Journal of Perinatology (2012) 32, 218–226
r 2012 Nature America, Inc. All rights reserved. 0743-8346/12
www.nature.com/jp
ORIGINAL ARTICLE
Ultrasound for the assessment of bone quality in preterm
and term infants
B Rack1, E-M Lochmüller1, W Janni2, G Lipowsky3, I Engelsberger4, K Friese1 and H Küster5
1
Frauenklinik Innenstadt, Klinikum der Ludwig-Maximilians-Universita¨t, Mu¨nchen, Germany; 2Frauenklinik der Heinrich-HeineUniversita¨t, Düsseldorf, Germany; 3Kinderklinik, Klinikum der Ludwig-Maximilians-Universita¨t, Mu¨nchen, Germany, 4Kinderklinik
und Poliklinik der Technischen Universita¨t, Kinderklinik Schwabing, Mu¨nchen, Germany and 5Universita¨tskinderklinik,
Ernst-Moritz-Arndt-Universita¨t, Greifswald, Germany
Objective: As 80% of intrauterine bone mineralization takes place during
the last trimester of pregnancy, preterm infants should be supplemented
postnatally with optimal doses of calcium, phosphate and vitamin D.
Calcium and phosphate excretion in the urine may be used to monitor
individual mineral requirements, but are sometimes difficult to interpret.
The objective of this study was to assess the value of quantitative ultrasound
(QUS) for the analysis of bone status in neonates.
Study Design: All admissions to three independent tertiary neonatal
intensive care units were studied. In 172 preterm and term infants with a
gestational age between 23 and 42 weeks (mean 33.8±5.0) and a birth
weight from 405 to 5130 g (mean 2132±1091 g) bone status was
evaluated prospectively by quantitative ultrasound velocity using a
standardized protocol. Infants were followed in regular intervals up to
their first discharge home. While measurements were conducted in weekly
intervals initially (n ¼ 55), 2-week intervals were regarded as sufficient
thereafter due to limited changes in QUS values within the shorter period.
Infants with a birth weight below 1500 g were followed during outpatient
visits until up to 17 months of age.
Result: The intra-individual day-to-day reproducibility was 0.62%. QUSvalues from the first week of life correlated significantly with gestational
age and birth weight (r ¼ 0.5 and r ¼ 0.6; P<0.001). Small-forgestational-age infants showed lower values for QUS than appropriate-forgestational-age infants allowing for their gestational age. Follow-up
measurements correlated positively with age and weight during the week
of measurement (r ¼ 0.2 and r ¼ 0.4; P ¼ 0.001). Comparing bone
quality at 40 weeks of age in infants born at term versus infants born at
24 to 28 weeks, preterm infants showed significantly lower QUS than term
infants (P<.0001).There was a significant correlation of QUS with serum
alkaline phosphatase (P ¼ 0.003), the supplementation with calcium,
phosphate and vitamin D (P< 0.001 each), as well as risk factors for a
Correspondence: Dr B Rack, Department of Gynecology and Obstetrics, Klinikum Innenstadt,
Ludwig-Maximilians-Universität München, Maistr. 11, 80337 München, Germany.
E-Mail: [email protected]
Received 27 December 2010; revised 11 May 2011; accepted 11 May 2011; published online
16 June 2011
reduced bone mineralization. No correlation was found between QUS and
calcium or phosphate concentration in serum or urine.
Conclusion: QUS is a highly reproducible, easily applicable and
radiation-free technique that can be used to monitor bone quality in
individual newborns. Further prospective randomized-trials are necessary
to evaluate, if therapeutic interventions based on QUS are able to prevent
osteopenia of prematurity.
Journal of Perinatology (2012) 32, 218–226; doi:10.1038/jp.2011.82;
published online 16 June 2011
Keywords: quantitative ultrasound; preterm infants; neonate; bonedensity; osteopenia
Introduction
As 80% of bone mineralization in neonates occurs throughout the
last trimester, it is not surprising that 39% of all neonates, born
prematurly with a birth weight of less than 1500 g, suffer from
osteopenia.1 Because optimal postnatal nutritional
supplementation may modify the degree of osteopenia, the clinical
management would profit from a non-invasive and easily
applicable method that is well tolerated even by extremely
immature preterm infants and capable of providing quantitative
information on bone mineral status. However, no generally
accepted method is yet available for this purpose, especially as most
techniques used for bone-density measurement in adults require
ionizing radiation. Currently used methods for diagnosis of
osteopenia in preterm infants are based on either serum values for
calcium, phosphate and alkaline phosphatase (AP), or in some
countries, on excretion of calcium and phosphate in the urine.
Although the first method is invasive and adds to the constant
iatrogenic blood loss in very-low-birth-weight infants, the second is
non-invasive but much less reproducible even if results are
corrected for urine concentration.2,3 In addition, both methods
reflect bone mineral turnover that does not necessarily correlate
with bone mineral density or mechanical strength. Although
Bone quality in neonates by ultrasound
B Rack et al
219
conventional X-rays are not sensitive enough to guide calcium and
phosphate supplementation, dual-energy X-ray absorptiometry
(DXA) is a recognized method to assess the skeletal status in the
neonate.4 But again, this technique is subject to several limitations.
Apart from limited availability, need for neonatal transport, and
concerns on the cumulative radiation dosage associated with
repeated measurements, the variables provided by DXA (bone
mineral content and areal bone mineral density) are substantially
confounded by the changes in size and shape of the skeleton as
well as in the amount of soft tissue occurring during the rapid
growth of neonates.
