Bone Mineral Content of Term and Premature Infants
Longitudinal Changes in the
William B. Pittard III, MD; Kitty M. Geddes, Millage C. Miller, PhD; Bruce W. Hollis, PhD \s=b\ With the use of photon absorptiometry, bone mineralization was measured
at birth and 8 and 16 weeks after delivery in 12 very-low-birth-weight premature (mean \m=+-\SD gestational age, 31 \m=+-\1.5 weeks) infants who required minimal medical support. Simultaneously, 19 healthy term infants were studied. Throughout the study, each neonate received modified 84\p=n-\kJ/30mL formula containing no added calciferol. The recommended daily allowance (400 IU) of calciferol was given to each infant as an oral supplement. Serum 25-hydroxyvitamin D, calcium, phosphorus, and parathyroid hormone concentrations were monitored biweekly and were normal. Bone mineral content and bone width significantly differed at birth between the term and premature infants. However, by 16 weeks after delivery, the premature infants had exceeded the bone mineral status of the term infants at birth, and their bone mineral content was not significantly lower than that of the term infants. These data indicate improved bone mineralization as compared with previously reported data from very-low-birth-weight neonates.
(AJDC. 1990;144:36-40)
? ickets of prematurity or osteopenia isa well-recognized disorder of bone mineralization among very-low-birthweight (VLBW) neonates, particularly those with chronic illness that requires extended and complex medical therapy. Currently, its cause is thought to be multifactorial, with a deficiency in calciAccepted for publication August 1, 1989. From the Departments of Pediatrics (Drs Pittard and Hollis and Ms Geddes) and Biometry (Ms Sutherland and Dr Miller), Medical University of South Carolina, Charleston. Reprint requests to Division of Neonatology, Department of Pediatrics, Medical University of South Carolina, 171 Ashley Ave, Charleston, SC
29425 (Dr Pittard).
BSN; Susan E. Sutherland, MS;
and phosphorus intake representing major causative factor. Several groups of investigators have hypothe¬ sized that premature neonates with minimal illness who received no parenteral nutrition would demonstrate im¬ proved bone mineralization compared with VLBW neonates who required more extensive medical support.17 To test this hypothesis, a longitudinal as¬ sessment of bone mineralization in VLBW neonates who required only minimal medical support was under¬ taken. um
a
PATIENTS AND METHODS
body weight was not measured for each neo¬ nate throughout the entire 16-week study, for the hospitalized premature infants, oral intake until discharge from the hospital was recorded and used to derive mineral intakes from manufacturer's data sheets. To facilitate a consistent daily intake of calciferol, each neonate received throughout the study a modified 84-kJ/30 mL formula that contained 0.51 g/L of calcium, 0.39 g/L of phosphorus, and no added calciferol. The rec¬ ommended daily allowance (400 IU) of calci¬ ferol was then given to each infant in a glycerol carrier as a daily oral supplement. Since compliance with calciferol dosage is difficult to document after discharge from the hospi¬
25-hydroxyvitamin D concentra¬ monitored biweekly. These 25-hydroxyvitamin D determinations were performed as previously described.9 Serum calcium and phosphorus levels as well as parathyroid hormone (PTH) concentrations were measured biweekly throughout the study.10"12 Normal adult values (range) for PTH and 25-hydroxyvitamin D using these assays in our laboratory are 0.29 to 0.85 ng/mL and 25 to 125 nmol/L, respective¬ ly. Because of volume requirements for lab¬ tal,
serum
tions
Thirty-one study neonates, 19 term and 12 premature infants,
were
enrolled in the
study within 24 hours of birth. Women with a history of a normal term pregnancy were chosen at random from those delivering at the Medical University of South Carolina, Charleston, between January 1986 and March 1988. Mothers of premature infants delivered during the same period were also contacted at random after the presence of minimal lung disease was clinically estab¬ lished in their infant. The investigation was approved by the University Human Re¬ search Committee, and written parental con¬ sent was obtained for each newborn before the study began. Neonatal gestational age was initially de¬ termined from the history of the mother's last menstrual period and was confirmed with a Dubowitz examination within 36 hours of birth. Weight for both term and prema¬ turely delivered infants was determined bi¬ weekly with a standard balance scale. The premature infants were generally fed via orogastric tube supplemented with intra¬ venous glucose water during the first 2 weeks of life; otherwise all infants were fed by nipple. No infant received parenteral fluid containing amino acids, calcium, or phos¬ phate salts or lipid. Although the exact daily volume of formula intake per kilogram of
were
oratory assessment, the serum ionized calci¬ concentration was measured only in term infants, and the total serum calcium level was um
measured in premature infants. Normal adult values for ionized calcium using the assay in our laboratory are 1.20 to 1.30 mmol/L. Phosphorus concentrations us¬ ing the assay in our laboratory are consistent with values previously reported for neo¬
