Brain (1976), 99, 375-386
LANDING FROM AN UNEXPECTED FALL AND A VOLUNTARY STEP by RICHARD GREENWOOD and ANTHONY HOPKINS (From the Department of Neurology, St. Bartholomew's Hospital, London ECJA 7BE)
THE control of movement during stepping has been analysed extensively in animals. Using decerebrate cats Orlovsky (1972) has elucidated integrating mechanisms originating in brain-stem centres. Engberg and Lundberg (1969) have shown how this activity may be superimposed on spinal reflexes. In 1971 Melvill Jones and Watt (1971a) reported observations on EMG activity of gastrocnemius in man during a single downward step. They showed that there is a build up of EMG activity prior to contact with the ground, suggesting that deceleration on landing is brought about by a preprogrammed response rather than by activity resulting reflexly from any mechanical stretch on landing. They went on to examine muscle activity evoked in gastrocnemius in response to a short unexpected fall (1971ft). They found a burst of activity beginning about 75 ms after the onset of fall. They suggested that this activity originated in the otolith apparatus and inferred that it was concerned in the preprogrammed control of landing; after about 195 ms voluntary control of muscle activity would result in a 'truly comfortable, well-controlled landing.' Melvill Jones (1973) has thus gone on to suggest that the timing of pre-contact extensor activity during a downward step, hopping or running might depend in part on vestibular cues of vertical head movement which occurs during these activities (Muybridge, 1901). We have shown (Greenwood and Hopkins, 1974) that muscle activity in soleus during fall occurs as two peaks. We have shown that the onset of the first peak is related to release and suggested that it is analogous to a startle response, probably originating in the otolith apparatus, as it is modified in subjects with absent labyrinthine function (Greenwood and Hopkins, 1976). Watt (1976) has confirmed this observation in labyrinthectomized cats. We also showed (1976) that, by contrast to the initial peak, the second peak of activity during fall is related to landing and accurately timed to occur prior to the moment of landing. Furthermore the initial peak is absent if the subject is allowed to release himself. It thus seemed unlikely that the initial peak is similar to pre-contact activity during a voluntary downward step.
Downloaded from http://brain.oxfordjournals.org/ by guest on March 21, 2016
INTRODUCTION
R. GREENWOOD AND A. HOPKINS
376
We here report further results on the characteristics of the second peak of activity during a downward fall and compare it with pre-contact activity during a downward step. METHODS
LANDING
FIG. 1. The apparatus used to study muscle activity during unexpected falls in man. Separation of the magnet triggered the crystal-controlled oscillator. Counterbalancing masses (M) could be added to the system to reduce the acceleration during fall.
Downloaded from http://brain.oxfordjournals.org/ by guest on March 21, 2016
The apparatus used to suspend subjects, to time release and landing and to record EMG activity during the fall has been previously described (Greenwood and Hopkins, 1976). In the present experiments the system was sometimes modified, as shown in fig. 1, by a counterbalancing mass (M) to reduce acceleration during the fall. In experiments without a counterbalancing mass the time taken for a fall from a given height was found to be slightly greater than the calculated time for free fall from that height using Newtonian physics. It was found that the duration of the fall from a given height was the same for both a subject and for lead weights of the same mass. We therefore concluded that acceleration during a fall was slightly decreased by friction and inertia within the system, and not by relative movement of the limbs of the subject during fall. A body at rest is acted on in the longitudinal (z) axis by the acceleration of gravity (go = 9-81 ms J )Large letter G is a symbol recommended by Gell and his panel (1961) as a unit to express inertial resultant to whole body acceleration in multiples of the magnitude of the acceleration of gravity, so at rest the body is exposed to an acceleration of +1G X . Footward acceleration produces a negative Gz, and the inertial resultant is the difference between the two. Because of the degree of friction and inertia within the system the maximal footward acceleration obtained was about 8-8 ms~2 (about 0-9 go). The inertial resultant acceleration is therefore + 1GZ —0-9 G I = 0 1 G r . Acceleration during fall was always calculated using the elapsed time and the
LANDING AFTER DOWNWARD MOTION
377
RESULTS
Unexpected Falls
