The purpose of the current study was to evaluate the relationship between exogenous testosterone administration and Achilles tendon injury. Our results demonstrate that, when compared to a matched control cohort, the rate of Achilles tendon injury and subsequent surgery was significantly higher in patients prescribed TRT. Furthermore, this trend was found in both males and females even after statistical control for several medical comorbidities, most notably hypogonadism. These results demonstrate an important correlation between testosterone use and Achilles tendon injury and point to future clinical implications of testosterone prescription and usage. Much of prior literature has focused on the effects of anabolic steroids on tendon structure and health, but this study is the first to retrospectively analyze rates of Achilles tendon injuries and subsequent surgery in patients prescribed TRT.

In the Achilles tendon, testosterone has been linked to increased tendon stiffness, a reduced relaxin response, and inhibited matrix metalloproteinase (MMPs) activity, all of which negatively affect tendon integrity after repeated strain [21, 22]. At a molecular level, tendon health is maintained by a balance of MMPs and tissue inhibitors of matrix metalloproteinases (TIMPs) that is crucial for tissue healing, and the use of exogenous steroids has been demonstrated to significantly alter this balance, leading to collagen dysplasia in rats and a consequent increase in Achilles tendon injury [23,24,25]. In multiple studies on human AAS users, increased rates of tendon rupture were discovered compared to non-AAS users [12, 13, 26]. Multiple case reports on rare tendon avulsions and ruptures have also linked AAS usage of the patients to the cause of the tendon injury [27,28,29]. In the context of testosterone’s biochemical effects on tendon health and structure and its rising usage and prescription, it imperative to better understand the musculoskeletal complications associated with exogenous testosterone in order to minimize patient harm.

This study reports an increased odds of 1.38 in Achilles tendon injury in males and an increased odds of 1.44 in females with a minimum of three consecutive months of prescription TRT. While the current body of literature regarding prescription TRT and musculotendinous injury is scarce, a recent study by Testa et al. in a similar analysis demonstrated a significant increase in rotator cuff tears and revision rotator cuff repair [14]. A second similar study published by Smith et al. evaluated rates of rotator cuff repair (RCR) in patients that were sex-hormone deficient and they found an 89% increase in rates of RCR in males who were testosterone deficient and a 48% increase in estrogen-deficient females [30]. Although Smith et al. do not account for testosterone supplementation in sex-hormone-deficient patients, their study does allude to the complex relationship between sex-hormone derangements and musculotendinous health, which is in accordance with the present study. The present study builds on this literature by adjusting for previously diagnosed hypogonadism and other comorbidities associated with musculoskeletal injury when evaluating and comparing rates of Achilles tendon injury and subsequent surgery in patients prescribed consistent TRT, identifying increased odds of Achilles tendon injury and subsequent surgery regardless of hypogonadism status and after controlling for other comorbidities.

Prior studies observed increased rates of injury in male patients compared to female patients. Testa et al. found increased odds of rotator cuff tears in both male and female patients prescribed TRT, with males having a higher aOR of 4.73 in comparison to 3.49 for females [14]. In another study on the association of TRT with rates of distal biceps tendon injury (BTI), Rebello et al. demonstrated that male patients prescribed TRT were 4.68 times more likely to experience distal BTI, while females prescribed TRT were 1.82 times more likely [31]. There are many speculative explanations for these findings, including the prevalence weight training, strain from higher force loads across the tendon, or rates of aerobic exercise participation [32,33,34,35]. Our study observed increased odds of general Achilles tendon injury in both men and women prescribed TRT, but in contrast to the above studies, the association was similar between men and women. However, when only considering full-thickness Achilles tendon ruptures, previous literature has identified a 5:1 ratio of men to women in the occurrence of rupture, which may be more consistent with other published literature on tendon injuries and TRT [32]. Additionally, other studies have demonstrated the decreased stiffness of the Achilles tendon in women, including Intziegianni et al. who measured an increased tendon cross-sectional area deformation under a lower applied force in female tendons, pointing to a more compliant and less injury-prone tendon that could be associated with the decreased rates of Achilles tendon rupture in women [32, 33]. Nevertheless, it remains difficult to truly explain this association without further understanding of the interplay between testosterone and estrogen in maintaining tendon health and as the effects of estrogen on tendon size, stiffness, and collagen content are still debated, this interconnection is likely multifaceted and remains a topic for additional research [36,37,38].

