TRT and Injury: Does One Lead to Another?

Does Testosterone Increase the Likelihood of Tendon Injury?

Why are men reporting tendon injuries after getting on TRT? Isn’t testosterone supposed to make the body more resilient? Answers here.

Anabolic steroid users have long suspected a link between tendon injuries and steroid use. There are two general theories as to how this could occur:

Theory 1: Steroids cause the muscles to get bigger and stronger while the tendons comparatively remain the same, leading to eventual injury.

Theory 2: High doses combined with heavy lifting cause direct weakening or stiffening of the tendons, leading to injury.

However, researchers are unable to form any firm conclusions about testosterone and tendon injuries after a comprehensive review of the scientific literature (1).

But What About Doctor-Prescribed TRT?

This is where the plot thickens. Recent retrospective case-control studies reveal significantly higher odds of tendon injuries in those taking exogenous testosterone versus those who are not. Men taking testosterone were nearly five times more likely to suffer a distal biceps tendon injury within one year of initiating therapy (2).

Researchers found similar results when evaluating the odds of quad tendon injury. The ratio of injury was nearly six times greater in those taking testosterone (3).

Weird, right? You probably wouldn’t expect these results among doctor-prescribed TRT users if most steroid users weren’t reporting the same tendon problems.

So, Testosterone Causes Tendon Injury?

Not so fast. There are reasons not to draw such a firm conclusion. Aside from more obvious limitations, like this type of study design doesn’t allow causation to be established, it’s important to note potential confounders. A confounder is a variable that may affect the outcome (tendon health) and has a relationship to the potential cause, like testosterone use in this case.

First, even though these studies also included women, the odds ratios for tendon injury were either indistinguishable between those taking testosterone and those not taking testosterone, or they were only slightly elevated in an older age group. (However, this wasn’t consistent and well within the range to be explained by several forms of research bias (4).)

If we aren’t seeing this across both sexes, it indicates a more complicated picture than simply testosterone equaling greater odds of tendon injury.

The Role of Aromatase Inhibitors

One weakness of these studies where researchers found an association (in men) between testosterone administration and tendon injury was the lack of consideration for concomitant medication use. In layman’s terms, that’s the use of other meds simultaneously.

Using aromatase inhibitors (AIs), despite being prudent and necessary for some, can also cause harmful changes to the integrity of tendons, making the users more prone to injury (5, 6). For those new to this topic, aromatase inhibitors prevent testosterone from converting to estrogen.

While probably underreported, AIs alone are known to cause tendon injuries. More recent in vitro research suggests testosterone and estrogen have an almost yin-and-yang relationship regarding tendon integrity and function (6). When estrogen isn’t present in sufficient concentrations, tendons may become more prone to injury.

This might also explain the association between anabolic steroid use and tendon injuries in those cases where endogenous testosterone is suppressed and no exogenous source of testosterone is present.

It also explains why women generally had no association between tendon injury and testosterone administration. They have much higher estrogen levels than men and rarely receive AIs as part of any TRT program. These medications are typically reserved for breast cancer treatment.

The Last Word

The evidence for TRT causing tendon injury isn’t as strong as some research has shown. I’d bet the association is due to AI use. If your clinician is administering an AI as part of your TRT, be sure he’s monitoring your estrogen levels and report any joint pain to him.




  1. Jones IA, Togashi R, Hatch GFR 3rd, Weber AE, Vangsness CT Jr. Anabolic steroids and tendons: A review of their mechanical, structural, and biologic effects. J Orthop Res. 2018 Nov;36(11):2830-2841. doi: 10.1002/jor.24116. Epub 2018 Sep 5. PMID: 30047601.

  2. Rebello E, Albright JA, Testa EJ, Alsoof D, Daniels AH, Arcand M. The use of prescription testosterone is associated with an increased likelihood of experiencing a distal biceps tendon injury and subsequently requiring surgical repair. J Shoulder Elbow Surg. 2023 Jun;32(6):1254-1261. doi: 10.1016/j.jse.2023.02.122. Epub 2023 Mar 12. PMID: 36918119.

  3. Meghani O, Albright JA, Testa EJ, Arcand MA, Daniels AH, Owens BD. Testosterone Therapy Is Associated With Increased Odds of Quadriceps Tendon Injury. Clin Orthop Relat Res. 2023 Jul 4. doi: 10.1097/CORR.0000000000002744. Epub ahead of print. PMID: 37404114.

  4. Grimes DA, Schulz KF. False alarms and pseudo-epidemics: the limitations of observational epidemiology. Obstet Gynecol. 2012 Oct;120(4):920-7. doi: 10.1097/AOG.0b013e31826af61a. PMID: 22996110.

