Constant mineralization density distribution in

Bone 32 (2003) 316 –323
www.elsevier.com/locate/bone
Constant mineralization density distribution in cancellous human bone
P. Roschger,a,* H.S. Gupta,a,b A. Berzlanovich,c G. Ittner,d D.W. Dempster,e,f P. Fratzl,b
F. Cosman,e,g M. Parisien,f R. Lindsay,e,g J.W. Nieves,e,h and K. Klaushofera
a
Ludwig Boltzmann Institute of Osteology, 4th Medical Department, Hanusch Hospital & UKH-Meidling, Vienna, Austria
Erich Schmid Institute of Materials Science, Austrian Academy of Sciences and University of Leoben, Leoben, Austria
c
Institute of Forensic Medicine, University of Vienna, Vienna, Austria
d
Traumatology Hospital Meidling, Vienna, Austria
e
Helen Hayes Hospital, New York State Department of Health, West Haverstraw, NY, USA
f
Department of Pathology, Columbia University, New York, NY, USA
g
Department of Medicine, Columbia University, New York, NY, USA
h
Department of Epidemiology, Columbia University, New York, NY, USA
b
Received 19 April 2002; revised 15 November 2002; accepted 19 November 2002
Abstract
The degree of mineralization of bone matrix is an important factor in determining the mechanical competence of bone. The remodeling
and modeling activities of bone cells together with the time course of mineralization of newly formed bone matrix generate a characteristic
bone mineralization density distribution (BMDD). In this study we investigated the biological variance of the BMDD at the micrometer
level, applying a quantitative backscattered electron imaging (qBEI) method. We used the mean calcium concentration (CaMean), the most
frequent calcium concentration (CaPeak), and full width at half maximum (CaWidth) to characterize the BMDD. In none of the BMDD
parameters were statistically significant differences found due to ethnicity (15 African–American vs. 27 Caucasian premenopausal women),
skeletal site variance (20 ilium, 24 vertebral body, 13 patella, 13 femoral neck, and 13 femoral head), age (25 to 95 years), or gender.
Additionally, the interindividual variance of CaMean and CaPeak, irrespective of biological factors, was found to be remarkably small (SD
⬍ 2.1% of means). However, there are significant changes in the BMDD in the case of bone diseases (e.g., osteomalacia) or following
clinical treatment (e.g., alendronate). From the lack of intraindividual changes among different skeletal sites we conclude that diagnostic
transiliac biopsies can be used to determine the BMDD variables of cancellous bone for the entire skeleton of the patient. In order to quantify
deviations from normal mineralization, a reference BMDD for adult humans was calculated using bone samples from 52 individuals.
Because we find the BMDD to be essentially constant in healthy adult humans, qBEI provides a sensitive means to detect even small changes
in mineralization due to bone disease or therapeutic intervention.
© 2003 Elsevier Science (USA). All rights reserved.
Keywords: Bone mineralization density distribution; Cancellous human bone; Adult human bone; Quantitative backscattered electron imaging
Introduction
The mechanical properties of bone are established by the
structural organization at different hierarchical levels, including organ, tissue, and material level [1]. It has been
shown that bone diseases can affect different levels of bone
* Corresponding author. Ludwig Boltzmann-Institute of Osteology,
UKH-Meidling, Kundratstrasse 37, A-1120 Vienna, Austria.
E-mail address: [email protected] (P. Roschger).
structure, such as trabecular architecture in osteoporosis
[2,3], the degree of mineralization in osteomalacia [4], and
the nanostructure of the composite material in osteogenesis
imperfecta [5–7,40]. Clinical evaluations of biochemical
markers, radiography, DEXA, QCT, pQCT, and NMR are
not always sufficient to make correct clinical diagnoses. In
such cases, bone biopsies (usually transiliac bone) are taken
from the patients for histological studies. Additionally,
these bone biopsies provide the opportunity to examine the
bone quality at different hierarchical levels by other physical methods, e.g., (a) ␮CT and magnetic resonance micro-
8756-3282/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved.
doi:10.1016/S8756-3282(02)00973-0
P. Roschger et al. / Bone 32 (2003) 316 –323
imaging (MR␮I) [3,8,41], which are powerful tools to evaluate 3D-structural parameters of trabecular bone, (with a
spatial resolution down to 15 ␮m achievable by commercially available CT-scanners); (b) microradiography [9 –12]
and quantitative backscattered electron imaging (qBEI)
[4,13–15], which quantify the degree of mineralization
within sectioned bone regions; (c) synchrotron radiation
microtomography, which permits the analysis of the 3D
structure together with degree of mineralization [16] in
bone; (d) scanning small angle X-ray scattering (scanningSAXS) [17,18], which enables the determination of mineral
particle size, shape, and orientation; (e) Fourier transform
infrared microspectroscopy (microFTIR) [19 –21], which
can provide local information on variations in mineral:
matrix ratios, in carbonate:phosphate ratios of the mineral,
in crystallinity, and in collagen cross-linking; and (f) Raman
spectroscopy [22], which provides insight into molecular
variations in the structure of the mineral. We focus on the
assessment of the local calcium distribution within the trabecular bone matrix using qBEI [4,23]. The degree of mineralization is a crucial factor for bone quality, because it
modifies the elastic modulus of the matrix material—the
higher the degree of mineralization, the higher the stiffness
of the material [24].
