Anthropometric adjustments are helpful in the interpretation of BMD and BMC Z-scores of pediatric patients with Prader-Willi syndrome.

Anthropometric adjustments of bone measurements are necessary in Prader-Willi syndrome patients to correctly assess the bone status of these patients. This enables physicians to get a more accurate diagnosis of normal versus abnormal bone, allow for early and effective intervention, and achieve better therapeutic results. INTRODUCTION: Bone mineral density (BMD) is decreased in patients with Prader-Willi syndrome (PWS). Because of largely abnormal body height and weight, traditional BMD Z-scores may not provide accurate information in this patient group. The goal of the study was to assess a cohort of individuals with PWS and characterize the development of low bone density based on two adjustment models applied to a dataset of BMD and bone mineral content (BMC) from dual-energy X-ray absorptiometry (DXA) measurements. METHODS: Fifty-four individuals, aged 5-20 years with genetically confirmed PWS, underwent DXA scans of spine and hip. Thirty-one of them also underwent total body scans. Standard Z-scores were calculated for BMD and BMC of spine and total hip based on race, sex, and age for all patients, as well as of whole body and whole-body less head for those patients with total-body scans. Additional Z-scores were generated based on anthropometric adjustments using weight, height, and percentage body fat and a second model using only weight and height in addition to race, sex, and age. RESULTS: As many PWS patients have abnormal anthropometrics, addition of explanatory variables weight, height, and fat resulted in different bone classifications for many patients. Thus, 25-70 % of overweight patients, previously diagnosed as normal, were subsequently diagnosed as below normal, and 40-60 % of patients with below-normal body height changed from below normal to normal depending on bone parameter. CONCLUSIONS: This is the first study to include anthropometric adjustments into the interpretation of BMD and BMC in children and adolescents with PWS. This enables physicians to get a more accurate diagnosis of normal versus abnormal BMD and BMC and allows for early and effective intervention.

Anthropometric models of bone mineral content and areal bone mineral density based on the bone mineral density in childhood study.

UNLABELLED: New models describing anthropometrically adjusted normal values of bone mineral density and content in children have been created for the various measurement sites. The inclusion of multiple explanatory variables in the models provides the opportunity to calculate Z-scores that are adjusted with respect to the relevant anthropometric parameters. INTRODUCTION: Previous descriptions of children's bone mineral measurements by age have focused on segmenting diverse populations by race and sex without adjusting for anthropometric variables or have included the effects of a single anthropometric variable. METHODS: We applied multivariate semi-metric smoothing to the various pediatric bone-measurement sites using data from the Bone Mineral Density in Childhood Study to evaluate which of sex, race, age, height, weight, percent body fat, and sexual maturity explain variations in the population's bone mineral values. By balancing high adjusted R(2) values with clinical needs, two models are examined. RESULTS: At the spine, whole body, whole body sub head, total hip, hip neck, and forearm sites, models were created using sex, race, age, height, and weight as well as an additional set of models containing these anthropometric variables and percent body fat. For bone mineral density, weight is more important than percent body fat, which is more important than height. For bone mineral content, the order varied by site with body fat being the weakest component. Including more anthropometrics in the model reduces the overlap of the critical groups, identified as those individuals with a Z-score below -2, from the standard sex, race, and age model. CONCLUSIONS: If body fat is not available, the simpler model including height and weight should be used. The inclusion of multiple explanatory variables in the models provides the opportunity to calculate Z-scores that are adjusted with respect to the relevant anthropometric parameters.

Fitting of bone mineral density with consideration of anthropometric parameters.

