The effect of two intensities of massage on H-reflex amplitude.

The control of spasticity (hypertonicity) is an important aspect in the physical rehabilitation of neurologically impaired persons such as those with head injury, stroke, spinal cord injury, and multiple sclerosis. Physical therapists use a variety of modalities to control spasticity. Spasticity has been defined as

... a motor disorder characterized by a

velocity-dependent increase in tonic

stretch reflexes (muscle tone) with

exaggerated tendon jerks, resulting

from hyperexcitability of the stretch

reflex, as one component of the upper

motoneuron [syndrome.sup.(485)][l]

Prolonged muscle stretching, application of tendon pressure, and muscle tapping are commonly used clinical tools for the reduction of spasticity.[2] Among the manual inhibitory techniques currently used by therapists, a number have been studied experimentally, including the use of tendon pressure,[3-5] muscle stretching,[6] massage,[7-9] and muscle tapping.[10] All of these manual techniques have been shown to decrease H-reflex amplitude - an indirect measure of spinal motoneuron excitability.[11]

Massage has only recently been subjected to scientific scrutiny. Recent studies conducted in our laboratories[7-9] have clearly demonstrated that one-handed petrissage decreases peak-to-peak H-reflex amplitude (suggestive of a decrease in motoneuron excitability) in neurologically healthy subjects. Furthermore, these studies[8,9] demonstrated that the inhibitory response produced by the massage is specific to the muscle being examined and is not due to the effects of reciprocal inhibition, temperature change, changes in nerve conduction velocity, or gender differences. One aspect of motoneuron excitability that has not yet been studied in relation to massage is the differential effect of varying intensities of pressure on the recorded inhibitory response.

Leone and Kukulka[5] investigated the effect of different levels of static pressure applied to the Achilles tendon on H-reflex amplitudes in patients with stroke. They reported no difference in the amount of inhibition (decrease in H-reflex amplitude) produced following the application of 5 or 10 kg of static pressure to the Achilles tendon. The fact that this study was performed on a neurologically impaired sample could explain the lack of differential response observed between the two levels of pressure.

The intermittent nature of the pressure applied during petrissage, the form of massage used in this study, may cause a passive stretch in the Achilles tendon similar to that produced by the application of intermittent tendon pressure.[4] The similarity between them modes of treatment may explain in part the reduction in our H-reflex amplitude reported in our studies.[7-9] Belanger et al[10] reported a decrease in H-reflex amplitude with muscle tapping (a form of intermittent pressure) in a neurologically healthy sample. An initial attempt to standardize the pressure exerted during muscle tapping was made by these authors, albeit unsuccessfully.

Recently, an increase in interest regarding methods of quantifying the amount of manual pressure exerted during physical therapy interventions has appeared in the literature. Several studies have developed specific measurement systems to quantify the applied manual pressure in both human[12,14] and animal[15] models. Researchers have also investigated the intersubject[12,13] and intrasubject[12] reliablity of measurements obtained during the use of these techniques, as well as the effect of feedback[16] and on the learning of these skills.

In our previous studies,[8,9] we quantified the rhythm of the massage application 0.5 Hz), but not the pressure. The quantification of the pressure exerted during massage should not be overlooked for a number of reasons. Such quantification is important in (1) standardizing the massage application; (2) investigating the differential effects of varying intensities of pressure on the previously observed inhibitory response; and (3) finding a clinically effective, yet comfortable, level of pressure to be exerted in the clinical setting.

The purposes of this study were (1) to compare the effects of two levels of massage-light massage (LM) and deep massage (DM)-on the reduction of H-reflex amplitudes, (2) to investigate the presence of any gender effects in relation to massage, and (3) to describe a method used to quantify the amount of pressure exerted during the application of massage. We expected that both levels of massage would result in a decrease in the peak-to-peak H-reflex amplitude as compared with control values, that DM would result in a greater inhibition of the H-reflex than would the superficial massage, and that no difference m response between male and female subjects would be observed. The results of this study will serve to advance an ongoing study designed to investigate the effect of massage in a neurologically impaired group of subjects.

