- Email: [email protected]
Body weight, static and dynamic skinfold thickness in small premature infants during the first month of life
Early Human Development, Elsevier
217
8 (1983) 217-224
EHD 00503
Body weight, static and dynamic skinfold thickness in small premature infants during the first month of life Sergio A. Bustamante, University
Patricia Jacobs and John A. Gaines
of Arizona,Health Sciences Center, Tucson, AZ 85724, U.S.A. Accepted
for publication
5 April 1983
Summary
The growth of prematurely born infants is different from the growth of fetuses of the same age remaining in utero. This is in part due to changes in body composition that occur after birth. In search for a practical and reliable method to assess the growth of small prematures, we analyzed data obtained in two anthropometric studies that included 180 premature infants of 750- 1750 g weight at birth, and we studied the relationships between weight, static skinfold thickness (SSFT) and dynamic skinfold thickness (ASFT, i.e. the percentage of change in skinfold thickness between 15 and 60 s after application of the Harpenden caliper). The results show that the SSFT increases steadily after birth in spite of a significant decrease in weight and ASFT. Whether it contains fat or not, the fold of the skin is increasing in thickness at a time when by weight alone, one would have considered that there was no growth. The nutritional implication of this finding remains to be studied. Serial correlations of measures obtained at each period indicate that weight and SSFT have a good correlation to same measures in subsequent weeks (P < 0.01). ASFT, however, showed only a weak correlation (P = 0.05). The ASFT follows the general pattern of known changes in total body water, but it is not accurate enough to determine changes in individual infants; further studies are thus needed to find a practical method to evaluate changes of body composition and its relevance in the measurement of growth of premature infants. skinfold nutrition
thickness;
0378-3782/83/$03.00
body
weight;
body
composition;
0 1983 Elsevier Science Publishers
B.V.
premature
infants:
growth;
218 Introduction The growth and nutrition of premature infants have been studied extensively over the past several decades, and yet optimal postnatal growth for small premature infants remains to be defined. Measurements of growth which reflect body composition changes that occur after birth are needed; the procedure should be simple enough for everyday practice. Even though body weight is a fairly simple and accurate measurement, by itself, it does not tell the whole story about relative changes in body composition that are important in the assessment of growth [17]. Methods like antipyrine space [8], corrected bromide space [7], the distribution of deuterium in the body fluids [lo] and other methods involving markers of water in its compartments within the body [ 1 l] have been used to study the changes of body composition that occur in both premature and full term infants after birth, and yet these methods are invasive and complex enough to prevent their use in the everyday care and assessment of growth of newborn infants. Brans et al. proposed the existence of a relationship between skinfold thickness and body composition [2]. They suggested that if proven reliable, skinfold thickness measurements (SFT) could help to define optimum growth of premature infants. In an attempt to verify its usefulness, we analyzed a collection of weight and SFT measurements originated during two studies of the anthropometric growth of premature infants of birth weights between 750 and 1750 g.
Subjects and Methods Premature newborn infants with birth weights between 750 and 1750 g were included in two studies of anthropometric growth carried out at Wesley Medical Center, Wichita, KS, from 1976 to 1978 [5], and at the University of Arizona Health Sciences Center, Tucson, AZ, from 1978 to 1980 [6]. A total of 180 infants were studied. Subjects were selected for both studies according to the following criteria: (1) appropriate for gestational age; (2) free of non-bacterial infectious disease; (3) free of major congenital anomalies; (4) no post hemorrhagic hydrocephalus; (5) no congenital heart disease; (6) no major surgery; and (7) parental informed consent obtained. The protocols were approved by the Human Subjects Committee of each institution. The infants had their weights recorded daily to the nearest 5 g on a standard nursery scale or on an electronic balance. The measurements of skinfold thickness (SFT) were made at two sites - mid-triceps (SFT,) and subscapular (SFT,) - using a Harpenden caliper (British Indicators Ltd., St. Albans, U.K.). Readings were made to the nearest 0.05 mm at 15 s (SFT,,) and 60 s (SFT,,) after the application of the caliper. The dynamic change of the SFT (ASFT) was defined as the percentage of change measured from 15 to 60 s according to the following formula [2]:
dsm
= 100 x (sFT,,
Measurements
- SFT,,)/SFT,,
of the SFT
were done
by four
observers,
each
one
trained
and
219
