by others (cf reference 5 to 8). Most previous studies, however, have compared various casein-predominant formulas to one another. None has systematically compared the degree of tyrosinemia in preterm infants fed "high-protein" and "low-protein" diets to that found in preterm infants fed human milk, and in many studies the protein quality was not specified. The hyperphenylalaninemia and tyrosinemia and tyrosyluria that persists for four to six weeks may be the result of a delayed maturation ofp-hydroxyphenylpyruvic acid oxidase? The infants in the present study were supplemented with approximately 5.5/zg/kg/day of folic acid and 10,2 mg/kg/day of ascorbic acid. Perhaps massive doses of these two vitamins should have been given. Systematically controlled studies of folate ~ and of ascorbate,,O. 1, however, suggest that the tyrosinemia and tyrosyluria found in preterm infants fed high-protein diets is not corrected by massive doses of either. In any case, there is no evidence that such megavitamin therapy would affect the high concentrations of other amino acids. In our opinion, the megavitamin approach in this situation can be likened to wearing a life jacket in order to go swimming with ones shoes on. Plasma concentrations of phenylalanine as high as those found in phenylketonuria were observed in a few infants fed the high-protein, casein-predominant formula (F,). The mean concentrations of plasma tyrosine, however, were far greater than those of phenylalanine, and the highest concentrations of tyrosine persisted for four to six weeks. Some individual infants fed formula F,
Milk protein: Phenylalanine and tyrosine
359
had plasma tyrosine concentrations that were 30 times the maximum found in infants fed pooled human milk. Furthermore, a five- to tenfold increase over "basal excretion" of urinary tyrosine and its metabolites phydroxyphenyllactic and p-hydroxyphenylpyruvic acid has been reported in infants fed a diet essentially the same as the 1.5 gm/dl (18:82) formula. 12 The fact that the concentrations of tyrosine were higher than those of phenylalanine is compatible with the known presence of phenylalanine hydroxylase in immature human liver. 13 In a number of studies an attempt has been made to relate later intellectual or neurologic impairment of children who were born prematurely to the feeding of a highprotein diet in early infancy. Most of these studies presume that any such impairment would be mediated by tyrosine, both because of the high-tyrosine concentrations found and because of the structural relationship of tyrosine and its acidic metabolites to phenylalanine and its acidic metabolites. In light of the extent and severity of other imbalances in amino acid metabolism that accompany tyrosinemia and tyrosyluria, 1-:' however, it seems advisable to consider that other metabolic deviations may also have untoward effects rather than just those most readily measured, i.e.. tyrosinemia and phenylalaninemia. The multiplicity of pathogenetic possibilities to explain brain dysfunction associated with disordered amino acid metabolism has been reviewed critically. TM Definitive answers to the question of the role of highprotein intakes and the resultant metabolic imbalances on brain development are not available, but the following should be pointed out: (1) With one exception, '~ all the studies of the potential neurologic sequelae of highprotein diets in preterm infants have been retrospective and thereby limited. (2) None of these studies has systematically compared the various quantities and qualities of protein currently included in milk formulas for the human infants, at least for the term human infant. The data presented here suggest that some so-called lowprotein diets are not really "low" and that quality (composition of the protein) as well as quantity of protein must be taken into account in the evaluation of the dietary intake of the young infant. We suggest that the definition of tyrosinemia, and of other aminoacidemias perhaps, should be revised sharply downward. For example, compare the definitions of tyrosinemia as > 15 mg (82 ~moles/dl)? > 20 mg/dl (110 #moles/dl)." or > 10 mg/dl (55/~moles/dl) TM with the fact that not a single determination in any infant fed human milk in this study wa s greater than 20 #moles/dl (Fig. 4) and that the average tyrosine concentration of such infants was 6 to 9 #moles/dl. Before an etiologic role for
360
Rassin el al.
