Effect of Ankle-Foot Orthosis Alignment and Foot-PlateLength on the Gait of Adults With Hemiplegia

810

PROSTHETICS, ORTHOTICS, DEVICES

Effect of Ankle-Foot Orthosis Alignment and Foot-Plate Length on the Gait of Adults With Poststroke Hemiplegia Stefania Fatone, PhD, Steven A. Gard, PhD, Bryan S. Malas, MHPE, CO ABSTRACT. Fatone S, Gard SA, Malas BS. Effect of anklefoot orthosis alignment and foot-plate length on the gait of adults with poststroke hemiplegia. Arch Phys Med Rehabil 2009;90: 810-8.

Objective: To investigate the effect of ankle-foot orthosis (AFO) alignment and foot-plate length on sagittal plane knee kinematics and kinetics during gait in adults with poststroke hemiplegia.

Design: Repeated measures, quasi-experimental study.

Setting: Motion analysis laboratory.

Participants: Volunteer sample of adults with poststroke hemiplegia (n⫽16) and able-bodied adults (n⫽12) of similar age.

Interventions: Subjects with hemiplegia were measured walking with standardized footwear in 4 conditions: (1) no AFO (shoes only); (2) articulated AFO with 90° plantar flexion stop and full-length foot-plate– conventionally aligned AFO (CAFO); (3) the same AFO realigned with the tibia vertical in the shoe– heel-height compensated AFO (HHCAFO); and (4) the same AFO (tibia vertical) with ¾ length foot-plate–¾ AFO. Gait of able-bodied control subjects was measured on a single occasion to provide a normal reference.

Main Outcome Measures: Sagittal plane ankle and knee kinematics and kinetics.

Results: In adults with hemiplegia, walking speed was unaffected by the different conditions (P⫽.095). Compared with the no AFO condition, all AFOs decreased plantar flexion at initial contact and mid-swing (P⬍.001) and changed the peak knee moment in early stance from flexor to extensor (P⬍.000).

Both AFOs with full-length foot-plates significantly increased the peak stance phase plantar flexor moment compared with no AFO and resulted in a peak knee extensor moment in early stance that was significantly greater than control subjects, whereas the AFO with three-quarter length foot-plate resulted

From the Prosthetics Research Laboratory and Rehabilitation Engineering Research Program, Departments of Physical Medicine and Rehabilitation (Fatone, Gard) and Biomedical Engineering (Gard), Northwestern University; Jesse Brown Veterans Affairs Medical Center (Gard) and Moira Tobin Wickes Orthotics Program, Children’s Memorial Hospital (Malas), Chicago, IL. Presented to the International Society for Prosthetics and Orthotics, July 29 – August 30, 2007, Vancouver, BC, Canada; the American Academy of Orthotists and Prosthetists, March 21–24, 2007, San Francisco, CA; the American Academy of Orthotists and Prosthetists, March 1– 4, 2006, Chicago, IL; the American Academy of Orthotists and Prosthetists, March 16 –19, 2005, Orlando, FL; International Society for Prosthetics and Orthotics, August 1– 6, 2004, Wanchai, Hong Kong, China; the Gait and Clinical Movement Analysis Society, April 21–24, 2004, Lexington, KY; and the American Congress of Rehabilitation Medicine, October 23–26, 2003, Tucson, AZ. Supported by the Office of Research and Development (Rehabilitation R&D Service), Department of Veterans Affairs (merit review #A2676I) and administered by the Jesse Brown VA Medical Center.

No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.

Reprint requests to Stefania Fatone, PhD, NUPRL & RERP, 345 E Superior St, Room 1441, Chicago, IL 60611, e-mail: s-fatone@northwestern.edu.

0003-9993/09/9005-00534$36.00/0 doi:10.1016/j.apmr.2008.11.012

Arch Phys Med Rehabil Vol 90, May 2009

in ankle dorsiflexion during stance and swing that was significantly less than control subjects.

Conclusions: These findings suggest that when an articulated AFO is to be used, a full-length foot-plate in conjunction with a plantar flexion stop may be considered to improve early stance knee moments for people with poststroke hemiplegia.

Key Words: Gait; Hemiplegia; Orthotic devices; Rehabilitation; Stroke. © 2009 by the American Congress of Rehabilitation Medicine TROKE IS ONE OF THE leading causes of serious longS term disability in the United States with a reported 4.5 million stroke survivors of whom 15% to 30% are permanently

