The transmission of a mechanical vibratory stimulus to the muscle-tendon structures has been promoted over the past 2 decades as an effective method for improving athletic performance (11), neuromuscular function (22), balance (21), and flexibility (10). It has been shown that vibration excites the primary spindle endings (2) and produces a tonic vibration reflex in the exposed muscle (5). Vibration can be applied directly to the targeted tendon (12) or muscle belly (9) or indirectly through a part of the body in either the lower (19) or upper limbs (11). Although vibratory platforms have become the most common tools to indirectly stimulate lower-limb muscles and tendons via the feet (19), some systems have been specifically designed to apply an indirect vibratory stimulus to the upper-body muscles via the hands (3,10,14).
At present, 3 upper-limb vibration systems have been successful in producing gains in neuromuscular performance (3,11,14,16). In the first system, the vibratory stimulus is produced by an electromotor device (11) or electromagnetic actuator (13) and transmitted to a handle through an isotonic cable pulley resistance system or through belts. The second systems involves vibratory dumbbells (1,3), which are commercially available and are capable of vibrating in the 0–30 Hz range with an amplitude of 2–3 mm around a horizontal axis and a maximum load of 5 kg. The third system consists of 1 (16) or 2 vibrating motors (14) that are fixed directly on a rigid bar through which the vibratory stimulus is transmitted. The use of vibratory bars (VBs) has recently been proposed as an interesting tool for improving neuromuscular performance (14,16) because damping is low at the grip level (the vibration source is mounted directly on the bar), contrary to vibrating cables/belts, and because VBs seem capable of providing effective vibration accelerations, frequencies, and displacements under loading conditions that clearly exceed the 5 kg of the vibratory dumbbells (14,16).
Recent methodological studies have highlighted their mechanical limitations of commercially available vibratory platforms (15,17). It is indeed possible that the frequency and amplitude generated by a vibratory platform differ from the preset values, in particular when the subject is performing dynamic exercise and/or is using additional weights (18). Only 1 study has explored the mechanical performance of a VB for upper-body exercise (14), but exclusively in a single loading condition (15 kg) and for 1 type of exercise (bench press). One of the most striking findings of that previous study was that increasing the preset frequency above 30 Hz did not result in a large increase in VB frequency and acceleration. Moreover, inconsistent information was provided about vibration directions with respect to the hand-arm system (16,22) because vibration displacement and its direction are 2 of the main factors that affect human response (7). Thus, considering the lack of knowledge on upper-limb vibration exercise, we aimed to design a novel VB that would allow to exercise with a consistent vibration superimposed under different loading conditions, and with a wider vibration frequency range compared with similar devices (14).
The aim of the present study was to analyze the influence of different vibration frequencies, loading conditions, and exercise types on the mechanical behavior of a novel VB, as assessed with an accelerometer mounted on the VB. Based on previous whole-body vibration research (15,17), we hypothesized the following: (a) increasing vibration frequency (from 20 to 50 Hz) would result in a progressive increase in VB acceleration and frequency with minor effects on displacement; (b) adding weight to the VB (almost up to the maximum sustainable load [MSL]) would have marginal or no effect on VB frequency, acceleration, and displacement; and (c) altering the type of static exercise (pulling vs. pushing) would not affect VB frequency, acceleration, and displacement. If these hypotheses are confirmed, the validity of the novel VB could be established.
Materials and Methods
Experimental Approach to the Problem
A repeated-measures (single session) design consisting of 18 different measurements (presented in a random order) was adopted to analyze the influence of 3 different loading conditions (20, 50, and 80% of the MSL), 3 preset vibration frequencies (f in) (20, 35, and 50 Hz), and 2 types of static exercise (lying row: pulling and bench press: pushing) on the mechanical behavior of the VB. For each of the 18 measurements (3 loads × 3 f in × 2 exercises), the subjects were asked to hold the VB as steadily as possible for 10 seconds with an elbow joint angle of 90°. A 3-axis 10g accelerometer was fixed at 1 extremity of the VB to measure 3-dimensional accelerations. The dependent variables were vibration root mean square (RMS) acceleration (a RMS), peak-to-peak displacement (D), and frequency (f out).
