Effect of Whole-Body Vibration on Calcaneal ...

Effect of Whole-Body Vibration on Calcaneal Quantitative Ultrasound Measurements in Postmenopausal Women: A Randomized Controlled Trial Lubomira Slatkovska, Joseph Beyene, Shabbir M. H. Alibhai, Queenie Wong, Qazi Z. Sohail & Angela M. Cheung Calcified Tissue International and Musculoskeletal Research ISSN 0171-967X Volume 95 Number 6 Calcif Tissue Int (2014) 95:547-556 DOI 10.1007/s00223-014-9920-1

1 23

Your article is protected by copyright and all rights are held exclusively by Springer Science +Business Media New York. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23

Author's personal copy Calcif Tissue Int (2014) 95:547–556 DOI 10.1007/s00223-014-9920-1

ORIGINAL RESEARCH

Effect of Whole-Body Vibration on Calcaneal Quantitative Ultrasound Measurements in Postmenopausal Women: A Randomized Controlled Trial Lubomira Slatkovska • Joseph Beyene • Shabbir M. H. Alibhai • Queenie Wong Qazi Z. Sohail • Angela M. Cheung

•

Received: 3 July 2014 / Accepted: 14 October 2014 / Published online: 12 November 2014 Ó Springer Science+Business Media New York 2014

Abstract The purpose of this study was to examine the effect of whole-body vibration (WBV) on calcaneal quantitative ultrasound (QUS) measurements; which has rarely been examined. We conducted a single-centre, 12-month, randomized controlled trial. 202 postmenopausal women with BMD T score between -1.0 and -2.5, not receiving bone medications, were asked to stand on a 0.3 g WBV platform oscillating at either 90- or 30-Hz for 20 consecutive minutes daily, or to serve as controls. Calcium and vitamin D was provided to all participants. Calcaneal broadband attenuation (BUA), speed of sound, and QUS index were obtained as pre-specified secondary endpoints at baseline and 12 months by using a Hologic Sahara Clinical Bone Sonometer. 12-months of WBV did not improve QUS parameters in any of our analyses. While most of our analyses showed no statistical differences between the WBV groups and the control group, mean

calcaneal BUA decreased in the 90-Hz (-0.4 [95 % CI -1.9 to 1.2] dB MHz-1) and 30-Hz (-0.7 [95 % CI -2.3 to 0.8] dB MHz-1) WBV groups and increased in the control group (1.3 [95 % CI 0.0–2.6] dB MHz-1). Decreases in BUA in the 90-, 30-Hz or combined WBV groups were statistically different from the control group in a few of the analyses including all randomized participants, as well as in analyses excluding participants who had missing QUS measurement and those who initiated hormone therapy or were \80 % adherent. Although there are consistent trends, not all analyses reached statistical significance. 0.3 g WBV at 90 or 30 Hz prescribed for 20 min daily for 12 months did not improve any QUS parameters, but instead resulted in a statistically significant, yet small, decrease in calcaneal BUA in postmenopausal women in several analyses. These unexpected findings require further investigation.

L. Slatkovska S. M. H. Alibhai Q. Wong A. M. Cheung (&) Osteoporosis Program, University Health Network/Mount Sinai Hospital, 200 Elizabeth Street, 7 Eaton North, Room 221, Toronto, ON, Canada e-mail: [email protected]

J. Beyene Department of Clinical Epidemiology & Biostatistics, McMaster University, Hamilton, ON, Canada

L. Slatkovska Q. Wong Q. Z. Sohail A. M. Cheung Women’s Health Program, University Health Network, Toronto, ON, Canada L. Slatkovska S. M. H. Alibhai A. M. Cheung Institute of Medical Science, University of Toronto, Toronto, ON, Canada J. Beyene A. M. Cheung Dalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada

S. M. H. Alibhai Q. Z. Sohail A. M. Cheung Department of Medicine, University of Toronto, Toronto, ON, Canada S. M. H. Alibhai A. M. Cheung Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, ON, Canada Q. Wong A. M. Cheung Centre of Excellence in Skeletal Health Assessment, Joint Department of Medical Imaging, University of Toronto, Toronto, ON, Canada

123

Author's personal copy 548

Keywords Whole-body vibration Menopause Quantitative ultrasound Calcaneus Randomized controlled trial

Introduction Whole-body vibration (WBV) therapy involves the transmittance of mechanical vibrations to the musculoskeletal system by means of an oscillating platform. Various types of WBV platforms exist on the market and have been investigated in clinical trials. These platforms vary in terms of vibration frequencies (Hz) and vertical accelerations (g) [1]. Duration of the WBV protocol, treatment frequency (number of treatments per day and rest periods between treatments), and treatment duration have also varied in previous investigations [1]. Therefore, currently it is unclear which WBV regimen produces the most desired effect of WBV on the skeleton with minimal deleterious effects on the rest of the body [1]. The calcaneus is the closest skeletal site to the WBV platform. When standing on a platform with vertical accelerations, the vibrations are transmitted through the feet to the weight-bearing skeleton, and typically become weaker as the distance from the platform increases, because of the cushioning (or dampening) provided by major joints and soft-tissue [2, 3]. The calcaneus, however, is separated from the source of vibration by only a thin layer of soft tissue, and thus is directly in contact with the platform accelerating upwards. It is almost entirely composed of trabecular bone, which is more metabolically active and may respond faster to treatment than cortical bone. Changes in the calcaneal bone are commonly measured using quantitative ultrasound (QUS), which projects ultrasound waves through the heel, and thereby collects different information about bone material properties than bone densitometry tools such as dual-energy X-ray absorptiometry (DXA) and high-resolution peripheral quantitative computed tomography (HR-pQCT). QUS parameters, broadband attenuation (BUA) and speed of sound (SOS), provide relatively good estimates of calcaneal BMD, and BUA in particular may also reflect trabecular microarchitectural properties [4, 5]. Yet the effect of WBV therapy on the calcaneus has rarely been examined using QUS [1, 6, 7]. Randomized controlled trials (RCTs) of WBV therapy in postmenopausal women have primarily examined hip and lumbar spine areal BMD obtained with DXA [6–12]. A statistically significant, although clinically small, increase was seen at the hip in two trials [8, 10], but none of the trials found a significant effect at the lumbar spine. Volumetric BMD at the distal tibia obtained with HR-pQCT showed no significant changes in postmenopausal women

