Intratendinous pressure changes in the Achilles tendon during stretching and eccentric loading: Implications for Achilles tendinopathy

Mechanical overload is considered the main cause of Achilles tendinopathy. In addition to tensile loads, it is believed that the Achilles tendon may also be exposed to compressive loads. However, data on intratendinous pressures are lacking, and consequently, their role in the pathophysiology of tendinopathy is still under debate. Therefore, we aimed to evaluate the intratendinous pressure changes in the Achilles tendon during stretching and eccentric loading. Twelve pairs of human cadaveric legs were mounted in a testing rig, and a miniature pressure catheter was placed through ultrasound‐guided insertion in four different regions of the Achilles tendon: the insertion (superficial and deep layers), mid‐portion, and proximal portion. Intratendinous pressure was measured during three simulated loading conditions: a bent‐knee calf stretch, a straight‐knee calf stretch, and an eccentric heel‐drop. It was found that the intratendinous pressure increased exponentially in both the insertion and mid‐portion regions of the Achilles tendon during each loading condition (p < 0.001). The highest pressures were consistently found in the deep insertion region (p < 0.001) and during the eccentric heel‐drop (p < 0.001). Pressures in the mid‐portion were also significantly higher than in the proximal portion (p < 0.001). These observations offer novel insights and support a role for compression in the pathophysiology of Achilles tendinopathy by demonstrating high intratendinous pressures at regions where Achilles tendinopathy typically occurs. To what extent managing intratendinous pressure might be successful in patients with Achilles tendinopathy by, for example, avoiding excessive stretching, modifying exercise therapy, and offering heel lifts requires further investigation.

Mechanical overload is considered the main cause of Achilles tendinopathy.
In addition to tensile loads, it is believed that the Achilles tendon may also be exposed to compressive loads. However, data on intratendinous pressures are lacking, and consequently, their role in the pathophysiology of tendinopathy is still under debate. Therefore, we aimed to evaluate the intratendinous pressure changes in the Achilles tendon during stretching and eccentric loading. Twelve pairs of human cadaveric legs were mounted in a testing rig, and a miniature pressure catheter was placed through ultrasound-guided insertion in four different regions of the Achilles tendon: the insertion (superficial and deep layers), mid-portion, and proximal portion. Intratendinous pressure was measured during three simulated loading conditions: a bent-knee calf stretch, a straight-knee calf stretch, and an eccentric heel-drop. It was found that the intratendinous pressure increased exponentially in both the insertion and mid-portion regions of the Achilles tendon during each loading condition (p < 0.001). The highest pressures were consistently found in the deep insertion region (p < 0.001) and during the eccentric heel-drop (p < 0.001). Pressures in the mid-portion were also significantly higher than in the proximal portion (p < 0.001). These observations offer novel insights and support a role for compression in the pathophysiology of Achilles tendinopathy by demonstrating high intratendinous pressures at regions where Achilles tendinopathy typically occurs. To what extent managing intratendinous pressure might be successful in patients with Achilles tendinopathy by, for example, avoiding excessive stretching, modifying exercise therapy, and offering heel lifts requires further investigation.