Fourteen healthy men (mean ± SD: age, 25 ± 5 years; height, 179 ± 5 cm; body weight, 74 ± 6 kg) volunteered to participate in the study. They were physical education students with previous experience in supervised resistance training (at least 2 sessions per week during the 2 years preceding the study), but not in vibration exercise. Excluding criteria were joint implants, recent operative wounds, and history of severe musculoskeletal problems in the upper body. Subjects were required to abstain from moderate-high intensity exercise, alcohol, and caffeine for 12 hours before testing. The study was conducted in accordance with the Declaration of Helsinki and was approved by the local ethics committee. Written informed consent was obtained from all participants before inclusion.
The VB consists of a 3-phase electric vibrating motor (Italvibras M3/45-S02, 50 Hz, Fiorano Modenese, Italy) fixed in a cylindrical central body with 2 bars welded to its lateral sheets. The outer diameter of the central body of the VB is 150 mm, and the diameter of the left/right bars is 30 mm. The length and mass of the VB are 1.10 m and 8.8 kg, respectively. The grip area was covered with 1-mm thick nonslip grip foam. An inverter (Omron Sysdrive 3G3JV, single phase 230 V, 0.55 kW) was used to reduce the speed of the motor (3000g) according to the required f in. The VB was developed to induce rotational vibrations around the mediolateral axis (X-axis) by vibrating in both vertical (Z-) and anteroposterior (Y-) axes (Figure 1). The 2 extremities of the VB were attached by means of carabineers to the guide rails of a pneumatic resistance system (Keiser Half Rack, Fresno, CA, USA), which allowed pneumatic pulleys to roll seamlessly with the movement of the bar (Figure 1). The pneumatic system allowed adjustment of the external vertical load in the range 0.1–100 kg, with 0.1 kg steps.
For measurements during the lying row (pulling) exercise, subjects lay prone on a horizontal bench (width 0.28 m; length 1.90 m) elevated approximately 0.50 m over another horizontal bench (Figure 1A). For measurements during the bench press (pushing) exercise, subjects lay supine on the lower horizontal bench (seat height around 0.46 m) with their knees flexed approximately at an angle of 90° (Figure 1B).
Assessment of Maximum Sustainable Load
The experimental session started with a standardized warm-up consisting of 5 minutes of cycling on an ergometer without resistance and arm/shoulder mobilization exercises. Before MSL assessment, subjects were also familiarized with the protocol and the equipment. During the assessment of MSL, they were asked to maintain the exercise position as steadily as possible for 5 seconds with an elbow joint angle of 90°, as verified with an electronic goniometer (SG110; Biometrics, United Kingdom). Previously documented training experience was used as guidance for selecting the initial test weight. The load was progressively increased by 10 kg with each successful attempt until the subject was unable to maintain the isometric position for at least 5 seconds. Then, the load was decreased 5 kg and a last attempt was made. The load corresponding to the last successful attempt was considered as the MSL. Subjects self-selected the duration of rest periods, with a minimum of 5 minutes between attempts and 10 minutes between exercises. Pulling and pushing MSL were evaluated in a random order between subjects. Subjects were asked to hold the VB with a pronated grip and not to modify the wrist position during the assessments. They were also encouraged to maintain a comfortable and constant grip force throughout the procedure. Before and after each test trial, the VB was maintained by 2 experimenters to avoid fatigue (Figure 1A). No verbal encouragement was provided throughout testing.
At least 15 minutes after MSL assessment, subjects were asked to hold the VB as steadily as possible for 10 seconds with an elbow joint angle of 90° (Figure 1) and with loads of 20, 50, and 80% MSL and f in of 20, 35, and 50 Hz. Exercises were assigned with the same order established during the assessment of MSL. The order in which the loads and vibration frequencies were assigned were randomized between subjects. The duration of rest phases was self-selected on an individual basis, with a minimum of 3 minutes between measurements and 10 minutes between exercises. No verbal encouragement was provided throughout testing. Subjects were requested to report any unusual symptoms (e.g., discomfort, queasiness, tingling) during vibration exposure.
Three-Dimensional Acceleration Measurement
To measure the acceleration in the X-, Y- and Z-axes, a 3-axis 10g accelerometer (Mega Electronics Ltd., Kuopio, Finland) was fixed with adhesive tape to one extremity of the VB (Figure 1), close to the right hand of the subjects, and connected to a 14-bit AD converter (ME6000 Biomonitor, Mega Electronics Ltd.). The signals were sampled with a frequency of 2000 Hz. Static calibration of the accelerometer was achieved by placing the transducer on a horizontal surface, corresponding to a value of 1.0g. The transducer was then rotated by 180° angle to give the value of −1.0g. This process was conducted for each axis (X, Y, and Z) taking a calibration reading at each axis point, that is, 2 measurement orientations giving a total of 6 measured data points. During the measurements, the vertical alignment of the accelerometer was visually inspected to avoid undesirable tilts. An offset equal to the earth's gravity was subtracted from all acceleration signals in a vertical direction to make all signals start at 0 m·s−2.