123

L. Slatkovska et al.: Effect of WBV on Calcaneal QUS Measurements

in two trials [12, 13]. To our knowledge, calcaneal assessment using QUS was performed in only one RCT of WBV in postmenopausal women [7]. A significant improvement in calcaneal BUA was found in response to twice-weekly, 6-minute sessions of WBV at C1 g and 12.5 Hz (3.4 %, p = 0.05), but not in response to twiceweekly, 15-minute sessions of WBV at 0.3 g and 30 Hz (-0.8 %, p = 0.44) or no WBV (-3.1 %, p = 0.08) [7]. However, no significant between-group differences in BUA changes were found and SOS was not reported [7]. Furthermore, the trial was small (n = 47) with short follow-up duration (8 months), and vitamin D adequacy was not documented [7]. We conducted a 12-month RCT in 202 postmenopausal women who were provided with calcium and vitamin D supplements, and compared effects of daily 20-minute WBV at 0.3 g and 90- or 30-Hz with no WBV (Vibration Study). We have previously reported on our main outcomes [12]; no effect of WBV was found on distal tibial volumetric BMD and parameters of bone microstructure assessed by HRpQCT, or hip and spine areal BMD assessed by DXA [12]. Calcaneal QUS outcomes were collected as pre-specified secondary endpoints and examined separately, as we expected the calcaneus to receive a more intense WBV stimulus due to its proximity to the oscillating platform than our primary endpoint location (distal tibia). Our a priori hypothesis was that year-long, daily 20-min WBV therapy will improve calcaneal QUS outcomes. In this paper, we are reporting the results of the QUS outcomes of the vibration study.

Methods Trial Design, Setting, and Randomization A 12-month, superiority RCT with three parallel arms was conducted at the Postmenopausal Health Research Clinic of Toronto General Hospital, University Health Network, Toronto, Canada. Recruitment started in October 2006 and finished in November 2008 when the target sample size was achieved. Calcaneal QUS measurements were obtained as pre-specified secondary endpoints at baseline and 12 months. HR-pQCT and DXA outcomes were collected and reported previously [12]. A computer-generated block-randomization scheme with 1:1:1 allocation ratio and block size of 12 was used to assign eligible participants to receive one of three interventions: WBV at 0.3 g and 90 Hz, WBV at 0.3 g and 30 Hz, or no WBV (control group). Sealed envelopes containing participant number and group allocation were opened sequentially at baseline after eligibility criteria were satisfied and baseline calcaneal QUS outcomes were

Author's personal copy L. Slatkovska et al.: Effect of WBV on Calcaneal QUS Measurements

collected. Sham WBV was not utilized in controls and QUS outcome assessment was not blinded. Participants knew whether or not they were controls, but were unaware of the 90-Hz versus 30-Hz group assignment. The trial was approved by the University Health Network research ethics board, registered at the ClinicalTrials.gov (#NCT00420940) and funded by the Physicians’ Services Incorporated Foundation. Participants Potential participants were recruited in the Greater Toronto Area primarily by using posted flyers, word of mouth and our postmenopausal health newsletter. Women were eligible if they had experienced cessation of menses 1 or more years prior and their lowest BMD T score at the lumbar spine, femoral neck, or total hip was between -1.0 and -2.5. We excluded women with a BMD T score greater than -1.0, because previous research has shown that less-dense bones may have a greater response to WBV [11, 14, 15]. Other exclusion criteria included osteoporosis (BMD T score of B-2.5); fragility fracture after age 40; secondary causes of bone loss; other metabolic bone diseases or diseases affecting bone metabolism; history of active cancer in the past 5 years; body mass of C90 kg; knee or hip joint replacements; spinal implants; use of hormone therapy in the past 12 months, raloxifene or teriparatide in the past 6 months, or bisphosphonates for C3 months or within the past 3 months; chronic glucocorticoid, anticoagulant or anticonvulsant therapy; inability to tolerate WBV for 20 consecutive minutes at screening; and expected changes in physical activity levels or out-of-town travels for more than four consecutive weeks. Interventions and Adherence Participants randomized to the 90- or 30-Hz groups were given WBV platforms synchronously oscillating at a frequency of 90- or 30-Hz, respectively, with a peak acceleration or magnitude of 0.3 g (peak-to-peak displacement of \50 lm), provided by Juvent Regenerative Technologies Corporation, Riveria Beach, Florida [16]. At baseline, the participants were instructed to stand erect on the oscillating platform at home for 20 consecutive minutes daily for 12 months, with neutral posture at the neck, lumbar spine, and knees, wearing socks or barefoot, and without excessive foot or body movements. Self-reported adherence to WBV was obtained at 6 months and feedback was provided. Actual adherence was extracted from each WBV platform at 12 months by using an internal clock that recorded the date, time, and duration of every session. Percentage of adherence to WBV was calculated on the basis of total cumulative duration of WBV performed during the study [(total minutes