During an unexpected fall lasting for more than 200 ms (200 mm) EMG activity in soleus follows a constant pattern as previously described (Greenwood and Hopkins, 1974). There is a period of electrical silence after release; then follows an initial burst or peak of activity which is a startle response related to release; then follows, after a variable interval, a second peak related to landing. A typical experiment is shown in fig. 2. If the fall was from a height of less than 200 mm no clear second peak was seen. The characteristics of both peaks were examined during falls at reduced acceleration. (a) Five subjects were examined during falls from a constant height (500 mm) at various accelerations with a counterbalancing mass, and 5 during falls from various heights at a constant acceleration of approximately 3-9 m s J (+0-6 GJ. Our findings are illustrated in figs. 3 and 4. (i) The timing of the onset, maximum and termination of the initial peak was the same during falls at accelerations from 8-8 ms~2 down to about 4-9 ms~2 (+0-1 G z to +0-5 G J (fig. 3A). At accelerations less than 4-9 ms~2 activity tended to occur only between 140 and 180 ms, the timing of the maximum, until, during falls at accelerations less than about 20 m s J (+0-8 G J, no initial peak was seen. The onset of the second peak was often ill-defined but its maximum was related to the timing of landing, as in free fall, whether the duration of the fall depended
Downloaded from http://brain.oxfordjournals.org/ by guest on March 21, 2016
height of the fall, but an approximate estimate of the desired mass (M) to produce any desired acceleration (a) could be obtained from the formula :• (M.-M).g, M.+M where M, is the mass of the subject and go is the acceleration of gravity. Auditory and visual cues of release were excluded in some experiments by a randomly varied electronic sound delivered through headphones and by a blindfold. In some experiments while the subject was blindfold the height of the fall could be shortened or lengthened to 300 or 500 mm by introducing or removing a platform beneath the subject without warning. In 5 subjects soleus EMG was recorded through silver disc electrodes during repeated single downward steps of 150 mm down on to a 200 mm platform. EMG activity was also integrated (Devices Signal Processor and Audio Unit, type 4010) resetting the integrator every 10 ms. In these experiments, a variable distance above the platform a light beam was broken by the subject's descending foot. This signal was used to trigger a Digitimer which controlled the remaining systems. The moment of contact was recorded by means of contact plates on the subject's shoe and the platform. Sometimes when visual and auditory cues were removed as above, the platform was removed without warning in between downward steps so that the step was unexpectedly increased to 350 mm. Each subject performed 20 steps at about one step every three seconds. Visual and auditory cues were then excluded and he performed a further 20 steps. He was then informed that the platform would be removed unexpectedly at intervals and continued stepping until 10 unexpectedly long steps had been performed. The Ethical Committee of St. Bartholomew's Hospital approved these experiments and subjects gave their informed consent.
378
R. GREENWOOD AND A. HOPKINS
SOLEUS EMG
0.5mV
t
100ms ONSET OF FALL
IDII LANDING
FIG. 2. Upper trace: soleus EMG during unexpected free fall from 800 mm. Lower trace: integrated activity from the same muscle. Integrator reset every 20 ms.
upon varying the acceleration (fig. 3A) or the height (fig. 3B) of the fall. This maximum always occurred between 40 and 140 ms prior to the moment of landing, as during free fall. Thus there was a period of decreasing EMG activity just prior to landing. (ii) The amount of integrated soleus activity in the initial peak, and the amplitude of the maximum of the second peak (chosen because the start of the second peak was often ill-defined) were less in falls at lesser accelerations (fig. 4A and B). At accelerations below about 20 ms~2 (+0-8 G J no initial peak was seen whereas the second peak was always seen to occur prior to landing, even at accelerations of less than 20 ms-2. (iii) The amount of activity in the initial peak and the amplitude of the maximum of the second peak were not related to the height of fall at constant accelerations (fig. 3B). (b) To enable us to investigate the effect of restricting afferent inputs on the second peak, good separation of the first and second peaks was required. This was done by using falls at reduced accelerations (3-9 ms-2, +0-6 G J rather than high free falls, so as to avoid excessive discomfort to the subject on landing.
