Objective Calcaneal apophysitis, is a relatively common cause of heel pain in children. Very few randomised studies have evaluated treatment options. This trial compared the effectiveness of currently employed treatment options for the relief of pain and disability associated with calcaneal apophysitis.
Design Factorial 2×2 randomised comparative effectiveness trial with 1, 2, 6 and 12-month follow-up.
Setting Participants were recruited from the caseload of podiatrists at Monash health and Peninsula Health.
Participants Children aged 8–14 years with clinically diagnosed calcaneal apophysitis.
Interventions Treatment factor 1: two different types of in-shoe orthoses: a heel raise or prefabricated orthoses. Treatment factor 2: footwear replacement or no footwear replacement.
Outcomes Our primary outcome was functional disability, the secondary outcomes were pain and ankle dorsiflexion range.
Results A total of 133 children and their parents responded to the recruitment advertisement, 124 participated in the trial.
At the 1 and 2-month follow-up points, there was a main effect of the shoe insert (heel raise) in only the physical domain for the Oxford ankle foot questionnaire (p=0.04). At the 6 and 12-month follow-up points, there was no main effect or interaction effect for any outcome measure.
Conclusion This trial indicates at the 2-month time point there is a relative advantage in the use of heel raises over prefabricated orthoses for the treatment for calcaneal apophysitis. At 12 months there was no relative advantage to any one of the investigated treatment choices over another. Therefore, if a physical impact is experienced for greater than 2 months, the selection of treatment choice may defer to clinical judgement, cost-minimisation and or patient preference.
Trial registration number ACTRN12609000696291.
Statistics from Altmetric.com
Calcaneal apophysitis, also known as Sever's disease is a condition that causes pain at the posterior aspect of the heel. It may present in children between the ages of 8 and 15 years,1 ,2 during the years of skeletal development where the apophysis of the heel is open.3 ,4 The child often reports the pain to be worse during and postactivity, and the severity of the pain may cause a limp or avoidance of heel contact during gait on the affected limb(s).5–7 The diagnosis of calcaneal apophysitis is not by imaging but rather by clinical assessment via the ‘squeeze test’ (medial and lateral compression of the heel)8 and by the exclusion of other potential differential causes through thorough history and/or clinically indicated diagnostic tests.9 Even though calcaneal apophysitis is a self-limiting condition, previous research has shown that this condition adversely affects the health-related quality of life of symptomatic children.10
There are several reported theories addressing the cause of calcaneal apophysitis.6 ,11 ,12 These differing theories concur that trauma takes place to the apophyseal tissues through movement of apophysis relative to the diaphysis through either passive tension of the insertion of the gastroc/soleus complex or through compression or impact forces. Treatment modalities therefore have been implemented to address either of these with varied degrees of success.13
Heel raises are a commonly treatment modality.14 This treatment modality may address the trauma caused by the passive tension of the gastrocnemius/soleus complex, caused by relative lengthening of the tibia during rapid growth.6 ,12 Physical activity modification or restriction is commonly used to treat calcaneal apophysitis.11 It is thought that overuse trauma is caused by active contraction of the gastrocnemius/soleus complex and may be influenced by abnormal foot posture. Some authors within the literature have used this theory to argue that having a pronated foot posture increases the generated forces required by the gastrocnemius/soleus complex during activity,15 ,16 thus accentuating the force being transferred through the apophysis. This theory explains why some clinicians prescribe orthoses. Lastly, clinicians often recommend footwear modification or replacement. It is believed that the supportive or cushioned footwear decreases the trauma at the apophysis by absorbing the ground reaction force at during walking, running and jumping impact.17
Despite these theories, there is further uncertainty surrounding the causative mechanisms of calcaneal apophysitis. High levels of athletic activity18 ,19 and obesity (increased weight or body mass index) have both been identified as risk factors,20 which would support theories related to active contraction of the gastrocnemius/soleus complex and that of higher ground reaction forces.20 One cross-sectional observational study demonstrated no significant difference in the amount of foot pronation between children with and without calcaneal apophysitis. However, the study did identify a difference in the amount of ankle dorsiflexion range of motion of the symptomatic foot in children with calcaneal apophysitis, therefore supporting the passive length theory.21
Previous research has reported the effect of treatment modalities over a short period of time (4–12 weeks). Heel raises address the length and balance theory of the tibia and Achilles tendon, and has been reported to provide some initial pain relief.22 Similarly, use of shock absorbing material in footwear may be an appropriate treatment option to address the impact force. Another commonly reported treatment modality is the use arch supporting devices to address the biomechanical theory of the pronated foot type. The mechanism of action for pain relief may be as a result of shock absorption and the device properties distributing ground reaction forces from the heel to the medial longitudinal arch of the foot. Similar devices have been examined in a prospective cross-over study resulting in some pain relief during the trial period.14 There are no definitive results about superiority of any of these treatments leading to a limited understanding of the mechanism or development of calcaneal apophysitis.
