Study Finds Increased Risk of CVD from TRT

[Jinzang]

A newly published study found an increased risk of stroke and myocardial infarction from testosterone replacement therapy. The risk was elevated for the first two years of treatment, suggesting that there is a subpopulation at risk, presumably those with existing heart disease. The journal article is paywalled, but the abstract is available, here is most of it:

"Using the UK Clinical Practice Research Datalink, we formed a cohort of men aged 45 years or older with low testosterone levels and no evidence of hypogonadotropic or testicular disease, between 1995 and 2017. Hazard ratios (HRs) and 95% confidence intervals (CIs) of a composite of ischemic stroke/transient ischemic attack and myocardial infarction were estimated using time-dependent Cox proportional hazards models, comparing current use of TRT with nonuse."

"The cohort included 15,401 men. During 71,541 person-years of follow-up, 850 patients experienced an ischemic stroke/transient ischemic attack/myocardial infarction (crude incidence rate 1.19 [95% confidence interval (CI), 1.11-1.27] per 100 persons per year). Compared with nonuse, current use of TRT was associated with an increased risk of the composite outcome (HR 1.21; 95% CI, 1.00-1.46). This risk was highest in the first 6 months to 2 years of continuous TRT use (HR 1.35; 95% CI, 1.01-1.79), as well as among men aged 45-59 years (HR 1.44; 95% CI, 1.07-1.92)."

 

[Systemlord]

This study is only reporting a finding without finding the causality. Consider the poor treatment of men on TRT, no proper testing which is common in the UK.

 

[Rain27]

Wat type testosterone usage?

 

[Rain27]

Wat type testosterone they used

 

[Rain27]

Undecanoat
Enanthate
Cipionato
Gel ?

 

[Nelson Vergel]

I bet in the UK, TRT-induced high hematocrit management is not common. I also bet that no subject had therapeutic phlebotomy. Also, I bet many men were still hypogonadal on TRT.


The title should be:

"Study Finds Increased Risk of CVD from Poorly Managed TRT"

 

[Joe Sixpack]

Compared with nonuse, current use of TRT was associated with an increased risk of the composite outcome (HR 1.21; 95% CI, 1.00-1.46).

Is this trying to tell me that TRT users had a 21% higher chance of adverse outcomes than non users?

 

[xqfq]

This was a cohort study done using health data that’s available in the UK. So I don’t think researchers gave a specific TRT protocol, etc. They searched this database for people on TRT and then looked at their outcomes over a period of time. This is important because there could be huge variations in terms of how treatment was delivered and how health was monitored.

The abstract notes that the risk increase was only seen in the first two years of TRT. This makes me think that — if this risk increase was due to TRT usage and not factors in the hypogonadal subpopulation — that it’s related to increases in hematocrit and blood pressure (rather than, say, dyslipidemia / lowering of HDL in some men in some cases).

As we know, hypogonadal men are at increased CVD risk. So there may have been a selection problem here, meaning that this subpopulation may have already been at higher risk than baseline, regardless of TRT usage. And this risk could have built over many years, meaning TRT may not have done much to reduce the aggregate risk. I would want to see if the study accounted for this with a control group consisting of hypogonadal men not on TRT, but I suspect that data simply wasn’t available in their data set as “TRT usage” may have been their “is hypogonadal?” check. But I’d need the full article to see .

Regardless of control group, men seeking out TRT may also have other characteristics that make them more likely to get CVD. In simple terms, they might be sicker or more unhealthy than the baseline population, and possibly sicker than even the baseline hypogonadal population. Because this wasn’t an interventional study (the researchers didn’t select people and then give them TRT), this is a possibility.

 

[Jinzang]

Presumably, these are all men under the British NHS. The standard for treatment is strict and those with low testosterone are likely to have a host of other health problems, including pre-existing heart disease.

 

[Joe Sixpack]

This was a cohort study done using health data that’s available in the UK. So I don’t think researchers gave a specific TRT protocol, etc. They searched this database for people on TRT and then looked at their outcomes over a period of time. This is important because there could be huge variations in terms of how treatment was delivered and how health was monitored.