disabled.1,2 Stroke often results in dysfunction of 1 side of the body (hemiplegia). The gait of persons with hemiplegia is less metabolically efficient and leads to increased falls compared with able-bodied persons.3-8 Problems with poor balance, instability in stance, hypertonicity, inappropriate and involuntary posturing of the foot and ankle, and recurvatum and instability at the knee have led to the recommendation that orthoses be incorporated in the lower limb management of patients after stroke.9 Bowker et al10 described orthoses as acting directly if they surround the segment or joint they are attempting to influence or indirectly if they attempt to modify the external forces acting on a joint beyond their physical boundaries. For example, it has been shown in able-bodied adults11-13 and children with cerebral palsy14 that the position of the ground reaction force vector relative to the knee joint axis may be manipulated by altering inclination of the tibia relative to the vertical by using an appropriate combination of AFO and footwear. Controlling AFO alignment may not only assist function of the ankle-foot complex, but may also influence and improve knee function as well. In the absence of strong evidence, it has been suggested by experts that a nonarticulated AFO may be used to control mild recurvatum or instability of the knee, an articulated AFO with plantar flexion stop may be used to control knee recurvatum, and an AFO with dorsiflexion stop may be used to control knee flexion instability.9 It may also be possible to influence the knee by manipulating the AFO foot-plate length. Foot-plate length may influence moments at the knee by altering the moment arm of the ground reaction force. Commonly used foot-plate lengths include the full-length foot-plate that extends distal to the toes and the three-quarter length foot-plate that ends proximal to the metaList of Abbreviations AFO CAFO COP HHCAFO ¾ AFO

ankle-foot orthosis conventionally aligned ankle-foot orthosis center of pressure heel-height compensated ankle-foot orthosis ¾ length heel-height compensated ankle-foot orthosis

ANKLE-FOOT ORTHOSIS ALIGNMENT AND FOOT-PLATE LENGTH, Fatone

tarsal heads. Clinically, the length of the foot-plate may be determined by, among other things, the degree of toe-clawing15 or the volume that can be accommodated in the toe-box of the shoe. The potential effect of AFO foot-plate length on gait has yet to be explored.

AFOs are presently the most widely used orthoses in the United States, accounting for 26% of clinical practice by certified orthotists, double that of any other type of orthosis.16 However, in spite of their widespread prescription and usage, there has been limited research undertaken, especially with respect to their use in the management of stroke.9 An international, multidisciplinary consensus conference on the orthotic management of stroke concluded that, with the exception of improving walking speed and cadence, which was supported by a moderate level of evidence, indications for an articulated AFO could only be recommended by experts as a good practice point based on clinical experience due to a lack of scientific evidence.9 Articulated AFOs with plantar flexion stops are frequently provided for patients with hemiplegia.17-20 Therefore, the purpose of this study was to investigate the effect of AFO alignment and foot-plate length on sagittal plane knee kinematics and kinetics during gait in adults with poststroke hemiplegia.

It was hypothesized that alignment of the plantar flexion stop would influence moments occurring at the knee during loading response such that a more vertical alignment of the AFO calf section would lessen the initial knee flexor moment compared with a more anteriorly inclined alignment of the AFO calf section. It was also hypothesized that a relatively rigid fulllength foot-plate would allow for progression of the COP further distally along the foot compared with a three-quarter length foot-plate, influencing moments at the knee. Data from the subjects with hemiplegia are compared with those of ablebodied control subjects of similar age walking at similar speeds. This provides a frame of reference for identifying the effects of orthotic intervention.21 METHODS The Northwestern University Institutional Review Board approved this study and informed consent was obtained from all subjects prior to their participation. Persons with hemiplegia after stroke who volunteered to participate in this study were primarily outpatients of rehabilitation programs in the Chicago area. Selection of subjects with hemiplegia was based on the following inclusion criteria: presence of hemiplegia after stroke; a minimum of 24 months poststroke; age 40 to 70 years; no major involvement of the contralateral limb; currently wearing or had previously worn an articulated AFO. All participants with hemiplegia were initially assessed, measured, and had an impression taken by the same orthotist for a custom-molded, polypropylene, articulated AFO with 90° plantar flexion stop and free dorsiflexion. Articulation was achieved with Tamarack Flexure Jointsa and a mechanical plantar flexion stop of reinforced polypropylene (fig 1). The impression for the AFO was taken with the ankle in a neutral position; that is, with the tibia and foot aligned at a 90° angle.

Polypropylene 3/16in (4.8mm) thick was used in the fabrication of all AFOs and a dorsal ankle strap was attached to the AFO when it appeared necessary to maintain proper foot and ankle alignment in the orthosis. Footwear was standardized for the subjects with hemiplegia with each participant receiving a pair of extra depth leather shoes.b The footwear used had a heel to forefoot sole thickness difference of approximately 1cm.

We used shoe heel-height to manipulate AFO alignment because it was felt that this would result in realistic alignment changes. The same orthosis and shoes were used to test 3

811

Fig 1. Example of the custom-molded, polypropylene, articulated AFO with 90° plantar flexion stop, free dorsiflexion, and full-length foot-plate used in this study. Inset indicates attachment of the reflective marker to the AFO ankle joint for gait analysis.

orthotic conditions in the following order for all subjects: In condition 1, which was termed the CAFO, the plantar flexion stop was set at 90° (tibia-foot angle) and the foot-plate was flat and full-length, extending distal to the toes. When placed in the shoe, the calf section was inclined 5° to 7° anteriorly due to the difference in heel to forefoot sole thickness. In condition 2, which was termed the HHCAFO, the plantar flexion stop was adjusted by removing material from the reinforced area posterior to the calcaneus (the plantar flexion stop) so that the calf section was vertical in the shoe, a change in alignment of 5° to 7°. Foot-plate length was unaltered. In condition 3, which was termed the ¾ AFO, the foot-plate was trimmed to three-quarter length, ending proximal to the metatarsal heads. Ankle alignment was unaltered.

