| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 79 No. 4 April 1998, pp. 1702-1716
Copyright ©1998 by the American Physiological Society
Department of Physiological Science, Laboratory of Neuromotor Control, University of California, Los Angeles, California 90095-1568
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Smith, Judith L., Patricia Carlson-Kuhta, and Tamara V. Trank. Forms of forward quadrupedal locomotion. III. A comparison of posture, hindlimb kinematics, and motor patterns for downslope and level walking. J. Neurophysiol. 79: 1702-1716, 1998. To gain further insight into the neural mechanisms for different forms of quadrupedal walking, data on postural orientation, hindlimb kinematics, and motor patterns were assessed for four grades of downslope walking, from 25% (14° slope) to 100% (45°), and compared with data from level and downslope walking at five grades (5-25%) on the treadmill (0.6 m/s). Kinematic data were obtained by digitizing ciné film, and electromyograms (EMGs) synchronized with kinematic records were taken from 13 different hindlimb muscles. At grades from 25 to 75%, cycle periods were similar, but at the steepest grade the cycle was shorter because of a reduced stance phase. Paw-contact sequences at all grades were consistent with lateral-sequence walking, but pace walking often occurred at the steepest grades. The cats crouched at the steeper grades, and crouching was associated with changes in fore- and hindlimb orientation that were consistent with increasing braking forces and decreasing propulsive forces during stance. The average ranges of motion at the hindlimb joints, except at the hip, were often different at the two steepest slopes. During swing, the range of knee- and ankle-joint flexion decreased, and the range and duration of extension increased at the ankle joint to lower the paw downward for contact. During stance the range of flexion during yield increased at the ankle joint, and the range of extension decreased at the knee and metatarsophalangeal joints. Downslope walking was also associated with EMG changes for several muscles. The hip extensors were not active during stance; instead, hip flexors were active, presumably to slow the rate of hip extension. Although ankle extensors were active during stance, their burst durations were truncated and centered around paw contact. Ankle flexors were active after midstance at the steeper slopes before the need to initiate swing, whereas flexor and extensor digit muscles were coactive throughout stance. Overall the changes in posture, hindlimb kinematics, and activity patterns of hindlimb muscles during stance reflected a need to counteract external forces that would accelerate angular displacements at some joints. Implications of these changes are discussed by using current models for the neural control of walking.
The assessment of quadrupedal walking forms in the cat has proven to be an inexhaustible area to study the different functions of skeletal muscles in the production of limb motion and to gain insight into the neural mechanisms required for locomotor control. Accounts of these studies are provided in the INTRODUCTION of companion papers to this study (Carlson-Kuhta et al. 1998 Except where noted, the methods were identical to those described in the study of upslope walking in the companion paper (Carlson-Kuhta et al. 1998
Cycle period data and gait patterns: grades 25-100%
Data from four cats (cats 1 and 3-5) were used to determine the average cycle periods and hindlimb kinematics for overground, downslope walking. Several trials at each grade (25-100%) were filmed for each cat (Fig. 1). During a single trial each cat usually completed 5-6 steps, walking at its own speed down the walkway. Criteria used to select individual hindlimb steps from the film records are described in Carlson-Kuhta et al. (1998)
Posture, hindlimb orientation, and stride length: grades 25-100%
Typical postures for level and downslope walking are compared in Fig. 1. During downslope walking, particularly at the steeper grades, the cats assumed a crouched posture with the trunk lowered. We measured hip height from the walkway (Fig. 3A, Hh) at key points in the step cycle, and these measurements were used to assess the level of the hindquarter crouch. Hip height was similar at paw contact and liftoff, and average hip height decreased with increasing slope (Fig. 4, Hh). At the steepest slope, hip height decreased an average of 26% from that of level treadmill walking.
Hindlimb kinematics: grades 25-100%
Kinematic data for each hindlimb joint are illustrated in Fig. 5. For level walking and walking at the 25% grade, the patterns of angular displacement were stereotypical and similar for all cats; thus averaged records, such as those illustrated in Fig. 5, resembled kinematic records from individual steps. At the steeper slopes, however, the kinematic data for the knee and ankle joints tended to be more variable from step to step for all cats, as indicated by wider than usual standard deviation bands for the knee-joint data in Fig. 5B (50 and 100% grades) and the ankle-joint data in Fig. 5C (50% grade). The tendency for greater variability at the steeper slopes is also indicated by the frequency of SD >10° in Table 2, which lists the angular positions of the hindlimb joints at key transitions of the step cycle, and Table 3, which lists the ranges of motion for the major phases of the step cycle.
SWING.
At the onset of swing, the hip, knee, and ankle joints were more flexed at the steeper slopes (Table 2, PO) and this increased flexion was consistent with the crouched posture and the reduction in the length of the hindlimb axis at liftoff (Fig. 4, LXp). The range of hip-joint flexion at the beginning of swing was the same at all grades; however, the range of flexion decreased at the knee and ankle joints at the two steeper slopes (Table 3, B-C: F). In contrast, the range of flexion at the metatarsophalangeal (MTP) joint increased twofold from 0% to the 25% grade but increased no further at the steeper grades (Table 3D, F-swing).
STANCE.
