Postnatal attainment of intrauterine macromineral accretion rates in low birth weight infants fed fortified human milk

Richard J. Schanler, MD, and Steven A. Abrams, MD

From the U.S. Department of Agriculture/Agricultural Research Service Children's Nutrition Re- search Center, and the Section of Neonatology, Department of Pediatrics, Baylor College of Medicine, Houston, Texas

Hypothesis: Provision of more b ioava i lab le mineral sources as human milk sup- plements enables very low birth weight (VLBW) infants to meet the intrauterine accret ion rate for calc ium and phosphorus. Design: Comparison of currently formulated human milk fortifier with previous formulation. Setting: Neonatal level II and III nurseries Patients: Twenty-six healthy, VLBW infants, whose mothers chose to breast-feed. Interventions: We tested the effects of two formulations designed for VLBW infants as human milk supplements and differing primarily in their quanti ty and source of Ca, P, and magnesium. The study interval began with a milk intake of 100 ml • kg -4 • day -4 and ended when a body weight reached 2.0 kg. Main outcome measures: Net absorption and retention of Ca, P, and Mg during a nutrit ional ba lance study conduc ted once during the study interval, growth during the entire study interval, and bone mineral content of the radius were measured at the beginning and end of the study interval. Results: The newer Ca g luconate-g lycerophosphate preparat ion (given to group CaGP) resulted in greater net absorption and retention of Ca and P (p <0.01) than in infants given Ca phosphate (group CaTB). Mg retention was greater than (in group CaGP) or equivalent to (in group CaTB) the intrauterine accret ion rate. Radius bone mineral content was signif icantly greater in group CaGP than in groupCaTB (p <0.001). Volumes of the fortif ied human milk prep- arat ion needed to meet the needs for gain in body weight were higher in group CaGP than in group CaTB (p <0.004). Conclusions: Intrauterine accret ion rates for Ca and P can be ach ieved when VLBW infants are fed human milk supplemented with Ca g luconate-g lycero- phosphate. Supplementat ion of human milk with Mg may not be indicated. In this study, greater intakes of Ca and P, and not improvements in bioavai labi l i ty, result in improved net retention and bone mineral content of VLBW infants. (J PE- DIATR 4995;126:441-7)

Supported in part by the General Clinical Research Center, Bay- lor College of Medicine/Texas Children's Hospital Clinical Re- search Center, by grant No. MO 1-RR-00188, National Institutes of Health, and Mead Johnson Nutritional Division, Evansville, Ind. Partial funding also has been provided from the U.S. Department of Agriculture/Agricultural Research Service under Cooperative Agreement No. 58-6250-1-003.

The contents of this publication do not necessarily reflect the views or policies of the U.S. Department of Agriculture, nor does men-

tion of trade names, commercial products, or organizations imply endorsement by the U.S. Government.

Submitted for publication April 12, 1994; accepted Sept. 8, 1994.

Reprint requests: Richard J. Schanler, MD, Children's Nutrition Research Center, Baylor College of Medicine, 1100 Bates St., Houston, TX 77030.

Copyright © 1995 by Mosby-Year Book, Inc. 0022-3476/95/$3.00 + 0 9/23/60452

4 4 1

4 4 2 Schanler and Abrams The Journal of Pediatrics March 1995

The need for mineral and, perhaps, protein supplementation of human milk in the nutritional support of the very low birth weight infant is recognized. 15 The data indicate that without mineral supplementation during hospitalization, bone mineralization and long-term outcomes such as growth, particularly in length, are adversely affected. 6s We ob- served a "catch-up" period of bone mineral content of the distal radius by 2 years of corrected age in healthy, former VLBW infants who received fortified human milk during their hospitalization. 9 The methods used at that time to fortify human milk did not provide optimal retention of calcium and phosphorus. 1° We inferred from those data that had we improved the retention of Ca and P during the period of most rapid bone mineral accretion, catch-up min- eralization would have been achieved earlier. 9, 11 Recently, investigators have suggested that supplemental Ca and P be continued beyond hospitalization for VLBW infants.12-14 It is unclear whether continued supplementation is necessary if in-hospital efforts at providing appropriate net retention of Ca and P are optimal. Our previous studies, however, in- dicated lower bioavailability of calcium phosphate, com- pared with other sources, for use as a human milk fortifi- er. 10

