All vibratory devices, as exemplified in Fig. 1 (left panel) were constructed using steel DC micro devices (model 1319006SR – Faulhaber® – Switzerland) with a maximal nominal tension of 6V, maximal no-load speed of 13,100rpm and weighing 12g each. An eccentric load (weighing 0.5g; 1cm of diameter) was attached to each micro device axis. The rotational velocity of the eccentric mass, and therefore the vibration frequency, is linearly determined by the amount of electrical potential difference given to each micro device (details in the Results section). Hence, electrical wires were fixed to each device using ordinary soldering. The sets (devices plus eccentric masses) were embedded in a custom cylindrical enclosure made of non-toxic polyvinyl chloride (PVC), measuring 4.5cm×2cm×2cm, weighing 12g. In total, each vibratory device weighed between 27g and 31g. The designed configuration allowed vibration only in 2-D directions (X=anteroposterior; Z=mediolateral).

Figure 1.

Vibratory devices (left panel) and logical control system (right panel). (A) Micromotor; (B) eccentric mass; (C) PVC cylindrical enclosure; (D) power wires; (E) control board; (F) vibratory devices power output; (G) control board power input; (H) USB Zigbee controller module.

(0.25MB).

In the RCVibro System, each vibratory device is controlled by a custom electronic board (Fig. 1 – right panel) measuring 12cm×12cm and controlled by a Zigbee network protocol (radio transmission). The Zigbee controller module may be connect to any computer, via USB (Fig. 1). The custom-made board (powered by a 5V AC adapter) can control up to eight different vibratory devices and allows through ordinary hyperterminal software, the manipulation of both vibration onset moment and frequency. This manipulation can be applied to all vibratory devices simultaneously or individually. Also, the vibration frequency of each device is manipulated by the command “at+x=y” in a hyperterminal software, whereas “x” is the target channel and “y” is the percentage of the maximal given electric potential difference (100% is equal to 5V). This board can be attached on a belt or in a backpack, which allows the participant to move freely.

Since the devices axis rotational velocity is dependent on the given electric potential difference, it was necessary, for each vibratory device (six were separately tested), to achieve a percentage of electric potential difference/vibration frequency curve. Therefore, a three-dimensional optoelectronic movement analysis system was used (OPTOPTRAK Certus – NDI®). This system has an accuracy of 0.1mm and a resolution of 0.01mm. Thus, the OPTOPTRAK system is sensitive enough to detect the vibratory movement. Since frequencies higher than 218Hz were not expected (maximal no-load speed of 13,100rpm) and to respect the Nyquist theorem, a sample acquisition rate of 500Hz was used. In each device, an infrared emitter diode (IRED) was fixed using adhesive tape. No emissors movement was observed during the assessments. The vibratory devices were fixed in a vertical position, allowing free movement in two directions (X and Z).

The electric potential difference for each vibratory device was increased in series of 7% of total, resulting in eleven points series (14–84%; lower percentages were excluded, since, in some devices, they did not promote any movement; higher percentages were not used to avoid any mechanical damage on vibratory devices). A 5s vibration period was assessed for each electric potential difference. The position of each marker was offline filtered using a low-pass 4th order Butterworth filter with a 10Hz cut-off frequency. The 5s periods was then split into six 0.5-s periods, with 0.5-s overlap (the first and last second of data collection were excluded): 1–1.5s, 1.5–2.0s, …, 3.0–3.5s, 3.5–4.0s. For each period, the power spectrum of each marker position was determined using power spectral density (PSD) analysis through specific-designed MatLab (MathWorks®) codes. The power spectrum had a resolution of 0.66Hz.

Afterwards, we determined the peak power frequency for each 0.5-s, for all electric potential difference percentages for both X and Z directions. This procedure allowed us to determine: (i) the percentage of electric potential difference/vibration frequency curve for each device (Fig. 2); (ii) the vibration variability within the 5s data-sampling period (through the coefficient of variation ([standard deviation*100]/mean).

Figure 2.

Percentage of electrical potential difference/vibration frequency curves achieved for one vibratory device on each test session. Bold diamonds: Test; Open circles: Re-Test 1; Crosses: Re-Test 2; Gray triangles: Re-test D2.

(0.07MB).

In order to assess the system test–retest reliability in the same day, these procedures were repeated three times, named Test, Re-Test 1 and Re-Test 2. They were performed two hours apart from each order. Additionally, in order to assess the test-retest reliability between different days, a fourth test (named Re-Test D2) was repeated 24h after Test. For the reliability (defined as the extent to which scores have not changed for repeated measurements),15 we considered the mean vibration frequency between the six 0.5-s periods. For each device, to assess the reliability within days, we used ICC(2,6) procedures among Test, Re-Test 1 and Re-Test 2; to assess the system reliability between days, we used the ICC(2,6) procedures between Test and Re-Test D2.16,17 The standard error of measurement (SEM) was also determined. Finally, we verified the Minimal Detectable Change (MDC) for each device using the following formula: MDC=SEM×1.96×√2. The MDC reflects the value needed to be changed at 95% of confidence to be considered a real change and not a difference given by a measurement error.