nates.13
Bone mineral content (BMC) and bone width (BW) were measured with the photon
absorptiometric system (Lunar Radiation, Ine, Madison, Wis), in which a collimated, 3-mm-diameter photon beam from a low-ac¬ tivity (radioactive iodine 125) source was passed beneath the distal third of the right radius. The site of bone mineral assessment was at the junction of the middle and distal thirds of the radius. This region of the radius has the least variation in mineral content,
Downloaded From: http://archpedi.jamanetwork.com/ by a University of California - San Diego User on 06/07/2015
thus minimizing error generated by measur¬ ing slightly different bone sites throughout a longitudinal study.1 Measurements of BMC and BW were made within the first 2 days of life and subsequently at 8 and 16 weeks of age. Reproducibility (without repositioning the arm) for the five scans performed for each determination of BMC and BW was reflected by a mean ± SD coefficient of variation of 1.1±0.7% and 0.8±0.5%, respectively. The "repositioning error" determined by repeat¬ ing measurements of BMC and BW after repositioning the arm is small. The correla¬ tion coefficients between measurements be¬ fore and after repositioning of the arm were .99 for both BMC and BW. Infant weights and gestational ages were correlated with the bone mineralization de¬ terminations. Furthermore, figures demon¬ strating BMC and BW changes from 29 to 58 weeks after conception were constructed from the birth values obtained from all 31 study infants plus the 8- and 16-week values from only the 19 full-term infants. A sigmoidal model was selected consistent with nor¬ mal neonatal growth. Specifically, the data were fitted by a cubic polynomial using leastsquares error estimation procedure. The 95% confidence intervals were established about the predicted polynomial to provide normal 95% upper and lower limits for BMC and BW. The 8- and 16-week mineralization values from the 12 premature infants were not used in developing these limits as these data may not reflect "normal" bone changes. They were, however, fit by a second cubic polynomial and plotted with the birth values for comparison with the limits for normal mineralization. All data were analyzed using analysis of variance methods with repeated measures on individual patients. These analyses were followed by standard Student's t tests for independent groups with Bonferroni adjust-
ments for multiple comparisons. Pearson correlation coefficients were used to deter¬ mine the association of variables. Curvilin¬ ear and linear regression lines relating the dependent variables, BMC and BW, to the independent variable, weeks after concep¬ tion, were calculated. The best fit was deter¬ mined by R2 values for the specific model used.11 Confidence intervals were calculated using the predicted values from the model ±1SD.
RESULTS The study patients included 1 term and 1 premature white infant and 18 term and 11 premature black infants. Seven of the term and 5 of the prema¬ ture infants were male. Ofthe 12 prema¬ ture infants, 3 were electively delivered secondary to maternal preeclampsia, and 9 were delivered after the spontane¬ ous but premature rupture of the amniotic membranes. The mean ( ± SD) gestational age for the term and premature study infants was 39.8 ± 1.1 weeks and 31.0 ±1.5 weeks, respectively. Similar¬ ly, the mean (±SD) birth weight for these infants was 3416 ±363 g and 1281 ±177 g, respectively. All infants
Table
1.—Daily Calcium and Phosphorus
of appropriate size for their gestational age15 and free of congenital abnor¬ malities. Furthermore, their postnatal growth was consistent with reported normal values.16 All infants were at or above birth weight and were being solely enterally fed 420 kJ/kg or more daily by 2 weeks of age. The mean daily calcium and phos¬ phorus intakes per kilogram of body weight for the hospitalized premature infants are shown in Table 1. The bi¬ weekly serum calcium, phosphorus, and PTH concentrations for all infants throughout the entire study are shown in Table 2. Although the phosphorus concentrations were significantly (P
Longitudinal Changes in the
William B. Pittard III, MD; Kitty M. Geddes, Millage C. Miller, PhD; Bruce W. Hollis, PhD \s=b\ With the use of photon absorptiometry, bone mineralization was measured
at birth and 8 and 16 weeks after delivery in 12 very-low-birth-weight premature (mean \m=+-\SD gestational age, 31 \m=+-\1.5 weeks) infants who required minimal medical support. Simultaneously, 19 healthy term infants were studied. Throughout the study, each neonate received modified 84\p=n-\kJ/30mL formula containing no added calciferol. The recommended daily allowance (400 IU) of calciferol was given to each infant as an oral supplement. Serum 25-hydroxyvitamin D, calcium, phosphorus, and parathyroid hormone concentrations were monitored biweekly and were normal. Bone mineral content and bone width significantly differed at birth between the term and premature infants. However, by 16 weeks after delivery, the premature infants had exceeded the bone mineral status of the term infants at birth, and their bone mineral content was not significantly lower than that of the term infants. These data indicate improved bone mineralization as compared with previously reported data from very-low-birth-weight neonates.
