Effects of solid ankle-foot orthoses with individualized ankle angles on gait for children with cerebral palsy and equinus

Abstract

PURPOSE:

For children with cerebral palsy (CP) and equinus, the conventional practice of setting the ankle angle in an ankle-foot orthosis (AA-AFO) at 90 ${}^{\circ}$ may not adequately accommodate gastrocnemius length/stiffness. Therefore, this study compared the effects of statically-optimized solid AFOs with individualized AA-AFOs (iAA-AFOs) and conventionally-prescribed AFOs on gait for children with CP and equinus.

METHODS:

Ten children with CP and equinus (15 limbs with AFOs), and 15 typically-developing (TD) children participated. For the children with CP, solid AFOs with iAA-AFOs (range $=$ 5 ${}^{\circ}$ –25 ${}^{\circ}$ plantarflexion) were compared with their usual AFOs using three-dimensional gait analysis. TD children walked in shoes only. Peak values and Gait Variable Scores (GVS) for joint and segment variables were calculated for stance phase. Responses were categorized using 90% confidence intervals relative to TD data, for each affected leg.

RESULTS:

Net responses to iAA-AFOs were positive for 60% of limbs and negative for 40%. Knee variables (GVS and peak extension, flexion, and midstance moment) were most positively affected, and foot-floor angle and vertical ground reaction force were most negatively impacted.

CONCLUSION:

Individualized AFO prescription and iAA-AFOs can impact gait biomechanics for some children with equinus, compared to conventionally-prescribed AFOs. Optimizing dynamic alignment for walking may further improve outcomes.

Keywords

Cerebral palsy individualized ankle-foot orthosis prescription AFO ankle angle equinus AFO tuning gait

1. Introduction

Cerebral palsy (CP) describes a group of permanent motor disorders resulting from an injury to the developing fetal or infant brain [1]. Amongst children with CP, one of the most common gait impairments is equinus [2] – a condition associated with triceps surae spasticity and static or dynamic contracture, planovalgus or cavovarus foot deformities, and gait impairments. These include excessive stance phase plantarflexion, instability, knee hyperextension, decreased shock absorption, and poor swing leg clearance [3, 4]. Muscle shortening may alter plantar flexor length-tension curves and impair the muscle’s mechanical ability to produce force [3, 5]. Foot deformity and altered biomechanics may cause musculoskeletal stress, strain, and pain (e.g., in feet and knees), affecting long-term ambulation and functional outcomes [3, 6, 7].

Ankle-foot orthoses (AFOs) are often prescribed to improve joint motion and gait patterns for children with equinus. Optimal AFO prescription is a complex and challenging process that aims to match the child’s characteristics (e.g., gait pattern, clinical findings) to the mechanical/design characteristics of the orthosis, such as AFO type (e.g., solid AFOs intended to prevent ankle motion, or articulated and flexible designs which allow motion), and the angle of the ankle in the AFO (AA-AFO) [8, 9, 10]. However, there is a lack of consensus regarding the best AFO type, AA-AFO, and method of clinical decision-making for children with equinus. Some authors recommend either solid or articulated AFOs with partially-defined indications [11]. Others (including a review article) provide more specific guidance – for example, AFOs should fully accommodate gastrocnemius length [9, 12, 13], and that articulated AFOs are contraindicated for children without appropriate gastrocnemius length and stiffness (i.e., resistance to passive stretch due to hypertonicity) to allow $>$ 10 ${}^{\circ}$ ankle dorsiflexion with the knee extended and the arch maintained in stance phase [10]. For children with abnormal stance phase shank segment kinematics, solid AFOs have been recommended to optimize shank motion and improve proximal mechanics by redirecting the ground reaction force (GRF) vector in relation to the lower extremity joints [13, 14, 15]. To optimize these effects, “tuning” the static alignment of AFO-footwear combinations (AFO-FCs) for the desired activity (e.g., standing or walking) has been recommended as standard practice [8, 10, 15, 16, 17, 18, 19]. This process includes optimizing joint and segment alignments including the shank-to-vertical angle (SVA), equalizing leg lengths, and footwear heel and sole modifications [10].

The AA-AFO is another aspect of the AFO prescription that can be individualized. Traditionally, AA-AFOs are set at 90 ${}^{\circ}$ (i.e., with the ankle in 0 ${}^{\circ}$ dorsiflexion), theoretically preventing or limiting ankle plantar flexor contracture. However, this convention may stem from an erroneous assumption that a 90 ${}^{\circ}$ ankle with vertical shank facilitates stance phase knee extension [10, 19]. Its benefits in terms of contracture prevention have not been substantiated [18]. Because the gastrocnemius muscle spans the subtalar, talocrural, and knee joints, a solid AFO with an individualized, plantarflexed AA-AFO may have the potential to accommodate the decreased gastrocnemius length and increased stiffness associated with equinus deformities better than conventional approaches [15, 20, 21].

If the AFO type and AA-AFO do not fully accommodate the length and stiffness/hypertonicity of the gastrocnemius musculotendinous unit, several compensations may occur. Knee extension may be limited at initial contact (IC) or during stance [10, 15, 22], and subtalar pronation may compensate for restricted ankle dorsiflexion. When the subtalar joint pronates, alignment of the talonavicular and calcaneocuboid joint axes allows more dorsiflexion at the mid-tarsal joint compared to when the subtalar joint is neutral or supinated [23]. Without adequate ankle passive range of motion (PROM), the forces applied during dorsiflexion motion are more likely to stretch the small, extensible midfoot ligaments than the Achilles tendon [22]. This makes it difficult to selectively target stretching forces to the ankle joint and suggests that an excessively dorsiflexed AA-AFO (relative to the available PROM) may promote hyperpronation and/or midfoot break, lever arm dysfunction, and future pain [22]. This rationale warrants discussion about inconsistent AFO type and AA-AFO section practices in the literature which inconsistently consider gastrocnemius length [9, 10, 15, 19].

Thus, for children with CP, orthotic intervention may be more effective when AA-AFOs are individualized based on gastrocnemius length and stiffness [8, 10, 19]; however, evidence-based guidelines are lacking [8, 23]. Even when plantarflexed AA-AFOs have been reported (e.g., references [8, 26, 27]), the rationale for determining the specific angles has not been described. Although a clinical algorithm has been proposed to determine the AA-AFO in a solid AFO [26], it has not been experimentally evaluated to date. Despite potential for solid AFO designs to address many of the biomechanical challenges associated with equinus, articulated and flexible designs remain prevalent in conventional prescription practices, habitually preferred by clinicians who assume they promote function and a more natural gait pattern [27, 28]. Therefore, this study compared the effects of two AFO conditions on gait for children with CP and equinus: 1) solid AFOs with optimized static alignment and individualized AA-AFOs (iAA-AFOs) that accommodated gastrocnemius length and stiffness of each involved limb, and 2) the child’s conventionally-prescribed AFOs (“Usual AFOs”). It was hypothesized that, compared to the Usual AFOs, the iAA-AFOs would promote foot and shank segment kinematics (i.e., foot-floor angle and SVA), and knee and hip joint kinematics and kinetics in the sagittal plane that were closer to reference values for age-matched typically-developing children. We did not expect that spatiotemporal values (e.g., velocity) would be impacted.

2. Methods

2.1 Participants

Participants were 10 children with CP (6–18 years; median $=$ 10 y 10 m), and 15 Typically-developing (TD) children (6–18 years; median $=$ 12 y 7 m). Children with CP, recruited from the province’s two rehabilitation facilities, met the following criteria: 1) aged 5–18 years; 2) diagnosis of spastic CP; 3) gastrocnemius contracture $\geqslant$ 5 ${}^{\circ}$ ; 4) wore AFO(s) $>$ 6 months; 5) able to walk 8 meters $\geqslant$ 20 times; 6) no orthopaedic surgery or botulinum toxin injection in $>$ 6 months. TD children met the following criteria: 1) aged 5–18 years; 2) born full-term; 3) no injury or condition affecting walking/balance. Children were excluded if their parent reported they would be unable to co-operate with the testing procedure.

The study was approved by the Research Ethics Boards of the Former Regina Qu’Appelle Health Region and University of Saskatchewan. All children provided informed assent and parents/guardians provided informed consent to participate.

2.2 Procedures

2.2.1 Clinical assessment and determination of individualized AA-AFO

A registered physiotherapist with 15 years’ pediatric rehabilitation experience (KJK) assessed each child’s lower limb alignment, gait, ROM, and tone [21, 29, 30]. AA-AFOs were determined according to a clinical algorithm [10, 21, 26]. Accordingly, as a proxy for gastrocnemius musculo-tendinous length, ankle dorsiflexion PROM was measured with the knee extended and the foot in fully pronated, neutral, and supinated positions. The most restricted measurement determined the iAA-AFO; the angle was further plantarflexed if necessary to optimize triplanar bony alignment of the foot or if excessive force was required to overcome gastrocnemius stiffness to maintain this optimal bony alignment [10, 21, 26]. Ankle dorsiflexion ROM in knee flexion was also assessed to compare the influence of the soleus and gastrocnemius (Appendix B). Dorsiflexion PROM was measured using an iPad (Apple Computers, Inc., Cupertino, CA, USA) and the goniometry application (DrGoniometer, CDM S.r.L, Este, Italy) shown to have a degree of reliability comparable to traditional goniometry [31]. The Gross Motor Function Classification System (GMFCS) [32] was used to classify the motor abilities of each participant with CP.