The objective of this investigation Was to compare, under similar conditions, the net Ca and P retention in VLBW in- fants who received a newly formulated human milk fortifier

I BMC Bone mineral content [

I VLBW Very low birth weight

containing calcium gluconate and glycerophosphate, with Ca and P retention when a previous formulation containing calcium phosphate was used as the principal mineral source. We hypothesized that more bioavailable mineral salts in the new preparation would improve absorption and promote Ca and P retention to meet intrauterine estimates. 15

M E T H O D S

Subjects. Healthy VLBW infants admitted to the neona- tal nurseries of Texas Children's Hospital were enrolled in the first week after birth if they were of appropriate size for gestational age, their mothers intended to breast-feed, and they were free from major cardiopulmonary and gas- trointestinal disorders. Two separate cohorts were studied under similar conditions: 15 infants, who received the newer Ca gluconate-glycerophosphate preparation (group CaGP), were studied from 1991 to 1992, and 11 infants, who received Ca phosphate (group CaTB), were studied in 1988 and 1989. Informed written consent was obtained from parents andthe protocol was approved by the review boards for human research at our institutions.

Study design. Growth, nutritional balance, serum indexes

of mineral and protein status, and bone mineralization were compared in the two groups. The two studies represented the evaluations of current (group CaGP) and previous (group CaTB) formulations (Table I) of Enfamil Human Milk Fortifier (Mead Johnson Nutritional Group, Evans- ville, Ind.). The initiation of enteral nutrition was at the discretion of the attending neonatologist. The protocol for orogastric feeding began with unfortified human milk, 20 to 25 ml • kg -1 • day -1, and advanced daily by that volume, as tolerated, on the basis of clinical examination results and gastric residual volume; When the intake reached 100 ml • kg -1 • day -1, the volume was maintained for 2 days while the concentration was increased by the addition of two packets, and then four packets, of fortifier per 100 ml. Sub- sequently, the intake of fortified human milk was advanced daily to maintain adequate body weight gain. Infants received only their own mothers' milk, and not donor milk, throughout the study. Vitamin D, 400 IU/day, was given to infants after complete enteral nutrition was achieved in group CaTB. Vitamin D was not given to infants in group CaGP. All infants received elemental iron, 2 rag . kg -1 • d -1, 2 weeks after complete enteral nutrition was achieved. The protocol for the lactation management of the mothers and the preparation of milk for feeding has been published previously. 16 The most recent milk available, ei- ther fresh or frozen, was used for fortification.

Anthropometric measurements were determined as the slope of serial determinations obtained during the interval beginning with a milk intake of 100 ml • kg -1 • day -1 and ending with a body weight of 2.0 kg. A 3- to 4-day nutri- tional balance study was initiated 2 to 3 weeks after com- plete enteral nutrition was achieved. One blood sample for the determination of Ca and P concentrations, alkaline phosphatase activity, blood urea nitrogen concentration, and albumin concentration was obtained during the week of the balance study. Bone mineral content of the distal one third of the left radius was measured at the start and the end of the study interval.

Analytic methods. Measurements of body weight were made with electronic scales, length with a rigid board pro- viding immobile head support and adjustable foot support, head circumference with a flexible stainless steel measure, and skin-fold thickness with Lange calipers. The methods for the balance studies have been described, l, 16 Carmine red (100 mg/kg) was given through an orogastric tube at the beginning and the end of the balance period to mark the fecal collection interval. Milk and feces were analyzed for energy by bomb calorimetry (Parr Instrument Co., Moline, Ill.) and for fat by gravimetric methods.16 Nitrogen in milk, urine, and feces was analyzed by semiautomated Kjeldahl methods (Kjeltec; Tecator Inc., Herndon, Va.). Milk and feces were ashed in a muffle furnace at 440 ° C for 14 hours.