This trial aimed to compare the effectiveness of common treatment modalities that address some of the suspected causative mechanisms of calcaneal apophysitis.
This trial was a factorial randomised comparative effectiveness trial; two factors (shoe insert and footwear) each with two levels (heel raise/prefabricated orthoses and current footwear/new athletic footwear, respectively). The study protocol has previously been published.23 In addition to the planned short-term evaluation as per the published protocol with a follow-up of 1 and 2 months, additional funding was received to perform further follow-up evaluations at 6 and 12 months. A CONSORT flow chart for the design of this trial is presented (figure 1). There was no control group as the principal recruitment population where those children presenting to a health service for treatment of this condition. The investigators felt that the feasibility of participant recruitment would be impacted to the extent that the trial would no longer be possible if a control group were part of the design. Data were collected during this trial to also examine this assumption.
Participants and setting
Children aged between 8 and 14 years were recruited from the caseload of podiatrists at Cardinia Casey Community Health Service (Monash Health) and Peninsula Health—Community Health. In addition, general advertisements were sent to local sporting clubs, primary schools, general practitioners and private allied health clinicians. The advertisement encouraged the parents of children reporting heel pain in this age group, to contact the local health service for eligibility screening and potential to participation in a research project about paediatric heel pain. Children were eligible to participate if they provided a subjective report of pain located at the calcaneal apophysis (ie, posterior aspect of heel) and with pain on palpation (positive calcaneal squeeze medial and lateral borders).8 Children were excluded from the trial if they had been diagnosed with a fracture or tumour of the foot/leg, a major lower limb orthopaedic condition and/or a previous diagnosis of infective, reactive or rheumatoid arthritis in the past 12 months.
The trial was conducted at Monash Health and Peninsula Health Services. These health services are in a geographical location with lower than average socioeconomic status compared to the rest of Australia. These areas have a documented lower than average rate of participation in sporting activities24 and higher rates of obesity.24 ,25
All the children participating in the trial received a standardised icing and stretching programme. The children were asked to ice the area of pain once a day for a total of 10 min during the initial stage of care (1 month). It was also recommended to continue the icing regime after sporting activities until the child was pain free. The stretching programme was initiated for all children at the 1 month follow-up appointment, after the acute phase of pain sufficiently subsided. The stretching programme consisted of the static weight-bearing gastrocnemius and soleus stretch with the lower limb in the position as demonstrated in figure 2A, B.26
The two levels of shoe orthoses that were investigated were:
Factor 1: Foot orthoses
The two chosen in-shoe interventions represent widely available treatments. It is not common for these interventions to be concurrently provided. These approaches are low in cost compared to other interventions such as semicustomised or customised foot orthoses. Heel raises (figure 3, 6 mm heel raise) were made from high-density (PE 180 kg/m3) ethylene vinyl acetate (EVA). The EVA heel raise aimed to reduce the available range of motion and decrease the activity of the gastrocnemius soleus-Achilles tendon complex on the calcaneo-Achilles attachment by elevating the calcaneus. Heel raises have been found to provide therapeutic relief in tendo Achilles bursitis, tenosynovitis of Achilles tendons, and postoperative management of ruptured Achilles tendon.27 The EVA material of the heel raise also provides shock absorption at heel contact. The prefabricated orthoses (figure 4, Prothotic W The Orthotic Laboratory) intervention was a polyurethane device that aimed to limit the pronation by inverting the rearfoot with a medial varus wedge combined with a small notch in the cuboid area.28 The cuboid notch is a small raised area at the cuboid area that is thought to provide a pronatory force to the mid-foot and prevent lateral slippage off the orthotic. The authors anticipated that the use of a medial varus wedging device was contraindicated with a Foot Posture Index -6 (FPI-6) equal or less than -1.23 If a child presented with an FPI equal or less than -1 then the child was excluded from the trial and offered alternative treatment through the health services. The orthoses were also covered in a 3 mm blown multidensity EVA cover (Multiform) which provided additional shock absorption. The children were provided with wearing instruction at fitting of the orthotic intervention.