Downloaded from http://brain.oxfordjournals.org/ by guest on March 21, 2016
50mV
INTEGRATED SOLEUS EMG
379
LANDING AFTER DOWNWARD MOTION
B
> E w
O
50
LU
2 40
CO
CO
UJ
3 30
O
tQ 20
CO Q LU
LU
LANDING FROM AN UNEXPECTED FALL AND A VOLUNTARY STEP by RICHARD GREENWOOD and ANTHONY HOPKINS (From the Department of Neurology, St. Bartholomew's Hospital, London ECJA 7BE)
THE control of movement during stepping has been analysed extensively in animals. Using decerebrate cats Orlovsky (1972) has elucidated integrating mechanisms originating in brain-stem centres. Engberg and Lundberg (1969) have shown how this activity may be superimposed on spinal reflexes. In 1971 Melvill Jones and Watt (1971a) reported observations on EMG activity of gastrocnemius in man during a single downward step. They showed that there is a build up of EMG activity prior to contact with the ground, suggesting that deceleration on landing is brought about by a preprogrammed response rather than by activity resulting reflexly from any mechanical stretch on landing. They went on to examine muscle activity evoked in gastrocnemius in response to a short unexpected fall (1971ft). They found a burst of activity beginning about 75 ms after the onset of fall. They suggested that this activity originated in the otolith apparatus and inferred that it was concerned in the preprogrammed control of landing; after about 195 ms voluntary control of muscle activity would result in a 'truly comfortable, well-controlled landing.' Melvill Jones (1973) has thus gone on to suggest that the timing of pre-contact extensor activity during a downward step, hopping or running might depend in part on vestibular cues of vertical head movement which occurs during these activities (Muybridge, 1901). We have shown (Greenwood and Hopkins, 1974) that muscle activity in soleus during fall occurs as two peaks. We have shown that the onset of the first peak is related to release and suggested that it is analogous to a startle response, probably originating in the otolith apparatus, as it is modified in subjects with absent labyrinthine function (Greenwood and Hopkins, 1976). Watt (1976) has confirmed this observation in labyrinthectomized cats. We also showed (1976) that, by contrast to the initial peak, the second peak of activity during fall is related to landing and accurately timed to occur prior to the moment of landing. Furthermore the initial peak is absent if the subject is allowed to release himself. It thus seemed unlikely that the initial peak is similar to pre-contact activity during a voluntary downward step.
Downloaded from http://brain.oxfordjournals.org/ by guest on March 21, 2016
INTRODUCTION
R. GREENWOOD AND A. HOPKINS
376
We here report further results on the characteristics of the second peak of activity during a downward fall and compare it with pre-contact activity during a downward step. METHODS
LANDING
FIG. 1. The apparatus used to study muscle activity during unexpected falls in man. Separation of the magnet triggered the crystal-controlled oscillator. Counterbalancing masses (M) could be added to the system to reduce the acceleration during fall.
Downloaded from http://brain.oxfordjournals.org/ by guest on March 21, 2016
The apparatus used to suspend subjects, to time release and landing and to record EMG activity during the fall has been previously described (Greenwood and Hopkins, 1976). In the present experiments the system was sometimes modified, as shown in fig. 1, by a counterbalancing mass (M) to reduce acceleration during the fall. In experiments without a counterbalancing mass the time taken for a fall from a given height was found to be slightly greater than the calculated time for free fall from that height using Newtonian physics. It was found that the duration of the fall from a given height was the same for both a subject and for lead weights of the same mass. We therefore concluded that acceleration during a fall was slightly decreased by friction and inertia within the system, and not by relative movement of the limbs of the subject during fall. A body at rest is acted on in the longitudinal (z) axis by the acceleration of gravity (go = 9-81 ms J )Large letter G is a symbol recommended by Gell and his panel (1961) as a unit to express inertial resultant to whole body acceleration in multiples of the magnitude of the acceleration of gravity, so at rest the body is exposed to an acceleration of +1G X . Footward acceleration produces a negative Gz, and the inertial resultant is the difference between the two. Because of the degree of friction and inertia within the system the maximal footward acceleration obtained was about 8-8 ms~2 (about 0-9 go). The inertial resultant acceleration is therefore + 1GZ —0-9 G I = 0 1 G r . Acceleration during fall was always calculated using the elapsed time and the
LANDING AFTER DOWNWARD MOTION
377
RESULTS
Unexpected Falls