Method

Subjects

Twenty neurologically healthy subjects (10 male, 10 female), with a mean age of 22.6 years (SD=1.9) volunteered to participate in this study. Participants were recruited from the staff and student populations of both Concordia University and the Universite de Montreal. All subjects signed an institutionally approved informed consent form prior to participation in the study. Criteria for admission were chosen in an effort to minimize the effects of extraneous variables known to have an effect on the H-reflex (dependent variable). Admission criteria were that subjects (1) have no history of neurological illness, (2) be less than 40 years of age," (3) not engage in strenuous physical activity for at least 12 hours prior to data collection,is (4) not ingest any substance containing alcohol or caffeine for at least 12 hours prior to data collection,[19] and (5) present no other condition (eg, orthopedic, urinary) that might inhibit them from remaining comfortably in the testing position throughout the period of data acquisition (approximately 60 minutes).

Massage Technique

The massage technique used in this study consisted of a one-handed petrissage technique.[20,21] The petrissage was performed in accordance with guidelines previously described in the literature[20,21] and in our previous studies.[7-9] One-handed petrissage consists of alternately grasping and lifting the muscle away from the bone in a circular motion, releasing the tissues on the downward pan of the motion.[20,21]

Pressure Quantification System

Under normal clinical conditions, massage does not require a precise quantification of the pressure applied. In a study comparing the effects of two intensities of massage, however, the quantification and standardization of these pressures are essential. As no validated system was available to us, we decided to construct a presssure quantification system (PQS) based on the work of Pedneault et al[12] and Stijns.[13]

The PQS used in this study was chosen both for its simplicity of construction and use and for the availability of all necessary materials in physical therapy departments. The measurement system consisted of a Jobst Intermittent Compression Unit (model 85-00)(*) and its accompanying upper-extremity (UE) pressure sleeve, a Wika pressure gauge (type 611.10),(dagger) and a two-way air valve used to regulate pressure. This system is illustrated in Figure 1. The Jobst UE sleeve was chosen for its approximation in size, when MM to the human calf the pressure gauge, which was used to provide visual feedback, was selected for its sensitivity to low pressures (0-7.5 kPa [0-30 in [H.sub.2O]). The sleeve was firmly attached to a table and was inflated to an initial pressure of 1.25 kPa (5 in [H.sub.2O]). The sleeve at this pressure was subjectively determined to be similar to the consistency felt in a neurologically healthy leg at rest. No formal trials for validity or reliability were undertaken. The PQS served simply as an indicator of the pressure applied. In the absence of any other pressure-measuring device, the PQS was considered adequate for the objectives of this study.

Standardization of the Massage

Prior to data collection, a physical therapist (JG) was trained, using her dominant hand (left), to apply two different preselected levels of massage (IM, DM). The selected pressures corresponded to 2.5 kPa (10 in [H.sub.2O]) for DM and 1.25 kPa for LM, as measured by the PQS.

The criteria pressures for DM and LM corresponded to values of 3.75 kPa (15 in [H.sub.2O]) and 2.5 kPa, respectively, on the pressure gauge (baseline at 1.25 kPa). These pressures were chosen empirically, both being assessed by a trained physical therapist as being similar to the pressures exerted during a "deep" or "light" massage administered in a clinical setting. Both levels of massage were performed at a rate of 0.5 Hz, as suggested in the literature[20] and as used in previous studies.[7,8] The rate was monitored using a stopwatch.

Training Criteria

Training criteria were established prior to the beginning of the study. The therapist was deemed to be trained when she could perform the following sequence of rates and pressures (within an accepted maximum error of [+ or -] 5%) without feedback. The sequence consisted of 10 DM manipulations applied to the pressure sleeve at the rate and pressure previously selected, followed by a 3-minute pause, followed by 10 LM manipulations at the appropriate rate and pressure.

Training Protocol

Prior to training, talcum powder was applied to the surface of the pressure sleeve to reduce friction during the training session and to more closely approximate the surface of the sleeve to that of a human leg. The training session consisted of three successive phases.

Phase 1 consisted of the massage manipulation performed by the therapist with visual feedback of both pressure and rate (by monitoring the pressure gauge and stopwatch) for the two levels of massage. This phase lasted approximately 5 minutes (Fig. 1).

Phase 2 was a repeat of phase 1 without the visual feedback. Verbal cues regarding the rate and pressure were given by an assistant. This phase also lasted approximately 5 minutes.

Phase 3 of the training session consisted of the massage manipulations as established by the preset training criteria with no feedback of any type. During the manipulations of the pressure sleeve, the assistant recorded the pressure and rate of application in order to ensure accurate replication of the preset criteria within [+ or -] 5%. Only one session was found to be necessary to fulfill the training criteria. Prior to each session of data collection, the therapist was required to retrain on the pressure sleeve according to the criteria previously established in order to ensure conformity to the standardized pressures.

