Relationship between structure and age in healthy Achilles tendons

Abstract

Age is an important factor to consider with Achilles tendon injury, as variability in tendon structure during developmental growth and aging influence lower limb function and mobility. However, the overlap in structural alterations with aging and Achilles tendon injury makes it unclear which structural changes are related to age separate from tendon pathology. The objective of this study was to determine the relationship between structure and age in healthy Achilles tendons. Healthy Achilles tendons from 389 children and adults (8–79 years) were included in this retrospective analysis. Achilles tendon morphology was assessed via B-mode ultrasound of Achilles tendon length, cross-sectional area (CSA), and thickness. Mechanical properties of Achilles tendon shear modulus and viscosity were assessed via continuous shear wave elastography. The relationship between Achilles tendon structure and age was determined using General Linear Models and White’s test of heteroscedasticity (to assess for unequal variance across the age span), controlling for sex, weight, and physical activity level. Healthy free Achilles tendon length (p=0.002), thickness (p<0.001), CSA (p<0.001), and viscosity (p=0.009) increased with age, supporting age-related changes in tendon structure that may limit its capacity to store and transfer energy in older adults. Full Achilles tendon length and CSA varied across the age span (p<0.05), suggesting the Achilles tendon undergoes natural aging processes seen with most musculoskeletal tissue. Normative data on Achilles tendon structure with age will contribute to our understanding and interpretation of Achilles tendon injury pathogenesis; aiding in the design of injury prevention and treatment strategies.

1. Introduction

Aging is a known risk factor for degenerative changes to the musculoskeletal system.^1,2 Yet, the relationship between age and Achilles tendon tissue structure remains largely unknown. While information on tendon growth during human maturation is scarce,³ the majority of Achilles tendon structural (i.e., morphology and mechanical properties) development seems to occur between the ages of 9 and 13^4–6 with low rates of cell renewal after the age of 17.⁷ Though presumably stable after adolescent growth, healthy tendon structure may vary with age in adulthood. Understanding how Achilles tendon structure varies with advancing age is crucial to distinguishing natural from pathological adaptations, as changes to Achilles tendon structure can alter tissue function^8,9 and may predispose individuals to tendon injury.

Tendons are viscoelastic tissue that exhibit both viscous (liquid resistance to motion) and elastic (resistance to permanent deformation) mechanical properties.¹⁰ In-vivo studies show that Achilles tendon stiffness (resistance to length change when force is applied) and water content reduces with older age,^1,11,12 creating a more compliant tendon with reduced functional capacity.^8,9,11,13 While there is inconclusive evidence for change in Achilles tendon CSA in older adults (65+ years),^1,14 accounting for physical activity level may nullify age-related effects on tendon size and stiffness.¹⁴ Achilles tendinopathy (a mechanical load-related tendon injury) is similarly associated with reduced in-vivo tendon stiffness and reduced lower limb functional performance,^15,16 but is marked by dramatic structural increases in CSA and water content.^16–18 Further, the incidence rate for Achilles tendinopathy increases with age, with the highest incidence occurring between 41–60 years in the general population¹⁹ and 35–64 years in the athletic population.²⁰

With such overlap in structural alterations between aging and Achilles tendon injury, it is unclear which changes are related to age separate from tendon pathology. Additionally, the lack of data from youth maturation stages makes it difficult to fully understand healthy Achilles tendon structural development from youth to older adulthood. No previous study has investigated Achilles tendon structure (morphology and mechanical properties) across child and adult age groups. Thus, the purpose of this study was to determine the relationship between structure and age in healthy Achilles tendons. It was hypothesized that (1) older age would be associated with lower Achilles tendon shear modulus (i.e., lower elasticity) and higher viscosity (i.e., lower water content), (2) Achilles tendon size (cross-sectional area and thickness) would have a positive relationship with age, and (3) Achilles tendon structure would vary across the age span.