2.2.2 Orthoses and footwear

For all children in the CP group, solid AFOs with iAA-AFOs were fabricated from 3/16” polypropylene according to standard practice of the orthotics department, with ankle trimlines anterior to malleoli. Apart from the AA-AFO (which was individualized according to the clinical examination), the design of these solid AFOs was standardized as much as possible to reduce the number of confounding variables without excessively impairing the child’s gait pattern. Each orthosis had tibial and ankle straps, and trimlines distal to metatarsophalangeal joints. The orthotist added an ethylene-vinyl acetate (EVA) heel wedge equal to the child’s iAA-AFO to create a vertical bench alignment (i.e., vertical alignment of the calf segment of the AFO with the AFO standing upright; Fig. 1). Also, the orthotist assessed the bench alignment according to standard practice, estimating a straight line along the calf segment of the AFO between the projected location of child’s lateral malleolus and fibular head. The physiotherapist (KJK) measured the ankle angle of each iAA-AFO to confirm that it matched the ROM measurement used to determine the iAA-AFO within 5 ${}^{\circ}$ .

Figure 1.

Photo of AFO with vertical bench alignment and the ankle angle in the AFO (AA-AFO) in 25 ${}^{\circ}$ of plantarflexion.

To reduce footwear-related gait variation, all participants except one wore Rebound mid-top skate shoes (DC Shoes, Inc.) in both orthotic conditions. These shoes had a 0 mm heel-sole differential (HSD) and were deep enough to accommodate multiple heel wedge heights. Participant 10 wore New Balance 636 shoes with 5 mm HSD (New Balance, Boston, MA, USA) due to discomfort in the study shoes. Shoes were laced securely to minimize AFO motion.

2.2.3 Static alignment of AFO-footwear combinations (AFO-FCs)

Each AFO-FC was statically aligned by the first author prior to gait testing. Participants stood on a force plate (50.8 cm $\times$ 46.3 cm, OR6-7, AMTI, Watertown, MA, USA) while the GRF vector was visualized using direct video overlay (Nexus 2.5, VICON, Centennial, CO, USA; A620FC Digital Camera, Basler AG, DE). EVA heel wedges were secured to each shoe to adjust the SVA (angle of anterior tibia relative to vertical; Fig. 2) until the sagittal plane GRF vector was aligned through the middle of the knee joint and foot [17, 33]. High-density plastazote shoe raises equalized leg lengths as indicated by leg length measurements and wedge heights. For the iAA-AFOs (because AFO trimlines were located distal to the MTP joints) point loading rockers made of 1/4” high-density plastazote were secured underneath both soles at 80% of shoe length in order to simulate third (forefoot) rocker [8, 21]. Final SVAs were measured with children standing, using the DrGoniometer application (Fig. 2), and GRF locations were re-confirmed. A 10–15-minute period of acclimatization was provided in each AFO-FC.

Figure 2.

Screen shot of shank to vertical angle (SVA) measurement using DrGoniometer app. Photo shows the AFO from Fig. 1 in a statically aligned AFO-footwear combination (AFO-FC). Labels indicate location of wedge (used to adjust SVA) and point loading rocker (PLR). SVA is measured as the angle between the dashed line along the anterior tibia on the app and vertical (i.e., 9.4 ${}^{\circ}$ inclined, as calculated from the measurement on the app).

Table 1

Variable names and descriptions for stance phase kinematic and kinetic variables

	Variable description	Variable name
Peak joint angles	Ankle dorsiflexion	AnkleDF
	Ankle plantarflexion	AnklePF
	Knee extension at IC	KneeIC
	Knee flexion	KneeFlex
	Knee extension	KneeExt
	Hip flexion	HipFlex
	Hip extension	HipExt
Peak SVA	SVA at IC	SVA_IC
	SVA at TMSt	SVA_TMSt
Peak FFA	FFA at IC	FFA_IC
	FFA at TMSt	FFA_TMSt
Peak joint moments	Knee moment at LR ${}^{\text{b}}$	KneeMom_LR
	Knee moment at TMSt	KneeMom_TMSt
	Knee moment at TS ${}^{\text{c}}$	KneeMom_TS
	Hip extension moment	HipExt_Mom
	Hip flexion moment	HipFlex_Mom
Spatio-temporal	Stride velocity	Stride_Velocity
variables	Stride length	Stride_Length
	Stride time	Stride_Time
	Stance percent	Stance_Percent
GVS	Ankle ROM	GVS_Ankle
	Knee ROM	GVS_Knee
	Hip ROM	GVS_Hip
	SVA	GVS_SVA
	FFA	GVS_FFA
	Knee moment	GVS_KneeMom
	Hip moment	GVS_HipMom
	Vertical force	GVS_VF

Note: Children with bilateral CP wore orthoses on both legs, and children with unilateral CP wore an AFO on the affected limb. F: Female; M; male; y: years; m: months; GMFCS: Gross Motor Function Classification System [29]; DF: Dorsiflexion range of motion (measured with knee extended); n/a $=$ not applicable (i.e., child did not wear an AFO on this limb and ROM was not tested).

Table 2

Characteristics of participants with CP

Child ID	Age	Sex	Diagnosis (affected limb if unilateral)	GMFCS	DF in supination		DF in neutral		DF in pronation
					Left	Right	Left	Right	Left	Right
1	17y 11m	F	Bilateral	II	$-$ 15 ${}^{\circ}$	$-$ 20 ${}^{\circ}$	$-$ 11 ${}^{\circ}$	$-$ 18 ${}^{\circ}$	$-$ 15 ${}^{\circ}$	$-$ 20 ${}^{\circ}$
2	6y 8m	M	Unilateral (right)	I	n/a	$-$ 15 ${}^{\circ}$	n/a	$-$ 17 ${}^{\circ}$	n/a	$-$ 15 ${}^{\circ}$
3	16y 7m	F	Bilateral	I	$-$ 6 ${}^{\circ}$	$-$ 18 ${}^{\circ}$	7 ${}^{\circ}$	1 ${}^{\circ}$	$-$ 5 ${}^{\circ}$	$-$ 20 ${}^{\circ}$
4	9y 10m	F	Bilateral	II	$-$ 4 ${}^{\circ}$	$-$ 12 ${}^{\circ}$	$-$ 7 ${}^{\circ}$	$-$ 4 ${}^{\circ}$	$-$ 5 ${}^{\circ}$	$-$ 10 ${}^{\circ}$
5	7y 7m	F	Unilateral (left)	I	$-$ 4 ${}^{\circ}$	n/a	$-$ 4 ${}^{\circ}$	n/a	$-$ 5 ${}^{\circ}$	n/a
6	14y 11m	M	Unilateral (right)	I	n/a	$-$ 8 ${}^{\circ}$	n/a	$-$ 4 ${}^{\circ}$	n/a	$-$ 12 ${}^{\circ}$
7	11y 10m	M	Unilateral (right)	I	n/a	$-$ 25 ${}^{\circ}$	n/a	4 ${}^{\circ}$	n/a	$-$ 25 ${}^{\circ}$
8	6y 5m	M	Unilateral (right)	I	n/a	$-$ 11 ${}^{\circ}$	n/a	0 ${}^{\circ}$	n/a	$-$ 15 ${}^{\circ}$
9	6y	F	Bilateral	II	$-$ 4 ${}^{\circ}$	$-$ 14 ${}^{\circ}$	10 ${}^{\circ}$	2 ${}^{\circ}$	$-$ 5 ${}^{\circ}$	$-$ 15 ${}^{\circ}$
10	17y	F	Bilateral	III	$-$ 18 ${}^{\circ}$	$-$ 10 ${}^{\circ}$	10 ${}^{\circ}$	2 ${}^{\circ}$	$-$ 20 ${}^{\circ}$	$-$ 12 ${}^{\circ}$

Note: Children with bilateral CP wore orthoses on both legs, and children with unilateral CP wore an AFO on the affected limb. F: Female; M; male; y: years; m: months; GMFCS: Gross Motor Function Classification System [29]; DF: Dorsiflexion range of motion; n/a $=$ not applicable (i.e., child did not wear an AFO on this limb and ROM was not tested).

2.2.4 Gait analysis

Three-dimensional lower limb kinematics were collected using an 8-camera motion capture system (Vicon Nexus, Centennial, CO, fs $=$ 100 Hz). Participants walked at a self-selected speed along a 10-meter walkway with two embedded force plates (OR6-7, AMTI, Watertown, MA, USA, fs $=$ 2000 Hz). Children with CP walked in Usual AFO and iAA-AFO conditions in randomized order. TD children walked with shoes only. For each condition, trials were conducted until six force plate contacts were obtained with each foot. Rest periods were provided as needed. Ankle axes were defined using an initial static pose without AFOs and subsequently tracked virtually using the fixed tibia marker clusters.

