Bone 39 (2006) 1059 – 1066

www.elsevier.com/locate/bone

Low-level mechanical vibrations can influence bone resorption

and bone formation in the growing skeleton
Liqin Xie a , Jeffrey M. Jacobson a , Edna S. Choi a , Bhavin Busa a , Leah Rae Donahue b ,
Lisa M. Miller c , Clinton T. Rubin a , Stefan Judex a,⁎
a
Department of Biomedical Engineering, Psychology A, 3rd Floor, State University of New York at Stony Brook, Stony Brook, NY 11794-2580, USA
b
The Jackson Laboratory, Bar Harbor, ME 04609, USA
c
National Synchrotron Light Source, Brookhaven National Laboratory, Upton, NY 11973-5000, USA
Received 10 February 2006; revised 9 May 2006; accepted 15 May 2006
Available online 7 July 2006

Abstract

Short durations of extremely small magnitude, high-frequency, mechanical stimuli can promote anabolic activity in the adult skeleton. Here, it
is determined if such signals can influence trabecular and cortical formative and resorptive activity in the growing skeleton, if the newly formed
bone is of high quality, and if the insertion of rest periods during the loading phase would enhance the efficacy of the mechanical regimen. Eight-
week-old female BALB/cByJ mice were divided into four groups, baseline control (n = 8), age-matched control (n = 10), whole-body vibration
(WBV) at 45 Hz (0.3 g) for 15 min day− 1 (n = 10), and WBV that were interrupted every second by 10 of rest (WBV-R, n = 10). In vivo strain
gaging of two additional mice indicated that the mechanical signal induced strain oscillations of approximately 10 microstrain on the periosteal
surface of the proximal tibia. After 3 weeks of WBV, applied for 15 min each day, osteoclastic activity in the trabecular metaphysis and epiphysis
of the tibia was 33% and 31% lower (P < 0.05) than in age-matched controls. Bone formation rates (BFR·BS− 1) on the endocortical surface of the
metaphysis were 30% greater (P < 0.05) in WBV than in age-matched control mice but trabecular and middiaphyseal BFR were not significantly
altered. The insertion of rest periods (WBV-R) failed to potentiate the cellular effects. Three weeks of either WBV or WBV-R did not negatively
influence body mass, bone length, or chemical bone matrix properties of the tibia. These data indicate that in the growing skeleton, short daily
periods of extremely small, high-frequency mechanical signals can inhibit trabecular bone resorption, site specifically attenuate the declining
levels of bone formation, and maintain a high level of matrix quality. If WBV prove to be efficacious in the growing human skeleton, they may be
able to provide the basis for a non-pharmacological and safe means to increase peak bone mass and, ultimately, reduce the incidence of
osteoporosis or stress fractures later in life.
© 2006 Elsevier Inc. All rights reserved.

Keywords: High-frequency mechanical stimuli; Peak bone mass; Bone quality; Mechanical strain

Introduction
Increasing peak bone mass during young adulthood is central
to optimizing skeletal health [1]. Generally, the amount of bone
present is inversely related to the risk of bone pathologies,
including osteoporotic and stress fractures [2,3]. Variability in
peak bone mass is modulated by genetics, life-style, and func-
tional load bearing [4]. The inherent scientific and ethical chal-
lenges of altering genetics, together with the potential of
pharmaceutical interventions for acute and chronic complications
⁎ Corresponding author. Fax: +1 631 632 8577.
E-mail address: stefan.judex@sunysb.edu (S. Judex).
8756-3282/$ - see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.bone.2006.05.012
[5], emphasize that lifestyle strategies such as diet and exercise
may present an alternative to promote bone quantity and quality.
In particular, strengthening the skeleton through exercise during
adolescence and early adulthood [6,7] may be a promising means
of reducing the incidence of skeletal fractures later in life.
Exercise can increase bone formation [8–10], decrease bone
resorption [9,10], raise peak bone mass [11], and enhance bone
strength [10]. Despite its non-pharmacological nature, skeletal
loading must also be approached with caution. The potential
attenuation of longitudinal bone growth in gymnasts [12], or the
high incidence of stress fractures in military recruits [13], ballet
dancers [14], and marathon runners [15] demonstrates that spe-
cific aspects of a mechanical loading regime may be harmful and
1060 L. Xie et al. / Bone 39 (2006) 1059–1066

