Acute and residual effects of vibratory stimulation
Power Plate Studies

Vibration applied to muscle or tendon induces a non-voluntary muscular contraction termed the `tonic vibration reXex’ (Eklund and Hagbarth, 1966). The voluntary impetus increases such a muscular contraction, and thus the maximum voluntary contraction can be facilitated (Matyas et al., 1986). Moreover, vibratory stimulation combined with a substantial voluntary eV ort was shown to elicit movement in neuromuscular patients who were unable to contract their paretic muscles (Hagbar th and Eklund, 1966). The technique is widely used in neurophysiology and physiotherapy (Granit, 1970; Bishop, 1974). Attempts to use vibratory stimulation in the training of athletes have been undertaken only recently (Nazarov and Spivak, 1987). A substantial increase in muscle strength was observed after 3 weeks of vibratory stimulation strength training when compared with regular strength training (Issurin et al., 1994).
Explosive strength, or the ability to develop force within a very short time, is of primary importance in many sports. Typical exercises for explosive strength training are characterized by fast muscular contractions with an external load of about 50± 70% of maximal strength (Vrijens, 1990). The immediate eV ect of such exercises can be assessed by the power which an athlete can generate in a movement. Several additional training techniques have been used to accentuate power training: the quick release technique, pre-stretching of active muscles before contraction, electrical stimulation and biofeedback. The objectives of these techniques are to improve upon previous achievements, to facilitate motor learning eV ects and to enhance muscular capacity (Torrey, 1985). Based on the results of a previous study (Issurin et al., 1994), it is likely that similar outcomes may also be achieved using vibratory stimulation.
The physical characteristics of the two groups of athletes (mean ± s)

Vibratory stimulation of the muscle tendon evokes an excitation of muscle sense organs (Brown et al., 1967). It has also been suggested that vibratory stimulation activates central nervous organization which is responsible for neuromotor control (Granit, 1970).
Another suggestion made recently concerns the dif- Wculty in achieving full muscle activation by voluntary eV ort during dynamic exercise when large muscle groups are involved (James et al., 1995). It is possible that, owing to vibration, the muscles will be par tially activated and their mobilization at the beginning of the eV ort will be faster. Therefore, it could be hypothesized that this additional vibratory excitation will stimulate the appropriate muscle group activation and the power exertion in explosive strength exercises. Moreover, an increased excitability of peripheral sense organs and the central nervous system may have a positive eV ect on the subsequent contractions. From an ethical point of view, vibratory stimulation exercises should be viewed as belonging to the group of so-called `non-conventional training’ methods, such as electrical muscle stimulation, velocity-assisted exercises (Maglischo, 1982) and computerized training machines (Torrey, 1985). Thus, superimposed vibration to the muscle may enhance its contraction (acute eV ect) or elicit post-stimulation facilitation (residual eV ect). The aim of this study was to establish the acute and residual eV ects of vibratory stimulation in explosive strength exercises.

Altogether, 28 male athletes aged 18± 42 years volunteered to participate in the study. They were divided into two groups (Table 1). The Wrst group consisted of athletes from the Israeli national judo, wrestling, weightlifting, gymnastics and track and Weld teams. These athletes regularly engaged in highly intensive power training. The second group consisted of amateur athletes par ticipating in club or college sports, such as basketball, volleyball, judo, weightlifting, body-building, boxing and track and Weld. The amateur athletes were also engaged in power exercises but not as extensively as their elite counterparts (2± 4 times a week). Because all of the athletes were familiar with power exercises, they were able to perform several repetitions with maximal eV ort and high reproducibility (see Table 2). This was one reason why elite and highly qualiWed athletes were enrolled as par ticipants.
The study was approved by the local ethics committee and informed consent was obtained from the par ticipants before the study began.
Instrumentation and tests:
The athletes performed bilateral biceps curl exercises in a sitting position on a `Schnell’ dynamic bilateral biceps machine (Schnell, Germany, D.B. Pat. 2213440). They were secured to the machine by pads placed at the elbow, chest and back (Fig. 1). The pulling action began from a position of maximal forearm extension and Wnished with the elbow at an angle of 90° (1.57 rad). The athletes were instructed to perform each repetition as quickly as possible.
The superimposed vibration during the exercise was transmitted to the muscles by a specially designed vibratory stimulation device (Issur in et al., 1994). It consists of an electromotor with a speed reduction and eccentric wheel. The load is held by a cable which is passed through the eccentric wheel via the pulleys (Fig. 1). The eccentric rotation elicited peak-to-peak oscillations of 3 mm with a frequency of 44 Hz. After vibration damping owing to cable transmission, the acceleration on the handle was about 30 m ´ s- 2 (RMS). Vibration from the two-arms handle was transmitted through the contracting muscles involved in the pulling action.
The power of the active phase of exercise was measured using a `Power Teach’ instrument (GE Sport S.A.S., Rome, Italy). Two probes were installed on the counterweight frame. The locations of the probes were established during the warm-up; the lower probe was placed 2 cm above the counterweight start position and the upper probe was placed opposite the Wnal counterweight position. Therefore, the probes covered the complete range of movement. The distance between the probes and the counterweight was transfer red to a microcomputer before the primary task. A magnetic element was Wxed to the counterweight. When the counterweight and magnet moved through the probes, electrical signals were generated and the time between the signals from the lower and upper probes was recorded. The mean power was computed as a product of force and velocity. The power of each repetition was shown to the performer on-line. After each set of exercises, the maximal and mean values were automatically recorded and displayed on the screen to an accuracy of 1 W.
The bilateral biceps curl exercise and instrumentation.