Instrumentation

H-reflex recordings were obtained from the distal portion of the right triceps surae (IS) muscle group using a pair of Meditrace silver-silver chloride disposable electrocardiographic (ECG) surface electrodes. (double dagger) Electrode placement was standardized along the midline of the posterior surface of the TS muscle at 1/16 the distance between the flare of the lateral malleolus md the distal popliteal crease.[22] Electrodes were positioned in parallel to the direction of the TS muscle fiber orientation with an interelectrode distance of 15 mm. A reference electrode Meditrace silver-silver chloride disposable ECG surface electrode) was positioned on the lateral malleolus of the same leg and secured with adhesive tape.

Stimulating electrodes were positioned as follows. The anode, a 2-cm-wide band, was placed around the distal thigh, just proximal to the superior pole of the patella the cathode, a stationary surface electrode, was fixed in position in the popliteal fossa following localization of the posterior tibial nerve with a hand-held electrode probe. The electrode placement for both the recording and stimulation configuration is illustrated in Figure 2.

H-reflexes and M-responses were elicited every 10 seconds by a 1-millisecond square-wave impulse to the posterior tibial nerve, delivered by a computer-driven Grass S88 stimulator[sub-section] and related stimulus isolation and constant current units[sub-section] (Grass SIU5 and CCU1, respectively). The recorded H-reflexes and M-responses were amplified (DISA model 15CO15~~) and band-pass filtered (3 dB down at 10 and 1,000 Hz) prior to being digitized & 5,000 Hz. Peak-to-peak amplitudes (in millivolts) of the H-reflex and M-response were recorded and analyzed.

Experimental Procedure

Subject positioning. All participants lay in a prone position on a padded wooden treatment table with their right lower extremity supported in a frame that maintained the leg in a standardized position of 17 degrees of knee flexion and 15 degrees of ankle plantar flexion. This position is illustrated in Figure 2. To control for the possible effect of the asymmetrical tonic neck reflex,[23] the head was always positioned to the left.

Preparation for data acquisition. In preparation for electrode placement, the distal region of the TS muscle and the popliteal fossa were shaved and then cleansed with a 70% isopropyl alcohol solution. Electrodes were positioned as described previously. A 5-minute period of low intensity stimulation was then initiated to allow the subjects to familiarize themselves with the electrical stimulation. During this 5-minute period, the therapist retrained on the PQS as described previously. Following this 5-minute period, an abbreviated H-reflex recruitment profile was generated in order to establish the control stimulation intensity to be used throughout the period of data collection. The selected stimulation intensity corresponded to an H-reflex amplitude of approximately 70% of the maximal M-response (Mmax) amplitude. This amplitude is somewhat higher than that previously used[7-9] (30%-50% of Mmax amplitude); however, this value was chosen to ensure that an M-response of sufficient amplitude would be accurately monitored for any changes. Alterations in the M-response would indicate that changes in the stimulation conditions had occurred and therefore interpretation of the results would be impossible.

Data Acquisition/Experimental Protocol

Data acquisition took place in a soundproof room with subdued lighting in order to minimize competing stimuli and to provide a relaxing atmosphere. All subjects were instructed to remain as still and as quiet as possible during the data-acquisition period. Ten trials were recorded during each of the five control conditions (C1, C2, C3, C4, C5) and during each of the two experimental conditions (DK, LM). The purpose of these five control conditions was to establish premassage and postmassage baseline values with which H-reflex amplitude changes could be compared. The experimental conditions were interspersed between C1 and C2 and between C3 and C4, during which one level of massage was administered. A diagram of the experimental protocol is presented in Figure 3.

Each condition (control and massage) was of a 3-minute duration. During both massage conditions, however data were collected only during the first 2 minutes, although the massage was applied throughout the entire 3-minute period. This procedure allowed the time necessary to prepare the computer so that conditions C2 and C4 could begin immediately following cessation of the massage. In order to minimize any carryover effects between the two levels of massage, the order of administration of the massage conditions was reversed with each successive subject. The sequence of massage condition reversal for subjects 1 through 4 is presented in Table 1.

