Effectiveness of reducing tendon compression in the rehabilitation of insertional Achilles tendinopathy: a randomised clinical trial

Abstract

Objective

To assess the effectiveness of low tendon compression rehabilitation (LTCR) versus high tendon compression rehabilitation (HTCR) for treating patients with insertional Achilles tendinopathy.

Methods

In an investigator-blinded, stratified randomised trial, 42 sport-active patients (30 males and 12 females; age 45.8±8.2 years) with chronic (> 3 months) insertional Achilles tendinopathy were allocated in a 1:1 ratio to receive LTCR or HTCR. Both rehabilitation protocols consisted of a progressive 4-stage tendon-loading programme, including isometric, isotonic, energy-storage and release and sport-specific exercises. The LTCR programme was designed to control Achilles tendon compression by limiting ankle dorsiflexion during exercise, eliminating calf stretching and incorporating heel lifts. The primary outcome was the Victorian Institute of Sports Assessment-Achilles (VISA-A) score at 12 and 24 weeks, which measures tendon pain and function and was analysed on an intention-to-treat basis using a linear mixed model. Significance was accepted when p<0.05.

Results

20 patients were randomised to the LTCR group and 22 to the HTCR group. Improvement in VISA-A score was significantly greater for LTCR compared with HTCR after 12 weeks (LTCR=24.4; HTCR=12.2; mean between-group difference=12.9 (95% CI: 6.2 to 19.6); p<0.001) and after 24 weeks (LTCR=29.0; HTCR=19.3; mean between-group difference=10.4 (95% CI: 3.7 to 17.1); p<0.001). These differences exceeded the minimal clinically important difference of 10.

Conclusions

In sport-active patients with insertional Achilles tendinopathy, LTCR was more effective than HTCR in improving tendon pain and function at 12 and 24 weeks. Consequently, LTCR should be considered in the treatment of insertional Achilles tendinopathy.

Trial registration number

ClinicalTrials.gov (ID: NCT05456620).

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Compared with midportion Achilles tendinopathy, exercise therapy for insertional Achilles tendinopathy is less effective, resulting in about 50% of patients proceeding with surgery. Since compressive overload might play a role in its pathogenesis, it has been proposed that reducing tendon compression during rehabilitation could yield better results.

WHAT THIS STUDY ADDS

• This randomised clinical trial demonstrates that reducing tendon compression (eg, limiting ankle dorsiflexion) during rehabilitation for insertional Achilles tendinopathy leads to improved clinical outcomes and higher patient satisfaction.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• In sport-active patients with insertional Achilles tendinopathy, treatment should include rehabilitation that limits Achilles tendon compression through patient education, heel lifts and progressive exercise therapy that controls the amount of ankle dorsiflexion.

Introduction

Achilles tendinopathy poses a substantial challenge for athletes, affecting up to 52% of elite runners throughout their careers.¹ It is characterised by localised pain, tendon thickening and impaired performance, leading to a loss of training time and, in some cases, the premature end of their sporting careers.² Achilles tendinopathy is categorised according to the pathological anatomical location into midportion Achilles tendinopathy, affecting the main body of the tendon, and insertional Achilles tendinopathy, affecting the tendon’s attachment to the calcaneus.³ Approximately one-third of cases involve insertional Achilles tendinopathy.⁴ This classification has important clinical implications, as insertional Achilles tendinopathy is generally more challenging to treat with exercise therapy than midportion Achilles tendinopathy.⁵,⁹ One hypothesis for the low success rate of exercise therapy in patients with insertional Achilles tendinopathy is that the Achilles tendon insertion is exposed to both high amounts of tensile and compressive loads during the end-range ankle dorsiflexion.¹⁰,¹³ Achilles tendon compression occurs during ankle dorsiflexion, as the insertion wraps around the superior border of the calcaneus, leading to substantial compressive load in this region.¹⁰ Moreover, since insertional Achilles tendinopathy is characterised by tendon thickening, the affected Achilles tendon experiences more compression during ankle dorsiflexion for the same amount of loading, potentially perpetuating pathology.^{14 15} Therefore, a commonly cited review advised reducing compressive loads during rehabilitation while applying enough tensile load to strengthen and stimulate the Achilles tendon.¹⁶ In this perspective, three recommendations to reduce tendon compression in patients with insertional Achilles tendinopathy are proposed. First, it is suggested to minimise calf muscle stretching and use calf muscle massage techniques to address muscle stiffness instead. Second, it is advised to limit the amount of dorsiflexion during tendon loading exercises. Third, heel lifts are recommended to reduce ankle dorsiflexion during daily and sports activities. Although there is a rationale for reducing tendon compression during rehabilitation for patients with insertional Achilles tendinopathy, no robust evidence currently supports this advice. Therefore, this study aimed to investigate the effectiveness of low tendon compression rehabilitation (LTCR) versus high tendon compression rehabilitation (HTCR) on clinical outcomes after 12 and 24 weeks in patients with insertional Achilles tendinopathy. We hypothesised that LTCR would yield better clinical outcomes than HTCR in sport-active patients with insertional Achilles tendinopathy. With the Victorian Institute of Sports Assessment–Achilles (VISA-A, scale 0–100 points) score as the primary clinical outcome, we hypothesised that LTCR would result in higher scores.