At paw contact the hindlimb joints were more flexed at the steeper slopes (Table 2, PC); this was consistent with the crouched posture at the beginning of stance and the reduction in the length of the hindlimb axis (Fig. 4, LXa). As illustrated in Fig. 5, the hip and MTP joints extended throughout most of stance, whereas the knee and ankle joints flexed (yield or E2 phase) and then extended (E3 phase).
INTRALIMB COORDINATION.
Intralimb coordination for proximal and distal pairs of adjacent joints was assessed by angle-angle plots (e.g., cyclographs). The hip-knee cyclographs (Fig. 7, A-B) show a progression of grade-related changes. In Fig. 7A, the progressive shift to the left was due to the crouched posture and a greater degree of flexion at the knee joint, whereas the diagonal shift in Fig. 7B was due to a progressive increase in the flexed position at the hip and knee joints. In Fig. 7A, the plot for downslope walking at 50% also increased in the vertical and horizontal dimensions. These expansions demonstrate the increases in the ranges of hip-joint flexion and knee-joint extension during swing; both of these changes were typical of cat 1 but not of the other cats.
Motor patterns of hindlimb muscles: grades 25-100%
EMG data were collected from cats 1 and 3-5 during level walking on the treadmill and downslope walking on the inclined walkway. Most muscles were examined for level walking and at all four grades, but four distal muscles (FHL, FDB, EDL, and EDB) were recorded for level walking and only at two grades, 25 and 75%. For each muscle, 10-25 steps were typically analyzed for each grade. Exemplar averaged EMG traces, triggered to paw liftoff, are illustrated in Figs. 8 and 9 for proximal and distal muscles, respectively. For comparison purposes, the durations of the EMG bursts were normalized to the cycle period.
MUSCLES WITH STANCE-RELATED ACTIVITY ONLY.
Four hindlimb muscles, each with extensor functions at knee (VL) or ankle (LG, PLT, FHL) joints had stance-related activity only. During level walking, these muscles were active just before paw contact and their activity continued throughout most of stance. At increased grades of downslope walking, the amplitude of the stance-related activity decreased and the relative duration of their bursts declined from 50-60% of the step cycle at the two lower grades to 15-17% of the cycle at the two steeper grades (Figs. 8 and 9).
MUSCLES WITH PRIMARILY SWING-RELATED ACTIVITY.
Activity of the ST and FDL was typically associated with swing-related actions. The ST had two bursts of activity that centered around paw liftoff and paw contact. The paw-liftoff burst (STpo) increased in amplitude and duration, particularly at the two steeper grades (Fig. 8C). At the 25% grade the STpo burst occupied 17-20% of the cycle, and the same burst occupied 40-50% of the cycle at the 100% grade. The amplitude of the paw-contact burst (STpc) was markedly lower than the amplitude of the STpo burst. Occasionally, low-level ST activity persisted throughout stance, and there was no pause separating the stance-phase activity from the onset of the STpo burst (Fig. 8C).
MUSCLES WITH STANCE- AND SWING-RELATED ACTIVITY.
Five muscles (EDB, FDB, EDL, TA, and IP) had both stance- and swing-related activity during downslope walking. The EDB, usually inactive during the stance phase of level walking, showed a progressive increase in stance as the grade increased (Fig. 9). The FDB, conversely, had bouts of activity during swing and stance-related activity during level and downslope walking. FDB activity progressively increased in amplitude with steeper inclines of downslope walking, particularly during stance (Fig. 9).
MUSCLES WITH NO ACTIVITY.
Although the two hip extensor muscles (ABF and ASM) were active during the stance phase of level walking, neither muscle was active during downslope walking at grades of 25-100% (Fig. 8). These results were unexpected and prompted us to study downslope walking at lower grades on the treadmill to ascertain if these muscles were inactive during downslope walking at any grade or if the lack of activity was associated with grades of 25% and higher.
Downslope treadmill walking: grades 5-25%
Two cats were tested on the treadmill at five grades (5-25%). For all grades, the average cycle periods were similar and ranged from 608 to 713 ms, with 65 ± 3% of the cycle period devoted to stance. EMG records are shown in Figs. 10 and 11.
Slope-related changes in posture, gait, and hindlimb dynamics
During downslope walking the body mass must be moved downward and forward, both aided by gravity, without slipping on the inclined walkway. We covered the walkway surface with a thin, nonskid mat (see METHODS) (see also Carlson-Kuhta et al. 1998 Motor pattern changes associated with downslope walking
Based on the assumption that the requirement for propulsive forces would be less during the stance phase at the steeper slopes of downslope walking, we anticipated that extensor activity would decrease over the range of grades tested. Our EMG data, as well as force-buckle and EMG data reported by Herzog et al. (1993) MOTOR PATTERNS OF HIP-JOINT MUSCLES DURING STANCE.
The hip joint extended throughout stance of downslope walking, and the range of extension was similar at all grades tested; however, the primary, uniarticular extensor muscles of the hip joint (ABF and ASM) were inactive at all grades of overground downslope walking and at grades of treadmill downslope walking steeper than 15%. These results suggest that in contrast to level walking, extension at the hip joint may not be associated with an extensor muscle torque during the stance phase of downslope walking at grades MOTOR PATTERNS OF A KNEE-JOINT EXTENSOR DURING STANCE.