Table 1. Order of Control and Massage Conditions for Subjects 1 Through 4(a)
Subject No.   Order of Conditions    Massage Order(b)
1             C1-DM-C2-C3-LM-C4-C5   DM/LM
2             C1-LM-C2-C3-DM-C4-C5   LM/DM
3             C1-DM-C2-C3-LM-C4-C5   DL/LM
4             C1-LM-C2-C3-DM-C4-C5   LM/DM
(a) DM = deep massage, LM = light massage, C1-C5=control
conditions.
(b) This sequence of massage condition reversal continued for
all subjects (N=20) in order to
minimize any possible carryover effects between the two
intensities.

Data Analysis

Group descriptive statistics (mean, standard deviation, and standard error of the mean) were calculated for both H-reflexes and M-responses for each of the seven conditions. The means of the 10 trials for each condition for each subject were entered into a 2 x 7 (gender X condition) analysis of variance (ANOVA)for repeated measures on the condition factor. Peak-to-peak H-reflexes and M-responses were analyzed in order to test for both gender and condition effects. The Newman-Keuls post hoc procedure was used to locate statistical differences. In order to identify variations in H-reflex amplitude during the two massage conditions and during each control condition immediately following the massage condition (ie, C2 and C4), four one-way repeated-measures ANOVAS were computed on the 10 individual trials in each condition. All statistical analyses were performed with the SYSTAT analysis program.(#) A significance level of .01 was used for all statistical analyses.

Results

M-Response

The mean peak-to-peak amplitudes for the 10 M-responses elicited during each of the seven conditions for the combined group ranged from 0.63 to 0.69 mV (Tab. 2). The ANOVA for gender and conditions (Tab. 3) revealed no significant differences attributable to gender (F=0.10; df= 1,18; P>.01), conditions (F=2.20; df=6,108; P>.01), or the interaction between gender and conditions (F=0.90; df=6,108:. P>.01). The stability of the M-response throughout the seven conditions indicates that no change occurred in the recording conditions during the period of data acquisition.

[TABULAR DATA OMITTED]

Table 3. Analysis-of Variance for Peak-to-Peak M-Response Amplitude
Source           df    SS      MS    F       P
Gender (G)        1    0.18   0.18   0.10   .761
Error            18   34.27   1.90
Condition (C)     6    0.04   0.01   2.20   .049
GxC               6    0.02   0.00   0.90   .499
Error           108    0.36   0.00

H-reflex

The descriptive statistics for the H-reflex amplitudes are presented in Table 4 and Figure 4. The corresponding ANOVA results are presented in Table 5. No statistical differences were located between the responses of the male and female subjects for the peak-to-peak H-reflex amplitudes (F=1.44; df=1,18; P>.01). A significant conditions effect was obtained (F=73.86; df=6,108; P<.01). Subsequent post hoc analysis determined that the H-reflex amplitudes obtained during both the DM and the LM conditions were significantly reduced (49% and 39%, respectively) relative to the amplitudes obtained during the control conditions. In addition, the H-reflex amplitude obtained during the DM condition was significantly reduced in comparison with that obtained during the LM condition (P<.01), thus indicating a differential response to the two experimental levels of pressure. The gender X conditions interaction was not significant (F=2.05; df=6,108; P>.01).

[TABULAR DATA OMITTED]

Table 5. Analysis-of-variance Results for Peak-to-Peak H-Reflex
Amplitude
Source          df    SS       MS      F       P
Gender (G)        1    50.34   50.34    1.44   .246
Error            18   631.23   35.07
Condition (C)     6   352.43   58.74   73.86   .0001
GxC               6     9.78    9.78    2.05   .650
Error           108    85.86   85.86

The data collected during the 10 trials of each of the DM, C2, LM, and C4 conditions were amw. Four oneway repeated-measures ANOVAS were performed. The ANOVA for the condition following the LM condition revealed no significant difference among trials (F=1.71; df=9,144; P>.01). A significant difference, however, was detected during the condition immediately following the DM condition (F=3.95; df=8,144; P<.01). A Newman-Keuls past hoc analysis revealed this difference to be between the first trial following the DM condition and all other trials in that condition P<.01), with the first trial being significantly reduced compared with the other trials in that condition. No significant difference was found among the 10 trials of the DM or the LM condition (F=1.58; df=9,144; P>.01 and F=1.30; df=9,144; P>.01, respectively).

Discussion

This study compared the changes in the amplitude of the H-reflex during the application of two intensities of massage to the right triceps surae muscle of 20 neurologically healthy subjects. As expected, a significant mean inhibitory response in the triceps surae muscle motoneuron excitability (as evidenced by a decrease in the peak-to-peak H-reflex amplitude) was recorded during the application of massage, as compared with the values obtained during the control conditions. This response was consistent in all subjects.

