DXA Evaluation of Infants and Toddlers

Fig. 8.1

Spine BMD-for-age reference curve for infants , birth to 12 months. Taken from Gallo et al. [58]. Data were obtained on a Hologic QDR 4500A (n = 59)

Fig. 8.2

Median and ±2 standard deviation curves for lumbar spine BMC (a) and aBMD (b) for age for boys (solid lines) and girls (dashed lines). Taken from Kalkwarf et al. [79]. Data were obtained on a Hologic Discovery A fan-beam (n = 307)

Hazell and coworkers compared forearm DXA data from 57 children aged 2–5 years obtained on a Hologic 4500 Discovery and found these data to agree with a portable DXA (pDXA) [80]. Precision estimates of forearm bone measurements using the current generation of DXA devices in infants and toddlers are not currently available in the literature.

Proximal femur scans have been performed on children as young as 2–3 years of age [81–84] and the whole femur in infants up to 12 months of age [58, 85]. The analysis of these scans is difficult due to the short femoral neck, low bone density, and the shape of the trochanter, leading the ISCD to recommend that the proximal femur scan not be obtained for clinical purposes in children <5 years of age.

Bone Measurement Issues

Equipment and Software Differences

Despite the extensive use of DXA in infants , discrepancies exist among studies in the reported normative values for BMC, aBMD, and body composition. In a large part these discrepancies are due to differences in the acquisition and analysis software among DXA manufacturers and the various software versions. Whole body BMC values vary significantly between pencil- and fan-beam devices as well as between adult and pediatric software analyses [86]. Hammami and coworkers compared the differences among adult, pediatric, and infant software analyses on a Hologic QDR4500A (Fig. 8.3) [87]. They found that the adult software significantly underestimated BMC, and overestimated aBMD, compared to the infant software while there was some improvement using the pediatric software [87].

Fig. 8.3

Relationships between adult, pediatric , and infant whole body software for fan-beam DXA. Measurements using adult software (dashed line) are more than twice the measurements obtained using infant software. Results using pediatric software (dotted line) also are greater than infant software results, but not to the same extent as use of adult software. The solid line is the line of identity. Modified from Hammami et al. [87]

The Hologic QDR-1000W and QDR-1500 were the original pencil beam devices that had an infant whole body application and are now obsolete. These original pencil beam devices were replaced by the QDR-2000 that could operate in both the single beam and fan-beam modes. Eventually the QDR-4500 replaced the QDR-2000. The different DXA devices, models, and software versions are summarized in Table 8.1 and reference data for different devices are given in Table 8.2.

Table 8.1

DXA densitometers using pediatric or infant software

Manufacturer	Model^a (production years)		Software versions	Published data by software version
Hologic Quantitative Digital Radiography (QDR)	QDR-1000 & 1000 W (1987–1994)	Pencil-beam	Various	v5.35 (PWB)—Brunton et al. [123] v5.47 (PWB)—Picaud et al. [124] v5.56—Brunton et al. [123] v5.56—(PWB) Specker et al. [7] v5.64P (IWB)—Zia-Ullah et al. [77] v4.57Q (infant spine)—Zia-Ullah et al. [77] v5.71p (IWB)—Koo et al. [125] Infant spine—Koo et al. [78]
	QDR-150 0R (1987–1994)	Pencil-beam	Various	v5.67P (IWB)—Avila-Diaz et al. [126]
	QDR-2000 (1987–1994)	Pencil-/fan-beam	Various	IWB v5.64—Picaud et al. [124] IWB v5.73p—Koo et al. [127]
	QDR-2000+ (1992–1994)	Pencil-/fan-b eam	Various	IWB v5.73P—Zia-Ullah et al. [77] Infant spine v4.76Q—Zia-Ullah et al. [77] v5.71p—Koo [125, 127]
	QDR-4000 (1997–2003)		Various	Cannot do infant whole body
	QDR-4500 series (1995–2004)	Fan-beam	V9.x-Apex 4.x (if upgraded)	IWB—Koo et al. [127, 128] IWB vKH6—Hammami et al. [87, 104]
	Delphi series (2000–2004)	Fan-beam	V11.2-Apex 4.x (if upgraded)	IWB Elite v11.2—Gallo et al. [58]
	Discovery series (2003-pres)	Fan-beam	V12.0-v12.7 Apex 2.x Apex 3.x (aka V13.x) Apex 4.x (aka V14.x) Apex 5.x (aka V15.x)	Infant spine fast array v12.7—Kalkwarf et al. [79] IWB APEX 13.2.1—Gallo et al. [85]
GE/Lunar	DPX	Pe ncil-beam
	Prodigy	Fan-beam		IWB v12.10—Godang et al. [88] IWB v12.10– Ay et al. [89]
	iDXA	Fan-beam		IWB v11-30.062 enCore2007 [129]

