Interfibrillar shear behavior is altered in aging tendon fascicles

Abstract

Tendon elongation involves both stretching and sliding between adjacent fascicles and fibers. Hence, age-related changes in tendon matrix properties may alter sliding behavior and thereby affect injury thresholds. The objective of this study was to investigate the effects of age on interfibrillar shear behavior in partial cut tendon fascicles. Cine microscopic imaging was used to track deformation patterns of intact and partial cut fascicles from mature (9 months, n=10) and aged (32 months, n=10) rat tail tendons. Finite element (FE) models coupled with experimental data provided insight into age-related changes in tissue constitutive properties that could give rise to age-dependent behavior. Intact fascicles from aged tendons exhibited a 28% lower linear region modulus and reduced toe region when compared to fascicles from mature tendons. Partial cut tendon fascicles consistently exhibited a shearing plane that extended longitudinally from the tip of the cut. Both mature and aged fascicles exhibited distinct failure that was observable in differential displacement across the shearing plane. However, aged fascicles exhibited 11–20% higher grip-to-grip strain at failure and tended to exhibit more variable and greater differential displacement at failure, when compared to mature fascicles. FE models suggest that this age-related change in shear behavior arises from a reduction in interfibrillar shear modulus with age. These data suggest that aging alters interfibrillar failure mechanisms and hence may contribute to the increased propensity for injury that is commonly seen in older tendons.

1. Introduction

Tendon’s composition and structure enable it to transmit forces generated by muscles to the bony skeleton. Tendon’s ability to stretch and recoil during load-bearing can be attributed to its hierarchy of longitudinally-aligned collagenous structures, ranging from microfibrils on the nanoscale to fascicles on the microscale (Handsfield et al. 2016). The interaction between these structures are regulated by the proteoglycan-rich, non-collagenous matrix between hierarchical levels (Kastelic et al. 1978), which maintains structural integrity at each level. At the fascicle level, the interfascicular matrix enables sliding between adjacent collagen fascicles (Thorpe et al. 2015b). Likewise at the fiber level, collagen fibers both stretch and slide relative to each other to produce overall fascicle elongation (Screen et al. 2004; Cheng and Screen 2007). Mechanics at this level are of particular interest because tenocytes, the mechanosensitive cells of tendon, are situated in rows aligned between adjacent fibers (Kjaer 2004; Kjaer et al. 2009). The unique interplay between fibers, matrix, and tenocytes continuously evolves throughout an individual’s lifespan (Lavagnino et al. 2013).

Partial cut tendon fascicles have been effectively used to probe interfibrillar mechanics. One of the most intriguing observations is the distinct emergence of a shearing plane with loading that extends longitudinally from the tip of the partial cut (Szczesny et al. 2015). Interfibrillar shear failure occurs at relatively low loads, with the failure shear stresses estimated as 32 kPa in rat tail fascicles (Szczesny et al. 2015). Interfibrillar shear failure has negative implications on tenocyte viability. While tenocytes have the capacity to withstand normal physiological ranges of shear loading, abnormal or excessive loading causes alterations in shape, necrosis, or apoptosis (Provenzano et al. 2002; Scott et al. 2005; Thorpe et al. 2015a). A previous study of partial cut fascicles discovered evidence of extensive tenocyte death along the shear plane far from the partial cut (Kondratko-Mittnacht et al. 2015). Hence, exaggerated interfibrillar sliding has the potential to induce localized damage that can alter tendon remodeling and overall tendon behavior.

Aging is associated with changes in tendon structural and mechanical properties at multiple scales. One notable structural change is increased advanced glycation end-products (AGEs), which accompanies a reduction in tendon collagen content and turnover (Haut et al. 1992; Heinemeier et al. 2013; Svensson et al. 2016). By cross-linking at the molecular level, AGEs impart added stability and strength to collagen fibrils, thereby reducing collagen fiber sliding and increasing collagen fiber stretch in a compensatory manner (Li et al. 2013). Further, an accumulation of microdamage could compromise the non-collagenous matrix and alter sliding behavior at multiple scales. Both the interfascicular matrix and fascicles from aging horses exhibited reduced fatigue life and greater matrix disruption with repetitive cyclic loading (Thorpe et al. 2014). Structural changes likely are reflected in measured macroscale mechanical properties. For example, aging human tendon exhibited a decreased elastic modulus and strength relative to younger adult tendons (Kubo et al. 2003; Onambele et al. 2006; Stenroth et al. 2012). Similarly, energy-storing tendons exhibited less sliding with aging at the fascicular interface at physiological loads (Thorpe et al. 2012b). However, implications of aging on failure at the interfibrillar length scale are not well understood, but are potentially relevant for understanding increased injury rates in aged tendons (Clayton and Court-Brown 2008; Hess 2010; Albers et al. 2016).

