Skeletal effects of whole-body vibration in adult and aged mice


  • Michelle A. Lynch,

    1. Department of Orthopaedic Surgery, Washington University, St. Louis, Missouri 63110
    Search for more papers by this author
  • Michael D. Brodt,

    1. Department of Orthopaedic Surgery, Washington University, St. Louis, Missouri 63110
    Search for more papers by this author
  • Matthew J. Silva

    Corresponding author
    1. Department of Orthopaedic Surgery, Washington University, St. Louis, Missouri 63110
    2. Department of Biomedical Engineering, Washington University, St. Louis, Missouri 63110
    • Department of Orthopaedic Surgery, Washington University, St. Louis, Missouri 63110. T: 314-362-8585.
    Search for more papers by this author


Low-amplitude, whole-body vibration (WBV) may be anabolic for bone. Animal studies of WBV have not evaluated skeletal effects in aged animals. We exposed 75 male BALB/c mice (7 month/young-adult; 22 month/aged) to 5 weeks of daily WBV (15 min/day, 5 day/wk; 90 Hz sine wave) at acceleration amplitudes of 0 (sham), 0.3, or 1.0 g. Whole-body bone mineral content (BMC) increased with time in 7 month (p < 0.001) but not 22 month (p = 0.34) mice, independent of WBV (p = 0.60). In 7 month mice, lower-leg BMC increased with time in 0.3 and 1.0 g groups (p < 0.005) but not in the sham group (p = 0.09), indicating a positive WBV effect. In 22 month mice, there were no changes with time in lower-leg BMC (p = 0.11). WBV did not affect tibial trabecular or cortical bone structure (by µCT), dynamic indices of trabecular or cortical bone formation, trabecular osteoclast surface, or the mass of the reproductive fat pad (p > 0.05). Each of these outcomes was diminished in 7 month versus 22 month animals (p < 0.05). In summary, 5 weeks of daily exposure to low-amplitude WBV had no skeletal effects in aged male mice. The potential of WBV to enhance bone mass in age-related osteoporosis is not supported in this preclinical study. © 2009 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 28:241–247, 2010

Low bone mass contributes to skeletal fragility and age-related increases in fracture incidence in women and men.1 Physical interventions such as exercise and skeletal loading can potentially increase bone mass or reduce the loss of bone associated with aging. In particular, whole-body vibration (WBV) at low amplitude (≤1.0 g peak acceleration) and at 30 to 90 Hz frequency has been proposed as a safe intervention to enhance bone mass.2 Small clinical studies have demonstrated the potential of low-amplitude WBV to increase bone mass in adolescents and young adults,3, 4 and to prevent bone loss in postmenopausal women.5 In addition, results from studies of low-amplitude WBV in mice and rats showed anabolic effects on bone, such as increased rates of bone formation and enhanced trabecular bone volume.6–10

Most prior studies of WBV in rodents were done in growing (<4 month)9–12 or young adult (4–8 month)6, 7, 13 animals. There are no reports of WBV in aged animals, and thus no preclinical evidence to support the use of WBV as treatment for age-related osteoporosis. Prior studies of high-magnitude, low-frequency direct skeletal loading (e.g., tibial bending) have indicated that aged animals are less responsive to mechanical stimulation than young animals.14–16 We sought to determine the potential of WBV as an anabolic stimulus in the aged skeleton by comparing responses to WBV in young adult (7 month) and aged (22 month) male mice. Male mice exhibit loss of trabecular bone and cortical thinning with age,17, 18 and thus aged male mice have relevance to age-related osteoporosis in men.

Effects of WBV may depend on acceleration magnitude and frequency. We previously examined the effects of 5 weeks of daily WBV in young adult (7 month) male mice.6 Mice exposed to 0.3 g WBV (at 45 Hz) had no evidence of increased bone volume in the tibia compared to sham-loaded mice, a finding in contrast to studies in younger rodents that showed anabolic effects using 0.3 g, 45 Hz WBV.7, 9, 10 However, we did observe significantly higher trabecular bone volumes in mice in the 1.0 g WBV group compared to sham and 0.3 g groups,6 suggesting a magnitude-dependent response. Of note, a previous study in turkeys observed a progressive increase in labeled bone surface with increasing magnitude from 0.1 to 0.9 g.19 A frequency of 90 Hz showed positive effects in rodents,8, 11 and in one study 90 Hz was more effective than 45 Hz in stimulating bone formation in adult rats.13 Thus, available evidence suggests that a WBV magnitude of 1.0 g can produce a stronger anabolic response than 0.3 g, and that a frequency of 90 Hz can be anabolic in adult rodents.