| INTRODUCTION
Achilles tendinopathy is a debilitating injury that is common among athletes, especially those involved in running sports. Around 30% of all runners exhibit Achilles tendinopathy with an annual incidence of 7-9%. 1 Effective treatment is still a challenge and full recovery can take a year or even longer. As a result, athletes experience persistent pain, which further leads to reduced performance, loss of training time, and sometimes even the premature end of their sporting careers. 2 Achilles tendinopathy is classified according to its anatomical region and broadly includes insertional and mid-portion tendinopathy. 3 Several mechanisms are considered to play a role in the etiology of Achilles tendinopathy, yet a prominent role seems present for excessive overload. 2,4 Traditionally, the nature of this overload is thought to be purely tensile. 5 For this reason, previous biomechanical studies have mainly focused on this tensile load and the tensile mechanical properties of the Achilles tendon. [6][7][8] However, tendons may also be exposed to compressive loads. 5,9,10 This certainly seems to be the case for the Achilles tendon insertion, where external compression may occur as the tendon wraps around the posterosuperior prominence of the calcaneus during ankle dorsiflexion. In addition, some researchers suggest that internal compression may also occur in the mid-portion region of the Achilles tendon as a result of torsion or twisting. 5,11,12 The formation of fibrocartilage, which is typically found in histological examination of tendinopathic tissue, may be considered an adaptation to this compressive load. 11,13,14 However, to date, no quantitative data are available on the intratendinous pressure changes in the Achilles tendon in response to this potential compressive load during activities. Therefore, these assumptions are based purely on indirect measurements, 15 theoretical models 16,17 or histological research. 12,18,19 The aim of our study was threefold: (1) to investigate the influence of ankle dorsiflexion on the intratendinous pressure of the Achilles tendon, (2) to identify differences in intratendinous pressure between the insertion (superficial and deep layers), mid-portion and proximal portion, and (3) to identify differences in intratendinous pressure between stretching and eccentric loading. We hypothesized that the intratendinous pressure of the Achilles tendon would progressively increase with ankle dorsiflexion during stretching, primarily at the deep insertion region and that the intratendinous pressure would also be greater during eccentric loading.

| MATERIALS AND METHODS
This biomechanical study was designed to measure intratendinous pressure changes in the Achilles tendon during stretching and eccentric loading in both the insertion, mid-portion, and proximal portion. Approval was obtained by the Ethics Committee of the Ghent University Hospital (BC-09407).

| Specimen preparation and mounting
Twelve matched pairs of fresh frozen human cadaveric legs from six male and six female donors (mean age 77.9 ± 10 years, mean BMI 23.3 ± 5.1 kg/m 2 ) were collected from a tissue bank for experimental research. All donors gave their written consent. The specimens were obtained by hip disarticulation, underwent a full lower limb computed tomography (CT) scan, and were then placed in a −20°C freezer for storage. None of the specimens were excluded since none had evidence of any trauma sequelae, implant material, or bone deformities (e.g., Haglund's deformity). A 3D segmentation model of each bone was created based on the CT using Mimics software (Mimics software®, Materialise). Rigid markers linked to the femur and the tibia allowed for real-time position monitoring using OptiTrack cameras (OptiTrack®, NaturalPoint). Using these three-dimensional models and an in-housedeveloped software script, the flexion angles of the knee and ankle joint could be measured in real time. 20 Before testing, the specimens were thawed at room temperature for 24 h. The femur was sectioned 95 mm from the femoral hip center to pot the femur in a container using resin. To assure a correct anatomical position of the bones in the test setup, dedicated guides were designed and printed based on the preoperative CT scan. The proximal tendons of the medial and lateral gastrocnemius muscle were identified and stitched, using Ti-Cron™ braided polyester sutures. Note here that the tendons were not cut but only sewn up. A customized validated testing rig was used as an activity simulator in this study ( Figure 1).

K E Y W O R D S
Achilles tendon, compressive load, eccentric loading, intratendinous pressure, stretching, tendon biomechanics, tendon compression Compared with previous publications, an adjustment has been made by adding an ankle-foot holder, which allowed movement of the ankle joint in the sagittal plane using two actuators. 21,22 The major advantage of this setup was the preservation of the anatomical connection between the gastrocnemius muscle and the knee, allowing a more physiological tensile load of this bi-articular muscle on the Achilles tendon during ankle plantar-or dorsiflexion. The specimen's foot was inserted in the ankle-foot holder and fixated with three Steinmann pins. One pin was drilled through the medial and lateral calcaneus and one through the medial talus so that movement in the subtalar joint was impeded and only plantar and dorsal flexion could occur in the tibiotalar joint. Finally, a connection was made between the cable pulley system (two gastrocnemius actuators) and the lateral and medial head of the gastrocnemius, so that additional tensile load could be applied to the Achilles tendon. F I G U R E 1 Testing rig with a prepared specimen inserted F I G U R E 2 Overview of the four different Achilles tendon regions that were studied. Control of the catheter location was assessed using ultrasound imaging. The tip of the needle (bright foci) can be seen in the long (left) and the short axis (right) of each tendon region 2.2 | Intratendinous pressure determination A commercially available pressure sensor, the Mikro-Tip® catheter transducer (Model SPR-524; Millar Instruments), was used to measure intratendinous pressure of the Achilles tendon. The transducer is flexible and has highfidelity pressure sensors mounted at the tip, a natural frequency of ≥10 kHz, and a diameter of 1.16 mm. The device functions by the piezoresistive principle and relies on chip technology transforming pressure into an electronic signal. Data from the implanted sensor were recorded using a digital acquisition system (PowerLab 8/30, AD Instruments) and analyzed with LabChart 8 software. The reliability of this transducer for measuring total interstitial tissue pressure has been published before and is considered a gold standard in both positive and negative pressure regimes. 23 The same technique has already been successfully used in human tissue for the dynamic measurement of intraneural, intradiscal, and intramuscular pressure. [24][25][26] Before the pressure measurement was performed, calibration and equilibration were performed as advised by the manufacturer. A multipoint calibration of the measurement equipment at 0, 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100% of, based on pilot testing, the expected maximum intratendinous pressure, that is, 4000 mm Hg, was undertaken.