The Journal of Pediatrics Schanler and Abrams 4 4 3 Volume 126, Number 3

Table I. Nutrient composition of study human milk

fortifiers (per 4 gm powder)*

CaGP CaTB

Protein (gin) 0.7 0.7 Carbohydrate (gm) 2.7 2.7 Sodium (mg) 7 7 Potassium (mg) 15.6 15.6 Chloride (mg) 17.7 17.7 Calcium (mg) 90 60 Phosphorus (mg) 45 33 Magnesium (mg) 1.0 0 Zinc (#g) 710 310 Copper (#g) 62 80 Manganese (#g) 4.7 9.0 Vitamin A (IU) 950 780 Vitamin D (IU) 210 260 Vitamin K (#g) 4.4 9.1 Biotin (ug) 2.7 0.8 Vitamin C (mg) 12 24

*From manufacturer's product manual.

Urine and ashed milk and fecal samples were analyzed for Ca and magnesium by atomic absorption spectrophotome- try, and P colorimetrically. 1 Serum was analyzed for Ca, P, alkaline phosphatase activity, albumin, and blood urea ni-

trogen values by automatic analyzer (COBAS FARA II; Roche Diagnostic Systems, Inc., Montclair, N.J.). Bone mineral content of the distal one third of the left radius was

determined by single-photon absorptiometry (SP-2 scan- ner; Lunar Radiation Corp., Madison, Wis.).

Data analyses. Group comparisons were made by one-

way analysis of variance. Relationships between variables were tested by linear regression analysis. Confounding variables such as postnatal age, birth weight, gestation, vi- tamin intake, and time of initiation of enteral feedings were tested as potential covariates in the analyses. The level of statistical significance was set at p <0.01. Data are ex- pressed as mean ___ SD.

R E S U L T S

Infants in the two studies were of similar birth weight and gestational age (Table II). There was no difference between groups in the age at which enteral feedings were initiated or when complete enteral feedings were achieved. The study interval duration (39 + 15 vs 46 + 9 days), rate of weight gain (20 + 4 vs 19 + 3 g • kg -1 • day-l), and increments in length and head circumference were similar in the two

groups. Differences between group CaGP and group CaTB were observed for the average of triceps, biceps, subscapu- lar, and thigh skin-fold thicknesses: 0.15 + 0.09 versus 0.36 ___ 0.06 mrn/wk (p <0.001), respectively. Infants in group CaGP required greater fluid intake to achieve targeted growth rates (10 to 15 gm • kg -a • day -a) than did

Table I I , Infant characteristics

Group CaGP Group CaTB (n = 15) (n = 11)

Birth weight (gm) 1148 ___ 263 1041 _+ 117 Gestational age (wk) 28 _+ 2 28 _+ 1 Enteral feeding 10 + 5 14 + 7

initiated (day) Complete enteral 19 _+ 6 22 + 9

feeding (day) Balance study

Postmenstrual age 33 + 2 34 + l (days)

Body weight (gm) 1498 + 250 1507 + 166 Milk intake 172 + 10" 159 ___ 5

(gm • kg -1 • day -1)

Values are expressed as mean -+ SD *p <0.005.

those in group CaTB (p <0.001). The intake of parenteral nutrition and supplemental iron was similar in the two groups. Vitamin D intake differed between groups CaGP and CaTB: 550 versus 1000 IU/day, respectively. When

tested as covariates, differences between groups in postna- tal age, birth weight, gestation, vitamin intake, and time of initiation of enteral feedings did not affect subsequent analyses.

At the time of the balance study, the two groups of infants were similar with respect to body weight and postmenstrual age (Table II). The groups had similar average heart and

respiratory rates and carmine red fecal transit times during the balance study. Differences in the N, Ca, P, and Mg

concentrations in the fortified human milk were reflected in differences in the net intake of these nutrients (Table III). Metabolizable energy intake (gross energy intake - fecal and urinary energy losses) was similar in the two groups. Differences between groups for fat absorption and the net retention of N, Ca, P, and Mg were significant (Table IV).