We had two objectives in the current study: to determine if the effects of low-amplitude WBV would differ between adult and aged male mice, and to determine if a magnitude effect exists for WBV loading at 90 Hz. We hypothesized that aged mice are less responsive to loading than young adult mice, and that 1.0 g is more anabolic than 0.3 g.


Animals and WBV Loading

Following approval by our institutional Animal Studies Committee, we obtained 82 male BALB/c mice (Aged Rodent Colony; National Institute of Aging). Mice were 7 or 22 months of age at the start of the experiment and were assigned at random to loading groups based on magnitude of applied acceleration: 0 (sham), 0.3, or 1.0 g, with 13–15 mice per age and loading group. Mice were exposed to WBV or sham loading for 15 min/day, 5 days/week for 5 weeks using a custom-built device that has been described previously.6 Briefly, mice were placed in groups of four to five in a plastic cage attached to a rigid platform. The platform was driven in the vertical direction from below by an electromagnetic actuator connected to a function generator and power supply/amplifier. The vibration frequency was 90 Hz (sine wave). An accelerometer attached to the center of the plate was used to monitor peak acceleration; the amplitude was adjusted at the start of each session to the desired value. The acceleration waveform was centered at 0, therefore for a magnitude of 0.3 g the plate oscillated between ±0.3 g (0.6 g peak-to-peak). For sham loading, no power was supplied to the actuator. Mice were injected with calcein green (7 mg/kg, i.p.; Sigma) on days 5 and 15 to label bone forming surfaces.

Seven mice died prior to completion of the study, four from accidental anesthesia overdose and three for unknown reasons while in their cages. These included two 7 month mice (one from the 0 g loading group, one from the 1.0 g group) and five 22 month mice (two 0 g, one 0.3 g, two 1.0 g). Deaths did not appear to be related to WBV. No data from these mice are reported. At the completion of the 5 week study, the remaining 75 mice (n = 12–13 per group) were euthanized by CO2 asphyxiation. Bilateral tibiae were dissected, fixed overnight in 4% paraformaldehyde, then placed in 70% ethanol.

DXA Scanning

At baseline (0 weeks), 2.5, and 5 weeks, body masses were recorded, and mice were scanned using DXA (GE Lunar PIXImus). Mice were anesthetized by i.p. injection of ketamine (100 mg/kg BW) and xylazine (10 mg/kg), and two scans per mouse were obtained. For the first scan, the entire mouse was placed on the positioning tray; from this a whole-body analysis (excluding the skull) was performed. For the second scan, the left leg was placed approximately in the plane of the positioning tray; from this a lower-leg analysis (below the knee) was performed. Bone mineral content (BMC) and areal bone mineral density (aBMD) were determined for whole-body and lower-leg regions of interest.

Fat Pad Mass

Immediately postmortem, bilateral reproductive (epididymis) fat pads were removed from each mouse and weighed. In pilot studies, we determined that these were the only fat pads that could be reproducibly harvested, especially in the 22 month mice.


The left tibiae were scanned by µCT at the proximal metaphysis to assess trabecular bone, and at the mid-diaphysis (3 mm proximal to the tibiofibular junction) to assess cortical bone. Bones were suspended in agarose, placed in a 16 mm diameter plastic tube, and scanned at 16 µm voxel size (standard resolution, 55 kV, 134 µA, 150 ms integration; µCT 40, SCANCO Medical). Analysis was performed using the manufacturer's evaluation software and a fixed global threshold of 28% of maximum gray-scale value (280/1000). The trabecular volume of interest included all tissue within the endocortical margin, spanning a longitudinal height of 0.48 mm (30 slices) starting just distal to the proximal growth plate. Bone volume (BV/TV), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), and trabecular number (Tb.N) were computed based on the direct method of calculation (not based on stereological models).20 Apparent volumetric BMD (vBMD) was computed based on the entire tissue volume (TV), and tissue mineral density (TMD) was computed based on the bone volume (BV). In both cases these are calibrated to the manufacturer's hydroxyapatite (HA) phantom. Cortical TMD was determined over 30 slices spanning 3 mm at the mid-diaphysis.