The middle 5-second portion of each measurement phase was considered for further analysis. Data processing and analysis routines were written using Matlab (The MathWorks Version 184.108.40.2064 [R2010b]; Natick, MA, USA). The RMS estimation of the signals provided by the accelerometer was computed in a bandwidth of 16–200 Hz so as to consider only vibration-induced VB accelerations and to exclude the fluctuations in limb position that are mainly confined between 1 and 15 Hz (6). To achieve this, the power spectrum of each axis was estimated using a windowed periodogram (20) with a Hanning window. The RMS value was computed as the square root of the integration of the power spectrum inside the bandwidth of 16–200 Hz. Moreover, the corresponding frequency component (f out) was estimated as the peak frequency of the power spectrum in the above-mentioned bandwidth.
In addition, after visual inspection, we fitted a sine function through all test acceleration signals (using the method of nonlinear least squares and the algorithm of the trust region), and R 2 was calculated to quantify how successful the fit was in explaining the variation of the high-frequency vibration. Overall, the mean ± SD R 2 was 0.84 ± 0.11. We assumed sinusoidal VB motions across conditions and calculated the VB peak-to-peak displacement (D) at the location of the accelerometer using the following set of equations (18):
where, a RMS is the average rate of change in velocity of the acceleration signal, a peak is the maximal rate of change in velocity of the acceleration signal, and f out is the frequency component (Hz).
As the VB induced rotational vibrations around the mediolateral X-axis by vibrating in the vertical (Z-) and the anteroposterior (Y-) axes (Figure 1), the vector of the a RMS and D was computed by quadratic combination of the values of the Y- and Z-axes as follows:
where b Y and b Z represent the a RMS or D values for Y- and Z-axes, respectively. The vector value represents the rotational vibration magnitude provided by the VB.
Kolgomorov-Smirnov tests confirmed that all data were normally distributed. For all measurements because f out did not differ significantly between Z- and Y-axis (as verified with paired t-tests), f out data were collapsed across axes. Three-way analyses of variance for repeated measures were used to assess the effect of load (20, 50, and 80% MSL), f in (20, 35, and 50 Hz), and exercise type (pulling and pushing) on a RMS, D, and f out. When main effects were significant, Bonferroni tests for multiple comparisons were applied. Percent differences in a RMS, D, and f out between the different conditions were also calculated. The level of significance was set at p ≤ 0.016 (Bonferroni correction). Data are presented as means and SDs. Statistical analyses were conducted using SPSS v13 (SPSS Inc., Chicago, IL, USA).
None of the subjects reported any unusual symptoms as a result of vibration exposure. Mean ± SD MSL values were 86 ± 8 kg and 79 ± 14 kg for pulling and pushing exercises, respectively. Figure 2 shows an example of VB acceleration in X-, Y-, and Z-axes with the f in preset at 35 Hz. The vibration acceleration was mainly in the vertical Z-axis and the anteroposterior Y-axis and behaved approximately as a sine wave. Mediolateral X-axis a RMS were less than 0.11g at any f in. Therefore, we only present the main results for Y- and Z-axes.
Overall, a RMS ranged from 0.48 ± 0.01 to 2.74 ± 0.24g and from 0.36 ± 0.02 to 2.63 ± 0.43g for the Y- and Z-axis, respectively. Similarly, D ranged from 1.00 ± 0.10 to 1.14 ± 0.12 mm and from 0.72 ± 0.06 to 0.99 ± 0.15 mm for the Y- and Z-axes, respectively. Thus, a RMS and D data were lower for the loaded Z-axis than for Y-axis, with an average difference of 18% (range, 4–37%).