  5. Mitsimponas N, Klouva E, Tryfonopoulos D, Grivas A, Demiri S, Koumakis G, Gouveris P. Aromatase Inhibitor-Associated Tendinopathy and Muscle Tendon Rupture: Report of Three Cases of This Exceedingly Rare Adverse Event. Case Rep Oncol. 2018 Aug 17;11(2):557-561. doi: 10.1159/000491874. PMID: 30186139; PMCID: PMC6120376.

  6. Chidi-Ogbolu NS. The Effects of Hormonal Manipulation on Ligament Function. University of California, Davis; 2022.


As n=1, my experience matches your thoughts on estrogen being a possible culprit. I was both low T and very low E. Went on TRT and my joints have improved tremendously with my E getting to a higher value and the T/E ratio getting more optimal.

Not a miracle by any stretch due to some permanent sports related orthopedic issues, but for sure a noticeable effect in every day life.

Excellent Article Cy! Love seeing critical thinking skills applied rather than blind faith.

Some additional points for people on TRT/Testosterone to research and consider:

  1. If high Testosterone levels alone were causing increased tendon injury, then shouldn’t the bulk of tendon injurues be seen in young males with high Testosterone levels? Anytime I hear a hysterical “Testosterone causes X!!!”, apply the same logic. (This may not apply to supraphysiolgic doses.)
  2. The reference range for testosterone and estrogen levels are based on average people with lower than what most here would probably consider optimal testosterone levels (especially if you’re doing TRT or using Testosterone). When you bring up Testosterone levels, Estrogen levels should naturally come up as well. Keeping Estrogen artificially low via aromatase inhibitors while simultaneously boosting Testosterone levels has a range of potentially negative effects.
  3. Women’s TRT dose for Testosterone is a tiny fraction of men’s. (On the order of 1/20th.) I’ve never heard of women using an AI as part of TRT FWIW, but it would be interesting to see if there’s any correlation between tendon injuries and women taking AI’s for breast cancer.
  4. Some TRT doctors flat out refuse to prescribe any AI because of the harmful effects. For those susceptible to estrogen-related side effects they’ll recommend the supplement DIM (Diindolylmethane).

I’ be been on TRT for about a year and a half (current age 57). Having read some negative things about estrogen blockers, I reduced frequency of taking Anastrozole to once per month, rather than the prescribed weekly. I found that I felt better and had a more even libido.
About 8 months ago I had an ice hockey accident wherein I tore my left triceps tendon 100% and my right one 40%. I am now mostly healed from surgery, etc, but have been concerned about future tendon injury. This article makes me think differently about causation. I was concerned about TRT being the culprit, but I do think that suppressing E makes more sense in terms of increasing the chances of tendon damage. I currently am about 8 weeks in on not taking Anastrozole at all, and I feel fine. I think I will stick with this course and see how it plays out.

What is your estrodol coming in at?

1 Like

I can’t believe this went unnoticed.

Testosterone inhibits collagen synthesis.

Exogenous use means there’s no natural pulse, meaning daily waves of up and downs and of course, probably a tendency to be abnormally high and/or supra physiological amounts.

Glad you enjoyed it RJ!

Interestingly enough, women have a higher rate of ligament ruptures than men but fewer muscle injuries. It’s speculated that this may be due to the differences in sex steroids.

There seems to be an interesting interplay between T and E where both are needed. The T increases collagen cross-linking making the tissue stronger but a certain level of estrogen is needed for this to occur; at the same time, estrogen decreases collagen degradation. You really do need both as you note.

On the AI’s, they can be used in rare cases but I think they are dosed excessively (i.e., should be dosed in mcg quantities instead of mg for TRT purposes). I was initially skeptical of DIM but the literature is actually pretty compelling.

1 Like

T does decrease collagen synthesis but it increases cross-linking making it stronger, albeit stiffer. Estrogen inhibits collagen degradation and at higher concentrations, inhibits cross-linking, making the tendon more lax. You really need both. T without E, you get stiff but brittle tendons. E without sufficient T, you get lax tendons that aren’t as strong. The former places more strain on the muscle, the latter placing more strain on the connective tissue. It explains the sexual dimorphism in injury rates for ligaments and muscle.

Even with exogenous use of T, you’re still getting estrogen formation and since the aromatase isn’t saturated, at least not at TRT concentrations, the increase is fairly linear. Thats, why I find exogenous T not to be a good explanation for an increases injury rate. You have to remove the estrogen from the equation. You see tendon injuries when aromatase inhibitors alone are used as well.

1 Like

I love diving into this shit and all the physiology thanks for the info!I had no idea bout the cross linking and the relation of it to those 2 hormones. I wonder if theres there any relation to hyper mobility?