may contribute to the risk of skeletal fragility. Despite the im-
portance of appropriate exercise for the optimal development of
the growing skeleton, the diminishing time committed to physical
activity programs in children and adolescents [16] has reduced the
likelihood that bone mass can be augmented purely by exercise in
a substantial subpopulation of children and adolescents.
To use bone's sensitivity to mechanical signals as a means of
enhancing bone quantity and/or quality during skeletal growth, it
will be necessary to employ strategies that are safe, effective, short
in duration, and can achieve high compliance. In the adult skel-
eton, recent studies have indicated that bone is sensitive to very
low-level mechanical signals, induced non-invasively through
whole-body vibrations (WBV). These high-frequency (>20 Hz)
mechanical signals produce vertical whole-body oscillations of
less than 50 μm and generate strains (deformation) in cortical
bone two orders of magnitude below those associated with
physical activity. With as little as 10 min day− 1, these low-level
mechanical signals can promote bone formation [17], enhance
bone morphology [18], increase bone strength [19], and attenuate
the negative effects associated with catabolic stimuli [20].
The safety profile of these low-magnitude accelerations is
favorable. While safety concerns for the musculoskeletal system
arise when WBV exceed 1 g in magnitude [21], the International
Safety Organization describes no evidence of any acute or
chronic complications of 20 to 90 Hz vibration when exposure
falls below 0.56 g [22]. In the human skeleton, short-duration
low-amplitude WBV may be anabolic and/or anti-catabolic as
the prevention of bone loss in postmenopausal women [18] or
the increase in bone density in children with disabling conditions
[20] suggests. The mechanisms by which these positive effects
were achieved have not been elucidated. Using a murine model,
here, the following novel research questions were addressed: (1)
is WBV capable of increasing cortical and/or trabecular bone
formation in the growing skeleton that is subjected to high levels
of physical activity? (2) Is WBV capable of reducing cortical
and/or trabecular bone resorption? (3) Can the cellular effects of
WBV be potentiated with the inclusion of multiple “rest
intervals”? (4) Is bone formed during the loading regimen of
high chemical quality? (5) Can WBV have detrimental effects on
endochondral ossification during growth?
in previous studies in which their insertion caused a beneficial effect, albeit at
much larger load magnitudes [23–25].
At 8 weeks of age, mice are reproductively mature but have not reached peak
BMD. To enable measurement of dynamic indices of bone formation, mice were
injected (i.p.) with calcein (15 mg kg− 1) on days 15 and 20 of the experimental
protocol. After the 3-week experimental duration, the right tibia was harvested
and submerged in 70% ethanol for microcomputed tomography (μCT), histo-
morphometry, and chemical composition analyses (Fig. 1). The left tibia was
fixed overnight in 10% neutral-buffered formalin for staining of osteoclastic
resorption via tartrate-resistant acid phosphatase (TRAP, Fig. 1). The length of
the right tibia was measured with digital calipers.
Bone strain measurement
Cortical surface bone strains generated in the proximal tibia during 0.3 g,
45 Hz WBV were measured in two additional BALB mice. Because of the small
size of the mouse tibia at 8 weeks of age, adult animals were used. Under
isoflurane anesthesia, a miniature single-element strain gage (1 mm gage length,
120 Ω, TML Gages, Kenkyujo, Japan) was implanted on the antero-medial
surface of the proximal tibia (cyanoacrylate). Upon recovery from surgery (1–2 h),
and with the animal standing on the vibrating plate, strain data were collected over
two 10-s trials. Strain gage signals were amplified (SX500, Syminex Inc, Mt.
Arlington, NJ) with an excitation of 4 V and a 1000× gain and acquired at a
sampling rate of 1000 Hz. This setup collected in vivo strain data at a resolution of
approximately 0.5 microstrain. Strain data were plotted (Origin 7.5, OriginLab,
Northampton, MA), and the difference between consecutive bottom–top peaks,
averaged across the 10-s period, defined the peak-to-peak magnitude of the
oscillatory signal.
Microcomputed tomography
To evaluate potential changes in trabecular and cortical bone morphology
over the 3-week experimental period, the right proximal tibia of all mice was
scanned (μCT 40, Scanco Medical, SUI) at a resolution of 6 μm. The epiphysis
(180 μm in length) and metaphysis (600 μm in length) were defined according to
precise landmarks that attempted to maximize trabecular volume within each
region (Fig. 1). Trabecular bone was separated from cortical bone with manually
drawn contour lines. Cortical bone was analyzed from the metaphysis (sur-
rounding the trabecular volume of interest) and from three diaphyseal regions:
the proximal diaphysis (defined at 40% of total length), the middiaphysis (at
50%), and the distal diaphysis (at 60%). The values for sigma, support and
threshold (parameters required for the evaluation), were set at 0.5, 1, 240 for the

Methods

Experimental design

Experimental procedures were approved by Stony Brook's Institutional

Animal Care and Use Committee. Thirty-eight 8-week-old female BALB/cByJ
(BALB) mice were obtained from The Jackson Laboratory (Bar Harbor, ME)
and randomly divided into four groups: (1) baseline control (BC, n = 8), (2) age-
matched control (AC, n = 10), (3) mice subjected to WBV at 45 Hz, 0.3 g for
15 min day− 1, 5 days week− 1 (WBV, n = 10), and (4) mice subjected to the same
frequency and peak acceleration of the mechanical signal (45 Hz, 0.3 g) but
with the insertion of 10-s rest periods [23] after each second of vibration (WBV-
R, n = 10). Thus, WBV-R mice were loaded for a total of 165 min during which
45 cycles of loading in 1 s were repetitively followed by 10 s of rest. Aside from
the “rest intervals”, WBV and WBV-R groups shared the same mechanical
parameters including peak accelerations and acceleration patterns, force
magnitude, strain frequency, and number of loading cycles (40,500) per day. Fig. 1. Synopsis of the site-specific measurements performed in the epiphysis,
The length and frequency of the resting periods were similar to those described metaphysis, and diaphysis.
 L. Xie et al. / Bone 39 (2006) 1059–1066 1061