Methods

Study design

The GhenTendon study was a parallel-group, investigator-blinded, stratified, block-randomised clinical trial (RCT) that included recreational and competitive sport-active patients with insertional Achilles tendinopathy who were not informed of the study hypothesis. The trial was conducted at the Department of Sports Medicine at Ghent University Hospital. The study protocol was registered on ClinicalTrials.gov (ID: NCT05456620) before the inclusion of the first participant (online supplemental study protocol). All participants gave written informed consent prior to participation. Findings are reported according to the Consolidated Standards of Reporting Trials (CONSORT) guidelines and fulfil the requirements of the Template for Intervention Description and Replication checklist, the Consensus on Exercise Reporting Template checklist and the Checklist for Statistical Assessment of Medical Papers.¹⁷,²⁰

Patient and public involvement

Patients and public representatives were not involved in designing or conducting the study protocol. The study was announced through an online advertisement to healthcare providers and the public.

Equity, diversity and inclusion statement

Participant inclusion was based on preset criteria and all patients who met the criteria were eligible to participate regardless of sex, race/ethnicity or socioeconomic status. We included both males and females in our research and explored the impact of sex on our primary outcomes. We recognise that our study excluded individuals with physical disabilities that prevented them from engaging in exercise therapy. Furthermore, the author group of this study consisted of members of different sexes and included junior, mid-career and senior researchers from various disciplines, such as sports medicine, orthopaedics and rehabilitation sciences.

Patients

Between November 2022 and December 2023, athletic participants were recruited through advertisements placed on hospital/university facilities and social media platforms. Participants were eligible for inclusion if they met the following inclusion criteria: age 18–60 years old, diagnosis of chronic insertional Achilles tendinopathy, with symptoms for more than 3 months but less than 3 years, a VISA-A score<80 points, performing running-based sports at least 2 times a week and being able to comply with an exercise programme.^{21 22} Diagnosis of insertional Achilles tendinopathy was confirmed if: (1) pain at the insertional region (ie, within 2 cm of the insertion on the calcaneus) could be reproduced on palpation and during a single leg-hop test and (2) ultrasonography showed the presence of intratendinous hypoechoic changes and/or tendon thickening and/or the presence of intratendinous power Doppler flow.³ Additionally, lateral weight-bearing radiographs were performed to detect a Haglund’s deformity and/or intratendinous calcifications. However, since focal calcifications, a Haglund’s deformity and an enlarged retrocalcaneal or subcutaneous bursa often coexist in patients with insertional Achilles tendinopathy and may also be present in asymptomatic individuals, participants were not excluded if these ultrasonographic or radiographic features were present.³²³,²⁸ Participants were excluded in case of a history of Achilles tendon rupture or surgery, coexisting midportion Achilles tendinopathy or ankle pathology (eg, arthritis), treatment with physiotherapy, local injection(s), extracorporeal shockwave therapy or orthotics in the past 3 months, presence of metabolic, endocrine or rheumatological disorders, use of fluoroquinolone therapy in the preceding 2 years and pregnancy.

Randomisation and blinding

Randomisation was conducted by using the independent, computer-generated Research Electronic Data Capture (REDCap system) software hosted at Ghent University.^{29 30} The treatment allocation for each participant was provided after confirming their eligibility and completing baseline assessments. To ensure a balanced distribution of participants across both groups, the random allocation sequence was generated with a 1:1 ratio to LTCR or HTCR using varying block sizes. Stratification was performed based on the preinjury activity level of the patients, categorised by the Cincinnati Sports Activity Scale (CSAS) as level 1 or level 2, and by tendon pain severity, determined by the VISA-A score (<50 or ≥50 points). Allocation concealment was ensured by entrusting the randomisation list to a sports physiotherapist (RC), who was not involved in conducting the follow-up measurement. The allocation sequence remained concealed until patients were enrolled and assigned to interventions. Throughout the entire data collection period, the main investigator (LP) remained blinded to the assigned treatments. Therefore, patients were instructed to discuss the exercises with their assigned physiotherapist and to consult an independent sports physician (AB) for other therapy-related questions.

Sample size

An a priori sample size calculation was conducted on the primary outcome measure, the VISA-A score, using G*Power, V.3.1.9.7 and an independent t-test. To detect a minimal clinically important difference (MCID) of 10 points in the VISA-A score with 90% power, a SD of 8.4 points, an alpha level of 0.05 and accounting for a 20% dropout rate, a minimum of 40 participants is required.⁶³¹,³⁵ This sample size is comparable to those used in other rehabilitation RCT studies on Achilles tendinopathy.^{21 22 36}

Interventions

Patients were randomly allocated to the LTCR or HTCR for 24 weeks. Both interventions included progressive exercise therapy and adjunct therapy (figure 1). The exercise therapy was carried out at home or in a local gym in addition to supervised rehabilitation sessions every week. A detailed overview of these rehabilitation protocols is available in the information brochures (online supplemental material).