| Tendon region determination
The pressure sensor was positioned under ultrasound guidance by an experienced sonographer in four different regions of the Achilles tendon, namely the proximal portion (30 mm distal to the myotendinous junction of the medial gastrocnemius muscle), the mid-portion (25 mm proximal to the posterosuperior calcaneal border), and the superficial and deep insertions (5 mm distal to the posterosuperior calcaneal border in the ½ superficial and ½ deep insertion layers, respectively). The locations related to the latter two distances were chosen because Achilles tendinopathy typically occurs in these regions. 27,28 Ultrasonographic imaging was performed using a GE Logic S8 ultrasound system (GE Healthcare) with a highresolution linear ML6-15 transducer (range 4-15 MHZ, MSK pre-set). A 17-gauge catheter was inserted percutaneously and advanced until the tip of the needle was at the desired location and centered in the mediolateral and anteroposterior directions, which was evaluated both longitudinally and transversally ( Figure 2). After the needle was removed, the flexible pressure sensor was inserted through the catheter, which in turn was withdrawn so the pressure sensor was completely isolated.

| Testing protocol
Before testing, 10 preconditioning cycles of both knee flexion (0°-120°) and ankle flexion (−15° plantarflexion to 15° dorsiflexion) were performed. The experiments were conducted in three different loading conditions. In the first phase (bent-knee calf stretch), the ankle joint was placed in 15° plantarflexion (−15°), and a squat motion was performed dynamically until a tibiotalar angle of +15° was achieved (see the Video S1). In the second phase (straightknee calf stretch), the knee was fixed in full extension, and the ankle was moved from a plantar flexed position (−15°) to dorsiflexion (+15°), tilting the ankle-foot holder upward. Finally, in the third phase (eccentric heel-drop), the knee was still fixed in full extension, but an additional continuous tensile load was applied to the Achilles tendon by activating the two gastrocnemius actuators. To preserve the integrity of the specimens, a downscaling by a factor five was applied to the average tensile force acting on the Achilles tendon during an eccentric heel-drop, resulting in a total tensile load of 200 N (100 N for each head of the gastrocnemius). 29 Intratendinous pressure measurements were recorded at each tendon region during the ankle motion cycle from plantar flexion (−15°) to dorsiflexion (15°) and back. Two cycles of each loading condition were performed without removing the catheter, and the mean intratendinous pressure was determined. After each series of measurements, ultrasonography was used to confirm that the pressure sensor did not move.

| Statistical analysis
Statistical analysis was performed using the IBM® SPSS® 28.0 Statistics software package. The effect of ankle flexion angle, tendon region, and loading condition on the intratendinous pressure of the Achilles tendon was investigated with a linear mixed model. Analysis was executed with the cadaver as a random factor and ankle flexion angle, tendon region, and loading condition as main fixed factors. Since the raw data showed an exponential relationship between intratendinous pressure and ankle flexion angle, a log transformation of the data was done. The residuals were checked for normal distribution and homoscedasticity. For the significant fixed factors, estimates of the fixed effects were determined and pairwise comparisons of the resultant estimated geometric means of the intratendinous pressure were performed using Bonferroni correction. After back-transforming the logarithmic values to the original scale, the relative differences in pressure were defined at 15° of ankle dorsiflexion with their 95% confidence interval (CI). The level of significance was set at p value ≤0.05.

