Depression of Hoffmann reflexes following voluntary contraction and implications for proprioceptive neuromuscular facilitation therapy.

Proprioceptive neuromuscular facilitation (PNF) relaxation techniques often utilize voluntary contraction by patients in an attempt to induce a subsequent phase of muscle relaxation. [1] Such voluntary contraction commonly is used to precede manual stretching of the same muscle group (eg, as in the hold-relax technique) in the hope of reducing stretch reflex responses and thereby minimizing resistance to muscle elongation. [2]

Previous research has revealed both excitatory and inhinitory effects on neurons and muscles in response to prior activity. Excitatory effects include (1) postcontraction sensory discharge (from muscle spindles); (2) post-tetanic twitch (force) potentiation; (3) post-tetanic potentiation of monosynaptic reflexes and of motor end-plate potentials; and (4) postcontraction facilitation of stretch, tendon tap, and vibration reflexes. These effects have been previously reviewed in the literature. [3]

Inhibiutory effects were described by Sherrington and Forbes [4] as a component of successive induction. They observed that, following reflex activation of an agonist, a reversal process sometimes occurred, involving intagonist facilitation and reciprocal inhibition of the agonist. This process is called successive induction. [4] Forbes noted that the presence of either subsequent "augmentation" (facilitation) or subsequent depression (inhibition) of agonist activity was dependent on the strength of the reflex stimulus. These reflex findings were generalized to voluntary contractions during the development of PNF, in the hope of utilizing successive induction to increase agonist strength and promote antagonist relaxation.

Changes in muscle excitability are difficult to quantify, especially when the subject is at rest. The status of the motoneurons (MNs) controlling the contractile state of a muscle, however, can be monitored indirectly. Hoffmann reflexes (H-reflexes) [5,6] can ve used to measure changes in the reflex excitability of an MN pool. An electrical stimulus is delivered to a peripheral nerve, action potentials in afferent fibers result in excitatory postsynaptic potentials to the MNs, and a reflex muscle twitch may be evoked. The amplitude of the triphasic electromyographic (EMG) response associated with this twitch is measured to provide a quantitative indication of the number of motor units that were reflexly recruited. The neuronal circuitry involved in this reflex is controversial. [7]

The response of an MN pool is determined by the summation of all inputs impinging on the MNs. The Amplitude of an evoked reflex will therefore depend partly on the background excitability level of the MN pool. In the presence of a constant background input, a give stimulus should elicit a consistent reflex response. If inhibitory influences are present at the time the reflex is elicited, however, the response of the MN pool may be depressed, resulting in reflex activation of fewer MNs and a smaller reflex muscle twitch. [8] According to this reasoning, if H-reflex amplitudes are reduced following therapeutic procedures (such as PNF relaxation techniques), we can infer that these techniques haves produced an inhibition of the reflex exitability of the MN pool and its associated muscle.

We therefore used H-reflexes in our study to assess the presence of reflex inhibition following voluntary contraction. The purpose of the study was to determine the onset latency, duration, and magnitude of the postcontraction reflex depression. This information should increase the knowledgeability of the therapist in optimally applying therapeutic techniques such as the hold-relax technique by optimally timing the hold phase prior to the desired relaxation interval.

Method

Subjects

Sixteen female subjects, aged 22 to 25 years ([unkeyable]=23.6, SD=0.96), with no known neurological disease volunteered for this study. All subjects signed informed consent forms prior to participation in the study.

Procedure

Subject positioning. Each subject was positioned prone on a treatment plinth (Fig. 1), with the left foot secured by an ankle cuff and turnbuckles to a footplate instrumented with a force transducer. (*) To facilitate access to the tibial nerve, the knee was flexed about 30 degress and the ankle was plantar flexed about 15 degrees. The signal from the force plate was amplified with a high-impedance differential amplifier (+) and stored on an FM tape recorded. (++)

Electromyographic recording apparatus. Surface EMG electrodes (+) were positioned over three muscles: (1) the soleus muscle to record the H-reflex (electrode position was standardized as the lower electrode edge at one third the distance from the lateral malleolus to the fibular head on the calf posterior midline), (2) the triceps surae muscle to signal the end of voluntary plantar flexion (adjacent electrode positioned immediately proximal to the soleus muscle electrode), and (3) the tibialis anterior (TA) muscle to assess any reciprocal activity (electrode positioned over the muscle belly visible during voluntary dorsiflexion). The EMG signals were amplified by on-site preamplifiers and a high-impedance differential amplifier. The triceps surae muscle EMG signal was also "integrated" by means of electronic root-mean-square (RMS) processing. High and low frequency cutoffs of 75 Hz and 5 kHz, respectively, were used for the raw EMG signals, and a 110-millisecond time constant was used for the RMS processing.