These results support those reported in our earlier studies[7-9] and reemphasize the findings of an inhibitory response produced during the application of massage. A noticeable difference, however, exists between the magnitude of the inhibitory response recorded m this study and that of a previous study conducted in our laboratory.[7] In this study, the DM (2.50 kPa) produced a 49% reduction in H-reflex amplitude compared with control values, and the LM (1.25 kPa) produced a 39% reduction, whereas a 71% reduction in H-reflex amplitude was recorded previously.[7] This difference in response to massage may possibly be attributed to a difference in the pressure exerted by the therapist in our previous study.[7] Unfortunately, this hypothesis is difficult to evaluate because of the lack of quantification of pressure in that study. More importantly, however, is that our analyses also revealed a statistically significant differential response between the two levels of massage administered, with DM resulting in a significantly greater inhibitory response in comparison to LM (49% and 39% reduction, respectively). This finding suggests that the mechanisms involved in the inhibition may be regulated by pressure-sensitive receptors.

To our knowledge, only one previous study[5] has addressed the question of a differential response in H-reflex amplitude to pressure. In that study, Leone and Kukulka applied different levels of static pressure (5 kg and 10 kg) to the Achilles tendon in a neurologically impaired sample, but no differences in the resulting H-reflex amplitudes were reported. These findings cannot be directly compared with those of this study because of several important methodological differences between the two studies. The pressure Leone and Kukulka used was static in nature and was applied solely to the Achilles tendon. in contrast, our study involved a neurologically healthy sample, and the pressure used was dynamic in nature and was applied to the muscle belly.

The massage technique used in this study undoubtedly activates a wide spectrum of afferents, including both cutaneous and muscular mechanoreceptors, as mentioned in studies dealing with similar manual modalities.[3-5,10] It is difficult, however, to determine which of these receptors is responsible for the inhibition recorded during massage. Based on the current literature, it would appear that the mechanism responsible involves a combination of both types of receptors.

The role of the cutaneous mechanoreceptors in the amplitude changes of the H-reflex m both neurologically healthy[24,25] and neurologically impaired[26,27] persons has been studied. These studies demonstrated that desensitizing the skin by the application of a topical anesthetic resulted in an increase in the H-reflex amplitude. the role of the cutaneous mechanoreceptors was interpreted as having an inhibitory effect on the H-reflex amplitude. Consequently, it may be assumed that these cutaneous mechanoreceptors, presumably activated during massage, exert an inhibitory effect on the central nervous system during massage, as well as during other manual techniques (eg, application of tendon pressure, muscle tapping).

The role of these cutaneous mechanoreceptors in the depression of the H-reflex during muscle stretching was also examined and subsequently refuted by several authors.[6,28,29] They concluded that the inhibitory response to muscle stretch was mediated primarily by the activity of secondary muscle spindle afferents. Belanger et al[10] have also suggested that the reduction in H-reflex amplitude observed during their interventions (muscle stretching, muscle tapping) could be mediated by muscular mechanoreceptors (golgi tendon organs, muscle spindles).

Pressure applied to the Achdles tendon pruduces a passive stretch in the soleus muscle similar to that found during muscle stretching. The grasping and lifting nature of the massage technique applied to the calf undoubtedly causes a passive stretch in the soleus muscle. Based on the conclusions of Etnyre and Abraham[28] and Burke et al,[29] it is possible that the inhibition recorded during massage is also mediated y the secondary spindle afferents. The results of this study provide some preliminary evidence that pressure mechanoreceptors may play a role in the inhibitory response recorded. This evidence does not preclude the possible contributing effects of cutaneous mechanoreceptors in the observed response.

The PQS developed for this study is a preliminary attempt to quantify the pressures applied during the application of a dynamic manual modality and, as such, possesses certain weaknesses. Among the limitations of the system are that the sleeve, once inflated, assumes a uniform pressure throughout, and its response to an externally applied pressure is correspondingly uniform. A human limb, in contrast, consists of tissues of different densities (muscle tissue, adipose tissue, connective tissue) and therefore manifests nonuniform pressures throughout in response to external pressure. Consequently, one cannot state with any firm degree of confidence that the levels of pressure produced on the PQS are directly transferable to the pressures produced on a human limb. Nonetheless, this system does provide a usable model of the pressure requirements needed to compare two intensities of massage. The system is simple in construction and readily available to physical therapists working in areas of rehabilitation in which the amount of pressure exerted in the treatment of a client needs to be qualified. Furthermore, it ensures a more objective standardization of the massage technique between therapists, as well as provides feedback to both the physical therapy student learning this technique and the Clinical Instructor teaching it.