cortical and trabecular metaphysis and diaphysis and 1, 2, 240 for the epiphysis
because of greater noise in this region.
For trabecular regions, bone volume fraction (BV·TV− 1), trabecular sepa-
ration (Tb.Sp), trabecular thickness (Tb.Th), trabecular number (Tb.N),
connectivity density (Conn.D), geometrical degree of anisotropy (DA), and
the structural model index (SMI) were determined. For cortical bone, bone area
(Ct.Ar) and areas of the endocortical (Ec.En or bone marrow area) and periosteal
(Ps.En or periosteal area) envelopes were calculated.
Histomorphometry
Following tomographic scanning, the right proximal and diaphyseal tibia were
embedded in methyl methacrylate resin (MMA) using a standard protocol. The
proximal specimens were sectioned longitudinally in the center with a microtome to
yield 5-μm frontal sections (RM 2165 microtome, Leica, Bensheim, Germany) while
diaphyseal samples were sectioned (40 μm) with a diamond wire saw (Well Diamond
Wire Saws, Norcross, GA). The evaluated regions in the epiphysis (trabecular bone)
and metaphysis (trabecular and cortical bone) were similar to the regions scanned by
μCT and spanned 200 μm in length in the epiphysis and 800 μm in length in the
secondary spongiosa of the metaphysis, starting 400 μm distal from the border of the
physis (Fig. 1). Because of a lack of consistent double labels at the periosteal surface
of the metaphysis and at the endocortical surface of the middiaphysis, indices of
cortical bone formation were only quantified at the endocortical surface of the
metaphysis and the periosteal surface of the middiaphysis. Histomorphometry
software (Osteomeasure, OsteoMetrics Inc., Atlanta, GA) was used to trace the
fluorescent labels and bone surfaces. Mineral apposition rate (MAR, μm day− 1) was
calculated as the distance between double labels divided by the labeling interval.
Mineralizing surface with bone surface as referent (MS·BS− 1, %) was obtained by
adding the ratio of double-labeled surface to bone surface (dLS·BS− 1) to 50% of the
ratio of single-labeled surface to bone surface (sLS·BS− 1, %). A standard measure of
bone formation rates (BFR·BS− 1, μm year− 1) was calculated by multiplying MS·BS-
1 by MAR, with 1 year as the referent.
The left proximal tibia was fixed fresh in 10% neutral-buffered formalin
overnight and decalcified in 2.5% formic acid (pH 4.2) for 4 days. Upon
dehydration, samples were embedded in glycol methacrylate (GMA) according
to the manual of the JB-4 embedding kit (Polysciences, Warrington, PA). Frontal
sections (7 μm) were produced with a microtome and stained for TRAP activity.
Hexazotization was achieved by mixing equal amounts of 4% NaNO2 and 4%
pararosaniline solution. Naphthol-ASTR-phosphate (Sigma, St. Louis, MO) was
used as a substrate, and the enzyme reaction was carried out in the presence of
tartrate (10 mM) to demonstrate TRAP activity (pH 5 in 0.1 M acetate buffer).
Sections were counterstained with methyl green to improve contrast. In bone,
macrophages, osteoclasts, and chondroclasts can express TRAP [29] but our
analysis was selective for osteoclasts because the TRAP signal intensity of
macrophages is much lower [29,30] and the anatomical regions under study
were void of chondroclasts [29]. The ratio of osteoclast surface (Oc.S) to bone
surface (BS) was determined for trabecular and cortical bone surfaces in the
metaphysis and epiphysis using commercially available histology software
(Osteomeasure). The epiphyseal and metaphyseal regions that were analyzed
matched those of the histomorphometric analyses described above.
To investigate whether the application of high-frequency vibrations affected
histological measures of bone growth, the growth plate was divided into three
different zones: reserve zone, proliferative zone, and hypertrophic zone. The
metaphyseal and epiphyseal borders of the growth plates were defined by the
extent of methyl green staining of the cartilage. The hypertrophic-proliferative

Chemical composition of newly formed bone

To establish the quality of the newly formed bone in the vibrated- and age-
matched control groups, high-resolution in situ analysis of collagen and mineral
content and composition was performed on metaphyseal cortical and trabecular
bone by synchrotron infrared microspectroscopy (SIRMS) [26]. Synchrotron light
(Beamline U10B, The National Synchrotron Light Source, Brookhaven National
Laboratory, Upton, NY) and a Nicolet Magna 860 spectrometer (ThermoNicolet,
Madison, WI) were coupled to a Continuum IR microscope (ThermoNicolet,
Madison, WI) and MCT-B (Mercury Cadmium Telluride) detector. The epi-
fluorescent microscope was modified such that it allowed the visualization of the
calcein labels during SIRMS data collection [26]. Spectra were collected from the
5-μm frontal metaphyseal sections in transmission mode, over the frequency range
of 4000–400 cm− 1, 256 scans per point, at 4 cm− 1 spectral resolution, with the
apertures size set at 12× 12 μm.
Chemical measurements were taken from two randomly selected cortical and
trabecular regions that were enclosed by double calcein labels, thus limiting the
analysis to tissue that had formed during the last 7 days of the experimental period.
With five spatially adjacent 12×12 μm spectra acquired from each region, a total of
ten trabecular and ten cortical chemical spectra were averaged for each sample.
Infrared spectra were analyzed for phosphate-to-protein (integrated area of
phosphate peak at 500–650 cm− 1 to the amide I peak at 1595–1510 cm− 1),
carbonate-to-protein ratio (integrated area of carbonate peak at 905–825 cm− 1 to the
amide I peak at 1595–1510 cm− 1), and carbonate-to-phosphate ratio. A collagen
cross-linking parameter was calculated from the intensity ratio of peaks at 1660 and
1690 cm− 1. The acid phosphate-to-total-phosphate content was obtained by dividing
the peak height at 538 cm− 1 by the integrated area of the phosphate peak at 500–
650 cm− 1. Crystallinity, an indicator of crystal size perfection, was determined as the
ratio of stoichiometric (603 cm− 1) to non-stoichiometric phosphate (563 cm− 1) [27]. Fig. 2. (a) Accelerometer recording of the vertically oscillating vibrating plate,
producing peak accelerations of 0.3 g at 45 Hz. (b) Simultaneous recording from
a longitudinal strain gage (attached to antero-medial surface of the tibia) while
Assessment of osteoclastic activity and thickness of growth plate the mouse was subjected to the mechanical signal. WBV caused extremely small
strain oscillations on the order of ten microstrain (peak to peak). (c) Standing on
Tartrate-resistant acid phosphatase (TRAP), as an indicator of osteoclastic an inactive plate induced longitudinal normal strains of less than one microstrain
by-products, was stained in situ by previously verified standard methods [28]. (peak to peak).
1062 L. Xie et al. / Bone 39 (2006) 1059–1066