As is shown by microradiography [10 –12] and backscattered electron imaging [4,13,15,23], the bone matrix is not
uniformly mineralized, but exhibits a range of mineral concentration, which is determined primarily by the duration of
secondary mineralization of the individual bone packets.
The differences in the degree of mineralization within a
certain bone region can be quantified by measurement of the
frequency distributions of calcium concentrations detected.
We designated such a distribution as the bone mineralization density distribution (BMDD), synonymous to bone
mineral density distribution used previously [4]. For the
reasons noted above, BMDD indirectly provides information on the bone turnover rate. For instance, a high bone
turnover rate will result in a higher contribution of less
mineralized matrix and will shift the distribution to lower
mineral concentration values. Additionally, the distribution will get broader, because of the increase in the
heterogeneity in mineralization [25,26]. In contrast, when
bone turnover is reduced, e.g., by treatment with an
antiresorptive drug [19,27,28], the distribution shifts to
higher calcium concentration values and becomes more
narrow, the sharpening indicating that the mineralization
is more homogeneous.
To gain insight into the mineralization pattern of normal
bone and to establish the BMDD as a diagnostic tool to
detect deviations from normality, we studied the BMDD of
trabecular bone from different skeletal sites from healthy
individuals of variable ethnicity, gender, and age by quantitative backscattered electron imaging (qBEI) [4] in the
scanning electron microscope.
317
Materials and methods
Samples
Human bone biopsies were obtained from the sources
listed below. No evidence of metabolic bone disease or post
mortal alterations in mineralized tissue was present in any
of the groups of bone samples designated P1 to P4 in the
following: P1: transiliac bone samples from autopsies of 20
individuals (7 males, 13 females, ages 30 – 85 years) from a
previous study [4]; P2: L4-vertebral samples from autopsies
of 24 individuals (16 males, 11 females, ages 5–95 years);
and P3: transiliac biopsies from a previous study [29] of 42
premenopausal women (volunteer; 27 white, 15 black) of
similar ages (ages 26 –37 years). Subjects were within 20%
of ideal weight for height, and had no endocrine, renal,
hepatic, or other disorders. They also had low or moderate
alcohol and tobacco use (P4: cancellous bone samples of
patella, femoral head, and femoral neck, from 13 individuals
each (nine males and four females, aged 39 – 65 years)).
Additionally, an example of severe deviations from the
normal BMDD in a transiliac biopsy from a patient (female,
39 years) with osteomalacia caused by celiac disease is
shown. A case with only subtle changes in BMDD is represented by the BMDD from an osteoporotic patient (female, 63 years) treated for 2 years with alendronate (10
mg/day) (for more details see recent work [28]). The study
was approved by the Institutional Ethical Review Board
(Institute of Forensic Medicine, Vienna, Austria) for the
autopsy samples P1, P2, and P4, and by the Institutional
Ethical Review Boards of Helen Hayes Hospital, Columbia—Presbyterian, Mount Sinai, and St. Lukes—Roosevelt
Medical Centers for P3 [29].
The samples were fixed in 70% ethanol (P3) or ethanol/
formalin solution 70:30 v/v (P1, P2, and P4) and dehydrated
in a graded series of ethanol solutions [4]. Subsequently,
they were embedded in polymethylmethacrylate (PMMA),
and the surfaces were made planar and parallel to each other
by grinding and polishing (PM5 Logitech, Glasgow, Scotland). A thin carbon layer was deposited by vacuum evaporation (Argar SEM Carbon coater, Argar Scientific Limited, Essex, UK) for qBEI analysis.
Bone mineralization density distribution (BMDD)
measurements
The mineral content of bone in these samples was measured by a quantitative backscattered electron imaging
(qBEI) method [4,23]. This technique relies on the fact that
the amount of electrons backscattered from a thin surface
layer (1–2 microns thick) of a sectioned bone region is
directly proportional to the weight concentration of calcium
within the underlying interaction volume of the beam electrons. The samples were analyzed using a digital scanning
electron microscope with a four-quadrant semiconductor
backscattered electron detector (DSM 962, Zeiss,
318
P. Roschger et al. / Bone 32 (2003) 316 –323
frequency (given as percentage of bone area) of occurrence
of pixels of a certain gray level, or Ca concentration (Fig.