During an unexpected fall lasting for more than 200 ms (200 mm) EMG activity in soleus follows a constant pattern as previously described (Greenwood and Hopkins, 1974). There is a period of electrical silence after release; then follows an initial burst or peak of activity which is a startle response related to release; then follows, after a variable interval, a second peak related to landing. A typical experiment is shown in fig. 2. If the fall was from a height of less than 200 mm no clear second peak was seen. The characteristics of both peaks were examined during falls at reduced acceleration. (a) Five subjects were examined during falls from a constant height (500 mm) at various accelerations with a counterbalancing mass, and 5 during falls from various heights at a constant acceleration of approximately 3-9 m s J (+0-6 GJ. Our findings are illustrated in figs. 3 and 4. (i) The timing of the onset, maximum and termination of the initial peak was the same during falls at accelerations from 8-8 ms~2 down to about 4-9 ms~2 (+0-1 G z to +0-5 G J (fig. 3A). At accelerations less than 4-9 ms~2 activity tended to occur only between 140 and 180 ms, the timing of the maximum, until, during falls at accelerations less than about 20 m s J (+0-8 G J, no initial peak was seen. The onset of the second peak was often ill-defined but its maximum was related to the timing of landing, as in free fall, whether the duration of the fall depended
Downloaded from http://brain.oxfordjournals.org/ by guest on March 21, 2016
height of the fall, but an approximate estimate of the desired mass (M) to produce any desired acceleration (a) could be obtained from the formula :• (M.-M).g, M.+M where M, is the mass of the subject and go is the acceleration of gravity. Auditory and visual cues of release were excluded in some experiments by a randomly varied electronic sound delivered through headphones and by a blindfold. In some experiments while the subject was blindfold the height of the fall could be shortened or lengthened to 300 or 500 mm by introducing or removing a platform beneath the subject without warning. In 5 subjects soleus EMG was recorded through silver disc electrodes during repeated single downward steps of 150 mm down on to a 200 mm platform. EMG activity was also integrated (Devices Signal Processor and Audio Unit, type 4010) resetting the integrator every 10 ms. In these experiments, a variable distance above the platform a light beam was broken by the subject's descending foot. This signal was used to trigger a Digitimer which controlled the remaining systems. The moment of contact was recorded by means of contact plates on the subject's shoe and the platform. Sometimes when visual and auditory cues were removed as above, the platform was removed without warning in between downward steps so that the step was unexpectedly increased to 350 mm. Each subject performed 20 steps at about one step every three seconds. Visual and auditory cues were then excluded and he performed a further 20 steps. He was then informed that the platform would be removed unexpectedly at intervals and continued stepping until 10 unexpectedly long steps had been performed. The Ethical Committee of St. Bartholomew's Hospital approved these experiments and subjects gave their informed consent.
378
R. GREENWOOD AND A. HOPKINS
SOLEUS EMG
0.5mV
t
100ms ONSET OF FALL
IDII LANDING
FIG. 2. Upper trace: soleus EMG during unexpected free fall from 800 mm. Lower trace: integrated activity from the same muscle. Integrator reset every 20 ms.
upon varying the acceleration (fig. 3A) or the height (fig. 3B) of the fall. This maximum always occurred between 40 and 140 ms prior to the moment of landing, as during free fall. Thus there was a period of decreasing EMG activity just prior to landing. (ii) The amount of integrated soleus activity in the initial peak, and the amplitude of the maximum of the second peak (chosen because the start of the second peak was often ill-defined) were less in falls at lesser accelerations (fig. 4A and B). At accelerations below about 20 ms~2 (+0-8 G J no initial peak was seen whereas the second peak was always seen to occur prior to landing, even at accelerations of less than 20 ms-2. (iii) The amount of activity in the initial peak and the amplitude of the maximum of the second peak were not related to the height of fall at constant accelerations (fig. 3B). (b) To enable us to investigate the effect of restricting afferent inputs on the second peak, good separation of the first and second peaks was required. This was done by using falls at reduced accelerations (3-9 ms-2, +0-6 G J rather than high free falls, so as to avoid excessive discomfort to the subject on landing.
Downloaded from http://brain.oxfordjournals.org/ by guest on March 21, 2016
50mV
INTEGRATED SOLEUS EMG
379
LANDING AFTER DOWNWARD MOTION
B
> E w
O
50
LU
2 40
CO
CO
UJ
3 30
O
tQ 20
CO Q LU
LU