Shoe and foot, defined as two different coordinate systems, were both tracked with the same shoe-based marker cluster. The shoe coordinate system was generated from a static pose with the dorsal axis aligned vertically and the medio-lateral axis perpendicular to the plane defined by the dorsal axis and the heel-to-toe vector. The foot coordinate system was then established by rotating the shoe coordinate system about its medio-lateral axis based on the AA-AFO and any attached heel wedges. For the TD participants, shoe and foot coordinate systems were coincident.

Ankle joint kinematics were defined as the relative angle between shank and foot coordinate systems. SVA was calculated between the long axis of the shank and vertical axis in the sagittal plane. Foot-floor angle was defined between the heel-toe axis of the shoe coordinate system and horizontal plane.

Kinematic and force plate data were combined using standard inverse dynamics techniques [34] to calculate 3D joint moments for the ankle, knee, and hip. Segment parameters for inverse dynamics calculations were estimated using published data [35].

2.3 Analyses

Motion analysis data were processed and visually inspected using Vicon Nexus 2.0 (Centennial, CO, USA) and custom routines in MATLAB R2017a (MathWorks, Natick, MA, USA). Stance phase data from three to six trials were ensemble averaged for each affected limb and condition and for each TD participant. Kinetic data were not collected for participant 9 because short step lengths precluded force plate contact with only one foot. Data from TD participants were combined into three age-bands (6–8, 10–13, and 15–18 years).

Peak ankle, knee, and hip kinematics and joint moments and individual gait variable scores (GVS) [36] were calculated for the entire stance phase (Table 1). GVS, a gait quality index, was calculated as the mean root-mean-square difference between each trial of the child’s stance phase data and the mean of the appropriate TD age-band group stance phase data. The GVS was selected due to its potential to provide more information about changes in the variables of interest [37], an advantage over global gait indices which have not demonstrated adequate sensitivity or specificity to detect changes with orthotic intervention [38].

Because of wide within-sample variability, group means were not emphasized [39]. For the children with CP, data for each affected limb during stance phase were analyzed visually using descriptive statistics. For each limb, the mean, SD, and confidence interval (90%CI) were calculated for Usual AFO and iAA-AFO conditions. A difference between conditions was defined as no overlap of 90%CIs. Differences were considered to indicate positive responses for the iAA-AFO if iAA-AFO condition values were closer to the mean of TD data than the Usual AFO condition. iAA-AFO values farther from the TD mean than the Usual AFO condition were considered negative responses. Overlapping 90%CIs indicated no difference between conditions (i.e., equivocal). If values for both conditions were within one SD of the TD mean, the results were also considered equivocal (within normal limits). The number of responses in each category (positive, negative, and equivocal) were summed for each limb for all 25 kinematic, kinetic, and spatiotemporal variables combined. Limbs were considered net positive responders if more variables were affected positively than negatively. Conversely, net negative responders demonstrated more negative than positive responses. Cohen’s d was calculated using a pooled SD as an effect size measure quantifying the magnitude of each positive and negative response. Effect sizes of 0.8 or greater were considered large effects [40].

Wilcoxon signed-rank test compared ankle dorsiflexion (Ankle_DF), plantarflexion (Ankle_PF) and Ankle_GVS between conditions to confirm the effectiveness of AA-AFO experimental manipulation. Ankle plantarflexion was reported as negative while dorsiflexion was defined as positive.

Statistical analyses were completed using SPSS Statistics version 24.0 (IBM Corp., Armonk, NY, USA), and MATLAB R2017a (MathWorks, Natick, MA, USA). Statistical significance was set at an $\alpha$ level of 0.05.

Table 3
Characteristics of the Usual AFOs and iAA-AFOs, and description of adjustments made during the static alignment process

	Usual AFO ${}^{\text{a}}$						iAA-AFO orthosis ${}^{\text{b}}$
Child (limb)	AFO type	AA-AFO ${}^{\text{c}}$ (height of heel wedge)	Other features	Additions		SVA	iAA-AFO ${}^{\text{c}}$ (height of heel wedge)	Other features	Additions		SVA
				Heel ${}^{\text{d}}$	Through raise				Heel ${}^{\text{d}}$	Through raise
1 (R)	Articulated ${}^{\text{e}}$	0 ${}^{\circ}$	–	0	0	20 ${}^{\circ}$	$-$ 15 ${}^{\circ}$ (4 cm)	Ribs	0	0	12 ${}^{\circ}$
1 (L)	Articulated ${}^{\text{e}}$		–	0	0	17 ${}^{\circ}$	$-$ 20 ${}^{\circ}$ (4.5 cm)	Ribs	0	0	11 ${}^{\circ}$
2 (R)	Solid ${}^{\text{f}}$	$-$ 10 ${}^{\circ}$ (2.4 cm)	Trim lines distal to MTPs	5 ${}^{\circ}$ (1 cm)	1 cm	9 ${}^{\circ}$	$-$ 15 ${}^{\circ}$ (3 cm)	–	10 ${}^{\circ}$	5 mm	10 ${}^{\circ}$
3 (R)	Solid ${}^{\text{f}}$	$-$ 5 ${}^{\circ}$ (1.5 cm)	–	2.5 ${}^{\circ}$ (5 mm)	0	10 ${}^{\circ}$	$-$ 20 ${}^{\circ}$ (4.5 cm)	–	0	7 mm	12 ${}^{\circ}$
3 (L)	SMO ${}^{\text{f}}$	0 ${}^{\circ}$	–	0	0	10 ${}^{\circ}$	$-$ 5 ${}^{\circ}$ (1.5 cm)	–	0	1.4 cm	13 ${}^{\circ}$
4 (R)	Semi-flexible ${}^{\text{g}}$	$-$ 3 ${}^{\circ}$ (0.7 cm)	0.5 cm medial forefoot post; SMO insert, forefoot strap	0	0	17.5 ${}^{\circ}$	$-$ 5 ${}^{\circ}$ (0.8 cm)	12 ${}^{\circ}$ forefoot equinus	5 ${}^{\circ}$	1 cm	18 ${}^{\circ}$
4 (L)	Semi-flexible ${}^{\text{g}}$	$-$ 10 ${}^{\circ}$ (1 cm)	SBA 5 ${}^{\circ}$ recline; 0.5 cm medial forefoot post; SMO insert, forefoot strap	0	0	10 ${}^{\circ}$	$-$ 10 ${}^{\circ}$ (1.2 cm)	15 ${}^{\circ}$ forefoot equinus	5 ${}^{\circ}$	1 cm	14 ${}^{\circ}$
5 (L)	Solid ${}^{\text{f}}$	0 ${}^{\circ}$	Trim lines distal to MTPs; SBA 10 ${}^{\circ}$	7 ${}^{\circ}$ (1.3 cm)	1.4 cm on right leg	11 ${}^{\circ}$	$-$ 15 ${}^{\circ}$ (1.7 cm)	–	5 ${}^{\circ}$	2 cm on right leg	9 ${}^{\circ}$
6 (R)	Semi-flexible ${}^{\text{g}}$	0 ${}^{\circ}$ (1 cm)	SMO insert with 1 cm lateral forefoot post	0	5 mm	8 ${}^{\circ}$	$-$ 12 ${}^{\circ}$ (3 cm)	–	0	3 cm on left leg	10 ${}^{\circ}$
7 (R)	Carbon fiber ${}^{\text{h}}$	0 ${}^{\circ}$	–	0	1.5 cm on left leg	10 ${}^{\circ}$	$-$ 25 ${}^{\circ}$ (6.5 cm)	–	5 ${}^{\circ}$	5 cm on left leg	12 ${}^{\circ}$
8 (R)	Articulated ${}^{\text{e}}$	0 ${}^{\circ}$	Toe strap	0	5 mm on left leg	10 ${}^{\circ}$	$-$ 15 ${}^{\circ}$ (3 cm)	–	0	3 cm on left leg	10 ${}^{\circ}$
9 (R)	Articulated ${}^{\text{e}}$	0 ${}^{\circ}$	–	0	1 cm	5 ${}^{\circ}$	$-$ 15 ${}^{\circ}$ (2.5 cm)	–	0	0	10 ${}^{\circ}$
9 (L)	Articulated ${}^{\text{e}}$	0 ${}^{\circ}$	–	0	0	10 ${}^{\circ}$	$-$ 5 ${}^{\circ}$ (1 cm)	–	0	1.5 cm	10 ${}^{\circ}$
10 (R)	Articulated ${}^{\text{e}}$	0 ${}^{\circ}$	Casted in calcaneovarus with	0	0	5 ${}^{\circ}$	$-$ 12 ${}^{\circ}$ (2.5 cm)	Ribs	5 ${}^{\circ}$ (1.4 cm)	0	17 ${}^{\circ}$
10 (L)	Articulated ${}^{\text{e}}$		15 ${}^{\circ}$ lateral forefoot &	0	0	5 ${}^{\circ}$	$-$ 20 ${}^{\circ}$ (3.5 cm)	Ribs	5 ${}^{\circ}$ (1.4 cm)	0	14 ${}^{\circ}$
			hindfoot post