2. Methods

2.1. Participants

This study was a retrospective analysis (Level III) of uninjured Achilles tendons from studies performed by the Delaware Tendon Research Group from November 2014 to June 2024, including an ongoing clinical trial (ClinicalTrials.gov identifier: NCT03523325) comparing the effect of exercise treatment for Achilles tendinopathy in women and men and a completed clinical trial (ClinicalTrials.gov identifier: NCT04816188) determining the feasibility of exercise therapy and pain-guided activity modification for children with heel pain. A full description of study populations is provided in Supplemental Table 1. All participants signed an informed consent, or child assent and parental permission, and study procedures were approved by the University of Delaware Institutional Review Board.

Participants in this study were healthy adults and children, as well as adults with unilateral Achilles tendinopathy and children with unilateral heel pain aged 7–65+. The right limb of healthy participants was used for analysis. The contralateral, uninjured limb of participants with unilateral Achilles tendinopathy was used for analysis. Achilles tendon health of each participant was assessed through palpation, single-leg hopping, and B-mode ultrasound imaging to rule out symptomatic and asymptomatic tendon injury.^21,22 Healthy Achilles tendon inclusion criteria included all the following: no reported pain on palpation, no reported pain during single-leg hopping, and tendon thickness less than 7 mm on B-mode ultrasound imaging. Thickness criterion was based on the average 5 mm thickness of healthy adult Achilles tendon²³ and greater than 2mm thickening indicating Achilles tendon pathology.²⁴

2.2. Procedures

Participants completed a survey for general demographics and the Physical Activity Scale (PAS)²⁵ or Patient-Reported Outcomes Measurement Information System (PROMIS) Pediatric Physical Activity Short Form²⁶ questionnaire. These validated questionnaires are designed to assess self-reported capability to perform physical activity. Child and adult physical activity level scores were adjusted to a 10-point scale due to differences in questionnaire scoring range (Adjustment: [original score - scale minimum / scale range] * maximum rescale score of 10). Participant height and weight were also measured using a standard scale.

2.2.1. Achilles Tendon Morphology

Achilles tendon morphology was assessed using B-mode ultrasound imaging acquired with a LOGIQ e ultrasound scanner (10MHz, 3.5 cm depth; GE Healthcare, Chicago, IL) and wideband linear array probe (5.0 – −13.0 MHz). Images of the full Achilles tendon length, free tendon length, thickness, and cross-sectional area (CSA) were taken and measured using previously described reliable procedures (Figure 1).^18,27,28 Briefly, participants were positioned prone with their feet in a relaxed position off the edge of an examination table. Long-axis extended field-of-view image of the full Achilles tendon length was measured from calcaneal insertion to gastrocnemius myotendinous junction. Long-axis extended field-of-view image of the free Achilles tendon length was measured from calcaneal insertion to soleus myotendinous junction. From the free tendon length image, Achilles tendon thickness was measured at approximately 2 cm proximal to the calcaneal insertion. Achilles tendon CSA was taken in short-axis view at the visually thickest portion of the tendon. Morphological measures were averaged across three images for analysis.

Figure 1.

Extended field-of-view long-axis image of the Achilles tendon (a) with full tendon length and free tendon length measures (dashed lines). Long-axis image of the Achilles tendon (b) with thickness measure (solid line) between tendon borders (dashed lines). Short-axis image of the Achilles tendon (c) with cross-sectional area measure (dashed line). Color maps of Achilles tendon shear modulus (d) and viscosity (e). MTJ=myotendinous junction.