1b). These histograms, designated as BMDD, have a bin
width of a single grey level step resulting in a resolution in
Ca concentration of 0.17 wt%. In order to perform quantitative comparisons and statistical analysis from different
BMDDs, we made a further data reduction and discarded
the information of each single histogram bin and introduced
BMDD parameters characterizing the total distribution: the
weighted mean calcium concentration (CaMean), the most
frequent Ca concentration (CaPeak), and the width of the
distribution (CaWidth) (Fig. 1b). For the determination of
CaMean, a technical precision (interassay variance) of 0.3%
was achieved by this method [4].
Statistical analysis
Fig. 1. (a) An example of a digital gray level backscattered electron image
of a trabecular bone sample (iliac crest, black woman (age 33 years). (b)
Corresponding bone mineralization density distribution (BMDD) including
the BMDD parameters. Cai, Fi are coordinates of a histogram data point;
Cai is showing the calcium weight percent of the ith histogram bin (on the
x axis); and Fi the corresponding bin height, the frequency of appearance
(at the y axis). The peak indicates the most frequently occurring calcium
concentration and its coordinates are CaPeak and FPeak. CaWidth ⫽ the full
width at half maximum of the histogram. CaMean ⫽ weighted mean Ca
concentration.
In order to examine the effect of different biological
variables on the BMDD systematically, we investigated the
influence of individuality, ethnicity (African–Americans vs.
Caucasians), skeletal site (ilium, vertebra, patella, femoral
neck, and femoral head), age (25 to 95 years), and gender on
BMDD using the qBEI method.
Unpaired t tests were used to test for equality of means.
Significance for nonequality was set at P ⬍ 0.05. ANOVA
of repeated measurements usually applied to cases for analysis of treatment effects in an individual at different time
points was employed to search for differences between the
three skeletal sites studied within an individual at one time
point. Linear regression analysis with age was performed
for BMDD variables in the age range from 25 to 97 years.
Statistical analyses were carried out with the R 1.2.2 package for statistical computation [30] and the statistical software StatView 4.5 (Abacus Concepts, Inc. Berkeley, CA,
USA).
Results
Interindividual variance
Oberkochen, Germany), at a working distance of 15 mm.
Probe current was adjusted to 110 ⫾ 0.4 pA and electron
beam energy to 20 keV. The pixel resolution of the digital
image was 4 microns/pixel at magnification of 50⫻, with a
resolution of 256 gray levels. For each sample, 5– 6 regions
of the same size (2 ⫻ 2.5 mm wide) were scanned, using a
scan speed of 100 s per frame. Carbon and aluminum were
used for gray level references and osteoid and hydroxyapatite were employed as references to convert gray level
values into calcium wt% values. Thus, the single digital
backscattered electron (BE) images display the specific spatial variations of Ca concentration within the imaged bone
region (Fig. 1a). In order to get more general quantitative
data about the variations of mineralization in the bone
matrix, we discarded the spatial information and produced
gray level histograms from the BE images to show the
The BMDD variables for 27 premenopausal white females (group P3) of approximately the same age, similar
weights, and with similar health status were determined at
the same skeletal site (iliac crest) and were averaged (Table
1). The standard deviations were less than 2.3% for CaMean
and CaPeak, and less than 12.5% for CaWidth for this group of
individuals.
Ethnic origin influence
A group of 27 white and 14 black (age range 26 to 37
years) premenopausal females (group P3) was analyzed.
Fig. 2 compares the average BMDD for white and black
women. t Tests between the BMDD parameters in the two
ethnic groups showed no significant difference in CaMean,
P. Roschger et al. / Bone 32 (2003) 316 –323
319
Table 2
Intraindividual skeletal site influence on BMDD
Table 1
Interindividual and ethnic origin influence on BMDD
BMDD
parametera
Whiteb
Blackb
Pc
Skeletal sitea
CaMean [wt% Ca]
CaPeak [wt% Ca]
CaWidth [wt% Ca]
CaMean (wt% Ca)
CaPeak (wt% Ca)
CaWidth (wt% Ca)
21.87 (⫾2.29%)
22.67 (⫾2.16%)
3.62 (⫾12.15%)
21.87 (⫾1.60%)
22.60 (⫾1.42%)
3.35 (⫾10.15%)
0.99
0.66
0.34
Patella
Femoral
neck
Femoral
head
22.35 (⫾2.28%)
22.32 (⫾1.79%)
22.96 (⫾1.79%)
22.89 (⫾1.40%)
3.01 (⫾10.30%)
2.99 (⫾6.35%)
22.38 (⫾1.92%)
22.92 (⫾1.75%)
2.92 (⫾10.96%)
a
Mean values with SD given as percentage of means in parentheses.
Two groups of premenopausal women, 27 white and 15 black.
c
t Test P values indicating no significance between groups.
b
CaPeak, and CaWidth (Table 1). The differences between
means were found to be less than 0.3% for CaMean and
CaPeak and 7.5% for CaWidth.
Intraindividual skeletal site influence
a
Three skeletal sites from the same individual, studied in 13 cases:
one-way repeated measures ANOVA revealed no statistically significant
differences among sites (P ⫽ 0.87). Mean values with SD given as
percentage of means in parentheses.
no significant change within this age range (Table 3 and Fig.