Factor 2: Footwear
The two levels of footwear to be investigated were:
Current footwear worn by participant.
New athletic footwear provided by trial.
The first condition in this factor entailed no direction for modification of current footwear. The children were requested to continue wearing their most commonly worn footwear. This may have been their school shoes, runners or fashion footwear.
The alternate condition was provision of new athletic footwear. This involved the supply of a shoe with a firm heel counter, dual density EVA midsole and rearfoot control provided by Adidas Australia. All athletic footwear provided were the same model (Supernova). The footwear replacement intervention was provided to the child at no cost.
All children who received the new footwear were given standardised shoe wearing instructions. Schools within the trial area generally allowed the students to wear athletic footwear as the chosen footwear, if not, the treating podiatrist supplied a letter of support for the footwear and liaised with the school if required.
The primary outcome measure for the trial was the Oxford ankle foot questionnaire.29 This scale measured the disability associated with calcaneal apophysitis from the perspective of the child and the parents and contains ‘physical’ (6 items, Cronbach's α=0.92, parent-child intraclass correlation coefficient (ICC)=0.72), ‘school and play’ (4 items, Cronbach's α=0.89, parent-child ICC=0.73) and ‘emotional’ (4 items, Cronbach's α=0.86, parent- child ICC=0.72) and footwear (single item) domain areas.29
The secondary outcome measurements were the Faces pain scale30 and the weightbearing lunge test.31 The Faces pain scale is a seven-point verbal rating scale that was used to measure severity of pain in the morning, overall pain, pain while wearing the intervention and pain when not wearing the intervention. The test-retest reliability data for the Faces pain scale have previously been examined among 6-year-old children yielding a rank correlation coefficient of 0.79.30 The Faces pain scale is a commonly used assessment tool within the Australian health service, and the participating investigators were familiar with the tool.
The weightbearing lunge tests31 ,32 were used as the clinical measure of the available ankle range of motion. The test was conducted with a leg straight then knee bent position and angle determined with a digital inclinometer. These measures were collected to determine any change in ankle dorsiflexion during the face-to-face assessment period (baseline, 1 month and 2-month). The intra-rater reliability of experienced raters conducting this test has been shown to be high when using a digital inclinometer (average ICC=0.88, average 95% limits of agreement=–6.6° to 4.8°). The intra-rater reliability of an inexperienced rater has also been demonstrated to be good to high when using a inclinometer (ICC=0.77, 95% limits of agreement=–9.1° to 8.3°).31 The inter-rater reliability has been shown to range ICC=0.80 and 0.86 between novice and experienced raters.32 This measure has also been commonly used with children.33 ,34
Data were also collected from the participants including age, gender, height and weight, FPI-6.35 The Child Leisure and Activity Survey Study (CLASS) was utilised to calculate a snap shot (1 week duration) of the participants activity levels from the previous week.36 Participant compliance was recorded by completion of a star chart by the child (with parent observation) within their school diary that logged daily compliance with the allocated shoe insert intervention and completion of ice application and stretching. Parents of the participating children were also asked at the initial assessment prior to randomisation, but following collection of baseline data whether they would have been willing to continue to participate in the trial if their child hypothetically were to be allocated to an inactive control group. A yes or no response was recorded.
