| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Orthopaedics, Campus Box 7055, University of North Carolina,
Chapel Hill, NC 27599.
2
Department of Pediatrics, Campus Box 7525, University of North Carolina,
Chapel Hill, NC 27599.
3
School of Public Health, Campus Box 7400, University of North Carolina, Chapel
Hill, NC 27599.
4
Department of Research, A.I. duPont Hospital for Children, 1600 Rockland Rd.,
Box 269, Wilmington, DE 19899.
5
School of Nursing, University of Pennsylvania, 420 Guardian Dr., Philadelphia,
PA 19104.
6
Present address: Department of Gastroenterology and Nutrition, Children's
Hospital, 747 Fifty-Second St., Oakland, CA 94609.
7
Department of Radiology, Campus Box 7510, University of North Carolina, Chapel
Hill, NC 27599.
8
Department of Allied Health Sciences, Campus Box 7120, University of North
Carolina, Chapel Hill, NC 27599.
9
Departments of Radiology and Pediatrics, Jefferson Medical College, 1020
Walnut St., Philadelphia, PA 19107.
10
Department of Medical Imaging, A. I. duPont Hospital for Children, Wilmington,
DE 19899.
Received July 3, 2001;
accepted after revision August 21, 2001.
Address correspondence to R. C. Henderson.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. The study was a cross-sectional, single-observational assessment of 256 healthy children and adolescents between the ages of 3 years and 18 years 6 months (mean, 10 years 5 months). Bone mineral density was measured in the nondominant proximal femur, lumbar spine, and both distal femurs using dual X-ray absorptiometry.
RESULTS. We found that bone mineral density increases with age in the cortical, cancellous, and mixed regions of the distal femur, similar to the findings with other regional analyses of bone density. Bone density in the distal femur correlates very highly with bone density in the proximal femur and slightly less well with bone density in the lumbar spine.
CONCLUSION. In pediatric patients who have deformities, have experienced trauma, or have undergone surgical procedures that prevent reliable measures of bone density in the proximal femur, bone mineral density may be measured in the distal femur and interpreted relative to the bone mineral density findings in healthy age- and sex-matched controls.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Some pediatric conditions associated with osteoporosis may also be associated with serious contractures that prevent proper positioning for DXA measures of bone mineral density in the proximal femora. Children with conditions such as cerebral palsy, juvenile rheumatoid arthritis, spina bifida, and muscular dystrophy commonly have contractures of the hips, knees, or both that prevent the children from lying flat in the supine position. Osteoporotic fractures in children with these conditions most commonly involve the lower limbs, particularly the femora and rarely the spine [2,3,4,5,6,7]. Therefore, in children with these conditions, a problem arises because bone mineral density cannot be reliably measured in the proximal femora, and measurement of bone mineral density in the lumbar spine is an unreliable indicator of bone status in the clinically important femora.
To address this problem, researchers developed a technique for DXA assessment of bone mineral density in the distal femur projected in the lateral plane that has been successfully used in children with cerebral palsy [8]. It is an excellent technique to use in children with contractures of the lower limbs because bone mineral density is measured in the bone at greatest risk of fracture, and even children with severe contractures can be assessed while lying comfortably on their side. The purpose of this study was to provide the normal reference data important in clinical practice to the interpretation of bone mineral density measures made in the distal femur of children and adolescents.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A questionnaire was used to screen for general good health in each participant. Excluded from the study were those children or adolescents who had past or present medical conditions, were receiving medications, or had sustained physical injuries that could potentially affect bone mineral density. Each child's (barefoot) height was measured with a stadiometer; each child's weight (in light clothing) was measured with a digital scale. The Tanner stage was reported by the caregiver of the younger or self-reported by the older children and adolescents using a series of annotated pictures that depict the stages of pubertal development [9].