Prior to data collection, we allowed the participants 2 weeks of accommodation to each orthotic condition. Data for each condition were collected on 3 separate occasions each lasting approximately 2 hours. Subjects were instructed that they could rest as necessary at any time during the testing session. All gait data were acquired in a motion analysis laboratory equipped with an 8-camera real-time motion capture systemc and 6 forceplatesd embedded flush in the floor of a 10m walkway.

Gait analyses were performed with subjects walking under 4 conditions: a baseline, shoe only (no AFO) condition and the 3 AFO conditions described previously (CAFO, HHCAFO, and Arch Phys Med Rehabil Vol 90, May 2009

812

ANKLE-FOOT ORTHOSIS ALIGNMENT AND FOOT-PLATE LENGTH, Fatone

¾ AFO). For each condition, data were collected with the subjects walking at their normal self-selected walking speed.

During the gait analyses, reflective markers were taped to the skin over palpable anatomic landmarks using the modified Helen Hayes marker set,22 which allowed us to acquire 3-dimensional data from the pelvis and both lower limbs. Markers were located on the dorsum of the foot between the second and third metatarsals immediately proximal to the metatarsal heads, the posterior calcaneus, lateral malleoli, lateral femoral condyles, right and left anterior superior iliac spines, and the sacrum at the superior aspect of the L5/sacral interface. Wand markers were placed on the lateral aspects of the thigh and calf.

Because the AFO obscured the landmarks required for identification of the anatomical ankle joint axis, the base plate of ankle markers were attached using the proximal screw of the mechanical ankle joints, the axis of which was aligned with the anatomical axis by the orthotist during fabrication of the orthosis (see fig 1, inset). The marker was suspended from the proximal screw so that the center of the marker was located in line with the gap between the shank and foot segments of the AFO. Heel (calcaneal) markers were placed 1cm higher than toe/dorsal markers to account for angulation of the foot within the shoe caused by the heel of the shoe. The same investigator placed all markers on subjects for all gait analyses. Force data were acquired simultaneous with motion data.

We recorded normal reference gait data from able-bodied control subjects of similar age using the same instrumentation and marker placement. Control subjects were recruited primarily by word of mouth and with the assistance of an agingresearch registry. Able-bodied subjects were assessed on a single occasion while walking in comfortable gym shoes at 5 self-selected speeds: normal, slow, very slow, fast, and very fast. Collecting data at 5 self-selected speeds is standard protocol for data included in our able-bodied database.

EVa RealTime softwarec was used to determine the 3-dimensional position of each marker relative to the laboratory coordinate system during each frame of a given trial. A static standing trial using additional markers on the medial malleoli and medial femoral condyles was recorded in order to calculate joint centers. The raw coordinate data were filtered using a second order bi-directional low pass filter with an effective cutoff frequency of 6Hz.23 Orthotrak softwarec was used to calculate temporospatial parameters, kinematics, and kinetics.

All moments reported in this study refer to internal moments.

COP excursion was calculated as the distance between the first COP point and the last COP point for each gait cycle.

For all subjects, we analyzed a minimum of 3 walking trials for each condition and speed. The exact number of trials recorded per subject was contingent on acquiring at least 3 clean forceplate strikes per foot. Clean forceplate strikes were ones in which only 1 foot contacted a forceplate, without being over the edges of the plate. Data were normalized based on the entire gait cycle, averaged, and then included for group analysis.

Statistical analysis indicated that some variables lacked homogeneity of variance and were not normally distributed.

Hence, nonparametric tests were used for all analyses involving data from subjects with hemiplegia. Friedman tests were used to determine if there were differences between repeated measures at alpha equal to .05. If the Friedman test was significant, the Wilcoxon signed-rank test with Bonferroni correction for multiple comparisons was used with alpha equal to .008. Mann-Whitney and Kruskal-Wallis tests with Bonferroni correction were used to determine if there were differences between independent groups (control subjects versus subjects Arch Phys Med Rehabil Vol 90, May 2009

with hemiplegia and comparisons between subgroups of subjects with hemiplegia) at alpha equal to .0125.

RESULTS Twenty-two people with hemiplegia were enrolled in this study with 16 completing the protocol. Six subjects withdrew because they found paid employment (#17), became ill (#22), lost the study shoes (#14), or changed their mind about participating (#3, #4, #10). The participants ranged in age from 43 to 66 years (mean, 53⫾7y) with a mean time since stroke of 7⫾4 years. Data from 12 healthy, able-bodied control subjects (mean age, 57⫾8y) were collected for comparison. Descriptive information regarding the research participants is summarized in table 1.

Because kinematic and kinetic data are affected by walking speed, walking speeds were analyzed first.21 Regardless of condition, the participants with hemiplegia walked with a significantly slower median self-selected normal walking speed than the control subjects (P⬍.000) (fig 2). There was no significant difference in median walking speed between control subjects walking at their very slow self-selected walking speed and subjects with hemiplegia walking at their normal self-selected walking speed (Pcontrol:no AFO⫽.307, Pcontrol:CAFO⫽.164, Pcontrol:HHCAFO⫽ .353, Pcontrol:3/4 AFO⫽.710). Also, when comparing orthotic conditions, there were no significant differences in self-selected normal walking speed (P⫽.095). All subsequent kinematic and kinetic data were compared using normal self-selected walking speed for the participants with hemiplegia and a very slow self-selected walking speed for the control subjects.