The net N and Mg retention in group CaGP surpassed, and in group CaTB was equivalent to, the intrauterine accretion rate (N, 23 mmol, or 325 nag • kg -1 • day-l; Mg, 0.11

mmol, or 2.7 mg• kg -1 • day-l). The intrauterine accre-

tion rate for Ca (2.5 to 3.0 retool, or 100 to 120 mg• kg -1 • day -1) was approximated and for P (2.3 retool,

or 70 mg • kg -~ • day -a) was surpassed by infants in group CaGP, but in group CaTB both Ca and P retention were far below the intrauterine accretion rate.15 No differences in Ca

bioavailability were observed. The percentage of Ca ab- sorptionwassimilaringroupsCaGPandCaTB:60% + 13% versus 59% _+ 19%, respectively. The percentage of P absorption differed between groups CaGP and CaTB: 94% _+ 6% versus 74% + 12%, respectively; p <0.001.

The percentage absorption of Ca was significantly corre- lated with the percentage absorption of Mg (r = 0.71; p

4 4 4 Schanler and Abrams The Journal of Pediatrics March 1995

T a b l e III. Intakes of energy and selected nutr ients

Group CoGP Group CoTB (n = 15) (n = 11) p 95% Cl

Gross energy (kJ) 567 + 95 515 + 32 0.095 -3 , 107 Nitrogen (retool) 41 ± 6 33 ± 3 <0.001 5, 12 Fat (gin) 6.0 + 1.3 5.0 ___ 0.8 0.049 0.1, 2 Calcium (mmol) 4.6 ± 0.41 3.0 _+ 0.3 <0.001 1.3, 1.8 Phosphorus (mmol) 4.1 + 0.35 2.1 ± 0.27 <0.001 1.7, 2.2 Magnesium (retool) 0.31 +_ 0.03 0.25 _+ 0.03 0.001 0.03, 0.08

Conversion factors: 1 kJ = 0.239 kcal; 1 mmol N = 14 mg N; 1 mmol Ca = 40 mg Ca; 1 mmol P = 31 mg P; 1 mmol Mg = 24 mg Mg. Values (except p values and CIs) are expressed as mean -+ SD and are in units per kilogram of body weight per day. CI, Confidence interval of differences.

T a b l e IV. Metabol izable energy, fat absorption, and retent ion of N, Ca, P, Mg

Group CaGP Group CaTB (n = 15) (n = 11) p 95% Cl

Metabolizable energy (kJ • kg -1 • day -1) 449 __ 112 459 ± 38 0.800 -77, 58 Fat absorption (%) 64 + 21 84 ± 12 0.007 6, 34 Net retention (mmol • kg -1 • day -1)

Nitrogen 28 ± 5 23 ± 3 0.016 1, 8 Calcium 2.6 + 0.9 1.6 +_ 0.4 0.003 0.4, 1.6 Phosphorus 3.2 + 0.6 1.5 _+ 0.3 <0.001 1.4, 2.1 Magnesium 0.1 + 0.06 0,1 _+ 0.04 0.068 -0.001, 0.1

Conversion factors: 1 kJ = 0.239 kcal; 1 mmol N = 14 mg N; 1 mmol Ca Values (except p values and CIs) are expressed an mean ± SD. CI, Confidence interval of differences.

<0.001) and fat (r = 0.56; p = 0.003) bu t not with age or

the intake of Ca, P, Mg, fat, or vi tamin D. The following

factors significantly (p <0.01) affected net Ca retention: Ca

intake (r = 0.62), P re tent ion (r = 0.63), Mg retent ion

(r = 0.80), and fat absorption (r = 0.48). Fecal excretion of

fat correlated significantly with fecal excretion of Ca

(r = 0.68; p <0.001) and M g (r = 0.62; p = 0.001). As a

proportion of intake, the ur inary concentrat ion of Ca was

lower in group C a G P than in group CaTB: 3% + 2% ver-

sus 7% _+ 6%, respectively (p = 0.02); the ur inary concen-

t ra t ion of P was significantly greater in group C a G P than

in group CaTB: 16% + 9% versus 5% + 6%, respectively

(p = 0.003). Serum P and a lbumin concentrat ions differed

between groups at the t ime of the balance study (Table V).

Significant relationships were observed between net P

re tent ion and the serum P concentra t ion (r = 0.55); p

<0.004). The range in serum Ca concentrat ions was 2.3 to

2.7 m m o l / L (9.1 to 10.9 mg/d l ) . A serum Ca concentrat ion

>2.6 m m o l / L (10.5 m g / d l ) was observed in one infant in

group C a G P and in three infants in group CaTB.