All children presenting to the participating health services with heel pain were screened for trial eligibility by the treating podiatrist. The parents and participants who met the trial inclusion criteria were provided with a written and verbal explanation of the trial and parents were asked to provide written consent for their child to participate. Following consent, baseline assessments were completed prior to randomisation.
Randomisation was undertaken using a permuted-block randomisation approach stratified by site. Randomisation blocks of four or eight participants were generated and randomly selected. The assessor not involved in recruitment or assessment, completed the randomisation and provided each site with opaque, sealed envelopes. The treatment conditions in the envelopes were provided as per the random allocation sequence following completion of the baseline assessment. The four treating podiatrists had the group allocation concealed from them to eliminate allocation bias.
Primary and secondary outcome measurements were undertaken at initial presentation and at 1 and 2-month follow-up. At the 6 and 12-month follow-up only the primary outcome measure was undertaken. The initial, 1 and 2-month assessments took place face to face, the 6 and 12-month follow-up consultations were completed by telephone by a clinician who was blinded to group allocation. In the case of non-attendance, the Oxford ankle foot questionnaire was either posted to the child and parent or a phone consultation was provided to complete the questionnaire.
It was considered that a minimum clinically important change in the Oxford ankle foot questionnaire in any domain was seven points and that based on previous work, the maximum SD in any domain was six points.29 Given this, and that the trial had one preintervention measure and four postintervention measures, it was deemed that a sample size of n=27 per factorial trial cell (ie, total trial n=108) would have >90% power to detect a significant difference of seven points in any simple contrasts undertaken, assuming a correlation between assessment points within individual participants is r=0.7. To account for a possible 15% dropout and incomplete assessments, a total of 124 participants were recruited.
Participant data were initially explored with means (SD) or frequency (%). All outcome measures were collated and the median, (IQR) or means (SD) calculated. The intervention factors (heel raise and new shoes, orthoses and new shoes, heel raises and own shoes, orthoses and own shoes) were examined over time (baseline 1 and 2 month time period) on the primary and secondary outcome measures using a generalised estimating equation (GEE). Only the primary outcome measure was examined at the 6 and 12-month period using GEE analysis. This approach was suitable for the analysis of the longitudinal data as it has been shown to produce unbiased effect estimates with appropriate precision in the presence of missing data without the need for data imputation techniques and does not involve list wise deletion of participant data where missing data are present.37 The primary outcome data were categorical and not checked for distribution prior to analysis due to the robust nature of GEE analysis. We examined main effects for each intervention factor separately and then a combined interaction effect between the intervention factors. The intervention factor and the combined interaction effect were considered significant at the level of p<0.05.
A total of 133 children and their parents responded to the recruitment advertisement or were recruited from the participating clinics. Nine children were excluded due to a suspected diagnosis of juvenile idiopathic arthritis (n=1), FPI–6 score >–1 (n=4), resolved pain prior to initial assessment (n=3) and a participant making up pain to avoid school activity (n=1). All remaining 124 participants (52 girls, 42%, 72 boys, 58%) assented to participate in the trial and parents consented. Participant characteristics at baseline, broken down by group, are presented (table 1). There were 115 of the 124 (93%) participants (children) who indicated that they would not continue with this trial if they were allocated to a control group with no treatment.
Activity and leisure time (min/week) was collected for 88 participants. The total mean (SD) physical activity minutes per week were 669 (460) min/week and leisure activity minutes per week were 1443 (1089) min/week. There were no differences between the groups in the amount of time spent participating in physical activity (p>0.05) or leisure (p>0.05).
Participant compliance with the intervention, ancillary treatments and adverse events are presented in table 2. Adherence to the use of the allocated interventions appeared to decrease over the 12-month follow-up period of this trial.
At the 1 and 2-month time point there was a main effect of shoe insert (heel raises vs prefabricated orthoses) for the child in the physical domain of the Oxford ankle foot questionnaire (p=0.04). There was neither main effect of footwear nor a shoe insert by footwear interaction, for the primary or secondary outcome measures (table 3).