The abstract notes that the risk increase was only seen in the first two years of TRT. This makes me think that — if this risk increase was due to TRT usage and not factors in the hypogonadal subpopulation — that it’s related to increases in hematocrit and blood pressure (rather than, say, dyslipidemia / lowering of HDL in some men in some cases).

As we know, hypogonadal men are at increased CVD risk. So there may have been a selection problem here, meaning that this subpopulation may have already been at higher risk than baseline, regardless of TRT usage. And this risk could have built over many years, meaning TRT may not have done much to reduce the aggregate risk. I would want to see if the study accounted for this with a control group consisting of hypogonadal men not on TRT, but I suspect that data simply wasn’t available in their data set as “TRT usage” may have been their “is hypogonadal?” check. But I’d need the full article to see .

Regardless of control group, men seeking out TRT may also have other characteristics that make them more likely to get CVD. In simple terms, they might be sicker or more unhealthy than the baseline population, and possibly sicker than even the baseline hypogonadal population. Because this wasn’t an interventional study (the researchers didn’t select people and then give them TRT), this is a possibility.

Very good writeup my friend. I agree with you.

I enjoyed your answer, although, what I was asking about is the meaning of this specific language...
Compared with nonuse, current use of TRT was associated with an increased risk of the composite outcome (HR 1.21; 95% CI, 1.00-1.46).

Does HR 1.21 mean that the TRT user group experienced a 21% increase in adverse events?

PS, Sorry I wasn't more clear in wording my question.

 

[xqfq]

Joe Sixpack: yes, that's my understanding of what the "hazard ratio" means.

Does anyone know what the commonly used injectable testosterone protocols would have been in the UK during the study period? These (current!) British Society for Sexual Medicine guidelines state that testosterone enanthate should be given "every 2–3 weeks."

 

[Nelson Vergel]

Joe Sixpack: yes, that's my understanding of what the "hazard ratio" means.

Does anyone know what the commonly used injectable testosterone protocols would have been in the UK during the study period? These (current!) British Society for Sexual Medicine guidelines state that testosterone enanthate should be given "every 2–3 weeks."

At that dosing regimen, most men would be hypogonadal at week 3, which explains the higher rate of cardiovascular events.

Another Study Review Debunks Two Prior Flawed Papers That Linked Testosterone and Heart Attacks

Does Testosterone Increase the Chance for a Heart Attack?

 

[Vince Carter]

This is case in point of why I don't read studies and dismiss the posts of guys that fall back on them. They all seem so flawed and you don't have to look far to find the gaping hole(s).

 

[Deleted member 16042]

Very good writeup my friend. I agree with you.

I enjoyed your answer, although, what I was asking about is the meaning of this specific language...
Compared with nonuse, current use of TRT was associated with an increased risk of the composite outcome (HR 1.21; 95% CI, 1.00-1.46).

Does HR 1.21 mean that the TRT user group experienced a 21% increase in adverse events?

PS, Sorry I wasn't more clear in wording my question.

"hazard ratio" is not a direct comparison of how many from group A has adverse events compared to how many from group B have events in a given time period. It is a mathematical manipulation of raw data Hazard Ratio Calculator - calculate hazard ratio, HR confidence intervals and p-value.

For example, if you have 1 person out of 100 that has an event who is not on TRT and 2 out of 100 that have an event on TRT, that would be a hazard ratio of 2, so you would have double the potential hazard compared to control group.

But in direct comparative terms you would have only a 1% higher comparative risk in terms of side by side comparison of the raw data.

To me, side by side comparison of data is much more realistic and important than hazard ratio or relative risk or relative risk reduction. These are generally clever methods to make data comparison seems much more dramatic.

There is another kind of flaw in the hazard ratio calculation: I ran a couple quick calcs on that calculator. It does not take into consideration how many are in the sample groups, i.e. whether 100 people in treatment vs 100 people in control vs 1000 in treatment vs 1000 in control. In either case if 12 in treatment group has adverse event and 10 in control do, the ratio is 1.2. To me this mathematical representation of ratio is pretty meaningless without seeing the actual raw data.

 

[madman]

Joe Sixpack: yes, that's my understanding of what the "hazard ratio" means.