UNLABELLED: A new model describing normal values of bone mineral density in children has been evaluated, which includes not only the traditional parameters of age, gender, and race, but also weight, height, percent body fat, and sexual maturity. This model may constitute a better comparative norm for a specific child with given anthropometric values. INTRODUCTION: Previous descriptions of children's bone mineral density (BMD) by age have focused on segmenting diverse populations by race and gender without adjusting for anthropometric variables or have included the effects of anthropometric variables over a relatively homogeneous population. METHODS: Multivariate semi-metric smoothing (MS(2)) provides a way to describe a diverse population using a model that includes multiple effects and their interactions while producing a result that can be smoothed with respect to age in order to provide connected percentiles. We applied MS(2) to spine BMD data from the Bone Mineral Density in Childhood Study to evaluate which of gender, race, age, height, weight, percent body fat, and sexual maturity explain variations in the population's BMD values. By balancing high adjusted R (2) values and low mean square errors with clinical needs, a model using age, gender, race, weight, and percent body fat is proposed and examined. RESULTS: This model provides narrower distributions and slight shifts of BMD values compared to the traditional model, which includes only age, gender, and race. Thus, the proposed model might constitute a better comparative standard for a specific child with given anthropometric values and should be less dependent on the anthropometric characteristics of the cohort used to devise the model. CONCLUSIONS: The inclusion of multiple explanatory variables in the model, while creating smooth output curves, makes the MS(2) method attractive in modeling practically sized data sets. The clinical use of this model by the bone research community has yet to be fully established.

Estimation of bone strength from pediatric radiographs of the forearm.

OBJECTIVE: Bone strength is a function of both material and architectural properties. However, bone geometry or architecture, which determines the distribution of bone, is an underappreciated determinant of bone strength. The aim of the study was to evaluate the contribution of only architecture to bone strength. METHODS: We used 2-D (planar) geometric information from radiographs of human radii to construct 3-D finite-element models. To transition from 2-D to 3-D (volume) space, we assumed that all bone cross-sections were elliptical in shape. The finite-element models were subjected to cantilever loading to determine the locations in the bone with the highest propensity to fracture (points of maximum stress). The finite-element-analysis results of the models generated from radiographs of both normal (18) and temporary-brittle-bone-disease (11) infants were subjected to a receiver operating curve analysis. The area under the receiver operating curve was used to evaluate the power of a given bone-strength indicator in segregating the two populations. The actual choice of the material properties (Young's modulus or Poisson's ratio) was not critical for this study, since the finite element analyses were designed to capture the difference in the bone strength of the two populations only based on their architecture. Therefore, the material properties were assumed to be the same in both the normal and TBBD populations. RESULTS: The area under the curve of the bending load required to cause fracture among the two populations was 0.82. Other bone-strength indicators, such as average section modulus, cortical thickness and bone length, were associated with an area under the curve of 0.75, 0.73 and 0.63, respectively. CONCLUSION: The results of the finite-element-analysis suggest that the temporary-brittle-bone-disease population has an altered bone geometry, which increases susceptibility to fracture under normal bending loads.

Thresholding technique for accurate analysis of density and geometry in QCT, pQCT and microCT images.

Computed tomography (CT) is widely used in the assessment of bone parameters in live patients and animals as well as bone samples. Quantitative analysis requires the segmentation of the bone from the surrounding tissue, and most segmentation methods rely on some type of thresholding technique. The aim of this communication is to highlight the influence of threshold selection on various bone parameters and recommend appropriate thresholds. Two types of information are of interest in bone analysis from images: geometric parameters and density parameters. We know from imaging theory that blurring is an inherent byproduct of all imaging methods. Depending on the threshold used for segmentation, the object boundary moves in space due to the sloping edge. It is, thus, critical to select the threshold that creates an object boundary that reflects the actual object size. Similarly, due to blurring, the imaged density shows erroneous values at the object boundaries. Such values must not be included for an accurate representation of the object density. Using a pQCT scanner and a bone phantom with known density and geometry, we show that the thresholds for geometry and density are different. The threshold for accurate geometric segmentation was 49% of the difference of the density between the adjacent tissues. The threshold for accurate density assessment was 95% of the maximum density value of the bone. These specific thresholds are valid only for the scanner tested; however, the principle for selecting the thresholds is valid across scanner platforms and scale of imaging.