The raw soleus muscle EMG signal was displayed on a Tektronix model 7313 analog oscilloscope (*1) for H-reflex monitoring during the experiment and recorded on FM tape (++) for later analysis. This EMG signal was also led to an audio amplifier. (*2) The raw TA muscle EMG signal was monitored on the same scope and recorded on tape.

The triceps surae muscle RMS EMG signal was led to an electronic voltage level detection (VLD) circuit that provided a positive output pulse whenever EMG voltage dropped below a preset level (ie, a mean of 80.33% [+ or -] 12.22% of voluntary contraction EMG voltage). This VLD pulse served as the signal for the end of the voluntary contraction, initiating the nerve stimulation. Both RMS EMG activity and VLD pulse signals were displayed to the subject on a Tektronix model 5110 digital oscilloscope (*1) and recorded on tape.

Nerve stimulation apparatus. An acrylic plastic stimulating electrode housing, [9] containing the cathode (a 5-mm stainless-steel ball covered by moistened gauze and felt), was strapped to the subject's left popliteal fossa, over the tibial nerve. This device ensured stability of electrode location and pressure on the nerve. The anode was a 5- X 5-cm moistened pad and metal plate placed under the anterior left thigh.

The H-reflex stimulus consisted of 1-millisecond square-wave pulses delivered by a Grass S88 stimulator (*2) through a Grass SIU5 stimulus isolation unit. (*2) Initiation of stimulation was controlled by the VLD pulse, used as input to a D100 Digitimer, (#) which controlled timing of subsequent H-reflex stimuli. These Digitimer pulses were also recorded on tape.

Preliminary data collection. After the subject was positioned in the apparatus, each experiment began by determining the subject's maximal voluntary contraction (MVC) peak EMG voltage level for 3 seconds of isometric plantar flexion (best of 3-5 trials) to choose the target intensity for the experimental voluntary isometric plantar-flexion contractions. Enoka et al [10] have shown that there is no significant difference in H-reflex depression following contractions at 50% or 100% of MVC. Therefore, 65% to 75% of each subject's MVC peak EMG voltage level was chosen to be the target contraction level. This target range was drawn on the oscilloscope screen, and subjects were then responsible for monitoring their own triceps surae muscle EMG activity, keeping voluntary contractions within the desired range. Previous research has shown no significant difference in reflex depression for contractions lasting 2.5 or 5.0 seconds [10] or for 10 to 20 seconds [11] A contraction duration of 3 seconds (with a 1-minute rest between trials) was therefore chosen to minimize fatigue.

For each subject, an H-reflex recruitment curve was generated with the subject at rest. A stimulus intensity was selected to produce the largest H-reflex possible without an M-wave. Prior to each trial, resting H-reflexes were elicited to ensure that the reflex had returned to control levels from any postcontraction depression produced by previous trials. If necessary, small adjustments were made in stimulus intensity to maintain an appropriate resting H-reflex amplitude.

Experimental protocol. Each trial consisted of resting (control) H-reflexes, followed by a single voluntary plantar-flexion contraction, then postcontraction H-reflexes. It was desirable to test H-reflex depression at several times shortly after contraction. The MN pool, however, needs time to recover from previous stimuli. To avoid depression of the H-reflex amplitude by the preceding reflex, H-reflexes should not be elicited at intervals of less than 5 or 10 seconds. [12]

To overcome this problem and allow reflex testing at various times shortly after voluntary contraction, subjects performed multiple trials, with reflex stimulation beginning at one of five different postcontraction delays (0.05, 0.1, 0.5, 1, or 5 seconds) on each trial. Following the initial stimulus, stimuli continued every 10 seconds (post-delay interval) for 1 minute, resulting in a total of seven postcontraction H-reflexes per trial. The postcontraction times at which these seven reflexes were elicited for any given delay are shown in Table 1.