supervised by the senior author (S.A.B.). The first SFT measurement was made in every case within 48 h of birth; subsequent measurements were made every week thereafter unless the attending neonatologist considered that the baby was too unstable to be moved. For statistical analysis we have taken one measurement of weight per week; the one that corresponded to the day the weekly skinfold thickness was measured. Infants included in the study were in an incubator or under a radiant heater. Those infants under radiant heaters also had a plastic cover and added humidity to help prevent insensible water loss [ 121. The feeding protocol used for both studies was as follows: On day 1, infants were given 10% dextrose in water with sodium 3 mEq/kg/day at a rate of 75-100 ml/kg/day (amounts in inverse relationship to birth weight). Sodium used was either bicarbonate (for pH less than 7.30) or chloride. On day 2, 1 g of amino acids per kg/day and 1 ml of multivitamin preparation was added to the parental solution and the volume increased by 15-25 ml/kg/day. On day 3 and every day thereafter, amino acids were increased by 0.5 g to a maximum of 2.5 g/kg/day, and the volume of parenteral fluid increased by 15-25 ml to a maximum of 150 ml/kg/day. On day 4, infants that required less than 0.30 FIO,, and whose respiration rate was less than 80/min, were started on enteral nutrition. We used either own mothers’ milk or a soy-based infant formula (Isomil@ Ross Laboratories, Columbus, OH) given by intermittent gastric gavage every 3 h, starting with approximately 25% of the total fluid to be given while decreasing the parenteral fluids accordingly. On day 5, 40-60% of intake was provided by enteral nutrition. On day 6, up to 80% was given by enteral route and by day 7, total enteral nutrition was achieved. Infants were started on enteral feedings when the conditions of day 4 were met; not all the infants met the day 4 criteria at the same postnatal age. However, all infants had to be on 81 kcal/lOO ml caloric density formula by day 21 or were excluded from the study (see below). The infant’s own mother’s milk or soy-formula was given until the infant showed a gain in weight of at least 10 g per day. The feedings were then changed to an 81 kcal/lOO ml caloric density formula until the infant reached a weight of 2100 g. Similac 24@, Similac LBW@ and Similac Special Care@ (Ross Laboratories, Columbus, OH) were used as 81 kcal/lOO ml caloric density formulas. Statistical
analysis
An initial evaluation of the data in groups of 250 g (shown in Table II) offered no advantage over the presentation of the total study population as one group. The interrelationship of the measurements did not change with the size of the infant, only an insignificant wider scatter of the data was seen in the smaller infants. The data gathered was analyzed in three stages corresponding to three foci of interest. First, interrelations among the measurements recorded at birth (weight and gestational age), and within the first 48 h (SFT, ASFT) were examined with simple linear correlations, and scatter plots were generated to assess for non-linearities. Second, changes in each measurement over time (growth) were assessed with interperiod paired t-tests, subsequent to multivariate t2 tests of the existence of
220
interperiod differences. Third, interrelations among individual curves of the several measurements were examined. This was done by first converting each weight, SFT and ASFT(x) to a Z-scoring using mean (X) and standard deviation (s) as norms for each period using the formula: Z = (x - X)/s. This maneuver in effect stabilized the variances over time preserving only information on each subject’s relative position with respect to the rest of the group, on each measurement, at each time period. Regression analysis of changes in relative weight, SFT and ASFT was then used to study interperiod differences between Z-scores to quantify each subject’s trajectory directly relative to the group on each measurement over each recorded period. Results Of the 180 infants followed in this study, 92 were males and 87 females; one infant had no recorded sex. We noted differences of measurements between the two institutions (data not shown; available upon request) but used the pooled data because separation had no effect on the results of the study. The weight and SFT,, had a positive correlation with gestational age (r = 0.735, SFT, 0.452, SFT, 0.404) while the ASFT showed a fair degree of negative correlation (r = ASFT, - 0.329, ASFT, - 0.311) (Table I). There was no significant correlation between sex and any of the measurements taken. The ASFT measurements of mid-triceps and subscapular sites show only a moderate correlation with each other (r = 0.485) and, in turn, an inverse correlation to weight at birth as well as to the SFT, with the heavier infants showing smaller ASFT and the smaller infants the larger ASFT. The patterns of change (growth) are summarized in Table II. There was a significant decrease in weight during the first week of life matched by an increase observed in the second week; by the end of the second week the group had regained birth weight; thereafter, weights continue to increase. The SFT showed no decrease but rather a steady increase from birth and throughout the observation period while the ASFT showed a significant drop in the TABLE
I
Correlation
among
measurements Gestational
obtained
in the first 48 h
wt.
SFT,
SFT,
ASFT,
ASFT,
0.990 0.735
0.638 0.452 0.636
0.590 0.404 0.580 0.762
- 0.370 -0.329 -0.371 - 0.392 - 0.341 -
- 0.376 -0.311 - 0.382 - 0.365 -0.418 - 0.485 -
age Birth weight Gestational age
wt. SFT, SFT, A SFT, A SFT,
0.735 -
Wt. = weight; SFT, = skinfold thickness mid-triceps; SFT, = skinfold thickness subscapular; dynamic skinfold thickness mid-triceps; ASFT, = dynamic skinfold thickness subscapular.
ASFT, =
II
1002+ 10925
18 14
3 wks
4 wks
= dynamic
925*
16
2 wks
168
1486+123 1627+ 1847k 1962k161
64 64 58 30
2 wks
3 wks
4 wks
(n=67)
2.50 + 0.49 2.88 f 0.53 3.13kO.55 3.34 & 0.58
59 61 58 29
site measurements
given, subscapuiar
SD.
b Mid-triceps
f
age in weeks
measurements
skinfold
were similar.
expressed
as a percent
of change in measurements
117
2.93 _+0.73 117
1622_+347
121
4 wks
thickness in mm; ASFT
163 158
I 2.46 _+0.6
2.75 f 0.66
163 158
I 360 k 283 1510*350
136 149
2.05 + 0.4 1
58
2.20 f 0.49
1k 2.68 7.8
7.34*
61
136
9.48 k 2.76 59
149
13.88+3.19
52
obtained
at 15 and 60 s.