Figure 1. Overview of the LTCR and HTCR treatments, including progressive exercise and adjunct therapy. An example exercise is depicted for each stage. The entire exercise programme is available in the online supplemental material. HTCR, high tendon compression rehabilitation; LTCR, low tendon compression rehabilitation.

The progressive exercise therapy of both groups had a similar structure, comprising individual home-based and supervised therapy, lasting up to 12 weeks and consisting of 4 consecutive stages in which patients performed 3 isometric, isotonic, energy-storage and release or sport-specific exercises on a daily basis. Within each stage, exercises were further categorised into low, moderate and high loads based on their intensity. Stage 1 consisted of solitary isometric exercises (starting with 5 repetitions of 30 s at 50% of maximum voluntary contraction (MVC) and progressing to 5 repetitions of 60 s at 70% MVC). Stage 2 consisted of high-load isometric exercises of stage 1 every first day and new isotonic exercises every second day (starting with 4 sets of 15 repetitions at 65% of the 1 repetition maximum (1RM) and progressing to 4 sets of 6 repetitions at 85% of 1RM). Stage 3 consisted of high-load isometric exercises every first day, high-load isotonic exercises every second day and new energy-storage and release exercises every third day (starting with 3 sets of 10 repetitions at 75% 1RM and progressing to 3 sets of 6 repetitions at 85% 1RM). Stage 4 involved running and jumping exercises performed every 2–3 days, with the high-load isometric exercises of stage 1 continuing on the days these sport-specific exercises were not performed. The major difference between both rehabilitation programmes was the amount of Achilles tendon compression which was minimised in the LTCR by limiting the amount of ankle dorsiflexion. In contrast, in the HTCR, patients were instructed to perform all exercises in end-range ankle dorsiflexion. Progression to each subsequent stage was determined using individualised criteria which were assessed weekly through self-reported REDCap questionnaires. The criteria were as follows: (1) pain less than 5 on the 11-point Numerical Pain Rating Scale both during exercise therapy, 1 hour after completion, and the following morning; (2) no increase in average pain and stiffness from week to week; and (3) completion of each stage for at least 2 weeks, with high-load exercises performed for a minimum of 1 week.^{2 37} Throughout the rehabilitation, all patients received individual weekly supervision from trained physiotherapists at the Department of Sports Medicine of Ghent University Hospital, who were informed on the study rationale and protocol by the research team. Furthermore, all participants received a written manual and video database to guide the home-based therapy. Based on the progression criteria, the supervising physiotherapists determined the exercise intensity and when participants could advance to the next stage. Furthermore, participants were instructed to avoid Achilles tendon-loading sports during stages 1–3. According to the progression criteria, the minimum time required to return to sports, and thus complete the allocated rehabilitation programme, was 8 weeks. During this return to sports period, it was recommended to perform maintenance exercises from stages 1 and 2 twice per week.

The adjunct therapy in both groups included education on insertional Achilles tendinopathy from a sports physiotherapist (RC), covering the pathology, the relationship between tendon load and pain, the benefits of progressive exercise therapy to improve the load-bearing capacity of the musculotendinous unit and the importance of pain-monitoring. However, in the LTCR group, specific attention was given to the role of compression in the pathogenesis of insertional Achilles tendinopathy and the importance of reducing this load during rehabilitation. In the HTCR group, Achilles tendon compression was not mentioned. Second, the LTCR group performed daily calf muscle massages with a therapy ball to address calf muscle stiffness.^{38 39} In contrast, daily calf stretching was prescribed for the HTCR group, which also reduces calf muscle stiffness but induces high levels of Achilles tendon compression.^{10 40} Third, only patients in the LTCR group received two pairs of Clearly Adjustable 12 mm heel lifts designed to reduce tendon compression during daily activities and running. The heel lifts were provided in small, medium or large sizes based on each participant’s shoe size. During phase 4 of the programme, the height of the heel lifts was gradually reduced: initially maintained at full height for low-load running, then reduced to half height for moderate-load running, and finally removed entirely during high-load running. This same progression of height adjustment was also applied to daily activities.

Treatment credibility and expectancy

Since participants were aware of their assigned intervention, their perceptions about the treatment are relevant and might impact our outcomes of interest. Therefore, patients completed the Credibility/Expectancy Questionnaire (CEQ) after being assigned to their intervention group.⁴¹

Outcomes

Outcomes were selected according to a recent consensus study which reported nine core domains for tendinopathy research.⁴² These outcomes were assessed at baseline, 12 weeks and 24 weeks.