AA-AFO: Ankle angle in the AFO; SVA: Shank to vertical angle (optimized and measured in standing); PF: Plantarflexion; DF: Dorsiflexion; SMO: Supramalleolar orthosis; MTP: Metatarsophalangeal joints; PLR: Point loading rocker. ${}^{\text{a}}$ All Usual AFOs had a 90 ${}^{\circ}$ SBA, trim lines proximal to MTPs, and PLRs were added at 80% where trimlines were distal to MTPs, unless otherwise specified. ${}^{\text{b}}$ All iAA-AFOs were solid AFOs, with trimlines distal to MTPs, PLR at 80%, toes 90 ${}^{\circ}$ to shank, and vertical bench alignment unless otherwise specified. ${}^{\text{c}}$ Negative numbers indicate a plantarflexed AA-AFO; positive values indicate a dorsiflexed AA-AFO. ${}^{\text{d}}$ All footwear had a heel-sole differential (HSD) of 0 mm, except for Participant 10, whose HSD was 5 mm. ${}^{\text{e}}$ Articulated AFOs were custom-fabricated from 3/16” polypropylene with DF assist and PF stop at 0 ${}^{\circ}$ , AA-AFO rests in 25 ${}^{\circ}$ due to DF assist joints, and vertical bench alignment unless otherwise specified. ${}^{\text{f}}$ Custom-fabricated from 3/16” polypropylene. ${}^{\text{g}}$ FlexiSport, Cascade DAFO, Ferndale, WA. ${}^{\text{h}}$ ToeOFF ${}^{\@setsize{\tiny}{6pt}{\vpt}{\@vpt}\textregistered}$ , Allard USA, Rockaway, NJ.

3. Results

3.1 Physical assessment

Physical examination findings for each participant with CP are presented in Table 2 and the Appendix. Median dorsiflexion PROM was $-$ 12 ${}^{\circ}$ in supination (range $=-$ 4 ${}^{\circ}$ to $-$ 25 ${}^{\circ}$ ), and 0 ${}^{\circ}$ in both neutral (range $=$ 10 ${}^{\circ}$ to $-$ 18 ${}^{\circ}$ ) and pronation (range $=$ 8 ${}^{\circ}$ to $-$ 20 ${}^{\circ}$ ).

Table 4
Individual variable response profiles by limb according to net response: A. Limbs with net positive response; B. Limbs with net negative response. Effect size (Cohen’s d) values inside red and green boxes indicate magnitude of positive and negative responses

Positive response

Negative response

Equivocal response

Not tested

A.
Child (limb)		1 (R)	2 (R)	4 (R)	6 (R)	7 (R)	8 (R)	9 (L)	10 (R)	10 (L)
Knee	KneeIC			3.3		1.7
	KneeFlex			2.8				2.7		3.0
	KneeExt	4.0				2.1		1.7	2.2	17.7
	GVS_Knee		1.5			2.1		2.6	2.2	8.6
	KneeMom_LR				1.7	3.7		X		4.1
	KneeMom_TMSt				2.1	$-$ 2.7		X	3.5	2.2
	KneeMom_TS							X	2.3
	GVS_KneeMom		2.8					X	2.2	2.0
Hip	HipFlex			3.0		2.1		2.4
	HipExt				1.6	$-$ 1.6	2.1		$-$ 1.9	$-$ 3.9
	GVS_Hip								$-$ 2.8	$-$ 3.6
	HipExtMom					1.9		X	2.1
	HipFlexMom							X
	GVS_HipMom				$-$ 2.0			X	2.4
Segments	SVA_IC			2.7		2.7
	SVA_TMSt			2.3		4.6
	FFA_IC					4.7				2.8
	FFA_TMSt					4.5				$-$ 3.8
	GVS_SVA					4.7				$-$ 2.4
	GVS_FFA		1.8							$-$ 4.5
GRF	GVS_VF			$-$ 2.5				X	1.9
Spatio-temporal	StrideVelocity
	StrideLength					1.7
	StrideTime
	StancePercent
% Positive		4%	12%	20%	12%	48%	4%	24%	32%	28%
% Negative		0%	0%	4%	4%	8%	0%	0%	8%	20%
% Equivocal		96%	88%	76%	84%	44%	96%	76%	60%	52%
Net response

B.
Child (limb)		1 (L)	3 (R)	3 (L)	4 (L)	5 (L)	9 (R)
Knee	KneeIC	3.9
	KneeFlex	3.1	$-$ 3.1		$-$ 2.9
	KneeExt	$-$ 6.8			1.8		$-$ 1.4
	GVS_Knee	$-$ 3.7			$-$ 1.6
	KneeMom_LR	$-$ 4.7	$-$ 2.5				X
	KneeMom_TMSt		3.4				X
	KneeMom_TS						X
	GVS_KneeMom	$-$ 2.6			$-$ 2.3		X
Hip	HipFlex
	HipExt
	GVS_Hip
	HipExtMom					$-$ 2.4	X
	HipFlexMom	$-$ 1.8					X
	GVS_HipMom						X
Segments	SVA_IC
	SVA_TMSt		$-$ 2.9
	FFA_IC		$-$ 2.1		6.4
	FFA_TMSt		$-$ 2.7		2.2
	GVS_SVA		$-$ 2.6
	GVS_FFA	$-$ 2.2	$-$ 3.2	$-$ 2.7	$-$ 1.6
GRF	GVS_VF	$-$ 2.0			$-$ 2.2		X
Spatio-temporal	StrideVelocity	$-$ 2.3
	StrideLength	$-$ 2.4
	StrideTime
	StancePercent
% Positive		8%	4%	0%	12%	0%	0%
% Negative		36%	28%	4%	20%	4%	6%
% Equivocal		56%	68%	96%	68%	96%	94%
Net response

Figure 3.

Representative knee flexion/extension kinematics for two limbs for both AFO conditions during stance phase (from initial contact to terminal contact): A) positive response (participant 10, right limb); and B) negative response (participant 1, left leg). Shaded area denotes data from age-matched typically developing children $\pm$ 1SD, and grey dotted line denotes 2SD.

3.2 Participant AFO-FC characteristics/static alignment

Participants wore orthoses on 15/20 limbs. Table 3 describes the Usual orthoses. All except three Usual AFOs had 0 ${}^{\circ}$ AA-AFOs (or in the case of the articulated and semi-flexible AFOs, did not allow plantarflexion beyond 0 ${}^{\circ}$ ). SVAs while standing ranged from 5 ${}^{\circ}$ to 20 ${}^{\circ}$ inclined (median $=$ 10 ${}^{\circ}$ ). The range of iAA-AFOs was $-$ 5 ${}^{\circ}$ to $-$ 25 ${}^{\circ}$ (median $=-$ 15 ${}^{\circ}$ ). SVAs ranged from 9 ${}^{\circ}$ to 18 ${}^{\circ}$ (median $=$ 12 ${}^{\circ}$ ) inclined while standing.

3.3 Kinematics and kinetics

3.3.1 Ankle kinematic variables

Median AnkleDF was 12 ${}^{\circ}$ (range $=-$ 6 ${}^{\circ}$ to 45 ${}^{\circ}$ ) in the Usual AFOs and $-$ 12 ${}^{\circ}$ (range $=-$ 20 ${}^{\circ}$ to 3 ${}^{\circ}$ ) in the iAA-AFOs. Median AnklePF in the Usual AFOs was $-$ 2 ${}^{\circ}$ (range $=-$ 18 ${}^{\circ}$ to 16 ${}^{\circ}$ ) and $-$ 19 ${}^{\circ}$ (range $=-$ 9 ${}^{\circ}$ to $-$ 33 ${}^{\circ}$ ) in the iAA-AFOs. Differences between conditions were significant for AnkleDF (Z $=-$ 3.124, $p=$ 0.002) and AnklePF (Z $=-$ 2.897, $p=$ 0.004). GVS_Ankle in Usual (median $=$ 9; range $=$ 4–25) and iAA-AFO (median $=$ 24; range $=$ 11–32) conditions were significantly different (Z $=-$ 3.067, $p=$ 0.002).

3.3.2 Knee, hip, segment, and spatiotemporal variables

Of the 15 limbs, 9 (60%) were net positive responders, 6 (40%) were net negative responders (Table 4). For these 25 variables, 12/15 limbs demonstrated $\geqslant$ 1 positive response (median $=$ 3 positive responses; range $=$ 0–12), and 11/15 limbs demonstrated $\geqslant$ 1 negative response (median $=$ 1 negative response; range $=$ 0–9).

Variables most often affected positively were: KneeExt (40%), GVS_knee (33%), KneeMom_TMSt (31%), and KneeFlex (27%) (Fig. 3, Table 4, Appendix). GVS for the foot-floor angle (GVS_FFA) (33%), GVS for vertical GRF (GVS_VF) (23%), and HipROM (20%) were most often negatively affected. Spatiotemporal variables were minimally impacted, and effects were greater at the knee than hip. Effect sizes for positive and negative responses ranged from 1.5 to 17.7 and $-$ 1.4 to $-$ 6.8 respectively.