8.22_+ 1.92
8.24 * 2.36
8.53 _+2.78
10.17*3.14
15.35 * 3.57
8.45 + 1.85
1.91
1.58 35
166
thickness
2.34 + 0.36
52
8.06 + 2.43 7.40+
41
158
a Gestational
SSFT = static skinfold
29
3.20f0.63
3 wks
(n = 180)
1334+261 1251k247
180 163
193
2.96 _+0.56
2 wks
1 wk
2 days
30.6 k 0.5 =
All groups
79 151
1615+
67
2 days
1 wk
1500- 1750
32.2 + 1.48 a
183
35
38
41
1552+ 1790*
41
3 wks
4 wks
9.95 * 3.10 8.04 + 2.49
44
2.50k0.55
1369 k 149
45
2 wks
(n=46)
44
15.71 k3.29 41
2.05 + 0.36
35
67 2.20_+0.40
1263_+103
43
35
1367k
46
1 wk
186
41
8.57k2.12
38
1+ 0.60
2.6
38
1378+
39
4 wks
2 days
9.69 + 3.05 9.04 + 2.48
41 41
2.06+0.34 2.34 f 0.45
41 41
107
1104+ 1215k146
16.49k3.62 10.67 + 3.29
1.89
34
s.90*
36
15
41
30.8 f 1.63 a
1250-1499
10.10+2.57
17
9.54k2.15
12.60 f 3.40
13 18
17.46+_3.61
15
Mean f S.D.
b
41
1.84+0.27
2.15+0.42
15
1.91+0.29
2.OOf0.39
18
36
1.84kO.36
17
34
1.65 + 0.22
13
n
mid-triceps
3 wks
99
78
129
122
100
73
1.63 f 0.20
15
ASFT
2 wks
1 wk
(n=48)
1119+ 1012+
48 43
2 days
29.0+
1.14’
851k
13
1wk
1000-1249
(n=l9)
1.25a
2 days
28.6 f
750-999
’
Mean k S.D.
n
885+
19
63
” Mean f S.D.
SSFT mid-triceps
to weight group and time after birth
Weight (g)
thickness in relation
Time
skinfold
Weight group
Static and dynamic
TABLE
222 TABLE
III
Correlations
of measurements
between
periods
Time
Weight
SM;
SFT,
A SFT,
A SFT,
First 48 h to 1 wk I wkto2wks 2 wks to 3 wks 3 wks to 4 wks
0.944 0.969 0.980 0.984
0.75 1 0.834 0.886 0.897
0.678 0.837 0.902 0.912
0.289 * 0.364 0.393 0.299 *
0.074 * * 0.237 * 0.367 0.466
Abbreviations same as in Table I. * Borderline correlation at P > 0.05, P > 0.0 1. ** No significant correlation at P = 0.05.
first week followed by a less significant drop the second week and stabilization thereafter for the rest of the study period. The statistical evaluation of how subjects behave within the group is summarized in Table III. We found the weight and both SFT, and SFT, to have good to very strong correlations. The values tended to increase with time indicating that subjects maintain the same relative position within the groups with less mixing occurring with each period of observation. Thus, babies in the high side of the distribution within of the the group remain in the high side and so on. The serial correlations measurements of ASFT resulted in lower values but still significant (except for the ASFT, between first 48 h and 1 week). Discussion To paraphrase Dancis’ remark, the weight curve of the normal premature infant is probably as well known as any pediatric datum [9]. Since the publication of Dancis’ weight grid, several growth charts have been published, most notably Babson’s [l]. These charts, however, do not account for a major factor in the assessment of growth of the premature infant: the water share of the body mass and fluctuations between intracellular and extracellular compartments related to gestational age [7], postnatal age and postnatal nutrition [3]. Water is the largest single constituent of all living matter. The proportion of change in weight relative to total body water is therefore an important variable in the assessment of growth of premature infants. The total body water varies from approximately 86% in the fetus at 30 weeks of gestational age to 65% in the newborn at term [ 131. Furthermore, both the total body water and the extracellular water compartment show an appreciable loss over the first days of life, a change that occurs irrespective of gestational age at birth [2,14]. Tissue accretion as opposed to water retention and the amount of water in extracellular vs. intracellular spaces are changes not reflected by weight as the only measurement of growth. Indeed, these factors may change with gestational age and with the feeding practices of different nurseries [3]. The use of calipers to evaluate body fat in nutrition studies has been established for some time [ 151. The use of ASFT as a measure of subcutaneous interstitial water has been proposed by Brans et al. [2]. The relationship of the static and dynamic skinfold measurements to changes in weight of small premature infants 750- 1750 g at birth has not been reported.