Figure 3 shows a RMS, D, and f out of the VB over the 3 f in conditions. As expected, increasing vibration frequency from 20 to 50 Hz resulted in a progressive increase in a RMS and f out, as indicated by a significant main effect of f in (both p ≤ 0.001). Post hoc tests revealed significant differences on a RMS and f out between the 3 f in conditions. A significant main effect of f in was also observed for D (p ≤ 0.001), and this latter was significantly lower at 20 and 35 Hz compared with 50 Hz (mean differences lower than 5.9%). Mean f out were 9.0 ± 0.6%, 12.0 ± 1.5%, and 16.9 ± 1.7% lower than the preset f in values of 20, 35, and 50 Hz, respectively. Thus, an f in of 20 Hz resulted in an f out of around 18 Hz; an f in of 35 Hz resulted in an f out close to 31 Hz; and an f in of 50 Hz resulted in an f out of approximately 42 Hz.
Figure 4 shows a RMS, D, and f out of the VB over the 3 loading conditions. A significant main effect of load was observed for a RMS (p = 0.014) and f out (p = 0.002), but not for D. Post hoc tests showed that a RMS at 50% MSL was significantly lower compared with 20% (p = 0.007) and 80% MSL (p = 0.009), with mean differences lower than 4.2%. In a similar way, f out was significantly higher at 20% MSL than at 50% (p = 0.003) and 80% MSL (p = 0.010), with mean differences lower than 1.7%.
Figure 5 shows a RMS, D, and f out of the VB for pulling and pushing exercises. No main effect of exercise was detected for a RMS (p = 0.477), D (p = 0.046), and f out (p = 0.059).
The main findings of this study were that (a) increasing vibration frequency from 20 to 50 Hz resulted in a progressive increase of VB a RMS and f out with smaller variations of D; (b) adding weight to the VB (progressive overload from 20 to 80% MSL) did not affect D and minimally but significantly affected a RMS and f out values; and (c) altering the type of exercise (pushing vs. pulling) had no significant effect on a RMS, D, and f out.
As expected, increasing vibration frequency from 20 to 50 Hz resulted in a progressive increase of VB a RMS and f out with small variations of D. The VB showed a higher and more consistent mechanical output performance compared with a previous custom built VB (14) that was unable to provide a large increase in frequency and acceleration when the preset frequency increased above 30 Hz. The VB used in the present study achieved a f out of around 42 Hz, which is clearly higher than the ∼30 Hz provided by the bar developed by Moras et al. (14). These authors concluded that the limited mechanical performance of their VB was due to structural and functional limitations in the manufacturing system. The present results and those of Moras et al. (14) suggest that during vibration exercise with VB, and probably with other devices designed to apply an indirect vibratory stimulus to the upper-body muscles via the hands, differences could be expected between the preset f in and the f out measured at the nearest place of transmission to the hands. In fact, the present VB provided f out of around 18, 31, and 42 Hz when preset f in of 20, 35, and 50 Hz were selected. However, it is interesting to emphasize that during pilot testing under unloaded conditions (pulling exercise), the f out provided by the present VB was only 6% lower than the preset f in. This means that most of the differences between f out and f in were produced by overloading the VB up to low loads of 20% MSL (range, 12–20 kg), without major changes despite after overloading the VB up to heavy loading conditions of 80% MSL (average, 67 kg), as revealed by the present results. These inconsistencies between f in and f out seem also to affect several commercially available vibratory platforms, in which the f out were normally within 6% of the preset f in values under both unloaded and loaded conditions (15,17).
Increasing vibration frequency from 20 to 50 Hz was associated to a small though significant increase of D (∼0.08 mm). Previous studies focusing on professional synchronous vibratory platforms showed larger variations on the maximum vertical displacement when the preset f in was increased from 25 to 50 Hz (a reduction of ∼0.2 mm) in both low (1.2 mm peak-to-peak) and high (2.2 mm peak-to-peak) amplitude modes (15), or from 20 to 55 Hz (a reduction of ∼0.9 mm) with respect to the theoretical peak-to-peak displacement of 4 mm (17). In the same way, the VB used by Moras et al. (14) showed a reduction on the vertical peak-to-peak displacement ranging from 6.1 to 4.6 mm when the preset f in was increased from 20 to 50 Hz. Thus, from a practical point of view, the small variation of D with the increase of the preset f in could be considered as a normal mechanical behavior of the VB.