In contrast, quantitative ultrasound (QUS) parameters were
shown to depend on bone density, bone mechanical properties5 and
trabecular bone microstructure.6 Recent studies have suggested that
QUS parameters are also associated with other factors of bone
strength, such as the geometry and porosity of cortical bone.7 Both
retrospective and prospective clinical studies have shown that QUS
has similar abilities of predicting fracture at the proximal femur
and spine in adults as DXA.8 Although QUS is now widely used in
adults and represents a relatively established method for evaluating
bone mineral status in the context of diagnosing osteoporosis, QUS
has only been employed in mostly small study samples of newborn
infants and children.9–17 However, none of these studies provide
longitudinal data in regular intervals on QUS measurements in
preterm infants simultaneously accounting for growth changes and
increase of soft tissue. Therefore, we initiated a trial for the
evaluation of QUS following both preterm and term infants in
predefined intervals until the age of 17 months. By using a
standardized measurement technique, we could minimize the effect
of soft tissue growth and thus assess factors influencing early bone
development.
The objective of this study was to evaluate the usefulness of QUS
to monitor mineral supplementation in a large number of
neonates comprising the whole spectrum of term and preterm
infants. Specific objectives were to correlate QUS measurements in
newborns with their gestational age and birth weight, follow-up
bone development by QUS after birth and correlate QUS
measurements with laboratory measures of bone metabolism,
nutritional supplies and risk factors of osteopenia of prematurity.
Methods
QUS measurement
Ultrasound transmission velocity (m s1) was measured using the
Osteoson KIV (Minhorst, Meudt, Germany). The device has two
probes of 10 mm diameter, one transmitting broadband ultrasound
of 0.1 to 0.6 MHz, the other serving as receiver. These probes are
positioned on a caliper that allows the attached computer to
automatically calculate the exact distance between the two probes.
Before each measurement, the device was calibrated against air
to give a velocity of 346 m s1. In each neonate, separate
measurements were done at four different locations on one side of
the patient’s body that was randomly chosen before the first
measurement. Measurement locations were the middle of the
diaphysis of the long bones (humerus, radius/ulna combined,
femur and tibia) because they calcify early in fetal development
and their cylindrical shape and homogeneous bone density allows
repositioning. For each of the four locations, 50 measurements were
done within 1 to 2 min without change of position. Of these
50 measurements, the median of the 10 fastest values was calculated
to include only those travelling through bone. Subsequently, the
mean of all four medians from all four measurement locations was
determined and used for further analysis.
In the measurement of neonates, fixed probe distances of 20 or
35 mm were used according to their size. The remaining distance
between the probes and the skin of the neonate was filled with a
precursor with a velocity identical to soft tissue in neonates
(1560 m s1) (Figure 1). This velocity was calculated as the mean
of 55 measurements of soft tissue conducted at the upper leg in 16
infants from 604 to 3060 g (1624±559 g). By this calculation
model, we selected values most likely representing bone, as
ultrasound waves propagate fastest in dense tissue. The decision for
this measuring method was based on experiments using isolated
bones, where we observed a decrease of transmission velocity by
61% when the surrounding soft tissue, simulated by a gel layer, was
increased from 5 to 20 mm. This increase in soft tissue is one of
the most pronounced effects of body growth in neonates. Therefore,
we limited this effect of soft tissue growth by maintaining the
amount of surrounding tissue at a fixed valueFdepending on the
size of the infantFat 20 or 35 mm. Consequently, values for 20
and 35 mm distances are not comparable.
Patients
After receiving informed consent from the parents, 602
measurements were performed on 172 newborn infants hospitalized
in three independent level three neonatal care units and one
Figure 1 Quantitative ultrasound measurement with precursor at the forearm of
a neonate.
Journal of Perinatology
Bone quality in neonates by ultrasound
B Rack et al
220
Table 1 Patient characteristics
Number of patients
Gestational age (weeks)
Birth weight (g)
SGA (n; %)
Gender (F/M)
Endotracheal ventilation (days)
Nasal ventilation (days)
Parental nutrition (days)
Steroid treatment (days)
Treatment with diuretics (days)
Preterm infants
Term infants
122
32 (24–36)
1570 (405–4010)
31 (26%)
65/57
5 (1–55)
10 (1– 315)
8 (1–64)
15 (1–50)
19 (3–80)
50
40 (37–42)
3410 (1742–5130)
8 (16%)
16/34
0
1
3 (1–7)
0
0
Abbreviations: F, female; M, male; SGA, small-for-gestational-age.
Values are median (range) for gestational age and birth weight.
healthy baby nursery at the Ludwig-Maximilian-University and the
Technical University of Munich (Table 1), if parents gave informed
consent and study personnel was available. Neonates had a
gestational age between 24 and 42 weeks (mean±s.d: 33.8±5.0)
and a birth weight between 405 and 5130 g (2132±1091 g);
40 neonates (23%) were small-for-gestational-age (birth weight
p10th percentile) (SGA). Measurements during the first week
of life were available in 114 neonates, whereas 58 measurements
were started after the first week.
Neonates were fed with a combined parenteral and oral regimen
starting immediately after birth. Parenteral protein and fat in 10%
glucose were increased within 5 days from 1 g kg1 per day up to
2.5 and 3.5 g kg1 per day, respectively. Adjustments for glucose,
electrolytes and vitamins were made according to the infants’ needs
and international guidelines. As oral nutrition, preferably breast
milk was fed enriched with 400 to 800 I.E. of vitamin D and
fortifier (FM85, Nestle, Frankfurt, Germany). If not available,
breast milk was substituted by preterm formula. The median
weekly intake of calcium was 20 mmol kg1 per week, of
phosphate 16 mmol kg1 per week and of vitamin D 2151 I.E.
per week.