223
Our study population was selected on the basis of two similar protocols used to assess the effects of different formulas in the anthropometric growth of small premature infants [5,6]. The general health of these small premature infants naturally has considerable variations but we expected that those uncontrolled variations were limited by the specified selection criteria (vide supra). The differences of measurements observed between the two centers did not relate to sex, ethnic origin, caliper use or any other parameter tested, but may have been due to inter-observer variation [4]. The separated groups, however, show the same trends consistently and thus are presented together. As is commonly seen in premature infants, an initial weight loss is followed by a steady gain maintained thereafter. The initial weight loss may be due in part to known changes of body water occurring after birth [7]. As a group, premature infants lose weight at approximately the same rate as their ASFT decreases during the first week of life. The ASFT continues to decline during the second week to a steady state that was then maintained for the duration of the study. These changes follow the same general direction and shape of curves of total body water reported by most authors [ 1 I] and suggests the ASFT changes may be due to changes in total body water. The SFT measurements follow a different pattern than body weight, whether SFT represents fat, skin thickness or a combination of both [ 131. These tissues seem to continue growing after birth regardless of what happens to the body weight and to the ASFT. The relevance of this finding has to do with the utilization of nutrients that constitute the skinfold (whether it contains fat or not). The competition for nutrients that the growth of some organs may impose over other organs indicates perhaps that the nutrition of small prematures should be modified to meet the demands of specific growth of different organs. We were disappointed with the apparent criss-crossing of individual lines within the ASFT measurements and consider these only weak indicators of where the individual stands within the weight or the SFT groups. This apparent lack of correlation, however, may be due to the method used and the time sequence of the measurements taken. The phenomenon of water distribution and its relationship with the ASFT may require that this measurement be done with a different methodology, i.e. study the dynamic curve rather than the two points in time of the ASFT, taking the measurements at shorter intervals to assess its relationship to already known changes in water compartments [ 161. In summary Studying the relationship between weight, static SFT and ASFT we have come to the following conclusions: 1. The double layer of skin and subcutaneous tissue thickness continues to increase after birth irrespective of changes of body weight. 2. The ASFT changes are comparable to reported changes in total body water that occur after birth, but with the method used ASFT does not appear reliable to assess individual variations. 3. Weight and SFT appear as more reliable indicators of change (growth) than the ASFT.
224
Based on these observations, it is apparent that further studies are needed to find a practical method of assessment of body composition and water compartments so that we can better understand and evaluate the dynamics of growth and development of prematurely born infants. Acknowledgements The authors acknowledge with gratitude Otakar Koldovsky’s reviews of the manuscript, and Sheila Schreck and Rosemary Branson for technical assistance. This work was supported in part by grants from Ross Laboratories and Wesley Research Foundation. References 1 Babson, S.G. (1970): Growth of low-birth-weight infants. J. Pediatr. 77, 1 l-18. approach to body 2 Brans, Y.W., Sumners, J.E., Dweck, H.S. and Cassady, G. (1974): Noninvasive composition in the neonate: Dynamic skinfold measurements. Pediatr. Res. 8, 215-222. G. (1976): Feeding the 3 Brat-q Y.W., Sumners, J.E., Dweck, H.S., Bailey, B.A. and Cassady, low-birthweight infant: orally or parenterally?: II. Corrected bromide space in parenterally supplemented infants. Pediatrics 58, 809-815. 4 Branson, R.S., Vaucher, Y.E., Harrison, G.G., Vargas, M. and Thies, C. (1982): Inter- and intra-observer reliability of skinfold thickness measurements in newborn infants. Hum. Biol. 54, 137- 143. S.A. (1980): Growth and development of low-birth-weight infants fed experimental 5 Bustamante, formula and Similac 24. In: Meeting Nutritional Goals for Low-Birth-Weight Infants. Ross Clinical Research Conference, Columbus, OH. S.A. and S&reck, S. (1979): Growth of low-birth-weight infants fed Isomil or Similac 24 6 Bustamante, LBW. In: Ross Clinical Research Conference Low-Birth-Weight Infants fed Isomil, pp. 59-64. Editors: W.L. Bachhuber, J.D. Benson, M.C. Dame and I.M. Gray. Ross Laboratories, Columbus, OH. 7 Cassady, G. (1970): Bromide space studies in infants of low birth weight. Pediatr. Res. 4, 14-24. 8 Cassady, G. and Milstead, R.R. (1971): Antipyrine space studies and cell water estimates in infants of low birth weight. Pediatr. Res. 5, 673-682. 9 Dancis, J., O’Connell, J.R. and Halt, L.M. (1948): A grid for recording the weight of premature infants. J. Pediatr. 33, 570-572. 10 Flexner, L.B., Wilde, W.S., Proctor, N.K., Cowie, D.B., Vosburgh, G.J. and Hellman, L.M. (1947): The estimation of extracellular and total body water in the newborn infant with radioactive sodium and deuterium oxide. J. Pediatr. 30, 413-415. 11 Friis-Hansen, B. (1957): Changes in body water compartments during growth. Acta Paediatr. Stand. 46, Suppl. 110. 12 Hammarlund, K., Sedin, G. and Stromberg, B. (1982): Transepidermal water loss in newborn infants. VII. Relation to post-natal age in very pre-term and full-term appropriate for gestational age infants. Acta Paediatr. Stand. 71, 369-374. 13 McGowan (1979): The fat and water content of the dead infants’ skinfold. Pediatr. Res. 13, 1304- 1306. 14 Maclaurin, J.C. (1966): Changes in body water distribution during the first two weeks of life. Arch. Dis. Childh. 41, 286-291. 15 Tanner, J.M. and Whitehouse, R.H. (1955): The harpenden skinfold caliper. Am. J. Physiol. Anthropol. 13, 743-746. 16 Thortoh, C.J., Shannon, D.L., Hunter, M.A. and Brans, Y.W. (1982): Dynamic skinfold thickness measurements: a noninvasive estimate of neonatal extracellular water content. Pediatr. Res. 16, 989-994. 17 Widdowson, E.M. (1974): Changes in body proportions and composition during growth. In: Scientific Foundations of Pediatrics, pp. 153- 163. Editors: J.A. Davis and J. Dobbing. W.B. Saunders Company, Philadelphia.