Anthropometric measures included the determination of height, body mass and bicep girth (i.e. mid-upper arm circumference), according to Tittel and Wutscherk (1972).
Study design:
Two separate series of biceps curl exercises were performed in random order by each athlete. Each series consisted of three sets with three repetitions in each set. In one series, the exercise was performed with vibratory stimulation in the second set; in the other series, the exercise was performed without vibratory stimulation. The maximal and mean power values of three repetitions were recorded after each set.
The athletes performed a general warm-up for 5± 7 min, including indoor running (2± 3 min), general calisthenics (1± 2 min) and exercises for the upper extremities (2 min). They then performed 8± 10 repetitions of the biceps curl with a low to medium load (20± 40% of body weight) to adapt to the exercise and equipment.
Then, 3± 5 attempts were performed at increasing weight to determine the one-repetition maximum value. The athletes were then allowed to rest for 15 min, during which anthropometric measures were taken and informed consent was obtained.
A weight equivalent to 65± 70% of the one-repetition maximum value was selected. Two series of exercises were performed, with the interval between them allowing full recovery (8± 15 min); the duration of the rest period was determined by the athletes. The exercise rate within a set was approximately one repetition every 2 s; the period of rest between sets was 2± 3 min. The athletes were asked to perform each repetition with maximal eV ort.
Data analysis :
The acute eV ect of vibratory stimulation was assessed as the diV erence between the power values in the second set with vibratory stimulation and in the Wrst set without vibratory stimulation. Similarly, the residual acute eV ect was assessed as the diV erence between the power values in the third (after vibratory stimulation) and Wrst (before vibratory stimulation) sets. These diV erence values in the Wrst and second series were subjected to repeated-measures analysis of var iance with group (elite vs amateur athletes) as a between-participants factor. SigniWcance was accepted at P < 0.05. Paired t-tests and Pearson product± moment correlations were computed to establish diV erences and relationships between the two series for maximal and mean power.

Maximal and mean power and two series
Results :
The means and standard deviations of maximal and mean power in the Wrst set of each series were compared using paired t-tests (Table 2). No signiWcant diV erence between the two series was found for the elite or amateur groups. The test± retest correlation coeYcient between the two series was 0.97 for maximal power and 0.97 for mean power of the biceps curl exercises.
The repeated-measures analysis of variance showed that mode of exercise (with vs without vibratory stimulation) had a signiWcant eV ect for mean power (F1,26 = 59.2, P < 0.001) and for maximal power (F1,26 = 56.3, P < 0.001). Also, the group factor (elite vs amateur) resulted in a signiWcant eV ect for maximal power (F1,26 = 4.41, P < 0.04). These eV ects are shown in Figs 2 and 3.
In the elite athletes, vibratory stimulation resulted in an average gain in maximal power of 30.1 ± 15.3 W and in an average gain in mean power of 29.8 ± 16.6 W; these values correspond to increases of 10.4% and 10.2% respectively. The series without vibratory stimulation revealed a non-signiWcant decrease in these values of 1.1 and 2.6 W, respectively. In the amateur athletes, the gains in maximal and mean power owing to vibratory stimulation were 20.0 ± 16.9 and 25.9 ± 18.9 W respectively; these values correspond to increases of 7.9% and 10.7% respectively. The maximal and mean power decreased by 7.4 W without vibratory stimulation. We also observed that the immediate acute eV ect in maximal power was signiWcantly greater in the elite than in the amateur athletes (F1,26 = 7.32, P < 0.01).
Similar analyses of variance were applied to the mean and maximal power diV erences between the third and Wrst sets in the two modes of exercise (with vs without vibratory stimulation). Group (elite vs amateur exercise mode and the interaction eV ects were all nonsigniWcant (P > 0.05) (see Fig. 4). Therefore, vibratory stimulation in the second set resulted in an insigniWcant residual eV ect in the third set.