(AJDC. 1990;144:36-40)
? ickets of prematurity or osteopenia isa well-recognized disorder of bone mineralization among very-low-birthweight (VLBW) neonates, particularly those with chronic illness that requires extended and complex medical therapy. Currently, its cause is thought to be multifactorial, with a deficiency in calciAccepted for publication August 1, 1989. From the Departments of Pediatrics (Drs Pittard and Hollis and Ms Geddes) and Biometry (Ms Sutherland and Dr Miller), Medical University of South Carolina, Charleston. Reprint requests to Division of Neonatology, Department of Pediatrics, Medical University of South Carolina, 171 Ashley Ave, Charleston, SC
29425 (Dr Pittard).
BSN; Susan E. Sutherland, MS;
and phosphorus intake representing major causative factor. Several groups of investigators have hypothe¬ sized that premature neonates with minimal illness who received no parenteral nutrition would demonstrate im¬ proved bone mineralization compared with VLBW neonates who required more extensive medical support.17 To test this hypothesis, a longitudinal as¬ sessment of bone mineralization in VLBW neonates who required only minimal medical support was under¬ taken. um
a
PATIENTS AND METHODS
body weight was not measured for each neo¬ nate throughout the entire 16-week study, for the hospitalized premature infants, oral intake until discharge from the hospital was recorded and used to derive mineral intakes from manufacturer's data sheets. To facilitate a consistent daily intake of calciferol, each neonate received throughout the study a modified 84-kJ/30 mL formula that contained 0.51 g/L of calcium, 0.39 g/L of phosphorus, and no added calciferol. The rec¬ ommended daily allowance (400 IU) of calci¬ ferol was then given to each infant in a glycerol carrier as a daily oral supplement. Since compliance with calciferol dosage is difficult to document after discharge from the hospi¬
25-hydroxyvitamin D concentra¬ monitored biweekly. These 25-hydroxyvitamin D determinations were performed as previously described.9 Serum calcium and phosphorus levels as well as parathyroid hormone (PTH) concentrations were measured biweekly throughout the study.10"12 Normal adult values (range) for PTH and 25-hydroxyvitamin D using these assays in our laboratory are 0.29 to 0.85 ng/mL and 25 to 125 nmol/L, respective¬ ly. Because of volume requirements for lab¬ tal,
serum
tions
Thirty-one study neonates, 19 term and 12 premature infants,
were
enrolled in the
study within 24 hours of birth. Women with a history of a normal term pregnancy were chosen at random from those delivering at the Medical University of South Carolina, Charleston, between January 1986 and March 1988. Mothers of premature infants delivered during the same period were also contacted at random after the presence of minimal lung disease was clinically estab¬ lished in their infant. The investigation was approved by the University Human Re¬ search Committee, and written parental con¬ sent was obtained for each newborn before the study began. Neonatal gestational age was initially de¬ termined from the history of the mother's last menstrual period and was confirmed with a Dubowitz examination within 36 hours of birth. Weight for both term and prema¬ turely delivered infants was determined bi¬ weekly with a standard balance scale. The premature infants were generally fed via orogastric tube supplemented with intra¬ venous glucose water during the first 2 weeks of life; otherwise all infants were fed by nipple. No infant received parenteral fluid containing amino acids, calcium, or phos¬ phate salts or lipid. Although the exact daily volume of formula intake per kilogram of
were
oratory assessment, the serum ionized calci¬ concentration was measured only in term infants, and the total serum calcium level was um
measured in premature infants. Normal adult values for ionized calcium using the assay in our laboratory are 1.20 to 1.30 mmol/L. Phosphorus concentrations us¬ ing the assay in our laboratory are consistent with values previously reported for neo¬
nates.13