217
8 (1983) 217-224
EHD 00503
Body weight, static and dynamic skinfold thickness in small premature infants during the first month of life Sergio A. Bustamante, University
Patricia Jacobs and John A. Gaines
of Arizona,Health Sciences Center, Tucson, AZ 85724, U.S.A. Accepted
for publication
5 April 1983
Summary
The growth of prematurely born infants is different from the growth of fetuses of the same age remaining in utero. This is in part due to changes in body composition that occur after birth. In search for a practical and reliable method to assess the growth of small prematures, we analyzed data obtained in two anthropometric studies that included 180 premature infants of 750- 1750 g weight at birth, and we studied the relationships between weight, static skinfold thickness (SSFT) and dynamic skinfold thickness (ASFT, i.e. the percentage of change in skinfold thickness between 15 and 60 s after application of the Harpenden caliper). The results show that the SSFT increases steadily after birth in spite of a significant decrease in weight and ASFT. Whether it contains fat or not, the fold of the skin is increasing in thickness at a time when by weight alone, one would have considered that there was no growth. The nutritional implication of this finding remains to be studied. Serial correlations of measures obtained at each period indicate that weight and SSFT have a good correlation to same measures in subsequent weeks (P < 0.01). ASFT, however, showed only a weak correlation (P = 0.05). The ASFT follows the general pattern of known changes in total body water, but it is not accurate enough to determine changes in individual infants; further studies are thus needed to find a practical method to evaluate changes of body composition and its relevance in the measurement of growth of premature infants. skinfold nutrition
thickness;
0378-3782/83/$03.00
body
weight;
body
composition;
0 1983 Elsevier Science Publishers
B.V.
premature
infants:
growth;
218 Introduction The growth and nutrition of premature infants have been studied extensively over the past several decades, and yet optimal postnatal growth for small premature infants remains to be defined. Measurements of growth which reflect body composition changes that occur after birth are needed; the procedure should be simple enough for everyday practice. Even though body weight is a fairly simple and accurate measurement, by itself, it does not tell the whole story about relative changes in body composition that are important in the assessment of growth [17]. Methods like antipyrine space [8], corrected bromide space [7], the distribution of deuterium in the body fluids [lo] and other methods involving markers of water in its compartments within the body [ 1 l] have been used to study the changes of body composition that occur in both premature and full term infants after birth, and yet these methods are invasive and complex enough to prevent their use in the everyday care and assessment of growth of newborn infants. Brans et al. proposed the existence of a relationship between skinfold thickness and body composition [2]. They suggested that if proven reliable, skinfold thickness measurements (SFT) could help to define optimum growth of premature infants. In an attempt to verify its usefulness, we analyzed a collection of weight and SFT measurements originated during two studies of the anthropometric growth of premature infants of birth weights between 750 and 1750 g.
Subjects and Methods Premature newborn infants with birth weights between 750 and 1750 g were included in two studies of anthropometric growth carried out at Wesley Medical Center, Wichita, KS, from 1976 to 1978 [5], and at the University of Arizona Health Sciences Center, Tucson, AZ, from 1978 to 1980 [6]. A total of 180 infants were studied. Subjects were selected for both studies according to the following criteria: (1) appropriate for gestational age; (2) free of non-bacterial infectious disease; (3) free of major congenital anomalies; (4) no post hemorrhagic hydrocephalus; (5) no congenital heart disease; (6) no major surgery; and (7) parental informed consent obtained. The protocols were approved by the Human Subjects Committee of each institution. The infants had their weights recorded daily to the nearest 5 g on a standard nursery scale or on an electronic balance. The measurements of skinfold thickness (SFT) were made at two sites - mid-triceps (SFT,) and subscapular (SFT,) - using a Harpenden caliper (British Indicators Ltd., St. Albans, U.K.). Readings were made to the nearest 0.05 mm at 15 s (SFT,,) and 60 s (SFT,,) after the application of the caliper. The dynamic change of the SFT (ASFT) was defined as the percentage of change measured from 15 to 60 s according to the following formula [2]:
dsm
= 100 x (sFT,,
Measurements
- SFT,,)/SFT,,
of the SFT
were done
by four
observers,
each
one
trained
and
219