Within both male and female cohorts in this study, the highest adjusted odds were seen in the aged 66 to 75 cohorts. As hypogonadism increases with age, the presence of this trend in the oldest age cohort regardless of patient hypogonadism status indicates that there is likely a separate explanation for this observation. One such reason is that tendon physiology and structure can change significantly due to aging. In a study by Pardes et al., Achilles tendon properties were measured and compared among rats of various ages, with increased passive stiffness and decreased range of motion seen in older rats. Older rats were also found to have Achilles tendons with lower maximum stresses and lower elasticities compared to those of younger rats, additional properties that make a tendon more susceptible to injury [39]. Aging has a similar impact on tendons in humans, and causes additional biological changes such as decreased cellularity, reduced proliferation and activity of tenocytes, and decreased organization of collagen fibers that all weaken tendon structure [40]. The association of TRT with decreased obesity, improved energy levels, and increased self-esteem in patients with hypogonadism implicates a possibility for increased physical activity in these patients as well, which, in combination with the biomechanical and physiologic changes in tendon structure, could explain the increased odds of Achilles tendon injury in the oldest cohorts [41]. The interplay between degenerative tendon changes and participation in physical activity of each age group could also determine the acuity of the Achilles tendon injury sustained, thereby affecting the odds of undergoing surgery.

The decision to undergo surgery is also heavily impacted by the physician’s evaluation of a patient’s healing potential, which decreases with age [40]. Thus, despite having the highest odds of Achilles tendon injury, the aged 66 to 75 TRT cohorts did not have significant increases in likelihood of surgery possibly due to physician preference of nonsurgical treatment for older patients regardless of TRT prescription. Among the remaining age-sex TRT cohorts, both males and females aged 56 to 65 had significant odds of undergoing surgery for their Achilles tendon injury, which, speculatively, may be a result of an increase in participation in high-intensity exercise along with less physician preference for nonsurgical treatment out of concern for age-related complications of surgery in comparison to the oldest cohorts [42]. As this trend is not present for the male cohorts aged 56–65, in which the odds were significantly higher than the control cohorts, there likely is an additional layer of complexity in the relationship among sex, comorbidities, and TRT prescription that could explain the increased odds of surgical intervention for Achilles tendon injury for specific age-sex cohorts.

As with all retrospective administrative claims database studies, this study is not without limitations. As a retrospective and observational study, the results of this study cannot be used to make any causal inference between TRT and Achilles tendon injury. In addition, while this study does use one-to-one matching on patient demographics and the Charlson comorbidity index and subsequently controls for several comorbid medical conditions in the multivariable analyses, there are several other factors that place patients at increased risk of injury that could not be captured in this study, such as patient baseline activity level. Similarly, this study could not control for the specific Achilles tendon injury each patient incurred; there is no specific CPT code for Achilles tendon rupture and therefore we could not directly look at these injuries, which is an important entity to be considered more closely in future studies. We were unable to address additional informative factors including degree of injury, muscle atrophy, and degree of muscle retraction in potential ruptures, as well as details regarding the surgery, including specific method used, surgeon experience, or the location of the operation. We are likewise unable to determine the specific dosing of TRT, route of administration, changes in dosing over time, compliance with the prescribed dosing, or the testosterone levels of the patient over time. Therefore, this study is unable to draw any conclusions on the dose-related effect of TRT or the effect of route of administration on Achilles tendon injury and is only able to determine that there is an association between a minimum of three months of testosterone replacement therapy and increased rates of Achilles tendon injury. Finally, while this study does capture over 400,000 patients who filled prescriptions for TRT, this study only includes data from a single insurer, Humana Inc., and may not be representative of the greater national population that would include patients insured through other private companies, Medicare, or Medicaid. Notably, the much smaller sample sizes of female cohorts make it difficult to obtain a comprehensive perspective on the female population that is prescribed TRT, limiting insight into the sex differences in TRT- associated Achilles tendon injury.