Evaluation of cortical bone by computed tomography.

The purpose of this study was to determine the minimum thickness of cortical bone required for the accurate measurement of cortical material density by computed tomography (CT) and to establish normal reference values. A phantom with several wall thicknesses of bone-like material was constructed to simulate various cortical widths. The CT density at each level of thickness was measured on a GE 9800 CT scanner and on the OsteoQuant, a special CT scanner optimized for the measurement of bone in the extremities. The minimum width required to attain the correct material density was determined for each scanner. Additionally, the material density and width of the cortex in the radius and/or femur were measured by CT in 761 healthy subjects, ages 4-84 years. The minimum thickness necessary for an accurate density evaluation of the walls of the phantom by CT was 2-2.5 mm; below these thresholds the values fell in a linear way relative to width. In humans, the material density of cortical bone in the appendicular skeleton was not influenced by height or weight, and the values were similar for all subjects, as long as the cortical width was above 2-2.5 mm. The cortical width increased with age up to 30 years and decreased from 50 years on. We conclude that the material density of cortical bone in the appendicular skeleton can be measured accurately by CT if the thickness of the cortex exceeds 2-2.5 mm.

Correction of scatter in computed tomography images of bone.

A cylindrical aluminum/Plexiglas phantom representing trabecular bone surrounded by various amounts of cortical bone was constructed. Measurements of this phantom using a computed tomography scanner with a 125I photon source demonstrated errors of 0% to 28% in the density of trabecular bone. Two contributing factors are identified: scatter and exponential edge-gradient effect. A simple first-order correction is developed to correct for the scatter-induced error. Relative to the exponential edge-gradient effect, which contributes up to 3.4% error over the range of cortical thicknesses measured, the correction procedure reduces the scatter-induced error to a level of -0.66% to +0.61%. The consistency of the optimized correction parameters with the physical model as well as the effect of scatter measured by the same phantom on a GE 8800 scanner are shown.

A variable-resolution rotate-only computed tomography scanner.

The Rotoscan is a computed tomography scanner that combines the advantages of variable geometric resolution and adjustable size of measurement diameter of translate-rotate scanners with the improved speed of rotate-only scanners. Because of the small number of only 26 detectors used for this scanner, a special data collection scheme of multiple rotations with interleaved detector positions was employed. In order to avoid angular data interpolation after reordering of the projections from the fan- to a parallel-beam geometry, the detectors were incrementally moved at a right angle to the centerline of the fan rather than rotated about the source. The measurement time of 40 s for one cross-section is comparable to that of second-generation systems. However, for longer measurement diameters, the measurement time for second-generation systems increases, whereas that of the Rotoscan remains constant.

Performance evaluation of density measurements of axial and peripheral bone with x-ray and gamma-ray computed tomography.

We examined sources of error in bone measurements made with computed tomography (CT) using a whole-body scanner (GE 8800) and a peripheral-bone CT scanner (developed at the University of Alberta). We investigated the influence of various factors on trabecular bone density: homogeneity and noise in the image plane, linearity of calibration, body size, effects of cortical bone, and the image analysis procedure. With the GE 8800 scanner, the precision (SD) of measurements of a single vertebra is expected to be +/- 1.65% (noise: +/- 0.22%, calibration: +/- 1.3%, analysis: +/- 1%); the accuracy, excluding consideration of marrow fat, varied between -2.7 and +7.3% (compact-bone thickness: 2-5%, body size: -2.5 - +1.5%, calibration: -0.47 - +0.77%). With the peripheral-bone CT scanner, the total precision error (+/- 0.53%) was dominated by noise, with only a minor contribution from the analysis procedure (+/- 0.04%); accuracy varied between -0.6 and +3.4% (effect of cortical bone: up to 3.0%; changes in size of object: -0.59 - +0.4%). The magnitude of these errors was determined under 'ideal' conditions, mostly through phantom measurements; therefore, the errors represent optimistic lower limits in clinical application. Furthermore, measurements of density of cortical bone were not reliable for bone thicknesses of less than about 4 mm with the GE 8800 scanner and less than about 1.5 mm with the peripheral scanner.