The VL was active during stance of downslope walking at all grades tested on the treadmill and the walkway, and there was little change in the amplitude or duration of the VL burst. The VL EMG data and the absence of knee flexor muscle activity during the stance phase of downslope walking suggest that the extensor muscle torque at the knee joint might have a similar profile over a wide range of downslope grades. Preliminary data from R. Gregor's laboratory (Smith et al. 1997 STANCE MOTOR PATTERNS OF DISTAL MUSCLES.
The duration of extensor muscle activity at the ankle joint (LG, PLT, and FHL) decreased markedly at the steeper grades. These data suggest that the duration of the extensor muscle torque, which predominates the kinetic profile of the ankle joint during level walking (Fowler et al. 1993 Neural control of slope walking by a multifunctional CPG
In 1981 Grillner proposed a model for a spinal central pattern generator (CPG) that included separate control units for flexor and extensor muscles at each hindlimb joint. He hypothesized that each unit contained all the neural elements required (except for tonic facilitation by descending supraspinal fibers) to generate locomotor-like bursts without sensory feedback or patterned input from supraspinal centers. One advantage of Grillner's model is that the array of units can be connected in different ways to program various walking forms or even different cyclic motions such as walking and scratching (also see Gelfand et al. 1988
CPG UNITS FOR THE CONTROL OF HIP- AND KNEE-JOINT MUSCLES.
The excitatory coupling between the hip and knee extensor units, illustrated in Fig. 12A, creates a robust extensor synergy that is typical of forward and backward walking (in contrast to Grillner's 1981 prediction) (see Buford and Smith 1990 CPG UNITS FOR THE CONTROL OF ANKLE-JOINT AND DIGIT (FOOT) MUSCLES.
The extensor muscles (LG, PLT, and FHL) of the ankle joint participate in the extensor synergy with the hip and knee extensors during forward walking, and they are reciprocally active with the flexor muscles (TA and EDL) of the ankle joint. These patterns of activity are unchanged for upslope walking; thus the circuits in Fig. 12, A and C, are similar. It is important to note that we have placed the control of the FHL and the EDL muscles in the CPG units that are consistent with their actions at the ankle joint (which are opposite to their actions at the digits). These placements are consistent with our EMG data and with Engberg's (1964) classification of the FHL as a "physiological extensor" and the EDL as a "physiological flexor."
MULTIFUNCTIONAL CPG AND FUZZY CONTROLLERS.
Although the unit-burst model may account for the reorganization of muscle synergies, it does not account for the precise timing or the relative output of activity to each muscle within a synergy. It is probable that the CPG output to motoneuron pools is reinforced by sensorimotor responses at the spinal level that can be gated by higher centers for specific tasks. For example, discharges of group I afferents from ankle extensors facilitate the recruitment of extensor muscles at the hip, knee, and ankle during locomotion induced by stimulation of the mesencephalic locomotor region, but the same discharges inhibit extensor recruitment when the cat is at rest (Guertin et al. 1995
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; Trank et al. 1996), as well as in recent reviews focused on the neural control of locomotion (Rossignol 1996; Stein and Smith 1997) and limb dynamics (Zernicke and Smith 1996). Here we detail the changes in posture, hindlimb orientation, and kinematics, as well as the motor patterns of selected hindlimb muscles associated with downslope walking at grades from 5 to 100%. These data are compared with similar data for upslope walking (Carlson-Kuhta et al. 1998) and crouched walking (Trank et al. 1996), as well as forward and backward walking on a level surface (Buford and Smith 1990; Buford et al. 1990; Perell et al. 1993; Pratt et al. 1996; Trank and Smith 1996).
; Smith et al. 1994, 1996; Trank and Smith 1995) and in one rapid publication (Smith and Carlson-Kuhta 1995).
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
). Six laboratory-raised cats [Felis domesticus; 4 male (4.0-4.5 kg) and 2 female (3.0-4.2 kg)] were trained to walk at moderate speeds (0.6-0.7 m/s) on a motorized treadmill (0.3 × 0.8 m) enclosed with Plexiglas. Four of these cats (1 and 3-5) were also subjects for the upslope study in the companion paper (same identification numbers). The cats were also trained to walk down a walkway (1.85 × 0.29 m) with Plexiglas sides (0.33 m) from an elevated level platform (0.56 × 0.72 m). The walkway was inclined at one of four different grades: 25% (14° incline; Fig. 1A), 50% (26.6°; Fig. 1B), 75% (37°), and 100% (45°; Fig. 1C).
View larger version (16K):
[in a new window]
FIG. 1.
Comparison of postures for level and downslope walking at 3 grades [25% (A), 50% (B), and 100% (C).] Posture drawings for cat 5 were traced from the first paw-contact frame of the reference hindlimb (rH; here, the left hindlimb). For each pair of tracings, the unshaded drawing for downslope walking was superimposed over the shaded figure for level walking by aligning the measurement line for Hh, a line drawn from the hip-joint marker perpendicular to the walkway surface (also see Hh in Fig. 3A).