Primary outcome

The primary outcome was disability, measured via the VISA-A questionnaire at 12 and 24 weeks. This reliable and self-administered patient-reported outcome evaluates Achilles tendon pain, function and ability to play sports. The maximum score is 100 points, with lower scores indicating greater disability, and the most used MCID being 10 points.^{32 34}

Secondary outcomes

Secondary key outcomes included subjective patient satisfaction, return to sports rate, the Visual Analogue Scale (VAS, scale 0–100 mm) for pain during 10 unilateral hops (VAS-HOP) and daily activities over the past 7 days (VAS-ADL), the number of repetitions in the single-leg heel-raise (SLHR) test and the Achilles tendon insertion’s anteroposterior diameter which was assessed by ultrasonography in longitudinal view (online supplemental figure S2).¹⁴⁴³,⁴⁵ Patient satisfaction was categorised as excellent, good, moderate or poor.⁴⁶ The return to sports rate was categorised as a return to desired sports at the preinjury level, return to desired sports but not at the preinjury level, return to sports but not the desired sports or no return to sports.⁴⁶ Additional secondary outcomes included the Lower Extremity Functional Scale (LEFS), the EuroQol instrument (EQ-5D-5L), the Tampa Scale for Kinesiophobia (TSK-17) and Achilles tendon’ Doppler signal scored by the Modified Ohberg Scale.⁴⁷ For patients with bilateral symptoms, all outcome parameters were obtained for the most painful side which was defined through the VISA-A questionnaire at baseline. Additionally, exercise adherence, the use of discouraged cointerventions (including non-steroidal anti-inflammatory drugs, shockwave therapy, orthotics and local injections) and adverse events (such as Achilles tendon ruptures, calf muscle strains, stress fractures and ankle ligament sprains) were monitored weekly through self-reported questionnaires in REDCap. Exercise adherence was assessed by recording the number of days patients performed their prescribed exercises during the preceding week and used to calculate overall exercise adherence as the percentage of completed daily exercises out of the total prescribed exercise days.

Statistical methods

The primary analysis was performed by the principal investigator (LP) who was blinded to group allocation, using the intention-to-treat approach. The online supplemental statistical analysis plan was uploaded on ClinicalTrials.gov (ID: NCT05456620) before the completion of the study. The normality of the data was explored visually through histograms and Q-Q plots. Descriptive statistics summarised participant characteristics using mean (SD), median (IQR) or number (percentage), as appropriate. Baseline measurements of participants’ credibility and expectations, obtained using the CEQ and potentially confounding the treatment outcomes, were compared between groups using independent t-tests. Longitudinal data were analysed using linear mixed-model analyses with a random intercept model for continuous outcomes. The independent variables included treatment (LTCR, HTCR), time (baseline, 12 weeks, 24 weeks) and treatment-by-time interaction. Predetermined adjustments were made for potential covariates, including age, sex, body mass index, symptom duration, CSAS, VISA-A score and the presence of an associated Haglund’s deformity and/or enlarged retrocalcaneal or subcutaneous bursa.⁴⁸ Bonferroni corrections were applied for five post hoc multiple comparisons: three comparisons at each time point between the two treatment groups and two comparisons over time within each treatment group and the reported p values are Bonferroni-adjusted. Mean differences at 12 and 24 weeks were estimated using the treatment-by-time interaction term, with the associated 95% CI and p value. Categorical variables were analysed using Fisher’s exact test. Patient satisfaction was dichotomised into dissatisfied (moderate/poor) and satisfied (excellent/good).⁴⁹ Return to sports was dichotomised into no return to the desired sport and return to the desired sports, regardless of reaching the preinjury level.⁴⁹ Finally, differences in exercise adherence were assessed by a t-test. An additional sensitivity analysis was conducted for the primary endpoint to account for missing data due to follow-up loss, exploring the findings’ robustness. Multiple imputation was not performed because the missing outcome data was assumed not to be random but dependent on the difference in effectiveness between both rehabilitation protocols and thus related to the actual outcome. Instead, VISA-A scores from the last observation were carried forward for participants who dropped out, and missing values were imputed under worst-case and best-case scenarios (online supplemental table S1).⁵⁰ In the worst-case scenario (favouring the HTCR), missing values were replaced by the group mean+2 SD in the HTCR group and by the group mean−2 SD in the LTCR group at each time point. This approach was reversed in the best-case scenario (favouring the interventional group).⁵⁰

Results

Between October 2022 and December 2023, 122 participants contacted our research group regarding the trial and were subsequently screened. A total of 42 patients were enrolled in the trial, with 20 allocated to the LTCR group and 22 to the HTCR group. Figure 2 shows the flow of participants through the trial. Four patients (9.5%) were lost to follow-up: one in the LTCR group and three from the HTCR group. There were no significant differences in baseline characteristics between groups (table 1).

Figure 2. The CONSORT flow diagram. CONSORT, Consolidated Standards of Reporting Trials; HTCR, high tendon compression rehabilitation; IAT, insertional Achilles tendinopathy; LTCR, low tendon compression rehabilitation; MAT, midportion Achilles tendinopathy.

Table 1. Baseline characteristics of the LTCR and HTCR groups^*.