I’m more curious if there’s any link between your info and the type of tendon structure you get from winstrol. Supposedly winny causes brittle tendons because the enhanced collagen synthesis is so fast. Yet it’s still an androgen acting on the same receptors as test…

If what u say is the case I would want to know more about the sampled selections. If these were active men, (I’m not going to say athletes so we can include the casuals and weekend warriors) vs say the office couch potato type), one might assume it is the muscle strength (possibly CNS as well) outpacing the collagen production. This seem to be pretty common with prolonged steroid use, but I’m making that statement anecdotally and not based on any hard data. We might even see men who get back into lifting or other sport after being sedentary from decreased energy levels and try to pickup where they left off way too fast

High T or high T ratio still seem obvious to me even based on your statements. Say AI alone is taken. You will still end up with High T even tho it’s not exogenous right? Idk how high it would be w.out the loss normally attributed to estrogen conversion. With little or no estrogen present the HTPA feedback should be calling for even more test production. Plunging the system into similar conditions.


The studies conducted regarding testosterone and tendon injury that I cover basically consisted of reviewed medical records. The authors didn’t control for a number of potential confounders. They basically looked for men and women between a large age range (they later stratified by age groups) who had recently started testosterone within the last year. They didn’t control for any other medications or lifestyle factors. They did indicate an effort to match controls with those taking testosterone by using a morbidity index score and they also had exclusion criteria for conditions that might make one prone to tendon injury. Outside of that however, it’s a study design I find to be ripe for confounders. I also noticed that the authors were very broad in their scope for formulations of testosterone which can be problematic.

One of the studies I referenced actually looked at T, alone and with an aromatase inhibitor. T alone (i.e., allowing for aromatization to estradiol) improves the mechanical properties of ligaments, in vitro at least. This is abolished when you add in an AI, potentially making the tissue more prone to injury. That makes a pretty strong case for estrogen, at least at a certain concentration, being necessary for T to have a positive impact upon tendon mechanics. This is what I was referring to in the article as the two hormones having a sort of yin-yang relationship. You have to have both at a sufficient concentration or you end up with deterioration/weakness. This, along with several other lines of evidence makes me believe it’s testosterone, without sufficient estrogen levels that leads to increased rates of tendon injury, not simply elevated testosterone. To your point, AIs will raise endogenous testosterone in men via inhibition of negative feedback at the estrogen receptor (due to lack of estradiol). However, SERMs like clomiphene or tamoxifen (via ER-antagonism) as well as hCG (LH agonist) all increase endogenous testosterone production in men to a similar or even greater extent than AIs, yet they are not associated with an increased risk of tendon injury so this causes me to believe it’s the lack of estrogen with T that causes issues as the SERMs and hCG are not interfering with the ability of aromatase to catalyze the formation of T to E. As I’ve said before, E2 will rise linearly with T, even with exogenous administration as aromatase isn’t saturated (at least not at TRT doses) so this allows for the counterbalancing to continue, even with higher levels of T. It’s when this is disrupted via an AI that issues occur. Another point is the lack of any increase in injury rates in women who were taking testosterone which we would have expected if simply increasing testosterone alone results in tendon injury.

Regarding the anecdotal reports of certain androgens like stanozolol being more likely to cause ruptures, I think that fits with the same notion above. I suspect that androgens which do not aromatize (i.e., therefore not providing any estrogenic activity to counter the androgenic effects in connective tissue) are more likely to cause these injuries, particularly if they are used in isolation without any source of testosterone (i.e., either endogenous or exogenous) to act as a substrate for estrogen formation. In other words, someone taking something like stanozolol alone will experience suppressed endogenous testosterone production and consequently, suppressed estrogen formation, resulting in only androgenic activity in connective tissues (i.e., similar to the in vitro experiments where T is given with AI; you get androgenic activity with no estrogenic activity to counterbalance these effects).

1 Like

Awesome reply! Thanks for the in depth review.

I think that’s the nail in the coffin for the argument. Especially considering what you mentioned about substances like winny causing suppression of T but not being “wet”. Make perfect sense. There’s nowhere else to go that I can see.

You mentioned E rises at a linear rate to T. I kinda knew that already. But I’m wondering if you know whether E also falls at a linear rate or do they have different half lives/metabolism rates? I’ve never been able to find an answer on that

No problem! As far as E and it’s metabolism and elimination, I’ve seen figures for a plasma half-life of 20 to 30 minutes but this is based upon IV administration of E2 so not directly related to its elimiation after aromatization from tesosterone. But, I suspect it’s probably similar. Now, there’s also the issue of enterhepatic circulation which prolongs the half-life substantially to several hours or more. One idea would be a study where they would administer radiolabeled T and then measure the elimination of any subsequently formed E and its elimation rate to get a more definitive answer.