). A Newman-Keuls posthoc test was used to test differences between cell means. All statements that indicate a quantitative difference between data sets are based on a significance level of P 
0.05.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
. The five best walking steps were analyzed for each cat at each grade; thus a total of 80 steps were assessed. The downslope data were compared with data from level treadmill walking (5 steps per cat).
View this table:
TABLE 1.
Cycle period measurements for the reference hindlimb
, was the same at all grades, the time intervals between contacts were variable, leading to a variety of support combinations during the gait cycle.
View larger version (17K):
[in a new window]
FIG. 2.
Gait diagrams for four individual steps [25% (A), 50% (B), 75% (C) and 100% (D)] from cat 1. Each diagram is referenced to the step cycle of the rH (the hindlimb facing the camera). Bars represent the stance phase for each limb. cH, contralateral hindlimb; iF, forelimb ipsilateral to rH; and cF, contralateral forelimb.
). Pace walking tended to be more common at the two steeper grades, and this tendency is illustrated in Fig. 1. At the 25% grade (Fig. 1A), the left hindpaw made contact just after the left forepaw lifted off, and the hind- and forepaw are spatially very close; this is typical of a lateral sequence walk. At the steeper grades the spatial distance between the two ipsilateral paws often increased as forepaw contact occurred soon after the hindpaw made contact (Fig. 1C); this is typical of pace walking.
View larger version (15K):
[in a new window]
FIG. 3.
Measurements of hindlimb orientation and stride length. A: limb measurements for cat 4 at the 50% grade for paw liftoff. 
, anterior iliac crest (C), hip joint (H), knee joint (K), ankle joint (A), metatarsophalangeal (MTP) joint of 5th digit (M), and distal phalanx of 4th digit (D) - - -, line segments represent the pelvis (C-H), thigh (H-K), shank (K-A), paw (A-M), and digits (M-D). Limb orientation measurements at paw liftoff are LXp [leg axis (H-D)] and Hh [hip height] perpendicular distance from walkway to H; Dp, distance from Hh line to D, parallel to the walkway; and Øp, limb axis angle at paw liftoff. B: limb orientation measurements at 0 and 100% grades are compared; data are 5 steps averaged at each grade for cat 3. Paw contact (PC) and paw liftoff (PO) measurements are distinguished by the ending letters "a" (anterior orientation) and "p" (posterior orientation), respectively. Calibration bars, 1 cm.
View larger version (22K):
[in a new window]
FIG. 4.
Changes in hindlimb orientation measurements at paw contact and liftoff for downslope walking; see Fig. 3 for abbreviations. 
, , 
, and 
: paw liftoff measurements, labeled p (posterior orientation); 
, 
, 
, and 
: paw-contact measurements, labeled a (anterior orientation). Each data point represents the average of 20 steps (5 steps/cat); all data were normalized and expressed as a percentage of the value for level walking (- - -, 100%).
View larger version (19K):
[in a new window]
FIG. 5.
Typical hindlimb kinematics for level (0% grade) and downslope walking. Graphs begin and end with paw liftoff; , average time of paw contact. - - -, SD lines for downslope data, based on 5 steps for each trace (cat 3); see Fig. 5 in Carlson-Kuhta et al. (1998) for the SD bands of the level walking data.
View this table:
TABLE 2.
Angular positions of the hindlimb joints at step cycle transitions
View this table:
TABLE 3.
Ranges of motion during phases of the step cycle
), and the range of flexion was usually small, ranging from 1 to 9°.
View larger version (17K):
[in a new window]
FIG. 6.
Flexion of the MTP joint around paw contact for downslope walking. Drawings taken from a single step at the 75% grade (cat 4); to reduce the vertical size of the illustration, the sloped walkway was reoriented to the horizontal. The 4 paw drawings were traced from alternate film frames (1 frame = 10.3 ms); 1 illustrates the plantar angle measured for the MTP. The numbered dots on the data trace correspond to the enumerated paw drawings. MTP flexion began before paw contact (PC) as the digits move downward and flexion continued into stance as the tarsals moved downward; also see text. The dashed horizontal line was added as a referent to the vertical motion of the ankle-joint dot.
View larger version (39K):
[in a new window]
FIG. 7.
Interjoint coordination for level and downslope walking at 4 grades. Angle-angle plots in A and B (hip-knee coordination) and C and D (ankle-MTP coordination) are the averages from 5 steps for each grade. For all cyclographs, the step cycle starts and ends with paw liftoff (PO) and reads in a counterclockwise direction, as indicated by the arrows; paw contact is marked by the arrowheads labeled PC. Data points are plotted for each film frame (10-ms intervals). Data in A and C are from cat 1, B and D, cat 4.
View larger version (28K):
[in a new window]
FIG. 8.
A composite of average electromyogram (EMG) records from proximal muscles. Windows of data were triggered from paw liftoff (|) and then averaged over 13-15 steps. EMG records with similar cycle periods were selected for each grade; average cycle periods were 779 ± 71 ms forlevel treadmill walking (0.5-0.6 m/s), 656 ±45 ms for 50% grade, and 442 ± 15 ms for the 100% grade. 