At the 6 and 12-month time point only the primary outcome measure was examined, there was no main effect of shoe insert (heel raise vs prefabricated orthoses), or footwear (usual footwear vs athletic footwear) nor an insert by footwear interaction effect (table 4) with the exception of the footwear subscale of the Oxford ankle foot questionnaire. There was a difference in the footwear domain between footwear intervention groups (usual footwear vs athletic footwear), however, the direction of this difference was opposite for children (p=0.02, favouring usual footwear) than for their parent (p=0.01, favouring athletic footwear). This single item domain asks participants to rate the extent to which the foot condition affected their ability to wear their preferred footwear.
Summative data for the primary and secondary outcomes across the follow-up time points are presented in the supplementary tables provided (see online supplementary tables S1–S3). The improvement over time for the oxford ankle foot questionnaire domains (child and parent) across all four-treatment groups is shown within figures 5 and 6.
This trial identified a benefit in heel raises over prefabricated orthotics at the 1 and 2-month point (table 3) as measured by the physical and school and play domains of the Oxford ankle foot questionnaire. At the 6 and 12-month point no relative advantage of any of the treatment approaches was identified. The only significant finding at the end point of the trial was for the footwear domain of the Oxford ankle foot questionnaire. These results were inconsistent in direction between child and parent.
There are two broad implications of these findings. First, for clinicians, this means they cannot be certain that one of these interventions can adequately be used to treat all their patients with calcaneal apophysitis. Similarly to previous studies22 we have demonstrated that the use of heel raises at the initial stages of therapy may aid children with symptoms to remain physically active compared to the orthotic; however, there were similar results at the 12-month point for heel raises and orthotics. Therefore, the selection of treatment choice may defer to clinical judgement, cost minimisation or patient preference.
Second, in terms of understanding the aetiology of this condition, this trial indicates that no one single mechanism is likely to explain the causation of the condition in all cases. It may be that the different mechanistic theories are true in some cases but not others, meaning individualised treatments would be required. Appropriate initial and review assessment allows clinicians to identify and monitor concerns such as ankle range of motion, abnormal foot posture and footwear deterioration or style therefore allowing customisation of treatment.
Strengths and limitations
An important strength of this trial was the additional follow-up periods of 6 and 12 months. This was added to the original trial protocol23 after additional funding was received. Previously, the longest follow-up period for any study of a single intervention approach for calcaneal apophysitis was 12 weeks.22 This is the first trial to follow children with calcaneal apophysitis for 12 months. All other studies investigating calcaneal apophysitis report pain resolution follow treatment within 6 weeks to 3 months. This trial provides evidence that calcaneal apophysitis may not fully resolve postintervention and may be intermittent in nature. The finding of children having pain up to, 10 months prior to inclusion and continued physical impact at the 12 months time point supports this.
An additional strength of this trial is the sampling/recruitment. Previous studies have focused their sampling on sports clinics or academies, which targeted particular subgroups of children. Previously published baseline data for this group of participants did not identify level of activity associated with pain.38 This current trial recruited from a large geographical area with a diverse range of participants perhaps highlighting that calcaneal apophysitis present in very active children and less active (non-sporty) children.
A limitation in the capacity to extrapolate our trial findings is the inability to understand the mechanisms underpinning the treatment effect. While the results indicate that no treatment modality was more effective than the other over the 12-month period, we cannot say that they were ineffective as no control group (inactive group) was included in the design. The limitations to the chosen treatment modalities included:
Insufficient height of the heel raise or a decrease in height from to compression of the heel raise due to body weight over the 12-month period.
Change in shock absorption of size of footwear over the 12 months. Replacement footwear was not issued if the midsole compressed, the child outgrew or misplaced the footwear.