The combination of the five postcontraction delays with the seven postdelay intervals allowed assessment of H-reflexes at 35 different times post-contraction (Tab. 1). The mean precontraction reflex measurement (time 0.0) was also utilized in the data analysis, making a total of 36 reflex times. Each subject performed five trials at each of the five postcontraction delays (25 trials total) in a randomized order, providing 175 (35x5) postcontraction H-reflex measurements per subject.

Data Analysis

H-reflex data. The raw H-reflexes were digitized off-line at a rate of 20,480 samples per second by playing the FM tape onto a digital oscilloscope, using cursors to measure the peak-to-peak (maximal positive peak to maximal negative peak) reflex amplitude. To allow comparisons across subjects, the measured amplitudes were normalized by expressing them as a percentage of each subject's resting H-reflex amplitude. [10] The normalized reflex amplitudes for each postcontraction time were averaged across the five trials for all subsequent analyses.

To test the main effect of postcontraction time elapsed (the independent variable) on H-reflex amplitude (the dependent variable), a two-way (16 subjext x 36 times) analysis of variance (ANOVA) was performed, using a repeated-measures design.

In the presence of significant F values, planned post hoc paired t tests (with Bonferroni's correction for multiple comparisons) were made of least-squares means. To determine the total duration of H-reflex depression, pairwise comparisons were made of the mean precontraction H-reflex amplitude (control) with each postcontraction H-reflex amplitude (dependent variable), using only postcontraction times from 0.05 to 25 seconds (independent variable). Second, to determine the time of maximal H-reflex depression, all possible pair-wise comparisons were made of H-reflex amplitudes (dependent variable0 for postcontraction times of 0.0, 0.05, 0.1, 0.5, 1, and 5 seconds (independent variable).

Force data. Force data from five trials (all 5-second postcontraction delay trials) were analyzed for each subject to assess fatigue. Force data were digitized off-line at a rate of 20,480 Hz by playing the FM tape onto a ditigal oscilloscope, using cursors to measure (1) the reflex twitch peak force, (2) the reflex twitch half-relaxation time (time required for force to decline from peak to 50% of peak), (3) the voluntary contraction peak force, and (4) the MVC peak force.

For statistical analysis, data were averaged across subjects. A separate two-way (5 trialsx7 postcontraction reflex times) ANOVA was performed for each dependent variable (peak force and force half-relaxation time), using a repeated-measures design. All statistical analyses were performed using the Statistical Analysis System (SAS) package.(**)

Tibialis anterior muscle EMG activity was monitored in an attempt to control for possible confounding variable. If subjects had unintentionally contracted their dorsiflexors during or after their voluntary plantar-flexion contractions, reciprocal inhibitory effects from the dorsiflexors could conceivably have affected the soleus muscle H-reflex. The TA muscle EMG activity was therefore visually examined to rule out such undesirable contractions. The data revealed minimal activity, consistent with rest. These data, therefore, were not analyzed further.

Results

H-Reflex Data

H-reflex amplitudes were profoundly depressed following voluntary contraction (P<.0001) (Tab. 2). Figure 2 illustrates this postcontraction depression in a single subject's data. Mean values across subjects and trials (Tab. 3, Fig. 3) demonstrated that H-reflexes already exhibited depression by 0.05 second postcontraction (335 of control amplitudes, or a 67% decrease) and were more depressed at 0.1, 0.5, and 1 second postcontraction (17%, 19%, and 24% of control amplitudes, or 83%, 81%, and 76% decreases, respectively). Partial recovery occurred by 5 seconds postcontraction (70% of control amplitudes, or a 30% decrease), and near-total recovery (90% of control amplitudes, or a 10% decrease) occurred by 10.05 seconds postcontraction.

Post hoc tests to determine the total duration of reflex depression revealed that H-reflex amplitudes for each time from 0.05 to 5 seconds postcontraction were significantly depressed relative to control values (P<.0001), but were no longer significantly different beyond 15 seconds postcontraction (with the exception of 10.05 seconds postcontraction) (Tab. 4). Therefore, the duration of the depression was about 10 seconds. Other post hoc tests to determine the duration of the phase of maximal reflex depression revealed that H-reflex amplitudes at 0.1, 0.5, and 1 second postcontraction were not significantly different from one another (P>.05) (Tab. 5), indicating a plateau in the maximum H-reflex depression lasting from 0.1 to 1 second postcontraction.

