Acta PEdiatr Scand 79: 729-736, 1990
Concentration of Twelve Plasma Proteins at Birth in Very Low Birthweight and in Term Infants STAFFAN K. T. POLBERGER,' GORAN FEX2 and NIELSC. R. RAIHA' From the Departments of IPaediatrics and 2Clinical Chemistry, University of Lund, Malmo General Hospital, Sweden
ABSTRACT. Polberger, S. K. T., Fex, G. and Riiiha, N. C. R. (Departments of Paediatrics and Clinical Chemistry, University of Lund, Malmo General Hospital, Sweden). Concentrations of twelve plasma proteins at birth in very low birthweight and in term infants. Acta Paediatr Scand 79: 729, 1990. Plasma samples obtained at birth from 70 very low birth weight (VLBW) infants (gestational age 24 to 34 weeks) and from 20 term infants were analysed for concentrations of 12 different proteins. The plasma concentrations of albumin, transthyretin (TTR), retinol-binding protein (RBP), vitamin D-binding protein, apolipoprotein A I, fibronectin, orosomucoid and a,-antichymotrypsin were significantly lower in the VLBW infants than in the term infants, whereas the values of a-fetoprotein (AFP) were significantly higher in the VLBW infants. No differences were found between the two groups for apolipoprotein A 11, apolipoprotein B and transferrin. Birth asphyxia and sex had no influence on the measured plasma protein concentrations. The plasma concentrations of apolipoprotein A I and A I1 were significantly lower in small-forgestational age (SGA), VLBW infants compared with appropriate-for-gestational age (AGA), VLBW infants. Possible acute inflammation (defined as raised concentrations of orosomucoid or a,-antichymotrypsin) was associated with significantly higher values of vitamin D-binding protein in both VLBW and term infants, suggesting that this protein may act as an acute phase protein in newborn infants. Key words:plasma proteins, electroimmunoassay, VLBW infants, term infants.
It has long been recognised that fetal plasma protein synthesis is induced at different gestational ages for different proteins (1). During pregnancy, the concentrations of most plasma proteins increase in the fetus with a few exceptions (e.g. a2macroglobulin and AFP). Since the studies by Gitlin et al. (2), several reports have presented data on plasma proteins in newborn infants, however, most of these studies were performed on cord blood from term infants only. For most of the plasma proteins studied, there are few data on concentrations in VLBW infants at birth, especially subdivided data considering various gestational ages. As part of a nutritional study (3), we have measured concentrations at birth of 12 different plasma proteins in term and VLBW infants, the latter group of infants all having their gestational ages assessed by an ultrasound fetometry in early pregnancy. MATERIAL AND METHODS In a prospective study at the neonatal units in Malmo and Lund, the concentrations of 12 plasma (EDTA) proteins were determined in 70 VLBW infants and 20 term infants. Ten of the plasma proteins were selected to serve as possible indicators of nutritional status in VLBW infants subsequently fed varying protein and energy intakes, while orosomucoid and a,-antichymotrypsin were chosen as indicators of possible inflammatory reaction. In all infants, the gestational ages were calculated from an early ultrasound fetometry performed in weeks 16 to 18. This gave a reliable estimation of gestational age of 2 3 days (SD) expressed as completed gestational weeks (4). Blood was drawn from routine indwelling umbilical catheters (artery or vein) in the VLBW infants as soon as installed (within 2 hours from birth)
730 S. Polberger et al.
Acta Paediatr Scand 79
number 25
I
15 10
5 gestational weeks
0
24
26
Fig. 1. The gestational age distribution of the VLBW infants ( n = 70).
and always before blood or plasma transfusions were given and before enteral feeding was initiated. Before blood sampling 1.5-2 ml of blood was withdrawn from the catheter to avoid dilution with infusion fluids. Cord blood was obtained from the term infants immediately after delivery. Plasma (250-350 pl) was isolated and dispensed in small 600 p1 plastic tubes and stored at -70°C for 2-6 months before analysis. Samples from two groups of infants were studied: 1 . Term, AGA infants (n=20) gestational age: mean ? SD, (range) 40? 1 (38-4 ) weeks, birthweight: 3 590'. 500 (2 560-4 650) g. 2. VLBW infants (n=70), gestational age: 28?2 (24-34) weeks, birth weight: 1 l o t 2 4 0 (600-1 500) g, with the subgroups:
a. AGA infants < 29 weeks ( n = 3 l), b. AGA infants 2 29 weeks ( n =26) and c. SGA infants 2 2 9 weeks 13) (defined as birthweight less than - 2 S D according to Swedish growth charts (5)).