Table 1
Mean (±SD) body mass and indicators of longitudinal bone growth in the tibia of control and vibrated groups
Index BC (n = 8) AC (n = 10) WBV (n = 10) WBV-R (n = 10)
Initial body mass (g) 19.1 ± 2.3 20.5 ± 1.6 19.8 ± 2.3 19.6 ± 2.4
Final body mass (g) n/a 22.0 ± 1.2* 21.4 ± 1.6* 20.7 ± 1.5*
Length of tibia (mm) 15.7 ± 0.6 16.8 ± 0.6** 16.6 ± 0.5 16.5 ± 0.6
Thickness of growth plate (μm) 166.8 ± 23.9 128.8 ± 6.2*** 136.4 ± 11.4 137.8 ± 11.2
Proliferative zone (μm) 63.0 ± 9.7 44.9 ± 5.0*** 50.4 ± 8.7 49.2 ± 10.3
Hypertrophic zone (μm) 68.9 ± 14.6 52.0 ± 7.6*** 56.2 ± 8.0 55.1 ± 9.0
Reserve zone (μm) 34.6 ± 8.9 32.9 ± 1.9 30.2 ± 5.2 30.0 ± 3.8
* Significant difference between initial and final body mass (P < 0.05).
** Significant difference between AC and BC group (P < 0.05).
*** Significant difference between AC and BC group (P < 0.01).

boundary was defined as the point at which the flattened cell lacunae of the
proliferative zone exhibited a proximal–distal height that was approximately
equal to or greater than the transverse dimension [31]. The boundary of each
zone was traced and the thickness was analyzed with software (Osteomeasure).
Statistical analysis
All data were expressed as mean ± SD. The two vibrated groups (WBV and
WBV-R) were contrasted with the age-matched control group via a least significant
difference test. This test was only used when a one-way ANOVA indicated
significant differences between the three groups (and thus protecting against an
increased likelihood of incurring type I errors). To enhance the interpretation of
potential differences between the three age-matched groups, age-related differences
in bone quantity, quality, and metabolic activity of the two control groups (AC and
BC) were compared to each other by two-tailed t tests (as a secondary assessment).
Statistical significance was set at 5% (SPSS 13.0, Chicago, IL).
Results
Strain magnitudes induced by WBV in the metaphyseal cortex
The accelerometer attached to the vibration plate confirmed
the sinusoidal nature of the vertically oscillating vibration plate
(Fig. 2a). Concurrent in vivo recordings from a strain gage
attached to the cortical metaphysis demonstrated transmissibility of
the mechanical signal into the tibia as indicated by the sinusoidal
strain pattern at the same frequency (Fig. 2b). Standing on an
inactive plate induced strains on the order of 1 με (Fig. 2c). The
vibratory oscillations applied at a frequency of 45 Hz and peak
accelerations of 0.3 g induced peak bone strain oscillations at the
antero-medial surface of the tibia on the order of 10 με. In contrast,
vibration of the strain gage itself (while attached to a stiff PMMA
block) produced strain magnitudes that were less than the noise
level of the system. Recorded strain signals were consistent across
trials and mice. Average (±SD) peak-to-peak strain magnitudes
were 10.5 ± 2.2 με and 11.6 ± 1.9 με for the two trials in one
animal, and 11.4 ± 1.4 με and 10.2 ± 1.4 με in the second animal
(Fig. 2b).
Effect of WBV on body mass and indices of skeletal growth
There were no differences in mean body mass between the
four groups at the beginning of the study, and each of the three
experimental groups (AC, WBV, WBV-R) gained similar

Table 2
Indices of bone formation and bone morphology in the proximal and middiaphyseal tibia (mean ± SD)
Index BC (n = 8) AC (n = 10) WBV (n = 10) WBV-R (n = 10)
−1
Trabecular metaphysis MAR (μm day ) 2.44 ± 0.26 2.11 ± 0.29* 2.10 ± .033 2.02 ± 0.39
BFR·BS− 1 (μm year− 1) 236 ± 39 177 ± 45* 160 ± 39 166 ± 55
BV·TV− 1 (%) 18.2 ± 3.0 18.9 ± 3.3 18.5 ± 2.3 17.8 ± 2.9
Tb.Th (μm) 35.2 ± 2.6 39.7 ± 2.6*** 40.1 ± 2.7 39.9 ± 3.1
Trabecular epiphysis MAR (μm day− 1) 1.61 ± 0.35 0.95 ± 0.37*** 1.15 ± 0.49 1.02 ± 0.27
BFR·BS− 1 (μm year− 1) 208 ± 111 61 ± 37*** 69 ± 35 50 ± 20
BV·TV− 1 (%) 27.6 ± 4.7 31.9 ± 3.7* 33.1 ± 4.4 32.4 ± 4.2
Tb.Th (μm) 45.9 ± 5.2 52.0 ± 3.9* 51.4 ± 3.1 52.1 ± 3.0
Cortical metaphysis MAR (μm day− 1) 8.21 ± 1.87 4.13 ± 1.0*** 5.31 ± 1.09 5.27 ± 1.38
BFR. BS− 1 (μm year− 1) 2695 ± 616 1358 ± 328*** 1764 ± 332** 1730 ± 453
Ps.En (mm2) 2.68 ± 0.45 2.90 ± 0.34 2.97 ± 0.25 2.91 ± 0.20
Ct.Ar (mm2) 1.14 ± 0.13 1.23 ± 0.17 1.24 ± 0.10 1.21 ± 0.08
Ec.En (mm2) 1.54 ± 0.33 1.67 ± 0.31 1.73 ± 0.21 1.70 ± 0.21
Cortical middiaphysis MAR (μm day− 1) 2.45 ± 0.42 2.01 ± 0.89 1.89 ± 0.76 1.96 ± 0.90
BFR·BS− 1 (μm year− 1) 399 ± 122 208 ± 92*** 191 ± 114 185 ± 76
Ps.En (mm2)a 0.79 ± 0.10 0.84 ± 0.08 0.82 ± 0.07 0.83 ± 0.06
Ct.Ar (mm2)a 0.51 ± 0.06 0.57 ± 0.05* 0.55 ± 0.05 0.55 ± 0.05
Ec.En (mm2)a 0.28 ± 0.0.05 0.27 ± 0.04 0.27 ± 0.03 0.27 ± 0.04
a
Only middiaphyseal data are shown as results at the other two diaphyseal locations were qualitatively similar.
* Significant difference between AC and BC group (P < 0.05).
** Significant difference between WBV and AC group (P < 0.05).
*** Significant difference between AC and BC group (P < 0.01).
 L. Xie et al. / Bone 39 (2006) 1059–1066 1063