Differences between conditions for ankle angle at initial contact, maximum dorsiflexion in stance, and ankle angle at midswing were significant (P⬍.000) (fig 3A). Compared with walking without an AFO, all AFO conditions significantly decreased the plantar flexion angle to neutral at initial contact (P⬍.000) and mid-swing (P⬍.001) (figs 4A and 4C). The AFOs with full-length foot-plate (CAFO and HHCAFO) significantly increased dorsiflexion in late stance compared with no AFO, whereas the ¾ AFO did not (fig 4B). However, there were no significant differences between the 3 AFO conditions for ankle angle at initial contact (PCAFO:HHCAFO⫽.148, PCAFO:3/4 AFO⫽.015, PHHCAFO:3/4 AFO⫽ .163), maximum ankle dorsiflexion in stance (PCAFO:HHCAFO⫽ .569, PCAFO:3/4 AFO⫽.012, PHHCAFO:3/4 AFO⫽.017), and ankle angle at mid-swing (PCAFO:HHCAFO⫽.278, PCAFO:3/4 AFO⫽.012, PHHCAFO:3/4 AFO⫽.134).

Compared with control subjects, ankle angle at initial contact was significantly more plantar flexed in the no AFO condition, significantly more dorsiflexed in the CAFO condition (see fig 4A), but no different with the heel-height compensated alignment regardless of foot-plate length (Pcontrol:HHCAFO⫽.286 and Pcontrol:3/4 AFO⫽.642). Maximum stance phase dorsiflexion and mid-swing ankle angles for the AFOs with full-length foot-plates were no different to that of control subjects (Pcontrol:CAFO⫽.307 and Pcontrol:HHCAFO⫽.330).

Maximum stance phase dorsiflexion was less than that of control subjects for the no AFO and ¾ AFO conditions (see fig 4B), whereas mid-swing ankle angle was significantly less than control subjects for the no AFO condition (see fig 4C).

The peak stance phase ankle plantar flexor moment was significantly different for all conditions (P⫽.005) (fig 3B).

The AFOs with full-length foot-plate significantly increased the plantar flexor moment in stance compared with no AFO (fig 4D), whereas the ¾ AFO did not (Pno AFO:3/4 AFO⫽.134).

The peak plantar flexor moment for all AFO conditions were no different from that of control subjects (Pcontrol:CAFO⫽ .095, Pcontrol:HHCAFO⫽.086, Pcontrol:3/4 AFO⫽.016), whereas

813

ANKLE-FOOT ORTHOSIS ALIGNMENT AND FOOT-PLATE LENGTH, Fatone Table 1: Information Regarding Participants With Hemiplegia (nⴝ16)* Subject Number

Sex Men/Women

Age (y)

Years Since Stroke

1 2 5 6 7 8 9 11 12 13 15 16 18 19 20 21

M M M M W M M W M W M M M W W W

46 55 58 58 66 64 53 44 48 45 56 65 43 48 50 52

8 6 2 16 7 2 10 6 14 10 6 2 3 6 16 7

Hemiplegia Mean ⫾ SD Control Mean ⫾ SD

6 Women 10 Men 4 Women 8 Men

53.2⫾7.5 57.1⫾8.5

7.6⫾4.6 NA

Involved Side Right/Left

R R L L L R R L L L L R R L L L 6 Right 10 Left NA

Height (cm)

Mass (kg)

171 169.5 172 184 157 181.5 165 166 185.5 165 188.5 175 183 165 171.5 144

111 77 87 93.5 58 96.5 90.5 77.5 83 81 157 105 105 72.5 72 58.5

171.5⫾11.6 174.5⫾9.9

89.1⫾23.9 80.8⫾13.0

NOTE. Mean data for able-bodied control subjects are shown at the bottom of the table (n⫽12). Subjects 3, 4, 10, 14, 17, and 22 are not included because they withdrew from the study.

Abbreviation: NA, not applicable.

in the no AFO condition it was significantly less than controls (see fig 4D).

Compared with control subjects, the subjects with hemiplegia had significantly less peak knee flexion in swing (P⬍.000) (fig 3C and 4F), but there was no difference in minimum knee angle in stance (Pcontrol:no AFO⫽.837, Pcontrol:CAFO⫽.599, Pcontrol:HHCAFO⫽ .537, Pcontrol:3/4 AFO⫽.347) (fig 4E). Peak knee extensor moments in the first half of stance phase were significantly greater than control subjects only for the AFOs with full-length foot plates (Pcontrol:no AFO⫽.450, Pcontrol:3/4 AFO⫽.029) (figs 3D and 4G). There was no difference between control subjects and subjects with hemiplegia for peak knee moments in the second half of stance (Pcontrol:no AFO⫽.837, Pcontrol:CAFO⫽.537, Pcontrol:HHCAFO⫽ .537, Pcontrol:3/4 AFO⫽.371) (fig 4H).