| RESULTS
The results of the mixed model analysis are displayed in Table 1 and show that intratendinous pressure of the Achilles tendon is influenced by both ankle flexion angle (F = 9404.88, p < 0.001), tendon region (F = 1269.29, p < 0.001), and loading condition (F = 435.98, p < 0.001). Interaction effects between tendon region and ankle flexion angle (F = 235.05, p < 0.001), between loading condition and ankle flexion angle (F = 101.49, p < 0.001), and between tendon region and loading condition (F = 10.96, p < 0.001) are also identified. The estimates of the significant main and interaction fixed effects are presented in Table 2.

| Effect of ankle dorsiflexion
A positive correlation is identified between the ankle flexion angle and intratendinous Achilles tendon pressure (p < 0.001) ( Table 2). Figures 3 and 4 show that intratendinous pressure increases exponentially in the mid-portion, superficial insertion, and deep insertion during ankle dorsiflexion. Visually, the intratendinous pressure remains relatively stable in plantar flexion and starts to increase markedly from 0° onwards.

| Effect of tendon region
As shown in Table 2, compared with the proximal portion, higher intratendinous pressures were found in both insertion and mid-portion regions (p < 0.001), regardless of the loading condition. In addition, since a significant interaction effect was also found between the tendon region and ankle flexion angle, meaning the slope of the pressure curve is influenced by the tendon region, an estimate of this effect was also determined. It was also found that, compared with the proximal portion, intratendinous pressure increased more in both insertion and mid-portion regions during dorsiflexion (p < 0.001). Finally, the pairwise comparisons in Table 3 demonstrate that for each loading condition, intratendinous pressure was consistently higher in the deep insertion than in any other tendon region at 15° of ankle dorsiflexion. For example, during the straight-knee calf stretch, intratendinous pressure was significantly higher in the deep insertion compared with the proximal portion (36.60 times higher, p < 0.001), mid-portion (3.75 times higher, p < 0.001), and superficial insertion (2.92 times higher, p < 0.001). Furthermore, intratendinous pressure was significantly higher in the mid-portion compared with the proximal portion during both the bent-knee calf stretch (7.45 times higher, p < 0.001), straight-knee calf stretch (9.77 times higher, p < 0.001), and eccentric heeldrop (8.52 times higher, p < 0.001).

| Effect of loading condition
As shown in Table 2, compared with the bent-knee calf stretch, higher intratendinous pressures were found during both the straight-knee calf stretch and the eccentric heel-drop (p < 0.001), regardless of the tendon region. In addition, compared with the bent-knee calf stretch, intratendinous pressure increased more during the straightknee calf stretch and the eccentric heel-drop (p < 0.05). Finally, pairwise comparisons in Table 4 show that, for each tendon region, intratendinous pressure at 15° of ankle dorsiflexion differed significantly between each loading condition with the order of highest to lowest pressure being consistently the eccentric heel-drop, straightknee calf stretch, and bent-knee calf stretch (p < 0.001). For example, in the deep insertion, intratendinous pressure was significantly higher during the eccentric heeldrop than during the bent-knee calf stretch (