The left tibiae were then dehydrated in ascending concentrations of ethanol (70–100%) and embedded in methylmethacrylate (Sigma). The plastic block was divided into proximal and distal pieces. From the proximal piece, we cut two midsagittal sections (100 µm thickness) using a saw microtome (Leica SP 1600). From the distal piece, we cut two mid-diaphyseal sections at a site 3 mm proximal to the distal tibiofibular junction. Sections were mounted on glass slides, coverslipped, and visualized on an inverted microscope (DP-30; Olympus) with a 100 W mercury-halogen light source at a 10X objective using a fluorescein isothiocyanate (FITC) filter set. Brightfield and fluorescent images were captured and analyzed using commercial software (BIOQUANT OSTEO). We determined bone formation indices as per histomorphometry standards21: single- (sLS/BS) and double-labeled (dLS/BS) bone surface, mineralizing surface (MS/BS), mineral apposition rate (MAR), and bone formation rate (BFR/BS). For samples with no detectable double label, MAR and BFR/BS were treated as missing data.22 For the proximal sections, the region of interest (ROI) was the metaphyseal trabecular bone beneath the growth plate (same ROI as for µCT analysis). For the diaphyseal sections, we analyzed the endocortical (Ec) and the periosteal (Ps) surfaces separately. We also assessed diaphyseal morphology: bone area (Ct.B.Ar), medullary area (Ec.Ar), and cortical width (Ct.Wi). Analysis was performed on duplicate sections and values averaged.

The right tibiae were decalcified in 14% EDTA and embedded in paraffin. Midsagittal sections (5 µm) were cut and stained for tartrate resistant acid phosphatase (TRAP).23 Trabecular osteoclast surface (Oc.S/BS) was determined as the fraction of bone surface covered by TRAP-positive cells.

Data Analysis

Body mass and DXA outcomes were analyzed by repeated measures ANOVA (Statview; SAS Institute) with time (0, 2.5, 5 weeks) as a repeated factor, and loading group (sham, 0.3 g, 1.0 g) and age (7, 22 month) as between factors. Two-factor ANOVA (loading group, age) was performed at each time point and for postmortem outcomes. Unless otherwise stated, p values refer to the ANOVA results. In some cases ANOVA indicated a significant interaction between age and another factor, in which case the 7 and 22 month groups were reanalyzed separately. Posthoc multiple comparisons (Tukey-Kramer) were done when ANOVA results were significant. Significance was defined as p ≤ 0.05. Data are reported as mean ± SD.


Modest reductions in body mass were observed during the 5 weeks, dependent on age (age-time, p = 0.023; Table 1). Mice in the 7 month age group did not lose body mass (p = 0.97); mice in the 22 month group lost an average of 3% (p = 0.002). An interaction occurred between loading group and loss of mass in the 22 month mice (p = 0.001); mice in the 0.3 g group did not lose mass, while mice in the sham and 1.0 g groups did. No group lost more than 1.5 g (∼5%). Mice in the 22 month group had lower body mass than mice in the 7 month group at each time point (p < 0.001), and the mass of the fat pad (measured postmortem) was 45% less in 22 month mice than in 7 month mice (p < 0.001; Table 1). Importantly, fat pad mass was unaffected by loading (p = 0.44).

Table 1. Body Mass at 0, 2.5, and 5 Weeks and Fat Pad Mass at 5 Weeks
 7 Month22 Month
Sham0.3 g1.0 gSham0.3 g1.0 g
  • n = 12–13.

  • *

    Twenty-two month different from 7 month, p < 0.001 by ANOVA.

  • Different from 0 week.

  • Different from 2.5 week; p < 0.05 by Tukey-Kramer test.