Measurements of QUS were repeated in 7-day intervals for the
initial 55 infants and in and 14-day intervals thereafter. On the day
of these follow-up measurements, the neonates’ weights ranged
from 600 to 7480 g (2078±979 g) and their postconceptional
maturity, calculated as gestational plus postnatal age, from 25 to
68 weeks (34.5±5.2) (Table 3). Infants were followed until their
discharge home. Preterm infants with a birth weight below 1500 g
were followed until up to 17 months of postconceptional age. The
median time of observation between the first and last measurement
was 5.8 weeks (1 to 45) and the mean number of examinations 3
per neonate (1 to 13) (Table 4). To determine day-to-day
reproducibility of the ultrasound Osteoson KIV, QUS was measured
on two consecutive days in five neonates chosen at random from
Journal of Perinatology
the collective described above with a median age of 29 weeks (27.3
to 35.1).
Detailed information on amount and composition of calcium,
phosphate and vitamin D supplementation of the infants, as well as
calcium and phosphate concentrations in serum and urine and AP
in serum were acquired from the patients’ charts. All clinical
chemistry assays were conducted with the Integra 800 (Roche
Diagnostics, Basel, Switzerland). The laboratory measurements
were conducted at regular intervals according to clinical routine of
the individual unites, but were not a mandatory part of the study
protocol.
Statistical analysis
Cases with missing clinical or laboratory data and QUS
measurements under treatment with diuretics (n ¼ 21) were
excluded from the analysis concerned. For the correlation of
QUS with laboratory or clinical risk factors for reduced bone
mineralization, change of QUS per week was calculated (alteration
of QUS over time (DQUS)), to analyze relative QUS changes
independent of absolute postconceptional age and weight. All data
were tested for normal distribution. Pearson and Spearman
correlation coefficients were calculated for normally distributed and
not normally distributed data, respectively. To compare categorical
variables, the w2-test was employed. The two-tailed t-test was used
to calculate the differences of the mean of independent samples
that had continuous variables. In summary, 10 variables were
analyzed for their influence on QUS values. Cox’s regression
analysis was used for multivariate analyses, to disclose the
significant factors from univariate analysis that determine bone
quality. The variables were entered forward stepwise. P values
<0.05 were considered to be significant. All calculations were done
with the computer software ‘Statistical Package for the Social
Sciences 12.0’ (SPSS, Chicago, IL, USA).
Results
QUS measured in the first week of life
In 114 neonates, QUS measurements were performed within the first
week of life. These values are as close as possible to intrauterine
bone mineralization and those least altered by postnatal influence
of nutrition or illness. Gestational age of these neonates ranged
from 24 to 42 weeks (mean±s.d: 35.9±4.5, interquartile range
32 to 40 wks) and weight from 600 to 5010 g (2353±1048 g,
interquartile range 1521 to 3163 g) (Table 2). There was a highly
significant positive correlation between QUS and gestational age
(r ¼ 0.53; P<0.001 both 20 and 35 mm) as well as birth weight
(r ¼ 0.59 for 20 mm and r ¼ 0.51 for 35 mm; both P<0.001). QUS
values measured with a probe distance of 20 mm were higher
compared with those found with a probe distance of 35 mm.
For both SGA and appropriate-for-gestational-age (AGA) infants,
QUS increased with gestational age and birth weight (P ¼ 0.001).
Bone quality in neonates by ultrasound
B Rack et al
221
Table 2 QUS measurements of preterm and term infants measured during their first week of life
Gestational age
(weeks)
24–28
29–32
33–36
37–40
>40
All infants
SGA infants
20 mm distance
(m s1)
35 mm distance
(m s1)
20 mm distance
(m s1)
35 mm distance
(m s1)
20 mm distance
(m s1)
1725±17
1743±25
1766±28
1793±17
1772±54
1655
1661±7
1687±20
1703±16
1716±16
1726±18
1748±24
1771±30
1804±11
1810
1655
1661±7
1687±20
1705±15
1716±17
1717
1725±19
1759±24
1784±17
1734
Weight (g)
All infants
20 mm distance
(m s1)
p1000
1001–1500
1501–2000
2001–2500
2501–3000
3001–3500
3501–4000
4001–5000
AGA infants
1721±10
1747±23
1759±22
1780±34
1781±32
1810
AGA infants
35 mm distance
(m s1)
1676±17
1688±21
1701±21
1706±14
1703±16
1719±23
20 mm distance
(m s1)
1726±12
1745±17
1760±27
1779±40
1784±27
1810
Number of
examinations (n)
35 mm distance
(m s1)
11
25
44
41
17
1672
1695±21
1711±15
SGA infants
20 mm distance
(m s1)
1673±20
1689±24
1697±21
1706±14
1703±16
1719±23
35 mm distance
(m s1)
1716±5
1753±35
1758±14
1783±11
1772±54
Number of
examinations (n)
20 mm distance
(m s1)
1682±14
1686±13
1713±21
9
18
25
25
28
14
12
7
Abbreviations: AGA, appropriate-for-gestational-age; QUS; quantitative ultrasound; SGA, small-for-gestational-age.
QUS values are shown as mean and s.d. for all four measurement sites.
With 20 mm probe distance, SGA infants showed significantly lower
QUS values compared with AGA when correlated to gestational age
(P ¼ 0.014) (Figure 2), but not when correlated to birth weight
(P ¼ 0.54). Repeated day-to-day measurements showed a 0.62%
variation coefficient for QUS. Twelve infants were large for
gestational age (980 to 5130 g; 26 to 42 weeks). QUS values were
comparable to AGA infants of the same age and birth weight.