| DISCUSSION
The principal findings of this study are threefold: (1) intratendinous pressure in the Achilles tendon increased exponentially with ankle dorsiflexion during stretching in both the insertion and mid-portion regions, (2) intratendinous pressures increased most in the deep insertion layers, and (3) ankle dorsiflexion during eccentric loading resulted in higher intratendinous pressures than during stretching. To the best of our knowledge, this is the first study to report intratendinous pressure changes in the Achilles tendon during stretching and eccentric loading. In contrast to the large number of studies examining the intratendinous strain changes in the Achilles tendon during mechanical loading, the intratendinous pressure changes in the Achilles tendon have hardly been investigated. 6,7 Chimenti et al. investigated the compressive strain of the Achilles tendon insertion during ankle dorsiflexion, using ultrasound elastography, but this non-invasive measurement technique only provides a rough, indirect estimate of the intratendinous pressure. 15 Recently, Matsui et al. 30 examined the extratendinous contact pressure between the Achilles tendon and the calcaneus during ankle dorsiflexion. However, since the external pressure exerted on the outside of biological tissue material is not translated into a homogeneous internal pressure distribution, it is uncertain to what extent this measured pressure corresponds to the intratendinous pressure observed by the tenocytes. 31 These tenocytes, according to the mechanotransduction model, are the primary mechanosensors that regulate the extracellular matrix based on the mechanical load applied to them. Therefore, they are also considered prominent actors in the pathophysiology of tendinopathies. 11,32 We believe that measuring interstitial pressure in tendons, as done in this study, would provide a more accurate estimate of the effective compressive load detected by tenocytes.
The first hypothesis was supported by the exponential increase in intratendinous pressure of the Achilles tendon insertion with ankle dorsiflexion during stretching. As expected, the increase was significantly higher in the deep tendon insertion compared with all other tendon regions. These findings are consistent with the in vivo study by Chimenti et al. 15  in the deep insertion region of the Achilles tendon during ankle dorsiflexion compared with the superficial insertion. A reasonable explanation is the mechanical impingement that occurs when the Achilles tendon wraps around the posterosuperior prominence of the calcaneus during ankle dorsiflexion, resulting in external compression, which was recently demonstrated in the in vitro study by Matsui et al. 30 In that study, based on published figures, the average contact pressure was approximately 0.85 MPa at 15° of ankle dorsiflexion, corresponding to 6375.52 mm Hg. Although this value is higher than the average pressure in the deep insertion region at this angle in our study, it is in the same order of magnitude, and this difference can be explained by the fact that we determined intratendinous pressure instead of extratendinous contact pressure. Moreover, compared with that study, our results provide an additional relevant clinical finding, as it was shown quantitatively that the increase in intratendinous pressure with ankle dorsiflexion was higher in the deep insertion region than in the superficial insertion region. This is consistent with the biomechanical model of Pauwels, 17 who described that theoretically, the intratendinous pressure, caused by the compressive forces in tendons wrapped around bones, should gradually decrease toward the peripheral tendon layers. Furthermore, our findings support the hypothesis of several other investigators that compression is indeed a relevant load for the Achilles tendon insertion. 2,5,9 It is in this deep tendon region, where we consistently observed the highest intratendinous pressures, that pathological changes of an insertional Achilles tendinopathy are generally found and chondrogenic metaplasia occur. 5,19 According to the mechanotransduction theory, this formation of fibrocartilage can be seen as an adaptation to excessive pressure on the tenocytes in wrap-around tendons. 32,33 A rather F I G U R E 3 Intratendinous pressure of the Achilles tendon (mm Hg) as a function of ankle flexion angle (°), comparing the four tendon regions for each loading condition. The data are presented as estimated geometric means with their corresponding limit of the 95% confidence interval. *p ≤ 0.05 indicates significantly higher intratendinous pressures compared with the proximal portion. † p ≤ 0.05 indicates significantly higher intratendinous pressures compared with the mid-portion. ‡ p ≤ 0.05 indicates significantly higher intratendinous pressures compared with the superficial insertion F I G U R E 4 Intratendinous pressure of the Achilles tendon (mm Hg) as a function of ankle flexion angle (°), comparing the three loading conditions for each tendon region. The data are presented as estimated geometric means with their corresponding limit of the 95% confidence interval. *p ≤ 0.05 indicates significantly higher intratendinous pressures compared with the bent-knee calf stretch. † p ≤ 0.05 indicates significantly higher intratendinous pressures compared with the straight-knee calf stretch striking finding, at first glance, is that intratendinous pressure also increased markedly in the mid-portion region with ankle dorsiflexion. This observation can be explained by the unique anatomical structure of the mid-portion region, which is characterized by a spiral movement of the tendon fibers of approximately 90°, with the medial fibers rotating posteriorly and the posterior fibers rotating laterally from proximal to distal. 34,35 Traction on the tendon T A B L E 3 Pairwise comparisons of estimated mean intratendinous pressure between the different tendon regions for each loading condition may cause torsion in this region as the subtendons twist around each other, similar to wringing out a towel. 11,36 Indeed, it has already been shown that chondral metaplasia also occurs in chronic mid-portion tendinopathies, and this was partially explained by the torsion-induced internal compression. 12,18 Finally, a discrete increase in intratendinous pressure is also noted in the proximal part of the Achilles tendon, which can potentially be explained by the Poisson effect. This indicates that elongation of a tendon leads to a decrease in the cross-sectional area, proportional to the longitudinal strain. In essence, this means that the Achilles tendon becomes proportionally thinner when it is stretched. 37 Nevertheless, two findings of this study suggest that the impact of this Poisson effect on the strong intratendinous pressure increase in both the mid-portion and insertion of the Achilles tendon is relatively limited. First, the intratendinous pressure increase with ankle dorsiflexion in the linear structure of the proximal portion was significantly lower than in the twisted structure of the midportion of the Achilles tendon. Second, it has already been shown that the longitudinal strain on the Achilles tendon insertion is lower in the deep layers than in the superficial layers, whereas we see the opposite with respect to intratendinous pressure. 6 Despite these two arguments, it is not possible to give a conclusive answer because the exact influence of the Poisson effect on the intratendinous pressure changes since tensile strains of the tendon were not determined.