The overall effect of intersubject variability on H-reflex amplitude was also ] significant (P<.0001) (Tab. 2), demonstrating that each subject exhibited a unique postcontraction H-reflex recovery time course. Plots of each subject's data revealed that all subjects exhibited postcontraction reflex depression, with slight variations in its time course. [13] This intersubject variability is reflected in the standard deviations of the pooled H-reflex amplitudes (Tab. 3, Fig. 3). Post hoc tests were not performed on this variable. We believe that reliability of the H-reflex measurements has been supported in the literature. [14,15]

Force Data

The mean reflex muscle twitch peak force was equivalent to 5.92% (SD=3.23%) of the MVC peak force. The ANOVA revealed no significant difference in muscle twich peak force over the seven postcontraction times sampled, either within single trials or between trials (P>.05) (Tab.6). The mean muscle twitch half-relaxation time was 69.22 milliseconds (SD=18.62), with no significant difference in this variable over time within single trials or between trials (P>.05) (Tab.7). We did not examine the reliability of these measurements within the configuration of this study, but similar measurements are routinely obtained in our laboratory with a high level of reproducibility.

Discussion

Magnitude of Reflex Depression

The intense postcontraction depression of H-reflex amplitudes observed in this study provides support for the theory underlying PNF relaxation techniques. These results also support previous findings of postcontraction H-reflex depression. [10,11]

It was difficult to compare the magnitude of maximal H-reflex depression found in this study with that of other investigations. Although these data were not provided in the text of the reports of the previous studies, the magnitude of depression can be inferred from their illustrations. Figure 2 in the article by Enoka et al [10] indicates a maximum postcontraction depression to about 35% of control amplitudes (65% decrease). Figure 2 in Schieppati and Crenna's [11] article indicates depressed reflex amplitudes approaching 0% of control amplitudes (complete suppression) for the fastest contraction relaxation time (0.3 second), but depression to only 45% of control amplitudes (55% depression) for show releases (2.5 seconds). Rate of relaxation following contraction was not controlled in the study of Enoka et al or in our study.

In our study, average maximal post-contraction H-reflex depression reached 16.7% of control amplitudes (83.3% decrease), despite a conservative interpretation of the data. Raw H-reflex were often observed to be completely absent (100% depression) immediately following contraction (0.05-1 second). The presence of H-reflexes with amplitudes smaller than the resolution of the equipment, however, could not be ruled out. Therefore, rather than assigning a zero to such amplitudes, a value corresponding to the equipment noise level (0.1 mV) was used. This practice was deemed necessary to avoid over-estimating the magnitude of H-reflex depression, but this conservative treatment also probably underestimated the actual magnitude of the depression effect. The magnitude of postcontraction reflex depression may thus be even greater than indicated by the statistical results (between 83.3% and 100% depression).

Time Course of Reflex

Depression

Out study extends earlier investigations by examining reflex changes at shorter latencies following contraction, demonstrating that reflex depression begins immediately (ie, 0.05 second) after contractile EMG activity declines. The duration of the phase of maximal reflex inhibition, however, is very brief. In our study, the period of maximum H-reflex depression lasted from 0.1 to 1 second postcontraction, with recovery to 70% of control reflex amplitudes already occurring within 5 seconds postcontraction.

Schieppati and Crenna's [11] illustrations indicate that the reflex had recovered to 50% within 1 second (in the case of rapid relaxation, lasting 0.3 second). Recovery to this level was delayed for about 3.5 seconds with slower relaxations (lasting 1 second). It was impossible to determine the duration of significant reflex depression in the study by Enoka et al, [10] because they did not sample the reflex at any times between their first stimulus (ie, 1-2 seconds postcontraction) and 5 seconds later, by which time the reflex had transiently recovered beyond control levels. From these results, it is important for therapists to recognize that significant postcontraction reflex inhibition is generally brief, probably present for only a few seconds.

Reflex depression recedes as recovery occurs. The time course of the recovery from postcontraction reflex depression, however, differed among our study and the studies of Enoka et al [10] and Shieppati and Crenna [11]. In our study, H-reflex amplitudes gradually recovered over the course of 10 to 15 seconds. From a minimum amplitude at 0.1 to 1 se4cond postcontraction, they reached 70% of control amplitudes by 5 seconds postcontraction and 90% of control amplitudes by 10.05 seconds postcontraction. Reflex depression of 82% to 86% of control amplitudes was present from 10.1 to 11 seconds postcontraction, but amplitudes remained above 90% thereafter for 65 seconds.