The prefabricated orthoses included were different to previously researched devices, as the Prothotic did not have a heel cup. Orthoses with a heel cup have previously been found to increase the heel pad thickness by centring the fat pad. This additional heel pad thickness subsequently reduces the heel peak pressures experienced by children with calcaneal apophysitis and providing pain relief.39
A limitation within the study is that inter-rater reliability was not established for the research team. The research team-met prior to the trial and all routinely used the weightbearing lunge test and FPI-6 measures. Another confounding factor is the provision of advice in using ice and stretching for all children participating in the initial stages. It is possible that these strategies may have had a beneficial, null, or harmful effect on the outcomes or may have had an interaction effect with some of the interventions examined.
Within the data analysis process it was noted that the 6 and 12-month data were not normally distributed and that data transformation was not undertaken. This statistical limitation is likely due to the large number of participants who had limited quality of life impact. In future longitudinal studies this limitation could be avoided by completing a survival analysis model technique and examining the time frame until symptom resolution.
Lastly, there were also a high number of males within the second (athletic footwear and heel raise) group, which may have contributed to the results. While it is unknown if there is a gender bias to reporting the pain related to calcaneal apophysitis, future studies should consider methods for group allocation to minimise this disparity.
Further research is required to examine the effectiveness of treatment for this condition. Use of control conditions should ideally be used, however, our trial has identified considerable practical concerns with this approach. Other interventions should also be considered including the use of ice or activity modification. Other studies have reported some success in pain management with prefabricated or custom-made orthoses, therefore these may also be considered.5 ,14 ,18 The large drop off in intervention use by the children, may be indicative of the natural intermittent nature of calcaneal apophysitis. Future research may consider more extensive long-term monitoring of the intervention that will provide important information about the natural history of the disease.
This trial indicates at the 2-month time point there was a relative advantage in the use of heel raises over prefabricated orthoses for the treatment for calcaneal apophysitis. At 12 months there was no relative advantage to any one of the investigated treatment choices over another in relation to the physical impact of calcaneal apophysitis. Therefore, if children are experiencing difficulties in engaging in physical activities as a result of calcaneal apophysitis, the selection of treatment choice may defer to clinical judgement, cost minimisation and or patient preference.
What are the findings?
This trial determined that there was an advantage in using heel raises during the first two months of treatment of calcaneal apophysis. There was no relative advantage to any one of the investigated treatment choices over another, over a 12-month period of time.
How might it impact on clinical practice in the future?
Selection of treatment choice should defer to a decision of cost-minimisation, patient preference and/or clinical assessment. Heel raises are a less expensive and proven beneficial treatment option in the first 2 months of treatment.
The authors thank Michelle Luscombe and Reshele Hunter assisted in data collection at Peninsula Health Service.
Correction notice This paper has been amended since it published Online First. Figure legends 5 and 6 have been corrected.
Contributors AMJ contributed to the trial design, recruitment, data collection, completed the data analysis, prepared the manuscript and reviewed the final manuscript. CMW contributed to the trial design, participated in data collection, assisted with data analysis, contributed to manuscript preparation/discussion and reviewed the final manuscript. TPH contributed to the trial design, assisted with data analysis, contributed to manuscript preparation/discussion and reviewed the final manuscript. All authors, external and internal, had full access to all of the data (including statistical reports and tables) in the trial and take responsibility for the integrity of the data and the accuracy of the data analysis.
Funding Monash Health Emerging Research Grant, Adidas Australia provided the footwear for the trial at no cost to the patient or trial and The Orthotic Laboratory provided the prefabricated orthoses at no cost to the patient or trial. Neither Monash Health Emerging Researcher grant committee, Adidas Australia or The Orthotics Laboratory had access or contributed to the research design, data collection, analysis or interpretation, in the writing of the report, and in the decision to submit the article for publication.
Competing interests AMJ, CMW and TPH have support from Peninsula Health, Monash Health and Monash University for the submitted work; AMJ, CMW and TPH have no relationships with Adidas Australia that might have an interest in the submitted work in the previous 3 years.
Patient consent Obtained.
Ethics approval Monash Health: HREC Ref: 09271B. Ethics Approval Peninsula Health Service: HREC/09/PH/65.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Full data set is available from the corresponding author. Consent was not obtained but the presented data are anonymised and risk of identification is extremely low.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.