549

of WBV performed at any time during study participation) 7 (total study days 9 20 min) 9 100]. We chose to examine a magnitude of 0.3 g, because lower WBV magnitudes (0.3 vs. 0.6 g) were previously found to be more effective on bone in adult female mice [17]. Further, we compared 90- and 30-Hz frequencies, since at 0.2 g, WBV at 90-Hz was shown to be more effective than WBV at 45-Hz in ovariectomized rats [18], and no RCTs up to date have compared high versus low frequencies. Finally, we chose a dose of 20 consecutive minutes a day, because WBV at 0.3 g and 30-Hz was found to have no significant effect on hip or spine areal BMD in postmenopausal women treated for 10 min twice daily for 12-months, while a significant increase in trabecular BMD was found at the femur in adult ewes treated for 20 consecutive minutes 5 days a week for 12 months [11, 19]. Our WBV protocol was considered safe based on the International Organization for Standardization recommendations in industries that use machinery involving vibration (ISO 2631) [2, 11, 20]. Control participants were asked not to use WBV therapies. Calcium and vitamin D supplements were provided to all participants at baseline and 6 months, so that their total daily intakes from diet plus supplements approximated 1,200 mg and 1,000 IU, respectively, as estimated by a validated recall questionnaire [21]. Calcium and vitamin D intakes were additionally assessed at 12 months using the same validated recall questionnaire [21], at which point self-reported estimates of overall adherence to calcium and vitamin D supplements were also obtained. Outcomes and Follow-Up Calcaneal QUS measurements were collected as secondary endpoints, because beneficial effects of WBV on bone were found to be more pronounced within the trabecular versus cortical bone tissue, and at weight-bearing skeletal sites located closer to the oscillating platforms in previous studies [22, 23]. BUA (dB MHz-1), SOS (m s-1), and QUS index (QUI, 0.41 9 [BUA ? SOS] - 571) were obtained at baseline and at 12 months using a Sahara Clinical Bone Sonometer (Hologic, Bedford, MA). QUS assesses the speed (i.e., SOS) and attenuation (i.e., BUA) of an ultrasound beam as it passes through the calcaneus, and QUI combines these two results linearly and re-scales them into heel BMD units. Therefore, both BUA and SOS reflect calcaneal BMD status: the denser the calcaneal bone, the greater the attenuation and speed of the ultrasound wave [4, 5, 24]. However, BUA has been found to also reflect trabecular microarchitecture status, possibly because as the ultrasound waves pass through bone they may become scattered and absorbed by the trabecular scaffolding [4, 5, 24].

123


A single, trained assessor performed calibration and measurements in the same room and using the same device for the entire duration of data collection. Calibration was performed on the day of measurement using a manufacturer-specific phantom, and if unacceptable quality control values were obtained, calibration was repeated until satisfactory. The quality control values for BUA and SOS were stable throughout our study period, with the exception of a slight upward drift in BUA in the last 2 months. During each measurement, all participants were asked to sit still in the same chair with the nondominant foot placed in a marked area on the device according to manufacturer instructions to minimize measurement error. If a measurement was indicated as invalid by the device, it was repeated up to three times. Two sources of error were identified and resulted in the exclusion of several QUS measurements from the analysis: (1) unsuccessful calibration on the day of measurement and (2) invalid measurement after three attempts as indicated by the device, often due to ankle edema. The root mean square coefficients of variation for short-term reproducibility of calcaneal BUA, SOS, and QUI measurements in our laboratory were 2.8, 0.2, 2.5 % and the corresponding least significant changes were 7.6, 0.7, 6.9 %, respectively, which is in agreement with other laboratories using the same QUS model [24]. Data on medical conditions, medications, and falls were collected at each study visit, and participants were also asked to inform us by telephone of any health changes they experienced during the study. Adverse events (defined as any untoward effects with an onset after baseline or worsening of an existing condition) were recorded by using the Common Terminology Criteria for Adverse Events version 3 from the US National Cancer Institute [25]. Total physical activity levels were estimated at baseline and 12-months from the daily activity metabolic index (AMI; kcal day-1) by using the Minnesota Leisure-Time Physical Activity Questionnaire [26]. Total physical activity levels were further divided into light, moderate, and heavy physical activity levels based on each activity’s metabolic index (light AMI = B4 kcal day-1; moderate AMI = 4.5–5.5 kcal day-1; heavy AMI = C6 kcal day-1). Statistical Analyses Between-group differences in absolute change from baseline (12 months—baseline) in calcaneal QUS outcomes were assessed by using one-way analysis of variance and a priori specified contrasts (90-Hz WBV vs. control, 30-Hz WBV vs. control, 90-Hz WBV vs. 30-Hz WBV, and combined 90- and 30-Hz WBV vs. control). Various multiple imputation models were used for missing QUS outcomes in the intent-to-treat approach

123


[27]. Participants with missing QUS outcomes (due to loss of follow-up or invalid, uncalibrated, or unattained QUS measurement) or those who initiated hormone therapy during the study were excluded from the per protocol approach. Finally, we also excluded participants with \80 % adherence to WBV from the per protocol data, as we hypothesized a priori to observe a greater effect of WBV in more adherent participants. The adherence threshold of 80 % was chosen a priori. Selfreported adherence to calcium and vitamin D supplements and 12-month changes in calcium and vitamin D intakes and physical activity levels were compared between groups using one-way analysis of variance. Sample size calculations for this RCT were based on our pre-specified primary outcome, tibial trabecular volumetric BMD, as outlined in our primary report [12]. All analyses were performed using SAS, version 9.3 (SAS Institute, Cary, NC) with a P \ 0.05 indicating statistical significance.