There were no significant differences between orthotic conditions for maximum knee angle in swing (P⫽.576) or minimum knee angle in stance (P⫽.141). There was a significant difference in peak internal knee moment during the first half of stance (P⬍.000) but not the second half of stance (P⫽.392). All AFO conditions significantly altered the peak internal knee

Fig 2. Self-selected normal and very slow median walking speeds shown for the control subjects (nⴝ12). Normal self-selected walking speed shown for the subjects with hemiplegia for each condition tested (nⴝ16). Variance indicated by first and third quartiles.

moment during the first half of stance phase compared with no AFO (see fig 4G) but there were no differences between AFO conditions (PCAFO:HHCAFO⬍.438, PCAFO:3/4 AFO⬍.039, PHHCAFO:3/4 AFO⫽.326). Note, however, that there was considerable variability in the peak knee moment during early stance for the No AFO condition. This was due in part to individual knee alignment: 7 of the 16 subjects in this study showed knee hyperextension when ambulating without an orthosis.

The effect of the AFO on knee moments was most pronounced for subjects who showed knee hyperextension during stance when ambulating without an AFO (figs 5 and 6). There was a significant difference in minimum knee angle in stance between the subjects who hyperextended and those who did not for all conditions (fig 6A). With regards to peak knee moments during the first half of stance phase, only the no AFO condition was significantly different between subjects who hyperextended and those who did not (Pno AFO⫽.001, PCAFO⫽.210, PHHCAFO⫽.210, P3/4 AFO⫽.114) (fig 6C).

For the subjects who hyperextended at the knee, all AFO conditions delayed the onset of hyperextension until single limb support (P⫽.078), decreased the magnitude of hyperextension (P⫽.037), and altered the internal knee moment in the first half of stance from flexor to extensor (P⫽.001) (see fig 5A and 5B).

However, post hoc analyses of the peak knee moments were unable to detect differences (Pno AFO:CAFO⫽.018, Pno AFO:HHCAFO⫽.018, Pno AFO:3/4 AFO⫽.018, PCAFO:HHCAFO⫽.398, PCAFO:3/4 AFO⫽ .018, PHHCAFO:3/4 AFO⫽.091).

For the subjects who did not hyperextend (see figs 5C and 5D), there was no significant difference in minimum knee angle in stance (P⫽.532) (see fig 6A) or peak knee moment during the first half of stance between conditions (P⫽.057) (see fig 6C).

Compared with control subjects, excursion of the COP was significantly less only for the no AFO condition (P⫽.000).

Excursion of the COP was significantly different for all conditions in the persons with hemiplegia (P⬍.000) (fig 7). COP excursion was significantly increased in all 3 AFO conditions compared with no AFO. Although there was no significant Arch Phys Med Rehabil Vol 90, May 2009

814

ANKLE-FOOT ORTHOSIS ALIGNMENT AND FOOT-PLATE LENGTH, Fatone

Fig 3. Mean ankle angle (A), ankle moment (B), knee angle (C), and knee moment (D) for the involved limb of the subjects with hemiplegia at normal self-selected walking speed (nⴝ16). Asterisks (*) indicate points in the gait cycle where the difference in peak angle or moment between conditions was significantly different. Vertical lines indicate mean toe-off for each condition. Internal moments are shown.

difference in COP excursion as a result of changing alignment (PCAFO:HHCAFO⫽.015) or foot-plate length (PHHCAFO:3/4 AFO⫽ .959), COP excursion was significantly longer in the CAFO compared with the ¾ AFO (PCAFO:3/4 AFO⫽.004) where both alignment and foot-plate length were different.

DISCUSSION This study investigated the effects of AFO alignment and foot-plate length on gait in adults with poststroke hemiplegia.

Gait analysis was used to assess the effect of subjects walking without an AFO (no AFO) and 3 articulated AFO conditions with different sagittal plane ankle alignments and foot-plate lengths. Compared with no AFO, all of the AFO conditions tested in this study were able to decrease plantar flexion of the ankle at initial contact and mid-swing, increase the early stance peak knee extensor moment and increase COP excursion.

These results were due to action of the plantar flexion stop, which mechanically restricted plantar flexion motion of the ankle and encouraged initial contact to occur with the heel rather than foot flat or fore-foot. This finding is consistent with previous findings that AFOs with resistance to plantar flexion, whether articulated or nonarticulated, improve the ankle angle at initial contact and mid-swing.24-26 In our study, these results were unaffected by the alignments and foot-plate lengths tested.

Because the plantar flexion stop restrains posterior tibial rotation with respect to the foot it was not surprising that the AFO had the most pronounced effect in subjects who showed knee hyperextension when ambulating without an orthosis.

Where hyperextension was present, there was a large flexor moment at the knee in the No AFO condition that was eliminated Arch Phys Med Rehabil Vol 90, May 2009

when an AFO was worn–all AFOs significantly altered the knee moment during the first half of stance from extensor to flexor.