C compact system, SL supine lateral feature added (C-arm rotates around patient), W whole body feature added, A research system with all features, IWB infant whole body

Software versions for Hologic QDR that affected pediatric measurements

V12.1 & V12.3—Auto Whole Body analysis added for subjects less than 40 kg. This replaced the former pediatric whole body analysis. Auto Whole Body analysis can be turned off to be identical to V12.01

V12.5—ISCD compatible reports with ISCD Pediatric reference data for <20 years added

V12.4, V2.5, V12.6—No changes in data collection during scan or analysis: BMD, BMC, and Area results are identical

APEX 2.2—Reference curves for children updated with Bone Mineral Density Childhood Study (BMDCS) data. Analysis method for <20 years changed for AP spine and hip to match BMDCS methods

APEX 3.0.1—Does not support “pediatric whole body” and does not provide legacy “Low Density Spine.” Auto Whole Body and Auto Low Density are used instead

APEX 3.1—Supports AP spine analysis for infants and toddlers birth to 3 years of age. Depends on the scan date set in the computer and date of birth entered

APEX 3.3—Added NHANES whole body bone and composition reference data for children

APEX 3.4—Reference data for whole body BMD versus age extended down to age 3 using Hologic Pediatric reference data for ages 3–6 and BMDCS references data at age 7. Report provides whole body subtotal BMD (excluding head) for ages <20 years

APEX 4.0—Infant Whole Body acquisition mode for infants aged birth to 1 year

Calibration of Infant Whole Body fat and lean changed on Discovery W, Wi, and Explorer models resulting in greater fat mass. Calibration for A models did not change

NHANES White Whole Body BMD versus age reference data is extended down to age 3 using the Hologic Pediatric reference data for ages 3–6 years and the BMD Children Study reference data at age 7 years

Whole body pediatric curves for lean vs. height and lean vs. subtotal BMC included (lean does not include BMC)

APEX 4.5.2 & 5.5.2—new “Infant Whole Body” analysis added. Excluded more bone pixels from body composition resulting in higher percent fat in most infants. The existing “Infant Whole Body” is still available but manufacturer recommends a transition