Bone mineral density was measured in the lumbar spine and both distal femurs. The nondominant proximal femur was also measured in the participants at the Chapel Hill and Philadelphia sites. For the distal femur scans, a previously reported [8] positioning protocol was used (Fig. 1). For a scan of the left distal femur, the child was placed on his or her left side. The left hip and knee could flexed to achieve any comfortable position, but the child had to be positioned on the table so that the femur was centered and parallel to the table. The right hip and knee were flexed forward in front of the left distal femur. Foam pads were used to comfortably support the torso and right leg as necessary so that the left distal femur could be placed in a true lateral position.
|
At the Chapel Hill site, subjects were measured on a 1000-W scanner (Hologic, Waltham, MA) in the high-precision mode. Subjects at the other two sites were measured on a Hologic 2000 scanner in the pencil-beam, high-precision mode. To confirm the reliability of pooling data from the three scanners, we measured a single spine phantom at each site with no significant differences observed.
Analysis of the Distal Femur Scans
Separate regions in the distal femur containing predominately cortical or
predominately cancellous bone were defined
(Fig. 2). This step is an
important aspect of the distal femur scanning technique because the
metabolisms in cortical and cancellous bone may differ considerably. We began
the analysis by determining the width of the femur distally at the growth
plate and the diameter of the diaphysis proximally. Region 1 was defined as a
rectangle that began at the antero-superior edge of the growth plate and
extended posteriorly one half of the width of the distal femur across the top
of the growth plate. The height of the rectangle was twice that of the
diaphyseal diameter. Region 1 did not extend fully to the back of the femur so
that we could exclude the more dense bone found in the posterior aspect where
the femoral condyles make the transition into the metaphysis. Thus, region 1
was defined to include almost exclusively cancellous bone.
|
We defined region 2 as extending proximally from the top of region 1 and encompassing the transition from metaphyseal to diaphyseal bone. Region 2 included the full width of the femur, with the same height as region 1. Region 3 (predominately cortical bone) was defined as extending proximally from the top of region 2 and having the same height as regions 1 and 2. The subregion forearm software version 5.73Q (Hologic) was used for the distal femur analyses done on the model 1000W DXA unit, and version 5.71 (Hologic) was used on the model 2000 DXA units. At each site, a single, specific individual—trained by a single technician at the University of North Carolina, Chapel Hill—was responsible for analyzing the scans.
Statistical Analyses
All analyses were performed using statistical software (version 8.01; SAS
Institute, Cary, NC). Weighted least squares methods were used to build the
predictive models for each of the three distal femur regions for both sexes,
using an individual child's age as the weighted variable. Z-scores for each
individual were calculated from the produced regression equations and weighted
variances.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Side-to-Side Differences
No significant differences in distal femur bone mineral density were found
between the subjects' dominant and nondominant sides. The median ± SE
absolute side-to-side difference was 3.5% ± 0.3% in region 1 (maximum
difference, 18.3%), 2.7% ± 0.2% in region 2 (maximum difference,
15.1%), and 2.5% ± 0.2% in region 3 (maximum difference, 12.2%). All
further analyses were conducted using the mean bone mineral density of the
left and right sides for each subject.
Sex Differences
Bone mineral density was compared in boys and girls as a function of age in
1-year intervals (Fig. 3) and
as a function of Tanner stage (Fig.
4). At no age did we find a statistically significant difference
between boys and girls in bone mineral density in any of the three regions of
the distal femur, although there is tendency for boys to have greater distal
femur bone mineral density than girls in late adolescence.
|
|
Racial Differences
The bone mineral density in regions 1-3 were compared for Caucasians,
African Americans, and the "other races" group. The densities of
the Caucasians and the other races group did not differ, so these two groups
were combined for all analyses. We found that bone mineral density in the
distal femur is greater in African Americans than in Caucasians and the other
races group at all ages, but we emphasize that the study cohort included only
25 African American children. The difference in densities between African
Americans and Caucasians ranged from 0.056 g/cm2 in region 3 among
girls to 0.124 g/cm2 in region 2 among boys (using best-fit
regression modeling with race as a variable). Given the limited number of
African American subjects, these densities should be considered only rough
estimates of the differences between African Americans and non-African
Americans.
Regression Equations
Table 1 shows the best-fit
regression equations for the observed relationship between age (expressed as
years) and bone mineral density (expressed as g/cm2) for
non-African American boys (n = 107) and non-African American girls
(n = 124) for each analysis region in the distal femur. The number of
African American subjects (n = 25) was not sufficient to develop
reliable equations for this racial group.
