How Much Is Enough? The Orthotic Management of Idiopathic Toe Walking

By Phil Stevens, MEd, CPO, FAAOP
silhouette of toddler walking

Toe walking can be associated with a number of comorbid diagnoses. Children with cerebral palsy, certain myopathies, and autism may adopt this gait pattern. However, this presentation also occurs in cases where these diagnoses have been ruled out and there are no other underlying orthopedic or neurologic causes. In such instances, an otherwise healthy child with normal muscle strength and adequate muscle length appears to simply choose this gait pattern. Frequently, such children can walk without this pattern when they choose to, but seem to prefer ambulating on their toes and balls of their feet. Idiopathic toe walking (ITW) is thus a diagnosis of exclusion. When all other possible causes have been excluded, the medical community can only shrug their collective shoulders and, uncertain of the underlying cause, attach the unsatisfying term idiopathic, a Greek word meaning "one's own suffering," or "a disease of its own kind."

Herrin and Geil summarize some of the additional observations associated with ITW, including the fact that it is only considered abnormal when it persists beyond two years of age, that it is more common in boys than girls, and that about 30 percent of patients who present with ITW have a family history of the presentation.1 While seemingly benign on the surface, ITW is aesthetically concerning to many families, decreases walking velocity, and places the child at greater fall risk. Left unaddressed, ITW can lead to future orthopedic problems, such as contracture of the Achilles tendon.1

It has been suggested that this voluntary gait deviation increases the child's sensory stimulation, increasing the support tone in the lower extremities during ambulation.1 In the absence of a clear understanding of why the deviation occurs, there is some debate about how best to manage it. Orthoses, passive stretching, serial casting, injections of botulinum toxin, and orthopedic surgery have all been advocated and explored.

Conflicting management strategies in lower-limb orthoses have recently emerged. Historically, ITW was viewed as a biomechanical problem with a biomechanical solution. A child who wanted to posture into relative equinus could be mechanically prevented from that alignment through the use of an articulated AFO with plantarflexion stops. More recently, less restrictive solutions have been attempted. Based on the premise that the causes of ITW appear to be related to sensory perceptions, less restrictive orthoses and rigid footplates have been used to alter the sensory inputs affecting the foot and ankle. This article reviews recent literature on potential treatment strategies in the management of ITW.

Herrin and Geil

A formal trial between two opposing theories about the orthotic management of ITW was recently published by Herrin and Geil.1 The authors describe the recruitment and subsequent observations of 18 children diagnosed with ITW between the ages of two and eight years old. Exclusions from this convenience sample included the presence of any neurologic condition, prior treatment of an Achilles tendon contracture (such as with serial casting or surgery), and the presence of a plantarflexion contracture.

Subjects were randomly assigned to one of two treatment modalities. The AFO group received a custom articulated AFO, fabricated from 1/8-inch polypropylene with a plantarflexion stop, which prevented plantarflexion beyond a 90 degree alignment. Tamarack ankle joints allowed unrestricted motion into dorsiflexion beyond 90 degrees. The AFOs had a full-length toeplate with medial and lateral footplate trim lines proximal to the metatarsal heads. The foot orthotic (FO) group received custom FOs, fabricated from 1/8-inch puff and 1/8-inch Poron® with a cork base and a full-length, firm, carbon fiber footplate.1 Guardians of the children in both cohorts were asked to ensure that the children wore the assigned devices at all times over a six-week period, except during sporting activities and sleep.

child up on balls of feet

The recruited subjects presented with an average age of five years. Eleven of the 18 subjects had begun toe walking at the same time they began walking. For the remaining subjects, ITW was acquired later, such that the average age the subjects began walking was 12 months and the average age of toe walking was 18 months. The average range of passive dorsiflexion was about 4 degrees and the parents of all 18 subjects confirmed that their children could walk normally if asked to do so.

Gait Lab Data

This last statement, one that resonates with anyone who works in the space of pediatric gait training and rehabilitation, appears to have undercut the authors' attempts to objectively quantify the presence and extent of toe walking associated with the two interventions in the setting of a formal gait laboratory. Even though all 18 subjects had been diagnosed with ITW, when taken through the formal processes and environment of the gait lab, fully three-quarters of their collective baseline steps were characterized by a normal initial contact with the heel. Given the young age of the participants, the reality that most of them had likely been told repeatedly by their parents to walk correctly or stop walking on their toes in the months and years leading up to the study, and the alien formalities of a gait lab, it is hardly surprising that-for the most part-these children who could walk normally when asked to do so, largely did.

Upon their return to the gait lab six weeks later, the subjects were recorded a second and third time, both with and without their respective interventions. For the AFO contingent, where the brace mechanically prevented them from gaining initial contact with any part of their foot other than the heel, 100 percent of their steps were characterized with a heel-first initial contact. When the devices were removed, 91 percent of their steps retained this characteristic, with 6 percent of their collective steps initiating contact with the ground on a flat foot and only 2 percent initiating contact with the toes.1

For the contingent with the foot orthoses and carbon footplates, even though their interventions failed to mechanically prevent initiating contact with the toes, a reasonable effect was observed. The percentage of steps in which ground contact was initiated with the heel increased from 73 to 87 percent. This corrective element in their gait was largely preserved when the devices were removed, with 88 percent of the subsequent steps demonstrating initial contact with the heel.