Characteristics	LTCR group (n=20)	HTCR group (n=22)
Demographics and anthropometrics
Age, mean (SD), years	41.6±10.6	42.6±11.2
Sex, male	14 (70)	16 (73)
Body mass index, mean (SD), kg/m2	24.88±2.65	24.09±2.36
CSAS, prior to the onset of insertional Achilles tendinopathy
Level I (4–7 days/week)
100	5 (25)	4 (18)
95	1 (5)	2 (9)
90	0 (0)	0 (0)
Level II (1–3 days/week)
85	5 (25)	2 (9)
80	9 (45)	14 (64)
Insertional Achilles tendinopathy characteristics
Symptom duration, median (IQR), weeks	46.0 (13.8–73.5)	40.0 (14.0–96.3)
VISA-A score, mean (SD), points	60.4±10.9	60.6±12.3
Affected side
Unilateral, left/right, n (%)	7 (41)/10 (59)	9 (50)/9 (50)
Bilateral, n (%)	3 (15)	4 (18)
Ultrasonographic and radiographic features
Tendon thickness, mean (SD), mm	5.5±1.0	5.7±1.0
Doppler signal (MOS), mean (SD)	2.1±1.5	2.0±1.6
Enlarged bursa(e), n (%)	4 (20)	5 (23)
Tendon calcifications, n (%)	6 (30)	8 (36)
Haglund’s deformity, n (%)	4 (20)
Therapy prior to the trial, n (%)
None	13 (65)	13 (59)
Physical therapy	4 (20)	4 (18)
NSAIDs	0 (0)	1 (5)
Shockwave therapy	1 (5)	3 (14)
Corticosteroid injection	1 (5)	0 (0)
Traumeel injection	3 (15)	1 (5)
PRP-injections	2 (10)	1 (5)
Orthotics	1 (5)	3 (14)

^*Data are shown as mean±SD for continuous variables or as % for categorical variables.

CSAS, Cincinnati Sports Activity Scale; HTCR, high tendon compression rehabilitation; kg, kilograms; LTCR, low tendon compression rehabilitation; m, metre; mm, millimetre; MOS, Achilles tendon’ doppler signal scored by the Modified Ohberg Scale; n, number; NSAID, non-steroidal anti-inflammatory drugs; PRP, platelet-rich plasma; VISA-A, Victorian Institute of Sports Assessment-Achilles; y, years.

Primary outcome

The interaction term treatment-by-time was significant (F=6.030; p=0.004) for the VISA-A score, indicating that the change in scores over time differed between the two groups. The estimated mean VISA-A score increased significantly (p<0.001) in both groups. In the LTCR group, the score increased from 59.8 (95% CI: 55.1 to 64.4) at baseline to 84.2 (95% CI: 79.6 to 88.9) at 12 weeks and to 88.8 (95% CI: 84.2 to 93.5) at 24 weeks. In the HTCR group, the score increased from 59.1 (95% CI: 54.4 to 63.7) at baseline to 71.3 (95% CI: 66.7 to 76.0) at 12 weeks and to 78.4 (95% CI: 73.8 to 83.1) at 24 weeks (table 2). The between-group difference in VISA-A score was 12.9 points (95% CI: 6.2 to 19.6) at 12 weeks and 10.4 points (95% CI: 3.7 to 17.1) at 24 weeks, in favour of the LTCR group. These between-group differences in VISA-A reached the predetermined MCID of 10. The trajectories of the VISA-A score are shown in figure 3. Additionally, both the worst-case and best-case scenarios in the sensitivity analysis confirmed the observed effect of the main analysis at both 12 and 24 weeks, leading to similar conclusions regarding the treatment effect (online supplemental table S2).

Table 2. Main outcome measures at 12 and 24 weeks in the LTCR and HTCR groups^*.