A special purpose x-ray fan-beam CT scanner for trabecular bone density measurement in the appendicular skeleton.

A special purpose x-ray CT scanner with the capability of scanning objects 75-220 mm in diameter with constant relative geometrical resolution has been developed. The data collection scheme for the scanner uses multiple rotations of a linearly shifted, asymmetric fan beam permitting user-defined variable resolution. Details of hardware and the calibration procedures for the scanner are described and the methods used to measure trabecular bone density (TBD) in the peripheral skeleton are outlined. The standard error of estimate (SEE) of a calibration line of pixel value as a function of K2HPO4 concentration was determined to be 0.07%. The short-term, in vivo precision of the TBD determination, by repeated measurements of a volunteer with repositioning between each measurement, was +/- 0.67% (coefficient of variation (CV] with a 50s scan time and a radiation dose of less than 20 mR per slice.

Comparison of absorptiometric evaluations from total-body and local-regional skeletal scans.

The most common measurement sites for dual-energy absorptiometry (DXA) in clinical practice are posteroanterior (PA) spine and femur. However, other skeletal regions may provide different bone density information. The purpose of this study was to establish the least number of DXA measurements needed to obtain complete information about bone. A total of 262 normal female subjects, 8-50, were measured on a Lunar DPX-L scanner under total body, PA spine, lateral spine, and femur protocol. Forearm measurements were performed with a Lunar SP2 single-photon absorptiometry scanner. The various measurements were compared based on a linear regression model. The correlation coefficients for bone mineral density (BMD) between adjacent vertebrae were 0.92-0.95, and the associated standard errors of the estimate (SEE) were 4.5-5.5%. Total-body BMD can best predict BMD of the trunk, arms, and legs (SEE<4.3%), but least that of the lateral view of the spine (SEE>13.9%). BMD values of the leg from total-body scans predict those from the femoral neck with an error of 9.0%, and those of the trochanteric region with 11.1%. The error between adjacent vertebrae (6%) is considered acceptable, then a total-body measurement combined with a lateral view of the spine and a femur scan are adequate.

Tibial bone density loss in spinal cord injured patients: effects of FES exercise.

A group of 37 spinal cord injured (SCI) patients underwent bone density measurements at the distal and proximal end of the tibia by a special computed tomography scanner, the OsteoQuant. Fifteen of these patients had follow-up measurements while enrolled in a lower-limb exercise training program with functional electrical stimulation (FES). The pre-exercise measurements revealed a strong correlation (0.88 < or = r < or = 0.90) of trabecular, subcortical, and cortical bone density between the distal and proximal ends of the tibia. The expected bone density loss during the first two years post injury (as calculated from the regression lines of bone density vs. time post injury) amounted to 51.5% for trabecular, 44.2% for subcortical, and 32.7% for cortical bone. No major bone density loss was calculated after 7 years post injury. Analysis of the bone density data during the FES exercise program revealed various degrees of loss. However, the rate of bone loss for this FES exercise group was less than expected from the regression lines. The reduction of bone loss was between 0.2 and 3.3% per year, and was significant (p < 0.05) for all bone parameters at the distal end and for trabecular bone density at the proximal end of the tibia. These bone density measurements revealed a potentially positive effect of FES exercise intervention for the rehabilitation of SCI patients.

Race and sex differences in bone mineral density and geometry at the femur.