Follow-up measurements of QUS
The effect of postnatal changes on bone mineral density was
examined by 602 follow-up measurements in 172 neonates between
25 and 68 postconceptional weeks (mean±s.d: 34.5±5.2)
(Table 3). QUS values ranged from 1621 to 1832
(1710±15 m s1). Both for 20 mm probe distance (r ¼ 0.197;
P ¼ 0.001) and for 35 mm probe distance (r ¼ 0.129; P ¼ 0.037),
there was a significant correlation of QUS with increasing
postconceptional age. Additionally, QUS during follow-up correlated
highly significant with the weight on the day of measurement
(r ¼ 0.377 for 20 mm and r ¼ 0.363 for 35 mm; both P ¼ 0.001).
When we compared bone quality at the age of 40 weeks in infants
born at term and infants born at 24 to 28 weeks, preterm infants
showed significantly lower QUS (1720 m s1; s.d 24 m s1) than
term infants (1785 m s1; s.d 27 m s1); P<0.001). None of the
infants experienced a fracture during the observation period.
In 60 newborns, repeated follow-up examinations (X3) were
available. In all 32% of these neonates were small for gestational
age. We observed an initial decrease of QUS during the first weeks
of life followed by an increase. QUS changes correlated significantly
with the infants’ week of life at measurement (P ¼ 0.026 for
20 mm and 0.024 for 35 mm). QUS changes depended also on the
maturity of the infants. Less mature infants showed a longer and
more distinct decrease of QUS (Figure 3). Furthermore, in preterm
infants QUS development from week 1 to 3 of life correlated
positively with QUS velocity at term (P ¼ 0.035).
Correlation of QUS with laboratory parameters and risk factors
In all 172 patients, calcium and phosphate excretion in the urine
was available. Calcium concentration in urine was found to be
0.1 to 13.5 mmol l1 (median 1.4 mmol l1) and phosphate
concentration in urine between 0.0 and 37.9 mmol l1 (median
3.3 mmol l1). There was no correlation of alteration of QUS over
time (DQUS) with calcium concentration in urine (P ¼ 0.47)
or phosphate concentration in urine (P ¼ 0.995). Calcium
concentration in serum was found to be 1.26 to 2.9 mmol l1
(median 2.4 mmol l1), and phosphate concentration in serum
between 0.91 and 8.23 mmol l1 (median 2.05 mmol l1). As
expected, calcium concentration in serum and phosphate
concentration in serum were not related to DQUS.
Journal of Perinatology
Bone quality in neonates by ultrasound
B Rack et al
222
Table 3 Follow-up QUS measurements of SGA and AGA infants
Postgestational
age (weeks)
All infants
20 mm
distance
(m s1)
24–28
29–32
33–36
37–40
41–44
45–48
49–52
53–68
35 mm
distance
(m s1)
1717±20
1729±26
1740±34
1745±32
1731±35
Weight (g)
1668±17
1673±24
1678±26
1685±35
1670±12
1671±8
1686±18
20 mm
distance
(m s1)
1715±21
1733±25
1742±34
1750±32
1742±61
All infants
20 mm
distance
(m s1)
p1000
1001–1500
1501–2000
2001–2500
2501–3000
3001–3500
3501–4000
4001–5000
>5000
AGA infants
1709±22
1728±27
1738±29
1749±34
1748±33
1797±19
SGA infants
20 mm
distance
(m s1)
20 mm
distance
(m s1)
35 mm
distance
(m s1)
1724±16
1719±27
1734±35
1739±31
1728±24
1668±17
1674±24
1682±26
1692±34
1671±12
1672±10
1683±19
20 mm
distance
(m s1)
1710±19
1730±24
1739±31
1750±35
1746±33
1797±19
1640
1665±20
1670±24
1680±25
1690±24
1694±28
1697±30
1679±17
2±2
3±2
4±3
6±5
8±7
16±6
25±5
32±7
19
80
274
165
48
7
3
6
1664±22
1668±25
1664±31
1663
1668
1700
AGA infants
35 mm
distance
(m s1)
Number of
Postconceptional
examinations (n)
age (weeks)
SGA infants
20 mm
distance
(m s1)
1640
1668±19
1673±25
1678±23
1690±25
1694±28
1697±30
1679±17
20 mm
distance
(m s1)
1707±25
1723±32
1737±27
1746±33
1759±35
(1–6)
(1–10)
(1–14)
(1–17)
(1–18)
(5–22)
(21–30)
(24–40)
Number of
examinations (n)
35 mm
distance
(m s1)
32
80
162
182
74
30
21
14
7
1657±18
1664±21
1692±34
1692±21
Abbreviations: AGA, appropriate-for-gestational-age; QUS; quantitative ultrasound; SGA, small-for-gestational-age.
QUS values are shown as mean and s.d. including measurements during first week of life.
Table 4 Details of follow-up measurements in preterm and term infants
Gestational week at birth (weeks)
24–28
29–32
33–36
37–40
41–44
Number of infants
31
41
49
37
14
Correspondingly, AP with values from 165 to 2500 U l1
(median 462 U l1) was compared with 167 QUS measurements.
In our patient group, AP was neither related to calcium
concentration in urine (P ¼ 0.84) nor to calcium concentration in
serum (P ¼ 0.28) or phosphate concentration in serum
(P ¼ 0.68). However, there was a highly significant inverse
correlation of AP with phosphate concentration in urine
Journal of Perinatology
Duration of follow-up (weeks)
15
10
3
1
2
(1–46)
(1–54)
(1–21)
(1–3)
(1–5)
Postconceptional age at last follow-up
measurement (weeks)
38
38
37
39
42
(26–64)
(31–67)
(33–52)
(37–41)
(41–45)
(P ¼ 0.001). Also, QUS values for both probe distances correlated
inversely with AP (P ¼ 0.003). In contrast, DQUS was not related
to AP (P ¼ 0.765).