Does anyone know what the commonly used injectable testosterone protocols would have been in the UK during the study period? These (current!) British Society for Sexual Medicine guidelines state that testosterone enanthate should be given "every 2–3 weeks."

Cardiovascular and cerebrovascular safety of testosterone replacement therapy among aging men with low testosterone levels: a cohort study

Simone Y. Loo, Laurent Azoulay, Rui Nie, Sophie Dell'Aniello, Oriana Hoi Yun Yu, Christel Renoux





RESULTS
After applying all exclusion criteria, our final cohort comprised 15,401 hypogonadal men (figure 1) with a mean age of 60.4 years (standard deviation (SD) 9.6 years) (table 1). TRT was prescribed to 4,485 patients (29.1%) on a least one occasion during a mean follow-up time of 4.7 years (SD 3.7 years), with a majority of patients initiated on testosterone gels/creams (56.8%) and injections (33.6%).






DISCUSSION
Our study is notable for specifically evaluating the cardiovascular and cerebrovascular safety of TRT among men with low testosterone levels in the context of aging. Given the recent trends in which these medications are prescribed in the absence of underlying disease, our findings are therefore generalizable to a significant number of aging, but otherwise healthy men who may be considering treatment. Our analyses were also designed to circumvent the biases which may have influenced previous findings. Mainly, a time-varying TRT exposure definition was used to eliminate the potential for immortal time bias, and we further estimated the rate of events as a function of duration of TRT use. Despite these strengths, several limitations to our study must also be considered. First, while patients with testicular and hypogonadotropic disease were excluded, our cohort may still include a mix of patients with low testosterone levels due to other etiologies that may cause functional hypogonadism including chronic opioid or glucocorticoid use, systemic illness and obesity. However, this heterogeneity reflects the prescribing patterns that have been observed in several countries and our results are still informative with respect to the safety of TRT in this population. Second, in defining hypogonadism using laboratory test values, we were not able to account for known diurnal fluctuations in endogenous testosterone levels. Nevertheless, we assume that clinicians are adherent to local guidelines for testosterone testing, which recommend blood sampling and testosterone testing in the mornings only and during a fasting state. Third, our definition of TRT exposure relied on prescriptions issued by general practitioners, and we were not able to determine patients’ adherence to their prescribed treatment. However, given the consistency of our results using a variety of exposure definitions, we expect the impact of any exposure misclassification to have been minimal. Fourth, in the stratified analyses, our study was underpowered to draw a firm conclusion on a potential effect modification by age or prior history of cardiovascular disease. Thus, these prespecified secondary analyses should be interpreted with caution. In addition, it was only possible to examine the risk with all-cause mortality as we did not have information on death from cardiovascular causes. Finally, although our analyses accounted for numerous potential confounders, including lifestyle habits and timevarying confounders, residual confounding is possible given the observational nature of the study.


Our findings suggest that TRT may be associated with an increased risk of ischemic stroke/transient ischemic attack myocardial infarction among aging men with low testosterone levels, although the overall risk difference was low. Patients may be especially susceptible in the first two years after treatment initiation. Further large and methodologically sound observational studies should be conducted to reaffirm these results. Until such a time, the potential harms of TRT should be critically weighed against its perceived and expected benefits, and caution is warranted when prescribing these medications to aging but otherwise healthy men.

 

[madman]

This is one of the more recent papers regarding trt and cardiovascular risk!

Testosterone replacement therapy and cardiovascular risk


Key points to keep in mind:

•The TRAVERSE trial, the first trial of testosterone therapy that is adequately powered to assess cardiovascular events, began in 2018, and its findings might take a decade to become available.

•Until the results of the TRAVERSE trial are available, clinicians should individualize testosterone treatment after having an informed discussion with their patients about the risks and benefits of testosterone replacement therapy.