Body mass (gm)
0 week*31.7 ± 2.131.3 ± 2.333.2 ± 2.030.1 ± 3.729.4 ± 3.228.7 ± 3.1
2.5 week*31.4 ± 2.431.5 ± 2.733.1 ± 1.930.0 ± 4.029.3 ± 3.228.0 ± 3.0
5 week*30.7 ± 2.731.8 ± 2.433.6 ± 1.628.5 ± 4.629.8 ± 3.427.4 ± 3.1
Fat pad mass (gm)
5 week*0.24 ± 0.140.26 ± 0.110.29 ± 0.090.13 ± 0.100.19 ± 0.110.12 ± 0.07

Whole-body BMC was unaffected by WBV loading (p = 0.60), but did increase with time in an age-dependent manner (time-age interaction, p < 0.001). Mice in the 7 month group had progressive increases in whole-body BMC during the study (p < 0.001), whereas mice in the 22 month group did not (p = 0.34) (Fig. 1A; Table 2). Similar though smaller changes were observed in aBMD (data not shown). Importantly, temporal changes in BMC and aBMD were unaffected by loading, with no significant differences between loading groups at any time point. BMC was significantly greater in 22 month mice than in 7 month mice at 0 and 2.5 weeks (p < 0.05), but by 5 weeks this difference was no longer significant (p = 0.15).

Figure 1.

Serial DXA scans revealed temporal increases in BMC in some groups. (A) In 7-month-old mice, whole-body BMC increased an average of 4% from 0 to 2.5 weeks and an additional 2% from 2.5 to 5 weeks (p < 0.05) independent of WBV. Whole-body BMC did not change with time in 22 month mice. (B) In 7 month mice, there were loading-dependent increases in the bone mass of the lower leg. Lower-leg BMC increased by 5% from 0 to 5 weeks in both 0.3 and 1.0 g WBV groups (p < 0.05) but did not increase significantly in the sham group, indicating a positive loading effect. In 22 month mice, there were no significant increases in lower-leg BMC over time. a, different from 0 week; b, different from 2.5 week; p < 0.05.

Table 2. Whole-Body and Lower-Leg Bone Mineral Content (BMC) at 0, 2.5, and 5 Weeks
  7 Month22 Month
Sham0.3 g1.0 gSham0.3 g1.0 g
  • n = 12–13.

  • *

    Twenty-two month different from 7 month, p < 0.05 by ANOVA.

  • Different from 0 week.

  • Different from 2.5 week; p < 0.05.

Whole-body0 week*0.543 ± 0.0350.535 ± 0.0420.563 ± 0.0300.605 ± 0.0640.597 ± 0.0550.597 ± 0.039
BMC2.5 week*0.562 ± 0.0360.551 ± 0.0460.587 ± 0.0340.597 ± 0.0730.594 ± 0.0500.592 ± 0.053
(gm)5 week0.580, ± 0.0400.564 ± 0.0520.592 ± 0.0280.596 ± 0.0710.600 ± 0.0490.590 ± 0.055
Lower-leg0 week*0.0456 ± 0.00420.0437 ± 0.00400.0472 ± 0.00310.0484 ± 0.00670.0480 ± 0.00610.0481 ± 0.0046
BMC2.5 week0.0460 ± 0.00440.0445 ± 0.00510.0491 ± 0.00360.0483 ± 0.00680.0480 ± 0.00540.0480 ± 0.0050
(gm)5 week0.0464 ± 0.00420.0458, ± 0.00500.0494 ± 0.00350.0484 ± 0.00660.0493 ± 0.00650.0488 ± 0.0060

Lower-leg BMC also increased with time in an age-dependent manner (age-time interaction, p = 0.03) with evidence of a modest, positive effect of WBV loading in 7 month mice. Similar to the whole-body analysis, mice in the 7 month group had progressive increases in lower-leg BMC during the study (p < 0.001), whereas mice in the 22 month group did not (p = 0.11). Moreover, an interaction was found between time and loading in the 7 month mice (p = 0.05), and posthoc analysis indicated that lower-leg BMC in the 7 month mice increased significantly with time in the 0.3 g (p = 0.0014) and 1.0 g (p < 0.001) groups, but not in the sham group (p = 0.09) (Table 2; Fig. 1B). Despite these relative differences, at the end of the study lower-leg BMC in 7 month mice did not differ between loading groups (p = 0.51). In the 22 month mice, lower-leg BMC did not increase with time in the sham (p = 0.97) or 1.0 g (p = 0.52) groups, and a slight increase in the 0.3 g group did not reach significance (p = 0.051). Lower-leg BMC was significantly greater in 22 month mice than in 7 month mice at 0 weeks (p < 0.05), but not at the other time points. Lower-leg aBMD did not depend on time, age, or loading (data not shown).