Bone resorption

Normal growth within the 3-week experimental period,

Fig. 3. The percentage of surfaces undergoing osteoclastic resorption (Oc.
S·BS− 1) in tibial trabecular bone of the metaphysis and epiphysis. BC = baseline
control; AC = age-matched control; WBV = whole-body vibrations for 15 min
day− 1; WBV-R = mechanical stimulation for 15 min day− 1 including rest
periods. (a) Significant difference between AC and BC. (b) Significant difference
between WBV and AC group (P < 0.05).
amounts of body mass (6–8%, P < 0.05) over the 3-week
experimental protocol (Table 1). Comparisons between baseline
(BC) and age-matched (AC) control mice indicated that the tibia
extended, on average, its length by 7% over the 3-week
experimental period (P < 0.01) but 15 min per day of WBV also
did not influence tibial growth (Table 1). The thickness of the
growth plate, proliferative zone, and hypertrophic zone was
23%, 29%, and 25%, respectively, smaller (P < 0.01) in AC
than in BC mice, but there were no significant differences
between AC, WBV, or WBV-R groups (Table 1).
Bone formation
As compared to baseline controls, indices of bone formation,
including mineral apposition (MAR) and bone formation rates
(BFR·BS− 1), declined by up to 70% in the tibial metaphysis,
epiphysis, and middiaphysis of age-matched control animals over
the course of the experimental period (Table 2). Mechanical
vibrations attenuated this decline only in the cortical metaphysis
where endocortical bone formation rates (BFR·BS− 1) were 30%
greater (P < 0.05) in WBV mice when compared to age-matched
controls (Table 2). Rest inserted loading (WBV-R) failed to sig-
nificantly influence parameters of bone formation in any anatom-
ical region as indicated by the lack of differences in BFR between
these mice and age-matched controls. There were no significant
differences in indices of bone formation between WBV and
WBV-R mice.
quantified as the difference between AC and BC, decreased the
prevalence of osteoclastic activity (Oc.S·BS− 1) in trabecular
bone by 25% (P < 0.01) in the tibial epiphysis. In contrast, the
levels of bone resorption were not different in the metaphysis
between 8 weeks and 11 weeks of age (Fig. 3). In addition to the
age-related 25% drop in Oc.S·BS− 1 in the epiphysis, 15 min of
daily WBV further decreased the levels of osteoclastic activity
as indicated by the 31% (P < 0.01) difference in Oc.S·BS− 1
between AC and WBV mice (Fig. 3). Even though age itself had
no effect on Oc.S·BS− 1 in the metaphysis, WBV was associated
with a decline in resorptive activity in this region. Oc.S·BS− 1
was 33% (P < 0.01) smaller in WBV than in AC mice (Fig. 3).
Modification of this vibratory stimulus to include intermittent
rest periods (WBV-R) resulted in osteoclastic activity levels that
were statistically not different from those of AC or WBV mice.
In cortical bone, no differences in Oc.S·BS− 1 were observed
between any of the groups.
Bone morphology and chemical matrix properties
Normal growth over the 3-week period led to a 16% greater
(P < 0.01) trabecular bone volume fraction (BV·TV− 1) in the
epiphysis but not the metaphysis of the tibia. Trabecular
thickness (Tb.Th) was 13% greater in AC than BC mice in both
the epiphysis and metaphysis (P < 0.05 each; Table 2). While
normal growth did not influence cortical bone morphology in
the metaphysis, the cortical area of the middiaphysis was 12%
greater (P < 0.05) in AC than in BC mice. Between the three
experimental groups subjected to either 3 weeks of normal cage
activity or 3 weeks of cage activity superimposed by a vibration
based regimen, there were no significant differences in trabec-
ular and cortical bone morphology in any region of the tibia
(Table 2).
None of the six chemical matrix properties probed in newly
formed metaphyseal trabecular and cortical bone exhibited sig-
nificant differences between baseline control and age-matched
mice (Table 3). Neither 15 min of daily continuous WBV nor rest-
inserted WBV affected the inorganic properties of trabecular or
cortical bone formed during the experimental period. However,
mice subjected to WBV without rest displayed a 10% smaller
collagen cross-linking ratio (1660·1690− 1) in trabecular bone and