supervised by the senior author (S.A.B.). The first SFT measurement was made in every case within 48 h of birth; subsequent measurements were made every week thereafter unless the attending neonatologist considered that the baby was too unstable to be moved. For statistical analysis we have taken one measurement of weight per week; the one that corresponded to the day the weekly skinfold thickness was measured. Infants included in the study were in an incubator or under a radiant heater. Those infants under radiant heaters also had a plastic cover and added humidity to help prevent insensible water loss [ 121. The feeding protocol used for both studies was as follows: On day 1, infants were given 10% dextrose in water with sodium 3 mEq/kg/day at a rate of 75-100 ml/kg/day (amounts in inverse relationship to birth weight). Sodium used was either bicarbonate (for pH less than 7.30) or chloride. On day 2, 1 g of amino acids per kg/day and 1 ml of multivitamin preparation was added to the parental solution and the volume increased by 15-25 ml/kg/day. On day 3 and every day thereafter, amino acids were increased by 0.5 g to a maximum of 2.5 g/kg/day, and the volume of parenteral fluid increased by 15-25 ml to a maximum of 150 ml/kg/day. On day 4, infants that required less than 0.30 FIO,, and whose respiration rate was less than 80/min, were started on enteral nutrition. We used either own mothers’ milk or a soy-based infant formula (Isomil@ Ross Laboratories, Columbus, OH) given by intermittent gastric gavage every 3 h, starting with approximately 25% of the total fluid to be given while decreasing the parenteral fluids accordingly. On day 5, 40-60% of intake was provided by enteral nutrition. On day 6, up to 80% was given by enteral route and by day 7, total enteral nutrition was achieved. Infants were started on enteral feedings when the conditions of day 4 were met; not all the infants met the day 4 criteria at the same postnatal age. However, all infants had to be on 81 kcal/lOO ml caloric density formula by day 21 or were excluded from the study (see below). The infant’s own mother’s milk or soy-formula was given until the infant showed a gain in weight of at least 10 g per day. The feedings were then changed to an 81 kcal/lOO ml caloric density formula until the infant reached a weight of 2100 g. Similac 24@, Similac LBW@ and Similac Special Care@ (Ross Laboratories, Columbus, OH) were used as 81 kcal/lOO ml caloric density formulas. Statistical
analysis
An initial evaluation of the data in groups of 250 g (shown in Table II) offered no advantage over the presentation of the total study population as one group. The interrelationship of the measurements did not change with the size of the infant, only an insignificant wider scatter of the data was seen in the smaller infants. The data gathered was analyzed in three stages corresponding to three foci of interest. First, interrelations among the measurements recorded at birth (weight and gestational age), and within the first 48 h (SFT, ASFT) were examined with simple linear correlations, and scatter plots were generated to assess for non-linearities. Second, changes in each measurement over time (growth) were assessed with interperiod paired t-tests, subsequent to multivariate t2 tests of the existence of
220
interperiod differences. Third, interrelations among individual curves of the several measurements were examined. This was done by first converting each weight, SFT and ASFT(x) to a Z-scoring using mean (X) and standard deviation (s) as norms for each period using the formula: Z = (x - X)/s. This maneuver in effect stabilized the variances over time preserving only information on each subject’s relative position with respect to the rest of the group, on each measurement, at each time period. Regression analysis of changes in relative weight, SFT and ASFT was then used to study interperiod differences between Z-scores to quantify each subject’s trajectory directly relative to the group on each measurement over each recorded period. Results Of the 180 infants followed in this study, 92 were males and 87 females; one infant had no recorded sex. We noted differences of measurements between the two institutions (data not shown; available upon request) but used the pooled data because separation had no effect on the results of the study. The weight and SFT,, had a positive correlation with gestational age (r = 0.735, SFT, 0.452, SFT, 0.404) while the ASFT showed a fair degree of negative correlation (r = ASFT, - 0.329, ASFT, - 0.311) (Table I). There was no significant correlation between sex and any of the measurements taken. The ASFT measurements of mid-triceps and subscapular sites show only a moderate correlation with each other (r = 0.485) and, in turn, an inverse correlation to weight at birth as well as to the SFT, with the heavier infants showing smaller ASFT and the smaller infants the larger ASFT. The patterns of change (growth) are summarized in Table II. There was a significant decrease in weight during the first week of life matched by an increase observed in the second week; by the end of the second week the group had regained birth weight; thereafter, weights continue to increase. The SFT showed no decrease but rather a steady increase from birth and throughout the observation period while the ASFT showed a significant drop in the TABLE
I
Correlation
among
measurements Gestational
obtained
in the first 48 h
wt.
SFT,
SFT,
ASFT,
ASFT,
0.990 0.735
0.638 0.452 0.636
0.590 0.404 0.580 0.762
- 0.370 -0.329 -0.371 - 0.392 - 0.341 -
- 0.376 -0.311 - 0.382 - 0.365 -0.418 - 0.485 -
age Birth weight Gestational age
wt. SFT, SFT, A SFT, A SFT,
0.735 -
Wt. = weight; SFT, = skinfold thickness mid-triceps; SFT, = skinfold thickness subscapular; dynamic skinfold thickness mid-triceps; ASFT, = dynamic skinfold thickness subscapular.
ASFT, =
II
1002+ 10925
18 14
3 wks
4 wks
= dynamic
925*
16
2 wks
168
1486+123 1627+ 1847k 1962k161
64 64 58 30
2 wks
3 wks
4 wks
(n=67)
2.50 + 0.49 2.88 f 0.53 3.13kO.55 3.34 & 0.58
59 61 58 29
site measurements
given, subscapuiar
SD.
b Mid-triceps
f
age in weeks
measurements
skinfold
were similar.
expressed
as a percent
of change in measurements
117
2.93 _+0.73 117
1622_+347
121
4 wks
thickness in mm; ASFT
163 158
I 2.46 _+0.6
2.75 f 0.66
163 158
I 360 k 283 1510*350
136 149
2.05 + 0.4 1
58
2.20 f 0.49
1k 2.68 7.8
7.34*
61
136
9.48 k 2.76 59
149
13.88+3.19
52
obtained
at 15 and 60 s.