Trabecular bone structure and formation were not affected by WBV loading, but did differ between young adult and aged mice. Measures of bone structure (BV/TV, Tb.Th, Tb.N, Tb.Sp) and density (vBMD, TMD) were not significantly different between loading groups, and there were no significant interactions between age and loading (p > 0.05; Table 3). Dynamic indices of trabecular bone formation, MS/BS (p = 0.87) and BFR/BS (p = 0.53; Fig. 2B), were unaffected by WBV loading, and effects of loading on MAR (15% less in the 1.0 g group than the sham group) did not reach significance (p = 0.055). Osteoclast surface, an index of resorption, was also unaffected by loading (p = 0.95). All measures of trabecular bone structure and formation were diminished in 22 month mice compared to 7 month mice (p < 0.05). For example, BV/TV and BFR/BS were 65 and 35% less, respectively, in 22 month versus 7 month mice (p < 0.001).

Table 3. Trabecular Morphology, Density, and Indices of Bone Formation from Proximal Tibial Metaphysis after 5 Weeks of Loading
 7 Month22 Month
Sham0.3 g1.0 gSham0.3 g1.0 g
  • n = 12–13 except as noted.

  • *

    Twenty-two month different from 7 month, p < 0.05 by ANOVA.

Tb.Th* (mm)0.068 ± 0.0080.069 ± 0.0060.066 ± 0.0060.054 ± 0.0070.059 ± 0.0060.055 ± 0.004
Tb.N* (1/mm)5.2 ± 0.25.1 ± 0.25.1 ± 0.24.8 ± 0.14.8 ± 0.34.7 ± 0.2
Tb.Sp* (mm)0.206 ± 0.0100.208 ± 0.0100.209 ± 0.0080.227 ± 0.0100.228 ± 0.0140.227 ± 0.016
vBMD* (mg/cm3)239 ± 34239 ± 25227 ± 26131 ± 18137 ± 24131 ± 24
TMD* (mg/cm3)792 ± 61797 ± 55794 ± 51843 ± 58837 ± 73838 ± 71
MS/BS* (%)10.1 ± 3.99.5 ± 3.49.5 ± 3.27.1 ± 2.58.5 ± 4.87.5 ± 4.4
MAR* (μm/day)0.86 ± 0.220.88 ± 0.190.76 ± 0.170.63 ± 0.24 (n = 7)0.68 ± 0.200.51 ± 0.25 (n = 10)
Oc.S* (%)7.3 ± 3.2 (n = 11)9.2 ± 3.58.5 ± 2.5 (n = 11)4.9 ± 2.7 (n = 10)3.5 ± 1.9 (n = 11)4.0 ± 2.4
Figure 2.

Trabecular (A) bone volume and (B) bone formation rate at the proximal tibial metaphysis. Twenty-two month mice had significantly less bone and diminished bone formation rate compared to 7 month mice (p < 0.01), but WBV did not influence trabecular bone structure or dynamic indices of bone formation.

Cortical bone structure and formation were unaffected by WBV loading (p > 0.05 for all outcomes), but did differ between young adult and aged mice. Cortical area and thickness were 10 and 25% less, respectively, in 22 month versus 7 month mice, while medullary area was 50% greater (p < 0.001; Table 4). Rates of cortical bone formation were extremely low; only 4 of 75 specimens had nonzero double-labeled periosteal surface, and only 3 of 75 had double-labeled endocortical surface. Periosteal mineralizing surface was less in 22 month versus 7 month mice (p = 0.004); endocortical mineralizing surface did not differ between ages.

Table 4. Cortical Morphology, Density, and Indices of Bone Formation from Tibial Diaphysis after 5 Weeks of Loading
 7 Month22 Month
Sham0.3 g1.0 gSham0.3 g1.0 g
  • n = 12–13.

  • *

    Twenty-two month different from 7 month, p < 0.05 by ANOVA.