Adding weight to the VB up to 80% MSL did not affect D and minimally affected a RMS and f out values. This proves, at least in part, the validity of our novel VB for upper-body exercise because its mechanical performance was little influenced by overloading. To our knowledge, this is the first attempt to evaluate the impact of adding weight to a VB or to any other device for upper-body vibration exercise. Thus, no clear comparisons are possible with previous studies. It is nevertheless worth mentioning that the above-mentioned studies of Pel et al. (15) and Preatoni et al. (17) highlighted the limitations of vibratory platforms in the presence of additional load. Concretely, Pel et al. (15) found that vertical acceleration of 2 professional synchronous vibratory platforms was reduced by 10–12% with conventional body weights of 62 and 81 kg. They also reported that when volunteers of 62, 81, and 100 kg loaded a home-use plate capable of inducing rotational vibration, the decline in horizontal vibrations was as high as 20% and 40% for the X- and Y-axis, respectively. Differences between theoretical and actual peak-to-peak displacement and peak acceleration were above 20% in the study of Preatoni et al. (17) and increased with increasing external load and frequency. Our results showed a slight but significant decline (4.2%) of αRMS only when loading the VB from 20 to 50% MSL, but no difference between 20 and 80% MSL. In agreement with Pel et al. (15), we observed that loading the system affected vibration acceleration primarily in the loaded plane (Z-axis), probably due to the added eccentricity in the vertical direction.
Altering the type of exercise (pulling vs. pushing) had no significant effect on a RMS, D, and f out. Our results suggest that the present VB is able to preserve its mechanical behavior during different static exercises in the vertical plane. We considered the pulling and pushing exercises because they are commonly used during upper-body strength training and rehabilitation programs and because of clear differences in hand-handle coupling actions (4). In contrast to the pushing exercise, in which the tension derived from vertical loading mainly acts on the palm (palm push) although low levels of finger grip force are needed, during the pulling exercise the tension acts mainly on the fingers, thus a stronger grip force (finger grip) is needed to hold the VB. These exercise specificities could have differently influenced the mechanical performance of the VB mainly because of differences in dynamic dampening of the vibrations, but this was not the case. Further research is required to confirm the consistency of the present results for movements outside the vertical plane and also for upper-body vibration devices other than VB.
The mean D and a RMS values observed with the present VB (respectively, ∼1.4 mm and ∼3.5g) were lower than those obtained with similar devices. Moras et al. (14) reported peak-to-peak displacement and maximal acceleration in the range of 4.6–6.1 mm and 3.8–9.8g, respectively. The device used by Poston et al. (16) provided 30 Hz and 4g with an amplitude of 1.1 mm, which resulted in a theoretical peak-to-peak displacement of 2.2 mm. Although both of these previous studies reported beneficial effects of vibration exercise on neuromuscular performance, some precaution is required due to the high peak accelerations that are produced, which could be potentially harmful to the human body (7). In fact, if one follows the safety rules to control the risks of hand-arm vibration in occupational vibration (8) and use the mean a RMS provided by the VB vibrating at an f out of 32 Hz (1.9g), the theoretical time to reach the daily exposure limit would be ∼35 minutes vs. 3–16 minutes with the VB of Moras et al. (14) and Poston et al. (16) vibrating at a similar frequency (∼30 Hz). Considering that exposure limit should not be exceeded to avoid potential health problems (8), the novel VB used in this study seems relatively safe. It is nevertheless important to remember that such vibration safety guidelines do not necessarily apply to sport training and physical rehabilitation, although an adaptation is urgently required (especially for upper-body vibratory devices).
A limitation of this study was related to the vertical accelerometer alignment. Although subjects were asked to hold the VB as steadily as possible and not to modify the wrist position during measurements, and accelerometer alignment was visually inspected to avoid undesirable tilts, we assume that a perfect vertical alignment was difficult to achieve.
The results of this methodological study establish the validity of a novel VB under a wide range of frequencies, loading conditions, and exercise types. This legitimates the safe and effective use of VB in the context of upper-body vibration training and rehabilitation programs. Concretely, practitioners can progressively and effectively increase the exercise stimulus to the neuromuscular system by modulating the vibration frequency (in the effective range, 18–42 Hz), the external resistance (by loading the bar almost up to the MSL), and the exercise type (lying row and bench press) with confidence, that is, without expecting the mechanical performance of the VB to be excessively altered, as recently demonstrated for commercially available whole-body vibration plates (15,17).
The authors express thanks to the participants for their time and effort. The authors would also like to thank Raul Cabello for help in data collection during the study. This research was supported by the European Social Fund and the Commission for Universities and Research of the Catalan Government's Department of Innovation, Universities, and Enterprises. We disclose professional relationships with companies or manufacturers that will benefit from the results of the present study and state that the results do not constitute endorsement of the product by the authors or the National Strength and Conditioning Association.