Loading condition (I) Tendon region (J) Tendon region
The second hypothesis was supported by higher intratendinous pressures in the Achilles tendon insertion during the eccentric heel-drop, compared with both calf stretches. As already suggested in the literature, the combination of tensile stress and impingement of the Achilles tendon during ankle dorsiflexion seems to lead to higher external compression. 5 Moreover, higher pressures were also found during the straight-knee stretch compared with the bent-knee stretch. This can be explained by the fact that during the bent-knee stretch, the restraining effect of the bi-articular gastrocnemius muscles on ankle dorsiflexion is fully eliminated around 20° of knee flexion. 38 As a result, the tensile stress on the Achilles tendon from this knee flexion angle and onward will come solely from the uni-articular soleus muscle. This is reflected in lower tensile stress and, consequently, lower intratendinous pressure. Although Matsui et al. 30 also showed that the extratendinous pressure on the Achilles tendon increased with the application of tensile force, these pressures were only determined in a plantigrade position (at 0° of ankle dorsiflexion), and their maximum applied tensile force in this position (70 N) was significantly lower than used in our study (200 N). In addition, our study also provides an insight into pressure changes in the other tendon regions. Indeed, also in the mid-portion region of the Achilles tendon, a clear difference in intratendinous pressures was seen between the different loading conditions. As mentioned earlier, this can be explained by the torsion-induced internal compression that will be more pronounced as more tensile stress is applied to the Achilles tendon. Finally, a similar trend was observed in the proximal portion, but this traction-induced pressure increase appears to be rather negligible, both in terms of absolute values and certainly in relation to the mid-portion and insertion regions. Nevertheless, it confirms the computational model of Lavagnino et al. 39 that the cellular mechanical stress, perpendicular to the long axis of the tendon, increases as the strain rate increases. It has been suggested that this may be the cause of the fluid excretion and volume loss of tendons in response to high tensile loading.
Some limitations should be considered for this biomechanical study. There are inherent limitations to cadaveric studies that may impede data interpretation, such as the fact that the injury history of the Achilles tendon was unknown. Also, despite careful preservation, the quality and age of the specimens may not exactly reproduce natural conditions and may behave differently from the target clinical population. In addition, although miniaturized pressure catheters were used, the implanted sensor itself may result in a space-occupying process, resulting in higher-measured pressures than would be the case in natural in vivo conditions. Finally, some other relevant factors that may influence intratendinous pressure were not investigated. First, the test setup in this study only allowed for dorsi-and plantarflexion of the ankle joint to be performed. An analysis of the influence of in-or eversion on intratendinous pressures could not be determined. However, it is hypothesized that excessive eversion may accentuate the torsion-induced internal compression in the mid-portion region of the Achilles tendon. 40 Second, to keep the gastrocnemius muscles intact and mimic their physiological conditions as closely as possible, it was decided not to perform a dissection. Therefore, the underlying soleus muscle could not be separately stitched and consequently could not be activated during the eccentric heel-drop. Although the activation of the soleus muscle may have a significant effect on the intratendinous pressure in the mid-portion region (as it may increase the torsion-induced internal compression), it has recently been shown that the contribution of the soleus muscle to the total tensile load acting on the Achilles tendon during an eccentric heel-drop with a straight knee is relatively limited. 41 Third, the degree of torsion of the Achilles tendon, which is variable, was not determined. 34,35 Fourth, the morphology of the calcaneus was not taken into account, although it has been shown that calcaneal hyperconvexity is associated with insertional Achilles tendinopathy. 42 Logically, impingement-induced external tendon compression will be more pronounced in cadavers with a more convex morphology of the calcaneus. Nevertheless, we plan to further investigate the full field of intratendinous Achilles tendon pressures, including determining the effect of these possible contributing factors.