Bone mineral content (BMC) and bone width (BW) were measured with the photon
absorptiometric system (Lunar Radiation, Ine, Madison, Wis), in which a collimated, 3-mm-diameter photon beam from a low-ac¬ tivity (radioactive iodine 125) source was passed beneath the distal third of the right radius. The site of bone mineral assessment was at the junction of the middle and distal thirds of the radius. This region of the radius has the least variation in mineral content,
Downloaded From: http://archpedi.jamanetwork.com/ by a University of California - San Diego User on 06/07/2015
thus minimizing error generated by measur¬ ing slightly different bone sites throughout a longitudinal study.1 Measurements of BMC and BW were made within the first 2 days of life and subsequently at 8 and 16 weeks of age. Reproducibility (without repositioning the arm) for the five scans performed for each determination of BMC and BW was reflected by a mean ± SD coefficient of variation of 1.1±0.7% and 0.8±0.5%, respectively. The "repositioning error" determined by repeat¬ ing measurements of BMC and BW after repositioning the arm is small. The correla¬ tion coefficients between measurements be¬ fore and after repositioning of the arm were .99 for both BMC and BW. Infant weights and gestational ages were correlated with the bone mineralization de¬ terminations. Furthermore, figures demon¬ strating BMC and BW changes from 29 to 58 weeks after conception were constructed from the birth values obtained from all 31 study infants plus the 8- and 16-week values from only the 19 full-term infants. A sigmoidal model was selected consistent with nor¬ mal neonatal growth. Specifically, the data were fitted by a cubic polynomial using leastsquares error estimation procedure. The 95% confidence intervals were established about the predicted polynomial to provide normal 95% upper and lower limits for BMC and BW. The 8- and 16-week mineralization values from the 12 premature infants were not used in developing these limits as these data may not reflect "normal" bone changes. They were, however, fit by a second cubic polynomial and plotted with the birth values for comparison with the limits for normal mineralization. All data were analyzed using analysis of variance methods with repeated measures on individual patients. These analyses were followed by standard Student's t tests for independent groups with Bonferroni adjust-
ments for multiple comparisons. Pearson correlation coefficients were used to deter¬ mine the association of variables. Curvilin¬ ear and linear regression lines relating the dependent variables, BMC and BW, to the independent variable, weeks after concep¬ tion, were calculated. The best fit was deter¬ mined by R2 values for the specific model used.11 Confidence intervals were calculated using the predicted values from the model ±1SD.
RESULTS The study patients included 1 term and 1 premature white infant and 18 term and 11 premature black infants. Seven of the term and 5 of the prema¬ ture infants were male. Ofthe 12 prema¬ ture infants, 3 were electively delivered secondary to maternal preeclampsia, and 9 were delivered after the spontane¬ ous but premature rupture of the amniotic membranes. The mean ( ± SD) gestational age for the term and premature study infants was 39.8 ± 1.1 weeks and 31.0 ±1.5 weeks, respectively. Similar¬ ly, the mean (±SD) birth weight for these infants was 3416 ±363 g and 1281 ±177 g, respectively. All infants
Table
1.—Daily Calcium and Phosphorus
of appropriate size for their gestational age15 and free of congenital abnor¬ malities. Furthermore, their postnatal growth was consistent with reported normal values.16 All infants were at or above birth weight and were being solely enterally fed 420 kJ/kg or more daily by 2 weeks of age. The mean daily calcium and phos¬ phorus intakes per kilogram of body weight for the hospitalized premature infants are shown in Table 1. The bi¬ weekly serum calcium, phosphorus, and PTH concentrations for all infants throughout the entire study are shown in Table 2. Although the phosphorus concentrations were significantly (P