Results Participants Of the 1,126 subjects initially screened for eligibility, 202 postmenopausal women met our eligibility criteria and were randomly assigned to the 90-Hz WBV (67 participants), 30-Hz WBV (68 participants), or control (67 participants) groups (Fig. 1). Eligible participants were the same postmenopausal women as those examined in our primary report [12]. Relevant baseline characteristics did not significantly differ between groups and are summarized in Table 1. At the end of the trial, QUS outcomes were missing in 25 participants due to drop-out (n = 7), unattained final measurement (n = 4), invalid measurement (n = 7), and unsuccessful calibration (n = 7). Two participants (1 each in the 90-Hz WBV and control groups) started hormone therapy, but returned for the final assessment. In addition, adherence to WBV was not obtained in three participants because their platform’s digital clock malfunctioned, and in five of the participants who dropped out. Most participants were either close to 100 or 0 % adherent and the median adherence based on the total cumulative duration of WBV was 79 % (interquartile range 41–91 %) for the 90-Hz WBV group and 77 % (interquartile range 55–86 %) for the 30-Hz WBV group. Furthermore, selfreported adherence to calcium and vitamin D supplements, 12-month changes in total daily calcium or vitamin D intakes, and 12-month changes in light, moderate, heavy, and total physical activity levels were similar between the three groups (data not shown).


551

Fig. 1 Participants’ progress through trial. WBV whole-body vibration, QUS quantitative ultrasound

Table 1 Baseline characteristics of the study population Baseline characteristics

90-Hz WBV group (n = 67)

30-Hz WBV group (n = 68)

Control group (n = 67)

Age (years), mean (SD)

60.5 (7.0)

59.6 (6.0)

60.8 (5.5)

Years since menopause, mean (SD)

10.2 (8.3)

10.8 (7.3)

10.5 (7.5)

European

55 (82)

48 (70)

54 (81)

Southeast Asian

8 (12)

14 (20)

10 (15)

Other

4 (6)

6 (9)

3 (4)

64.4 (10.6)

62.0 (10.5)

62.4 (9.5)

Ethnicity, n (%)

Mass (kg), mean (SD) Body mass index (kg m-2), mean (SD)

24.9 (4.0)

24.5 (3.6)

24.2 (3.4)

Height (m), mean (SD)

1.61 (0.06)

1.59 (0.06)

1.60 (0.06)

Total daily calcium intake (mg), mean (SD)a

1,538 (677)

Total daily vitamin D intake (IU), mean (SD)a 866 (582) Calcaneal quantitative ultrasound measurements, mean (SD)

1,399 (656)

1,352 (642)

778 (583)

808 (584)

BUA (dB MHz-1)

72.2 (13.0)

75.4 (14.7)

72.0 (12.9)

SOS (m s-1)

1,538.0 (28.3)

1,542.7 (23.9)

1,538.6 (23.5)

QUI

89.2 (16.3)

92.4 (14.8)

89.3 (14.4)

BUA broadband attenuation, QUI quantitative ultrasound index, SD standard deviation, SOS speed of sound, WBV whole-body vibration a

Total daily intake from diet plus patient’s own supplements, prior to providing study supplements

123



Table 2 Intent-to-treat analysis: between-group differences in absolute change from baseline in calcaneal quantitative ultrasound outcomes in the multiple imputation models Modela

1

2

3

4

Variables included in the model

Baseline and 12-month change in calcaneal BUA, SOS, and QUI plus baseline variablesb Baseline and 12-month change in calcaneal BUA, SOS, and QUI plus 12-month change in DXA and HR-pQCT outcomesc plus baseline variablesb 12-month change in calcaneal BUA, SOS and QUI, and in DXA and HR-pQCT outcomesc Baseline and 12-month change in calcaneal BUA, SOS, and QUI

Calcaneal quantitative ultrasound outcome

Between-group difference in absolute change from baseline

P value for pair-wise comparison

90-Hz WBV group– control group

30-Hz WBV group– control group

90-Hz WBV group versus control group

30-Hz WBV group versus control group

BUA (dB MHz-1)

-1.5

-1.8

0.144

0.102

SOS (m s-1)

-1.6

-0.9

0.379

0.620

QUI

-1.3

-1.1

0.220

0.305

BUA (dB MHz-1)

-1.8

-2.1

0.112

0.037

SOS (m s-1)

-1.2

-0.9

0.497

0.585

QUI

-1.2

-1.2

0.226

0.198

BUA (dB MHz-1)

-1.8

-2.3

0.087

0.026

SOS (m s-1)

-1.4

-1.8

0.444

0.361

QUI

-1.3

-1.6

0.210

0.123

-1.4

-1.8

0.245

0.080

SOS (m s )

-1.0

-1.0

0.552

0.569

QUI

-1.0

-1.1

0.326

0.252

BUA (dB MHz-1) -1

BUA broadband attenuation, DXA dual-energy X-ray absorptiometry, HR-pQCT high-resolution peripheral quantitative computed tomography, QUI quantitative ultrasound index, SOS speed of sound, WBV whole-body vibration a

Four multiple imputation models were used for missing QUS outcomes in 25 participants, by using different sets of variables in each model

b

Baseline variables included age, mass, height, body mass index, age at menarche and years since menopause

c

DXA outcomes included BMD at the femoral neck, total hip and lumbar spine, and HR-pQCT outcomes included trabecular BMD thickness, number and separation at the distal tibia; all were examined in our primary outcome report