Concerning nutritional factors influencing bone quality, we
found a significant inverse correlation of QUS with calcium,
phosphate and vitamin D supply during the week preceding the
measurement (P each <0.001). Among further risk factors for
Bone quality in neonates by ultrasound
B Rack et al
223
1850
per h) or intratracheal ventilation and treatment with
corticosteroids or diuretics were inversely related to QUS
(P each <0.001). The DQUS values, however, did not show any
association with nutrition or risk factors of reduced bone
mineralization.
UTG [m/s]
1800
1750
1700
1650
20
25
30
35
40
45
Gestational Age [weeks]
Figure 2 Correlation of quantitative ultrasound and gestational week during the
first week of life. Shown are mean QUS values and s.d. based on 114 single
measurements for all small-for-gestational-age (J) and appropriate-forgestational-age (’) infants born within a period of two gestational weeks and
measured with 20 mm probe distance.
1850
UTG [m/s]
1800
1750
1700
1650
1600
0
2
4
6
8
10
12
14
16
18
20
Age [weeks]
1850
UTG [m/s]
1800
1750
1700
1650
1600
0
2
4
6
8
10
12
14
16
18
20
Age [weeks]
Figure 3 (a)Correlation of quantitative ultrasound (QUS) with age according to
different gestational-age groups. Shown are QUS values according to different
gestational-age groups at birth (up to 27 weeks, black; 28 to 32 weeks, green; 33
to 37 weeks, blue; 37 weeks and higher, red). (b) Correlation of QUS with age in
infants born at term. QUS values in infants born at term are shown (small-forgestational-age in dark blue, appropriate-for-gestational-age in light blue).
reduced bone mineralization, prolonged duration of parenteral
nutrition (defined as an intravenous supply of lipids and/or
amniotic acids and/or 10% glucose in a dosage of more than 2 ml
Multivariate analysis
Linear regression analysis of QUS allowing to correct for
postconceptional age, weight at the time of examination, AP and
supply with calcium, phosphate, and vitamin D showed all
parameters except AP (P ¼ 0.90) and the supply of vitamin D
(P ¼ 0.96) as significant independent prognostic factors for
bone quality measured by QUS with 20 mm probe distance
(Table 5). For 35 mm probe distance, in contrast, only
postconceptional age could be confirmed as prognostic factor
for QUS (P ¼ 0.007).
Discussion
Osteopenia of prematurity develops particularly in very-low-birthweight infants, manifesting as reduced bone mineralization or
skull deformity and even fracture. A sufficient postnatal supply of
calcium, phosphate and vitamin D is an important prerequisite in
the prevention of this illness.18 However, an optimal
supplementation is difficult due to problems with oral feeding and
a reduced solubility of calcium and phosphate in nutrition.
Unfortunately, a diagnostic method to exactly adjust
supplementation to the individual requirement of each neonate is
still missing, as is any direct marker to monitor the consequences
of low supply: reduced bone quality. QUS is an established
osteodensitometric technique in the diagnosis of postmenopausal
osteoporosis and a good predictor of fracture risk.8,19 This
radiation-free technique not only provides information on bone
mineral content but also on the micro-architecture of bone, like
porosity, trabecular orientation, linkage and elasticity, which are
all crucial for the mechanical strength of bone.5,7
In our study, measurements on 114 term and preterm infants
within their first week of life, representing bone mineralization as
close as possible to the intrauterine situation and least altered by
postnatal influence of nutrition or illness, showed a highly
significant correlation of QUS with gestational age and birth
weight. During examination, the children did not show signs of
distress, for example, crying or other signs of discomfort. For
postnatal bone mineralization, our measurements on 172 neonates
showed a correlation of QUS with postconceptional age and the
weight on the day of measurement. These values are comparable to
those reported from children aged 1 to 6 years (1567 to
1832 m s1) whereas higher values are found in adults (1952 to
2297 m s1).20 Additionally, QUS increased in particular between
24 and 40 weeks of gestational age and levelled off thereafter. This
corresponds both to biochemical analyses of fetal bone and to
Journal of Perinatology
Bone quality in neonates by ultrasound
B Rack et al
224
Table 5 Multivariate regression analysis for QUS
20 mm probe distance
Postconceptional age on the day of measurement (weeks)
Weight on the day of measurement (g)
Alkaline Phosphatase (U l1)a
Supply with calcium (mmol kg1 per week)
Supply with phosphate (mmol kg1 per week)
Supply with vitamin D (I.E. per kg per week)
35 mm probe distance
Significance
95% CI
Significance
95% CI
0.006
<0.001
0.902
0.012
0.003
0.957
9.807 to 1.774
+0.028 to +0.086
0.032 to +0.029
+0.508 to +3.874
5.533 to 1.219
0.007 to +0.007
0.007
0.256
0.138
0.274
0.236
0.251
7.982 to 1.357
0.009 to +0.034
0.056 to +0.008
0.781 to +2.682
3.487 to +0.886
0.016 to +0.004
Abbreviations: AP, alkaline phosphatase; QUS; quantitative ultrasound.
Values for AP and supply with calcium, phosphate and vitamin D within 2 weeks before the individual measurement were used for calculation.
a
AP was available in 27% of QUS measurements (n ¼ 167).