, estimated times for paw contact. The semitendinosus (ST), anterior biceps femoris (ABF), and lateral gastrocnemius (LG) data are from cat 1 and the iliopsoas (IP), rectus femoris (RF), anterior semimembranosus (ASM), and vastus lateralis (VL) data are from cat 3, with 10-15 steps averaged for each trace. Level walking data in A are the same as the level walking data in Fig. 8A of Carlson-Kuhta et al. (1998). For each muscle, voltage calibration bars marked on the right side are the same for all records (A-C): VL (0.3 mV), RF and LG (0.2 mV), ABF (0.15 mV), IP and ST (0.1 mV), ASM (0.075 mV). Horizontal timescale, 100-ms intervals.
View larger version (28K):
[in a new window]
FIG. 9.
A composite of average EMG records from distal muscles. Windows of data were triggered from paw liftoff (|) and then averaged over 16-26 steps. EMG records with similar cycle periods were selected for each grade; average cycle periods were 711 ± 12 ms for level treadmill walking (0.5-0.6 m/s), 567 ± 30 ms for 25% grade, and 446 ± 9 ms for the 75% grade. , estimated times for paw contact. The FHL, EDL, and TA data are from cat 3 and the FDL, PLT, FDB, and EDB data were from cat 5. The data in A are the same as in Fig. 9A of Carlson-Kuhta et al. (1998). For each muscle, voltage calibration bars marked on the right side are the same for all 3 records (A-C): PLT (0.5 mV), FHL (0.25 mV), and FDL, FDB, EDB, EDL, and TA (0.125 mV). Horizontal timescale, 100-ms intervals.
View larger version (18K):
[in a new window]
FIG. 10.
Rectified-average EMG records for level (A, 0%) and downslope walking on the treadmill (0.6 m/s) at 2 grades of downslope walking (B, 10%; C, 20%). Windows of EMG from 10 steps were averaged by triggering from the time of paw liftoff (|); data are from cat 7. For each muscle, voltage calibrations are the same for all records: VL and ASM (0.2 mV), IP (0.4 mV), and ABF (0.6 mV). Horizontal timescale, 50-ms intervals.
View larger version (20K):
[in a new window]
FIG. 11.
Example of hindlimb kinematics and EMG for downslope walking at a 5% grade on the treadmill (0.6 m/s). Two steps are taken from cat 6; the record begins with PO of the 1st step and ends with PO of the 2nd step. |, each paw liftoff; , each paw contact; 
, marked reduction of hip extensor activity during the 2nd step; see text for details. Calibrations: horizontal, 100 ms; vertical, 0.2 mV.
). In this trial for cat 7, the ABF and ASM muscles were active in some steps but not in others with no apparent relationship to the hip-joint kinematics.
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
) to eliminate paw slippage at the steepest grade that occurred in one of two cats tested in a pilot study. With the nonskid mat the paws did not slide. Nonetheless, each cat adjusted its posture and limb orientation in ways that would increase the braking forces at the steeper grades. Although we did not assess stance-phase kinetics, our data provide some useful predictions.
). For upslope walking the hindlimb stride shifted in the opposite direction, and there was a greater range of stance-phase extension at all four hindlimb joints (Carlson-Kuhta et al. 1998); thus the percent of stance devoted to propulsion most likely increased.
; Hildebrand 1976). These authors suggest that pacing is only used by cats during novel or demanding situations, because pacing is associated with a greater roll of the body from side to side and is less stable than other gaits. Our data suggest that downslope walking at the steeper grades may be one of those demanding situations. In our testing setup each cat took only five or six steps down the steep walkway, and given the short distance and the fact that each cat crouched to lower its center of mass to increase its overall stability, decreases in lateral stability associated with pace walking may not have been at issue.
), and crouched walking (Trank et al. 1996), adjustments in the orientation of the hind- and forelimbs were different and well suited to each form of locomotion. For downslope walking at the steeper slopes, the hindpaw placement was extended rostrally, but during upslope and crouched walking there was a decrease in the anterior hindpaw placement. The forepaw placement during downslope walking was extended rostrally, whereas there was no apparent change in the forepaw placement during upslope and crouched walking. It is probable that each forelimb provided only a braking force for most, if not all, of the stance phase during downslope walking. This prediction is based in part on the observation that the orientation of the cats' forelimbs for downslope and backward level walking are similar. During the stance phase of backward walking, the forelimbs produce a continuous shear component ground-reaction force that would constitute a braking force during forward walking (see Fig. 5C of Perell et al. 1993). The function of the forelimbs during downslope walking deserves a comprehensive analysis.
for level and downslope (10° incline) walking, show that the recruitment of the MG and PLT, but not the soleus, were consistent with this prediction. We did not expect hip extensor muscles to be totally inactive at most grades of downslope walking, nor did we expect hip flexor muscles to be active during stance. Furthermore, we did not anticipate that the knee extensor EMG activity would be similar in amplitude at all grades. We focus here on these three unexpected findings, as well as the co-contraction of the digit muscles.
15%. For level walking an extensor torque at the hip joint is associated with the recruitment of hip extensor muscles during the first half of stance. At midstance, however, the hip-joint torque becomes flexor and the reversal is associated with the onset of RF activity and a period in which the resultant ground-reaction vector is oriented posterior to the hip joint (see Fig. 6A of Perell et al. 1993). As a consequence of these dynamics, power is generated at the hip joint for the first half of stance and then is absorbed during the second half of stance, as the rate of hip extension is slowed (see Fig. 9A of Perell et al. 1993).
