The goal of clubfoot treatment is to obtain a well-functioning foot. However, few studies exist which evaluate gross motor skills (GMS) in children with idiopathic clubfoot (IC).1–3 Minor delays in gross motor milestones have been reported in small children with clubfoot mainly regarding a later debut of independent walking.4,5 Children aged 7 years with IC demonstrate an increased prevalence of motor impairments.2 Poor motor skills seem to be associated with negative psychosocial well-being such as low self-esteem and decreased quality of life.6 Studies of gait analysis (GA) in children with IC report promising results regarding nonoperative treatments but gait deviations persist.7–9 Karol et al,1 evaluated the relationship between GA measurements and gross motor function in 5-year-old children and found that gait disturbances did not interfere with function. The initial severity of the clubfoot is an early predictor of surgical intervention but the relationship to GMS seems to be unclear.10 Moreover, the passive range of motion of the clubfoot is often described as an important outcome measurement but to our knowledge it has never been evaluated in relation to the child’s GMS.11
A recent study suggests that children with bilateral and unilateral clubfoot may represent differing severity in initial foot status.12 Even so, similar gait patterns have been reported between bilateral and unilateral clubfeet in children.13 In the contralateral foot, in children with unilateral IC, indications of foot adaptations have been found in gait and tiptoe rising.13,14 Even so, little is known of GMS in children with bilateral and unilateral clubfoot, respectively, and in relationship to typically developed (TD) children. Therefore, this study aimed to evaluate GMS in children with IC, focusing on the impact of foot involvement and asymmetries, and examine the associations between GMS, gait, passive foot motion, and initial severity in clubfeet.
We identified all children born and treated for IC in Stockholm County between 2005 and 2008 (n=122) for this study, approved by the regional ethical review board. The severity of the clubfeet was assessed using the Dimeglio Classification Score before treatment where each clubfoot is scored from 0 (benign) to 20 (very severe).15 The standard clubfoot treatment consisted of weekly above the knee serial castings and an Achilles tendon tenotomy during the first months of life. Thereafter, the corrected clubfoot was kept in a dynamic custom-made knee-ankle-foot orthosis (KAFO) with an open knee joint. Caregivers were instructed to have the child wear the KAFO for 23 hours per day during the first months and gradually reduce wear time to only nighttime and naps at one year of age. Bracing continued until the child was 4 to 5 years old, depending on the foot status. In case of recurrence, a second treatment period consisting of either casting or casting in combination with Tibialis anterior tendon transfer (TATT) were initiated. At ∼5 years of age, children participated in gross motor assessments and gait analyses at the motion laboratory to evaluate treatment outcome. At this visit, one of the motion laboratory physical therapists assessed passive foot range of motion using a standard protocol. Children were included in this study if they were 4.5 to 6.5 years of age and followed the standard clubfoot treatment program (including secondary treatment periods). The medical carts as well as the clinical findings at the time of the motion analysis were investigated for all participants, to ensure that only children with IC were included (ie, no postural or syndromic clubfeet). For comparison, 28 TD children in the same age span underwent the same procedure at the motion laboratory.
Five blinded assessors (2 orthopaedic surgeons and 3 physical therapists) rated the children’s GMS by videotapes using the Motion quality (MQ) domains from the Clubfoot Assessment Protocol (CAP). CAP is an assessment instrument developed for use in follow-up evaluation of children with clubfoot.16 The CAP consists of 5 domains: Mobility (7 items), Muscle function (2 items), Morphology (4 items), and MQ I (4 items) and II (2 items). In the MQ domains, the assessor visually assesses the quality of the items: running, walking, toe walking, heel walking, one leg stand, and one leg hop in the frontal plane. Each leg is rated separately on an ordinal scale; 0 (cannot), 1 (very deviant), 2 (deviant), 3 (slightly deviant), or 4 (within normal), when performing the task on a 10-meter walkway according to the manual. As an example, the item one leg hop is rated as 2 (deviant) if the child has problem with keeping balance; difficulty in keeping a straight path; insufficient propulsion power and/or either short hop stride or unregulated hop stride on the rated foot. The assessors were instructed to view the videos for each item, at most 3 times, without stopping or slowing down the tape. The intraclass correlation coefficient (ICC2,1) of the 5 assessors was calculated for the MQ-CAP scores (Table 1). Agreements of the 5 assessors were considered sufficient (range, 0.55 to 0.79). The median scores from the 5 assessors were used for the statistical analysis.