studies on absorption and retention of nutritional minerals in the
fetal skeleton during the last trimenon of pregnancy.21,22 These
studies have shown an exponential increase of bone mineral
density and mechanical breaking strength with gestational age
with a maximal bone mineralization rate at 36 weeks of pregnancy
followed by a decrease of transplacental supply of the fetus with
calcium and phosphate. Our data also confirm previous studies
that reported a correlation of bone quality with gestational age and
birth weight both by DXA 23–26 and QUS.9–15,27,28
Multivariate analysis confirmed postconceptional age as the
most relevant predictor of bone quality. Body weight was the
strongest predictor of QUS in the most immature infants but lost its
significance in term neonates. These results are consistent with
data published for X-ray absorptiometry and QUS in older
children.23,25,29 –33 However, postnatal bone mineralization in
neonates is discussed controversially. Gonnelli et al.13 monitored
longitudinal changes of bone quality in 73 term infants at birth
and 12 months of age. Corresponding to our data, the authors
reported a significant increase of QUS with age, when QUS was
adjusted for soft tissue thickness by a calculation model. In
contrast, in recent studies QUS correlated directly with gestational
age and birth weight, whereas the authors found a significant, but
inverse correlation of QUS with postnatal age.9,15,34,35 This was
attributed to mineral deficiency and lack of mechanical
stimulation, compared with the intrauterine situation. Yet, the
follow-up in our study exceeds previously published data, as our
patients were followed closely at predefined intervals for up to
4 months after birth and in greater intervals until the age of
17 months. These data revealed an initial decline of QUS during
the first weeks of life, which might reflect the negative correlation
in other studies. The duration of this decline correlated with the
maturity of the neonates, with a more prolonged decrease in the
most premature infants. Subsequently, however, QUS increased
continuously until reaching a plateau at a postconceptional age
of 40 weeks or a body weight of about 3500 g. Although a recent
report on 150 preterm infants, where longitudinal data were
collected until 14 months of age, also showed an initial QUS
Journal of Perinatology
decrease, with the lowest nadir and slowest recovery in the most
immature infants, the authors attributed this phenomenon to
changes in body mass index.10 However, we observed an identical
phenomenon using a measurement technique that accounts for
changes of body mass index and reflects bone quality at four
different sites. Therefore, this effect might be rather related to
maturity and reduced bone mineralization rate in very preterm
infants.
Tomlinson et al.15 published data on serial QUS measurements
in 18 preterm infants from birth until 37 weeks of age and found a
significant but inverse correlation of QUS and postnatal age.
Measurements made early in the postnatal period could not predict
bone quality at term in this study. Although the authors attributed
this observation to impaired bone strength, their technique did not
correct for soft tissue changes, which also would result in a
decrease of QUS values. Although we also found a decline in QUS
during the first weeks of life, initial bone quality development in
the first 3 weeks could predict QUS at term. Additionally, preterm
infants at term showed significantly reduced QUS compared with
neonates born at term. As in preterm infants the early postnatal
period seems to be essential for bone development; prolonged
parenteral supplementation might be beneficial.
In SGA infants, growth retardation is in many cases caused by a
generalized malnutrition due to placenta insufficiency.36 A possible
mechanism for lower age-dependent QUS values of SGA compared
with AGA infants in our study could be a calcium and phosphate
deficiency due to reduced placental transfer leading to diminished
bone mineralization. As this nutritional restriction does affect
growth, length and weight, as well as bone mineralization, this
difference could not be detected, when QUS was analyzed in
relation to body weight. These findings confirm published data
reporting a reduced bone mineral density in SGA compared with
AGA infants measured by X-ray absorptiometry and DXA.37,38 Two
studies using tibial speed of sound reported higher QUS values for
SGA compared with AGA infants, which might, however, be
explained by a soft-tissue effect that none of the working groups
corrected for.16,17
Bone quality in neonates by ultrasound
B Rack et al
225
The main cause of osteopenia of prematurity is the postnatally
reduced supply of calcium, phosphate and vitamin D. Several
studies provide evidence that an increased supply of these
substances can improve bone mineral density during the neonatal
period as measured by photonabsorption densitometry.18 In
contrast, our data show a significant, but inverse correlation of
QUS with the supply of calcium, phosphate and vitamin D. This
correlation in our study population can most likely be attributed to
the coincidence of an intentionally increased supplementation in
extremely preterm infants with low bone density, whereas term
infants with comparatively high bone density received less calcium
and phosphate per body weight. Although there was no standard
supplementation, nutrition was adjusted according to laboratory
values. Furthermore, the discrepancy to so far published data
might be due to the method of bone quality measurement, as most
available studies are based on radiological techniques measuring
bone mineral density. In contrast, QUS is also influenced by
mechanical and structural properties of bone, which goes beyond
sole bone mineral density.5,7 Increase in QUS might therefore not
be exclusively explained by mineral supply and bone mineral
content, but also reflect growing bones with relatively diminished
mineralization, which should display as reduced bone-mineral
density when using radiological techniques.
One of the parameters used to diagnose osteopenia of
prematurity are urinary excretion of calcium and phosphate as well
as AP. Although AP is not an established marker for bone turnover,
the usefulness of urinary calcium and phosphate excretion is
limited due to many confounders influencing the regulation of
renal mineral excretion especially in preterm neonates with
possibly impaired organ and enzyme function and under treatment
of a variety of different medications.39 Although urinary excretion
of calcium and phosphate were not related to QUS in our study, AP
correlated inversely with QUS, which confirms reports applying
both QUS and DXA that found AP to predict reduced bone density
with a sensitivity of 88% and a specificity of 71%.14,15,40,41 The lack
of correlation between AP or QUS and serum levels of calcium or
phosphate in our patients is most likely due to the stringent
regulation of these two minerals by a complex endocrine system to
keep their serum levelFessential for many organ
functionsFwithin normal limits. As expected, several risk factors
for reduced bone formation showed a relevant influence on QUS,
that is, the duration of parenteral nutrition or mechanical
ventilation and the treatment with diuretics or corticosteroids.42
Though we found a strong correlation of QUS with age and
weight of the neonates, inter-patient variability was high.