).
15%. Preliminary kinetic data from R. Gregor's laboratory (Smith et al. 1997) appear to support this prediction at three grades (25, 50, and 75%) of downslope walking. They found that a flexor muscle torque dominated the kinetic profile of the hip joint during the stance phase of downslope walking except for an extensor torque around paw contact that is brief in duration and usually small in magnitude. The finding of a flexor muscle torque is consistent with the lack of extensor muscle activity and the onset of flexor muscle activity at the hip joint during the early phase of stance.
) showed an extensor torque at the knee joint during the stance phase that was similar in magnitude and duration for three grades (25, 50, and 75%) of downslope walking. In contrast, they found that the peak muscle torque at the knee joint associated with upslope walking increased substantially for the same three grades (R. Gregor, unpublished data). These findings for upslope walking are consistent with a marked increase in VL EMG at the steeper grades of upslope walking (Carlson-Kuhta et al. 1998).
; Perell et al. 1993), occupies less of the stance phase of downslope walking, particularly at the steeper grades. The ankle flexor muscles (TA and EDL) remained reciprocally active with the ankle extensor muscles, and there was a tendency for EDL to be recruited around midstance at the steeper slopes.
), but for upslope walking at the 75% grade (Carlson-Kuhta et al. 1998), backward walking on a level treadmill (Trank and Smith 1996), and downslope walking at the 75% grade, the EDB is recruited for most of stance. The purpose of the FDB-EDB co-contraction during stance for these forms of walking is not clear and warrants further study.
). We illustrate in Fig. 12 how Grillner's unit-burst generator model can be used to account for changes in the organization of muscle synergies that distinguish downslope and upslope walking from level walking.
View larger version (25K):
[in a new window]
FIG. 12.
Three central pattern generator (CPG) configurations for the control of cat hindlimb muscles during level and slope walking. The CPGs are patterned after the model proposed by Grillner (1981) . Each large circle represents one unit-burst generator (called a unit) for the control of extensor (EX) or flexor (FX) muscles of the hip, knee, ankle, or foot joints. Striped circles represent units that act together during stance, whereas half-striped circles represent dual-phase units that generate stance-related and swing-related activity. In B, X represents lack of hip extensor activity during downslope walking, thus the inactivity of the hip extensor unit. Excitatory connections between units are represented by small open circles, whereas inhibitory connections are shown by small shaded circles. Hamstring muscles are placed in the knee flexor unit (indicated by ST). Muscles of the foot are placed according to their physiological functions described Engberg (1964); the FDB is a physiological extensor and the EDB is a physiological flexor. The placement of the EDL, FHL, and FDL, as well as other muscles hindlimb, are presented in the DISCUSSION.
), upslope walking (Fig. 12C), and crouched walking (Trank et al. 1996). During downslope walking, however, the extensor synergy is disrupted as the hip extensor muscles are not recruited at grades >20%. To model the lack of hip extensor activity at these grades, we have replaced the excitatory connections between the hip and knee extensor units with an inhibitory connection (Fig. 12B). With this altered connection and the inhibition provided by the hip flexor unit, the hip extensor unit will be inactive for most, if not all, of the step cycle (Fig. 12B, X).
modeled the two ST bursts by inhibiting the ST unit with inputs from the adjacent units (Fig. 12A); thus the ST unit was released from inhibition during a brief period at the end of stance and another at the end of swing. We could not use the same solution to model the two IP bursts because one burst coincided with the main period of extensor activity and the other with the main period of flexor activity. To model this we have added a diagonal excitatory connection between the knee extensor and the hip flexor units while maintaining the excitatory connections between the ankle and hip flexor units (Fig. 12B). By doing this we have created a "dual-phase" unit, one that generates both extensor- and flexor-related activity. This type of unit was not featured in Grillner's 1981 model.
). The mixed synergy for the locomotor CPG is modeled by a diagonal excitatory connection (Fig. 12B), and this design is similar to the diagonal excitatory connection between the knee extensor and ankle flexor units proposed by Carter and Smith (1986; see their Fig. 7B) for the paw-shake CPG. Pearson and Rossignol (1991) demonstrated that neither feedback nor input from supraspinal centers was necessary for the mixed synergy of paw shaking to be elicited. They elicited the VL-TA coactivity in a spinalized, immobilized preparation by squirting the hindpaw with water. With regard to fictive locomotion, the IP muscle has not been studied to any extent, and we do not know if the IP-VL coactivity can be generated centrally.
).
; Pearson and Rossignol 1991), but spontaneous locomotor-like motor patterns in the decorticated, immobilized cats are characterized by ST bursts that alternate with flexor bursts (Perret and Cabelquen 1976, 1980). That ST fictive motor patterns are preparation specific suggests that spinal networks may be reconfigured, as proposed in Fig. 12, by supraspinal input in the absence of feedback.
).