An increase in contraction strength induced by the tonic vibration reXex has been widely documented. Hagbar th and Eklund (1966), Johnston et al. (1970) and Arcangel et al. (1971) all reported that muscle force registered during isometric contractions increased because of local vibratory stimulation applied to the muscle or tendon. A similar result was noted by Armstrong et al. (1987), who administered 40 Hz superimposed vibration and registered an increase in grip force of 52%. These studies applied vibratory stimulation to muscles which contracted with low to intermediate levels of eV or t. Matyas et al. (1986) reported the facilitation of maximum voluntary contraction caused by 50 Hz tendon vibration in hemiplegic patients. Samuelson et al. (1989) reported a reduction in endurance of a maximal isometric contraction and a decrease in maximal force with 20 Hz superimposed vibration, in contrast to the results of the present study.
Three factors may be attributed to the acute vibratory stimulation eV ect: (1) the motor pool activation, (2) the frequency of vibratory stimulation and (3) the initial length of the stimulated muscles. Matthews (1966) and Brown et al. (1967) found that vibratory stimulation excites the primary aV erent endings of the muscle spindles which activate a-motoneurons. Unlike local vibratory stimulation, the low-frequency superimposed vibratory wave propagates from the distal links to muscles located proximally and activates a greater number of muscle spindles. Their discharge activates a larger fraction of the motor pool and recruits many previously inactive motor units into contraction.
There is evidence that an increase in vibration frequency evokes a proportional increase in muscle tension (Matthews, 1966). However, the high-frequency component of vibration is absorbed by soft tissues, whereas the low-frequency component propagates through the human body tissues (Pyykko et al., 1976). Therefore, on the one hand, the eV ect of vibratory stimulation depends on the frequency; on the other hand, low-frequency vibratory waves can only propagate through the kinetic chain to proximal muscle groups and activate them. It is likely that vibratory stimulation at a frequency of 40± 50 Hz may be optimal to combine two diV erent tasks: (1) transmission of vibration and (2) muscle activation before and during voluntary contraction (Issurin and Temnov, 1990).
It is known that stretched muscles are more sensitive to vibratory stimulation and contract more strongly (Eklund and Hagbar th, 1966; Johnston et al., 1970; Rohmert et al., 1989). In Samuelson and co-workers’ (1989) study, the superimposed vibration was administered during knee-joint extension in the sitting position with a knee angle of 90° (1.57 rad). Hence, the quadriceps muscle was not in a stretched position. This may be one reason why Samuelson et al. did not Wnd any facilitatory eV ect of vibration on maximum isometric contraction. Another reason may be the lower vibratory stimulation frequency of 20 Hz they used. In contrast, the present study was conducted with extremely stretched muscles before each repetition. This could be why we observed a power increase during vibratory stimulation.
Post-vibratory residual eV ects have also been widely documented in the literature. Arcangel et al. (1971) reported a substantial and signiWcant increase in the Achilles tendon reXex after 10 and 20 s tendon vibration. Cafarelli and Layton-Wood (1986) reported an improvement in force sensation in fresh muscles after short-term vibration. The reasons for such eV ects are probably associated with an increase in the sensitivity of the muscle receptors to excitation. Elevation of muscle temperatures resulting from the friction between vibrating tissues (Oliveri et al., 1989) and vibrationinduced increases in blood Xow (Wakim, 1985) may also contribute to the post-vibratory eV ect. In fact, the residual gain in power observed in this study was relatively small and not statistically signiWcant. Relatively short-term vibratory stimulation, as implemented in this study (6± 7 s), is probably not suYcient to aV ect subsequent muscle strength.
The diV erence in muscle response between the elite and amateur athletes was statistically signiWcant. The average gain in maximal power owing to vibratory stimulation was greater among the elite athletes. The reason for this marked diV erence may be associated with the higher sensitivity of muscle receptors and the central nervous system of elite athletes to additional stimulation.
In summary, the superimposed vibratory stimulation allowed a signiWcant facilitation of an explosive strength exertion. This approach may be useful in identifying the hidden reserves of an athlete and in augmenting an acute eV ect of power training