tyrosinemia and high-protein formulas in later neurologic impairment can be dismissed, the control group should be fed a formula closer in quantity and quality to human milk. In the only prospective study of neurologic outcome '7 neither tyrosine nor any other amino acid was measured, but preterm infants (birth weight < 1,700 gm) fed a 2% casein-predominant formula were compared to infants fed a 4% casein-predominant formula. An increased incidence of low IQ scores and of strabismus was found at three to five years of age in the group who had received the 4% protein diet. The intellectual performance of children who were born at term and had transient neonatal tyrosinemia on high-protein intakes was also reported recently. TM Although retrospective, the last study suggests that the term infant as well as the preterm infant, may be neurologically vulnerable. It would be of interest to compare later intellectual performance in children, who as preterm infants were fed human milk, with children, who as preterm infants were fed various highprotein formulas. It is our hope that the present study will provoke others to examine further the question of the requirements for the quantity and quality of the protein in the diet of the preterm infant and to include in the variables to be measured the adequacy of subsequent intellectual performance. We shall attempt to secure follow-up examinations of the infants in this s'tudy, although the number may be too small to permit firm conclusions. We are grateful for the expert assistance with the amino acid analysis and data analysis of Ms. kucille Donadio. We are grateful to Dr. Rudy Tomarelli, Mr. Rex Stribley, and Mr. Elmer Eckhardt of Wyeth Laboratories both for the preparation and for some analyses of the nutritional constitutents of the formulas. The authors are indebted particularly to Professors Sebastian B. Littauer (Emeritus) and Edward J. Ignall of Operations Research, Columbia University who gave considerable statistical advice during the planning, the early phases of data gathering, and the preparation of the manuscript. REFERENCES
1. Gaull GE, Rassin DK, Rfiiha NCR, and Heinonen K: Milk protein quantity and quality in low-birth-weight infants. III. Effects on sulfur amino acids in plasma and urine, J PEDIATR90:348, 1977. 2. Rahiha NCR, Heinonen K, Rassin DK, and Gaull GE: Milk protein quantity and quality in low-birth-weight infants: I.
The Journal of Pediatrics March 1977
3.
4.
5.
6. 7.
8.
9.
10.
ll.
12.
lY 14.
15.
16. 17.
18.
Metabolic responses and effects on growth, Pediatrics 57:659, 1976. Rassin DK, Gaull GE, Heinonen K, and R/iih/i NCR: Milk protein quantity and quality in low-birth-weight infants: II. Effects on selected aliphatic amino acids in plasma and urine, Pediatrics 59:407. 1977. Kretchmer N, Levine SA, and McNamara H: The in vitro metabolism of tyrosine and its intermediates in the liver of the premature infant, Am J Dis Child 93:19. 1957. Levine S, Marples E, and Gordon H: A defect in the metabolism of tyrosine and phenylalanine in premature infants: I. Identification and assay of intermediary products, J Clin Invest 20:199, 1941. Mathews J, and Partington NW: The plasma tyrosine levels of premature babies, Arch Dis Child 39:271, 1964. Avery ME, Clow CL, Menkes JH, Ramos A, Scriver CR, Stern L, and Wasserman BP: Transient tyrosinemia of the newborn, dietary and clinical aspects, Pediatrics 39:378, 1967. LaDu BN, and Gjessing LR: Tyrosinosis and tyrosinemia, in Stanbury JB, Wyngaarden JB, and Fredrickson DS, editors: The metabolic basis of inherited disease, ed 3, New York, 1972, McGraw-Hill Book Company, Ine, p 296. Partington MW, and Mathews J: The prophylactic use of folic acid in neonatal hypertyrosinemia, Pediatrics 39:776, 1967. Bakker HD, Wadman SK, vanSprang F J, van Der Heiden C, Ketting D, and DeBree PK: Tyrosinemia and tyrosyluria in healthy prematures: Time courses not vitamin C-dependent, Clin Chim Acta 61:73, 1975. Mohanram M, and Kumar A: Ascorbic acid and tyrosine metabolism in preterm and small-for-dates infants, Arch Dis Child 50:235, 1975. Fernbach SA, Summons RE, Pereira WE, and Duflield AM: Metabolic studies of transient tyrosinemia in premature infants, Pediatr Res 9:172, 1975. R~iih~i NCR: Phenylalanine hydroxylase in human liver during development, Pediatr Res 7:1, 1973. Gaull GE, Tallan HH, Lajtha A, and Rassin DK: Pathogenesis of brain dysfunction in inborn errors of metabolism, in Gaull GE, editor: Biology of brain dysfunction, vol 3, New York, 1975, Plenum Publishing Corp., p 47. Menkes JH, Welcher DW, Leir HS, Dallas J, and Gretsky NE: Relationship of elevated blood tyrosine to the ultimate intellectual performance of premature infants, Pediatrics 49:218, 1972. Partington MW: Neonatal tyrosinemia, Biol Neonate 12:316, 1968. Goldman HI, Goldman JS, Kaufman I, and Liebman OB: Late effects of early dietary protein intake on low-birthweight infants, J PEDIATR85:764, 1974. Mamunes P, Prince PE, Thornton NH, Hunt PS, and Hitchcock ES: Intellectual deficits after transient tyrosinemia in the term neonate, Pediatrics 57:675, 1976.