Calcaneal Quantitative Ultrasound Outcomes After 12 months, we found no improvement in BUA, SOS, or QUI in any of the WBV groups when compared to controls. Instead, in our intent-to-treat analyses (n = 202), we found statistically significant decreases in BUA in the 30-Hz WBV group when compared to the control group in 2 out of 4 multiple imputation models analyses (Table 2). We also found a trend in decrease in BUA in the 90-Hz WBV group compared to controls although not statistically significant. In the per protocol approach (n = 175), upon exclusion of participants with missing QUS measurement (n = 25) and those who initiated hormone therapy during the study (n = 2), statistically significant decreases in BUA were seen in the 30-Hz versus the control group and the 30- and 90-Hz WBV combined groups compared to the control group, but not in the 90-Hz versus the control group (Table 3). When we additionally excluded participants with \80 % adherence to WBV, a significant decrease in BUA was seen in the 90-Hz group versus the control group, but not in the 30-Hz versus the control group (Table 3). Using the per protocol data, the magnitude of WBV treatment effect, defined as the difference in mean BUA change between control group

123

(1.3 dB MHz-1 or 2.0 %) and 90-Hz (–0.4 dB MHz-1 or -0.2 %) or 30-Hz (-0.7 dB MHz-1 or -0.6 %) WBV groups, was -1.7 dB MHz-1 or -2.2 % for 90-Hz participants and -2.1 dB MHz-1 or -2.6 % for 30-Hz participants. Throughout all our analyses, the decrease in BUA did not significantly differ between 90- and 30-Hz WBV groups. Although, SOS and QUI showed a decreasing trend in the 30- and 90-Hz WBV groups as compared to the control group in all analyses, none were statistically significant (Tables 2 and 3). Adverse Events Several women in the 90- and 30-Hz WBV groups spontaneously reported minor foot-related problems that they attributed to WBV therapy. Some complained of plantar foot pain (two in 90-Hz and one in 30-Hz WBV group) lasting throughout the day, while others reported foot numbness (two in each 90- and 30-Hz WBV groups) or toe cramping (two in 90-Hz WBV group) that lasted briefly during or just after a WBV session. As summarized in our primary outcome report, no serious adverse events were caused by WBV, and quantitative analyses of various

90-Hz WBV group

30-Hz WBV group

Control group

Mean absolute change from baseline to 12 months (95 % CI)

P value for one-way ANOVA 90-Hz WBV group–control group

-1.0 (-2.3 to 0.4)

(-2.4 to 0.8)

(-4.1 to 0.7)

(-4.5 to 1.3)

-0.8

-1.7

-1.6

-0.7 (-2.3 to 0.8)

-0.4

(-1.9 to 1.2)

1.3

(-0.9 to 1.6)

0.4

(-2.7 to 1.9)

-0.4

(0.0 to 2.6)

-0.7 (-2.4 to 1.0)

(-4.9 to 3.5)

-0.8 (-2.9 to 1.3)

-0.9 (-4.0 to 2.2)

-0.7

-0.7 (-2.6 to 1.1)

-1.3

(-3.0 to 0.4)

1.3

0.4 (-0.9 to 1.6)

(-2.7 to 1.9)

-0.4

(0.0 to 2.6)

0.486

0.971

0.032

0.344

0.729

0.107

-1.7

-1.2 (-3.4 to 1.0)

(-4.4 to 3.8)

-0.3

(-4.8 to -0.5)

-2.6

(-3.2 to 0.8)

-1.2

(-4.8 to 2.4)

-1.2

(-3.8 to 0.3)

BUA broadband attenuation, QUI quantitative ultrasound index, SOS speed of sound, WBV whole-body vibration

Per protocol analysis excluded participants with missing outcomes and who started on hormone therapy

QUI

SOS (m s-1)

BUA (dB MHz-1)

C80 % adherent to WBV (90 Hz, n = 29; 30 Hz, n = 27; CON, n = 57)

QUI

SOS (m s-1)

BUA (dB MHz-1)

-1.0 (-3.3 to 1.2)

(-4.7 to 3.7)

-0.5

(-4.2 to 0.1)

-2.0

(-3.3 to 0.6)

-1.3

(-4.8 to 2.3)

-1.3

(-4.1 to 0.0)

-2.1

30-Hz WBV group–control group

-0.1 (-2.7 to 2.4)

(-4.6 to 5.0)

0.2

(-3.1 to 2.0)

-0.6

(-1.8 to 2.1)

0.1

(-3.5 to 3.6)

0.0

(-1.7 to 2.4)

0.4

90-Hz WBV group –30-Hz WBV group

-1.1 (-2.9 to 0.7)

(-3.8 to 3.0)

-0.4

(-4.1 to -0.6)

-2.3

(-3.0 to 0.4)

-1.3

(-4.3 to 1.8)

-1.2

(-3.7 to -0.1)

-1.9

Combined 90- and 30 Hz groups— control group

Between-group difference in absolute change from baseline (95 % CI)

Excluded participants with missing QUS outcomes and initiated hormone therapy (90 Hz, n = 58; 30 Hz, n = 60; CON, n = 57)

Calcaneal quantitative ultrasound outcome

Table 3 Per protocol analysis: absolute changes from baseline in calcaneal quantitative ultrasound outcomes

Author's personal copy

L. Slatkovska et al.: Effect of WBV on Calcaneal QUS Measurements 553

123


adverse events, including those involving the lower extremities, revealed no significant between-group differences [12].