Although not statistically significant, all AFO conditions also delayed the onset of hyperextension and decreased the magnitude of hyperextension. These findings are consistent with previous studies that have showed that nonarticulated or plantar flexion stop AFOs can control knee recurvatum and reduce the external knee extension moment during stance, especially if set in dorsiflexion or anterior tibial inclination.24,27 Although hyperextension was decreased by the AFOs, it was not eliminated entirely. Hyperextension was still present during single limb support and the knee remained significantly more extended for all conditions than the subjects who did not hyperextend (although the knee moment was no different in early stance between these 2 groups when all the AFOs were worn). Relative motion between the limb and AFO may account for some of the measured hyperextension because the knee marker was located on the anatomic knee joint axis and the ankle marker was located on the axis of rotation of the orthotic ankle joint. If compression of the calf musculature against the posterior shell of the orthosis occurred when the plantar flexion stop was maximally engaged, additional hyperextension may have been measured. It is also possible, and more likely, that although posterior motion of the tibia during stance may be halted by the AFO, forward momentum continues and is transferred to the thigh, contributing to knee hyperextension during double support.

In this study, use of an AFO did not alter walking speed compared with walking without one and there was no significant difference in normal self-selected walking speed between AFO conditions. Although some previous studies have sug-

ANKLE-FOOT ORTHOSIS ALIGNMENT AND FOOT-PLATE LENGTH, Fatone

815

Fig 4. Median ankle and knee data for the involved limb: (A) ankle angle at initial contact (IC); (B) peak ankle dorsiflexion (DF) in stance; (C) ankle angle at mid-swing; (D) peak stance ankle plantar flexor (PF) moment; (E) minimum (min) knee angle in stance; (F) maximum (max) knee angle in swing; (G) peak knee moment in first half of stance; and (H) peak knee moment in second half of stance. Variance indicated by first and third quartiles.

gested that AFOs increase speed compared with walking without an AFO,18,27-31 a systematic review by Leung and Moseley32 concluded only that AFOs may improve walking speed in people with stroke. It has been suggested that improvement in walking speed with an AFO is related to inclination of the tibia (ie, the shank-to-floor angle), with greater than 5° of anterior tibial inclination required to increase walking speed.33 The rationale for this conclusion is that controlling tibial progression in mid to late stance transfers forward momentum to the thigh, facilitating knee and hip extension in terminal stance and placing the limb in a more appropriate alignment for transfer of body weight to the contralateral limb and commencing swing

phase with the ipsilateral limb. The orthoses used in this study did not control tibial progression in mid to late stance.

With regard to our hypotheses, knee moments in early stance were no different between the 2 alignment conditions (CAFO and HHCAFO) and foot-plate length did not increase COP excursion (HHCAFO and ž AFO). The results suggest that the alignment changes used in this study and based on footwear heel-height may not have been of sufficient magnitude to have an impact at the knee.

Our results indicated that the AFOs with full-length footplate significantly increased stance phase dorsiflexion and plantar flexor moment compared with walking without an AFO so Arch Phys Med Rehabil Vol 90, May 2009

816

ANKLE-FOOT ORTHOSIS ALIGNMENT AND FOOT-PLATE LENGTH, Fatone

Fig 5. Mean knee angles and moments for the involved limb of (A, B) subjects who hyperextended at the knee (nⴝ7) and (C, D) subjects who did not hyperextend at the knee (nⴝ9). Vertical lines indicate mean toe-off for each condition. Internal moments are shown.

that they were no different to that of control subjects. The AFO with three-quarter length foot-plate resulted in significantly less than normal dorsiflexion in late stance. Perhaps the AFOs with full-length foot plate increased dorsiflexion and the plantar flexor moment in late stance by delaying or impeding heel rise more than the three-quarter length foot-plate, which may have allowed greater motion at the metatarsal break. The AFOs with full-length foot-plate also significantly increased the peak knee extensor moment in early stance compared with control subjects, whereas the peak knee moment in early stance in the ¾ AFO was no different to control subjects. It is unclear why this was the case. Further research is required to explore these results.

These findings have several clinical implications that warrant consideration when orthotic intervention is contemplated.

The results of this study suggest that subtle changes to the AFO sagittal alignment have relatively less effect on the knee moments during early stance than does the length of the AFO foot-plate. Based on these findings the orthotist should consider incorporating a full-length foot-plate in conjunction with a plantar flexion stop in order to limit the internal knee flexor moment in the early part of stance for people with hemiplegia.

This enhancement may have additional clinical application for persons with cerebral palsy and traumatic brain injury who present with a similar gait pattern. Additionally, all 3 AFO conditions decreased the internal knee flexor moment in early stance phase. Reducing this moment may decrease the risk for joint degeneration and ligamentous laxity that can occur as a result of prolonged knee flexor moments during stance phase.

Although this result is encouraging, the AFO conditions evaluated in this study did not entirely eliminate hyperextension at the knee during stance; rather they delayed the onset and reduced the magnitude of hyperextension. Perhaps more imArch Phys Med Rehabil Vol 90, May 2009

portantly, they altered the early stance phase knee moment so that it was no different from that of persons with hemiplegia without hyperextension ambulating with the same orthosis.