8.22_+ 1.92
8.24 * 2.36
8.53 _+2.78
10.17*3.14
15.35 * 3.57
8.45 + 1.85
1.91
1.58 35
166
thickness
2.34 + 0.36
52
8.06 + 2.43 7.40+
41
158
a Gestational
SSFT = static skinfold
29
3.20f0.63
3 wks
(n = 180)
1334+261 1251k247
180 163
193
2.96 _+0.56
2 wks
1 wk
2 days
30.6 k 0.5 =
All groups
79 151
1615+
67
2 days
1 wk
1500- 1750
32.2 + 1.48 a
183
35
38
41
1552+ 1790*
41
3 wks
4 wks
9.95 * 3.10 8.04 + 2.49
44
2.50k0.55
1369 k 149
45
2 wks
(n=46)
44
15.71 k3.29 41
2.05 + 0.36
35
67 2.20_+0.40
1263_+103
43
35
1367k
46
1 wk
186
41
8.57k2.12
38
1+ 0.60
2.6
38
1378+
39
4 wks
2 days
9.69 + 3.05 9.04 + 2.48
41 41
2.06+0.34 2.34 f 0.45
41 41
107
1104+ 1215k146
16.49k3.62 10.67 + 3.29
1.89
34
s.90*
36
15
41
30.8 f 1.63 a
1250-1499
10.10+2.57
17
9.54k2.15
12.60 f 3.40
13 18
17.46+_3.61
15
Mean f S.D.
b
41
1.84+0.27
2.15+0.42
15
1.91+0.29
2.OOf0.39
18
36
1.84kO.36
17
34
1.65 + 0.22
13
n
mid-triceps
3 wks
99
78
129
122
100
73
1.63 f 0.20
15
ASFT
2 wks
1 wk
(n=48)
1119+ 1012+
48 43
2 days
29.0+
1.14’
851k
13
1wk
1000-1249
(n=l9)
1.25a
2 days
28.6 f
750-999
’
Mean k S.D.
n
885+
19
63
” Mean f S.D.
SSFT mid-triceps
to weight group and time after birth
Weight (g)
thickness in relation
Time
skinfold
Weight group
Static and dynamic
TABLE
222 TABLE
III
Correlations
of measurements
between
periods
Time
Weight
SM;
SFT,
A SFT,
A SFT,
First 48 h to 1 wk I wkto2wks 2 wks to 3 wks 3 wks to 4 wks
0.944 0.969 0.980 0.984
0.75 1 0.834 0.886 0.897
0.678 0.837 0.902 0.912
0.289 * 0.364 0.393 0.299 *
0.074 * * 0.237 * 0.367 0.466
Abbreviations same as in Table I. * Borderline correlation at P > 0.05, P > 0.0 1. ** No significant correlation at P = 0.05.
first week followed by a less significant drop the second week and stabilization thereafter for the rest of the study period. The statistical evaluation of how subjects behave within the group is summarized in Table III. We found the weight and both SFT, and SFT, to have good to very strong correlations. The values tended to increase with time indicating that subjects maintain the same relative position within the groups with less mixing occurring with each period of observation. Thus, babies in the high side of the distribution within of the the group remain in the high side and so on. The serial correlations measurements of ASFT resulted in lower values but still significant (except for the ASFT, between first 48 h and 1 week). Discussion To paraphrase Dancis’ remark, the weight curve of the normal premature infant is probably as well known as any pediatric datum [9]. Since the publication of Dancis’ weight grid, several growth charts have been published, most notably Babson’s [l]. These charts, however, do not account for a major factor in the assessment of growth of the premature infant: the water share of the body mass and fluctuations between intracellular and extracellular compartments related to gestational age [7], postnatal age and postnatal nutrition [3]. Water is the largest single constituent of all living matter. The proportion of change in weight relative to total body water is therefore an important variable in the assessment of growth of premature infants. The total body water varies from approximately 86% in the fetus at 30 weeks of gestational age to 65% in the newborn at term [ 131. Furthermore, both the total body water and the extracellular water compartment show an appreciable loss over the first days of life, a change that occurs irrespective of gestational age at birth [2,14]. Tissue accretion as opposed to water retention and the amount of water in extracellular vs. intracellular spaces are changes not reflected by weight as the only measurement of growth. Indeed, these factors may change with gestational age and with the feeding practices of different nurseries [3]. The use of calipers to evaluate body fat in nutrition studies has been established for some time [ 151. The use of ASFT as a measure of subcutaneous interstitial water has been proposed by Brans et al. [2]. The relationship of the static and dynamic skinfold measurements to changes in weight of small premature infants 750- 1750 g at birth has not been reported.