There was no difference between groups in bone mineral

content at the s tar t of the study interval. However, bone

minera l content at hospital discharge differed between

group C a G P and group CaTB: 42 + 4 versus 31 + 5

m g / c m (95% confidence interval of differences = 5 to 16

= 40 mg Ca; 1 mmol P = 31 mg P; 1 mmol Mg = 24 mg Mg.

m g / c m ) , respectively (p <0.001). The age at the t ime of the

B M C differed (9 _+ 3 vs 6 -+ 1 wk in groups C a G P and

CaTB, respectively; p = 0.01). W h e n B M C was adjusted

for age, the difference between groups remained significant

(p = 0.002). The B M C obtained near hospital discharge

correlated positively with net retent ion of Ca (r = 0.58;

p = 0.003), P (r = 0.78; p <0.001), and Mg (r = 0.54;

p = 0.006).

D I S C U S S I O N

These data suggest tha t newer formulat ions available for

the fortification of h u m a n milk provide greater concentra-

tions of Ca and P, and with greater intakes the in t rauter ine

accret ion rates now are approached more closely and even

surpassed. The increased retent ion of Ca and P also was

manifested by greater B M C of the radius in group CaGP

than in group CaTB. Fur thermore , the addit ion of M g to

h u m a n milk resulted in net re tent ion of M g significantly

above in t rauter ine accretion rates.

Of concern to us was tha t despite greater fat intakes the

efficiency of fat absorption in infants fed the newer formu-

lation (group C a G P ) was significantly lower than in group

CaTB infants, and the larger fecal fat losses adversely af-

fected mineral retention. We also observed tha t to achieve

targeted growth rates, infants in group C a G P received sig-

The Journal of Pediatrics Schanler and Abrams 4 4 5 Volume 126, Number 3

Table V. Serum biochemical indexes

Group CaGP Group CaTB (n = 15) (n = 14) p 95% Cl

Sodium (mmol/L) 135 _ 2 135 _+ 1 0.91 -2, 2 Total CO 2 (mmol/L) 20 _+ 3 21 _+ 2 0.48 -3, 2 Ca (mmol/L) 2.5 _+ 0.1 2.6 _+_ 0.1 0.36 -0.1, 0.04 P (retool/L) 2.3 _+ 0.2 1.8 _+ 0.2 <0.001 0.3, 0.7 APA

(ttkat/L) 6.6 + 2.4 5.9 + 0.9 0.32 0.8, 2.2 (IU/L) 396 __+ 142 353 + 53 0.32 -47, 134

Albumin (gm/L) 35 + 4 28 _+ 3 <0.001 3, 10 Hematocrit 0.28 _+ 0.06 0.30 _+ 0.05 0.57 -0.07, 0.04

Conversion factors: 1 mmol Ca = 40 mg Ca; 1 mmol P = 31 mg P. Values (except p values and CIs) are expressed as mean _+ SD. CI, Confidence interval of differences; APA, alkaline phosphatase activity.

nificantly greater fluid intakes than those in group CaTB. Although similarities in weight, length, and head circum- ference measurements were observed, the greater fluid in- takes prescribed to meet these growth rates may be a con-

cern for the more compromised VLBW infant. It concerns us that volumes of milk in excess of 170 ml • kg -1 • day -I

were offered in an attempt to meet intrauterine needs for weight gain and for Ca and P. Although safe for healthy VLBW infants, nutritional recommendations must consider those infants whose fluid intakes are restricted and who are

unable to tolerate the high volumes of milk used in this study.17

Although the groups in our study were not evaluated concurrently, the data on group CaTB serve as a framework

to evaluate the newer formulation, CaGP. The comparisons are valid because similar environmental conditions, study population characteristics, and feeding patterns were prac- ticed in the neonatal units between 1988 and 1992. The re- ported differences between groups in this study would not be affected by differences in times at which the study was conducted.