			LTCR group		HTCR group		Between-group difference
Primary outcome measure
VISA-A score, points
Baseline			59.8 (55.1 to 64.4)		59.1 (54.4 to 63.7)
12 weeks			84.2 (79.6 to 88.9)		71.3 (66.7 to 76.0)		12.9 (6.2 to 19.6), p<0.001
24 weeks			88.8 (84.2 to 93.5)		78.4 (73.8 to 83.1)		10.4 (3.7 to 17.1), p=0.003
Within-group change (12 weeks)			24.4 (18.0 to 30.9), p<0.001		12.2 (5.8 to 18.7), p<0.001
Within-group change (24 weeks)			29.0 (22.6 to 35.5), p<0.001		19.3 (12.9 to 25.8), p<0.001
Secondary key outcome measures
Subjective patient satisfaction
12 weeks	Poor	Dissatisfied	1 (5.3)	4 (21.1)	0 (0.0)	10 (52.6)	p=0.046
	Moderate		3 (15.8)		10 (52.6)
	Good	Satisfied	7 (36.8)	15 (78.9)	6 (31.6)	9 (47.4)
	Excellent		8 (42.1)		3 (15.8)
24 weeks	Poor	Dissatisfied	0 (0.0)	1 (5.3)	1 (5.3)	8 (42.1)	p=0.009
	Moderate		1 (5.3)		7 (36.8)
	Good	Satisfied	9 (47.4)	18 (94.7)	6 (31.6)	11 (57.9)
	Excellent		9 (47.4)		5 (26.3)
Return to sports
12 weeks	No RTS	No RTS	1 (5.3)	1 (5.3)	1 (5.3)	6 (31.6)	p=0.045
	RTS, not to the desired sport		0 (0.0)		5 (26.3)
	RTS, not at preinjury level	RTS	16 (84.2)	18 (94.7)	12 (63.1)	13 (68.4)
	RTS, at preinjury level		2 (10.5)		1 (5.3)
24 weeks	No RTS	No RTS	0 (0.0)	1 (5.3)	0 (0.0)	5 (26.3)	p=0.090
	RTS, not to the desired sport		1 (5.3)		5 (26.3)
	RTS, not at preinjury level	RTS	12 (63.1)	18 (94.7)	10 (52.6)	14 (73.7)
	RTS, at preinjury level		6 (31.6)		4 (21.1)
VAS-HOP, mm
Baseline			32.5 (25.5 to 39.5)		27.8 (20.8 to 34.8)
12 weeks			6.2 (−0.9 to 13.3)		21.2 (14.2 to 28.2)		−15.0 (−25.0 to −4.9), p=0.004
24 weeks			5.4 (−1.6 to 12.4)		17.4 (10.4 to 24.4)		−12.0 (−22.1 to −1.9), p=0.020
Within-group change‡			−27.1 (−37.0 to −17.1), p<0.001		−10.4 (−20.4 to −0.5), p=0.037
VAS-ADL, mm
Baseline			44.4 (37.9 to 51.0)		39.4 (32.8 to 45.9)
12 weeks			8.6 (2.0 to 15.1)		18.6 (12.0 to 25.1)		−10.0 (−19.5 to −0.5), p=0.038
24 weeks			6.4 (−0.2 to 12.9)		16.8 (10.3 to 23.4)		−10.4 (−19.9 to −1.0), p=0.030
Within-group change‡			−38.0 (−47.3 to -28.8), p<0.001		−22.6 (−31.8 to −13.3), p<0.001
Achilles tendon APD, mm
Baseline			5.5 (5.0 to 6.0)		5.5 (5.0 to 6.0)
12 weeks			5.1 (4.6 to 5.6)		5.4 (4.9 to 5.9)		−0.3 (−1.0 to 0.5), p=0.457
24 weeks			4.9 (4.4 to 5.5)		5.4 (4.9 to 5.9)		−0.5 (−1.2 to 0.3), p=0.231
Within-group change‡			−0.6 (−0.8 to −0.4), p<0.001		−0.1 (−0.3 to 0.1), p=0.729
SLHR test, reps§
Baseline			17.3 (14.4 to 20.3)		18.9 (16.0 to 21.9)
12 weeks			28.4 (25.5 to 31.4)		26.7 (23.7 to 29.6)		1.7 (−2.5 to 6.0), p=0.412
24 weeks			28.9 (26.0 to 31.9)		27.2 (24.2 to 30.1)		1.7 (−2.5 to 6.0), p=0.412
Within-group change‡			11.6 (8.5 to 14.8), p<0.001		8.3 (5.1 to 11.4), p<0.001

^*Continuous values are presented as estimated adjusted means and estimated adjusted mean between-group differences (95% CI), analysed using mixed-model analysis. Data are adjusted for the following predefined baseline variables: age, sex, BMI, symptom duration, CSAS, VISA-A score and the presence of an associated Haglund’s deformity and/or an enlarged retrocalcaneal or subcutaneous bursa. Categorical values are presented as n (%) and analysed using Fisher’s exact test.

^†Significant interaction effect of time and treatment (p<0.05).

^‡Over 24 weeks.

^§Significant effect of time (p<0.05).

APD, anteroposterior diameter; BMI, body mass index; CSAS, Cincinnati Sports Activity Scale; HTCR, high tendon compression rehabilitation; LTCR, low tendon compression rehabilitation; mm, millimetre; reps, repetitions; RTS, return to sport; SLHR test, single-leg heel-raise test; VAS-ADL, Visual Analogue Scale during daily activities over the past 7 days; VAS-HOP, Visual Analogue Scale for pain during 10 unilateral hops; VISA-A, Victorian Institute of Sports Assessment-Achilles.

Figure 3. The time course of the VISA-A score in the LTCR and HTCR group. Data are the estimated means adjusted for the following predefined baseline variables: age, sex, BMI, symptom duration, CSAS, VISA-A score and the presence of an associated Haglund’s deformity and/or an enlarged retrocalcaneal or subcutaneous bursa. Error bars represent the 95% CI. *Indicates statistically significant difference (p<0.05) exceeding the minimal clinically important difference. BMI, body mass index; CSAS, Cincinnati Sports Activity Scale; HTCR, high tendon compression rehabilitation LTCR, low tendon compression rehabilitation; VISA-A, Victorian Institute of Sports Assessment-Achilles.