Discussion In our 12-month RCT of 202 postmenopausal women, WBV at 90- and 30-Hz did not improve BUA, SOS, or QUI. Instead, a statistically significant decrease in calcaneal BUA was found in participants who received 90- or 30-Hz WBV therapy at 0.3 g, compared to no WBV therapy in some, albeit not all of our analyses. This negative effect of WBV observed in our trial was unexpected and its clinical relevance is uncertain, for the decrease is small (90-Hz: -1.7 dB MHz-1 or -2.2 %; 30-Hz: -2.1 dB MHz-1 or -2.6 %). To put this into perspective for the Hologic Sahara device, the least significant change typically obtained for calcaneal BUA is 7.5–13.9 % for an individual [24], and the absolute difference in mean calcaneal BUA observed between postmenopausal women with and without osteoporotic fracture is -19.3 db MHz-1 [28]. We had expected an increase in calcaneal BUA, as the beneficial effects of WBV on bones, particularly those involving the trabecular tissue at weight-bearing sites located close to the oscillating platform, were previously reported in children and animal models [17–19, 29, 30]. Furthermore, in an RCT of postmenopausal women, a smaller decrease in calcaneal BUA was observed in response to twice-weekly 15-min sessions of WBV at 0.3 g and 30-Hz (-0.8 %) compared to no WBV (-3.1 %), and an improvement in calcaneal BUA was seen in response to twice-weekly 6-minute sessions of WBV at C1 g and 12.5 Hz (?3.4 %); however, these betweengroup differences were not statistically significant [7]. Though our results differ from this previous RCT, the discrepancy may be because we examined more participants (n = 202 vs. n = 47) and asked the participants to stand on the WBV platform more frequently (daily versus twice a week) over a longer period of follow-up (12 vs. 8 months) [7, 12]. We also examined calcaneal SOS and QUI, and thus offer additional information about potential WBV effect on the calcaneus [7, 12]. QUI is a mathematical sum of BUA and SOS and may provide more clinical insight due to its composite nature [24]. While bone acoustic properties are primarily influenced by the mineralized bone matrix and both SOS and BUA correspond to BMD changes, BUA is also thought to be especially influenced by the trabecular microarchitecture [4, 5, 31–33]. This is possibly because as sound waves pass through the bone they may become scattered and attenuated by the trabecular structure. Therefore, when a small

123


but significant effect is found in BUA alone, and not in SOS and QUI, it may be interpreted as trabecular architecture being affected in the absence of bone density changes. Finally, compared to the other trial, we observed increases in QUS parameters in the control group and not decreases as would be expected in a prospective follow-up of postmenopausal women [7, 12]. This increase was possibly due to vitamin D supplementation in our study [24]. However, similar increases would be expected in the WBV groups, since all participant were provided with vitamin D supplements as part of the trial. Several limitations existed in our trial. First, sample size and power calculations were not based on calcaneal QUS outcomes, because they were collected as secondary endpoints. In addition, women and the outcome assessor were not blinded to control intervention, since sham WBV was not provided due to limited funding and a lack of effective masking of true WBV by a sham platform [11]. Also, since the control group participants were not instructed to stand still for 20 consecutive minutes every day, it is unclear whether just standing on the vibration platform, regardless of WBV, could have contributed to the between-group differences in 12-month changes in BUA. Further, QUS outcomes were missing in 12 % of the participants and required replacements by the use of multiple imputation models in our primary analysis. Finally, numerous between-group comparisons were performed during various statistical approaches, thus increasing the likelihood of chance findings. However, our finding of a decrease in BUA with WBV was consistent across several analyses and the number of missing outcomes was similar between groups. In spite of these limitations, our results challenge the existing safety data of WBV in postmenopausal population if used long-term, and call for future research to consider this potential adverse effect and confirm our preliminary findings. If this negative effect is real, several mechanisms may explain it. First, the small decrease in calcaneal BUA in women receiving 90- or 30-Hz WBV may be due to minor bone damage caused by 20 consecutive minutes of daily WBV for 12 months, with an insufficient rest period between treatments [34, 35]. When women were standing on the WBV platform, their heel bones were hit by a small force (*18 N) of the platform accelerating upwards, 30–90 times per second consecutively for 20 min (i.e., 36,000–108,000 compressions), where only a thin layer of soft tissue but no major joints provided cushioning. Since the calcaneus is made up of mostly trabecular tissue and BUA decreased more than SOS in the 90- and 30-Hz WBV groups, there may be minute damage to the trabecular structure, similar to stress fractures which can occur with minimal but frequently repeated ground reaction forces such as walking. Second, perhaps the regulatory