In summary, while the in-device correction associated with the carbon footplates was not as consistent as that observed with the AFOs, upon their removal, the proportion of correct steps was nearly identical for both devices. Importantly, the trial period of six weeks was chosen in deference to the timelines established by the funding source on the project, the Orthotic and Prosthetic Education and Research Foundation, and was otherwise arbitrary. That is to say, there was no additional rationale for the six-week time period. Whether the efficacy of the devices might have been influenced by a longer utilization period is a matter of conjecture. Similarly, the follow-up observations, in which the children wore their shoes without their assigned devices, only provides insight in the immediate carryover effects of the interventions. The duration of the observed carryover effects is also left as a matter of conjecture.

Subjective Observations

Given the precarious nature of the gait lab observations with children capable of self-correcting their gait deviations, the subjective observations of their parents provide a second set of informative data. While the authors report no statistically significant differences in parental satisfaction with their children's gait or the efficacy of the interventions, the numbers are nevertheless worth consideration as their observations extend beyond their children's performance in the gait lab.

In viewing these numbers, the timing of their collection is relevant. Subjective feedback was solicited at the conclusion of the six-week trial period during which the subjects averaged ten and 11 hours per day of use with the AFOs and FOs, respectively. Thus, parental satisfaction with their children's gait and parental opinions on the efficacy of the treatments reflects their observations over the study period and perhaps immediately following when carryover effects remained evident.

For the AFO condition, 63 percent of the parents reported improved satisfaction with their children's gait following treatment, with an additional 25 percent reporting that their satisfaction had declined. However, fully half of the parents of children who wore AFOs described the treatment as ineffective, with an additional 13 percent describing the treatment as very ineffective.1 One possible interpretation of these observations is that these children walked quite well while wearing their AFOs but the long-term efficacy of the intervention was questioned by the majority of the parents.

child's lower legs and feet

For the FO condition, opposite trends were observed. The parents of half of these children reported that their satisfaction with their children's gait was unchanged following treatment, with only 38 percent reporting an improvement in this regard. However, a slight majority of the parents, 56 percent, described the treatment as effective, compared to 44 percent who described it as either ineffective or highly ineffective.1

One potential explanation might be found in the fact that parents found donning the FO to be easier than donning the AFO, and they felt the FO was more aesthetically appealing and less likely to cause skin breakdown than the more rigid, cumbersome AFO. It is possible that the effectiveness of each device was weighed against its relative size, restrictiveness, and associated impositions and that the FOs were more likely to be considered effective because of comparatively lowered expectations.

Back to the Lab

Some additional insights and considerations can be found in elements of the remaining gait lab data. For those subjects who were managed with AFOs, gait velocity was unquestionably quicker, both with the intervention and immediately after its removal. At the conclusion of the trial, the AFO contingent walked, on average, 11 percent faster than the FO contingent (1.36m/s and 1.22m/s, respectively).1 This benefit was obtained through a combination of increases in both cadence and step length and may help explain the trend toward improved parental satisfaction with their children's gait.

For the FO contingent, additional benefits were observed, but they require some explanation. In addition to an objective determination of how the children obtained initial contact for each of their recorded steps, the gait lab data allowed for objective measures of the timing of heel rise. This measurement captured an aspect of toe walking that appeared less prone to the compensatory corrections of the children.

Prior to the interventions, heel rise occurred prematurely for all subjects, occurring on average at about 20 percent of the way through the gait cycle. The mechanical correction of the AFO restrained this event until an average of 35 percent of the way through the gait cycle, but that correction did not last. Upon the removal of the device, the average timing of heel rise reduced to about 25 percent of the way through the gait cycle.

For the FO group, the effect of the intervention on this gait parameter was more measured, only delaying heel rise until roughly 26 percent of the way through the gait cycle. However, upon removal of the FOs, this improvement largely persisted, occurring about 25 percent of the way into the gait cycle. Ultimately, upon the removal of the interventions, the net effects of both interventions were about the same, delaying heel rise from 20 to 25 percent into the gait cycle.1 The difference is that there was a noticeable reversion of this parameter when the AFOs were removed, while it stayed about the same for the group with FOs. This is consistent with subjective parental feedback in the FO group, trending toward no change in parental satisfaction with their children's gait, but a general sentiment that the treatment was effective.


True to the nature of a pilot study, this effort may have raised more questions than it answered. However, some generalities can be drawn. By virtue of their mechanical restrictions, AFOs with plantarflexion stops can produce noticeable improvements in the gait of children with ITW-gait velocity increases, initial contact occurs consistently at the heel, and heel rise occurs later in the gait cycle. However, with the possible exception of velocity, these benefits appear to be temporary, with some level of reversion to pre-intervention gait patterns. By contrast, the altered sensory input associated with the FOs does not produce the immediate changes observed with the AFOs. However, the reduced benefits of the FOs, including increasing the likelihood of initial contact at the heel and delaying the timing of heel rise, appear to be better retained as carryover effects of the intervention. While the preferred means of managing ITW with orthoses requires more research, it appears that alterations to both gait biomechanics and sensory inputs may play a meaningful role.

Phil Stevens, MEd, CPO, FAAOP, is in clinical practice with Hanger Clinic, Salt Lake City. He can be reached at .


  1. Herrin, K., and M. Geil. 2016. A comparison of orthoses in the treatment of idiopathic toe walking: A randomized controlled trial. Prosthetics and Orthotics International 40 (2):262-9.