These duration results are in closest agreement with those of Schieppati and Crenna, [11] who also observed a gradual recovery, with no long-lasting (beyond 10 seconds) reflex depression. In contrast, Enoka et al [10] reported a transient recovery (to 107% of control values) at approximately 6 to 7 seconds postcontraction, followed by a plateau period of mild reflex depression (reflex amplitude about 80% of control values) from 10 to 30 seconds postcontraction, then gradual recovery from 30 to 45 seconds postcontraction. These different results may perhaps be explained by Schieppati and Crenna's observation that the time course of reflex depression is closely related to rate of relaxation from the preceding voluntary contraction. Slower relaxation was associated with less reflex depression and with more prolonged recovery times.

Neurophysiological Mechanisms

Several neurological processes potentially might be considered to be involved in postcontraction reflex depression, including (1) MN after-hyperpolarization (AHP), (2) recurrent inhibition, (3) muscle spindle pause, (4) Golgi tendon organ autogenic inhibition, and (5) presynaptic inhibition. Although our study did not specifically address these issues, the potential rile of each will briefly be considered.

Motoneuron AHP occurs following an action potential, making the MN less responsive to stimuli applied during this interval. The AHP in soleus muscle MNs produced by single action potentials lasts only about 90 milliseconds. [17] However, summation of AHPs from repetitive action potentials, such as would occur during voluntary contraction, would increase the magnitude and duration of MN depression over that produced by single potentials. [18,19]

Force Data

The force data provided the ability to rule out neuromuscular fatigue as a possible cause of postcontraction reflex depression. Fatigue is normally associated with a change in peak muscle twitch force and with a slowing of muscle twitch half-relaxation time. [16] Because neither of these processes was observed in the force data of our study, fatigue was not considered to be present and therefore could not be responsible for producing postcontraction reflex depression.

Therapeutic Implications

The findings of our study indicate that PNF relaxation techniques must be performed rapidly if the desired inhibitory effect is to be achieved. These results are compatible with current practice of techniques such as the hold-relax technique. The patient is usually taken to the end of his or her passive range of motion and resistance applied to a voluntary "hold" contraction, followed by an increase in the manual stretch. The results of our study indicate that these increments in stretch should be applied immediately after the voluntary contraction, preferably within the first second and certainly by 5 seconds postcontraction. The voluntary contraction can then be repeated to renew the inhibition, followed by successive increments in stretch. Therapists may also wish to consider that faster relaxation of the contraction appears likely to produce greater inhibition. [11] Although our study was performed with the subjects in the prone position, postcontraction reflex inhibition has also been observed with subjects in the sitting position. [10] The results of our study, therefore, may be generalized beyond our experimental protocol.

Limitations

The results of our study must be applied with at least two limitations in mind. The first limitation is that the relationship between H-reflex amplitude and clinical assessment of muscle excitability has not been established. Although it seems reasonable that changes in the electrically elicited H-reflex would reflect the same physiological phenomena underlying clinically observed changes in muscle excitability, the interpretation of the H-reflex is controversial. [7]

The second limitation is that our study was performed on neurologically normal subjects. The results are therefore generalizable only to neurologically normal patients, such as patients with only orthopedic disorders. Neurologically impaired patients may exhibit very different time patterns of postcontraction reflex depression. Nevertheless, therapists should consider these findings in planning their use of PNF treatment.

Motoneuron discharge also revokes recurrent inhibition (RI). Recurrent collaterals of the MN axon excite Renshaw interneurons that produce inhibition of the previously active MN pool. [20] This process is apparently more strongly influenced by activation of large MNs, preferentially suppressing small MNs. [21] Because small MNs are most responsive to stretch and monosynaptic reflex input [22] )and therefore presumably H-reflex), RI from large MNs activated by voluntary contraction could easily depress H-reflex excitability. Such depression would be most likely to occur following strong contractions that activate large MNs. Hultborn and Pierrot-Deseilligny [23] have determined that both AHP and RI play a role in the depression of H-reflexs produced by a submaximal conditioning stimulus at rest. These mechanisms, however, do not appear to be involved in postcontraction reflex depression. [13]

A postcontraction cessation of the muscle spindle afferent firing that normally occurs during contraction could also theoretically depress H-reflex amplitude. In the presence of normal gamma MN innervation, however, muscle spindle activation has been shown to persist following contraction (provided that muscle length is not increased), enhancing sensitivity to static and dynamic stretch. [24] This hypothesis is thus an unlikely explanation, because H-reflex would more likely be augmented than depressed.