Additionally, a s.d. of 15 m s1 for all measurements contrasted a
mean alteration of QUS values by 34 m s1 per month. Therefore,
we did not find it appropriate to define age- and weight-related
normal values for our study population. Weight and lengthdependent reference values for intrauterine growth have been
published by Ritschl et al10, based on QUS measurements of 132
preterm and term infants. In this study, however, a QUS increase of
only 12 m s1 during a weight gain from 500 to 5000 g was
contrasted by a mean s.d. of 22 m s1 per 500 g and limits the use
these reference values for the assessment of bone development in
clinical practice. In contrast, day-to-day reproducibility in our
study was excellent, follow-up measurements of individual patients
correlated well with changes in mineral supply, and, above all,
bone quality at term could be predicted by QUS development within
the first 3 weeks of life. This underlines the clinical relevance of
QUS for the monitoring of individual patients and might allow
randomized controlled studies to evaluate interventions to treat
osteopenia of prematurity.
In conclusion, this study finds QUS to be a highly reproducible,
easily applicable and radiation-free technique for the assessment of
bone quality in term and preterm infants. Though high intraindividual variation does not permit defining normal values, QUS
seems to be a valuable tool to assess bone quality when repeated
measurements in individual patients are used. For monitoring of
mineral needs, it might be more useful than the determination of
calcium and phosphate excretion in urine, which is dependent on
medication like diuretics often used in critically ill newborns. In
addition, QUS delivers information not only on bone
mineralization but also on bone quality and structure that are not
covered by mineral excretion. Therefore, its use especially in
infants with very low gestational age should allow early detection of
osteopenia better than currently available methods. Further
prospective randomized studies are necessary to evaluate, if
therapeutic interventions based on decreased bone quality
measured by QUS are able to reduce the incidence of osteopenia
of prematurity.
Conflict of interest
The authors declare no conflict of interest.
Acknowledgments
We are grateful to all infants and their parents who took part in this study, to all
nurses and doctors for their assistance, and to Heidi Weitmann Coleman for her
editorial support. The prototype Osteoson K IV was kindly provided by Minhorst,
Meudt, Germany
References
1 Rauch F, Schoenau E. Skeletal development in premature infants: a review of bone
physiology beyond nutritional aspects. Arch Dis Child Fetal Neonatal Ed 2002; 86:
F82–F85.
2 Sargent JD, Stukel TA, Kresel J, Klein RZ. Normal values for random urinary calcium to
creatinine ratios in infancy. J Pediatr 1993; 123: 393–397.
3 Tsang RC. Nutritional needs of the preterm infant: scientific basis and practical
guidelines. 2nd edition. Digital Education Publishing, Inc.: Cincinnati, OH, 2005.
Journal of Perinatology
Bone quality in neonates by ultrasound
B Rack et al
226
4 Gilsanz V. Bone density in children: a review of the available techniques and
indications. Eur J Radiol 1998; 26: 177–182.
5 Bouxsein ML, Radloff SE. Quantitative ultrasound of the calcaneus reflects the mechanical
properties of calcaneal trabecular bone. J Bone Miner Res 1997; 12: 839–846.
6 Fricke O, Tutlewski B, Schwahn B, Schoenau E. Speed of sound: relation to geometric
characteristics of bone in children, adolescents, and adults. J Pediatr 2005; 146:
764–768.
7 Sakata S, Barkmann R, Lochmuller EM, Heller M, Gluer CC. Assessing bone status
beyond BMD: evaluation of bone geometry and porosity by quantitative ultrasound of
human finger phalanges. J Bone Miner Res 2004; 19: 924–930.
8 Gluer CC, Eastell R, Reid DM, Felsenberg D, Roux C, Barkmann R et al. Association of
five quantitative ultrasound devices and bone densitometry with osteoporotic vertebral
fractures in a population-based sample: the OPUS Study. J Bone Miner Res 2004; 19:
782–793.
9 Rubinacci A, Moro GE, Boehm G, de Terlizzi F, Moro GL, Cadossi R. Quantitative
ultrasound for the assessment of osteopenia in preterm infants. Eur J Endocrinol
2003; 149: 307–315.
10 Ritschl E, Wehmeijer K, de Terlizzi F, Wipfler E, Cadossi R, Douma D et al. Assessment
of skeletal development in preterm and term infants by quantitative ultrasound.
Pediatr Res 2005; 58: 341–346.
11 Pereda L, Ashmeade T, Zaritt J, Carver JD. The use of quantitative ultrasound in
assessing bone status in newborn preterm infants. J Perinatol 2003; 23: 655–659.
12 Litmanovitz I, Dolfin T, Friedland O, Arnon S, Regev R, Shainkin-Kestenbaum R et al.
Early physical activity intervention prevents decrease of bone strength in very low birth
weight infants. Pediatrics 2003; 112: 15–19.
13 Gonnelli S, Montagnani A, Gennari L, Martini S, Merlotti D, Cepollaro C et al.
Feasibility of quantitative ultrasound measurements on the humerus of newborn
infants for the assessment of the skeletal status. Osteoporos Int 2004; 15: 541–546.
14 Nemet D, Dolfin T, Wolach B, Eliakim A. Quantitative ultrasound measurements of
bone speed of sound in premature infants. Eur J Pediatr 2001; 160: 736–740.
15 Tomlinson C, McDevitt H, Ahmed SF, White MP. Longitudinal changes in bone health
as assessed by the speed of sound in very low birth weight preterm infants. J Pediatr
2006; 148: 450–455.
16 McDevitt H, Tomlinson C, White MP, Ahmed SF. Quantitative ultrasound assessment of
bone in preterm and term neonates. Arch Dis Child Fetal Neonatal Ed 2005; 90:
F341–F342.
17 Littner Y, Mandel D, Mimouni FB, Dollberg S. Bone ultrasound velocity of infants born
small for gestational age. J Pediatr Endocrinol Metab 2005; 18: 793–797.