2.2.2. Achilles Tendon Mechanical Properties

Achilles tendon mechanical properties were calculated using continuous shear wave elastography (cSWE), a validated technique with excellent reliability,^29–31 that utilizes a SonixMDP Q+ ultrasound scanner (Ultrasonix, Vancouver, Canada) with a L14–5/38 probe, external actuator (mini-shaker type 4810, Bruel & Kjaer, Norcross, GA), and 128-channel external data acquisition unit. Details of this technique have been previously described.^29,30 Briefly, participants were positioned prone with their feet secured at 10° of dorsiflexion against a platform to remove slack from the tendon. The external actuator propagated shear waves along the length of the Achilles tendon at eleven known frequencies, while the ultrasound probe recorded raw radiofrequency data. For healthy participants, the ultrasound probe was placed on the free Achilles tendon region, just distal to the soleus myotendinous junction. For injured participants, the ultrasound probe was placed on the uninjured tendon at the location corresponding to the thickest portion of the injured tendon. Custom MATLAB code (R2023b, Mathworks, Natick, MA) was used to select the tendon as the sole region of interest (ROI) then quantify Achilles tendon linear displacement and shear wave speed.^29,31 From the shear wave speed, Achilles tendon mechanical properties of shear modulus and viscosity throughout the ROI were estimated using the Voigt model (Figure 1).²⁹ For each participant, shear modulus and viscosity measures were averaged across the ROI of three images for analysis.

2.3. Statistical Analysis

The relationships between age, a continuous predictor, and measures of Achilles tendon structure were tested using General Linear Models (GLMs) and White’s test of heteroscedasticity (to assess for unequal variance across the age span), while controlling for study population (i.e., study sample), sex, height, weight, and physical activity level. GLMs with robust errors were used for all structural variables that violated the assumption of equal variance. Effect size was calculated for each predictor using Cohen’s f; interpreted as: 0.10 – small effect, 0.25 – medium effect, 0.40 – large effect.³² All model variables were screened for outliers. Alpha was set a priori at p<0.05. All statistical analyses were performed using SPSS (Version 25, IBM Corp., Armonk, NY).

3. Results

Participant demographics are included in Table 1. Descriptive statistics for structural measures are provided in Table 2. Full Achilles tendon length and CSA violated the assumption of equal variance across the age span (p<0.05; Figure 2), thus robust errors were used for analysis.

Table 1:

Participant demographics, expressed as mean±SD [min-max] and ratio.

	All (n=389)	Children (n=24)	Adults (n=365)
Sex (M:F)	183:206	15:9	168:197
Age (year)	37±18 [8–79]	13±3 [8–17]	39±17 [18–79]
Height (cm)	171±11 [129–201]	159±16 [129–187]	171±10 [140–201]
Body Mass (kg)	79±21 [27–156]	53±14 [27–80]	80±20 [44–156]
BMI (kg/m2)	27±6 [16–59]	20±3 [16–26]	27±6 [17–59]
Measured Limb (R:L)	282:107	21:3	261:104
Symptomatic Limb (R:L)	107:119	3:2	104:117
PAL (adj AU)	5±2 [0–8]	4±2 [0–8]	6±2 [0–8]
PROMIS PA (t-score)		51±8 [32–68]
PROMIS PA (pt)		13±4 [4–20]
PAS (AU)			4±1 [1–6]

n=sample size; BMI=body mass index; PAL (adj)=score-adjusted physical activity level; AU=arbitrary units;

PROMIS PA=Patient-Reported Outcomes Measurement Information System Pediatric Physical Activity Short

Form; PAS=Physical Activity Scale. Note: Symptomatic limb assessed only in participants with unilateral injury.

Table 2:

Achilles tendon structure, expressed as mean±SD and median (IQR).

	All (n=389)	Children (n=24)	Adults (n=365)
Full tendon length (cm)	20.7±2.5 20.8 (3.4)	20.9±3.1 20.2 (3.0)	20.7±2.4 20.8 (3.5)
Free tendon length (cm)	6.1±1.8 5.8 (2.4)	6.1±1.8 5.8 (2.2)	6.1±1.8 5.7 (2.5)
Thickness (cm)	0.5±0.1 0.5 (0.1)	0.4±0.1 0.4 (0.1)	0.5±0.1 0.5 (0.1)
CSA (cm2)	0.6±0.2 0.5 (0.2)	0.5±0.1 0.5 (0.2)	0.6±0.2 0.5 (0.2)
Shear modulus (kPa)	98.2±17.3 96.4 (23.3)	99.5±25.8 101.5 (35.1)	98.1±16.6 96.1 (22.1)
Viscosity (Pa*s)	56.0±11.4 56.0 (16.6)	51.6±9.0 51.6 (16.5)	56.3±11.5 56.2 (16.8)

IQR=interquartile range; n=sample size; CSA=cross-sectional area

Figure 2.