), the FDL's activity patterns are similar to those of the FDB (Trank and Smith 1996). For this reason, we propose that the FDB and FDL may be controlled by the same unit in the locomotor CPG. Although the FDL has a flexor and an extensor burst during level walking, the FDL extensor-related activity is more of a facultative feature of its pattern during level walking (see O'Donovan et al. 1982; Trank and Smith 1996). Being facultative, it is likely that the FDL extensor-related burst is facilitated by external conditions triggered by sensory feedback, such as cutaneous stimulation (see Moschovakis et al. 1991). However, FDL activity also occurs during flexor- and extensor-related phases of fictive locomotion (Fleshman et al. 1984), suggesting that the FDL unit, similar to that proposed for the FDB unit in Fig. 12, is facilitated by central mechanisms that generate dual-phase patterns of activity.
). Taken together these results suggest that the CPG unit for the EDB may generate activity during the flexion or extension phase; this is consistent with our findings for slope walking and our CPG model in Fig. 12.
; McCrea et al. 1995; Whelan et al. 1995). A brief tap to the paw dorsum at the beginning of swing elicits a stumbling corrective response during forward but not backward walking (Buford and Smith 1993). For backward walking, a stumbling response is elicited when the plantar surface of the paw is tapped at the onset of swing. In both cases the tap mimics an obstruction to the paw's swinging motion, and the initial response withdraws the hindpaw from the unexpected object by ankle extension or flexion.
suggested that the nervous system may act like a "fuzzy controller" to combine multiple sensory inputs according to rules for the production of the appropriate sensorimotor responses. This analogy is particularly apt for discussing the control of hindlimb locomotion. The rules for effecting the transition from stance to swing, for example, might be different for different forms of walking. During level walking, hip flexion is initiated if the hip is extended and the ankle extensors are not contracting and the contralateral limb is loaded (bearing weight). The same if-and rule may apply to backward and upslope walking, but the weighting or combination of sensory inputs may be evaluated differently. The hip joint will be less extended when the ankle extensors stop contracting during backward walking (Buford and Smith 1990) and more extended when the ankle extensors stop contracting during upslope walking (Carlson-Kuhta et al. 1998). For downslope walking the ankle extensors may not be active after paw contact, but the hip continues to extend.
). By setting rules that reconfigure the CPG (see Fig. 12) and facilitate key interneurons for task-specific sensorimotor responses (see Fig. 8 of McCrea et al. 1995; Fig. 8 of Pearson and Collins 1993), inputs from higher centers would be able to control different forms of locomotion at the segmental level without providing an extensive set of motor instructions for each.
| |
ACKNOWLEDGEMENTS |
|---|
We thank S. Lauretz (Animal Health Technician) for assistance with surgery, animal care and training, and data collection; C. Chen for assisting with the gait analyses and working on many of the figures; S. Fornalski and P. Lim for assisting in the data collection and analysis of the treadmill downslope data; and K. Veling for assisting in the data collection and the preliminary analyses of the walkway slope data. We are also grateful to M. Orosz for developing the computer software used to analyze the kinematic variables illustrated in Fig. 3 and to R. Gregor at Georgia Institute of Technology for helpful suggestions on the DISCUSSION section related to limb dynamics.
This research was supported by National Institute of Neurological Disorders and Stroke Grant NS-19864 to J. L. Smith and a University of California, Los Angeles, Presidential Dissertation Year Fellowship to T. V. Trank.
| |
FOOTNOTES |
|---|
Address reprint requests to J. L. Smith.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  Y.-H. Chang, A. G. Auyang, J. P. Scholz, and T. R. Nichols Whole limb kinematics are preferentially conserved over individual joint kinematics after peripheral nerve injury J. Exp. Biol., November 1, 2009; 212(21): 3511 - 3521. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Maas, R. J. Gregor, E. F. Hodson-Tole, B. J. Farrell, and B. I. Prilutsky Distinct muscle fascicle length changes in feline medial gastrocnemius and soleus muscles during slope walking J Appl Physiol, April 1, 2009; 106(4): 1169 - 1180. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Donelan, D. A. McVea, and K. G. Pearson Force Regulation of Ankle Extensor Muscle Activity in Freely Walking Cats J Neurophysiol, January 1, 2009; 101(1): 360 - 371. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Lavrov, C. J. Dy, A. J. Fong, Y. Gerasimenko, G. Courtine, H. Zhong, R. R. Roy, and V. R. Edgerton Epidural Stimulation Induced Modulation of Spinal Locomotor Networks in Adult Spinal Rats J. Neurosci., June 4, 2008; 28(23): 6022 - 6029. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D.A. McVea and K.G. Pearson Contextual learning and obstacle memory in the walking cat Integr. Comp. Biol., October 1, 2007; 47(4): 457 - 464. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. J. Roberts, B. K. Higginson, F. E. Nelson, and A. M. Gabaldon Muscle strain is modulated more with running slope than speed in wild turkey knee and hip extensors J. Exp. Biol., July 15, 2007; 210(14): 2510 - 2517. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. McVea and K. G. Pearson Long-Lasting, Context-Dependent Modification of Stepping in the Cat After Repeated Stumbling-Corrective Responses J Neurophysiol, January 1, 2007; 97(1): 659 - 669. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. A. Rybak, N. A. Shevtsova, M. Lafreniere-Roula, and D. A. McCrea Modelling spinal circuitry involved in locomotor pattern generation: insights from deletions during fictive locomotion J. Physiol., December 1, 2006; 577(2): 617 - 639. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. R. Lammers, K. D. Earls, and A. R. Biknevicius Locomotor kinetics and kinematics on inclines and declines in the gray short-tailed opossum Monodelphis domestica J. Exp. Biol., October 15, 2006; 209(20): 4154 - 4166. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Gregor, D. W. Smith, and B. I. Prilutsky Mechanics of Slope Walking in the Cat: Quantification of Muscle Load, Length Change, and Ankle Extensor EMG Patterns J Neurophysiol, March 1, 2006; 95(3): 1397 - 1409. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. P. Ivanenko, R. E. Poppele, and F. Lacquaniti Spinal Cord Maps of Spatiotemporal Alpha-Motoneuron Activation in Humans Walking at Different Speeds J Neurophysiol, February 1, 2006; 95(2): 602 - 618. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Ekeberg and K. Pearson Computer Simulation of Stepping in the Hind Legs of the Cat: An Examination of Mechanisms Regulating the Stance-to-Swing Transition J Neurophysiol, December 1, 2005; 94(6): 4256 - 4268. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. I. Prilutsky, M. G. Sirota, R. J. Gregor, and I. N. Beloozerova Quantification of Motor Cortex Activity and Full-Body Biomechanics During Unconstrained Locomotion J Neurophysiol, October 1, 2005; 94(4): 2959 - 2969. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Gabaldon, F. E. Nelson, and T. J. Roberts Mechanical function of two ankle extensors in wild turkeys: shifts from energy production to energy absorption during incline versus decline running J. Exp. Biol., June 1, 2004; 207(13): 2277 - 2288. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. E. Higham and B. C. Jayne Locomotion of lizards on inclines and perches: hindlimb kinematics of an arboreal specialist and a terrestrial generalist J. Exp. Biol., January 15, 2004; 207(2): 233 - 248. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. E. Higham and B. C. Jayne In vivo muscle activity in the hindlimb of the arboreal lizard, Chamaeleo calyptratus: general patterns and the effects of incline J. Exp. Biol., January 15, 2004; 207(2): 249 - 261. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. B. Gillis and A. A. Biewener Effects of surface grade on proximal hindlimb muscle strain and activation during rat locomotion J Appl Physiol, November 1, 2002; 93(5): 1731 - 1743. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Jing and K. R. Weiss Interneuronal Basis of the Generation of Related but Distinct Motor Programs in Aplysia: Implications for Current Neuronal Models of Vertebrate Intralimb Coordination J. Neurosci., July 15, 2002; 22(14): 6228 - 6238. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Yakovenko, V. Mushahwar, V. VanderHorst, G. Holstege, and A. Prochazka Spatiotemporal Activation of Lumbosacral Motoneurons in the Locomotor Step Cycle J Neurophysiol, March 1, 2002; 87(3): 1542 - 1553. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. E. Burke, A. M. Degtyarenko, and E. S. Simon Patterns of Locomotor Drive to Motoneurons and Last-Order Interneurons: Clues to the Structure of the CPG J Neurophysiol, July 1, 2001; 86(1): 447 - 462. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Earhart and P. S. G. Stein Step, Swim, and Scratch Motor Patterns in the Turtle J Neurophysiol, November 1, 2000; 84(5): 2181 - 2190. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Matsuyama and T. Drew Vestibulospinal and Reticulospinal Neuronal Activity During Locomotion in the Intact Cat. II. Walking on an Inclined Plane J Neurophysiol, November 1, 2000; 84(5): 2257 - 2276. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. A. Abelew, M. D. Miller, T. C. Cope, and T. R. Nichols Local Loss of Proprioception Results in Disruption of Interjoint Coordination During Locomotion in the Cat J Neurophysiol, November 1, 2000; 84(5): 2709 - 2714. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Earhart and A. J. Bastian Form Switching During Human Locomotion: Traversing Wedges in a Single Step J Neurophysiol, August 1, 2000; 84(2): 605 - 615. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. H. Ting, S. A. Kautz, D. A. Brown, and F. E. Zajac Contralateral Movement and Extensor Force Generation Alter Flexion Phase Muscle Coordination in Pedaling J Neurophysiol, June 1, 2000; 83(6): 3351 - 3365. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J Quevedo, B Fedirchuk, S Gosgnach, and D A McCrea Group I disynaptic excitation of cat hindlimb flexor and bifunctional motoneurones during fictive locomotion J. Physiol., June 1, 2000; 525(2): 549 - 564. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Lamb and J. F. Yang Could Different Directions of Infant Stepping Be Controlled by the Same Locomotor Central Pattern Generator? J Neurophysiol, May 1, 2000; 83(5): 2814 - 2824. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Grasso, M. Zago, and F. Lacquaniti Interactions Between Posture and Locomotion: Motor Patterns in Humans Walking With Bent Posture Versus Erect Posture J Neurophysiol, January 1, 2000; 83(1): 288 - 300. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Carlson-Kuhta, T. V. Trank, and J. L. Smith Forms of Forward Quadrupedal Locomotion. II. A Comparison of Posture, Hindlimb Kinematics, and Motor Patterns for Upslope and Level Walking J Neurophysiol, April 1, 1998; 79(4): 1687 - 1701. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