4. Discussion

This study evaluated the effects of solid AFOs with iAA-AFOs and optimized static alignment on gait for children with CP and equinus. Results were mixed, with a net positive result for 5/10 children (60% of limbs), a net negative response for 2/10 children (40% of limbs), and one positive and one negative limb response for 3/10 children. While 9/10 children (12/15 or 80% of limbs) had some positive change with the iAA-AFOs, 8/10 children (11/15 or 73% of limbs) also showed some negative change. Although the results suggest that individualized prescription can potentially alter gait mechanics compared to conventionally prescribed AFOs, we were unable to identify the factors predicting or contributing to each child’s result. For those limbs that improved, the greatest effects were observed for knee kinematics and kinetics. Positive and negative effects for the hip, FFA, and SVA were observed with equal frequency, while spatiotemporal effects were equivocal. Our hypothesis was therefore partially supported, although the variation in individual responses makes it difficult to generalize about group response [41]. The results may assist clinicians in making more evidence-based, individualized orthotic decisions, and inform future research on this complex topic.

These results highlight potential proximal effects of a statically aligned solid iAA-AFO that considers gastrocnemius length and stiffness. The most frequently observed benefits – namely, more typical knee kinetics and kinematics at initial contact and throughout stance phase – may have important implications, given that many interventions for children with CP target knee ROM and gait kinematics. It is difficult to draw conclusions about the specific biomechanical differences between the Usual AFOs and iAA-AFOs, due to the variety of Usual AFO designs worn by participants and the individualized adjustments involved in statically aligning the iAA-AFOs; but in the iAA-AFO condition, it is likely that the solid AFO design and the individualized AA-AFOs both impacted gait outcomes. The solid iAA-AFOs may have facilitated better control of the GRF relative to the knee (and therefore, better control of shank and knee kinematics) in comparison to the more flexible Usual AFOs. At the same time, the iAA-AFO may have affected dynamic ankle and knee joint coupling; similarly, increased knee extension at initial contact has been observed after gastrocnemius-soleus tenotomies for equinus [42].

In practice, AFO-FC prescription is an iterative, multidisciplinary process [28], involving a series of opportunities for assessment, gait training, and biomechanical adjustments to the AFO-FC. Although the static alignment of the AFO-FCs was optimized during standing, limbs that responded negatively to the iAA-AFO might have benefitted from further AFO-FC tuning to optimize alignment for walking. While most participants demonstrated gait patterns that have been classified as amenable to tuning (e.g., knees extended and $<$ 20 ${}^{\circ}$ knee flexion at IC [8, 14]), the results show that some SVAs were not dynamically optimized in temporal midstance (TMSt) and highlight that static optimization for standing does not necessarily tune an AFO-FC for walking. These results are consistent with current definitions of tuning [20] but contradict reports that SVAs are equivalent during static standing and at TMSt [17]. For example, participant 1 demonstrated left knee hyperextension and insufficiently inclined SVA at TMSt that were not evident during the static alignment process (Fig. 3B); she may have benefitted from further adjustments to the SVA during stance phase to increase shank incline, or to the length of the PLR to optimize the location of the GRF during exit from midstance (e.g., [10, 20, 22, 25]). Therefore, standard practice following iAA-AFO fitting and static alignment should also include optimization of SVA alignment during walking [10, 15, 18, 43].

Adequate AFO stiffness also influences proximal mechanics [18, 44]. Although the solid AFOs were intended to prevent ankle motion, visual analysis of ankle joint kinematics in this study suggested that most solid AFOs were not stiff enough to maximize SVA control and optimize proximal effects. For some participants, excessive ankle plantarflexion at loading response ( $\geqslant$ 10 ${}^{\circ}$ ) may have contributed to increased knee hyperextension and shank recline, as insufficient stiffness prevented the AFO from facilitating normal stance phase shank inclination. In other cases, the iAA-AFO orthoses allowed $\geqslant$ 15 ${}^{\circ}$ of relative ankle dorsiflexion, potentially failing to optimize proximal benefits. As the design of the iAA-AFO orthoses reflects accepted orthotic practice, these observations imply that stiffer designs may better support the achievement of certain biomechanical goals.

The ubiquity of the 90 ${}^{\circ}$ AA-AFO in clinical practice was reflected in the large difference between each child’s conventionally prescribed Usual AA-AFO and his/her dorsiflexion PROM. As expected, differences between AFO conditions for AnkleDF and AnklePF variables were consistent with the AA-AFOs. Therefore, individualized AA-AFO prescription may require acknowledgement of potential trade-offs and prioritization of orthotic goals. For example, proximal kinematic gains may necessitate a more plantarflexed AA-AFO. This possibility could deter clinicians concerned about plantarflexed AA-AFOs causing or failing to address plantarflexion contractures [28]; however, AFOs have not been shown to increase ankle dorsiflexion ROM or prevent contractures [18]. Furthermore, ankle dorsiflexion PROM $<$ 0 ${}^{\circ}$ has been identified as a contraindication for 90 ${}^{\circ}$ AA-AFOs and limited dorsiflexion PROM (e.g., $<$ 10 ${}^{\circ}$ ; [12]) as a contraindication for articulated, free dorsiflexion AFOs [9, 10, 18, 19].

These results suggest that AFO prescription is more complex than the literature portrays. Recommendations based on gait pattern or topographical distribution may imply that orthotic prescription is simpler for children with higher levels of motor function [41]. For example, several authors suggest that articulated AFOs are more effective than solid AFOs for children with hemiplegia [41, 45]. Our results did not substantiate this presumption. However, all three children with hemiplegia who wore articulated AFOs demonstrated a net positive response to the solid iAA-AFO. Our results are consistent with reports that orthotic goals are not preferentially achieved for children in higher functioning (i.e., lower) GMFCS levels [46]; for example, only 57% (4/7) of limbs for children at GMFCS level I responded positively, comprising 44% of the positive responder group. In practice, selection of articulated vs solid AFOs appears inconsistent for children at lower GMFCS levels, while solid AFOs may be consistently recommended for children with more severe motor impairments [27].

Clinical algorithms have been used in medicine since the 1970s to describe and personalize intervention, and reduce trial-and-error decision making [47, 48]. Although clinicians endeavor to individualize AFO prescriptions for children with CP [27], decision-making is complex. Presently, this process is variable and relies on trial-and-error and anecdotal evidence [28], indicating a need for clinical practice guidelines [48]. Similar to other complex aspects of rehabilitation for children with CP (e.g., [44]), algorithms are likely to be valuable in the development of evidence-based guidelines for orthotic intervention.

The algorithm used to determine the iAA-AFO is unique for this purpose; however, as it is based on theoretical justification, clinical experience, and expert opinion, a systematic literature review and further research would help link recommendations with evidence. A future step may be to create an algorithm to guide visual gait analysis for tuning and biomechanical optimization following iAA-AFO fitting.

4.1 Limitations

The small sample size (which included children with both hemiplegia and diplegia) and variety of Usual AFO types preclude conclusions about the reasons underlying participants’ responses to the solid AFO with iAA-AFO. Replication with larger samples, comparison only to solid AFOs, and statistical modelling may help to identify factors affecting individual responses and reduce confounding factors such as differences between motor impairment (e.g., between GMFCS levels and unilateral vs bilateral involvement), and in AFO design. Experimental evaluation of the algorithm used to determine the AA-AFO would be beneficial to assess its reliability and validity. For example, the methodology assumes the ankle measurements are a proxy for gastrocnemius length; however, other tissues may be involved as well. Future research on individualized orthotic intervention should examine biomechanical, motor learning, and functional changes with task-oriented training along with the effects on community walking and participation.

Finally, the extent to which the AFOs controlled pronation or midfoot break is unknown. Dorsiflexion PROM was greater with the foot pronated than supinated for 12/15 limbs, and visually, most feet pronated during barefoot gait (Appendix). Uncontrolled pronation and midtarsal joint dorsiflexion may explain how AnkleDF values in the Usual AFOs surpassed clinically measured PROM for many participants. A large lever arm in terminal stance also contributed to stance phase ankle dorsiflexion; however, concomitant motion and chronic strain at joints other than the ankle (e.g., subtalar pronation or midtarsal dorsiflexion) [22, 23] could potentially contribute to long-term foot pain – a prevalent concern for adults with CP [6]. As clinicians may assume that dorsiflexion occurs at the talocrural joint in conventional 90 ${}^{\circ}$ AA-AFOs or articulated AFOs, this issue warrants further exploration for children with equinus. Other topics for future longitudinal research include the relationship between gastrocnemius length, foot deformity, and pain, along with the effect of plantarflexed AA-AFOs on contractures.

5. Conclusion

This study demonstrates that statically aligned solid AFOs with individualized AA-AFOs can impact gait for some children with CP and equinus, compared to conventionally prescribed AFOs. The observed positive and negative effects highlight the complexity of orthotic assessment and prescription, along with the potential to alter gait and address biomechanical goals by individualizing orthotic prescription. In addition to static optimization, objective evaluation to optimize dynamic alignment of iAA-AFOs for walking and consideration of AFO design stiffness may further improve outcomes. Future development of evidence-based guidelines that incorporate clinical algorithms is an important step toward consistent high-quality, individualized orthotic care for children with CP.

Footnotes

Acknowledgments

The authors would like to thank the children and families who participated and the orthotists at the Saskatchewan Abilities Council and Wascana Rehabilitation Centre who assisted with orthotic fabrication and fitting. Thanks also to Ali Bell for valuable statistical consultation and Aaron Awdhan for assistance with motion analysis testing. This work was funded by the Pedorthic Research Foundation of Canada and a joint award from the Physiotherapy Foundation of Canada and the Neurosciences Division of the Canadian Physiotherapy Association. The funders were not involved in any aspect of study design, data collection, analysis or interpretation, or in writing the publication.