Secondary outcomes

All data of the secondary key outcomes are shown in figure 4 and table 2. After 12 weeks, 78.9% (n=15) of patients in the LTCR group reported satisfaction with their clinical outcomes, compared with 47.4% (n=9) in the HTCR group. By 24 weeks, patient satisfaction had increased to 94.7% (n=18) in the LTCR group and to 57.9% (n=11) in the HTCR group. Satisfaction was significantly higher in the LTCR group at 12 weeks (p=0.046) and 24 weeks (p=0.009). Additionally, 94.7% (n=18) of the LTCR group had returned to their desired sport by 12 weeks, maintaining this rate through 24 weeks. In contrast, 68.4% (n=13) of the HTCR group had returned to their desired sport at 12 weeks, increasing to 73.7% (n=14) by 24 weeks. These differences favouring LTCR were significant at 12 weeks (p=0.045) and showed a trend towards significance at 24 weeks (p=0.090). Furthermore, a significant treatment-by-time interaction effect was observed for VAS-HOP (F=6.751; p=0.002), VAS-ADL (F=5.485; p=0.006), and Achilles tendon diameter (F=7.770; p<0.001). VAS-HOP decreased significantly in both the LTCR (p<0.001) and HTCR (p=0.037) groups during the trial. However, pain scores were significantly lower in the LTCR group at both 12 and 24 weeks, with mean differences of −15.0 mm (95% CI: −25.0 to −4.9; p=0.004) and −12.0 mm (95% CI: −22.1 to −1.9; p=0.020), respectively. Similar findings were observed for VAS-ADL, which also decreased significantly in both groups (p<0.001), but more so in the LTCR group at both 12 and 24 weeks, with mean differences of −10.0 mm (95% CI: −19.4 to −0.5; p=0.038) and −10.4 mm (95% CI: −19.9 to −1.0; p=0.030), respectively. Achilles tendon diameter decreased significantly in the LTCR group, with a mean reduction of −0.6 mm (95% CI: −0.8 to −0.4; p<0.001) over 24 weeks, while no significant changes were found in the HTCR group (mean change: −0,1 mm, 95% CI: −0.3 to 0.1; p=0.729). In contrast, no significant interaction effect (F=2.321; p=0.105) was found for the SLHR test. However, a significant time effect was observed (F=76.924; p<0.001), with both groups showing similar improvements. Over 24 weeks, SLHR repetitions increased by an additional 11.6 (95% CI: 8.5 to 14.8; p<0.001) in the LTCR group and 8.3 (95% CI: 5.1 to 11.4; p<0.001) in the HTCR group. Similarly, only significant time effects were observed across the additional secondary outcomes, including LEFS (F=40.686; p<0.001), EQ-5D-5L (F=29.983; p<0.001), TSK-17 (F=5.849; p=0.004) and Achilles tendon’ Doppler signal (F=6.104; p=0.004). A complete overview of the additional secondary outcomes is available in the appendix (online supplemental table S2).

Figure 4. Analysis of change in secondary key outcome measures in the LTCR and HTCR treatment groups. Categorical variables (A, B) were analysed using Fisher’s exact test. Continuous variables (C, D, E, F) were analysed with mixed-model and data are the estimated means adjusted for the following predefined baseline variables: age, sex, BMI, symptom duration, CSAS, VISA-A score and the presence of an associated Haglund’s deformity and/or an enlarged retrocalcaneal or subcutaneous bursa. Error bars represent the 95% CI. *Indicates statistically significant between-group difference (p<0.05). APD, anteroposterior diameter; HTCR, high tendon compression rehabilitation; LTCR, low tendon compression rehabilitation; mm, millimetre; reps, repetitions; SLHR test, single-leg heel-raise test; VAS-ADL, Visual Analogue Scale during daily activities over the past 7 days; VAS-HOP, Visual Analogue Scale for pain during 10 unilateral hops.

Treatment credibility and expectancy

There was no significant between-group difference between the LTCR and HTCR groups regarding the credibility (21.64 vs 22.80, p=0.305) and expectancy (18.88 vs 20.71, p=0.243) of the allocated intervention. Thus, patients in both groups considered their interventions credible and expected to benefit from them, indicating that these factors did not confound the treatment outcomes.

Adherence and adverse events

Exercise adherence was comparable between the LTCR (82%) and HTCR (84%) groups (p=0.567). Additionally, no patients in the LTCR or HTCR groups reported using cointerventions or experiencing adverse events during the trial period.

Discussion

The principal findings from this two-arm RCT demonstrate that for athletes with insertional Achilles tendinopathy, LTCR leads to greater improvements in pain and function over both the short term (12 weeks) and medium term (24 weeks) compared with HTCR. This is supported by a higher, clinically meaningful VISA-A score at both 12 weeks (12.9 points) and 24 weeks (10.4 points) in favour of the LTCR intervention. Notably, VISA-A improves the most during the first 12 weeks, when progressive tendon loading is applied, but appears to plateau between 12 and 24 weeks as weekly supervision ends and patients resume sports, suggesting that extended rehabilitation or additional recovery time may be beneficial for complete pain resolution. Moreover, LTCR shows higher satisfaction and a trend towards a higher return to desired sports rates. Based on these results, we recommend a (partially) supervised progressive tendon loading programme that limits tendon compression, combined with patient education and the use of heel lifts, as the basis of conservative treatment for patients with insertional Achilles tendinopathy. One suggested reason for the superiority of LTCR is that reducing tendon compression during rehabilitation is believed to be less provocative for the sensitised insertional region of the Achilles tendon, thereby facilitating the restoration of load-bearing capacity and function through progressive exercise therapy.^{16 51 52} Additionally, decreasing tendon compression while applying tensile load may inhibit fibrocartilaginous metaplasia, resulting in lowered glycosaminoglycan and proteoglycan synthesis and, consequently, decreased fluid accumulation.^{13 16 53 54} Indeed, solely the LTCR approach seems to effectively improve tendon structure, as demonstrated by the significant reduction in tendon thickness after 24 weeks (−0.6 mm), compared with baseline. This reduction in tendon thickness observed after fulfilling LTCR decreases tendon compression from calcaneal impingement during sporting activities, potentially reducing pain and disrupting pathology perpetuation.¹⁴ Consequently, our results provide a possible explanation for the fact that a reduction in Achilles tendon insertion thickness over time is reported to be associated with clinical improvement in patients with insertional Achilles tendinopathy.⁴⁴