The net retention data on VLBW infants fed the earlier formulation (group CaTB) are similar to those published by Ehrenkranz et a1.18 That study demonstrated similarities in

nutrient utilization between infants fed preterm formula and those fed fortified human milk; neither Ca nor P reten- tion in infants fed fortified human milk approximated intrauterine accretion. The data, however, demonstrated greater Ca retention in the preterm formula-fed infants. The newer formulation of fortified human milk evaluated in

our study provides a greater content of Ca and P, which re- sults in a greater net mineral retention and BMC.

By improving bone mineralization, we expect to avoid the potential for fractures in VLBW infants.19 The positive re- lation between radius BMC at hospital discharge and net mineral retention suggests that mineral supplementation is essential to the bone development of VLBW infants. Our

findings confirm the data of Horsman et al., 4 who reported

increments in BMC with Ca and P supplementation of for- mula for hospitalized VLBW infants. We speculate that the differences in BMC are a result of the differences in Ca and

P intakes between groups. It is unknown whether other dif- ferences, such as the Mg content of the two fortifiers, could account for these differences. Whether these differences will remain with long-term evaluation is unknown.

We observed that the high net retention of P in group CaGP surpassed the intrauterine accretion rate. is The uri-

nary losses of P also were large but not out of line with those of other milk-fed VLBW infants, l, lo Contrary to the data

on Ca, it appears that P bioavailability from the glycero- phosphate salt is relatively high. Thus, in further modifica- tions of the commercial fortifier, the Ca glycerophosphate portion may need to be reduced in conjunction with a pro- portional increment' in the Ca gluconate component. The use of the glycerophosphate salt also should be considered in other hypophosphatemic states.

The interaction among Ca, Mg, and fat is of interest. Fat intakes were slightly greater in group CaGP because of greater volumes of milk consumed. The higher fat intake was associated with a lower percentage of fat absorption. In

addition, the malabsorbed fat was associated with fecal

losses of Ca and Mg. We speculate that the high concen- tration of minerals in the milk formulation bound with fatty acids and promoted fecal soap formation. Interactions be- tween Ca and fatty acids have been reported during Ca supplementation of human milk for VLBW infants. 2° However, our data indicate differential effects of specifc

calcium salts on fat absorption. Whether this is an osmolar characteristic or a characteristic peculiar to Ca glycero- phosphate awaits laboratory study.

The vitamin D intake also differed between study groups. It is unlikely that the greater vitamin D intake affected Ca absorption; both groups absorbed approximately 60% of the Ca intake. Bronner et al. 21 indicated that preterm infants

4 4 6 Schanler and Abrams The Journal of Pediatrics March 1995

absorb Ca through a vitamin D-independent mechanism;

despite widely differing intakes of vitamin D, similarities in

Ca absorption were observed. The data on the coefficient of

Ca absorption in our study were similar to data summarized

by Bronner et al. and also to data reported by Hillman et

al. 22 Thus, for VLBW infants, further dosage increments

beyond a usual vitamin D intake probably do not improve

Ca absorption.

Our data indicate that given a bioavailable source of Ca,

the VLBW infant apparently absorbs a maximum of 60%

of Ca intake. Simply increasing intake will thus improve net

Ca absorption and retention significantly. Alternatively, Ca

absorption may be increased if fat absorption can be

improved in VLBW infants. We did not observe a difference

in the bioavailability of Ca from glycerophosphate com-

pared with the tribasic phosphate salt. This was unexpected

because initial observations in animals suggest that the or-

ganic salts, such as glycerophosphate, are highly soluble and

bioavailable. 23 Studies of parenteral nutrition indicate that

more Ca and P can be solubilized with the organic, glycero- phosphate salt. 24

Our data also readdress the optimal intake of Mg. 2 Group

CaGP had greater Mg intake (because Mg was added to the

newer formulation) and greater net Mg retention (signifi-

cantly more than the intrauterine accretion rate), whereas

net Mg retention in group CaTB was not significantly dif-

ferent from the intrauterine accretion rate. Although we

have compared our data with estimates for intrauterine ac-

cretion, which, at least for Mg, have not been verified, we

conclude that Mg supplementation of human milk for

VLBW infants may not be necessary.

The average skin-fold thickness in group CaGP was sig-

nificantly less than, and that of group CaTB was signifi-

cantly greater than, published values for triceps skin-fold

thickness measurementsY Body weight gain was similar in

the two groups, and we speculate, therefore, that lean body

mass was greater in group CaGP. The greater net N reten-

tion and serum albumin concentration in group CaGP sup-

port that speculation. These data further suggest that

higher N intakes may be appropriate for growing VLBW

infants.