Conflict of interest

The authors have no conflict of interest to report.

Appendix

Clinical assessment findings for participants with cerebral palsy: A. Clinical assessment findings for non-weightbearing lower extremity alignment ${}^{*}$ ; B. Modified Tardieu Scale (R1 and R2 angles) ${}^{**}$ ; C. Observational gait analysis ${}^{*}$ .

Child (limb)	Forefoot posture ${}^{\text{a}}$	Foot flexibility	Knee extension PROM ${}^{\text{b}}$	Thigh-foot angle ${}^{\text{c}}$	Femoral anteversion	Duncan ely test	Thomas test	LLD
A.
1 (R)	WNL	Flexible to neutral	$-$ 10 ${}^{\circ}$	$-$ 3 ${}^{\circ}$	yes	WNL	$-$ 5 ${}^{\circ}$	1 cm short
1 (L)	Equinus	Flexible to neutral	0 ${}^{\circ}$	WNL	WNL	WNL	WNL	–
2 (R)	Moderate MTA	Flexible	0 ${}^{\circ}$	0 ${}^{\circ}$	WNL	WNL	WNL	1.5 cm short
3 (R)	Equinus	Fixed	0 ${}^{\circ}$	WNL	WNL	WNL	$-$ 5 ${}^{\circ}$	2 cm short
3 (L)	WNL	Flexible to neutral	0 ${}^{\circ}$	WNL	WNL	WNL	$-$ 5 ${}^{\circ}$	–
4 (R)	Severe MTA	Fixed	10 ${}^{\circ}$	0 ${}^{\circ}$	Yes	Positive	$-$ 5 ${}^{\circ}$	–
4 (L)	Severe MTA	Fixed	0 ${}^{\circ}$	5 ${}^{\circ}$	WNL	WNL	WNL	0.5 cm short
5 (L)	WNL	Flexible	0 ${}^{\circ}$	0 ${}^{\circ}$	Yes	WNL	WNL	1 cm short
6 (R)	Equinus	Flexible	0 ${}^{\circ}$	2 ${}^{\circ}$	WNL	Positive	$-$ 5 ${}^{\circ}$	1 cm short
7 (R)	Hallux valgus; Mild MTA	Flexible	0 ${}^{\circ}$	15 ${}^{\circ}$	WNL	WNL	WNL	1.5 cm short
8 (R)	Severe MTA	Fixed MTA	0 ${}^{\circ}$	$-$ 5 ${}^{\circ}$	WNL	Positive	WNL	0.5 cm short
9 (R)	Moderate MTA	Flexible	10 ${}^{\circ}$	0 ${}^{\circ}$	WNL	WNL	$-$ 10 ${}^{\circ}$	1 cm short
9 (L)	Moderate MTA	Flexible	$-$ 5 ${}^{\circ}$	0 ${}^{\circ}$	Yes	WNL	WNL	–
10 (R)	Moderate MTA	Flexible to neutral	0 ${}^{\circ}$	5 ${}^{\circ}$	WNL	WNL	WNL	–
10 (L)	Moderate MTA	Flexible to neutral	0 ${}^{\circ}$	5 ${}^{\circ}$	WNL	WNL	$-$ 15 ${}^{\circ}$	0.5 cm short

${}^{*}$ Evaluation measures performed according to Cusick [29] and Owen [21] and standard physical therapy practice except where indicated. Findings were WNL on the unaffected side for children with unilateral CP. WNL: Within normal limits; FPA: Foot progression angle; MTA: Metatarsus adductus; LLD: Leg length discrepancy; PROM: passive range of motion (measured by goniometer). ${}^{\text{a}}$ Mild $=$ foot axis through 3 ${}^{\text{rd}}$ toe; moderate $=$ foot axis between 3 ${}^{\text{rd}}$ and 4 ${}^{\text{th}}$ toes; severe $=$ foot axis between 4 ${}^{\text{th}}$ and 5 ${}^{\text{th}}$ toes. ${}^{\text{b}}$ Positive values denote hyperextension and negative values denote contracture. ${}^{\text{c}}$ Positive values denote external torsion and negative values denote internal torsion.

B. Modified tardieu scale ${}^{**}$
Child (limb)	Hip extension		Hip abduction		Hip external rotation		Hip internal rotation		Knee extension		Knee flexion		Ankle DF at 90 ${}^{\circ}$ knee flexion		Ankle DF at 0 ${}^{\circ}$ knee extension
	R1	R2	R1	R2	R1	R2	R1	R2	R1	R2	R1	R2	R1	R2	R1	R2
1 (R)	WNL	45 ${}^{\circ}$	30 ${}^{\circ}$	30 ${}^{\circ}$	WNL	45 ${}^{\circ}$	0 ${}^{\circ}$	45 ${}^{\circ}$	WNL	WNL	$-$ 90 ${}^{\circ}$	$-$ 50 ${}^{\circ}$	0 ${}^{\circ}$	0 ${}^{\circ}$	$-$ 30 ${}^{\circ}$	$-$ 20 ${}^{\circ}$
1 (L)	WNL	45 ${}^{\circ}$	30 ${}^{\circ}$	30 ${}^{\circ}$	WNL	35 ${}^{\circ}$	0 ${}^{\circ}$	50 ${}^{\circ}$	WNL	WNL	$-$ 85 ${}^{\circ}$	$-$ 50 ${}^{\circ}$	$-$ 5 ${}^{\circ}$	0 ${}^{\circ}$	$-$ 30 ${}^{\circ}$	$-$ 15 ${}^{\circ}$
2 (R)	50 ${}^{\circ}$	WNL	25 ${}^{\circ}$	65 ${}^{\circ}$	WNL	65 ${}^{\circ}$	WNL	55 ${}^{\circ}$	WNL	WNL	$-$ 90 ${}^{\circ}$	$-$ 45 ${}^{\circ}$	0 ${}^{\circ}$	0 ${}^{\circ}$	$-$ 16 ${}^{\circ}$	$-$ 10 ${}^{\circ}$
2 (L)	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$
3 (R)	WNL	WNL	WNL	WNL	WNL	35 ${}^{\circ}$	WNL	WNL	WNL	WNL	$-$ 50 ${}^{\circ}$	$-$ 26 ${}^{\circ}$	$-$ 10 ${}^{\circ}$	5 ${}^{\circ}$	$-$ 25 ${}^{\circ}$	0 ${}^{\circ}$
3 (L)	WNL	WNL	WNL	WNL	WNL	45 ${}^{\circ}$	WNL	WNL	WNL	WNL	WNL	$-$ 32 ${}^{\circ}$	5 ${}^{\circ}$	10 ${}^{\circ}$	$-$ 10 ${}^{\circ}$	0 ${}^{\circ}$
4 (R)	WNL	WNL	WNL	30 ${}^{\circ}$	WNL	55 ${}^{\circ}$	WNL	70 ${}^{\circ}$	WNL	WNL	$-$ 80 ${}^{\circ}$	$-$ 40 ${}^{\circ}$	WNL	5 ${}^{\circ}$	$-$ 20 ${}^{\circ}$	$-$ 4 ${}^{\circ}$
4 (L)	WNL	WNL	WNL	30 ${}^{\circ}$	10 ${}^{\circ}$	40 ${}^{\circ}$	WNL	70 ${}^{\circ}$	WNL	WNL	$-$ 60 ${}^{\circ}$	$-$ 35 ${}^{\circ}$	WNL	5 ${}^{\circ}$	$-$ 20 ${}^{\circ}$	$-$ 7 ${}^{\circ}$
5 (R)	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	70 ${}^{\circ}$	$\backslash$	70 ${}^{\circ}$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$
5 (L)	WNL	WNL	WNL	WNL	WNL	75 ${}^{\circ}$	WNL	60 ${}^{\circ}$	WNL	WNL	$-$ 50 ${}^{\circ}$	$-$ 35 ${}^{\circ}$	WNL	WNL	$-$ 40 ${}^{\circ}$	5 ${}^{\circ}$
6 (R)	WNL	WNL	WNL	WNL	WNL	60 ${}^{\circ}$	WNL	40 ${}^{\circ}$	WNL	WNL	$-$ 60 ${}^{\circ}$	$-$ 45 ${}^{\circ}$	WNL	7 ${}^{\circ}$	$-$ 15 ${}^{\circ}$	$-$ 4 ${}^{\circ}$
6 (L)	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$
7 (R)	WNL	WNL	WNL	WNL	WNL	70 ${}^{\circ}$	WNL	70 ${}^{\circ}$	WNL	WNL	$-$ 55 ${}^{\circ}$	$-$ 45 ${}^{\circ}$	WNL	6 ${}^{\circ}$	$-$ 18 ${}^{\circ}$	0 ${}^{\circ}$
7 (L)	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$
8 (R)	WNL	WNL	WNL	WNL	WNL	WNL	WNL	WNL	WNL	WNL	$-$ 80 ${}^{\circ}$	$-$ 40 ${}^{\circ}$	15 ${}^{\circ}$	25 ${}^{\circ}$	$-$ 15 ${}^{\circ}$	2 ${}^{\circ}$
8 (L)	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$	$\backslash$
9 (R)	WNL	WNL	WNL	WNL	WNL	WNL	WNL	WNL	WNL	WNL	$-$ 50 ${}^{\circ}$	0 ${}^{\circ}$	0 ${}^{\circ}$	15 ${}^{\circ}$	$-$ 15 ${}^{\circ}$	8 ${}^{\circ}$
9 (L)	WNL	WNL	WNL	WNL	WNL	WNL	WNL	WNL	WNL	WNL	$-$ 40 ${}^{\circ}$	0 ${}^{\circ}$	0 ${}^{\circ}$	18 ${}^{\circ}$	0 ${}^{\circ}$	10 ${}^{\circ}$
10 (R)	WNL	WNL	WNL	WNL	WNL	60 ${}^{\circ}$	WNL	45 ${}^{\circ}$	WNL	WNL	$-$ 50 ${}^{\circ}$	$-$ 30 ${}^{\circ}$	WNL	10 ${}^{\circ}$	$-$ 15 ${}^{\circ}$	3 ${}^{\circ}$
10 (L)	WNL	WNL	WNL	WNL	WNL	50 ${}^{\circ}$	WNL	50 ${}^{\circ}$	WNL	WNL	$-$ 60 ${}^{\circ}$	$-$ 40 ${}^{\circ}$	WNL	10 ${}^{\circ}$	$-$ 30 ${}^{\circ}$	0 ${}^{\circ}$