Strengths and limitations

The strength of our study lies in conducting and reporting this RCT in accordance with the current CONSORT guidelines.¹⁷ Key aspects such as randomisation, investigator blinding, assessment of treatment credibility and expectancy (since the intervention could not be blinded to participants), control of confounding variables, strict inclusion criteria, use of validated outcome measures, a relatively low loss-to-follow-up rate (<10%), high exercise adherence (>80%) and analysis by intention-to-treat enhance the internal validity of the study. In addition, the implementation of state-of-the-art recommendations in tendinopathy management, such as patient education, a progressive tendon-loading programme that incorporates commonly used exercises, the combination of home-based exercises with weekly physiotherapist supervision and the use of commercially available heel lifts demonstrates the study’s generalisability, thereby further strengthening its external validity as well.⁵¹ However, several limitations should be acknowledged when interpreting the findings. First, we are unable to define the impact of individual components of the LTCR intervention on favourable clinical outcomes. Previous pilot data illustrated that solitary, non-progressive eccentric calf muscle training with limited ankle dorsiflexion might yield positive clinical results in patients with insertional Achilles tendinopathy.⁵⁵ More recently, it has been shown that using heel lifts alone also increases VISA-A scores by 7 points in the short term in this population.⁵⁶ However, this increase is substantially lower than the 29.0-point improvement observed in the LTCR group. Future research should aim to isolate the components of the LTCR and determine the extent of their individual contributions, which would be beneficial for designing the most optimal treatment regimen. Second, although participants were blinded to the study hypotheses and had similar beliefs and expectancies across both treatment groups, they were not blinded to the interventions themselves. This lack of blinding could have introduced bias when completing the patient-rated outcome measures. Third, as we only included sport-active patients with chronic symptoms (lasting more than 12 weeks), caution is warranted when generalising our findings to a non-sporting population or patients with more acute and/or reactive symptoms. Fourth, while the difference in VISA-A score improvements favouring LTCR exceeded the MCID at both 12 and 24 weeks, the lower bounds of the 95% CI fell below the MCID, reflecting some uncertainty about the size of the effect. Fifth, the TENDINopathy Severity Assessment–Achilles is now recommended as a superior outcome measure for Achilles tendinopathy, as the VISA-A may lack validity.⁵⁷ However, at the time this trial was conducted, the VISA-A was the standard and most widely used measure.

Comparison of findings with previous research

Our study is the first RCT to demonstrate that reducing tendon compression during insertional Achilles tendinopathy rehabilitation leads to superior clinical outcomes. Jonsson et al investigated the effectiveness of a modified 12-week Alfredson protocol that limited ankle dorsiflexion during eccentric calf training, and they reported a higher satisfaction rate of 67%.⁵⁵ This was a significant improvement compared with the previously reported 32% satisfaction rate with the traditional Alfredson protocol, which involved eccentric calf training with full ankle dorsiflexion.⁵ While their pilot study laid important groundwork in demonstrating the potential benefits of reducing tendon compression during insertional Achilles tendinopathy rehabilitation, our study strengthened internal validity by including a control group (HTCR), a larger sample size and a longer follow-up period. Additionally, we adopted a more comprehensive approach by incorporating a progressive tendon loading programme combined with education, calf muscle massage and heel lifts, which likely contributed to the higher satisfaction rate of 94.7% observed in the LTCR group.³ In a more recent study by Chimenti et al, which included patients with both insertional Achilles tendinopathy and midportion Achilles tendinopathy, a 20-point increase in VISA-A scores was reported following an 8-week progressive exercise therapy combined with education (from either a biopsychosocial or biomedical perspective).⁵⁸ In this study, tendon compression was minimised by limiting ankle dorsiflexion during exercise and using heel lifts, which may explain the improved outcomes compared with previous exercise-based rehabilitation studies.^{5 6} However, it remains unclear whether these positive results are truly due to LTCR or other factors, such as progressive tendon loading and/or education, as there was no HTCR group.

Conclusion

In sport-active patients with insertional Achilles tendinopathy, LTCR proved more effective than HTCR in reducing tendon pain and improving function at both 12 and 24 weeks. Therefore, these findings affirm the recommendation of progressive exercise therapy that limits ankle dorsiflexion, combined with education and heel lifts to reduce Achilles tendon compression during rehabilitation, could be considered the preferred conservative treatment for insertional Achilles tendinopathy.

We plan to disseminate the results to study patients and/or patient organisations.

Data availability statement

Data are available upon reasonable request.