4).
Gender influence
To test for differences in BMDD of cancellous bone
among different skeletal sites within the same individual,
the BMDDs from patella, femoral head, and femoral neck
were measured from 13 individuals (group P4). ANOVA of
repeated measurements revealed no significant differences
(P ⫽ 0.87) between the different skeletal sites (Table 2).
The similarity between the BMDDs at different skeletal
sites is demonstrated for one single individual in Fig. 3.
Because we observed no significant age dependence for
any of the BMDD parameters for males and females taken
as separate groups or taken as a combined group, the same
data set used in the previous analysis was now tested for
differences due to gender. t Tests revealed no significant
difference in population means between these two sets of 22
females and 13 males (Table 4).
Age influence
Iliac vs. vertebral bone
Because there was no ethnicity or skeletal site influence
in the BMDD variables, we included bone samples from
different ethnic origin and skeletal sites in this investigation.
The changes of CaMean, CaPeak, and CaWidth with age were
studied for a combination of sample groups P1, P2, and P3.
Twenty-two females (13 subjects from P1, five from P2,
four from P3) and 13 males (seven subjects from P1 and six
from P2) were used. Only subjects aged 25 years or older
were included. Linear regressions vs. age were fit for each
BMDD parameter. For both male and female, as well as
combined populations, CaMean, CaPeak, and CaWidth showed
To determine if a transiliac biopsy is representative of
trabecular bone in the spine, we compared CaMean, CaPeak,
and CaWidth between a set of transiliac bone samples (P1),
and a subset of the L4-vertebral samples (P2: 11 samples,
subjects older than 25 years). No significant difference in
BMDD between these two skeletal sites was detected (Table
5).
Fig. 2. Comparison of BMDD between black and white American women:
averaged BMDD histograms of iliac crest samples from 15 black and 27
white premenopausal women of mean age 32 years.
Reference BMDD
Because we found no statistically significant influence
due to biological factors in the BMDD parameters between
adult individuals, we calculated a typical BMDD for the
Fig. 3. Comparison of BMDD histograms from different skeletal sites
within a single individual.
320
P. Roschger et al. / Bone 32 (2003) 316 –323
Table 3
Age (25–97 years) influence on BMDD: linear regression analysis with age
Variable
Gender combinationa
Intercept
[wt% Ca]
CaMean
Male ⫹ female
Male
Female
21.71
21.58
21.49
CaPeak
Male ⫹ female
Male
Female
CaWidth
Male ⫹ female
Male
Female
a
b
R2
Pb
0.0076
0.0084
0.0144
0.1079
0.1683
0.1657
0.054
0.16
0.09
22.66
22.60
22.53
0.0054
0.0059
0.0075
0.0752
0.0960
0.1152
0.11
0.30
0.17
3.46
3.40
3.54
⫺0.0001
⫺0.0001
⫺0.0007
0.0002
0.0000
0.0023
0.97
0.98
0.85
Slope
[wt% Ca]/year
Population of 22 females and 13 males.
P values from ANOVA statistics for the regression. None of the variables were significantly correlated with age.
trabecular bone of normal adult human by averaging the
BMDDs of 52 samples without regard to age, gender, ethnicity, or skeletal site, as a reference with which pathological samples can be compared visually (Fig. 5). For averaging the BMDDs, all 20 samples from P1, 11 from P2 (five
female, six male), eight from P3 (four black and four white),
and 13 femoral head samples from P4 (nine males, four
females) were used.
Further, we calculated the means and standard deviations
of the 52 individual BMDD variables to use as normative
BMDD values for quantitative comparisons with pathological samples (Table 6). This allows deviations from normal-
ity to be quantified in terms of multiples of the SD of the
normative values. In order to have a parameter that is
especially sensitive to changes in the BMDD in the range of
low degree of mineralization, we introduced a low mineralization cutoff in the calcium concentration, in analogy to
a 95% confidence interval in statistics, such that 95% of the
bone area in a “normal” bone has a higher calcium concentration than this cutoff concentration. This cutoff value at
the fifth percentile of our reference BMDD was found to be
17.68 wt% Ca (Fig. 5). Accidentally, this value lies in the
range of Ca concentrations, which are achieved during primary mineralization of newly formed bone matrix [12,26].
Consequently, a new BMDD parameter CaLow was defined
as the percentage of bone area which is mineralized less
than 17.68 wt% Ca. Thus, CaLow for “normal” bone is ⬃5%
(Table 6). To demonstrate the utility of the normative
BMDD together with the BMDD parameters CaLow, as well
as CaMean, CaPeak, and CaWidth, in quantifying deviations
from normal BMDD, a case of osteomalacia and a case of
osteoporosis treated with alendronate [28] are shown in Fig.