From the GA session, motion data were collected with a 3-dimensional (3D) 8-camera system (Vicon, Oxford Metrics, UK) and 2 force plates (Kistler, Winterthur, Switzerland) with the usage of the Vicon Plug-in-Gait lower body marker placement.17 All children were instructed to walk in a self-selected speed. To present overall measurements of the gait pattern, the Gait Deviation Index (GDI) and the Gait Deviation Index-Kinetic (GDI-Kinetic) were calculated based on a mean of 3 (GDI) or 2 (GDI-Kinetic) representative gait trials for each child. Because of difficulties in clean force strikes, not all children had sufficient kinetic data. The GDI includes 9 gait curves from the pelvis and hip in all 3 planes, the knee and ankle in the sagittal plane and the foot progression angle in the transversal plane. GDI-Kinetic includes 6 internal moment curves from the hip, knee, and ankle in the sagittal and frontal plane and the total joint power curves from the hip, knee, and ankle. Both indices are developed to quantify how far the curves deviate from the control mean (calculated from the TD children in this study) to provide one overall score for each limb. A GDI/GDI-Kinetic score of ≥100 represents a gait pattern without pathology.
For the statistical analysis, children were divided into 3 groups: bilateral or unilateral clubfoot and TD children. The feet within each group were further subclassified as superior or inferior units depending on the total scores of the MQ-CAP (Fig. 1), where the greater scoring foot for each child would be classified as superior and the lesser scoring foot as inferior. If the child had the same score for both feet, the feet were randomly allocated to a unit. This subclassification was conducted to ensure statistical independence and to evaluate asymmetries with the assumption that most children, especially children with IC, have a better performing or favorite foot. In children with unilateral clubfoot, the inferior foot was always the clubfoot. For clarification, the units in this group were therefore labeled unilateral IC or contralateral (Fig. 1). For statistical comparisons, the sign test was used to compare the superior and inferior units within each group of children and the Mann-Whitney U test was used to compare outcomes across the different units between the groups. All units were compared with each other, with the exception of the contralateral unit which was not compared with the bilateral IC units as this was not considered of interest. The Kruskal-Wallis or Mann-Whitney U tests were used to compare demographic data between groups. The score of 2 (deviant) was identified as cut-off point in MQ-CAP to define gross motor deviations. In clubfeet, the relationships between GMS, gait, and foot status (MQ-CAP, GDI, GDI-Kinetic, passive range of motion of the foot and Dimeglio Classification Score) were analyzed with the Spearman ρ correlations. Criteria for the correlation coefficient was set as follows: poor ≤0.4, moderate 0.41 to 0.6, good 0.61 to 0.8, and >0.8 very good.18
Of the 122 children with IC, 47 children (22 bilateral, 25 unilateral) were included in this study. Reasons for exclusion were as follows: 41 children had missing MQ-CAP or missing gait data, 10 did not meet the age limit, 9 had not been treated with the standard clubfoot treatment program, and 15 caregivers did not give a written consent for the child to participate (Fig. 1). Five clubfeet had undergone a second treatment period (Table 2). Limited missing data were found concerning Dimeglio Classification Scores and passive range of motion of the clubfeet (Tables 2, 3). No statistical differences in initial foot severity or treatment data were found between the children with bilateral or unilateral clubfoot, nor in comparison with TD children in age, leg length, or sex (Table 2). However, in children with unilateral clubfoot a significant mean difference in leg length by 0.4 cm (P=0.02) were detected between the clubfoot and the contralateral foot.
Significant lower scores were found in MQ-CAP and GDI/GDI-Kinetic in the clubfeet, independent of unit, compared with TD (Table 3). The only exception was the bilateral IC superior unit that did not differ in one leg hop performance. Between the bilateral IC inferior and unilateral IC units, no significant differences were detected regarding GMS and gait. However, the bilateral IC superior unit had significantly better scores for walking, one leg hop, and the total MQ-CAP score compared with unilateral IC. The contralateral unit showed significant better scores in one leg hop and total MQ-CAP but lower scores in heel walking compared with TD inferior and superior, respectively. Yet, significant lower values were found in GDI, GDI-Kinetic, and foot motion compared with TD (Table 3).
Asymmetries between the feet were discovered within all 3 groups. In children with bilateral clubfoot, asymmetries were shown in heel walking, one leg stand and one leg hop, and the total MQ-CAP between the inferior and superior feet. In children with unilateral clubfoot, asymmetries between the clubfoot and the contralateral foot were evident in all MQ-CAP items and foot motion but not in GDI and GDI-Kinetic. In TD, the only asymmetries found were in one leg hop and the total MQ-CAP score (Table 3).