${}^{**}$ Measured in degrees using a universal goniometer according to Boyd and Graham [30]; DF: dorsiflexion; R: Right; L: Left; R1: angle of first resistance/catch; R2: full passive range of motion [30]; WNL: Within normal limits (i.e., WNL for R1 indicates no catch).

C. Gait observations
Child (limb)	Foot contact and knee posture at initial contact	Foot posture in stance phase	Foot progression angle in stance phase	SVA in stance ${}^{\text{d}}$	Gait classification ${}^{\text{e}}$
1 (R)	Forefoot	Pronation, midfoot break	$\sim$ 30 ${}^{\circ}$ internal	Excess incline	Group V – asymmetrical: Crouch gait on R; True Equinus on L
1 (L)	Forefoot; excess KF	Neutral; mild pronation	Neutral	WNL
2 (R)	Flatfoot	Inversion/supination, midfoot break, excessive pronation	Neutral	WNL	Type 2A – True equinus/plus neutral knee
3 (R)	Flatfoot, foot slap, and KF	Calcaneovalgus, pronates excessively	Neutral	WNL	Group I – True equinus
3 (L)	Heel and mild KF	Mild pronation	Neutral	WNL
4 (R)	Forefoot or flatfoot	Calcaneovarus	Internal	Excess incline	Group IV – Crouch in early to midstance
4 (L)	Heel	Calcaneovarus	Neutral	Excess incline
5 (L) Left	Forefoot	Equinus, inversion, hindfoot varus; rarely achieves heel	$\sim$ 45 ${}^{\circ}$ internal	Excess incline	Type IV
		contact
6 (R)	Flatfoot	WNL	Neutral	WNL	Type 2A – True equinus plus neutral knee
7 (R)	Flatfoot	Cavus with forefoot equinus; pronates excessively	$\sim$ 25 ${}^{\circ}$ external	Excess incline	Type 3 – True equinus with jump knee
8 (R)	Forefoot, knee and hip flexion	Supination	$\sim$ 40 ${}^{\circ}$ internal	Insufficient incline	Type 2B – True equinus/plus recurvatum knee – H
9 (R)	Forefoot, and with KHE	Cavus	$\sim$ 10 ${}^{\circ}$ –20 ${}^{\circ}$ internal	Excessive incline	Group III – Apparent equinus
9 (L)	Forefoot	Midfoot break	$\sim$ 10 ${}^{\circ}$ –20 ${}^{\circ}$ internal
10 (R)	Forefoot contact and KF	Midfoot break; hindfoot varus; supinates or pronates	$\sim$ 40 ${}^{\circ}$ internal	Insufficient incline	Group I – True equinus with recurvatum
10 (L)	Forefoot contact and KF	Midfoot break; supination	$\sim$ 20 ${}^{\circ}$ internal	Insufficient incline

R: Right; L: Left; KF: Knee flexion; KE: Knee extension; KHE: Knee hyperextension; DF: Dorsiflexion. ${}^{\text{d}}$ Gait classification according to Owen [10, 21], based on barefoot 3D shank kinematics in stance phase. ${}^{\text{e}}$ Gait classification according to Rodda and Graham [11], based on barefoot 3D ankle and knee kinematics in stance phase; classifications for spastic diplegia were used for participants with bilateral CP, and classifications for hemiplegia were applied for participants with unilateral CP.

References

Rosenbaum

, Paneth

, Leviton

, Goldstein

, Bax

, Damiano

, et al. A report: The definition and classification of cerebral palsy. Dev Med Child Neurol Suppl. 2007; 109: 8–14. doi: 10.1111/j.1469-8749.2007.tb12610.x.

Wren

TAL

, Rethlefsen

, Kay

. Prevalence of specific gait abnormalities in children with cerebral palsy: Influence of cerebral palsy subtype, age, and previous surgery. J Pediatr Orthop. 2005; 25(1): 79–83. doi: 10.1097/00004694-200501000-00018.

Davids

. Orthopedic treatment of foot deformities. In: Gage

, Schwartz

, Koop

, Novacheck

, eds. The Identification and Treatment of Gait Problems in Cerebral Palsy. London: Mac Keith Press; 2009. pp. 514–33.

Goldstein

, Harper

. Management of cerebral palsy: Equinus gait. Dev Med Child Neurol. 2001; 43: 563–9. doi: 10.1017/s0012162201001025.

Lieber

. Skeletal muscle structure, function and plasticity: the physiological basis of rehabilitation. 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2002.

Jahnsen

, Villien

, Aamodt

, Stanghelle

, Holm

. Musculoskeletal pain in adults with cerebral palsy compared with the general population. J Rehabil Med. 2004; 36(3): 78–84. doi: 10.1080/16501970310018305.

Murphy

, Molnar

, Lankasky

. Medical and functional status of adults with cerebral palsy. Dev Med Child Neurol. 1995; 37: 1075–84. doi: 10.1111/j.1469-8749.1995.tb11968.x.

Jagadamma

, Coutts

, Mercer

, Herman

, Yirrell

, Forbes

, et al. Optimising the effects of rigid ankle foot orthoses on the gait of children with cerebral palsy (CP) – an exploratory trial. Disabil Rehabil Assist Technol. 2015; 10(6): 445–51. doi: 10.3109/17483107.2014.908244.

Ridgewell

, Dobson

, Bach

, Baker

. A systematic review to determine best practice reporting guidelines for AFO interventions in studies involving children with cerebral palsy. Prosthet Orthot Int. 2010; 34(2): 129–45. doi: 10.3109/03093641003674288.

10.

Owen

. The importance of being earnest about shank and thigh kinematics especially when using ankle-foot orthoses. Prosthet Orthot Int. 2010; 34: 254–69. doi: 10.3109/03093646.2010.485597.

11.

Rodda

, Graham

. Classification of gait patterns in spastic hemiplegia and spastic diplegia: A basis for a management algorithm. Eur J Neurol. 2001; 8Suppl 5(3): 98–108. doi: 10.1046/j.1468-1331.2001.00042.x.

12.

Owen

. A proposed clinical algorithm for dorsiflexion free AFOFCs based on calf muscle length, strength, stiffness and skeletal alignment. In: American Academy of Orthotists & Prosthetists 41st Academy Annual Meeting and Scientific Symposium. 2015.

13.

Butler

, Nene

. The biomechanics of fixed ankle foot orthoses and their potential in the management of cerebral palsied children. Physiotherapy. 1991; 77(2): 81–8. doi: 10.1016/S0031-9406(10)63580-8.

14.

Butler

, Farmer

, Stewart

, Jones

, Forward

. The effect of fixed ankle foot orthoses in children with cerebral palsy. Disabil Rehabil Assist Technol. 2007; 2(1): 51–8. doi: 10.1080/17483100600662009.

15.

Meadows

, Bowers

, Owen

. Biomechanics of the hip, knee and ankle. In: Hsu

, Michael

, Fisk

, eds. AAOS Atlas of Orthoses and Assistive Devices. 4th ed. Philadelphia, PA: MOSBY Elsevier; 2008. pp. 299–309.

16.

Eddison

, Chockalingam

, Osborne

. Ankle foot orthosis-footwear combination tuning: An investigation into common clinical practice in the United Kingdom. Prosthet Orthot Int. 2015; 39(2): 126–33. doi: 10.1177/0309364613516486.

17.

Eddison

, Healy

, Needham

, Chockalingam

. Shank-to-vertical angle in ankle-foot orthoses: A comparison of static and dynamic assessment in a series of cases. J Prosthetics Orthot. 2017; 29(4): 161–7. doi: 10.1097/JPO.0000000000000141.

18.