5 and Table 6. In the case of osteomalacia, an increase of
CaLow (19.7 SD) and CaWidth (8.8 SD) combined with a
decrease in CaMean (⫺9.8 SD) and CaPeak (⫺8.3 SD) could
be detected. In contrast, the osteoporotic patient treated with
alendronate exhibited only minor differences in BMDD
compared to healthy individuals.
Table 4
Gender influence on BMDD
Fig. 4. Changes in the BMDD parameters CaMean, CaPeak, and CaWidth with
age for cancellous bone: iliac and L4-vertebral bone from a group of 22
females and 13 males aged ⬎ 25 years were studied. Solid lines are linear
regressions with age of combined male and female data. Ninety-five percent confidence intervals for the regression are shown by dashed lines
(corresponding regression analysis data are given in Table 3).
Variablea
Male (n ⫽ 13)
Female (n ⫽ 22)
Pb
CaMean [wt% Ca]
CaPeak [wt% Ca]
CaWidth [wt% Ca]
22.08 (⫾2.08%)
22.95 (⫾1.83%)
3.40 (⫾5.88%)
22.17 (⫾2.21%)
22.98 (⫾1.74%)
3.49 (⫾7.74%)
0.61
0.89
0.32
a
Mean values with SD given as percentage of means in parentheses.
t Test P values indicating no statistically significant differences between groups.
b
P. Roschger et al. / Bone 32 (2003) 316 –323
Table 5
Comparison of BMDD between iliacal and vertebral bone
BMDD parametera
Iliac crest (n ⫽ 20)
L4-vertebra (n ⫽ 11)
Pb
CaMean [wt% Ca]
CaPeak [wt% Ca]
CaWidth [wt% Ca]
22.20 (⫾2.21%)
23.05 (⫾1.56%)
3.41 (⫾7.04%)
21.98 (⫾2.27%)
22.81 (⫾2.10%)
3.56 (⫾6.74%)
0.23
0.13
0.10
a
Mean values with SD given as percentage of means in parentheses.
t Test P values indicating no statistically significant differences between sites.
b
Discussion
Numerous studies at organ and tissue levels have shown
that bone variables like bone mineral density (BMD), bone
mineral content (BMC), cortical bone area relative to trabecular area, bone volume, trabecular thickness, osteoid
perimeter, and thickness can exhibit distinct variations with
ethnicity [31–34], gender [2,35], age [2,31,32,36], and
menopausal status [31,32]. In contrast, at the material level,
we found that the BMDD is strikingly constant for trabecular bone. None of the biological factors such as interindividual variability, ethnic origin, skeletal site, age, or gender
were found to cause significant alterations in BMDD. To
our knowledge, this is the first systematic study of BMDD
that covers all these essential biological factors. The few
previous studies [4,10,13,14,37] on this topic are consistent
with the present results for cancellous bone. However,
BMDD in cortical bone can have remarkable variations
even within a single skeletal site (data not shown and
Portigliatti Barbos et al. [38]). The highly heterogeneous
osteonal structures found in cortical bone might be responsible for these variations. These are most likely generated by
locally different levels of bone turnover, perhaps as a response to exposure to varying mechanical stress. For the
systematical investigation of the normal cortical BMDD
with respect to all these biological factors, a suitable BMDD
evaluation method taking into account the heterogeneity of
osteonal structure has to be developed.
Ethnicity (African–American vs. white Americans) does
not seem to affect the material level. This makes it unlikely
that differences at the bone material level are contributing
essentially to the lower incidence for osteoporosis [38,42]
and hip fracture [39] in blacks compared to whites. Numerous studies comparing BMD and bone histomorphometry
between ethnic groups have been carried out in an attempt
to explain the difference in fracture incidence [29,31–34],
but the results are not conclusive. In a previous histomorphometric study [29], on the same bone samples used in the
present study, the authors found no difference in bone volume, microstructure, or turnover, but the bone formation
period and mineralization lag time were longer in blacks
compared to whites. However, surprisingly, these differences in dynamics of bone formation seem to have no
significant impact on the BMDD parameters measured in
trabecular bone.
321
The similarity of BMDD observed in cancellous bone at
different skeletal sites (patella, femoral neck, femoral head,
vertebral body, and iliac crest) is surprising. This suggests
that bone modeling and remodeling is biologically controlled in such a way that under physiological conditions
trabecular bone of similar mineralization pattern is produced, independent of the skeletal site and despite differences in mechanical loading conditions. In consequence,
deviations of BMDD as measured by qBEI in iliac crest
give strong evidence of pathological changes in the entire
skeleton, which confirms the feasibility of iliac crest as a
skeletal site for a routine diagnostic biopsy. Unfortunately
we could not directly compare the iliac vs. the vertebral sites
within the same individual but only between different individuals. However, because the interindividual variability
was so small, we believe the conclusion is valid.