INTRODUCTION: Differences in osteoporotic hip fracture incidence between American whites and blacks and between women and men are considered to result, in part, from differences in bone mineral density and geometry at the femur. The aim of this study was to quantify differences in femoral bone density and geometry between a large sample of healthy American white and black women and men. SUBJECTS AND METHODS: Healthy American white (n=612) and black (n=164) premenopausal women, aged 23 to 57 years, and healthy American white (n=492) and black (n=169) men, aged 20 to 63 years, had volumetric bone mineral density (vBMD) and geometry variables measured at the femur by computerized tomography (CT), and areal bone mineral density (aBMD) at femoral neck measured by dual X-ray absorptiometry (DXA). RESULTS: American blacks had higher vBMD at the femoral neck and femoral shaft cortex than American whites whereas femoral axis length and femoral neck area were not different. Men had lower vBMD at the femoral neck and femoral cortex than women but had greater femoral axis length and femoral neck area than women. The higher aBMD in American blacks than whites persisted after correction for measured area whereas the higher aBMD in men than women disappeared. CONCLUSIONS: At the femoral neck, American whites have lower bone density than American blacks but similar geometry. Women have higher bone density than men in both races but have smaller geometry variables. The differences in bone density may account in part for the differences in hip fracture incidence between American blacks and whites, whereas the differences in femur size may account for the differences in hip fracture rates between men and women.

A study of the long-term precision of dual-energy X-ray absorptiometry bone densitometers and implications for the validity of the least-significant-change calculation.

INTRODUCTION: Short-term precision is often quoted and used as the most important performance parameter of a dual-energy X-ray absorptiometry (DXA) scanner; however, long-term precision has a more profound impact on patient monitoring. Long-term precision refers to the combination of in-vivo precision errors and long-term equipment stability. METHODS: To monitor long-term equipment stability, a phantom was designed with four inserts ranging in bone-mineral density from 0.5 to 3.3 g/cm(2). This phantom was used to monitor the equipment stability of four modern fan-beam densitometers, two each from Hologic and GE/Lunar, over a 4-year period. Manufacturer-recommended quality assurance (QA) procedures were performed, and the scanners stayed within manufacturer-specified tolerances throughout the study. RESULTS AND CONCLUSION: During the 4-year period, the Hologic scanners were observed to cause clinically insignificant BMD shifts (maximum of 0.34%), whereas the GE/Lunar scanners revealed BMD shifts that were clinically significant (1.5% and 2.1%). As a result, using least-significant-change (LSC) calculations based only on short-term in-vivo precision studies for monitoring patients is not valid for the two GE/Lunar densitometers due to the poorer long-term stability they exhibited.

Recommendations for thresholds for cortical bone geometry and density measurement by peripheral quantitative computed tomography.

Peripheral quantitative computed tomography (pQCT) is widely used for clinical and research purposes. For accurate determination of bone geometry (bone cross-sectional area, cortical thickness, and cortical area), volumetric bone mineral density (vBMD) and cortical bone mineral content (BMC), it is important to select the appropriate thresholds. A Stratec XCT-2000 scanner was used to compare current standard practice with new optimized thresholds. Currently, a single threshold of 710 mg/mL for the measurement of cortical vBMD and geometry is used. We hypothesised that this threshold may not be optimal and used the European Forearm Phantom (EFP) and patient data to test more appropriate thresholds. A single slice (1.2 mm width, 0.4 mm pixel size) was made at section 4 of the EFP (representing the diaphyseal portion of a long bone). The EFP has a known cortical thickness of 2.5 mm and, therefore, the correct threshold for geometry would be that which measures cortical thickness as 2.5 mm. Thresholds were altered at approximately the 50% value between soft tissue (60 mg/mL) and peak density (879 mg/mL), and cortical thickness versus threshold was plotted; the correct threshold for geometry was 460 mg/mL. By expressing this threshold as a percentage of the range of density values in the EFP ([460-60]/[879-60] = 49%) and then applying this percentage to in vivo data, the optimum threshold for geometry can be determined: ([1240-79] x 0.49) + 79 = 648 mg/mL. For cortical vBMD of in vivo bone measurements at the midshaft site of the radius, thresholds were varied around the peak value (1240 mg/mL), and the threshold was set to that which gave a cortical density of 1240 mg/mL; the threshold for cortical density was, therefore, 1200 mg/mL. A subset of radius scans from a population of young healthy females was analyzed using the new thresholds (648 mg/mL for bone geometry, 1200 mg/mL for cortical vBMD) versus the current threshold (710 mg/mL). For bone geometry, the mean difference between the analysis based on the new threshold and that based on the manufacturer-recommended threshold ranged between 2.1% and 14% (total area = 2.1%, cortical thickness = 14%, cortical area = 3.7%). Although there was a 10% difference between the analysis based on the new threshold and that based on the manufacturer-recommended threshold, this difference was not systematic. Thresholds will significantly affect results obtained from pQCT. The current threshold of 710 mg/mL is inadequate for accurate determination of bone geometry and cortical vBMD. New thresholds of 648 mg/mL for geometry and 1,200 mg/mL for cortical vBMD should be used.