^aHologic models

Table 8.2

Publications providing normative data on healthy infants and toddlers

Reference	Device	Population N (male) age	Bone site	Mean ± SD
Location	Model software		Bone site	Mean ± SD
Whole body
Venkataraman et al. [130] Oklahoma City, OK Oklahoma Memorial Hospital	DXA—not specified	28 Newborns 1.46 ± 0.12 d	WB BMC (g) WB bone area (cm²)	80.5 ± 6.63 241 ± 13
Atkinson et al. [131] Hamilton, Ontario, Canada McMaster University	Hologic QDR 1000 W Pediatric WB v5.56	46 Neonates Age <1 m	WB BMC (g)	74 ± 13
Specker et al. [7] Cincinnati, Ohio University of Cincinnati	Hologic QDR 1000 W Pediatric WB v5.56	Infants Phase 1: 92 infants (39 M) Age 1, 3, 6 m Phase 2: 87 infants (37 M) Age 6, 9, 12 m	WB BMC (g) 1 m 3 m 6 m 9 m 12 m	91.9 ± 14.1 123.1 ± 16.9 161.8 ± 24.1 198.6 ± 28.7 236.2 ± 35.8
Koo et al. [132] Memphis, Tennessee University of Memphis	Hologic QDR 1000 W Pediatric WB v5.64P	130 Newborn and infant Age N 1–8 d 65 (37 M) 9–390 d 65 (34 M) 9–90 d 16 91–150 d 17 151–270 d 12 271–390 d 20	WB BMC (g) 1–8 d 9–90 d 91–150 d 151–270 d 271–390 d WB bone area (cm²) 1–8 d 9–90 d 91–150 d 151–270 d 271–390 d	68.2 ± 10.2 103.4 ± 21.4 137.1 ± 20.0 196.4 ± 26.6 253.2 ± 41.3 308 ± 26 431 ± 58 527 ± 45 650 ± 64 754 ± 88
Butte et al. [133] Houston, Texas Baylor College of Medicine	Hologic QDR 2000 Infant WB v5.56–5.71P	Newborns and infants Age N 0.5 m 76 (33 M) 12 m 74 (32 M) 24 m 72 (29 M)	WB BMC (g) 0.5 m Girls 0.5 m Boys 12 m Girls 12 m Boys 24 m Girls 24 m Boys	68 ± 12 68 ± 13 208 ± 31 221 ± 33 298 ± 48 321 ± 38
Avila-Diaz et al. [126] Mexico City, Mexico	Hologic QDR 1500 Infant WB v5.67P	48 Infants (26 M) 1–5 m Age N 33 ± 4 d 37 63 ± 4 d 35 94 ± 11 d 29 126 ± 10 d 18 147 ± 6 d 16	WB BMC (g) 1 m 2 m 3 m 4 m 5 m WB bone area (cm²) 1 m 2 m 3 m 4 m 5 m	82.6 ± 11.9 103.8 ± 16.0 111.0 ± 15.1 129.3 ± 17.7 134.0 ± 20.0 377.7 ± 36.6 437.2 ± 61.8 467.7 ± 39.3 499.3 ± 43.5 521.0 ± 46.0
Hammami et al. [104] Detroit, Michigan Wayne State University	Hologic QDR 4500A vKH6	73 Newborns (32 M) 3.0 ± 2.1 d	WB BMC (g) Bone area (cm²)	89.3 ± 14.1 371 ± 33
Ay 2011 et al. [89] Rotterdam, The Netherlands Erasmus Medical Center	GE lunar prodigy IWB v12.10	252 Infants (145 M) Boys: 6.4 ± 0.8 m Girls: 6.3 ± 0.7 m	WB BMC (g) Boys Girls	120.9 ± 23.5 110.5 ± 20.4
Gallo 2012 [58] Winnipeg, MB, Canada University of Manitoba	Hologic QDR 4500A Elite IWB v11.2	52 Infants (36 M) Age N 2–4 w 62 6 m 35 12 m 11	WB BMC (g) 2–4 w 6 m 12 m WB BMC less head (g) 2–4 w 6 m 12 m	76.0 ± 14.2 169.5 ± 29.0 227.0 ± 29.7 45.6 ± 8.1 87.7 ± 16.3 114.8 ± 21.9
Lumbar spine
Brallion et al. [134] Lyon, France Eduard Herriot Hospital	Hologic QDR 1000 UHR v4.10	10 Newborns 1 d	L1-L5 LS BMC (g) LS BMD (g/cm²)	2.34 ± 0.42 0.268 ± 0.030
Salle et al. [135] Lyon, France Eduard Herriot Hospital	Hologic QDR 1000 UHR v4.20	57 Newborns (29 M) Age ≤48 h	L1-L5 BMC (g) L1-L5 BMD (g/cm²)	Data presented graphically with normal range by weight (kg) and length (cm)
Kurl et al. [136] Kuopio, Finland Kuopio University Hospital	Lunar DPX Pediatric AP spine V3.8E	41 Infants (19 M) Age 2–6 m	L2-L4 BMC (g) Boys Girls L2-L4 Bone Area (cm²) Boys Girls L2-L4 BMD (g/cm²) Boys Girls	2.26 ± 0.58 1.93 ± 0.57 8.74 ± 1.20 8.27 ± 1.10 0.25 ± 0.04 0.23 ± 0.05
Zia-Ullah 2002 [77] Memphis, Tennessee University of Tennessee	Hologic QDR 1000 W IS v4.57Q	99 Infants (58 M) Age 1–391 d	L1-L4 BMC (g) Bone area (cm²) BMD (g/cm²)	2.2 ± 1.5 8.8 ± 4.2 0.231 ± 0.049
Ay 2011 [89] Rotterdam, The Netherlands Erasmus Medical Center	GE lunar prodigy IS v12.10	252 Infants (145 M) Boys: 6.4 ± 0.8 m Girls: 6.3 ± 0.7 m	LS BMC (g) Boys Girls LS BMD (g/cm²) Boys Girls	2.7 ± 0.5 2.6 ± 0.4 0.31 ± 0.04 0.33 ± 0.04
Gallo 2012 [58] Winnipeg, MB, Canada University of Manitoba	Hologic QDR 4500A Elite Auto low density spine v11.2	62 Infants (36 M) Age N 2–4 w 62 6 m 62 12 m 57	L1-L4 BMC (g) 2–4 w 6 m 12 m L1-L4 Bone area (cm²) 2–4 w 6 m 12 m L1-L4 BMD (g/cm²) 2–4 w 6 m 12 m	2.35 ± 0.42 3.59 ± 0.63 5.37 ± 1.02 8.86 ± 1.10 14.28 ± 2.01 17.67 ± 2.52 0.266 ± 0.044 0.252 ± 0.031 0.304 ± 0.044
Kalkwarf [79] Cincinnati, Ohio Children’s Hospital Medical Center	Hologic Discovery A Infant spine v12.7	307 (158 M) Age N 1–6 m 37 6.1–12 m 50 12.1–18 m 51 18.1–24 m 59 24.1–30 m 56 30.1–36 m 54		L1-L4 BMC (g) L1-L4 BMD (g/cm²) Data presented graphically by age and in table as Z-score

d day, m month, h hour, w week, WB whole body, BMC bone mineral content, BMD bone mineral density; IWB infant whole body