mechanisms involving either bone fluid flow or skeletal muscle activation, which may be responsible for increasing bone formation in response to WBV, were insufficient to compensate for small structural damage occurring at the calcaneus [36, 37]. This may be especially true in postmenopausal women, since they experience slower bone formation than resorption due to menopause, as compared to, for example, children and adolescents whose bone formation surpasses resorption [38]. Finally, since QUS measurements can be affected by changes in the heel soft tissue [24, 39], such as thickness or composition, these variables should also be collected in future research of WBV. It is plausible that at least part of the decrease in calcaneal BUA observed in our trial may have occurred due to heel soft tissue damage caused by WBV, rather than bone damage [24, 40, 41]. Several footrelated problems, such as pain and numbness, were spontaneously reported by the 90- and 30-Hz participants and attributed to WBV. In occupational settings, where drilling (with much higher magnitudes of vibration) is involved, prolonged exposures to vibration of the hands and feet were also found to cause injuries to the muscles, vasculature, and connective tissues [42, 43]. In conclusion, we found no beneficial effect of WBV on calcaneal QUS measurements in community-dwelling postmenopausal women receiving 0.3 WBV at 90 or 30 Hz, but instead a small but statistically significant decrease in calcaneal BUA in two out of four multiple imputation models, per protocol analysis and subgroup analysis of 80 % adherent participants. This potential negative effect needs to be confirmed in future research. In the absence of any clear beneficial bone effects at hip, spine, distal tibia, and calcaneus, we do not recommend WBV therapy at this time for the prevention of bone loss in postmenopausal women with low bone density. Acknowledgments The authors thank the women who volunteered their time and participated in this trial. We also thank OsTek Orthopaedics Inc. for their assistance in obtaining the platforms. In addition, we thank Alice Demaras, Diana Yau, Claudia Chan, Gail Jefferson, and Farrah Ahmed and our research volunteers and workstudy students who helped with various aspects of the study. Conflicts of Interest Please note that Lubomira Slatkovska, Joseph Beyene, Shabbir M. H. Alibhai, Queenie Wong, Qazi Z. Sohail, and Angela M. Cheung declare that they have no conflicts of interest. All authors made substantial contributions to the intellectual content of the paper. A peer-reviewed grant from the Physicians’ Services Incorporated Foundation funded this trial. Juvent Inc. supplied the WBV platforms and Jamieson Laboratories provided calcium and vitamin D supplements. None of these sources were involved in the study design, conduct, analysis, interpretation of the data, preparation of this manuscript, or decision to submit the manuscript for publication. Human and Animal Rights and Informed Consent All procedures performed in studies involving human participants were in

555 accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study.

References 1. Slatkovska L, Alibhai SM, Beyene J, Cheung AM (2010) Effect of whole-body vibration on BMD: a systematic review and metaanalysis. Osteoporos Int 21(12):1969–1980. doi:10.1007/s00198010-1228-z 2. Kiiski J, Heinonen A, Ja¨rvinen TL, Kannus P, Sieva¨nen H (2008) Transmission of vertical whole body vibration to the human body. J Bone Miner Res 23(8):1318–1325. doi:10.1359/jbmr. 080315 3. Rubin C, Pope M, Fritton JC, Magnusson M, Hansson T, McLeod K (2003) Transmissibility of 15-Hertz to 35-Hertz vibrations to the human hip and lumbar spine: determining the physiologic feasibility of delivering low-level anabolic mechanical stimuli to skeletal regions at greatest risk of fracture because of osteoporosis. Spine 28(23):2621–2627 4. Guglielmi G, Scalzo G, de Terlizzi F, Peh WC (2010) Quantitative ultrasound in osteoporosis and bone metabolism pathologies. Radiol Clin North Am 48(3):577–588. doi:10.1016/j.rcl. 2010.02.013 5. Sasso M, Haı¨at G, Yamato Y, Naili S, Matsukawa M (2008) Dependence of ultrasonic attenuation on bone mass and microstructure in bovine cortical bone. J Biomech 41(2):347–355 6. Verschueren SM, Bogaerts A, Delecluse C et al (2011) The effects of whole-body vibration training and vitamin D supplementation on muscle strength, muscle mass, and bone density in institutionalized elderly women: a 6-month randomized, controlled trial. J Bone Miner Res 26(1):42–49. doi:10.1002/jbmr. 181 7. Beck BR, Norling TL (2010) The effect of 8 mos of twice-weekly low- or higher intensity whole body vibration on risk factors for postmenopausal hip fracture. Am J Phys Med Rehabil 89(12):997–1009. doi:10.1097/PHM.0b013e3181f71063 8. Gusi N, Raimundo A, Leal A (2006) Low-frequency vibratory exercise reduces the risk of bone fracture more than walking: a randomized controlled trial. BMC Musculoskelet Disord 7:92 9. Iwamoto J, Takeda T, Sato Y, Uzawa M (2005) Effect of whole-body vibration exercise on lumbar bone mineral density, bone turnover, and chronic back pain in post-menopausal osteoporotic women treated with alendronate. Aging Clin Exp Res 17(2):157–163 10. Verschueren SM, Roelants M, Delecluse C, Swinnen S, Vanderschueren D, Boonen S (2004) Effect of 6-month whole body vibration training on hip density, muscle strength, and postural control in postmenopausal women: a randomized controlled pilot study. J Bone Miner Res 19(3):352–359 11. Rubin C, Recker R, Cullen D, Ryaby J, McCabe J, McLeod K (2004) 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 19(3):343–351 12. Slatkovska L, Alibhai SM, Beyene J, Hu H, Demaras A, Cheung AM (2011) Effect of 12 months of whole-body vibration therapy on bone density and structure in postmenopausal women: a randomized trial. Ann Intern Med 155(10):668–679, W205. Erratum in: Ann Intern Med. 2011 Dec 20;155(12):860. doi: 10.7326/ 0003-4819-155-10-201111150-00005 13. Russo CR, Lauretani F, Bandinelli S, Bartali B, Cavazzini C, Guralnik JM, Ferrucci L (2003) High-frequency vibration

123


14.

15.

16.

17.

18.

19.

20. 21.

22.

23.

24.

25.

26. 27. 28.