The objective of this study was to investigate the effects of age on interfibrillar shear behavior in partial cut tendon fascicles. This objective was achieved by comparing deformation patterns in partial cut fascicles from both mature and aged rat tail tendon specimens. Finite element (FE) models were used to provide insight into underlying material properties that could give rise to age-dependent behavior.

2. Methods

2.1. Specimen Preparation

A single ventral tail tendon was dissected from ten mature (9-month) and ten aged (32-month) male F344xBN rats. Animals were euthanized in a separate IACUC-approved study. Two fascicles were teased from each tendon and separated into a partial cut group and an intact group. For fascicles in the partial cut group, a 4.5× dissecting microscope (AmScope, Irvine, CA) and scalpel were used to introduce a mid-substance, partial-thickness cut into the fascicle. For the intact group, no cuts were made.

A grid was photobleached on the fascicles to visualize tissue deformation during loading. To do this, fascicles were stained in 56 μM (28 μg/ml) fluorescent stain (5-dichlorotriazinyl aminofluorescein) for 30 minutes. Fascicles were subsequently rinsed in 1 mL of PBS for 5 minutes to remove any unbound dye. A microgrid mesh with 38 μm square openings (TWP Inc., Berkley, CA) was glued to a glass microscope slide. This grid pattern size was chosen to ensure deformation patterns would be resolvable over the width of the fascicle. Fascicles were overlaid on the mesh for photobleaching on an epifluorescent, inverted microscope (Olympus IX-71, Tokyo, Japan) for 30 minutes, while hydration was maintained with PBS. After photobleaching was complete, fascicles were removed from the glass slide and placed into custom, 3D-printed grips.

2.2. Mechanical Testing

Fascicles were placed into a custom-made uniaxial testing device (Figure 1) mounted on a microscope stage and immersed in a PBS-filled bath. Throughout testing, images were acquired using an epifluorescent, inverted microscope, with a yellow-green filter cube (488/516 nm Sedat Laser Filter Set, Semrock, IDEX Health & Science, Rochester, NY) and a metal halide lamp (X-CITE, Excelitas, Waltham, MA). Two orthogonal snapshots were taken with the major axis in both a horizontal and vertical orientation by rotating the grips 90 degrees. Edges of the fascicle were identified in each orientation along the entire 3.5 mm field of view and averaged to produce fascicle width measurements using custom MATLAB code. Specifically, fascicle edges were identified by locating where image intensity transitioned from below to above 50% of the statistical intensity range of the entire image. Cross-sectional area was then calculated by representing the transverse cross-section as an ellipse. Cut depth was also assessed from the pre-testing image of the major axis using ImageJ (National Institutes of Health; Bethesda, MD).

Fig. 1.

Uniaxial tensile loading device for inverted microscope stage. Fascicles were imaged from beneath the stage while fascicle hydration was maintained with PBS

For mechanical testing, grips were secured to a fixed 44.5 N (10 lb) load cell with 3.11 × 10⁻⁴ N (7 × 10⁻⁵ lb) resolution (Futek LSB210, Irvine, CA) and linear actuator with 7 μm accuracy (Aerotech ACT165DL, Madison, WI). When samples were positioned into the grips, grips were less than 20 mm apart. To establish a consistent reference configuration, grip-to-grip length was measured with a low (148.6 ± 76.2 kPa) preload applied. Intact fascicles were preconditioned for 10 cycles to 4% grip-to-grip strain at 0.5 Hz. There was no visible deformation to the photobleached grid as a result of preconditioning intact fascicles. Partial cut fascicles were not preconditioned to preserve the resultant grid deformations surrounding the cut. All fascicles were subsequently stretched to 4% grip-to-grip strain at 0.01 s⁻¹, and the central 3.5 mm of the fascicle was imaged at 15 frames per second at 4×. By imaging the central portion of each fascicle, the field of view remained sufficiently far from the grips such that end effects were negligible. This testing procedure resulted in 2.54 μm per pixel resolution throughout testing. Load and displacement data were concurrently recorded at 100 Hz.