We conclude that intrauterine accretion rates for Ca and

P can be achieved while feeding VLBW infants fortified

human milk. Newer formulations should be targeted to

reduce the potential for fat malabsorption and promote

better mineral utilization so that fluid intake can be mini-

mized.

We thank Pamela Burns, RN, the neonatal nursery staff of the Neonatal Clinical Research Center and Texas Children's Hospital for their expertise, Charles Imo for laboratory assistance, and J. Kennard Fraley for database management.

R E F E R E N C E S

1. Schanler R J, Garza C, Smith EO. Fortified mothers' milk for very low birth weight infants: results of macromineral balance studies. J PEDIATR 1985;107:767-74.

2. Atkinson SA, Radde IC, Anderson GH. Macromineral bal- ances in premature infahts fed their own mothers' milk or for- mula. J PEDIATR 1983;102:99-106.

3. Greer FR, McCormick A. Improved bone mineralization and growth in premature infants fed fortified own mother's milk. J PEDIATR 1988;112:961-9.

4. Horsman A, Ryan SW, Congdon PJ, Truscott JG, Simpson M. Bone mineral accretion rate and calcium intake in preterm in- fants. Arch Dis Child 1989;64:910-8.

5. Polberger SKT, Axelsson IA, R~iih/i NCR. Growth of very low birth weight infants on varying amounts of human milk pro- tein. Pediatr Res 1989;25:414-9.

6. Abrams SA, Schanler R J, Garza C. Bone mineralization in former very low birth weight infants fed either human milk or commercial formula. J PEDIATR 1988;112:956-60.

7. Lucas A, Brooke OG, Baker BA, Bishop N, Morley R. High alkaline phosphatase activity and growth in preterm neonates. Arch Dis Child 1989;64:902-9.

8. Abrams SA, Schanler R J, Tsang RC, Garza C. Bone miner- alization in former very low birth weight infants fed either hu- man milk or commercial formula: one-year follow-up observa- tion. J PEDIATR 1989;114:1041-4.

9. Schanler R J, Burns PA, Abrams SA, Garza C. Bone miner- alization outcomes in human milk-fed preterm infants. Pedi- atr Res 1992;31:583-6.

10. Schanler RJ, AbramsSA, Garza C. Bioavailabilityofcalcium and phosphorus in human milk fortifiers and formula for very low birth weight infants. J PEDIATg 1988;113:95-100.

11. Congdon P J, Horsman A, Ryan SW, Truscott JG, Durward H. Spontaneous resolution of bone mineral depletion in preterm infants. Arch Dis Child 1990;65:1038-42.

12. Friel JK, Andrews WL, Matthew JD, McKim E, French S, Long DR. Improved growth of very low birthweight infants. Nutr Res 1993;13:611-20.

13. Lucas A, Bishop N J, King F J, Cole TJ. Randomised trial of nutrition for preterm infants after discharge. Arch Dis Child 1992;67:324-7.

14. Chan GM. Growth and bone mineral status of discharged very low birth weight infants fed different formulas or human milk. J PEDIATR 1993;123:439-43.

15. Ziegler EE, O'Donnell AM, Nelson SE, Fomon SJ. Body composition of the reference fetus. Growth 1976;40:329- 41.

16. Schanler RJ, Garza C, Nichols BL. Fortified mothers' milk for very low birth weight infants: results of growth and nutrient balance studies. J PEDIATR 1985;107:437-45.

17. American Academy of Pediatrics Committee on Nutrition: Nutritional needs of low-birth-weight infants. Pediatrics 1985;75:976-86.

18. Ehrenkranz RA, Gettner PA, Nelli CM. Nutrient balance studies in premature infants fed premature formula or fortified preterm human milk. J Pediatr Gastroenterol Nutr 1989;8:58- 67.

19. Koo WWK, Sherman R, Succop P, et al. Sequential bone mineral content in small preterm infants with and without fractures and rickets. J Bone Miner Res 1988;3:193-7.