356
TheJournalofPEDIATRICS
L
Milk protein quantity and quality in low-birth-weight infants IV. Effects on tyrosine and phenylalanine in plasma and urine
Well, appropriate-for-gestational age, low-birth-weight infants were divided into three gestational age groups and assigned randomly within each age group to one of five feeding regimens: pooled human milk (BM); formula 1 (F1) = 1.5 gm/dl protein, 60 parts bovine whey proteins: 40 parts bovine caseins," F~ = 3.0 gm/dl, 60:40; F3 = 1.5 gm/dl, 18:82; F4 = 3.0 gm/dl, 18:82. Plasma and urine concentrations of tyrosine and phenylalanine were far higher in the infants fed F1 to F~ especially F~ and Fo than in the infants fed BM. These findings offer further evidence for the limited capacity of the low-birth-weight infant to catabolize tyrosine. Infants fed F3 had significantly higher plasma tyrosine concentrations than infants fed F~ and those fed F4 had higher concentrations than those fed F~. Thus, increased plasma tyrosine concentrations in low-birth-weight infants are related directly both to the quantity and to the quality of the protein in their diets.
David K. Rassin, Ph.D.,* Gerald E. Gaull, M.D., Neils C,R. R~iih~i, M.D., and Kirsti Heinonen, M.D., New York, N. Y., and HelsinkL Finland
IN THIS PAPER, we report on the metabolism of phenylalanine and tyrosine in the same group of preterm infants reported on in the preceding paper ~ and in two papers published elsewhereY :~ The results give further evidence that incomplete biochemical development of the system for the metabolism of these aromatic amino acids 4 results in a limited ability to metabolize adequately protein loads that would be handled more readily in the adult. METHODS
AND
MATERIALS
T h e e x p e r i m e n t a l design, m e t h o d s , a n d c o m p o s i t i o n o f
the feeding regimens are as previously reported? -~ (See From the Department of Pediatric Research, New York State Institute for Basic Research in Mental Retardation, and Mount Sinai School of Medicine of the City University of New York, and the Children's Hospital, University of Hetsinki. Supported in part by Wyeth Laboratories, the Juselius Foundation, and the New York State Department of Mental Hygiene. *Reprint address: Department of Pediatric Research', N.Y.S. Institute for Basic Research in Mental Retardation, 1050 Forest Hill Road, Staten .Island, NY 10314. Vol. 90, No. 3, pp. 356-360
preceding paper for description of feeding regimens; human milk and four cow milk-based formulas.) RESULTS Phenylalanine in plasma and urine. The infants fed the 3.0 gm/dl protein formulas, especially the casein-predomSee related articles, pp. 348, 504, and 507.
Abbreviations used F: formula BM: pooled human milk
inant formula, had plasma phenylalanine concentrations two- to fourfold higher than those fed human milk (Fig. 1). T h e i n f a n t s f e d t h e 1.5 g m / d l p r o t e i n f o r m u l a s h a d i n t e r m e d i a t e values. T h e c o n c e n t r a t i o n s o f p h e n y l a l a n i n e in the u r i n e reflected t h o s e in t h e p l a s m a . * In a g e n e r a l *Complete tables with means _+ standard error of the mean for urine and plasma are available to interested readers directly from the authors. The variance of the mean concentrations of amino acids in the plasma and urine has not been indicated in the figures because they would become unreadable.
Volume 90
Milk protein." Phenylalanine and tyrosine
357
Number 3
PLASMA
PHENYLALANINE
PLASMA
TYROS/NE
90 o
18
80
0
16
70 14 60
o
.<
\-o
I0
50 40
8
0 6
..~s= ..-u---u "~
"'0 ....