2.3. Image Analysis

Speckle tracking was used to measure axial tissue displacements from the cine microscopic images (Chernak Slane and Thelen 2014). To do this, MATLAB was used to re-orient images by aligning the bottom edge of the fascicle with the horizontal axis of the images. Images were then cropped to the top and bottom fascicle edges at each frame. A rectangular array of nodes, spaced 79 μm horizontally and 28 μm vertically, was then positioned on the unloaded fascicle image. This resulted in 42 nodes along the fascicle length in the field of view and 9–19 nodes across the fascicle width, depending on the specimen dimensions. Rectangular regions of interest (ROI, 31 μm by 11 μm) were centered on each node. Nodal displacements between successive images were computed by finding the axial displacement that maximized the normalized cross-correlation of the ROI in one image with an ROI in the subsequent image (Chernak Slane and Thelen 2014). Sub-pixel displacement was obtained by using a cosine fit to the normalized cross-correlation function (Céspedes et al. 1995). Net nodal displacements were obtained by accumulating the measured motion across all images in a sequence.

Local fascicle strain was ascertained from the nodal displacement data. To do this, we first computed median nodal displacements of each column across the intact fascicle or across the intact portion of the partial cut fascicles. Local tissue strain was computed by linearly regressing the nodal displacements to the original nodal positions. Individual specimens achieved varying magnitudes of local stretch at 4% grip-to-grip strain. When calculating group statistics such as mean stress-stretch curves for each age group, analysis was conducted up to the stretch magnitude achieved by 70% of the specimens (1.0163 μm/μm). First Piola-Kirchhoff stress was calculated from the measured load divided by the reference cross-sectional area.

In partial cut fascicles, a shearing plane emerged with fascicle stretch. The visible shearing plane extended longitudinally from the tip of the cut (Figure 2, inset), so data were binned spatially as a function of distance from the cut. The location of the shearing plane was identified within each column of nodes by finding the successive nodes that exhibited the greatest relative motion. Shear deformation was then quantified at each frame by the differential displacement across the shearing plane. For this measure, the difference in displacements was taken between a span of four nodes (i.e., 112 μm) across the shearing plane. Reported differential displacements were binned by initial distance from the cut: 0–1 mm, 1–2 mm, and 2–3 mm. These were plotted as a function of grip-to-grip strain (Figure 2). An inflection point identification algorithm (Satopaa et al. 2011) was used to identify the frame of failure, defined as the frame at which the slope of the differential displacement abruptly increased. We then extracted the corresponding applied strain and differential displacement at failure within each bin.

Fig. 2.

Differential displacement by bin plotted as a function of grip-to-grip strain for a representative mature fascicle. Plotted lines are median curves, and the shaded regions are standard deviation. Differential displacement was calculated by taking the difference in absolute displacement between a span of 4 nodes across the shearing plane. Failure in each bin, denoted by a star, was defined as the inflection point of the differential displacement curve

2.4. Model Fitting and Finite Element Modeling

We modeled fascicles as a transversely isotropic, hyperelastic material (Quapp and Weiss 1997). We estimated constitutive model parameters using both intact fascicle stress versus local stretch behavior and partial cut fascicle differential displacement. A Poisson’s ratio of 0.2 was assumed (Safa et al. 2019). Sensitivity studies showed that results were insensitive to Poisson’s ratio. The deformation gradient was obtained from measured local stretch and assumed Poisson’s ratio because of the small strains at failure. Cauchy stress was calculated from the deformation gradient and first Piola-Kirchhoff stress (Holzapfel 2000). Mean stress-stretch data for each age group were fit to a transversely isotropic material model, with fibers assumed to only resist tension. The inter-fibrillar matrix was modeled as a neo-Hookean solid. Fibers were modeled with a toe region and a linear region. Cauchy stress along the loading axis resulted from the addition of the interfibrillar matrix and fibers:

$$\sigma =\{\begin{array}{ll}\hfill \raisebox{1ex}{$\mu $}\!\left/ \!\raisebox{-1ex}{$J$}\right.\left({\lambda }^2-1\right)+\raisebox{1ex}{${\lambda }_l$}\!\left/ \!\raisebox{-1ex}{$J$}\right.\text{ln}\left(J\right)\hfill & \text{ for }\lambda <1\hfill \\ \hfill \raisebox{1ex}{$\mu $}\!\left/ \!\raisebox{-1ex}{$J$}\right.\left({\lambda }^2-1\right)+\raisebox{1ex}{${\lambda }_l$}\!\left/ \!\raisebox{-1ex}{$J$}\right.\text{ln}\left(J\right)+c_3\left\{\text{exp}\left[c_4(\lambda -1)\right]-1\right\}\hfill & \text{ for }1\le \lambda <{\lambda }^{*}\hfill \\ \hfill \raisebox{1ex}{$\mu $}\!\left/ \!\raisebox{-1ex}{$J$}\right.\left({\lambda }^2-1\right)+\raisebox{1ex}{${\lambda }_l$}\!\left/ \!\raisebox{-1ex}{$J$}\right.\text{ln}\left(J\right)+c_5\lambda +c_6\hfill & \text{ for }{\lambda }^{*}\le \lambda \hfill \end{array}$$ (1)

Where σ is the Cauchy stress,μ is the shear modulus (and Lamé coefficient),J is the Jacobian (volume ratio),λ is the local (gage) stretch, λ_l is the second Lamé coefficient, λ* is the transition stretch between the toe and linear regions, c₃ and c₄ define the toe region behavior, c₅ is the linear region modulus c₆ and is the intercept (determined by enforcing continuity at λ*). A series of fits with various assumed shear moduli were made to the intact data using Eq. (1). First, the transition stretch was determined by identifying the point at which the stress-stretch curve deviated by greater than 10% from the linear fit of the final 5% of the curve. The toe and linear regions were then fit separately using MATLAB nonlinear least squares fitting.

Representative FE models of both mature and aged fascicles were generated to estimate the shear modulus. Plane strain, axisymmetric FE models were generated using mean sample dimensions and cut depths, with one model generated for the mature and one for the aged fascicles. Models were 20 mm long and were analyzed up to the average grip-to-grip strain at failure. Models were analyzed with fits to intact data that included different shear moduli. FE-predicted differential displacement in the 1–2 mm bin was calculated. The shear modulus that provided FE predictions matched to experimental results was taken as the mean shear modulus for that age group. Resulting FE-predicted shear stress maps were qualitatively evaluated. All models were analyzed in FEBio version 2.8 (Maas et al. 2012).

2.5. Statistics

Age-dependent differences in applied strain at failure and differential displacement at failure were analyzed using a mixed-factor ANOVA. Differences between age groups within each bin and between bins within each age group were tested (significant at p ≤ 0.05). Additionally, linear region modulus from constitutive model fitting was compared between age groups using a two-sample t-test (significant at p ≤ 0.05). Beyond linear modulus, age-dependent material behavior was not statistically compared because fits were made to mean data. All statistical tests were conducted in MATLAB (R2019a; MathWorks, Natick, MA).

3. Results

Rat weights significantly increased with age from 400.9 ± 19.0 grams to 463.1 ± 23.9 grams (p < 0.00001). However, there was no difference in fascicle cross-sectional area between fascicles from the two age groups (mature, 0.0865 ± 0.0368 mm²; aged, 0.0815 ± 0.0209 mm²; p = 0.7126). Grip-to-grip length of the taut fascicles did not differ between groups and averaged 19.72 ± 0.81 mm. Partial cut depths also did not significantly differ between age groups (mature, 34.5 ± 14.3% of fascicle width; aged, 40.6 ± 11.0% of fascicle width; p = 0.2971).

3.1. Intact Fascicle Model Fitting

Local stretch was substantially attenuated from grip-to-grip stretch (1.04 μm/μm) and did not exceed 1.02 μm/μm in all but one fascicle. Stress-stretch behavior and material constants were different for the two ages (Figure 3, Table 1), with the aged fascicles exhibiting greater compliance. When fitting individual specimen stress-stretch curves, there was a significant decrease in linear region modulus (c₅) with age from 1249 ± 330 MPa to 865 ± 346 MPa (p = 0.0204). In aged specimens, the toe region was also less distinct (Figure 3), which resulted in a lower exponential constant (c₄).

Fig. 3.

Group mean experimental Cauchy stress versus local stretch curves (triangles) were fit to a transversely isotropic, hyperelastic constitutive model (Eq. (1)) (lines). Error bars show experimental standard deviation

Table 1.

Fits to group average stress–stretch curves.