Primary afferent depolarization produced by Ia afferent terminals synapsing on other Ia afferents reduces muscle spindle Ia responsiveness to reflex stimuli, thereby producing presynaptic inhibition of MNs. [25] Ia afferent firing (achieved through alpha-gamma coactivation) during voluntary contraction could thereby cause a reduction in Ia and MN responsiveness to subsequent H-reflex stimuli. [11] This mechanism could further contribute to postcontraction H-reflex depression.

Golgi tendon organs are a known source of inhibition to MNs and are strongly activated by even minimal muscle contraction. [26] Persistence of their summated inhibitory postsynaptic potentials produced during contraction, therefore, could theoretically depress MN excitability. [27] However, Golgi Ib afferent firing is depressed following contraction [24,28] and should therefore produce MN disinhibition, making H-reflex augmentation more likely than depression.

In summary, it is possible that multiple neurological mechanisms may be involved in postcontraction reflex depression. The most reasonable explanation at this time appears to be presynaptic inhibition. Muscle fatigue,

Table 1. H-Reflex Stimulus Postcontraction Times (in Seconds) (a)

                 Post-Delay Interval (s)
Postcontraction  Post-Dealy Interval (s)
Delay (s)        0     10     20     30     40     50    60
0.05             0.05  10.05  20.05  30.05  40.05  50.05 60.05
0.1              0.1   10.1   20.1   30.1   40.1   50.1  60.1
0.5              0.5   10.5   20.5   30.5   40.5   50.5  60.5
1                1     11     21     31     41     51    61
5                5     15     25     35     45     55    65
  (a) There were a total of 35 postcontraction times for
delivery of H-reflex stimuli and measurement
of H-reflex amplitude (5 delaysX7 intervals=35).  The mean
precontraction H-reflex (time 0) was
also included in the data analysis, as a control.

Table 2. Two-Way Randomized Block Analysis-of-Variance Summary for Presence of Postcontraction Depression of H-Reflexes

Source    df   SS         MS       F      P
Subjects   15   28558.64  1203.91   9.79  .0001
Time       35  316686.56  9048.19  46.55  .0001
Error     525  102051.31   194.38
Total     575  447296.91   777.91

[TABULAR DATA OMITTED]

Table 4. Post Hoc Analysis of Duration of Postcontraction Depression of H-Reflexes

Post-        H-Reflex Amplitude
contraction  Least-Squares Mean
Time (s)     (% Precontraction)   P (a)
0.0          100.00 (control)     ...
0.05          33.01               .0001 (b)
0.1           16.70               .0001 (b)
0.5          19.45                .0001 (b)
1             24.12               .0001 (b)
5             69.52               .0001 (b)
10.05        90.49                .0543
10.1          82.46               .0004 (b)
10.5          86.28               .0056 (b)
11            85.42               .0032 (b)
15            90.66               .0588
20.05         96.61               .4927
20.1         94.41                .2573
20.5         93.35                .1778
21            92.16               .1125
25            93.39               .1804
  (a) Probability value for post hoc paired t test
with Bonferroni's correction for multiple comparisons.
  (b) H-reflex amplitude at each listed time was
compared with the precontraction control amplitude
(time 0.0).  Because 15 pair-wise comparisons
were planned, the critical one-tailed
probability value, with an experiment-wise
error rate of P[is less than or equal to]05, is
(.05/15) X2 = .0067.