Strikingly small changes at the bone material level were
found in the age range 25 to 97 years. This is in strong
contrast to the tissue level, where mass and microarchitecture are strongly age dependent. We found that CaMean,
CaPeak, and CaWidth did not significantly change within the
age range of 25 to 90 years. This is consistent with previous
findings [4,10,13]. In recent work [18], we demonstrated
that 70% of the overall increase in degree of mineralization
during life occurs within the first 4 years. For this reason, it
is not surprising that we did not find a significant age
dependence from 25 to 97 years in this work.
Interestingly, gender is also not a factor influencing bone
at the material level, but clearly is so at the tissue level.
Fig. 5. BMDD for 52 normal, adult humans and examples of deviations
caused by diseased bone: (a) The solid line indicates the average BMDD
for 52 individuals. Dashed lines show the 1 SD boundaries. CaLow is a
newly defined BMDD parameter and describes the percentage of bone area,
which is mineralized less than 17.68 wt% Ca (vertical dashed line). At this
cutoff Ca concentration the mean CaLow for the reference population is
⬃5% (see Table 6). (b) Solid line: reference BMDD; gray circles: BMDD
of a patient with osteomalacia; white triangles: BMDD of a patient with
osteoporosis treated with alendronate [28].
322
P. Roschger et al. / Bone 32 (2003) 316 –323
Table 6
Reference BMDD values and examples of deviations by disease (osteomalacia) or treatment (alendronate (ALN))
BMDD parameter
Referencea ⫾ SD
Osteomalaciab
ALN-treated osteoporosisb
CaMean [wt% Ca]
CaPeak [wt% Ca
CaWidth [wt% Ca]
CaLow [%]
22.20 ⫾ 0.45
22.94 ⫾ 0.39
3.35 ⫾ 0.34
4.93 ⫾ 1.57
17.79
19.72
6.34
35.89
22.12
22.53
2.95
4.29
a
b
(⫺9.8 SD)
(⫺8.3 SD)
(8.8 SD)
(19.7 SD)
(⫺0.15 SD)
(⫺1.1 SD)
(⫺1.36 SD)
(⫺0.45 SD)
Mean values with SD; n ⫽ 52.
Deviations from reference are given in units of SD of reference in parentheses.
However, it has been shown that males lose less trabecular
bone than females [2] at the tissue level, and cortical bone
area in males tends to remain constant, while in females a
reduction occurs [35]. In contrast, there appears to be no
gender-specific difference in the biological control of bone
matrix mineralization pattern.
The preceding results demonstrate the striking similarity
of the BMDDs of normal adults irrespective of ethnic
groups, skeletal site, age, and gender. Hence we have calculated reference values for the BMDD of healthy trabecular bone. These reference values can be used for comparison with values from abnormal bone samples, and thereby
the amount of deviations from normality can be given in
units of SD values of the reference BMDD parameters. The
new parameter, CaLow, was introduced to provide an additional tool to detect subtle alterations in the low range of
mineralization. The variance of CaMean and CaPeak based on
a set of 52 bone samples was found to be remarkably small
(SD ⬍ 2.1%). This small variability of BMDD makes it
especially suited to detect even small changes induced by
alterations in bone metabolism as illustrated by a patient
receiving an alendronate treatment. The distinct reduction in
CaWidth (⫺1.4 SD) is typical for an antiresorptive drug
[27,28] and reflects the highly significant (P ⬍ 0.001) effect
on the group of alendronate-treated osteoporotic patients
measured in the previous study [28]. Additionally, the lack
of any hypermineralization (CaMean ⫽ ⫺0.15 SD) of this
individual patient is also consistent with the behavior of the
entire group of patients [28]. Unfortunately, we had no bone
samples available to demonstrate the effect of alendronate
directly on the BMDDs before and after treatment on the
same individual. However, it can be expected, that in a
paired study, even smaller changes can be detected, because
no variations in base line characteristics can obscure a
potential effect of the treatment.
The crucial factors that are likely to be responsible for
the observed constancy of BMDD in trabecular bone are the
time course of the mineralization process of the organic
matrix up to a certain level of saturation and the regulation
of bone turnover within a relatively narrow range. The
mineralization of the organic matrix takes place in two
phases, with a rapid primary phase and a much slower,
secondary phase. Because the secondary phase takes approximately 6 months, the degree of mineralization also
reflects the age of the newly formed bone. Thus, the BMDD
can be considered also as an age distribution of the bone
matrix. It is remarkable that this dynamically generated
mineralization density pattern in trabecular bone was found
to have such low variation and therefore seems to be a
nearly constant biological parameter in normal adult humans.
Acknowledgments
We thank G. Dinst and P. Messmer for excellent technical assistance, and Dr. B.M. Grabner for very helpful
discussions of the manuscript. Supported in part by National
Institutes of Health Grants AR-31991 and DK-32333.
References
[1] Currey JD. The Mechanical Adaptations of Bones. Princeton, NJ:
Princeton University Press, 1984.
[2] Parfitt AM, Mathews CHE, Villanueva AR, Kleerekoper M, Frame B,
Rao DS. Relationships between surface, volume, and thickness of
iliac trabecular bone in aging and in osteoporosis. J Clin Invest
1983;72:1396 – 409.