This study also examined whether there was any carryover effect upon cessation of the massage application. A mean carryover effect was detected following the cessation of DM. This effect lasted less than 10 seconds, and the H-reflex amplitude subsequently returned to its baseline value. This finding differs from those of our previous studies[7-9] in which no carryover effect was noted, but is consistent with the results obtained during muscle stretching.[28].

We examined the effect of gender on the inhibitory response observed for two reasons: (1) to rule out any suggestion of a psychological component attributable to gender being associated with the decrease m H-reflex amplitude observed and (2) to compare our results with those of our previous work, which examined the effects of massage administered by a male therapist.[8] As expected, we found that the gender of the subject had no effect in relation to the responses observed during the massage application. These results are similar to those of an earlier study[8] in which the gender of the therapist was opposite to that of the therapist in this study. The fact that both studies found no significant difference with respect to gender suggests that neither the gender of the therapist nor that of the subject plays a role in the inhibitory response recorded. Thus, it is possible to conclude that the reduction in H-reflex amplitude produced during the application of massage is not related to any differential effect of gender or to any psychological effect related to gender.

Clinical Implications

The PQS described in this study could be useful to the clinician who requires feedback as to the pressurres being exerted during manual therapeutic techniques in order to control the amount of pressure being exerted. The amount of pressure applied during the application of massage, although only one of the many factors contributing to the effects produced, can be of importance. This importance lies not only in the comfort and safety of the procedure, but also in relation to the specificity of the results obtained. Because the reduction of spinal motoneuron excitability is pressure dependent, it is critical in the training of the therapist to produce a pressure that is both clinically effective, yet comfortable and safe, especially in the treatment of a patient who has a sensory deficit. One of the underlying purposes of the construction of the PQS was to enable us to establish the lowest, most clinically effective pressure at which to apply this technique to persons with insensate skin secondary to a spinal cord injury in order to prevent damage to the skin and atrophied muscle mass.

Conclusion

The results of this study demonstrate that the application of massage to the muscles of the calf produces a decrease in the peak-to-peak amplitude of the H-reflex in comparison with values obtained during control conditions. They also show that this decrease is pressure dependent, with deeper pressures producing a greater inhibitory response than lighter pressures. The inhibitory response was sustained for both levels of massage throughout the entire period of massage application. Only DM showed some carryover inhibition after cessation of the massage application, but this effect was transitory, with a return to baseline levels of H-reflex amplitude with the first 10 seconds. The results of this study can serve to define massage characteristics in an ongoing investigation of the effects of massage in persons with a lesion of the spinal cord.

(*) The Jobst Institute Inc, 653 Miami St, PO BOX 653, Toledo, OH 43694. (dagger) Wika Instruments Canada Ltd, 9783-4 Ave, Edmonton, Alberta, Canada T6E 5V8. (double dagger) Graphic Controls Canada Ltd, 215 Hebert, Gananoque, Ontario, Canada K7G 2Y7. (sub-section) Grass Medical Instruments Co. 101 Old Colony Ave, Quincy, MA 02169. (~~) Dantech, 100 Dynamic Dr, #103, Scarborough, Ontario, Canada N1V 5C4. (#) SYSTAT Inc, 1800 Sherman Ave, Evanston, IL 60201.