Ct.Ar* (mm2)0.78 ± 0.090.77 ± 0.060.77 ± 0.060.69 ± 0.080.70 ± 0.090.68 ± 0.10
Ec.Ar* (mm2)0.38 ± 0.040.34 ± 0.050.36 ± 0.030.54 ± 0.090.52 ± 0.100.56 ± 0.07
Ct.Wi* (mm)0.22 ± 0.010.23 ± 0.020.23 ± 0.010.17 ± 0.020.18 ± 0.020.17 ± 0.02
TMD (mg/cm3)1,213 ± 291,216 ± 271,211 ± 231,211 ± 371,200 ± 351,213 ± 45
Ec.MS/BS (%)6.2 ± 7.61.8 ± 3.84.2 ± 5.03.3 ± 3.07.1 ± 8.15.7 ± 7.9
Ps.MS/BS* (%)3.1 ± 4.75.1 ± 5.64.7 ± 6.61.7 ± 4.20.9 ± 1.91.1 ± 1.6


We examined the skeletal effects of daily exposure to low-amplitude (0.3 and 1.0 g), high-frequency (90 Hz) WBV in young adult and aged male mice. Our results indicate that WBV had few effects. WBV did not enhance the normal time-related accrual of whole-body BMC in 7 month mice, nor did it overcome the lack of increase seen in 22 month mice during the 5 week study duration. Similarly, WBV did not alter measures of trabecular bone structure, formation, or resorption at the proximal tibia, cortical bone structure and formation at the tibial diaphysis, or the mass of the reproductive fat pads. WBV did have modest, positive effects on accrual of BMC in the lower leg, but only in 7 month mice. In answer to our first hypothesis, our data indicate that 22 month mice were less responsive to WBV than 7 month mice, although this relative difference was seen in only one outcome (lower-leg BMC); overall, both ages were largely unresponsive. In answer to our second hypothesis, we did not find that 1.0 g/90 Hz WBV was more anabolic than 0.3 g/90 Hz WBV, as both were minimally effective as an anabolic stimulus in adult mice.

The response of aged animals to WBV has not been previously reported. Most WBV studies using intact (i.e., non-OVX) small animals were done with growing animals (≤4 month)7, 9, 10, 12 although some used young adults (6–8 month).6, 8 Mice have an average lifespan of about 2 years, and aged male mice exhibit changes in bone structure such as cortical thinning and trabecular bone loss.17, 18 We observed that, although 22 month male BALB/c mice have greater whole-body BMC than 7 month mice, they have structural evidence of osteopenia and lower rates of bone formation and resorption, supporting their use as a model of age-related osteoporosis. Previous studies that compared the mechanoresponsiveness of young and old animals reported conflicting results. Studies with high-magnitude, direct skeletal loading (ulnar compression in turkeys,14 tibial bending in rats16 and mice15) reported reduced mechanoresponsiveness in aged animals, but studies with exercise reported either no difference with age24 or enhanced responsiveness in aged animals.25, 26 Thus, mechanoresponsiveness in aged animals appears to depend on the type of stimulus. In our study, because both 7 and 22 month mice were relatively unresponsive, we cannot make a conclusion as to how aging influences the skeletal effects of WBV. Additional studies comparing young, growing animals and adults are needed to better understand the effects of age on responsiveness to WBV.

Other parameters that may affect responsiveness to WBV in mice are sex and inbred strain. Our study was done using male, BALB/c mice. Although the majority of WBV studies in rodents were done using females,7–10, 12, 13, 27 there are reports indicating positive responses in males.6, 11 We know of no reason why male mice would be less responsive, although direct comparisons between male and female animals have not been reported. We selected the BALB/c strain, in part, because a prior report indicated that they were responsive to WBV.7 Moreover, unpublished data from our lab indicate that adult, male BALB/c mice are responsive to high-strain tibial compression. Thus, while the choice of sex and strain may have contributed to the lack of response, we believe that this is unlikely.