9 ~
^.--'~o'
"'0. O"
o. ...e."
~ ""-"- O - -
". ""
~ "''" ~ ~
-0 . . . . v--.^
~~
4
.
~ " 1.5 ~ r
160:401
1.5%
30
(18:82)
9v . '~- 1.5e/= (18:82) 20
~"0 BM
I0
I
i
|
i
~
3
4 5 WEEKS
i
i
i
i
6
7
8
Fig. l. Effect of dietary regimen on the mean plasma concentration of phenylalaninc. The open circles (o) indicate p < 0.05 for the infants fed the formulas when tested for significance by Student's t test against the infants fed the pooled, banked, human milk (BM), The whey-predominant protein formulas (F,, F2) are indicated by the symbols 1.5% (60:40) and 3.0% (60:40), respectively, and the casein-predominant protein formulas (F3, F4) by the symbols 1.5%(18:82) and 3.0% (18:82). See text for details of feeding regimens. way, the concentrations of phenylalanine in the plasma and urine reflected the intake of phenylalanine (Table I), although the plasma phenylalanine concentrations of infants fed the high-protein, casein-predominant formula (F4) were disproportionately high compared to those of infants fed the high-protein, whey-predominant formula (F2). The plasma differences were noted by the first week of life. Only 2% of all plasma phenylalanine determinations of infants fed pooled human milk fell into the high-normal range (8 to 20 ~moles/dl) and not a single plasma determination fell into moderately (20 to 100 ~moles/dl) or greatly ( > 100 #moles/dl) increased ranges. Only infants fed the high-protein, casein-predominant formula had phenylalanine concentrations that were in the moderately or greatly increased range. Tyrosine in plasma and urine. The differences in plasma and urine tyrosine concentrations between infants fed pooled human milk and those fed the high-protein formulas, especially the casein-predominant formula, were even more striking than for phenylalanine (Fig. 2). These differences are compatible with the possibility that there is a limiting catabolic step distal to tyrosine, presumably at the p-hydroxyphenylpyruvic acid oxidase step? The average plasma tyrosine concentration of infants fed pooled human milk never rose above 8 /~moles/dl. In
i
|
i
I
2
3
i
i
i
i
4 5 WEEKS
i
6
7
8
Fig. 2. Effect of dietary regimen on the mean plasma concentration of tyrosine. Symbols as in Fig. 1. Table !. Tyrosine and phenylalanine intake from protein in five feeding regimens*
Pooled l F-.5% F, F.~ F, human 3.(~7~ 1.5% 3.0% milk (60.'40) (60.'40) (18:82) (18:82) Tyrosine 554t Phenylalanine 522 Total 1.076
628 681 1.309
1.258 1,353 2,611
762 753 1,515
1,522 1,516 3,038
*See preceding paper for composition of formulas, F, to F 4. tpmoles intake/kg/day.
striking contrast, the plasma tyrosine concentrations of infants fed the 3.0 gm/dl casein-predominant formula rose to a maximum four- to tenfold higher than the mean for infants fed human milk. Plasma tyrosine concentrations of infants fed the 1.5 gm/dl, whey-predominant formula were closest to those of infants fed human milk. The concentrations of tyrosine in the urine* reflected those in the plasma. It is instructive to compare the plasma tyrosine concentrations of the infants fed the casein-predominant formulas with those of the infants fed the whey-predominant formulas in order to distinguish differences due to protein quality from those due to protein quantity alone. Clearly the mean plasma tyrosine concentrations of the infants fed the casein-predominant formulas are disproportionately greater than those of the infants fed the whey-predominant formulas at the same protein concentration (compare Figs. 3, A and B, with Table I), The difference tends to disappear at about the fifth to
358
Rassin et aL
The Journal o f Pediatrics March 1977
90
80
PLASMA
40
~
PLASMATYROSINE
70
TYROSINE
1.5% (18:82)
3O Oi,**#'~ j W -1
o
2O
It,.,.,I O~
60
'-.. I0
50
~
" y " 1.5% (60z40)
o
-O
In tlJ 0 :E
I 2
40
I 4
I 6
I II
! I0
WEEKS
Fig. 3B. Effect of dietary regimen on the mean plasma concentration of tyrosine. The 1.5%(18:82) formula tested against the 1.5% (60:40) formula. The open circles (o) indicate p < 0.05.