Most gross motor deviations were discovered in one leg stand with deviations in between 68% and 91% of all clubfeet (Table 4). This was followed by one leg hop in the bilateral IC inferior and unilateral IC with deviations in 84% and 86% of the clubfeet. Although only 29% of the bilateral superior clubfeet deviated. Moreover, 52% to 64% of the children with clubfoot had deviations in performing toe walking and heel walking in at least one of the feet. Nearly no deviations were found in the contralateral and TD units except for the ability to stand and hop on one leg (Table 4).
Poor or moderate correlations were found between the MQ-CAP and GDI, GDI-Kinetic, foot motion, and initial severity in the clubfeet. The highest correlations found were between the items walking and heel walking with passive dorsal flexion of the foot and the Dimeglio Classification Score (Table 5).
This study reveals gross motor deficits and asymmetries in children 5 years of age with IC. Bilateral and unilateral clubfeet seem to deviate almost equally from TD feet; however, children with bilateral IC seem to develop their GMS noteworthy better on one foot. GMS showed poor to moderate relations with 3D GA, passive foot range of motion, and initial severity in clubfeet.
In our cohort, as many as 52% to 91% of the children with clubfoot had deviations in toe walking and heel walking, standing and hopping on one foot, in at least one foot. Contrary to our results, Karol et al1 concluded that the majority of children with clubfeet had average gross motor function with the Peabody Developmental Gross Motor Scale (Peabody) at age 5. The Peabody Scale assesses whole body development and ability, not specifying the quality of lower limb movements. For example, merging the ability to stand on one foot and performing sit-ups in the same domain. Andriesse and colleagues reported a prevalence of 35% of motor impairments in the children aged 7 years with IC using the Movement Assessment Battery for Children, still lower than the deficits noted in our study.2 The large number of gross motor deviations reported in our cohort is likely the result of the measurement used. The MQ-CAP assesses the quality of the motion for each lower extremity whereas, prior studies focused on the ability to complete a task.
Children with bilateral and unilateral clubfeet demonstrated similar GMS and gait patterns in the clubfeet. However, dissimilarities were detected in the bilateral superior performing clubfeet with better scores in the items walking and one leg hop and in the total MQ-CAP compared with unilateral clubfeet. Moreover, the ability to hop on one foot in the former group did not differ to TD. When considering asymmetries it seems that one foot develops substantial better than the other in children with IC as asymmetries were found to a higher extent compared with TD. These findings are consistent with our clinical impression that not only children with unilateral clubfoot have a favorite and better performing foot (the contralateral foot), but also children with bilateral clubfoot have a favorite foot when performing difficult gross motor tasks. In children with unilateral clubfoot the leg length discrepancy could have an influence of the results; however, we believe that the mean discrepancy of 0.4 cm is clinical insignificant.
The contralateral foot seems to develop GMS almost as the TD feet. The adaptations of the contralateral foot previously reported in tiptoe raising were not replicated in MQ-CAP.14 However, gait modifications of the foot were found (GDI and GDI-Kinetic) as well as small differences in passive foot motion when compared with TD children. It seems that the contralateral foot develops typically in GMS, and as previously reported in single gait parameters,13 but modifies with respect to overall gait pattern.
Only modest associations were found between GMS and gait, passive foot motion, and initial clubfeet severity. Even though passive dorsiflexion showed some relationship to MQ-CAP it cannot fully explain the gross motor deficits. Nor does the initial clubfoot severity seem to be a strong predictor of later GMS. The weak association indicates that gross motor measurements report a different outcome entity in clubfoot treatment. The results are in line with Karol et al1 and Andriesse et al2 findings of low interference between gait and foot status, respectively, with gross motor function. Other explanations, beyond musculoskeletal, of gross motor deficits might be considered such as motor coordination and focusing problems. Problems that today might be mistakenly concealed because of the child’s clubfoot.
GMS should be a component of treatment outcomes in children with clubfoot. The ability to stand or hop on one leg are 2 tasks that might be sufficient in evaluating GMS, as well as asymmetries, in children with IC. These tasks were also found to have the highest agreements with intraclass correlation coefficients of 0.66 and 0.79, respectively, between the assessors, which support their usability in the clinic. However, it needs to be taken into account that neither of the measurements used in this study evaluate the everyday activity or participation, for example, taking part in sport activities. Future research is necessary to understand the impact of the gross motor deficit in the children’s everyday living, psychosocial well-being, and the effect of motor skill training.