The OsteoQuant: an isotope-based CT scanner for precise measurement of bone density.

OBJECTIVE: We attempted to design and construct a computed tomography scanner with an in vivo precision of better than 0.5% for trabecular bone density of the radius. MATERIALS AND METHODS: A number of considerations involving physical limitations, stability of the system, and cost led to the development of the OsteoQuant, an isotope-based computed tomography scanner working on the translate-rotate principle. With 16 detectors providing a total of 128 projections and 256 data points per projection, the measurement time for one cross section is typically 90 s. Optimal for bone measurements in arms and legs, 125I was chosen as the photon source. The detectors are photomultipliers with Nal(TI) crystals employed in the counting mode. Usually, six to ten slices are measured at a given site, 2 mm apart from each other, and bone density is calculated for trabecular, subcompact, and compact bone. For repeat measurements, the evaluation sites are carefully matched, and the same volume of bone is analyzed at each measurement occasion. RESULTS: The long-term precision of the scanner, measured with a water cylinder, is 0.03%. This error includes the performance of the scanner hardware, calibration of the photon count rates, and reconstruction process. In vivo precision is influenced by additional factors such as slice positioning, patient cooperation, and bone contour detection. At the distal end of the tibia, trabecular bone density can be measured with a precision of 0.1%. The error for trabecular bone density in the radius is 0.3%. CONCLUSION: The OsteoQuant surpasses the design goals and represents an ideal instrument to assess small changes in bone density over time.

Bone density measurements by computed tomography in osteogenesis imperfecta type I.

The objectives of this study were (1) to determine whether there are differences in bone density in children versus adults with osteogenesis imperfecta type I (OI-type I) using computed tomography (CT) bone density measurements, (2) to determine whether there are differences in bone density between normal infants and infants with OI-type I using CT bone density measurements and (3) to determine whether CT bone density measurements could be helpful in investigating the infant with unexplained fractures. CT bone density measurements determine both the cortical bone density (CBD) and the trabecular bone density (TBD). CT bone density was determined using the OsteoQuant in 14 individuals with OI-type I who ranged in ages from 8 months to 45 years. The control groups consisted of over 1000 normal individuals, mostly adults, and included 7 normal infants who ranged in age from 10 months to 27 months. One of the individuals with OI-type I was a 4-month-old infant with multiple, unexplained fractures who had no other features of OI-type I and whose parents were accused of child abuse. Infants and children with OI-type I had low CBD and low TBD compared with normal controls, whereas adults with OI-type I had low TBD and high CBD when compared with controls. The one infant with multiple unexplained fractures and no other features of OI-type I had a bone density profile suggesting OI-type I with a low TBD and low CBD. Subsequent collagen analysis showed biochemical evidence of OI-type I. Individuals with OI-type I have abnormal CT bone density profiles that evolve over time from a low CBD and low TBD during infancy and childhood to a high CBD and low TBD during adulthood. This may explain the decreased frequency of fractures in individuals with OI-type I in adulthood compared with childhood. Individuals with OI-type I can present with only multiple unexplained fractures and have no other clinical features to strongly suggest the diagnosis. CT bone density measurements can be helpful in these atypical cases of OI-type I and should be considered in the investigation of the infant with unexplained fractures to help distinguish intrinsic bone disease from child abuse.