Relationship for age and Achilles tendon length (a), size (b), and mechanical properties (c) after controlling for study population, sex, height, weight, and physical activity level. Unadjusted individual values show the spread of tendon structure across the age span.

* significant relationship between structure and age (p<0.05)

# significant variability across age span (p<0.05)

Age was positively related to free Achilles tendon length (_adjR²=0.156, b=0.022, p=0.002, f=0.166), thickness (_adjR²=0.251, b=0.002, p<0.001, f=0.283), CSA (_adjR²=0.424, b=0.003, p<0.001, f=0.261), and viscosity (_adjR²=0.117, b=0.125, p=0.009, f=0.143) after controlling for study population, sex, height, weight, and physical activity level (Figure 2). Age was not related to full Achilles tendon length (_adjR²=0.440, b=0.006, p=0.448, f=0.041) or shear modulus (_adjR²=0.010, b=0.063, p=0.418, f=0.044). Analysis summaries are provided in Supplemental Tables 2–7.

3.1. Secondary Analysis

To identify structural relationships that may be specific to childhood or adulthood, a secondary analysis similar to the initial analysis was performed separately in children (<18 years old) and adults (>18 years old). All structural variables met the assumptions of equal variance across the child age span (p>0.05). Full Achilles tendon length and CSA violated the assumption of equal variance across the adult age span (p<0.05; Figure 3), thus robust errors were used for the analysis.

Figure 3.

Relationship for age and Achilles tendon length (a), size (b), and mechanical properties (c) after controlling for study population, sex, height, weight, and physical activity level in children (CH, under 18 years old) and adults (AD, over 18 years old).

* significant relationship between structure and age (p<0.05)

For children, age was negatively related to Achilles tendon CSA (_adjR²=0.558, b= −0.039, p=0.002, f=0.913) after controlling for study population, sex, height, weight, and physical activity level (Figure 3). Age was not related to full Achilles tendon length (_adjR²=0.592, b= −0.320, p=0.314, f=0.252), free Achilles tendon length (_adjR²=0.295, b= −0.342, p=0.163, f=0.354), thickness (_adjR²=0.119, b= −0.020, p=0.085, f=0.443), shear modulus (_adjR²=0.086, b= −3.826, p=0.369, f=0.224), or viscosity (_adjR²=0.041, b= −0.101, p=0.941, f=0.018).

For adults, age was positively related to free Achilles tendon length (_adjR²=0.168, b=0.022, p=0.003, f=0.164), thickness (_adjR²=0.245, b=0.002, p<0.001, f=0.299), CSA (_adjR²=0.427, b=0.003, p<0.001, f=0.277), and viscosity (_adjR²=0.108, b=0.125, p=0.011, f=0.144) after controlling for study population, sex, height, weight, and physical activity level (Figure 3). Age was not related to full Achilles tendon length (_adjR²=0.419, b= 0.005, p=0.504, f=0.038) or shear modulus (_adjR²=0.008, b=0.050, p=0.503, f=0.038). Secondary analysis summaries are provided in Supplemental Tables 8–19.