Table 5. Post Hoc Analysis of Time of Maximum Postcontraction Depression of H-Reflexes

Post-        H-Reflex Amplitude
contraction  Least-Squares Mean
Time (s)     (% Precontraction)   P (a)
0.0          100.00               .0001 (b)
0.05          33.01               .0010 (b)
0.1           16.70               .5772
0.5          19.45
1            24.12                .3437
5            69.52                .0001 (b)
  (a) Probability value for post hoc paired t tests
with Bonferroni's correction for multiple
comparisons.
  (b) H-reflex amplitude at each listed time was
compared with the amplitude at each time
before it.  Only five of these comparisons are
shown.  Because all 15 possible pair-wise comparisons
were planned, the critical two-tailed
probability value, with an experiment-wise
error rate of P[is lessthan or equal to].05, is (.05/15)=.0033.

Table 6. Two-Way Randomized Block Analysis-of-Variance Summary for Reflex Muscle Twitch Peak Force Data

Source     df   SS        MS    F      P (a)
Trial        4     46.10  0.12  0.08    .9882
Time         6    450.43  0.78  0.53   .8760
Trialxtime  24    323.65  0.14  0.10   1.0000
Error      464  65755.71  1.48
Total      498  66575.89  1.39
  (a) All values were nonsignificant (P[is greater than].05).

Table 7. Two-Way Randomized Block Analysis-of-Variance Summary for Reflex Muscle Twitch Half-Relaxation Time Data

Source     df   SS        MS      F         P (a)
Trial        4   2511.35  627.84  1.80      .1289
Time         6   2638.25  439.71  1.26      .2759
Trialxtime  24   5749.24  239.55  0.69      .8654
Error      328  11455.08  349.24
Total      362  12540.92  346.55
  (a) All values were nonsignifcant (P[is greater than].05).

reciprocal inhibition, muscle spindles, Golgi tendon organs, AHP, and RI do not appear to be involved, although they cannot be ruled out in our study.

Further basic science investigations of possible mechanisms underlying postcontraction reflex depression are warranted. Additional studies are also needed to determine the magnitude and time course of postcontraction reflex inhibition in patients with neurological disorders.

Conclusions

Following voluntary contraction, H-reflex amplitudes were strongly depressed (mean maximum decrease=83.3%), indicating a reduction in alpha MN reflex excitability. This depression was short-lasting, however, with maximal depression lasting less than 1 second and 70% recover occurring within 5 seconds. These results provide support for PNF relaxation technique claims of postcontraction inhibition, but indicate the need for speed when applying these techniques.

* Genisco Technology Corp, 650 Easy St, Simi Valley, CA 93065.

(+) Therapeutics Unlimited Inc, 2835 Friendship St, Iowa City, IA 52240.

(++) Hewlett-Packard Co, 1 310 Pruneridge Ave, Cupertino, CA 95014

(*1) Tektronix Inc, Howard Vollum Industrial Park, PO Box 500, Beaverton, OR 97077.

(*2) Grass Instrument Co, 101 Old Color Ave, PO Box 516, Quincy, MA 02169.

(#) Medical Systems Corp, 1 Plaza Rd, Greenville, NY 11548.

(**) SAS Computer Program, SAS Institute Inc, Gary, NC 27511.

References

[1] Voss DE, Knott M, Kabat H. The application of neuromuscular facilitation in the treatment of shoulder disabilities. Phys Ther Rev. 1955;33:536-541.

[2] Levine MG, Kabat H, Knott M, Voss DE. Relaxation of spasticity by physiological technics. Arch Phys Med. 1954;35:214-223.

[3] Hutton RS. Acute plasticity in spinal segmental pathways with use: implications for training. In: Komamoto M, ed. Neural and Mechanical Control of Movement. Kyoto, Japan: Yamaguchi-Shoten; 1984:99-113.

[4] Forbes A. Reflex inhibition of skeletal muscle. QJ Exp Physiol. 1912;5:149-186.

[6] Hugon M. Methodology of the H reflex in man. In: Desmedt JE, ed. New Developments in Electromyography and Clinical Neurophysilogy: Human Reflexes, Pathophysiology of Motor Systems, Methodology of Human Reflexes. Basel, Switzerland: S Karger AG, Medical and Scientific Publishers; 1973;3:277-293.

[7] Burke D. Mechanisms underlying the tendon jerk and H-reflex. In: Delwaide PJ, Young RR, eds. Clinical Neurophysiology in Spasticity. Amsterdam, the Netherlands: Elsevier Science Publishers BV; 1985:55-62.

[8] Lloyd DPC. Facilitation and inhition of spinal motoneurones. J Neurophysiol. 1946;9:421-438.