223
Our study population was selected on the basis of two similar protocols used to assess the effects of different formulas in the anthropometric growth of small premature infants [5,6]. The general health of these small premature infants naturally has considerable variations but we expected that those uncontrolled variations were limited by the specified selection criteria (vide supra). The differences of measurements observed between the two centers did not relate to sex, ethnic origin, caliper use or any other parameter tested, but may have been due to inter-observer variation [4]. The separated groups, however, show the same trends consistently and thus are presented together. As is commonly seen in premature infants, an initial weight loss is followed by a steady gain maintained thereafter. The initial weight loss may be due in part to known changes of body water occurring after birth [7]. As a group, premature infants lose weight at approximately the same rate as their ASFT decreases during the first week of life. The ASFT continues to decline during the second week to a steady state that was then maintained for the duration of the study. These changes follow the same general direction and shape of curves of total body water reported by most authors [ 1 I] and suggests the ASFT changes may be due to changes in total body water. The SFT measurements follow a different pattern than body weight, whether SFT represents fat, skin thickness or a combination of both [ 131. These tissues seem to continue growing after birth regardless of what happens to the body weight and to the ASFT. The relevance of this finding has to do with the utilization of nutrients that constitute the skinfold (whether it contains fat or not). The competition for nutrients that the growth of some organs may impose over other organs indicates perhaps that the nutrition of small prematures should be modified to meet the demands of specific growth of different organs. We were disappointed with the apparent criss-crossing of individual lines within the ASFT measurements and consider these only weak indicators of where the individual stands within the weight or the SFT groups. This apparent lack of correlation, however, may be due to the method used and the time sequence of the measurements taken. The phenomenon of water distribution and its relationship with the ASFT may require that this measurement be done with a different methodology, i.e. study the dynamic curve rather than the two points in time of the ASFT, taking the measurements at shorter intervals to assess its relationship to already known changes in water compartments [ 161. In summary Studying the relationship between weight, static SFT and ASFT we have come to the following conclusions: 1. The double layer of skin and subcutaneous tissue thickness continues to increase after birth irrespective of changes of body weight. 2. The ASFT changes are comparable to reported changes in total body water that occur after birth, but with the method used ASFT does not appear reliable to assess individual variations. 3. Weight and SFT appear as more reliable indicators of change (growth) than the ASFT.
224
Based on these observations, it is apparent that further studies are needed to find a practical method of assessment of body composition and water compartments so that we can better understand and evaluate the dynamics of growth and development of prematurely born infants. Acknowledgements The authors acknowledge with gratitude Otakar Koldovsky’s reviews of the manuscript, and Sheila Schreck and Rosemary Branson for technical assistance. This work was supported in part by grants from Ross Laboratories and Wesley Research Foundation. References 1 Babson, S.G. (1970): Growth of low-birth-weight infants. J. Pediatr. 77, 1 l-18. approach to body 2 Brans, Y.W., Sumners, J.E., Dweck, H.S. and Cassady, G. (1974): Noninvasive composition in the neonate: Dynamic skinfold measurements. Pediatr. Res. 8, 215-222. G. (1976): Feeding the 3 Brat-q Y.W., Sumners, J.E., Dweck, H.S., Bailey, B.A. and Cassady, low-birthweight infant: orally or parenterally?: II. Corrected bromide space in parenterally supplemented infants. Pediatrics 58, 809-815. 4 Branson, R.S., Vaucher, Y.E., Harrison, G.G., Vargas, M. and Thies, C. (1982): Inter- and intra-observer reliability of skinfold thickness measurements in newborn infants. Hum. Biol. 54, 137- 143. S.A. (1980): Growth and development of low-birth-weight infants fed experimental 5 Bustamante, formula and Similac 24. In: Meeting Nutritional Goals for Low-Birth-Weight Infants. Ross Clinical Research Conference, Columbus, OH. S.A. and S&reck, S. (1979): Growth of low-birth-weight infants fed Isomil or Similac 24 6 Bustamante, LBW. In: Ross Clinical Research Conference Low-Birth-Weight Infants fed Isomil, pp. 59-64. Editors: W.L. Bachhuber, J.D. Benson, M.C. Dame and I.M. Gray. Ross Laboratories, Columbus, OH. 7 Cassady, G. (1970): Bromide space studies in infants of low birth weight. Pediatr. Res. 4, 14-24. 8 Cassady, G. and Milstead, R.R. (1971): Antipyrine space studies and cell water estimates in infants of low birth weight. Pediatr. Res. 5, 673-682. 9 Dancis, J., O’Connell, J.R. and Halt, L.M. (1948): A grid for recording the weight of premature infants. J. Pediatr. 33, 570-572. 10 Flexner, L.B., Wilde, W.S., Proctor, N.K., Cowie, D.B., Vosburgh, G.J. and Hellman, L.M. (1947): The estimation of extracellular and total body water in the newborn infant with radioactive sodium and deuterium oxide. J. Pediatr. 30, 413-415. 11 Friis-Hansen, B. (1957): Changes in body water compartments during growth. Acta Paediatr. Stand. 46, Suppl. 110. 12 Hammarlund, K., Sedin, G. and Stromberg, B. (1982): Transepidermal water loss in newborn infants. VII. Relation to post-natal age in very pre-term and full-term appropriate for gestational age infants. Acta Paediatr. Stand. 71, 369-374. 13 McGowan (1979): The fat and water content of the dead infants’ skinfold. Pediatr. Res. 13, 1304- 1306. 14 Maclaurin, J.C. (1966): Changes in body water distribution during the first two weeks of life. Arch. Dis. Childh. 41, 286-291. 15 Tanner, J.M. and Whitehouse, R.H. (1955): The harpenden skinfold caliper. Am. J. Physiol. Anthropol. 13, 743-746. 16 Thortoh, C.J., Shannon, D.L., Hunter, M.A. and Brans, Y.W. (1982): Dynamic skinfold thickness measurements: a noninvasive estimate of neonatal extracellular water content. Pediatr. Res. 16, 989-994. 17 Widdowson, E.M. (1974): Changes in body proportions and composition during growth. In: Scientific Foundations of Pediatrics, pp. 153- 163. Editors: J.A. Davis and J. Dobbing. W.B. Saunders Company, Philadelphia.