Subject-specific biomechanics influences tendon strains in patients with Achilles tendinopathy

Key points

1. 01Bilateral heel rise and bilateral heel drop produced the lowest average mid-tendon strains (3.1% and 3.4%), while bilateral hopping produced the highest (11.5%).
2. 02Only unilateral heel drop and walking reached the group-average optimal strain range (6 +/- 1%) at peak muscle force.
3. 03At the individual level, optimal strain was reached during different exercises across participants, highlighting that one-size-fits-all protocols are inadequate.
4. 04Subject-specific muscle forces explained more variance in tendon strain than subject-specific material properties (lower RMSE for muscle forces model: 0.006 average, 0.151 peak vs. material model: 0.016 average, 0.223 peak).
5. 05Peak strain consistently occurred at the edge of the soleus sub-tendon in the mid-portion, the site of greatest pathological change in mid-portion Achilles tendinopathy.

How it was conducted

Design

Cross-sectional biomechanical study using subject-specific finite element (FE) models

Participants

21 adults with mid-portion Achilles tendinopathy (17 male, 4 female; mean age 49 +/- 13 years; VISA-A score 73 +/- 19)

FE model inputs

Subject-specific tendon geometry from 3D freehand ultrasound, elastic modulus from experimental stress-strain curve, and triceps surae muscle forces from 3D motion capture combined with musculoskeletal modelling in OpenSim

Exercises simulated

Six exercises in order of expected load: bilateral heel rise, bilateral heel drop, unilateral heel drop (knee extended), walking, unilateral heel drop (knee flexed), bilateral hopping

Primary outcome

Average maximum principal strain in the mid-portion of the Achilles tendon, ranked across exercises and compared to an optimal strain range of 5-7%

Comparison models

Three model types per participant: subject-specific (all features individualised), material model (subject-specific elastic modulus, generic muscle forces), muscle forces model (subject-specific muscle forces, generic elastic modulus)

What they found

• Generalized exercise ranking by average mid-portion strain: bilateral heel rise 3.1% +/- 1.0%, bilateral heel drop 3.4% +/- 0.9%, unilateral heel drop 6.6% +/- 2.3%, walking 6.9% +/- 2.0%, unilateral heel drop bent knee 7.8% +/- 2.3%, bilateral hopping 11.5% +/- 3.3%.
• All pairwise exercise comparisons were statistically significant (p < 0.05) except unilateral heel drop vs. walking (not significant).
• At the group level, only unilateral heel drop and walking fell within the optimal strain range (6 +/- 1%).
• At the individual level: 12 participants reached optimal strain during walking, 8 during unilateral heel drop, 7 during unilateral heel drop bent knee, 2 during bilateral hopping, 1 during bilateral heel rise, 1 during bilateral heel drop.
• Peak strain values in the mid-portion: bilateral heel rise 19.4% +/- 8.1%, bilateral heel drop 19.6% +/- 6.9%, unilateral heel drop 48.1% +/- 29.9%, walking 40.3% +/- 20.9%, unilateral heel drop bent knee 64.8% +/- 38.9%, bilateral hopping 96.3% +/- 46.2%.
• Peak strain comparisons: bilateral heel rise vs. bilateral heel drop not significant (p = 0.774); walking vs. unilateral heel drop not significant (p = 0.061); all other pairs significant (p < 0.05).
• Regression analysis: RMSE for average strain was 0.016 (material model) vs. 0.006 (muscle forces model); RMSE for peak strain was 0.223 (material model) vs. 0.151 (muscle forces model), indicating muscle forces model more closely matched the subject-specific model.

Limitations

• Frictional contact between sub-tendons was not modelled, which may underestimate non-uniform deformation in tendinopathic tendons where sliding is compromised.
• A hyperelastic (not viscoelastic) material was used, so loading rate effects on strain are not captured.
• The optimal strain threshold (approximately 6%) was derived from studies on healthy tendons (one in-vivo human study, one in-vitro animal study) and may not directly translate to tendinopathic tissue.
• FE models were not validated against controlled experimental strain measurements, limiting confidence in absolute strain magnitudes.

Why it matters

For patients

Patients with Achilles tendinopathy may need a rehabilitation programme built around their own muscle strength and tendon size rather than a generic protocol, because the exercise that delivers the right healing load differs from person to person.

For clinicians

Muscle force characteristics appear to drive tendon strain more than material stiffness, suggesting that monitoring individual force production capacity (for example via load-monitoring during exercise) may help guide exercise progression in Achilles tendinopathy rehabilitation.

For readers

This modelling study provides a ranked framework of six common Achilles tendon exercises from low to high strain, but highlights that population-level rankings do not reliably predict which exercise is optimal for any individual patient.