IS infant spine, LS lumbar spine

Whole Body Scans

There have been several types of software that have been used to obtain whole body scans, including infant, pediatric, and adult whole body scan software; the infant whole body scan software that does not come standard with the purchase of the device. The infant whole body scan takes about 3 min and bone edge detection algorithms are optimized to detect very low density bone. The pediatric software algorithm, which is now obsolete, was a default analysis used at body weights of less than 40 pounds. The acquisition and analysis algorithms for the infant, pediatric, and adult whole body scans differ and are not comparable [87].

Whole body scans of newborn infants with GE/Lunar Prodigy devices also have been reported [88, 89], with a scan time of about 6 min. It is unclear if the infant and regular whole body scan results converge at some body weight or are comparable. There are currently no studies that have performed head-to-head comparisons of devices from different manufactures and software versions for their impact on infant whole body bone measurements . GE/Lunar Prodigy normative values are about 30 % lower than data obtained on Hologic devices in infants 6 months of age.

Spine Scans

Scan acquisition is performed using the standard posterior-anterior (PA) spine scan. Hologic Apex software provides the ability to analyze spine scans for infants and toddlers by automatically determining the specific algorithm and bone edge detection thresholds. Bone edge detection thresholds are set based on age for children 36 months and younger, and by bone map evaluation for children older than 36 months. The algorithms were developed so that the same software can be used and scans compared across all ages. There are multiple scan modes (array, fast array, turbo) for acquisition of spine scans on the Hologic devices, but data comparing the various modes for infants and toddlers are lacking. Since it is not known whether these modes are comparable, it would be prudent to use the same scan mode on repeated measurements of the same child and when identifying appropriate normative data sets. The GE/Lunar densitometers also have the capability to perform spine measurements in infants and toddlers. Means ±1 standard deviation of spine BMC obtained from different publications are shown in Fig. 8.4 and illustrate differences in normal ranges that are observed among the different devices for various ages.

Fig. 8.4

Published normative data (means ±1 SD) for spine BMC in infants. Filled square = Hologic 1000 W; Plus Sign = Hologic 1500R; filled circle = Hologic 2000 (females); filled triangle = Hologic 2000 (males); filled diamond = Hologic QDR4500; open circle = GE/Lunar Prodigy (females); open triangle = GE/Lunar Prodigy (males). Different colors represent different studies

Performing a Scan in Infants or Toddlers

Infants are among the most challenging subjects to measure using DXA. Before measuring an infant it is helpful to feed and calm them and to place the infant on the scanning table in a clean diaper. If necessary, the child could be swaddled in a thin cotton sheet or blanket to reduce small involuntary movements. Subdued room lighting also may also help. Very young infants (i.e., <3 months of age) will usually sleep through the measurement and will require limited operator intervention (Fig. 8.5). Swaddling becomes ineffective for keeping infants motionless for the required 3–6 min as the infants become older and stronger. Stranger anxiety in infants around 8–9 months of age also makes this age especially difficult to scan. In the event of movement, DXA scans can be repeated with the hope of obtaining a movement-free image. Gallo et al. reported that when making up to two attempts, they obtained movement free whole body scans on 99 % of infants at 1 month of age compared to 81 % at 12 months of age [85]. Few studies report success rates obtaining movement free scans or the number of attempts to obtain one.

Fig. 8.5

Swaddling and positioning of an infant prior to whole body DXA measurement

Some investigators have used restraints on 6- and 12-month old infants scanned with fan-beam densitometers [89], but it is important the make sure that the restraints are radiolucent or scan results may be affected. Some studies have found an effect of multiple layers of cloth on bone measurements, which seems to be more of a problem with the fan-beam devices [87, 90].

Normative Data and Factors Affecting

Normative Data

In situations where a child’s growth is stunted or maturation is delayed, which is particularly true in children with chronic diseases , it may be more appropriate to determine whether BMC or aBMD results are appropriate for his or her body size by comparing the measurements with those of children of similar height or weight (see discussion above). However, these reference databases are not available on the DXA software and must be obtained from the pediatric literature.

Although numerous studies have included normative infant and toddler DXA data, the populations are typically small or highly selected based on specific population characteristics (Table 8.2). The largest studies providing age-specific reference data of the spine by fan-beam for infants and toddlers are those of Gallo et al. [58] (n = 62, 0–360 days; Fig. 8.1) and Kalkwarf et al. [79] (n = 308, 1–36 months; Fig. 8.2).

Only gold members can continue reading. Log In or Register to continue