Table 3
Chemical properties of newly mineralized bone in the metaphyseal tibia of control and vibrated groups (mean ± SD)
Index [10− 3] BC (n = 8) AC (n = 10) WBV (n = 10) WBV-R (n = 10)
−1
Trabecular metaphysis Phosphate·protein 804 ± 201 923 ± 162 877 ± 113 823 ± 110
Carbonate·protein− 1 61 ± 13 73 ± 15 70 ± 12 67 ± 11
Phosphate·carbonate− 1 76 ± 13 82 ± 16 80 ± 14 81 ± 12
Acid phosphate·total phosphate− 1 6.0 ± 0.4 6.0 ± 0.6 5.6 ± 0.5 5.7 ± 0.5
Crystallinity 818 ± 40 863 ± 68 856 ± 98 836 ± 80
Cortical metaphysis Phosphate·protein− 1 837 ± 126 908 ± 172 933 ± 137 932 ± 142
Carbonate·protein− 1 71 ± 11 77 ± 7 80 ± 10 71 ± 9
Phosphate·carbonate− 1 83 ± 12 83 ± 10 78 ± 5 78 ± 11
Acid phosphate·total phosphate− 1 5.8 ± 0.6 5.9 ± 0.8 5.6 ± 0.4 5.8 ± 0.6
Crystallinity 869 ± 47 848 ± 51 851 ± 47 838 ± 73
1064 L. Xie et al. / Bone 39 (2006) 1059–1066
Fig. 4. Mean (±SD) collagen (coll) cross-linking ratio of newly mineralizing
trabecular and cortical bone in the proximal metaphysis of the tibia (2–7 days
old). BC = baseline control; AC = age-matched control; WBV = whole-body
vibrations for 15 min day− 1; WBV-R = mechanical stimulation for 15 min
day− 1 including rest periods. (b) Significant difference between WBV and AC
group (P < 0.05).
a 11% smaller cross-linking ratio in cortical bone (P < 0.05;
Fig. 4). The collagen cross-linking ratio of WBV-R mice was not
significantly different from AC or WBV mice.
Discussion
To test the hypothesis that low-magnitude WBV can be sensed
in a growing skeleton without concomitant negative effects,
young mice were subjected to two different regimes of WBV for
3 weeks. Superimposing a 15-min bout of WBV onto daily
activities that produced cortical surface strains on the order of ten
microstrain attenuated the age-related decline in cortical meta-
physeal bone formation and reduced the resorptive activity in the
trabecular metaphysis and epiphysis. These site-specific and
compartment-specific effects were not achieved when periods of
“rest” were inserted into the loading paradigm. Chemical pro-
perties of the newly formed matrix and indices of longitudinal
bone growth were largely similar between vibrated and control
mice. As WBV affected the resorptive activity only in trabecular
bone and formative activity only in the cortical bone surrounding
it, it is possible that the enhanced trabecular bone density ob-
served in children with disabling conditions who were subject to
low-magnitude WBV was achieved primarily by a suppression of
osteoclastic activity [20]. Considering that the mice used in this
study were physically very active suggests that the skeleton may
not need to be in a pathologic state in order to be influenced by
these mechanical signals.
Compared to age-matched controls, 15 min per day of ex-
tremely small magnitude mechanical loads was sufficient to
reduce osteoclastic activity in both the epiphyseal and the meta-
physeal regions of the tibia by 30%, even though normal growth-
related changes in resorption were very different in these two
regions. In the epiphysis, osteoclast activity decreased over the
experimental period and WBV accentuated this decline. In the
metaphysis, osteoclast activity remained unchanged over the 3-
week period in control mice but WBV caused a reduction in
activity. While this is the first report suggesting that extremely
low-level mechanical signals can inhibit osteoclastic activity, it is
consistent with the ability of physical factors to mitigate the
recruitment and differentiation of osteoclasts [32], as well as the
ability of high-impact physical exercises to reduce levels of bone
resorption [9,10]. Future studies that will determine whether the
responsivity of specific bone cell populations to WBV is site-
specific and age-dependent may also provide critical clues to-
wards the identification of the mechanism by which bone senses
these extremely low magnitude stimuli.
The focus of this 3-week study was on changes in cellular
activity and the chemical quality of the newly formed bone. Thus,
the lack of observed morphological differences in trabecular and
cortical bone between vibrated and control groups, despite the
positive influences on bone resorption and formation, should not
be surprising and was likely associated with the short duration of
the protocol. Optimization of the signal, either by adjusting its
frequency, acceleration, or duration, may provide structural ben-
efits in a shorter period of time [33]. Longer term studies, par-
ticularly in humans, will have to establish whether altered cellular
activity will translate into enhanced bone morphology and strength
and whether gains achieved during late skeletal development can
be transferred into adulthood. Interestingly, preliminary data in-
dicate that extending the protocol from 3 weeks to 6 weeks
significantly enhances trabecular bone volume fraction (10%, P <
0.05) and trabecular thickness (5%, P < 0.05) in mice of the same
initial age and genetic strain [34].
Deformations induced in the tibial cortex of these mice by
WBV were extremely small (10 με) but similar in magnitude
when compared to larger animals subjected to WBV [35]. Given
the limitation of single-strain gage recordings, it can certainly be
argued that WBV could induce higher strains at other anatomical
locations, but a doubling or tripling of these signals would still be
orders of magnitude below those strains normally associated with
physical activity. Considering that the extremely small magnitude
of these mechanical deformations is typically associated with the
catabolic state of disuse, it is unlikely that bone cells can directly
sense the vibration induced matrix deformations. Alternative
hypotheses involve byproducts of matrix deformation, adjunct
physiologic systems including muscle and blood flow, or direct
vibratory effects on bone cells. Bone matrix deformations, for
instance, can increase intramedullary pressure, and the resulting
fluid pressure gradients in the lacunar–canalicular system are
strongly dependent on the frequency of the signal [36], perhaps
producing signals that are large enough to stimulate osteocytes or
cells lining the bone surface [37]. Similarly, muscle- and somato-
sensory receptors are known to respond to oscillatory stimuli