Quantitative measurement of bone density using gamma-ray computed tomography.

A special purpose gamma-ray computed tomography scanner has been developed for precise measurements of bone density in the human appendicular skeleton. Details of the scanner's hardware and of the software organization for system control and data analysis are given, together with an outline of the theoretical basis for conversion of measured linear attenuation coefficients to physical bone densities. Performance of the system was evaluated on bone-like phantoms. Clinically, a precision of +/- 0.5% is obtained for bone density determinations. This device is being used in experimental studies and clinical investigations.

Temporary brittle bone disease: association with decreased fetal movement and osteopenia.

Infants who present with multiple unexplained fractures pose a difficult diagnostic dilemma of child abuse versus intrinsic bone disease. Temporary brittle bone disease is a recently described disease characterized by a transient bone weakness in the first year of life which presents with multiple, unexplained fractures that can be confused with child abuse. The purpose of this study was to determine if there are common, historical features in infants with unexplained fractures that might suggest a basis for the fractures, and to determine if bone density measurements might indicate that such infants have low bone density. Medical records were reviewed in 33 infants who were referred for consultation for multiple unexplained fractures in which the parents and other caregivers denied wrongdoing. In 9 of the infants, radiographic absorptiometry and/or computed tomography bone density studies were performed. In 26 of these infants the diagnosis of temporary brittle bone disease was made. A normal collagen test was found in 17 of the 26 infants studied; 9 infants did not have a collagen test because the diagnosis of osteogenesis imperfecta was considered highly unlikely. In 25 of them there was a history of decreased fetal movement and/or intrauterine confinement. Bone density, as judged by plain X-ray films, was normal in all 26 cases, but when formally measured by radiographic absorptiometry or computed tomography, the bone density measurements were low in 8 of the 9 infants studied. These findings implicate decreased fetal movement and intrauterine confinement as contributing factors to temporary brittle bone disease and suggest that normal, unconstrained fetal movement during pregnancy is important for normal fetal bone formation. These findings support the model that bone formation and strength are dependent on the mechanical load placed on the bone. The results also demonstrate the usefulness of bone density measurements in evaluating the infant with multiple unexplained fractures to help distinguish nonaccidental injury from intrinsic bone disease.

Influence of fat on bone measurements with dual-energy absorptiometry.

In order to investigate the influence of fat on bone in dual-energy absorptiometry measurements, we evaluated a special phantom on the three scanners: Lunar DP3, Lunar DPX and Hologic QDR-1000. The phantom employed hydroxyapatite blocks of various thicknesses to simulate bone, water to simulate muscle and lucite to simulate fat. The lucite plates were arranged in one and two layers in three different configurations: over the whole measurement area, over the hydroxyapatite blocks only and at both sides of the hydroxyapatite blocks. For all scanners, no influence of fat could be demonstrated if it was homogeneously distributed over the whole measurement area. However, changes in area bone-density were observed if fat was distributed inhomogeneously over the measurement area. Fat over only the bone area reduced the measured bone values by 0.051 g/cm2 per cm fat layer. Fat over only the soft-tissue area increased the measured bone values by the same amount. These results apply to the Lunar DPX scanner. The results for the Lunar DP-3 scanner are similar; those for the Hologic QDR-1000 show a slightly smaller fat dependence of 0.044 g/cm2 per cm fat layer. The fat influences are not dependent on the amount of bone and only minimally on the soft-tissue thickness. A change of 50% in the fat content of the bone marrow will change the measured area bone-density of an averaged sized vertebra by 5-6% depending on scanner model. Inhomogeneous fat distribution in soft tissue, resulting in a difference of 2 cm fat layer between soft-tissue area and bone area, will influence the measured area bone-density by 9-10%.