20. Chappell JE, Clandinin MT, Kearney-Volpe C, Reichman B, Swyer PW. Fatty acid balance studies in premature infants fed

The Journal of Pediatrics .4bzug, Loeffelholz, and Rotbart 4 4 7 Volume 126, Number 3

human milk or formula: effect of calcium supplementation. J PEDIATR 1986;108:439-47.

21. Bronner F, Salle BL, Putet G, Rigo J, Senterre J. Net calcium absorption in premature infants: results of 103 metabolic bal- ance studies. Am J Clin Nutr 1992;56:1037-44.

22. Hillman LS, Johnson LS, Lee DZ, Vieira NE, Yergey AL. Measurement of true absorption, endogenous fecal excretion, urinary excretion, and retention of calcium in preterm infants by using a dual-tracer, stable isotope method. J PED1ATR 1993; 123:444-56.

23. Atkinson SA, Shah JK, Webber CE, Gibson 1L, Gibson RS. A multi-element isotopic tracer assessment of true fractional

24.

25.

absorption of minerals from formula with additives of calcium, phosphorus, zinc, copper, and iron in young piglets. J Nutr 1993;123:1586-93. Hanning RM, Mitchell MK, Atkinson SA. In vitro solubility of calcium glycerophosphate versus conventional mineral salts in pediatric parenteral nutrition solutions. J Pediatr Gastroen- terol Nutr 1989;9:67-72. Georgieff MK, Amarnath UM, Mills MM. Determinants of arm muscle and fat accretion during the first postnatal month in preterm newborn infants. J Pediatr Gastroenterol Nutr 1989;9:219-24.

Clinical and laboratory observations

Diagnosis of neonatal enterovirus infection by polymerase chain reaction

Mark J. Abzug, MD, Michael Loeffelholz, PhD, and Harley A. Rotbart, MD

From the Departments of Pediatrics (Infectious Diseases) and Microbiology, University of Col- orado School of Medicine and Children's Hospital, Denver; and Roche Molecular Systems, Branchburg, New Jersey

A 5-hour color imetr ic po iymerase chain react ion (PCR) assay was more sensi- tive than viral culture in identifying viral infect ion in initial serum (13/16 vs 5/16; p = 0.008) and urine (10/16 vs 5/16; p = 0.2) specimens from 16 enterovirus- in fected newborn infants, and rema ined more sensitive throughout their ill- nesses. C o m b i n e d sensitivity of serum and urine PCR was 14 of 16 (88%). Results of all acu te -phase PCR assays of serum and urine from four neonates with cul- tures negat ive for enterovirus were also negat ive . PCR assay of serum and urine faci l i tates rapid, a c c u r a t e diagnosis of neonata l enterovirus infections. (J PEDI" ATR 1995;126:447-50)

Neona ta l enterovirus infections are a common cause of ill-

ness and hospital izat ion 1 with significant morbidi ty and

morta l i ty rates. 2 Dist inguishing EV infection in a sick new-

born infant f rom other infectious and noninfectious diseases

tha t may have similar manifes ta t ions is challenging. 2 Cur-

rently, laboratory diagnosis depends on tissue cul ture isola-

Supported by grant RR-69 from the General Clinical Research Centers Program of the Division of Research Resources, National Institutes of health; the Kempe Research Center, Children's Hos- pital, Denver, Colo.,; and Roche Molecular Systems. Dr. Loeffel- holz is a full-time employee of, and Dr. Rotbart a paid consultant for, Roche Molecular Systems, whose products were used for the work discussed in the article.

Presented in part at the 10th Annual Clinical Virology Symposium and Annual Meeting of the Pan American Group for Rapid Viral

Diagnosis, Clearwater, FL, April 1994 and the 34th Interscience Conference on Antimicrobial Agents and Chemotherapy, Orlando, FL, Oct. 1994.

Submitted for publication July 22, 1994; accepted Sept. 15, 1994.

Reprint requests: Mark J. Abzug, MD, Infectious Diseases, Box B055, Children's Hospital, 1056 E. 19th Ave., Denver, CO 80218.

Copyright © 1995 by Mosby-Year Book, Inc. 0022-3476/95/$3.00 + 0 9/24/60612