[38,39] which may also increase capillary filtration [40,41]. In
support of the physiologic basis of this hypothesis, comparisons
between WBV and age-matched control mice used in this study
indicated that the high-frequency low-magnitude mechanical
signal significantly altered the number of blood vessels in the
soleus muscle [42].
The inclusion of a 10-s rest period after each second of WBV
rendered the low-level mechanical stimulus ineffective (with
respect to age-matched control mice), despite sharing identical
mechanical parameters such as acceleration magnitude, force
magnitude, strain frequency, or number of loading cycles per day.
These results are in contrast to recent data from mechanical loading
studies which suggest that the insertion of rest periods can sub-
stantially enhance bone's sensitivity and response to mechanical
 L. Xie et al. / Bone 39 (2006) 1059–1066
loading [23,25]. The discrepancy between these studies may lie
with the approximately 100 times lower strain magnitude used here.
Because of high cell refractory periods as well as fluid inertia and
viscosity, insertion of rest into mechanical regimes using higher
magnitude mechanical stimuli (∼ 1000 με) may increase bone fluid
flow, synchronize osteocytic activity, and enhance cell communi-
cation [23]. In contrast, the efficacy of extremely low-magnitude
WBV may be dependent on the temporal continuity of a large
number of loading cycles. A recently developed agent based model
[24] suggests that rest-inserted loading becomes effective only
when the strength of a mechanical stimulus surpasses a given
threshold upon which the insertion of rest periods between loading
bouts will allow the bone cells to “recover” from high levels of
loading. From our study, it appears that 45 consecutive cycles of a
0.3 g oscillating stimulus is not only below the threshold cycle at
which the insertion rest periods can would increase the efficiency of
mechanical loading but is also below the threshold required to
initiate a significant cellular response. Whether a different fre-
quency and duration of the rest intervals might have led to different
results remains to be determined.
A critical requirement of any effective biochemical or bio-
mechanical intervention for osteoporosis is that the bone which
is formed as a result of the treatment is of high bone quality.
While five out of six chemical properties of the newly min-
eralizing bone matrix were identical to the controls, the collagen
cross-linking ratio (1660·1690− 1) in the WBV group was lower
compared to age-matched controls, both for trabecular and
cortical bone. Mechanical and biologic consequences of smaller
cross-linking ratios are incompletely understood but have been
previously associated with less mature collagen fibers [43]. In
this study, chemical properties were probed in newly formed
bone that was 2–7 days old, and it is unclear whether the lower
ratio reflects an increase in the amount of reducible cross-links
(dihydroxylysinonorleucine, DHLNL-1690 cm − 1 ) as new
cross-links were being formed, or a decrease in the amount of
non-reducible cross-links as reducible cross-links were reduced
to Pyr-1660 cm− 1. Even though the mechanism and potential
consequences of differences in the collagen cross-linking ratio
will have to be studied further, previous data indicated that
differences in the total number of collagen cross-links have little
effect on the overall stiffness of the tissue [44].
Results from this 3-week study demonstrate that extremely low
levels of mechanical signals have the principal ability to reduce the
levels of trabecular bone resorption and to attenuate the declining
levels of cortical bone formation in the growing healthy skeleton in
a site-specific manner. The absence of any influence on bone
lengthening or bone quality may indicate the safe nature of this non-
invasive intervention and contrasts with the potential suppression
of longitudinal growth that high-intensity exercise or large loads at
high strain rates (high-impact loading) can impose on bone de-
velopment in animals [31]. These low-level WBV have previously
been shown to prevent the loss of bone mass in children and
postmenopausal women and future long-term animal and human
studies will need to establish the utility of these signals, perhaps
upon optimization, to enhance (and maintain) peak bone mass and
strength. If successful, they may ultimately serve to decrease the
incidence of osteoporotic or stress fractures later in life.
1065
Acknowledgments
Funding by the US Army Medical Research and Material
Command DAMD 17-03-1-0777 (SJ) and 17-01-1-0808
(LRD), the Whitaker Foundation RG-02-0564 (SJ), and NIH
AR 43498 (CTR) is greatly appreciated. We also would like to
thank Drs. Wei Lin, Russell Garman, and Shiyun Xu for their
expert technical advice.
References
[1] Sabatier JP, Guaydier-Souquieres G, Laroche D, Benmalek A, Fournier L,
Guillon-Metz F, et al. Bone mineral acquisition during adolescence and
early adulthood: a study in 574 healthy females 10–24 years of age.
Osteoporos Int 1996;6:141–8.
[2] Kannus P, Uusi-Rasi K, Palvanen M, Parkkari J. Non-pharmacological
means to prevent fractures among older adults. Ann Med 2005;37:
303–10.
[3] Nattiv A. Stress fractures and bone health in track and field athletes. J Sci
Med Sport 2000;3:268–79.
[4] Mora S, Gilsanz V. Establishment of peak bone mass. Endocrinol Metab
Clin North Am 2003;32:39–63.
[5] Burr DB, Miller L, Grynpas M, Li J, Boyde A, Mashiba T, et al. Tissue
mineralization is increased following 1-year treatment with high doses of
bisphosphonates in dogs. Bone 2003;33:960–9.
[6] Forwood MR, Baxter-Jones AD, Beck TJ, Mirwald RL, Howard A, Bailey
DA. Physical activity and strength of the femoral neck during the adolescent
growth spurt: a longitudinal analysis. Bone 2006;38:576–83.
[7] McKay HA, MacLean L, Petit M, Kelvie-O'Brien K, Janssen P, Beck T, et
al. “Bounce at the Bell”: a novel program of short bouts of exercise
improves proximal femur bone mass in early pubertal children. Br J Sports
Med 2005;39:521–6.
[8] Judex S, Zernicke RF. High-impact exercise and growing bone: relation
between high strain rates and enhanced bone formation. J Appl Physiol
2000;88:2183–91.
[9] Notomi T, Lee SJ, Okimoto N, Okazaki Y, Takamoto T, Nakamura T, et al.