30
20
DISTRIBUTION OF ALL TYROSINE VALUES
\
I0
3.0% 160:40) l
I
I
I
I
,!
2
4
6
8
I0
12
WEEKS
Fig. 3A. Effect of dietary regimen on the mean plasma concentration of tyrosine. The 3.0% (18:82) formula tested against the 3.0% (60:40) formula. The open circles (o) indicate p < 0.05.
60
'k
40
20 0
sixth postnatal week, which is consistent with further maturation of the p-hydroxyphenylpyruvic acid oxidase system. Differences in the frequency of high plasma tyrosine concentrations between feeding groups were even more striking for tyrosine than for phenylalanine (Fig. 4). Not a single plasma tyrosine determination in any infant fed human milk showed a concentration above our normal values and only 10% were in the high-normal range. In contrast, 22% of all plasma tyrosine concentrations in infants fed the high-protein, casein-predominant formula (F,) were in the greatly increased range ( > 100 t~moles/ dl), 23% were in the moderately increased range (20 to 100 /~moles/dl), and fewer than 10% were below 8 #moles/dl (Fig. 4). The distribution of the plasma tyrosine concentrations in the infants who received the other formulas were intermediate and directly reflected the intake of tyrosine. DISCUSSION For the entire period of study, the mean plasma and urine concentrations of phenylalanine and tyrosine were
3.0%(60:40)
1.5%(60:40)
80
80
BM
1.5%( 1 8 : 8 2 }
:5.0%(t8:82)
60
~
20
0
....9 ~ ~ t ~ , 1 0 0
/k.l/i ,10 ~ - ~ 2 [ N l ~ d l ~
,~
~ MOLES %
Fig. 4. Effect of dietary regimen on the distribution of all the individual determinations of concentration of plasma tyrosine. The determinations within each range are plotted as a percent of the total determinations for each particular feeding group: n - - 172, BM; n = 130, 1.5% (60:40); n = 127, 1.5% (18:82); n = 131, 3.0% (60:40): n = 123, 3.0% (18:82).
considerably greater in the infants fed formulas providing 4.5 gm protein/kg/day than in the infants fed pooled human milk (approximately 1.7 gm protein/kg/day). In infants fed formulas providing 2.25 protein/kg/day these concentrations were intermediate. In this regard, phenylalanine and tyrosine resembled most of the aliphatic:' and
Volume 90 Number 3
the sulfur-containing' amino acids (with the notable exception of taurine). The mean concentrations of tyrosine were far greater than those of phenylalanine and persisted for four to six weeks after birth. It is striking that the mean plasma tyrosine concentrations of infants who received the high-protein, caseinpredominant formula (F~) were considerably higher than those of the infants who received the high-protein, wheypredominant formula (FA (Fig. 3, A). Thus there is a metabolic difference which can be attributed to the quality of the protein, rather than to the quantity of protein alone. Furthermore, the difference in mean plasma tyrosine concentration is far higher than the 20% greater intake of tyrosine in the casein-predominant formula (Table I). A less striking, albeit highly significant, difference in mean plasma tyrosine concentrations attributable to protein quality was found when the two lowprotein formulas were compared to one another (Fig. 3, B). Both the severity and the duration of these metabolic sequelae in preterm infants fed high-protein, caseinpredominant formulas have been noted by others (cf reference 5 to 8). Most previous studies, however, have compared various casein-predominant formulas to one another. None has systematically compared the degree of tyrosinemia in preterm infants fed "high-protein" and "low-protein" diets to that found in preterm infants fed human milk, and in many studies the protein quality was not specified. The hyperphenylalaninemia and tyrosinemia and tyrosyluria that persists for four to six weeks may be the result of a delayed maturation ofp-hydroxyphenylpyruvic acid oxidase? The infants in the present study were supplemented with approximately 5.5/zg/kg/day of folic acid and 10,2 mg/kg/day of ascorbic acid. Perhaps massive doses of these two vitamins should have been given. Systematically controlled studies of folate ~ and of ascorbate,,O. 1, however, suggest that the tyrosinemia and tyrosyluria found in preterm infants fed high-protein diets is not corrected by massive doses of either. In any case, there is no evidence that such megavitamin therapy would affect the high concentrations of other amino acids. In our opinion, the megavitamin approach in this situation can be likened to wearing a life jacket in order to go swimming with ones shoes on. Plasma concentrations of phenylalanine as high as those found in phenylketonuria were observed in a few infants fed the high-protein, casein-predominant formula (F,). The mean concentrations of plasma tyrosine, however, were far greater than those of phenylalanine, and the highest concentrations of tyrosine persisted for four to six weeks. Some individual infants fed formula F,
Milk protein: Phenylalanine and tyrosine
359