[3] Wehrli FW, Gomberg BR, Saha PK, Song HK, Hwang SN, Snyder
PJ. Digital topological analysis of in vivo magnetic resonance microimages of trabecular bone reveals structural implications of osteoporosis. J Bone Miner Res 2001;16:1520 –31.
[4] Roschger P, Fratz P, Eschberger J, Klaushofer K. Validation of
quantitative backscattered electron imaging for the measurement of
mineral density distribution in human bone biopsies. Bone 1998;23:
319 –26.
[5] Camacho NP, Hou L, Toledano TR, Ilg WA, Brayton CF, Raggio CL,
Root R, Boskey AL. The material basis for reduced mechanical
properties in oim mice bones. J Bone Miner Res 1999;14:264 –72.
[6] Fratzl P, Paris O, Klaushofer K, Landis WJ. Bone mineralization in
an osteogenesis imperfecta mouse model studied by small-angle Xray scattering. J Clin Invest 1996;97:396 – 402.
[7] Traub W, Arad T, Vetter U, Weiner S. Ultrastructural studies of
bones from patients with osteogenesis imperfecta. Matrix Biol 1994;
14:337– 45.
[8] Borah B, Gross GJ, Dufresne TE, Smith TS, Cockman MD,
Chmielewski PA, Lundy MW, Hartke JR, Sod EW. Three-dimensional microimaging (MR␮I and ␮CT), finite element modeling, and
rapid prototyping provide unique insights into bone architecture in
osteoporosis. Anat Rec (new anat) 2001;265:101–10.
[9] Boivin GY, Chavassieux PM, Santora AC, Yates J, Meunier PJ.
Alendronate increases bone strength by increasing the mean degree of
mineralization of bone tissue in osteoporotic women. Bone 2000;27:
687–94.
P. Roschger et al. / Bone 32 (2003) 316 –323
[10] Boivin G, Meunier PJ. The degree of mineralization of bone tissue
measured by computerized quantitative contact microradiography.
Calcified Tiss Int 2002;70:503–11.
[11] Eschberger J, Eschberger DJ. Microradiography. In: von Recum FA,
editor. Handbook of biomaterials evaluation. New York: Macmillan
Publishing Company; 1986, p. 491–500.
[12] Jowsey J. Variations in bone mineralization with age and disease. In:
Frost HM, editor. Bone biodynamics. Boston MA: Little, Brown &
Company; 1964, p. 461–79.
[13] Boyde A, Elliott JC, Jones SJ. Stereology and histogram analysis of
backscattered electron images: age changes in bone. Bone 1993;14:
205–10.
[14] Reid SA, Boyde A. Changes in the mineral density distribution in
human bone with age: image analysis using backscattered electrons in
the SEM. J Bone Miner Res 1987;2:13–22.
[15] Skedros JG, Bloebaum RD, Bachus KN, Boyce TM, Constantz B.
Influence of mineral content and composition on graylevels in backscattered electron images of bone. J Biomed Mater Res 1993;27:57–
64.
[16] Nuzzo S, Lafage-Proust MH, Martin-Badosa E, Boivin G, Thomas T,
Alexandre C, Peysrin F. Synchrotron radiation microtomography
allows the analysis of three-dimensional microarchitecture and degree
of mineralization of human iliac crest biopsy specimens: effects of
edidronate treatment. J Bone Miner Res 2002;17:1372– 82.
[17] Rinnerthaler S, Roschger P, Jakob HF, Nader A, Klaushofer K, Fratz
P. Scanning small x-ray scattering analysis of human bone sections.
Calcified Tissue Int 1999;64:422–9.
[18] Roschger P, Grabner BM, Rinnerthaler S, Tesch W, Kneissel M,
Berzlanovich A, Klaushofer K, Fratzl P. Structural development of
the mineralized tissue in the human L4 vertebral body. J Struct Biol
2001;136:126 –3.
[19] Camacho NP, Rinnerthaler S, Paschalis EP, Mendelsohn R, Boskey
AL, Fratzl P. Complementary information on bone ultrastructure
from scanning small angle x-ray scattering and Fourier-transform
infrared microspectroscopy. Bone 1999;25:287–93.
[20] Ou-Yang H, Paschalis EP, Mayo WE, Boskey AL, Mendelsohn R.
Infrared microscopic imaging of bone: spatial distribution of CO32⫺.
J Bone Miner Res 2001;16:893–900.
[21] Paschalis EP, Verdelis K, Doty SB, Boskey AL, Mendelsohn R,
Yamauchi M. Spectroscopic characterization of collagen cross-links
in bone. J Bone Miner Res 2001;16:1821– 8.
[22] Freeman JJ, Wopenka B, Silva MJ, Pasteris JD. Raman spectroscopic
detection of changes in bioapatite in mouse femora as a function of
aged and in vitro fluoride treatment. Calcified Tiss Int 2001;68:156 –
62.