References

[1] Lance JW. Symposium synopsis. In: Feldman RG, Young RR, Koella WP, eds. Spasticity: Disordered Motor Control Chicago, Ill: Year Book Medical Publishers Inc; 1980. [2] Umphred DA. Neurological Rehabilitation. Toronto, Ontario, Canada: CV Mosby Co; 1984. [3] Kukulka CG, Fellows WA, Oehlertz JE, Vanderwilt SG. Effect of tendon pressure on alpha motoneuron excitability Phys Ther 1985;65:595-600. [4] Kukulka CG, Beckman SM, Holte JB, Hoppenworth PK Effects of intermittent tendon pressure on alpha motoneuron excitability. Phys Ther. 1986;66:1091-1094. [5] Leone JA, Kukulka CG. Effects of tendon pressure on alpha motoneuron excitability in patients with stroke. Phys Ther. 1988;68: 475-480. [6] Robinson KL, McComas AJ, Belanger AY. Control of the soleus motoneuron excitability during muscle stretch in man. J Neurol Neurosurg Psychiatry. 1982;45:699-704. [7] Morelli M, Seaborne DE, Sullivan SJ. Changes in H-reflex amplitude during massage of the triceps surae in healthy subjects. Journal of Othopaedic and Sports Physical Therapy. 1990;12:55-59. [8] Morelli M, Seaborne DE, Sullivan SJ. H-reflex modulation during manual muscle massage of human triceps surae. Arch Phys Med Rehabil. 1991;72:915-919. [9] Sullivan SJ, Williams LRT, Seaborne DE, Morelli M. Effects of massage on alpha motoneuron excitability. Phys Ther. 1991; 71:555-560. [10] Belanger AY, Morin S, Pepin P, et al. Manual muscle tapping decreases soleus H-reflex amplitude in control subjects. Physiotherapy Canada. 1989;41:192-196. [11] Schieppati M. The Hoffman reflex: a means of assessing spinal reflex excitability and its descending control in man. Prog Neurobiol. 1987;28:345-376. [12] Pedneault M, Blanchard Y, Lapalme S. Pression manuelle cutanee: etude de fidelite inter et intra-therapeutes. Physiotherapy Canada. 1990;42:130--135. [13] Stijns HJ. Recherche sur l'objective de la pression manuelle de massage. Werken van de BWVK-Travaux de la SSBK 1978;4:5". [14] Watson MJ, Burnett M. Equipment to evaluate the ability of physiotherapists to perform graded posterior-anterior cervical vertebral pressure type passive movements of the spine by thumb pressure. Physiotherapy. 1990; 76:611-614. [15] Olsen VL. Evaluation of joint mobilization treatment: a method. Phys Ther. 1987; 67:351-356. [16] Lee M, Moseley A, Refshauge K Effect of feedback on learning a vertebral joint mobitization skill. Phys Ther. 1990;70:97-104. [17] Sabbahi MA, Sedgewick EM. Age-related changes in monosynaptic reflex excitability. J Gerontology. 1982;37:24-32. [18] Bulbulian R, Darabos BL. Motoneuron excitability the H-reflex following exercise of high and low intensity. Med Sci Sports Exerc. 1985;18:697-702. [19] Eke-Okoro ST. The H-reflex studied in the presence of alcohol, caffeine, force and fatigue. Electromyogr Clin Neurophysiol. 1982;22: 579-589. [20] Tappan FM. Healing Massage Technique: Holistic, Claw and Emerging Methods. 2nd ed. East Norwalk, Conn: Appleton & Lange; 1988. [21] Hollis M. Massage for Therapies. Oxford, England: Blackwell Scientific Publications Ltd; 1987. [22] Morelli M, Sullivan SJ, Seaborne DE. Comparison of human triceps surae H-reflexes obtained from mid and distal recording sites. Electromyogr Clin Neurophysiol 1990; 30: 181-186. [23] Hayes KC, Sullivan J. Tonic neck reflex influence on tendon and Hoffmann reflexes in man. Electromyogr Clin Neurophysiol. 1976;16:250-261. [24] Sabbahi MA, Deluca CJ. Topical anesthesia: modulation of the monosynaptic reflexes by desensitizing the skin. Electroencephalogr Clin Neurophysiol. 1982;54:677-688. [25] Wolf SL, Minkowitz JA. Topical anesthetic - effects on the Achilles tendon and H-reflexes, I: able-bodied subjects. Arch Phys Med Rebabil 1989;70:531-536. [26] Sabbahi MA, Deluca CJ, Powers WR, Topical anaesthesia: a possible treatment method for spasticity. Arch Phys Med Rebabil. 1981;62:310-314. [27] Woff SL, Minkwitz JA. Topical anesthetics - effects on the Achilles tendon and H-reflexes, II: stroke patients. Arch Phys Med Rehabil. 1989;70:673-677. [28] Etnyre BR, Abraham LD. H-reflex changes during stretching and two variations of proprioceptive neuromuscular facilitation techniques. Electroencephalogr Clin Neurophysiol. 1986;63:174-179. [29] Burke D, Andrews C, Ashby P. Autogenic effects of static muscle stretch in spastic man. Arch Neurol. 1971;25:367-372.