Conclusions
Testosterone has an important role in cardiovascular physiology and metabolic health. Although epidemiological data are conflicting, some retrospective studies support a beneficial effect of testosterone replacement therapy on mortality, whereas other reports suggest that testosterone might increase the risk of serious cardiovascular events, a finding that is also supported by some randomized trials. Meta-analyses also report conflicting results and are limited by the inclusion of low-to-medium quality trials. Furthermore, to date, no published trials of testosterone replacement therapy have been adequately powered to assess cardiovascular events. Therefore, the TRAVERSE trial 41 will go a long way towards enabling an assessment of the cardiovascular safety of testosterone therapy. In the interim, clinicians should guide their patients in making informed decisions by having an open discussion about the current evidence on the cardiovascular safety of testosterone replacement therapy

 

[Joe Sixpack]

For example, if you have 1 person out of 100 that has an event who is not on TRT and 2 out of 100 that have an event on TRT, that would be a hazard ratio of 2, so you would have double the potential hazard compared to control group.

But in direct comparative terms you would have only a 1% higher comparative risk in terms of side by side comparison of the raw data.

To me, side by side comparison of data is much more realistic and important than hazard ratio or relative risk or relative risk reduction. These are generally clever methods to make data comparison seems much more dramatic.

Good point. This relative risk shannannigans is how they sell statins. It's very dishonest.

 

[Jinzang]

Here's more information on the study from an editorial in MedScape:

Renoux and colleagues' latest work provides no easy answers, despite a large cohort of 15,401 men, aged 45 years or older (mean 60.4 years), diagnosed with low testosterone without evidence of hypogonadotropic or testicular disease in the UK Clinical Practice Research Datalink database between 1995 and 2017.

Most patients received testosterone gels or creams (56.8%) and 33.6% received injections.

The results show a 21% increased risk for the composite of MI and ischemic stroke/transient ischemic attack (TIA) in current TRT users vs nonusers after adjusting for more than 20 potential confounders (hazard ratio

---

, 1.21; 95% confidence interval [CI], 1.00 - 1.46).

This corresponds to an adjusted risk difference of 2.4 events per 1000 persons per year, the authors reported July 18 in the American Journal of Medicine.

MI and stroke risk was highest in the first 6 months to 2 years of continuous use and mostly driven by an increased risk among men aged 45 to 59 years. Risk was also increased in men with and without prior cardiovascular disease (CVD), but the results did not reach statistical significance.

Surprisingly, current TRT use was associated with a lower risk for all-cause mortality (HR, 0.64; 95% CI, 0.52 - 0.78) and past use with an increased risk when compared with nonuse (HR, 1.72; 95% CI, 1.21 - 2.45).

"It's difficult to reconcile on one hand an increased risk of cardiovascular disease, which is one of the leading causes of mortality in older men, and at the same time finding a strong predictive effect on all-cause mortality," Renoux said.

"When you see such a strong protective effect on all-cause mortality, you have to begin to wonder because if it's too good to be true, it probably is," she said. "Because how many drugs can you claim have a 75% decrease risk in all-cause mortality?"

Reached for comment, Bradley Anawalt, an endocrinologist and chief of medicine at the University of Washington Medical Center in Seattle, said he was surprised the authors dismissed the all-cause mortality findings and highlighted a 2012 study he coauthored with "equally stringent statistical analyses," in which the risk for death was lower among men aged 40 years or older treated with testosterone (adjusted HR, 0.61; 95% CI, 0.42 - 0.88).

"There are plausible mechanisms; testosterone increases strength, which means perhaps these men didn't fall and break their hips or die from fractures," he said.

"To a priori say it couldn't be a drug effect that has the same biases and errors in logic someone might have if they say it couldn't have been the testosterone causing the heart attacks and strokes," Anawalt said.

As for the effect of testosterone on the composite of MI/stroke/TIA, Anawalt said that the absolute risk difference for just 2.4 events/1000 persons/year was very small. Exactly how TRT imparts CVD risks remains unclear, but the authors note it may enhance platelet aggregation.

 

[Joe Sixpack]

I also take comfort in reading that the highest risk was during the 6-24 month period when first starting TRT. I am well past that mark now. Probably a lot of us here are past that.

 

[Nelson Vergel]

I also take comfort in reading that the highest risk was during the 6-24 month period when first starting TRT. I am well past that mark now. Probably a lot of us here are past that.

Good point. That is when water retention (increases blood pressure due to increased blood volume) and hematocrit increase the fastest.