Despite these limitations, the findings of this study may have some important clinical implications, especially for the treatment of Achilles tendinopathies. There is a consensus that exercise therapy has a beneficial effect on the recovery of Achilles tendinopathies. 2 Since repair and remodeling of collagen is stimulated by loading the tendon, complete rest of an injured tendon seems to be detrimental. Essentially, the primary function of a tendon is to resist high tensile forces while transmitting forces from muscle to bone. This task requires strengthening of the Achilles tendon. However, the progressive formation of fibrocartilage in Achilles tendinopathy, resulting in a reduction in axial stiffness, makes the tendon more susceptible to degenerative ruptures under high tensile forces. 43 In addition, the increased content of glycosaminoglycan and proteoglycan in Achilles tendinopathies, resulting in swelling and reduced permeability, may likely lead to a higher increase in intratendinous pressure during ankle dorsiflexion. 5 Both of these histological features can be seen as a result of excessive pressure loading and may further perpetuate the pathology of Achilles tendinopathy. As these changes are adaptive and partially reversible over time, it is therefore essential that exercise therapy is aimed at protecting against compressive forces, but at the same time exerting sufficient tensile forces to stimulate collagen synthesis. 11 Based on our findings, an ankle range of motion (ROM) between −15° and 0° flexion seems to be a perfect "safe" zone to apply high tensile forces, both for the mid-portion and insertion regions of the Achilles. Furthermore, as already suggested, educating the patient about the importance of this compression, and offering heel lifts (inside or outside the shoe) to limit ankle dorsiflexion in daily activities also seem to be good approaches in the conservative treatment of Achilles tendinopathy. 2 Once the Achilles tendon adapts adequately and its volume decreases, compression can be gradually offered again in exercise therapy. Essentially, sufficient dorsiflexion of the ankle is necessary for various activities in both sports and daily life. Further interventional research is needed to determine whether limiting the compressive load could help to reduce the symptoms of Achilles tendinopathy and promote the healing process.

| PERSPECTIVE
Some widely cited articles suggest that compressive load might be an important factor in the development of tendinopathies. 5,9,11 A major argument is that tendinopathic changes occur in areas where tendons are theoretically exposed to high pressures. Applied to the Achilles tendon, the results of our study support this hypothesis. It was found that intratendinous pressure increases exponentially during ankle dorsiflexion, both in the mid-portion and insertion regions, due to tendon torsion and impingement, respectively. Moreover, higher intratendinous pressures were found in the deep insertion layers than in the superficial layers, while previous studies showed the opposite relationship with respect to the amount of tensile load. 6,7 Since tendinopathic changes occur mainly in the deep insertion layers, our results support the idea that compressive overload may be more detrimental for a tendon than tensile overload. Finally, it was confirmed that eccentric loading under ankle dorsiflexion results in higher intratendinous pressures than stretching. Based on these findings, we suggest treatment modifications that limit the amount of compression to improve recovery from Achilles tendinopathy. This can be done by avoiding excessive stretching, modifying exercise therapy that applies sufficient tensile load in a "safe" ROM, and providing heel lifts that limit ankle dorsiflexion during daily activities.