Although marker placement and tissue compression may have contributed to the measurement of knee angle, it may also mean that additional changes to the orthotic design are necessary to further reduce knee hyperextension during stance. Further reduction in hyperextension during double support may provide additional protection from damage to the knee joint complex.

Study Limitations There are several limitations to this study. One was that orthotic conditions were not randomized, potentially allowing for a series effect. Although fatigue during the course of a testing session could have influenced the results between trials for each orthotic condition, data for the different orthotic conditions were collected on separate occasions, attenuating the effect of fatigue on comparison of results between orthotic conditions. There is potential for bias in that the study population consisted of volunteer subjects recruited primarily from among outpatient rehabilitation facilities in the Chicago area and may have represented those with the greatest degree of mobility and motivation. Another limitation was that the study shows the effect of this particular AFO design without establishing that it was the most appropriate prescription for each individual participant. This AFO design was chosen for evaluation because it is commonly prescribed for persons with stroke (and had been previously prescribed to all participants in this study) and because prescriptive criteria for the orthotic management of stroke patients are based predominantly on clinical judgment. Further research is required to explore the effect of

ANKLE-FOOT ORTHOSIS ALIGNMENT AND FOOT-PLATE LENGTH, Fatone

817

Fig 6. Median knee data for the involved limb comparing subjects who hyperextended (HE) to those who did not (No HE): (A) minimum (min) knee angle in stance; (B) maximum (max) knee angle in swing; (C) peak knee moment in first half of stance; and (D) peak knee moment in second half of stance. Variance indicated by first and third quartiles.

different AFO designs on the sagittal plane ankle and knee kinematics and kinetics of persons with poststroke hemiplegia.

CONCLUSIONS This study showed that articulated AFOs with plantar flexion stops improve sagittal plane stance and swing phase ankle joint orientation and early stance phase knee moments in adults with hemiplegia after stroke, but did not increase walking speed. When hyperextension was present, the articulated AFOs with plantar flexion stops decreased the magnitude and delayed the onset of hyperextension. Changes made to alignment of the AFO were relatively small, but of a realistic magnitude because they were based on footwear heel-height. These findings suggest that when

Fig 7. Median excursion of the COP expressed as a fraction of foot length. Variance indicated by first and third quartiles.

an articulated AFO is to be used, a full-length foot-plate in conjunction with a plantar flexion stop may be considered to improve early stance knee moments for people with poststroke hemiplegia. Further research is required to explore the effect of different AFO designs on the sagittal plane ankle and knee kinematics and kinetics of persons with poststroke hemiplegia.

Acknowledgments: We thank Rebecca Stine, MS, for assistance in the collection of this data; Andrew Hansen, PhD, for assisting with calculation of COP data; and Richard Harvey, MD, the Clinical Neuroscience Research Registry at the Rehabilitation Institute of Chicago, and the Buehler Center for Aging, Health and Society at Northwestern University, for assistance with recruitment of subjects.

References 1. Centers for Disease Control. Prevalence of disabilities and associated health conditions among adults: United States, 1999.

MMWR Recomm Rep 2001;50:120-5.

2. American Heart Association. 2001 heart and stroke statistical update. Dallas: American Heart Association, 2000.

3. Detrembleur C, Dierick F, Stoquart G, Chantraine F, Lejeune T. Energy cost, mechanical work, and efficiency of hemiparetic walking. Gait Posture 2003;18:47-55.

4. Platts M, Rafferty D, Paul L. Metabolic cost of over ground gait in younger stroke patients and healthy controls. Med Sci Sports Exerc 2006;38:1041-6.

5. Lamontagne A, Stephenson J, Fung J. Physiological evaluation of gait disturbances post stroke. Clin Neurophysiol 2007;118:717-29.

6. Forster A, Young J. Incidence and consequences of falls due to stroke: a systematic inquiry. Br Med J 1995;311:83-6.

7. Jorgensen L, Engstad T, Jacobsen B. Higher incidence of falls in long-term stroke survivors than in population controls: depressive symptoms predict falls after stroke. Stroke 2002;33:542-7.

Arch Phys Med Rehabil Vol 90, May 2009

818

ANKLE-FOOT ORTHOSIS ALIGNMENT AND FOOT-PLATE LENGTH, Fatone

8. Wada N, Sohmiya M, Shimizu T, Okamoto K, Shirakura K. Clinical analysis of risk factors for falls in homeliving stroke patients using functional evaluation tools. Arch Phys Med Rehabil 2007;88:1601-5.

9. Condie E, Campbell J, Martina J, editors. Report of a consensus conference on the orthotic management of stroke patients. Copenhagen: International Society for Prosthetics and Orthotics; 2004.

Available at: http://www.ispo.ws/. Accessed October 16, 2008.

10. Bowker P, Condie DN, Bader DL, Pratt DJ, Wallace WA. Biomechanical basis of orthotic management. Oxford: Butterworth Heinemann; 1993.

11. Cook T, Cozzens B. The effects of heel height and ankle-footorthosis configuration on weight line location: a demonstration of principles. Orthot Prosthet 1976;30:43-6.

12. Lehmann J, Ko M, deLateur B. Knee moments: origin in normal ambulation and their modification by double-stopped ankle-foot orthoses. Arch Phys Med Rehabil 1982;63:345-51.