4. Discussion

This study evaluated the relationship between healthy Achilles tendon structure (morphology and mechanical properties) and age across the child to adult age span. In agreement with our hypothesis, findings revealed that healthy free Achilles tendon length, thickness, CSA, and viscosity increased with age after adjusting for sex, height, weight, and physical activity level. Full Achilles tendon length and CSA were observed to have unequal variance across the child to adult age span; suggesting that these structural measures vary with age. Secondary analysis revealed that adult aging may drive the current observations, as healthy Achilles tendon CSA decreased with age in children with no other significant structure-age relationships nor structural variability across the child age span. These normative data will aid in understanding changes in Achilles tendon structure that are related to aging separate from tendon pathology and contribute to further exploration of Achilles tendon developmental growth.

4.1. Association of Achilles Tendon Structure and Age

The observed increases in tendon size and viscosity support an age-related change in tendon structure that may alter its capacity to store and transfer energy in older adults. Older age was associated with longer free Achilles tendon length (insertion to soleus), greater thickness, greater CSA, and higher viscosity. While there are few human models of tendon morphological changes with age to compare findings, Onambele et al.³³ found a reduction in full Achilles tendon length in middle-aged and older adults (~46 and ~68 years, respectively) compared to younger adults (~24 years). In contrast, no significant association of full Achilles tendon length (insertion to gastrocnemius) and age were observed in the current study. The difference in findings may be due to the adjustment for participant sex in the current dataset. There could have also been discrepancies in tendon length measurement, as the Onambele et al.³³ study was unclear on its protocol. No study has shown an increase in free Achilles tendon length with age. This tendon lengthening may be caused by soleus muscle atrophy^34,35 and/or changes in resting ankle angle from altered muscle tone of the triceps surae with older age, though further investigation is warranted.

Our finding that tendon CSA increased with age is consistent with prior observations that the Achilles tendon increases in size with age.^{4–6,8,11,33,36} An increase in tendon size (mainly CSA) is a positive adaptation that increases mechanical loading capacity in developing tendons.⁶ The decrease in tendon size with older childhood age is contrary to previous observations by Mogi et al.^4,5 of Achilles tendon CSA increase during early adolescence (9–13 years), with no significant change to CSA in late adolescence (14–17 years); though the current study child sample size was small (n=24). In older adults, tendon size increase may be a compensatory mechanism in response to increased body mass (Table 1), reduced material/mechanical properties,^11,37,38 metabolic disorders, or other systemic conditions.^39,40 Future studies are needed to determine the variable growth pattern of Achilles tendon CSA in children and if increased tendon CSA in adulthood increases the risk of Achilles tendinopathy. As CSA had the strongest relationship to age, our results support the inclusion of this structural measure in ongoing tendon aging research.

The current findings demonstrate that healthy Achilles tendon viscosity increases with older age, as has been previously shown in adults.¹² Greater viscosity is likely due to decreased water content and greater collagen fiber cross-linking in response to advanced glycation end-product (AGE) accumulation in tendon with adult age.^1,41 The increased collagen cross-linking impairs tendon fiber movement when stretched and reduces water content in the tendon,^39,41 as the cross-links dehydrate collagen.^37,42 Reduced water content could also be attributed to a decline in proteoglycans in the ECM of older tendon,³⁷ but these findings have only been reported in animals. Conversely, Achilles tendinopathy is marked by a decrease in viscosity,^16,17 implying greater water content in the tendon. Thus, increased tendon water content may be a distinct structural adaptation of Achilles tendon pathology separate from aging.

Though shear modulus was not significantly associated with age in the current study, previous literature has found that older age relates to decreased stiffness and Young’s modulus when assessing the Achilles tendon during active isometric and ramped contraction.^8,9,11,36,38 However, as muscle strength decreases with age, it may be difficult to separate the effects of reduced contraction force from tendon properties.^14,37,38 Continuous shear wave elastography was used in the current study to measure shear modulus and viscosity in a passive state.^29,30 This method does not rely on muscle contraction, and thus, removes a confounding variable of assessing in-vivo Achilles tendon mechanical properties.