(n=
The distribution of gestational age of the 70 VLBW infants is presented in Fig. 1 . Most infants had a gestational age of 24 t o 30 weeks; only 9 infants were 2 3 1 weeks at birth. All infants below 29 weeks of gestation (n=31) were AGA and therefore we included only infants with a gestational age of 2 2 9 weeks in the comparison between AGA and SGA infants. No infant in the term group (n=20) had Apgar scores below 7 at 1 , 5, or 10 minutes, whereas 38 infants in the VLBW group ( n = 6 8 ) had Apgar scores < 7 at 1 minute, 14 at 5 minutes, and 4 at 10 minutes of age. Apgar scores were not obtained in 2 infants in the VLBW group. Determination of plasma proteins. The concentrations of all plasma proteins were determined in a total volume of only 100 pI of plasma, using electroimmunoassay (6). Plasma was usually diluted with 0.075 M barbital- 1.5 m M NaN, buffer pH 8.6. For analysis of fibronectin a second dilution with the same buffer containing 20 m M EDTA was performed and for analysis of apolipoproteins A1 and A I1 plasma was diluted with 8 M urea. The agarose used was Seakem from Marine Colloids (Locklands, ME) except for the assay of vitamin D-binding protein and fibronectin (agarose from Miles Seravac). Antisera t o all proteins except T T R (which was a goat antiserum) were raised in rabbits using the purified proteins as antigens. Seronorm" protein batch No. 103 (Nyegaard & Co, Oslo, Norway) was used as calibration standard for most of the proteins. For vitamin D-binding protein we used a standard from Behringwerke and for apolipoproteins A1 and B a standard was obtained from professor G Bondjers, Department of Internal Medicine, Sahlgrenska Hospital, Goteborg. For a,-antichymotrypsin and AFP we used the purified proteins as calibration standards. Pooled plasma from 20 term infants in two dilutions was used as an internal control. As no reliable standards were available for apolipoprotein A11 and fibronectin, the presented data on these proteins are expressed as per cent of an adult reference population (apolipoprotein A 11) or as percentage of the values in term infants (fibronectin). Statistical methods. Concentrations of most of the plasma proteins are expressed as arithmetic means and standard deviations (SD) except for orosomucoid, a,-antichymotrypsin and AFP, where the distributions showed pronounced skewness. In these cases, the average is presented as the median and the variation is described by means of a 95% tolerance interval obtained through applying the usual procedure t o the logarithms of the concentrations, and then taking antilogarithms. Significance of differences between groups were analysed by the
Plasma proteins in VLBW infants 73 1
Acta Paediatr Scand 79
Mann-Whitney U-test and correlations by Spearman's rank-correlation coefficient (r,?).This study was approved by the Ethics Committee at the University of Lund.
RESULTS Concentrations of ten plasma proteins in the VLBW and the term infants are given in Table 1. Except for apolipoprotein B and transferrin, the means and medians were significantly lower in the VLBW group compared with the term infants. Two more plasma proteins were studied in some of the infants. The concentration of apolipoprotein A I1 in the VLBW group ( n= 38) was 28 ? 9 9'0 (mean 2 SD) and in the term group (n=20) 27 2 1 1 9'0 @=0.70, non significant). The concentrations of AFP was analysed in 37 VLBW infants and 20 term infants. The medians (tolerance interval) were 320 (68-990) mg/l and 33 (10-120) mg/l, respectively. The difference between the two groups is highly significant (p 41 2 mg/l and/or a,-antichymotrypsin > 493 mg/l) with
Table 1. Concentrations of plasma proteins in VLBW and in term infants
Albumin TTR RBP Vit D-binding protein Apolipoprotein A I Apolipoprotein B Transferrin" Fibronectin Orosomucoid a,-antichymotrypsina
Unit
VLBW (n=70)
Term (n=20)
p
g/l mg/l mg/l mg/l mg/l mg/l
25.8 2 4 . 5 75222 23.828.3 214281 8392226 3772141 1.5820.53 62 2 30 110 (27-539)' 170 (41-821)'
41.3 2 3.2 121 2 3 0 3 5 . 5 2 13.0 215 2 57 9952156 4 1 0 2 108 1.62 2 0.54 100234 180 (83-412)' 283 (132-493)'
Concentration of Twelve Plasma Proteins at Birth in Very Low Birthweight and in Term Infants STAFFAN K. T. POLBERGER,' GORAN FEX2 and NIELSC. R. RAIHA' From the Departments of IPaediatrics and 2Clinical Chemistry, University of Lund, Malmo General Hospital, Sweden