had plasma tyrosine concentrations that were 30 times the maximum found in infants fed pooled human milk. Furthermore, a five- to tenfold increase over "basal excretion" of urinary tyrosine and its metabolites phydroxyphenyllactic and p-hydroxyphenylpyruvic acid has been reported in infants fed a diet essentially the same as the 1.5 gm/dl (18:82) formula. 12 The fact that the concentrations of tyrosine were higher than those of phenylalanine is compatible with the known presence of phenylalanine hydroxylase in immature human liver. 13 In a number of studies an attempt has been made to relate later intellectual or neurologic impairment of children who were born prematurely to the feeding of a highprotein diet in early infancy. Most of these studies presume that any such impairment would be mediated by tyrosine, both because of the high-tyrosine concentrations found and because of the structural relationship of tyrosine and its acidic metabolites to phenylalanine and its acidic metabolites. In light of the extent and severity of other imbalances in amino acid metabolism that accompany tyrosinemia and tyrosyluria, 1-:' however, it seems advisable to consider that other metabolic deviations may also have untoward effects rather than just those most readily measured, i.e.. tyrosinemia and phenylalaninemia. The multiplicity of pathogenetic possibilities to explain brain dysfunction associated with disordered amino acid metabolism has been reviewed critically. TM Definitive answers to the question of the role of highprotein intakes and the resultant metabolic imbalances on brain development are not available, but the following should be pointed out: (1) With one exception, '~ all the studies of the potential neurologic sequelae of highprotein diets in preterm infants have been retrospective and thereby limited. (2) None of these studies has systematically compared the various quantities and qualities of protein currently included in milk formulas for the human infants, at least for the term human infant. The data presented here suggest that some so-called lowprotein diets are not really "low" and that quality (composition of the protein) as well as quantity of protein must be taken into account in the evaluation of the dietary intake of the young infant. We suggest that the definition of tyrosinemia, and of other aminoacidemias perhaps, should be revised sharply downward. For example, compare the definitions of tyrosinemia as > 15 mg (82 ~moles/dl)? > 20 mg/dl (110 #moles/dl)." or > 10 mg/dl (55/~moles/dl) TM with the fact that not a single determination in any infant fed human milk in this study wa s greater than 20 #moles/dl (Fig. 4) and that the average tyrosine concentration of such infants was 6 to 9 #moles/dl. Before an etiologic role for
360
Rassin el al.
tyrosinemia and high-protein formulas in later neurologic impairment can be dismissed, the control group should be fed a formula closer in quantity and quality to human milk. In the only prospective study of neurologic outcome '7 neither tyrosine nor any other amino acid was measured, but preterm infants (birth weight < 1,700 gm) fed a 2% casein-predominant formula were compared to infants fed a 4% casein-predominant formula. An increased incidence of low IQ scores and of strabismus was found at three to five years of age in the group who had received the 4% protein diet. The intellectual performance of children who were born at term and had transient neonatal tyrosinemia on high-protein intakes was also reported recently. TM Although retrospective, the last study suggests that the term infant as well as the preterm infant, may be neurologically vulnerable. It would be of interest to compare later intellectual performance in children, who as preterm infants were fed human milk, with children, who as preterm infants were fed various highprotein formulas. It is our hope that the present study will provoke others to examine further the question of the requirements for the quantity and quality of the protein in the diet of the preterm infant and to include in the variables to be measured the adequacy of subsequent intellectual performance. We shall attempt to secure follow-up examinations of the infants in this s'tudy, although the number may be too small to permit firm conclusions. We are grateful for the expert assistance with the amino acid analysis and data analysis of Ms. kucille Donadio. We are grateful to Dr. Rudy Tomarelli, Mr. Rex Stribley, and Mr. Elmer Eckhardt of Wyeth Laboratories both for the preparation and for some analyses of the nutritional constitutents of the formulas. The authors are indebted particularly to Professors Sebastian B. Littauer (Emeritus) and Edward J. Ignall of Operations Research, Columbia University who gave considerable statistical advice during the planning, the early phases of data gathering, and the preparation of the manuscript. REFERENCES
1. Gaull GE, Rassin DK, Rfiiha NCR, and Heinonen K: Milk protein quantity and quality in low-birth-weight infants. III. Effects on sulfur amino acids in plasma and urine, J PEDIATR90:348, 1977. 2. Rahiha NCR, Heinonen K, Rassin DK, and Gaull GE: Milk protein quantity and quality in low-birth-weight infants: I.