[23] Roschger P, Plenk H Jr, Klaushofer K, Eschberger J. A new scanning
electron microscopy approach to the quantification of bone mineral
distribution: backscattered electron image grey-levels correlated to
calcium K␣⫺ line intensities. Scanning Microsc 1995;9:75– 88.
[24] Evans GP, Behiri JC, Currey JD, Bonfield W. Microhardness and
Young’s modulus in cortical bone exhibiting an wide range of mineral volume fraction, and in a bone analogue. J Mater Sci: Mater Med
1990;1:38 – 43.
[25] Kneissel M, Boyde A, Gasser JA. Bone tissue and its mineralization
in aged estrogen-depleted rats after long-term intermittent treatment
with parathyroid hormone (PTH) analog SDZ PTS 893 or human
PTH(1–34). Bone 2001;28:237–50.
[26] Roschger P, Grabner BM, Messmer P, Dempster DW, Cosman F,
Nieves J, Lindsay R, Bilezikian JP, Kurland ES, Shane E, Fratzl P,
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
323
Klaushofer K. Influence of intermittent PTH treatment on mineral
distribution in the human ilium: a paired biopsy study before and after
treatment. J Bone Miner Res 2001;16(Suppl1):S197.
Roschger P, Fratzl P, Klaushofer K, Rodan G. Mineralization of
cancellous bone after alendronate and sodium fluoride treatment: a
quantitative backscattered electron imaging study on minipig ribs.
Bone 1997;20:393–7.
Roschger P, Rinnerthaler S, Yates J, Rodan GA, Fratzl P, Klaushofer
K. Alendronate increases degree and uniformity of mineralization in
cancellous bone and decreases the porosity in cortical bone of osteoporotic women. Bone 2001;29:185–91.
Parisien M, Cosman F, Morgan D, Schnitzer M, Liang X, Nieves J,
Forese L, Luckey M, Meier D, Shen V, Lindsay R, Dempster DW.
Histomorphometric assessment of bone mass, structure, and remodeling: a comparison between healthy black and white premenopausal
women. J Bone Miner Res 1997;12:948 –57.
Ihaka R, Gentleman RR. A language for data analysis and graphics.
J Comput Graph Stat 1996;5:299 –314.
Han ZH, Palnitkar S, Rao DS, Nelson D, Parfitt AM. Effect of
ethnicity and age or menopause on the structure and geometry of iliac
bone. J Bone Miner Res 1996;11:1967–75.
Han ZH, Palnitkar S, Rao DS, Nelson D, Parfitt AM. Effects of
ethnicity and age or menopause on the remodeling and turnover of
iliac bone: implications for mechanisms of bone loss. J Bone Miner
Res 1997;12:498 –508.
Luckey MM, Meier DE, Mandeli JP, DaCosta MC, Hubbard ML,
Goldsmith SJ. Radial and vertebral bone density in white and black
women: evidence for racial differences in premenopausal bone homeostasis. J Clin Endocrinol Metab 1989;69:762–70.
Meier DE, Luckey MM, Wallenstein S, Clemens TL, Orwoll ES,
Waslien CI. Calcium, vitamin D, and parathyroid hormone status in
young white and black women: association with racial differences in
bone mass. J Clin Endocrinol Metab 1991;72:703–10.
Ruff CB, Hayes WC. Sex differences in age-related remodeling of the
femur and tibia. J Orthop Res 1988;6:886 –96.
Rauch F, Schoenau E. Changes in bone density during childhood and
adolescence: an approach based on bone’s biological organization.
J Bone Miner Res 2001;16:597– 604.
Boyde A, Jones SJ, Aerssens J, Dequeker J. Mineral density quantitation of the human cortical iliac crest by backscattered electron
image analysis: variations with age, sex and degree of osteoarthritis.
Bone 1995;16:619 –27.
Portigliatti Barbos M, Bianco P, Ascenzi A. Distribution of osteonic
and interstitial components in the human femoral shaft with reference
to structure, calcification and mechanical properties. Acta Anat 1983;
115:178 – 86.
Gyepes M, Mellins HZ, Katz IK. The low incidence of fracture of the
hip in the negro. J Am Med Assoc 1962;181:1073– 4.
Grabner B, Landis WJ, Roschger P, Rinnerthaler S, Peterlik H,
Klaushofer K, Fratzl P. Age- and genotype-dependence of bone
material properties in the osteogenesis imperfecta murine model
(oim). Bone 2001;29:453–7.
Laib A, Hildebrand T, Ha¨ uselmann J, Ru¨ egsegger P. Ridge number
density: a new parameter for in vivo bone structure analysis. Bone
1997;21:541– 6.
Moldawer M, Zimmermann SJ, Collins LC. Incidence of osteoporosis
in elderly whites and elderly negroes. J Am Med Assoc 1965;194:
859 – 62.

Download Report