ABSTRACT. Polberger, S. K. T., Fex, G. and Riiiha, N. C. R. (Departments of Paediatrics and Clinical Chemistry, University of Lund, Malmo General Hospital, Sweden). Concentrations of twelve plasma proteins at birth in very low birthweight and in term infants. Acta Paediatr Scand 79: 729, 1990. Plasma samples obtained at birth from 70 very low birth weight (VLBW) infants (gestational age 24 to 34 weeks) and from 20 term infants were analysed for concentrations of 12 different proteins. The plasma concentrations of albumin, transthyretin (TTR), retinol-binding protein (RBP), vitamin D-binding protein, apolipoprotein A I, fibronectin, orosomucoid and a,-antichymotrypsin were significantly lower in the VLBW infants than in the term infants, whereas the values of a-fetoprotein (AFP) were significantly higher in the VLBW infants. No differences were found between the two groups for apolipoprotein A 11, apolipoprotein B and transferrin. Birth asphyxia and sex had no influence on the measured plasma protein concentrations. The plasma concentrations of apolipoprotein A I and A I1 were significantly lower in small-forgestational age (SGA), VLBW infants compared with appropriate-for-gestational age (AGA), VLBW infants. Possible acute inflammation (defined as raised concentrations of orosomucoid or a,-antichymotrypsin) was associated with significantly higher values of vitamin D-binding protein in both VLBW and term infants, suggesting that this protein may act as an acute phase protein in newborn infants. Key words:plasma proteins, electroimmunoassay, VLBW infants, term infants.
It has long been recognised that fetal plasma protein synthesis is induced at different gestational ages for different proteins (1). During pregnancy, the concentrations of most plasma proteins increase in the fetus with a few exceptions (e.g. a2macroglobulin and AFP). Since the studies by Gitlin et al. (2), several reports have presented data on plasma proteins in newborn infants, however, most of these studies were performed on cord blood from term infants only. For most of the plasma proteins studied, there are few data on concentrations in VLBW infants at birth, especially subdivided data considering various gestational ages. As part of a nutritional study (3), we have measured concentrations at birth of 12 different plasma proteins in term and VLBW infants, the latter group of infants all having their gestational ages assessed by an ultrasound fetometry in early pregnancy. MATERIAL AND METHODS In a prospective study at the neonatal units in Malmo and Lund, the concentrations of 12 plasma (EDTA) proteins were determined in 70 VLBW infants and 20 term infants. Ten of the plasma proteins were selected to serve as possible indicators of nutritional status in VLBW infants subsequently fed varying protein and energy intakes, while orosomucoid and a,-antichymotrypsin were chosen as indicators of possible inflammatory reaction. In all infants, the gestational ages were calculated from an early ultrasound fetometry performed in weeks 16 to 18. This gave a reliable estimation of gestational age of 2 3 days (SD) expressed as completed gestational weeks (4). Blood was drawn from routine indwelling umbilical catheters (artery or vein) in the VLBW infants as soon as installed (within 2 hours from birth)
730 S. Polberger et al.
Acta Paediatr Scand 79
number 25
I
15 10
5 gestational weeks
0
24
26
Fig. 1. The gestational age distribution of the VLBW infants ( n = 70).
and always before blood or plasma transfusions were given and before enteral feeding was initiated. Before blood sampling 1.5-2 ml of blood was withdrawn from the catheter to avoid dilution with infusion fluids. Cord blood was obtained from the term infants immediately after delivery. Plasma (250-350 pl) was isolated and dispensed in small 600 p1 plastic tubes and stored at -70°C for 2-6 months before analysis. Samples from two groups of infants were studied: 1 . Term, AGA infants (n=20) gestational age: mean ? SD, (range) 40? 1 (38-4 ) weeks, birthweight: 3 590'. 500 (2 560-4 650) g. 2. VLBW infants (n=70), gestational age: 28?2 (24-34) weeks, birth weight: 1 l o t 2 4 0 (600-1 500) g, with the subgroups:
a. AGA infants < 29 weeks ( n = 3 l), b. AGA infants 2 29 weeks ( n =26) and c. SGA infants 2 2 9 weeks 13) (defined as birthweight less than - 2 S D according to Swedish growth charts (5)).