13. Lehmann JF, Condon SM, de Lateur BJ, Smith JC. Gait abnormalities in tibial nerve paralysis: a biomechanical study. Arch Phys Med Rehabil 1985;66:80-5.

14. Meadows B, Bowers R, Owen E. Biomechanics of the hip, knee, and ankle. In: Hsu J, Michael J, Fisk J, editors. AAOS Atlas of Orthoses and Assistive Devices. 4th ed. Philadelphia: Mosby Elsevier; 2008. p 299-309.

15. Miyazaki S, Yamamoto S, Kubota T. Effect of ankle-foot orthosis on active ankle moment in patients with hemiparesis. Med Biol Eng Comput 1997;34:381-5.

16. Whiteside S, Allen M, Barringer W, et al. Practice analysis of certified practitioners in the disciplines of orthotics and prosthetics. Alexandria: American Board for Certification in Orthotics and Prosthetics Inc; 2007. Available at: http://www.abcop.org/ Practice_Analysis.asp. Accessed October 16, 2008.

17. Tyson S, Thornton H, Downes A. The effect of a hinged anklefoot orthosis on hemiplegic gait: four single case studies. Physiother Theory Pract 1998;14:75-85.

18. Tyson SF, Thornton HA. The effect of a hinged ankle foot orthosis on hemiplegic gait: objective measures and users’ opinions. Clin Rehabil 2001;15:53-8.

19. Eberly V, Weiss W, Mulroy S, Gronley J, Perry J. Effect of AFO design on walking in individuals with hemiplegia and plantar flexion contractures from CVA. Gait Posture 2003;18(Suppl 2): 1-79.

20. Lee KH, Johnston R. Bracing below the knee for hemiplegia: biomechanical analysis. Arch Phys Med Rehabil 1973;54:466-70.

21. Hanlon M, Anderson R. Prediction methods to account for the effect of gait speed on lower limb angular kinematics. Gait Posture 2006;24:280-7.

Arch Phys Med Rehabil Vol 90, May 2009

22. Kadaba M, Ramakrishnan H, Wootten M, Gainey J, Gorton G, Cochran GV. Repeatability of kinematic, kinetic, and electromyographic data in normal adult gait. J Orthop Res 1989;7:849-860.

23. Winter D. Biomechanics and motor control of human movement.

2nd ed. New York: John Wiley and Sons Inc; 1990.

24. Gok H, Kucukdeveci A, Altinkaynak H, Yavuzer G, Ergin S. Effects of ankle-foot orthoses on hemiparetic gait. Clin Rehabil 2003;17:137-9.

25. Bleyenheuft C, Caty G, Lejeune T, Detrembleur C. Assessment of the chignon dynamic ankle-foot orthosis using instrumented gait analysis in hemiparetic adults. Ann Readapt Med Phys 2008;51: 154-60.

26. Yokoyama O, Sashika H, Hagiwara A, Yamamoto S, Yasui T. Kinematic effects on gait of a newly designed ankle-foot orthosis with oil damper resistance: a case series of 2 patients with hemiplegia. Arch Phys Med Rehabil 2005;86:162-6.

27. Lehmann J, Condon S, Price R, deLateur B. Gait abnormalities in hemiplegia: their correction by ankle-foot orthoses. Arch Phys Med Rehabil 1987;68:763-71.

28. Corcoran P, Jebsen R, Brengelmann G, Simons B. Effects of plastic and metal leg braces on speed and energy cost of hemiparetic ambulation. Arch Phys Med Rehabil 1970;51:69-77.

29. Weiss W, Mulroy S, Gronley J, Perry J, Boyd L. Rigid AFO impairs walking ability in individuals with hemiparesis from CVA. Gait Posture 2002;16(Suppl 1):S2.

30. Beckerman H, Becher J, Lankhorst GJ, Verbeek AL. Walking ability of stroke patients: efficacy of tibial nerve blocking and a polypropylene ankle-foot orthosis. Arch Phys Med Rehabil 1996; 77:1144-51.

31. Franceschini M, Massucci M, Ferrari L, Agosti M, Orsi M, Paroli C. Gait with custom-made orthosis in hemiplegic patients: preliminary data. Gait Posture 2001;13:139-40.

32. Leung J, Moseley A. Impact of ankle-foot orthoses on gait and leg muscle activity in adults with hemiplegia. Physiotherapy 2002;89: 39-55.

33. Bowers RJ. Non-articulated ankle foot orthoses. In: Condie E, Campbell J, Martina J, editors. Report of a consensus conference on the orthotic management of stroke patients. Copenhagen: International Society for Prosthetics and Orthotics; 2004. p 87-94.

Available at: http://www.ispo.ws/. Accessed October 16, 2008.

Suppliers a. Becker Orthopedic, 635 Executive Dr, Troy, MI 48083. b. PW Minor & Son, 3 Treadeasy Ave, Batavia, NY 14020. c. Motion Analysis Corporation, 3617 Westwind Blvd, Santa Rosa, CA 95403. d. Advanced Mechanical Technology Inc, 176 Waltham St, Watertown, MA 02472.