18 Specker BL, Beck A, Kalkwarf H, Ho M. Randomized trial of varying mineral intake on
total body bone mineral accretion during the first year of life. Pediatrics 1997; 99:
E12.
19 Bauer DC, Gluer CC, Cauley JA, Vogt TM, Ensrud KE, Genant HK et al. Broadband
ultrasound attenuation predicts fractures strongly and independently of densitometry in
older women. A prospective study. Study of Osteoporotic Fractures Research Group. Arch
Intern Med 1997; 157: 629–634.
20 Schonau E, Radermacher A, Wentzlik U, Klein K, Michalk D. The determination of
ultrasound velocity in the os calcis, thumb and patella during childhood. Eur J Pediatr
1994; 153: 252–256.
21 Shaw JC. Evidence for defective skeletal mineralization in low-birthweight infants: the
absorption of calcium and fat. Pediatrics 1976; 57: 16–25.
22 Wright LL, Glade MJ, Gopal J. The use of transmission ultrasonics to assess bone status
in the human newborn. Pediatr Res 1987; 22: 541–544.
23 Koo WW, Walters J, Bush AJ, Chesney RW, Carlson SE. Dual-energy X-ray
absorptiometry studies of bone mineral status in newborn infants. J Bone Miner Res
1996; 11: 997–102.
Journal of Perinatology
24 Salle BL, Braillon P, Glorieux FH, Brunet J, Cavero E, Meunier PJ. Lumbar bone
mineral content measured by dual energy X-ray absorptiometry in newborns and
infants. Acta Paediatr 1992; 81: 953–958.
25 Avila-Diaz M, Flores-Huerta S, Martinez-Muniz I, Amato D. Increments in whole
body bone mineral content associated with weight and length in pre-term and
full-term infants during the first 6 months of life. Arch Med Res 2001; 32:
288–292.
26 Chan GM, Armstrong C, Moyer-Mileur L, Hoff C. Growth and bone mineralization in
children born prematurely. J Perinatol 2008; 28: 619–623.
27 Littner Y, Mandel D, Mimouni FB, Dollberg S. Decreased bone ultrasound velocity in
large-for-gestational-age infants. J Perinatol 2004; 24: 21–23.
28 Littner Y, Mandel D, Mimouni FB, Dollberg S. Bone ultrasound velocity curves
of newly born term and preterm infants. J Pediatr Endocrinol Metab 2003; 16:
43–47.
29 Zadik Z, Price D, Diamond G. Pediatric reference curves for multi-site quantitative
ultrasound and its modulators. Osteoporos Int 2003; 14: 857–862.
30 Sundberg M, Gardsell P, Johnell O, Ornstein E, Sernbo I. Comparison of quantitative
ultrasound measurements in calcaneus with DXA and SXA at other skeletal sites: a
population-based study on 280 children aged 11 to 16 years. Osteoporos Int 1998; 8:
410–417.
31 Baroncelli GI, Federico G, Bertelloni S, de Terlizzi F, Cadossi R, Saggese G. Bone
quality assessment by quantitative ultrasound of proximal phalanxes of the hand in
healthy subjects aged 3–21 years. Pediatr Res 2001; 49: 713–718.
32 Lequin MH, van Rijn RR, Robben SG, Hop WC, van Kuijk C. Normal values for tibial
quantitative ultrasonometry in caucasian children and adolescents (aged 6 to 19
years). Calcif Tissue Int 2000; 67: 101–105.
33 Minton SD, Steichen JJ, Tsang RC. Bone mineral content in term and preterm
appropriate-for-gestational-age infants. J Pediatr 1979; 95: 1037–1042.
34 Mercy J, Dillon B, Morris J, Emmerson AJ, Mughal MZ. Relationship of tibial speed of
sound and lower limb length to nutrient intake in preterm infants. Arch Dis Child
Fetal Neonatal Ed 2007; 92: F381–F385.
35 Ahmad I, Nemet D, Eliakim A, Koeppel R, Grochow D, Coussens M et al. Body
composition and its components in preterm and term newborns: a cross-sectional,
multimodal investigation. Am J Hum Biol 2010; 22: 69–75.
36 Namgung R, Tsang RC, Sierra RI, Ho ML. Normal serum indices of bone collagen
biosynthesis and degradation in small for gestational age infants. J Pediatr
Gastroenterol Nutr 1996; 23: 224–228.
37 Namgung R, Tsang RC. Factors affecting newborn bone mineral content: in utero
effects on newborn bone mineralization. Proc Nutr Soc 2000; 59: 55–63.
38 Beltrand J, Alison M, Nicolescu R, Verkauskiene R, Deghmoun S, Sibony O et al. Bone
mineral content at birth is determined both by birth weight and fetal growth pattern.
Pediatr Res 2008; 64: 86–90.
39 Zanardo V, Dani C, Trevisanuto D, Meneghetti S, Guglielmi A, Zacchello G et al.
Methylxanthines increase renal calcium excretion in preterm infants. Biol Neonate
1995; 68: 169–174.
40 Backstrom MC, Kouri T, Kuusela AL, Sievänen H, Koivisto AM, Ikonen RS et al.
Bone isoenzyme of serum alkaline phosphatase and serum inorganic phosphate
in metabolic bone disease of prematurity. Acta Paediatr 2000; 89:
867–873.
41 Altuncu E, Akman I, Yurdakul Z, Ozdoğan T, Solakoğlu M, Selim N et al. Quantitative
ultrasound and biochemical parameters for the assessment of osteopenia in preterm
infants. J Matern Fetal Neonatal Med 2007; 20: 401–405.
42 Shrivastava A, Lyon A, Mcintosh N. The effect of dexamethasone on growth, mineral
balance and bone mineralisation in preterm infants with chronic lung disease. Eur J
Pediatr 2000; 159: 380–384.