[9] Simon JN. Dispositif de contention des electrodes de stimulation pour l'etude de Hoffmann chez l'homme. Electroencephalogr Clin Neurophysiol [Suppl]. 1963;22:174-176.

[10] Enoka RM, Hutton RS, Eldred E. Changes in excitability of tendon tap and Hoffmann reflexes following voluntary contractions. Electroencephalogr Clin Neurophysiol. 1980; 48:664-672.

[11] Schieppati M, Crenna P. From activity to rest: gating of excitatory autogenic afferences from the relaxing muscle in man. Exp Brain Res. 1984;65:448-457.

[12] Cook WA. Effects of low frequency stimulation on the monosynaptic reflex (H reflex) in man. Neurology. 1968;18:47-51.

[13] Moore MA. Depression of Simple and Conditioned Hoffmann Reflexes Following Voluntary Contractions. Iowa City, Iowa: The University of Iowa; 1987. Doctoral dissertation.

[14] McIlroy E, Brooke JD. Within-subject reliability of the Hoffmann reflex in man. Electromyogr Clin Neurophysiol. 1987;27:401-404.

[15] Mohan A, Kukulka CG. A statistical evaluation of sampling limitations of the H-reflex. Phys Ther. 1988;68:862. Abstract.

[16] Bigland-Ritchie B, Furbush F, Woods JJ. Fatigue of intermittent submaximal contractions: central and peripheral factors. J Appl Physiol. 1986;61:421-429.

[17] Burke RE. Motor unit types of cat triceps surae muscle. J Physiol (Lond). 1967;193:141-160.

[18] Gasser HS, Grundfest H. Action and excitability in mammalian A fibers. Am J Physiol. 1936;117:113-133.

[19] Ito M, Oshima T. Temporal summation of after-hyperpolarization following a motoneuron spike. Nature. 1962;195:910-911.

[20] Renshaw B. Influence of discharge of motoneurons upon excitation of neighboring motoneurons. J Neurophysiol. 1941;4:167-183.

[21] Eccles JC, Eccles RM, Iggo A, Ito M. Distribution of recurrent inhibition among motoneurons. J Physiol (Lond). 1961;159:479-499.

[22] Henneman E, Somjen G, Carpenter DO. Excitability and inhibitability of motoneurons of different sizes. J Neurophysiol. 1965;28:599-620.

[23] Hultborn H, Pierrot-Deseilligny E. Recurrent inhibition and after-hyperpolarization following motoneuronal discharge in the cat. J Physiol (Paris). 1979;297:253-266.

[24] Smith JL, Hutton RS, Eldred E. Post-contraction changes in sensitivity of muscle afferents to sttic and dynamic stretch. Brain Res. 1974;78:193-202.

[25] Eccles JC, Schmidt RF, Willis WD. Presynaptic inhibition of the spinal monosynaptic reflex pathway. J Physiol (Lond). 1962;161:282-297.

[26] Jansen JKS, Rudjord T. On the silent period and Golgi tendon organs of the solesus muscle of the cat. Acta Physiol Scand. 1964;62:364-379.

[27] Hufschmidt HJ. The demonstration of autogenic inhibition and its significance in human voluntary movement. In: Granit R, ed. Muscular Afferents and Motor Control. New York, NY: John Wiley & Sons Inc; 1966:269-274.

[28] Nelson DL, Hunton RS. Stretch sensitivity of Golgi tendon organs in fatigued gastrocnemius muscle. Med Sci Sports Exerc. 1986;18:69-74.

M Moore, PT, PhD, is Assistant Professor, Physical Therapy program, The University of Rhode Island, Kingston, RI 02881 (USA). She was a doctoral candidate in the Graduate Program in Physical Therapy, The University of Iowa, Iowa City, IA 52242, at the time this study was conducted. Address all correspondence to Dr Moore.

C Kukulka, PhD, PT, is Associate Professor. Graduate Program in Physical Therapy, The University of Iowa.

This article was adapted from a poster presentation at the 1988 Joint Congress of the American Physical Therapy Association and the Canadian Physiotherapy Association, June 12-16, 1988, Las Vegas, Nev.

This project was supported by NIH grant #NS24991 awarded to Dr Kukulka.

This study was approved by The University of Iowa Human Subjects Committee.

This article was submitted September 8, 1989, and was accepted November 20, 1990.