1. Bosco C, Cardinale M, Tsarpela O. Influence of vibration on mechanical power and electromyogram activity in human arm flexor muscles. Eur J Appl Physiol Occup Physiol 79: 306–311, 1999.
2. Burke D, Hagbarth KE, Lofstedt L, Wallin BG. The responses of human muscle spindle endings to vibration during isometric contraction. J Physiol 261: 695–711, 1976.
3. Cochrane DJ, Stannard SR, Walmsely A, Firth EC. The acute effect of vibration exercise
on concentric muscular characteristics. J Sci Med Sport 11: 527–534, 2008.
4. Dong RG, McDowell TW, Welcome DE. Biodynamic response at the palm of the human hand subjected to a random vibration. Ind Health 43: 241–255, 2005.
5. Eklund G, Hagbarth KE. Normal variability of tonic vibration reflexes in man. Exp Neurol 16: 80–92, 1966.
6. Elble RJ, Koller WC. Tremor. Baltimore, MD: Johns Hopkins, 1990.
7. Griffin MJ. Handbook of Human Vibration. London, United Kingdom: Academic Press, 1996.
8. HSE. Hand-arm vibration: The control of vibration at work regulations 2005. Guidance on Regulations. HSE BOOKS, 2005.
9. Humpries B, Warman G, Purton J, Doyle TLA, Dugan E. The influence of vibration on muscle activation and rate of force development during maximal isometric contractions. J Sci Med Sport 16–22, 2004.
10. Issurin VB, Liebermann DG, Tenenbaum G. Effect of vibratory stimulation training on maximal force and flexibility. J Sports Sci 12: 561–566, 1994.
11. Issurin VB, Tenenbaum G. Acute and residual effects of vibratory stimulation on explosive strength
in elite and amateur athletes. J Sports Sci 17: 177–182, 1999.
12. Luo J, McNamara BP, Moran K. A portable vibrator for muscle performance enhancement by means of direct muscle tendon stimulation. Med Eng Phys 27: 513–522, 2005.
13. Mischi M, Cardinale M. The effects of a 28-Hz vibration on arm muscle activity during isometric exercise. Med Sci Sports Exerc 41: 645–653, 2009.
14. Moras G, Rodriguez-Jimenez S, Tous-Fajardo J, Ranz D, Mujika I. A vibratory bar for upper body: Feasibility and acute effects on EMGrms activity. J Strength
Cond Res 24: 2132–2142, 2010.
15. Pel JJ, Bagheri J, van Dam LM, van den Berg-Emons HJ, Horemans HL, Stam HJ, van der Steen J. Platform accelerations of three different whole-body vibration devices and the transmission of vertical vibrations to the lower limbs. Med Eng Phys 31: 937–944, 2009.
16. Poston B, Holcomb WR, Guadagnoli MA, Linn LL. The acute effects of mechanical vibration on power output in the bench press. J Strength
Cond Res 21: 199–203, 2007.
17. Preatoni E, Colombo A, Verga M, Galvani C, Faina M, Rodano R, Cardinale M. The effects of whole-body vibration in isolation or combined with strength
training in female athletes. J Strength
Cond Res 26: 2495–2506, 2012.
18. Rauch F, Sievanen H, Boonen S, Cardinale M, Degens H, Felsenberg D, Roth J, Schoenau E, Verschueren S, Rittweger J. Reporting whole-body vibration intervention studies: Recommendations of the International Society of Musculoskeletal and Neuronal Interactions. J Musculoskelet Neuronal Interact 10: 193–198, 2010.
19. Roelants M, Verschueren SMP, Delecluse C, Levin O, Stijnen V. Whole-body-vibration-induced increase in leg muscle activity during different squat exercises. J Strength
Cond Res 20: 124–129, 2006.
20. Schuster A. On the investigation of hidden periodicities with application to a supposed 26 day period of meteorological phenomena. Terr Mag Atm Elect 3: 13–41, 1898.
21. Spiliopoulou SI, Amiridis IG, Tsigganos G, Economides D, Kellis E. Vibration effects on static balance and strength
. Int J Sports Med 31: 610–616, 2010.
22. Tripp BL, Faust D, Jacobs P. Elbow joint position sense after neuromuscular training with handheld vibration. J Athl Train 44: 617–623, 2009.