(n=
The distribution of gestational age of the 70 VLBW infants is presented in Fig. 1 . Most infants had a gestational age of 24 t o 30 weeks; only 9 infants were 2 3 1 weeks at birth. All infants below 29 weeks of gestation (n=31) were AGA and therefore we included only infants with a gestational age of 2 2 9 weeks in the comparison between AGA and SGA infants. No infant in the term group (n=20) had Apgar scores below 7 at 1 , 5, or 10 minutes, whereas 38 infants in the VLBW group ( n = 6 8 ) had Apgar scores < 7 at 1 minute, 14 at 5 minutes, and 4 at 10 minutes of age. Apgar scores were not obtained in 2 infants in the VLBW group. Determination of plasma proteins. The concentrations of all plasma proteins were determined in a total volume of only 100 pI of plasma, using electroimmunoassay (6). Plasma was usually diluted with 0.075 M barbital- 1.5 m M NaN, buffer pH 8.6. For analysis of fibronectin a second dilution with the same buffer containing 20 m M EDTA was performed and for analysis of apolipoproteins A1 and A I1 plasma was diluted with 8 M urea. The agarose used was Seakem from Marine Colloids (Locklands, ME) except for the assay of vitamin D-binding protein and fibronectin (agarose from Miles Seravac). Antisera t o all proteins except T T R (which was a goat antiserum) were raised in rabbits using the purified proteins as antigens. Seronorm" protein batch No. 103 (Nyegaard & Co, Oslo, Norway) was used as calibration standard for most of the proteins. For vitamin D-binding protein we used a standard from Behringwerke and for apolipoproteins A1 and B a standard was obtained from professor G Bondjers, Department of Internal Medicine, Sahlgrenska Hospital, Goteborg. For a,-antichymotrypsin and AFP we used the purified proteins as calibration standards. Pooled plasma from 20 term infants in two dilutions was used as an internal control. As no reliable standards were available for apolipoprotein A11 and fibronectin, the presented data on these proteins are expressed as per cent of an adult reference population (apolipoprotein A 11) or as percentage of the values in term infants (fibronectin). Statistical methods. Concentrations of most of the plasma proteins are expressed as arithmetic means and standard deviations (SD) except for orosomucoid, a,-antichymotrypsin and AFP, where the distributions showed pronounced skewness. In these cases, the average is presented as the median and the variation is described by means of a 95% tolerance interval obtained through applying the usual procedure t o the logarithms of the concentrations, and then taking antilogarithms. Significance of differences between groups were analysed by the
Plasma proteins in VLBW infants 73 1
Acta Paediatr Scand 79
Mann-Whitney U-test and correlations by Spearman's rank-correlation coefficient (r,?).This study was approved by the Ethics Committee at the University of Lund.
RESULTS Concentrations of ten plasma proteins in the VLBW and the term infants are given in Table 1. Except for apolipoprotein B and transferrin, the means and medians were significantly lower in the VLBW group compared with the term infants. Two more plasma proteins were studied in some of the infants. The concentration of apolipoprotein A I1 in the VLBW group ( n= 38) was 28 ? 9 9'0 (mean 2 SD) and in the term group (n=20) 27 2 1 1 9'0 @=0.70, non significant). The concentrations of AFP was analysed in 37 VLBW infants and 20 term infants. The medians (tolerance interval) were 320 (68-990) mg/l and 33 (10-120) mg/l, respectively. The difference between the two groups is highly significant (p 41 2 mg/l and/or a,-antichymotrypsin > 493 mg/l) with
Table 1. Concentrations of plasma proteins in VLBW and in term infants
Albumin TTR RBP Vit D-binding protein Apolipoprotein A I Apolipoprotein B Transferrin" Fibronectin Orosomucoid a,-antichymotrypsina
Unit
VLBW (n=70)
Term (n=20)
p
g/l mg/l mg/l mg/l mg/l mg/l
25.8 2 4 . 5 75222 23.828.3 214281 8392226 3772141 1.5820.53 62 2 30 110 (27-539)' 170 (41-821)'
41.3 2 3.2 121 2 3 0 3 5 . 5 2 13.0 215 2 57 9952156 4 1 0 2 108 1.62 2 0.54 100234 180 (83-412)' 283 (132-493)'