Adjusting for physical activity level did not nullify the age-related increases in healthy Achilles tendon size and viscosity, suggesting that aging still plays a role in Achilles tendon adaptations despite controlling for external factors like physical activity. However, as the adjusted models accounted for up to 42% and 12% of the variance in tendon size and viscosity, respectively, it is possible that these observed structural adaptations were influenced by changes in unmeasured innate factors such as hormones and metabolism during pediatric growth and adult aging. Changes in hormone levels have been shown to alter collagen synthesis in both males and females, resulting in changes to tendon mechanical properties.⁴³ Metabolic disorders of hyperglycemia and metabolic syndrome can cause low-grade microvascular inflammation and increased AGE accumulation in the tendon, which disrupt tendon homeostasis and impede tendon repair processes.^39,40

4.2. Variability in Achilles Tendon Structure with Age

From the observed structural changes with age, the Achilles tendon appears to follow a natural aging process seen with most musculoskeletal tissue. Healthy full Achilles tendon length and CSA varied across the age span, particularly in the older age range. Though previous data on the pediatric growth pattern of tendons suggests the tendon increases in size and stiffness to compensate for the rise in mechanical load demand during adolescent musculoskeletal development,^5,6,44 the current results for Achilles tendon adaptations indicated no variable growth during childhood. Further research with larger sample sizes is warranted to establish healthy Achilles tendon growth patterns in the pediatric population.

While tendon structure is presumed to be stable after puberty, our results indicate that changes associated with decreased tendon function,^8,9,11,13 such as reduced stiffness,^1,11,12 occur in adulthood. Older individuals had more variability in Achilles tendon CSA, with less variability in full Achilles tendon length. New molecular research shows that substantial cellular dysregulation occurs in humans at 44 and 60 years of age,⁴⁵ aligning with the current results and supporting a non-linear aging process with unique timepoints of variation. The variability in Achilles tendon structure observed in older individuals also aligns with the years associated with greater incidence of Achilles tendinopathy (41–60 years),¹⁹ suggesting that variation in Achilles tendon size and length with aging may be a risk factor for injury.

4.3. Limitations

Though the normative data on Achilles tendon structure from this study valuably contribute to the literature, there were limitations that should be considered when interpreting these findings. The current dataset came from several studies. Standard procedures and measures were used across studies with researchers trained to be reliable on all measures to reduce measurement and/or procedural errors. Statistically controlling for study population allowed the relation of age and Achilles tendon structure to be assessed with reduced bias. Yet, analyses revealed that differences in study population significantly influenced structural measures (Supplemental Tables 2–5). Thus, most relationships between age and Achilles tendon structure were also dependent on the type of participants included in each study. While participants were screened for Achilles tendon symptoms and structural pathology, it is possible that subclinical changes in Achilles tendon structure went undetected with the current ultrasound techniques and may have influenced outcomes measures. As discussed, our secondary analysis was limited by a small sample size of children (n=24), which made it difficult to draw strong conclusions on Achilles tendon developmental growth. Lastly, tendon mechanical properties were calculated using the Voight model which assumes linear shear stress. As tendon is a non-linear tissue, mechanical properties vary based on the direction of applied load. However, this non-linear effect is presumed negligible with small shear deformation and when shear waves are sent along a main anisotropy axis (i.e., parallel to tendon fibers) to maintain a proportional relationship between wave speed and shear modulus.⁴⁶

5. Conclusion

This was the first study to show the relationship between healthy Achilles tendon structure and age across the child to adult age span. The increase in tendon size and viscosity support age-related changes in Achilles tendon structure that may limit its capacity to store and transfer energy in older adults. Healthy Achilles tendon morphology varied across the age span, particularly in the older age range, suggesting the Achilles tendon undergoes natural aging processes seen with most musculoskeletal tissue. Normative data on tendon structure with age will contribute to our understanding and interpretation of Achilles tendon injury pathogenesis; aiding in the design of injury prevention and treatment strategies for Achilles tendon injuries across the lifespan.