In order to assess the proper functioning of the entire custom-made system and also, to ensure that the same effects observed in previous studies8,9,14 could be replicated by the RCVibro System, we assessed the CoP behavior during upright standing. Undergraduate and graduate students were invited to participate in this study through personal communication. A convenience sample of 15 subjects was determined. The CoP behavior of these fifteen healthy young adults (7 males, 8 females; mean (standard deviation) age of 26.29±6.46 years; 167.71±7.89cm of height; and 68.86±8.76kg of weight) was assessed while standing with eyes closed in two conditions: without and with vibration, applied bilaterally on the tibialis anterior muscle. This research has been approved by the local Ethical Board (Protocol: 083/2010 – Comitê de Ética em Pesquisa em Seres Humanos – IB, Universidade Estadual Paulista (UNESP), Rio Claro, SP) and an informed consent form was signed by all participants.

Two sets of trials were conducted, without randomization, maintaining the order: without (OFF) and with vibration (ON). This order was chosen to avoid any lasting vibration effect in OFF. Since only three trials were used (described ahead) and only healthy young participants were assessed, we do not believe in a fatigue-effect in the second round of trials (ON). During all attempts, the participants were asked to maintain an upright posture for 40s, standing on a force plate (AMTI®) with eyes closed (to enhance the vibratory effect14). The feet position was standardized for each participant and it kept consistent across all trials. The positioning of the CoP in relation to the force plate zero offset was assessed using the Balance Clinic software (AMTI®) with a sample rate of 200Hz. A resting time of 30s was given to participants between attempts. In total, three trials of each condition were performed, and the average of them was considered for analyzes.

For all conditions, the vibratory devices (position detailed below) were switched on at 10th second of data collection and a 30s vibration period was used to elicit postural effects. To investigate the effect of vibration on the CoP behavior, the CoP anteroposterior (A-P) position, CoP area and CoP mean velocity were assessed at two moments: at the close-to-last five seconds without vibration (4–9th seconds of data collection) and at the close-to-last five seconds of vibration (34–39th seconds of data collection). The CoP area was determined using the 95% confidence ellipse.18

Vibration was applied bilaterally on the tibialis anterior muscle belly. The devices used on the functioning test were randomly chosen from the six devices tested previously. All subjects used the same devices. The tibialis anterior muscle was chosen since this muscle is sensitive to local vibration.8,9,19 The devices were distant approximately 4.5cm proximal to the center of the muscle belly (adapted from the SENIAM Project19) and were positioned with the devices axis in parallel to muscle fibers.7,9 To locate the center of the muscle belly, participants seated on a chair and the center of the muscle was considered as the 1/3 of the distance between the head of the fibula and the tip of the medial malleolus.19 The center of the vibratory devices was positioned directly on the target spots. Extra care was taken to ensure that both vibratory devices were exclusively on the tibialis anterior muscle. For better fixation, ordinary elastic bands were used. A 100Hz vibration frequency was used to determine the vibration effects on the CoP dependent variables, since this frequency is successful to induce postural adaptations.7

The Statistica 7.0 (StatSoft®) software for Windows was used for statistical procedures and significant results were considered when p<0.05. Munro's classification for reliability coefficients was used to describe the degree of reliability: 0.26–0.49 reflects low correlation; 0.5–0.69 moderate correlation; 0.7–0.89 high correlation; 0.9–1.00 very high correlation.20 Student t-test for dependent variables was used to assess the difference in CoP behavior between the close-to-last five seconds without vibration and the close-to-last five seconds of vibration.

Fig. 2 shows the percentage of electrical potential difference/vibration frequency curves achieved for one vibratory device on Test, Re-Test 1, Re-Test 2 and Re-Test D2. The curves show a linear increase of vibration frequency with higher percentages of electrical potential differences. A similar pattern was observed for all devices. The vibration amplitude ranged between 0.7 and 1.2mm for different devices.

Table 1 presents the variability (CV), the ICC(2,6), its 95% Confidence Interval, the SEM and also the MDC results for each vibratory device in both X and Z directions. It can be noticed an extremely low variability in the vibration frequency within the five seconds testing period (CV<2.03%). All devices showed a very high level of agreement between the vibration frequencies achieved within the same day (ICC(2,6) ranging from 0.95 to 0.99) and between different days (ICC(2,6) ranging from 0.81 to 0.98) with some exceptions, such as in the device #3 (ICC(2,6), ranging from 0.63 to 0.71), discussed further. The complete results per device can be consulted in Table 2.

The CoP position behavior of one participant during one trial with and without vibration is shown in Fig. 3. It can be observed a forward CoP displacement after the vibration onset. On the other hand, Fig. 3 shows a CoP position maintenance without the use of vibration. The Student t-test results confirmed this observation, showing a mean of 1.10cm forward CoP displacement with vibration (OFF: −0.77±0.62cm and ON: 0.32±0.67cm; t(14)=3.34, p<0.01; effect size: 1.41; power: 0.999). Furthermore, vibration increased the CoP mean velocity (OFF: 1.21±0.30cm/s and ON: 1.46±0.46cm/s; t(14)=2.27, p<0.01; effect size: 0.40; power: 0.40). None influence of vibration was observed for CoP area (OFF: 1.20±0.91cm2 and ON: 1.41±1.04cm2; t(14)=1.16; p=0.26).

Figure 3.

The raw center-of-pressure (CoP) response to tibialis anterior vibration, in one subject. Bold arrow: vibration onset; gray bars: data acquisition periods.

(0.09MB).