We recognize that the treatment method used in our cohort is a limitation and makes it difficult to generalize our results to all children with IC. Moreover, we have no further information regarding the clinical status of the foot at the time of the motion analysis. However, the results of the GDI and passive range of foot motion are comparable with previous reports of children treated with the Ponseti method.8,9 Moreover, we believe our study population embodies a representative sample of children with IC as included children were sent to the motion laboratory as a part of the standard treatment program and that foot involvement and sex distribution, are similar to published studies.19 The use of 5 blinded assessors further strengthens the generalizability of our results. Finally, the outcome measurements used in our study were chosen to mainly reflect the functional clinical outcome. Other measurements like radiographs, outcome protocols (such as the International Clubfoot Score or Laaveg-Ponseti score), foot appearance and patient-reported outcome measurements might have been relevant to include. This highlights the question what characterizes a good clinical outcome. A clinical result can be defined in many different ways and a well-corrected clubfoot may not automatically imply a good functional level. This warrant further research within the field children with IC.
In summary, our results indicate gross motor deficits and asymmetries in 5-year-old children treated for IC, irrespective of foot involvement. The contralateral foot in children with unilateral clubfoot develops almost as TD children regarding GMS but modifies in gait and passive range of motion. GMS showed weak correlations to gait and foot status and should be considered a complement in clubfoot evaluation or follow-up in the clinic.
The authors thank the children and the families that participated in the study, as well as the assessors Christina Orefelt, Gunnar Hägglund, and Minna Petersson who helped with rating the children.
1. Karol LA, Jeans KA, Kaipus KA. The relationship between gait, gross motor function, and parental perceived outcome in children with clubfeet. J Pediatr Orthop. 2016;36:145–151.
2. Andriesse H, Westbom L, Hagglund G. Motor ability in children treated for idiopathic clubfoot. A controlled pilot study. BMC Pediatr. 2009;9:78.
3. Lohle-Akkersdijk JJ, Rameckers EA, Andriesse H, et al. Walking capacity of children with clubfeet in primary school: something to worry about? J Pediatr Orthop B. 2015;24:18–23.
4. Sala DA, Chu A, Lehman WB, et al. Achievement of gross motor milestones in children with idiopathic clubfoot treated with the Ponseti method. J Pediatr Orthop. 2013;33:55–58.
5. Garcia NL, McMulkin ML, Tompkins BJ, et al. Gross motor development in babies with treated idiopathic clubfoot. Pediatr Phys Ther. 2011;23:347–352.
6. Mancini VO, Rigoli D, Cairney J, et al. The elaborated environmental stress hypothesis as a framework for understanding the association between motor skills and internalizing problems: a mini-review. Front Psychol. 2016;7:239.
7. Jeans KA, Erdman AL, Jo CH, et al. A longitudinal review of gait following treatment for idiopathic clubfoot: gait analysis
at 2 and 5 years of age. J Pediatr Orthop. 2016;36:565–571.
8. Church C, Coplan JA, Poljak D, et al. A comprehensive outcome comparison of surgical and Ponseti clubfoot treatments with reference to pediatric norms. J Child Orthop. 2012;6:51–59.
9. Duffy CM, Salazar JJ, Humphreys L, et al. Surgical versus Ponseti approach for the management of CTEV: a comparative study. J Pediatr Orthop. 2013;33:326–332.
10. Goldstein RY, Seehausen DA, Chu A, et al. Predicting the need for surgical intervention in patients with idiopathic clubfoot. J Pediatr Orthop. 2015;35:395–402.
11. Graf A, Wu KW, Smith PA, et al. Comprehensive review of the functional outcome evaluation of clubfoot treatment: a preferred methodology. J Pediatr Orthop B. 2012;21:20–27.
12. Gray K, Barnes E, Gibbons P, et al. Unilateral versus bilateral clubfoot: an analysis of severity and correlation. J Pediatr Orthop B. 2014;23:397–399.
13. Lööf E, Andriesse H, Andre M, et al. Gait in 5-year old children with idiopathic clubfoot, a cohort study of 59 children, focusing on foot involvement and the contralateral foot
. Acta Orthop. 2016;87:522–528.
14. Maton B, Wicart P. Centrally adaptations in unilateral idiopathic clubfoot children following conservative treatment. J Electromyogr Kinesiol. 2005;15:72–82.
15. Dimeglio A, Bensahel H, Souchet P, et al. Classification of clubfoot. J Pediatr Orthop B. 1995;4:129–136.
16. Andriesse H, Hagglund G, Jarnlo GB. The clubfoot assessment protocol
(CAP); description and reliability of a structured multi-level instrument for follow-up. BMC Musculoskelet Disord. 2005;6:40.
17. Davis RB, Ounpuu S, Tyburski D, et al. A gait analysis
data-collection and reduction technique. Hum Mov Sci. 1991;10:575–587.
18. Landis R, Koch G. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159–174.
19. Wallander H, Hovelius L, Michaelsson K. Incidence of congenital clubfoot
in Sweden. Acta Orthop. 2006;77:847–852.