The Journal of Pediatrics March 1977
3.
4.
5.
6. 7.
8.
9.
10.
ll.
12.
lY 14.
15.
16. 17.
18.
Metabolic responses and effects on growth, Pediatrics 57:659, 1976. Rassin DK, Gaull GE, Heinonen K, and R/iih/i NCR: Milk protein quantity and quality in low-birth-weight infants: II. Effects on selected aliphatic amino acids in plasma and urine, Pediatrics 59:407. 1977. Kretchmer N, Levine SA, and McNamara H: The in vitro metabolism of tyrosine and its intermediates in the liver of the premature infant, Am J Dis Child 93:19. 1957. Levine S, Marples E, and Gordon H: A defect in the metabolism of tyrosine and phenylalanine in premature infants: I. Identification and assay of intermediary products, J Clin Invest 20:199, 1941. Mathews J, and Partington NW: The plasma tyrosine levels of premature babies, Arch Dis Child 39:271, 1964. Avery ME, Clow CL, Menkes JH, Ramos A, Scriver CR, Stern L, and Wasserman BP: Transient tyrosinemia of the newborn, dietary and clinical aspects, Pediatrics 39:378, 1967. LaDu BN, and Gjessing LR: Tyrosinosis and tyrosinemia, in Stanbury JB, Wyngaarden JB, and Fredrickson DS, editors: The metabolic basis of inherited disease, ed 3, New York, 1972, McGraw-Hill Book Company, Ine, p 296. Partington MW, and Mathews J: The prophylactic use of folic acid in neonatal hypertyrosinemia, Pediatrics 39:776, 1967. Bakker HD, Wadman SK, vanSprang F J, van Der Heiden C, Ketting D, and DeBree PK: Tyrosinemia and tyrosyluria in healthy prematures: Time courses not vitamin C-dependent, Clin Chim Acta 61:73, 1975. Mohanram M, and Kumar A: Ascorbic acid and tyrosine metabolism in preterm and small-for-dates infants, Arch Dis Child 50:235, 1975. Fernbach SA, Summons RE, Pereira WE, and Duflield AM: Metabolic studies of transient tyrosinemia in premature infants, Pediatr Res 9:172, 1975. R~iih~i NCR: Phenylalanine hydroxylase in human liver during development, Pediatr Res 7:1, 1973. Gaull GE, Tallan HH, Lajtha A, and Rassin DK: Pathogenesis of brain dysfunction in inborn errors of metabolism, in Gaull GE, editor: Biology of brain dysfunction, vol 3, New York, 1975, Plenum Publishing Corp., p 47. Menkes JH, Welcher DW, Leir HS, Dallas J, and Gretsky NE: Relationship of elevated blood tyrosine to the ultimate intellectual performance of premature infants, Pediatrics 49:218, 1972. Partington MW: Neonatal tyrosinemia, Biol Neonate 12:316, 1968. Goldman HI, Goldman JS, Kaufman I, and Liebman OB: Late effects of early dietary protein intake on low-birthweight infants, J PEDIATR85:764, 1974. Mamunes P, Prince PE, Thornton NH, Hunt PS, and Hitchcock ES: Intellectual deficits after transient tyrosinemia in the term neonate, Pediatrics 57:675, 1976.