Effects of resistance exercise training on mass, strength, and turnover of
bone in growing rats. Eur J Appl Physiol 2000;82:268–74.
[10] Hart KJ, Shaw JM, Vajda E, Hegsted M, Miller SC. Swim-trained rats have
greater bone mass, density, strength, and dynamics. J Appl Physiol
2001;91:1663–8.
[11] Bailey DA, Faulkner RA, McKay HA. Growth, physical activity, and bone
mineral acquisition. Exerc Sport Sci Rev 1996;24:233–66.
[12] Lindholm C, Hagenfeldt K, Ringertz BM. Pubertal development in elite
juvenile gymnasts. Effects of physical training. Acta Obstet Gynecol
Scand 1994;73:269–73.
[13] Lappe JM, Stegman MR, Recker RR. The impact of lifestyle factors on
stress fractures in female Army recruits. Osteoporos Int 2001;12:35–42.
[14] Bennell KL, Brukner PD. Epidemiology and site specificity of stress
fractures. Clin Sports Med 1997;16:179–96.
[15] Matheson GO, Clement DB, McKenzie DC, Taunton JE, Lloyd-Smith DR,
MacIntyre JG. Stress fractures in athletes. A study of 320 cases. Am J
Sports Med 1987;15:46–58.
[16] Karlsson MK. Physical activity, skeletal health and fractures in a long term
perspective. J Musculoskelet Neuronal Interact 2004;4:12–21.
[17] Judex S, Donahue LR, Rubin C. Genetic predisposition to low bone mass
is paralleled by an enhanced sensitivity to signals anabolic to the skeleton.
FASEB J 2002;16:1280–2.
[18] Rubin C, Recker R, Cullen D, Ryaby J, McCabe J, McLeod K. Prevention
of postmenopausal bone loss by a low-magnitude, high-frequency
mechanical stimuli: a clinical trial assessing compliance, efficacy, and
safety. J Bone Miner Res 2004;19:343–51.
[19] Judex S, Boyd S, Qin YX, Turner S, Ye K, Muller R, et al. Adaptations of
trabecular bone to low magnitude vibrations result in more uniform stress
and strain under load. Ann Biomed Eng 2003;31:12–20.
1066 L. Xie et al. / Bone 39 (2006) 1059–1066
[20] Ward K, Alsop C, Caulton J, Rubin C, Adams J, Mughal Z. Low magni-
tude mechanical loading is osteogenic in children with disabling con-
ditions. J Bone Miner Res 2004;19:360–9.
[21] Griffin MJ. Predicting the hazards of whole-body vibration—Considera-
tions of a standard. Ind Health 1998;36:83–91.
[22] International Standards Organization. Evaluation of human exposure to
whole-body vibration. ISO 2631/1.Geneva 1985.
[23] Gross TS, Poliachik SL, Ausk BJ, Sanford DA, Becker BA, Srinivasan S.
Why rest stimulates bone formation: a hypothesis based on complex adaptive
phenomenon. Exerc Sport Sci Rev 2004;32:9–13.
[24] Ausk BJ, Gross TS, Srinivasan S. An agent based model for real-time
signaling induced in osteocytic networks by mechanical stimuli.
J Biomech in press.
[25] Lamothe JM, Zernicke RF. Rest insertion combined with high-frequency
loading enhances osteogenesis. J Appl Physiol 2004;96:1788–93.
[26] Miller LM, Dumas P, Jamin N, Teillaud JL, Miklossy J, Forro L. Combining
IR spectroscopy with fluorescence imaging in a single microscope: bio-
medical applications using a synchrotron infrared source (invited). Rev Sci
Instrum 2002;73:1357–60.
[27] Miller LM, Vairavamurthy V, Chance MR, Mendelsohn R, Paschalis EP,
Betts F, et al. In situ analysis of mineral content and crystallinity in bone
using infrared microspectroscopy of the nu(4) PO(4)(3-) vibration.
Biochim Biophys Acta 2001;1527:11–9.
[28] Liu CC, Sherrard DJ, Maloney NA, Howard GA. Reactivation of inhibited
bone acid phosphatase and its significance in bone histomorphometry.
J Histochem Cytochem 1987;35:1355–63.
[29] Nordahl J, Andersson G, Reinholt FP. Chondroclasts and osteoclasts in
bones of young rats: comparison of ultrastructural and functional features.
Calcif Tissue Int 1998;63:401–8.
[30] Hayman AR, Bune AJ, Bradley JR, Rashbass J, Cox TM. Osteoclastic
tartrate-resistant acid phosphatase (Acp 5): its localization to dendritic cells
and diverse murine tissues. J Histochem Cytochem 2000;48:219–28.
[31] Robling AG, Duijvelaar KM, Geevers JV, Ohashi N, Turner CH. Modulation
of appositional and longitudinal bone growth in the rat ulna by applied static
and dynamic force. Bone 2001;29:105–13.
[32] Rubin J, Murphy TC, Fan X, Goldschmidt M, Taylor WR. Activation of
extracellular signal-regulated kinase is involved in mechanical strain
inhibition of RANKL expression in bone stromal cells. J Bone Miner Res
2002;17:1452–60.
[33] Judex S, Lei X, Han D, Rubin C. Low-magnitude mechanical signals that
stimulate bone formation in the ovariectomized rat are dependent on the
applied frequency but not on the strain magnitude. J Biomech in press.
[34] Xie LQ, Busa B, Donahue LR, Rubin C, Miller LM, Judex S. Low-level
mechanical vibrations enhance bone formation without loss of bone quality
in the growing skeleton. J Bone Miner Res 2005;20:349.
[35] Rubin C, Turner AS, Muller R, Mittra E, McLeod K, Lin W, et al. Quantity
and quality of trabecular bone in the femur are enhanced by a strongly
anabolic, noninvasive mechanical intervention. J Bone Miner Res 2002;17:
349–57.
[36] Turner CH, Forwood MR, Otter MW. Mechanotransduction in bone: do
bone cells act as sensors of fluid flow? FASEB J 1994;8:875–8.
[37] Han Y, Cowin SC, Schaffler MB, Weinbaum S. Mechanotransduction and
strain amplification in osteocyte cell processes. Proc Natl Acad Sci U S A
2004;101:16689–94.
[38] Roelants M, Delecluse C, Goris M, Verschueren S. Effects of 24 weeks of
whole body vibration training on body composition and muscle strength in
untrained females. Int J Sports Med 2004;25:1–5.
[39] Collins JJ, Chow CC, Imhoff TT. Stochastic resonance without tuning.
Nature 1995;376:236–8.
[40] Bryant JD. The effect of impact on the marrow pressure of long bones in
vitro. J Biomech 1983;16:659–65.
[41] Winet H. A bone fluid flow hypothesis for muscle pump-driven capillary
filtration: II. Proposed role for exercise in erodible scaffold implant
incorporation. Eur Cell Mater 2003;6:1–10.
[42] Murfee WL, Hammett LA, Evans C, Xie L, Squire M, Rubin C, et al. High-
frequency, low-magnitude vibrations suppress the number of blood vessels
per muscle fiber in mouse soleus muscle. J Appl Physiol 2005;98:
2376–80.
[43] 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.
[44] Vashishth D, Gibson GJ, Khoury JI, Schaffler MB, Kimura J, Fyhrie DP.
Influence of nonenzymatic glycation on biomechanical properties of cortical
bone. Bone 2001;28:195–201.