WBV studies have been done using a range of loading magnitudes and frequencies, but there have been few parameter studies, and no consensus exists on the most effective magnitude/frequency combination for stimulating bone formation.28 We focused on low-amplitude WBV of ≤1 g, which is likely to be safer for use in aged or osteoporotic individuals than WBV at >1 g. When the input acceleration exceeds 1 g, the movement of the body can become out of phase with the input and expose the subject to large dynamic loads, which would increase the risk of skeletal or other tissue damage.29 While it was not our goal to perform a parameter study, we did compare two acceleration magnitudes. WBV at 0.3 g has been reported to be effective in stimulating bone formation in young mice.7, 9, 10 We previously reported that 0.3 g did not cause increased trabecular BV/TV in adult mice compared to sham controls, whereas 1.0 g did.6 This finding was consistent with a report that increasing WBV magnitude in the 0.1 to 0.9 g range resulted in increasing labeled bone surface in turkeys.19 In the current study, we found no evidence of a differential response to the two magnitudes. We selected 90 Hz because our previous study in adult mice found inconsistent effects at 45 Hz,6 and a recent report indicated that 90 Hz was modestly effective in stimulating bone formation in ovariectomized adult rats, whereas 45 Hz was ineffective.13 One obvious difference is the increase in the number of loading cycles with 90 Hz (for the same daily exposure of 15 min), which should increase the anabolic stimulus.30 Nonetheless, the use of 90 Hz in the current study was largely ineffective.

The only positive response to WBV that we detected was an increase from 0 to 5 weeks in lower-leg BMC in 7 month animals, which was not observed in sham-loaded mice (Fig. 1B). However, this finding is at odds with the analysis of tibial bone formation and structure, which did not reveal differences between groups. The BMC increases were modest and likely represent the cumulative effects of small increases throughout the tibia, fibula, and foot. Any local changes in bone formation and structure at the proximal metaphysis and the mid-diaphysis might have been too small to detect, which raises the issue of statistical power. Posthoc analysis indicated that ANOVA could have detected differences between groups (in 7 month mice) of 15% in trabecular BV/TV and 6% in cortical area (n = 12–13, p = 0.05, power = 0.8). Thus, our study was adequately powered to detect differences that might be of structural importance.

Our study has several limitations. First, we examined only the tibia, despite the fact that the entire body is exposed to the WBV stimulus. In a previous WBV study,6 there was stronger evidence for effects at the proximal tibia than at other sites. Consistent with this site-specificity, in the current study WBV did not influence bone mass at the whole-body level, but we did see effects at the lower leg. The lower leg is in contact with the vibrating platform via the foot, and thus it is likely that this site is exposed to a greater level of stimulus than more proximal sites. A second limitation is that the study duration may not have been long enough to observe structural effects. Other studies of vibration reported increased measures of bone formation but no structural effects,7–9, 31 possibly because these studies were only 2 to 4 weeks in duration. Recent studies by Xie et al.9 found that 3 weeks of daily WBV (0.3 g/45 Hz) in growing mice failed to produce structural effects, whereas the same protocol applied for 6 weeks led to increased trabecular bone volume.10 Our study duration was 5 weeks, which was long enough to observe normal age-related changes in bone mass in 7 month mice and a modest increase in lower-leg BMC with WBV. Nonetheless, greater effects on bone mass and structure might be observed with a longer study. A third limitation is that we analyzed dynamic bone formation indices relatively early in the study period. We chose calcein injection times of 5 and 15 days to increase the likelihood of observing changes that occurred in the first 2 weeks, based on the results of short-duration WBV studies.7–9, 31 Labeling the bone at a later time, or with a longer interlabel interval, might have produced different results.

In summary, adult mice exposed to daily WBV for 5 weeks showed minimal skeletal responses. Whole-body BMC and tibial bone structure and bone formation rates were not positively affected by WBV. In 7 month mice, the BMC of the lower leg increased more in mice subjected to WBV than sham-loaded mice, suggesting that WBV has the potential to enhance lower-extremity bone mass in adults. However, this effect was not observed in 22 month mice, and overall we found no evidence that WBV enhances bone mass or structure in the aged skeleton. Our findings are comparable to other studies of WBV in small animals that reported modest effects that are often observed in some but not most outcome measures.6–10, 12, 13 Additional studies are needed to determine if WBV conditions exist that produce clinically significant benefits in the adult skeleton.


Presented in part at the 2008 annual meeting of the Orthopaedic Research Society. Supported by a grant from the National Institute of Arthritis, Musculoskeletal and Skin Diseases, National Institutes of Health (AR47867). Animals were housed in a facility supported by an NIH grant (NCRR C06 RR015502). We thank Crystal Idleburg for assistance with histology.