	λ*(-) Transition Stretch	c3 (MPa) Toe Region Scaling Constant	c4 (-) Toe Region Exponent Constant	c5 (MPa) Linear Region Modulus
Mature	1.0062	1.89	201.69	1219.16
Aged	1.0057	3.95	112.30	877.55

3.2. Partial Cut Fascicle Mechanics

A shearing plane emerged with fascicle stretch, creating distinct displacement and local strain patterns between the intact and cut portions of the partial cut fascicles (Figure 2, inset). Differential displacement in each bin across the shear plane initially grew slowly with axial elongation, before increasing abruptly at the defined failure point. The point of inflection in differential displacement, defined as the failure point in each bin, thus occurred at a different grip-to-grip strain in each bin, but this difference was not statistically significant (Figure 4a).

Fig. 4.

Boxplots of a) grip-to-grip strain at failure for failure defined in each bin and b) differential displacement between nodes spaced 112 μm apart at failure for failure defined in each bin binned by horizontal distance from the partial cut (* p < 0.05)

Mixed-factor ANOVA indicated there were significant differences between age groups (p = 0.0347). Aged fascicles exhibited significantly larger grip-to-grip strain at failure than mature fascicles in the regions closest to the cut (0–1 and 1–2 mm bins). Differential displacement at failure in aged specimens was more variable and tended to be larger (p = 0.1578) than that seen in the mature tendons (Figure 4b).

3.3. Non-fibrillar Matrix Behavior and Finite Element Modeling

A shear modulus,μ, of 3 MPa was sufficient for the FE model to replicate the average differential displacements in the 1–2 mm bin across the shear plane seen in the mature fascicles (Figure 5). This provided FE predictions within one standard deviation of experimental data for all bins (Supplementary Figure 1). A 17% lower shear modulus (2.5 MPa) was needed to predict the differential displacements in the 1–2 mm bin seen in the aged fascicles. This distinction resulted in the cut-induced shear stresses extending across a slightly larger span of the aged fascicles (Figure 6).

Fig. 5.

FE predictions (lines) for differential displacement up to failure in the 1–2 mm bin and experimental differential displacements in the 1–2 mm bin (triangles) demonstrate agreement, as this bin was used to fit the shear modulus. Error bars show experimental standard deviation. Inset FE models show displacement at failure up to 3 mm from the cut, with the dashed lines indicating the shearing plane. Variations in differential displacement patterns at failure are visible

Fig. 6.

In-plane FE-predicted shear stress in mature and aged fascicles. Partial cuts can be seen in the upper right of each model and dashed lines indicate the shearing plane. Top images show up to 8 mm from the cut, while bottom images show only 3 mm from the cut. Differences in shear stress pattern at failure are visible

4. Discussion

The objective of this study was to directly compare axial and shearing deformation patterns between mature and aged rat tail tendon fascicles. We observed age-related effects in both intact and partial cut tendons. Notably, intact aged fascicles exhibited a less pronounced toe region and increased compliance compared to mature fascicles. While coefficients for the intact toe region were different between age groups (Table 1), the response in the toe region was qualitatively similar (Figure 3). Conversely, the linear region was more compliant in the aged than mature fascicles. This suggests that the less pronounced transition between the toe and linear regions in aged fascicle may result primarily from the low linear region modulus in aged fascicles. Partial cut aged fascicles required greater elongation and exhibited more variable shear deformations at failure. Taken together, these results suggest that age-related changes in interfibrillar matrix likely alter both the macro scale tissue behavior and localized damage thresholds.

Age-related changes in intact fascicle behavior are generally consistent with prior studies. Aged fascicles were more compliant, exhibiting a 28% drop in the linear modulus. This effect has been seen at multiple levels of the tendon hierarchy (Kubo et al. 2003; Onambele et al. 2006; Thorpe et al. 2012b; Stenroth et al. 2012). Increased compliance in the linear region of aged fascicles likely resulted from the combined effects of both more compliant collagen fibrils and a lower interfibrillar shear modulus. Prior shear lag modeling suggests that fibril behavior and dimensions as well as interfibrillar matrix behavior contribute to the fascicle-level response (Ahmadzadeh et al. 2013). Further, it is known that age exaggerates the loss of functional integrity to the extracellular matrix, which predisposes tendons to injury (Dudhia et al. 2007). Numerous studies have also shown that it is the degeneration and limited repair of the tendon matrix that leads to decreased performance in older tendon (Dudhia et al. 2007; Yu et al. 2013; Ackerman et al. 2017). It is possible this degeneration and reduction in collagen content counters the expected increase in modulus from increased AGEs seen with aging. Specifically, a previous experimental model induced cross-linking and showed that this resulted in reduced collagen fiber sliding and increased collagen fiber stretch (Li et al. 2013); that our results showed increased sliding in aged fascicles strongly suggests that AGEs are either not present or are damaged. Aged fascicles also demonstrated a less distinct toe region, as evidenced by a lower c₄ value (Figure 3, Table 1). This mechanical change thus requires greater relative forces for fascicle extension during the initial portion of the stress-stretch curve. Similar results were found in a previous study of energy-storing equine tendons, which experienced greater relative loading at the interfascicular interface at an earlier point during tendon extension than seen in younger tendons (Thorpe et al. 2012b). While it is challenging to directly relate aging in animal models to that in humans, the approximate corresponding human ages for the mature and aged rats were 25 years old and 80 years old, respectively (Sengupta 2013). Thus, the fascicles in this study can be considered relevant to age-related tendinopathy in humans.