Bowers

, Ross

. A review of the effectiveness of lower limb orthoses used in cerebral palsy. In: Morris

, Condie

, eds. Recent developments in healthcare for cerebral palsy: Implications and opportunities for orthotics. Copenhagen Denmark: International Society for Prosthetics and Orthotics; 2009. pp. 235–97.

19.

Eddison

, Chockalingam

. The effect of tuning ankle foot orthoses-footwear combination on the gait parameters of children with cerebral palsy. Prosthet Orthot Int. 2013; 37: 95–107. doi: 10.1177/0309364612450706.

20.

Owen

. Defining what we do. J Prosthetics Orthot. 2018; 30(1): 2–4. doi: 10.1097/JPO.0000000000000175.

21.

Owen

. Pediatric Gait Analysis and Orthotic Management: Optimal Segment Kinematics and Alignment Approach to Rehabilitation [Course notes]. 2018.

22.

Karas

. Compensatory midfoot dorsiflexion in the individual with heelcord tightness: Implications for orthotic device designs. J Prosthetics Orthot. 2002; 14(2): 82–93.

23.

Johanson

, Dearment

, Hines

, Riley

, Martin

, Thomas

, et al. The effect of subtalar joint position on dorsiflexion of the ankle/rearfoot versus midfoot/forefoot during gastrocnemius stretching. Foot Ankle Int. 2014; 35: 63–70. doi: 10.1177/1071100713513433.

24.

Owen

. The point of “point loading rockers” in ankle-foot orthosis footwear combinations used with children with cerebral palsy, spina bifida and other conditions. Gait Posture. 2004; 20S: S61–112.

25.

Nuzzo

. High-performance activity with below-knee cast treatment, Part I: Mechanics and demonstration. Orthopedics. 1983; 6(6): 713–23. doi: 10.3928/0147-7447-19830601-06.

26.

Owen

. Proposed clinical algorithm for deciding the sagittal angle of the ankle in an ankle-foot orthosis footwear combination. Gait Posture. 2005; 22(Supplement 1): S38–39.

27.

Kane

, Lanovaz

, Musselman

. Physical therapists’ use of evaluation measures to inform the prescription of ankle-foot orthoses for children with cerebral palsy. Phys Occup Ther Pediatr. 2019; 39(3): 237–253. doi: 10.1080/01942638.2018.1463586.

28.

Kane

, Manns

, Lanovaz

, Musselman

. Clinician perspectives and experiences in the prescription of ankle-foot orthoses for children with cerebral palsy. Physiother Theory Pract. 2019; 35(2): 148–156. doi: 10.1080/09593985.2018.1441346.

29.

Cusick

. Legs & Feet: A review of Musculoskeletal Assessment Procedures for Children & Adults [videorecording]: Version 3.0. Placerville, CO: Progressive GaitWays, LLC; 2006.

30.

Boyd

, Graham

. Objective measurement of clinical findings in the use of botulinum toxin type A for the management of children with cerebral palsy. Eur J Neurol. 1999; 6: s23–35.

31.

Ferriero

, Vercelli

, Sartorio

, Muñoz Lasa

, Ilieva

, Brigatti

, Ruella

, Foti

. Reliability of a smartphone-based goniometer for knee joint goniometry. Int J Rehabil Res. 2013; 36(2): 146–51. doi: 10.1097/MRR.0b013e32835b8269.

32.

Palisano

, Rosenbaum

, Bartlett

, Livingston

. Gross Motor Function Classification System Expanded and Revised. CanChild Centre for Childhood Disability Research, McMaster University; 2007. [Cited 2018 May 1]. Available from: https://canchild.ca/system/tenon/assets/attachments/000/000/058/original/GMFCS-ER_English.pdf.

33.

Owen

. Shank angle to floor measures and tuning of Ankle-foot orthosis footwear-combinations for children with cerebral palsy, spina bifida and other conditions. University of Strathclyde, Glasgow; 2004.

34.

Winter

. Biomechanics and motor control of human movement. 4th ed. Hoboken, NJ: John Wiley & Sons, Ltd; 2009.

35.

Jensen

. Body segment mass, radius and radius gyration proportions of children. J Biomech. 1986; 19(5): 359–68. doi: 10.1016/0021-9290(86)90012-6.

36.

Baker

, McGinley

, Schwartz

, Beynon

, Rozumalski

, Graham

, et al. The gait profile score and movement analysis profile. Gait Posture. 2009; 30(3): 265–9. doi: 10.1016/j.gaitpost.2009.05.020.

37.

Cimolin

, Galli

. Summary measures for clinical gait analysis: A literature review. Gait Posture. 2014; 39(4): 1005–10. doi: 10.1016/j.gaitpost.2014.02.001.

38.

Danino

, Erel

, Kfir

, Khamis

, Hemo

, Wientroub

, et al. Do gait indices reflect the effect of ankle foot orthosis on gait impairment in patients with diplegic cerebral palsy. Dev Med Child Neurol. 2012; 54: 36–7.

39.

Damiano

. Meaningfulness of mean group results for determining the optimal motor rehabilitation program for an individual child with cerebral palsy. Dev Med Child Neurol. 2014; 56(12): 1141–6. doi: 10.1111/dmcn.12505.

40.

Landis

, Koch

. The measurement of observer agreement for categorical data. Biometrics. 1977; 33(1): 159–74.

41.

Buckon

, Thomas

, Jokobson-Huston

, Moor

, Sussman

, Aiona

. Comparison of three ankle-foot orthosis configurations for children with spastic diplegia. Dev Med Child Neurol. 2004; 46(9): 590–8. doi: 10.1017/s0012162204001008.

42.

Baddar

, Granata

, Damiano

, Carmines

, Blanco

, Abel

. Ankle and knee coupling in patients with spastic diplegia: Effects of gastrocnemius-soleus lengthening. J Bone Joint Surg Am. 2002; 84A: 736–44. doi: 10.2106/00004623-200205000-00006.

43.

Jagadamma

, Owen

, Coutts

, Herman

, Yirrell

, Mercer

, et al. The effects of tuning an ankle-foot orthosis footwear combination on kinematics and kinetics of the knee joint of an adult with hemiplegia. Prosthet Orthot Int. 2010; 34(3): 270–6. doi: 10.3109/03093646.2010.503225.

44.

Kerkum

, Buizer

, Van Den Noort

, Becher

, Harlaar

, Brehm

. The effects of varying ankle foot orthosis stiffness on gait in children with spastic cerebral palsy who walk with excessive knee flexion. PLoS One. 2015; 10(11): e0142878. doi: 10.1371/journal.pone.0142878.

45.

Aboutorabi

, Arazpour

, Ahmadi Bani

, Saeedi

, Head

. Efficacy of ankle foot orthoses types on walking in children with cerebral palsy: A systematic review. Ann Phys Rehabil Med. 2017; 60(6): 393–402. doi: 10.1016/j.rehab.2017.05.004.

46.

Wingstrand

, Hägglund

, Rodby-Bousquet

. Ankle-foot orthoses in children with cerebral palsy: A cross sectional population based study of 2200 children. BMC Musculoskeletal Disorders. 2014; 1–7. doi: 10.1186/1471-2474-15-327.

47.

Federer

, Taylor

, Mather

. Using evidence-based algorithms to improve clinical decision making: The case of a first-time anterior shoulder dislocation. Sports Med Arthrosc. 2013; 21(3): 155–65. doi: 10.1097/JSA.0b013e31829f608c.

48.

Sox

, Stewart

. Algorithms, clinical practice guidelines, and standardized clinical assessment and management plans: Evidence-based patient management standards in evolution. Acad Med. 2015; 90(2): 129–32. doi: 10.1097/ACM.0000000000000509.

49.

Fehlings

, Switzer

, Agarwal

, Wong

, Sochett

, Stevenson

, et al. Informing evidence-based clinical practice guidelines for children with cerebral palsy at risk of osteoporosis: A systematic review. Dev Med Child Neurol. 2012; 54(2): 106–16. doi: 10.1111/j.1469-8749.2011.04091.x.

Effects of solid ankle-foot orthoses with individualized ankle angles on gait for children with cerebral palsy and equinus

Abstract

PURPOSE:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Methods

2.1 Participants

2.2 Procedures

2.2.1 Clinical assessment and determination of individualized AA-AFO

2.2.2 Orthoses and footwear

2.3 Analyses

Table 3 Characteristics of the Usual AFOs and iAA-AFOs, and description of adjustments made during the static alignment process

3.1 Physical assessment

Table 4 Individual variable response profiles by limb according to net response: A. Limbs with net positive response; B. Limbs with net negative response. Effect size (Cohen’s d) values inside red and green boxes indicate magnitude of positive and negative responses

3.3 Kinematics and kinetics

3.3.1 Ankle kinematic variables

3.3.2 Knee, hip, segment, and spatiotemporal variables

4. Discussion

4.1 Limitations

5. Conclusion

Footnotes

Acknowledgments

Conflict of interest

Appendix

References

Table 3
Characteristics of the Usual AFOs and iAA-AFOs, and description of adjustments made during the static alignment process

Table 4
Individual variable response profiles by limb according to net response: A. Limbs with net positive response; B. Limbs with net negative response. Effect size (Cohen’s d) values inside red and green boxes indicate magnitude of positive and negative responses