Figure 5 is a representative
graph of the data showing bone mineral density as a function of age. Included
in Figure 5 are the regression
line ± 2 standard deviation lines derived from the equations in
Table 1.
|
|
Proximal Femur and Lumbar Spine Bone Mineral Density Versus Distal
Femur Bone Mineral Density
Measures of bone mineral density in the proximal femur (total), lumbar
spine (L1-L4), and each region of the three regions of the distal femur were
correlated for each child in the study groups at the Chapel Hill and
Philadelphia sites. Spearman correlational analyses found that bone mineral
density in all regions correlated well with bone mineral density in the other
regions. The densities in the three regions within the distal femur correlated
best with each other, nearly as well with the the density of the proximal
femur but less well with the density of the lumbar spine
(Table 2).
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The question arises in a convenience sampling of a healthy population—such as this study cohort—whether the study group is truly representative of the "normal" population and whether the findings in the study group can be reliably generalized to children in other regions of the country or the world. This concern is a valid one in regard to all data in the literature on bone mineral density in healthy children: we know of no large-scale studies on bone mineral density in children performed using random, population-based sampling.
The issue of reliably defining "normal" bone mineral density measures in the lumbar spine of children has been previously examined. Some differences are to be expected when one is comparing reference data for so-called normal bone density presented by different investigators in different geographic regions using small study groups. One group that compared six such references concluded that lumbar spine bone mineral density measurements for most healthy adolescent populations are comparable despite geographic diversity [12]. However, another group of investigators warned that use of different published reference data for the assessment of children can result in inconsistent diagnostic classification of bone mineral density findings [13]. In the latter study, their conclusion was based on categorization of subjects according to z-scores of less than -2.0 or greater than or equal to -2.0. The magnitude of differences among reference data sets is greatly increased if bone density is evaluated as a categorical variable, with normal and abnormal defined as an all-or-nothing threshold. Such an approach to the interpretation of bone mineral density measures is valuable in describing prevalence in a population but not appropriate in the evaluation of an individual.
We should note that DXA scanners do differ [14]. Therefore, the results of our study in which one brand of DXA scanners (Hologic) was used should not be directly applied to bone density measurements obtained with other brands of scanners.
In conclusion, assessment of bone mineral density in the distal femur of children at risk for osteoporosis is a useful alternative when deformity, fractures, or previous surgery prevent reliable assessment of bone mineral density in the proximal femur. The findings for the healthy study group reported herein can serve in clinical practice to reliably estimate age-normalized z-scores for distal femur bone mineral density measures in non-African American children.
Acknowledgments
We thank M. Gurka in the Department of Biostatistics at the University of
North Carolina, Chapel Hill, for the statistical analyses and D. Riggs and S.
Slezak for assistance with the DXA scanning of many of the patients.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Cohen, E. Lahat, T. Bistritzer, A. Livne, E. Heyman, and M. Rachmiel Evidence-Based Review of Bone Strength in Children and Youth with Cerebral Palsy J Child Neurol, August 1, 2009; 24(8): 959 - 967. [Abstract] [PDF]
|
|
|
|
 |
|
 M. L. Bianchi Diagnosis and Treatment of Bone Fragility in Childhood IBMS BoneKEy, September 1, 2008; 5(9): 323 - 335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A Ward, R. L Ashby, S. A Roberts, J. E Adams, and M Zulf Mughal UK reference data for the Hologic QDR Discovery dual-energy x ray absorptiometry scanner in healthy children and young adults aged 6-17 years Arch. Dis. Child., January 1, 2007; 92(1): 53 - 59. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. C. Henderson, R. K. Lark, M. J. Gurka, G. Worley, E. B. Fung, M. Conaway, V. A. Stallings, and R. D. Stevenson Bone Density and Metabolism in Children and Adolescents With Moderate to Severe Cerebral Palsy Pediatrics, July 1, 2002; 110(1): e5 - 5. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, dashed line) and Caucasian girls (
, solid
line). Mean ± standard error bars are given. At all ages,
densities for boys and girls did not differ significantly (p >
0.01). Plots for regions 1 and 3 were similar.