A partial cut induced exaggerated interfibrillar shearing with axial stretch. Consistent with others’ research, shear failure was observable at relatively low axial grip-to-grip strains (<3%, Figure 4a) (Szczesny et al. 2015). The corresponding failure shear stress has previously been estimated to be 32 kPa (Szczesny et al. 2015). Using a similar equilibrium-based approach to the previous study, failure shear stress magnitudes from this study were estimated as averaging 38.2 kPa and 35.2 kPa in mature and aged fascicles, respectively. However, while shear stress magnitudes were similar for the two groups, greater axial stretch was required to induce shear failure in the aged fascicles. This distinction likely results from the greater compliance exhibited by aged fascicles. Relative to mature fascicles, a 28% decline in linear region axial modulus and a 17% decline in shear modulus was needed to replicate experimental results. These two adaptations altered shear stress patterns near the cut at failure. In particular, the aged fascicle FE model predicted larger shear stress magnitudes near the cut and extended further along the fascicle length (Figure 5).

Mechanobiological responses are important to consider when evaluating the effects of interfibrillar shearing in fascicles. Previous research has demonstrated that tenocyte viability is compromised along the shear plane that emerges in partial cut fascicles (Kondratko-Mittnacht et al. 2015). Cell death at regions remote from the localized damage can adversely affect remodeling and repair of the tissue. Although cellular response was not observed in this study, it is likely that exaggerated deformations seen in aged specimens induce altered mechanobiological responses (Thampatty and Wang 2018) that can contribute to the degeneration and accumulated microdamage often seen in older tendons (Riley 2008; Thorpe et al. 2010).

A few limitations to the study should be acknowledged. First, the helical nature of fibers within tendon fascicles (Thorpe et al. 2013) can induce rotation with axial stretch. Our one-dimensional tracking method was not able to account for this. Additionally, variation was seen when comparing fascicle cross-sectional area from our study to previous reports for rat tail tendon fascicle cross-sectional area (Legerlotz et al. 2014; Szczesny et al. 2015; Lee and Elliott 2019). Differences were likely influenced by choice of rat breed, fascicle edge-detecting algorithm, and imaging technique. While our cross-sectional area measurements were lower than previously reported, the proportional drop in linear modulus (i.e., 28% reduction) between mature and aged remains comparable to past reports (Thorpe et al. 2012b). The measured mechanics may have been affected by the use of DTAF to visualize deformations. While the concentration used in this study was not exceedingly high, it is in the range that was previously established as potentially impacting fascicle mechanics (Szczesny et al. 2014). This study was conducted exclusively with rat tail tendon fascicles. It is known that interfascicular differences differentiate between energy-storing and positional tendon types (Thorpe et al. 2012a). For example, age-related microstructural strain changes were seen strictly in energy-storing tendons, while positional tendons displayed no significant changes. While this may be due to differences in stretch mechanisms, there was recently a direct comparison suggesting that both tendon types exhibit similar interfibrillar sliding mechanics (Lee and Elliott 2019). Finally, intact fascicles were preconditioned, while cut fascicles were not. Preconditioning was done to obtain a repeatable result in intact fascicles, but was not done in cut fascicles because failure occurred on the first load. This may have resulted in variation in fascicle behavior between the cut and intact cases, which are not reflected in our approach to fitting material behavior.

In summary, this study provides insight into changes in axial and shearing mechanics in the aging tendon fascicle. Aged specimens exhibited greater axial and shear compliance that resulted in exaggerated and more variable interfibrillar deformations at failure. These mechanical adaptations likely reflect compromised integrity of the interfibrillar matrix, which could be important to consider when developing tissue constructs and treatments for tendon injury in older individuals.