L. Slatkovska et al.: Effect of WBV on Calcaneal QUS Measurements training increases muscle power in postmenopausal women. Arch Phys Med Rehabil 84(12):1854–1857 Flieger J, Karachalios T, Khaldi L, Raptou P, Lyritis G (1998) Mechanical stimulation in the form of vibration prevents postmenopausal bone loss in ovariectomized rats. Calcif Tissue Int 63(6):510–514 Judex S, Donahue LR, Rubin C (2002) Genetic predisposition to low bone mass is paralleled by an enhanced sensitivity to signals anabolic to the skeleton. FASEB J 16(10):1280–1282 Rauch F, Sievanen H, Boonen S et al (2010) Reporting wholebody vibration intervention studies: recommendations of the International Society of Musculoskeletal and Neuronal Interactions. J Musculoskelet Neuronal Interact 10(3):193–198 Garman R, Gaudette G, Donahue LR, Rubin C, Judex S (2007) Low-level accelerations applied in the absence of weight bearing can enhance trabecular bone formation. J Orthop Res 25(6):732–740 Judex S, Lei X, Han D, Rubin C (2007) 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 40(6):1333–1339 Rubin C, Turner AS, Bain S, Mallinckrodt C, McLeod K (2001) Anabolism. Low mechanical signals strengthen long bones. Nature 412(6847):603–604 Griffin MJ (1998) Predicting the hazards of whole-body vibration—considerations of a standard. Ind Health 36(2):83–91 Hung A, Hamidi M, Riazantseva E, Thompson L, Tile L, Tomlinson G, Stewart B, Cheung AM (2011) Validation of a calcium assessment tool in postmenopausal Canadian women. Maturitas 69(2):168–172. doi:10.1016/j.maturitas.2011.02.016 Rubin C, Turner AS, Mallinckrodt C, Jerome C, McLeod K, Bain S (2002) Mechanical strain, induced noninvasively in the highfrequency domain, is anabolic to cancellous bone, but not cortical bone. Bone 30(3):445–452 Rubin C, Judex S, Qin YX (2006) Low-level mechanical signals and their potential as a non-pharmacological intervention for osteoporosis. Age Ageing 35(Suppl 2):ii32–ii36 Krieg MA, Barkmann R, Gonnelli S et al (2008) Quantitative ultrasound in the management of osteoporosis: the 2007 ISCD Official Positions. J Clin Densitom 11(1):163–187. doi:10.1016/j. jocd.2007 Common Terminology Criteria for Adverse Events, Version 3.0 (2006) Cancer Therapy Evaluation Program. http://ctep.cancer. gov/protocolDevelopment/electronic_applications/docs/ctcaev3. pdf. Accessed 15 Sept 2006 Wilson HW (1997) Minnesota leisure-time physical activity questionnaire. Med Sci Sports Exerc 29:S62–S72 Rubin DB (1996) Multiple imputation After 18? years. J Am Statist Assoc 91:473–489 Hans D, Genton L, Allaoua S, Pichard C, Slosman DO (2003) Hip fracture discrimination study: QUS of the radius and the calcaneum. J Clin Densitom 6(2):163–172

123

29. Prisby RD, Lafage-Proust MH, Malaval L, Belli A, Vico L (2008) Effects of whole body vibration on the skeleton and other organ systems in man and animal models: what we know and what we need to know. Ageing Res Rev 7(4):319–329. doi:10.1016/j.arr. 2008.07.004 30. Ward K, Alsop C, Caulton J, Rubin C, Adams J, Mughal Z (2004) Low magnitude mechanical loading is osteogenic in children with disabling conditions. J Bone Miner Res 19(3):360–369 31. Haı¨at G, Padilla F, Peyrin F, Laugier P (2007) Variation of ultrasonic parameters with microstructure and material properties of trabecular bone: a 3D model simulation. J Bone Miner Res 22(5):665–674 32. Njeh CF, Fuerst T, Diessel E, Genant HK (2001) Is quantitative ultrasound dependent on bone structure? A reflection. Osteoporos Int 12(1):1–15 33. Lin W, Xia Y, Qin YX (2009) Characterization of the trabecular bone structure using frequency modulated ultrasound pulse. J Acoust Soc Am 125(6):4071–4077. doi:10.1121/1.3126993 34. Ma R, Zhu D, Gong H et al (2012) High-frequency and lowmagnitude whole body vibration with rest days is more effective in improving skeletal micro-morphology and biomechanical properties in ovariectomised rodents. Hip Int 22(2):218–226. doi:10.5301/HIP.2012.9033 35. Rapillard L, Charlebois M, Zysset PK (2006) Compressive fatigue behavior of human vertebral trabecular bone. J Biomech 39(11):2133–2139 36. Judex S, Rubin CT (2010) Is bone formation induced by highfrequency mechanical signals modulated by muscle activity? J Musculoskelet Neuronal Interact 10(1):3–11 37. Ozcivici E, Luu YK, Adler B, Qin YX, Rubin J, Judex S, Rubin CT (2010) Mechanical signals as anabolic agents in bone. Nat Rev Rheumatol 6(1):50–59. doi:10.1038/nrrheum.2009.239 38. Reid IR (2006) Menopause. In: Favus MJ (ed) Primer on the Metabolic bone diseases and disorders of mineral metabolism. American Society for Bone and Mineral Research, Washington, DC 39. Lewiecki EM, Richmond B, Miller PD (2006) Uses and misuses of quantitative ultrasonography in managing osteoporosis. Cleve Clin J Med 73(8):742–766, 749–752 40. Chappard C, Camus E, Lefebvre F, Guillot G, Bittoun J, Berger G, Laugier P (2000) Evaluation of error bounds on calcaneal speed of sound caused by surrounding soft tissue. J Clin Densitom 3(2):121–131 41. Ikeda Y, Iki M (2004) Precision control and seasonal variations in quantitative ultrasound measurement of the calcaneus. J Bone Miner Metab 22(6):588–593 42. Sakakibara H, Hashiguchi T, Furuta M, Kondo T, Miyao M, Yamada S (1991) Circulatory disturbances of the foot in vibration syndrome. Int Arch Occup Environ Health 63(2):145–148 43. Thompson AM, House R, Krajnak K, Eger T (2010) Vibrationwhite foot: a case report. Occup Med (Lond) 60(7):572–574