Does Testosterone Replacement Therapy cause heart Attacks?

The short answer:

Like most things in medicine, the answer is complicated and we cannot say anything with 100% certainty.  For the most part, a well managed TRT regimen does not appear to increase the risk of developing heart disease, but the issue is still under at least some debate.  The overwhelming majority of evidence has shown testosterone replacement to be safe for the heart.  However, there have been a few studies (of questionable quality) that have shown an increased risk.  Continue reading to understand both sides of the story so you can make your own decision.  Something I hope you will realize, is that it is very dangerous and potentially harmful to draw conclusions from just a few studies, without looking at all the available data.

history of Testosterone Therapy and the FDA

To understand how TRT and heart disease became an issue, a brief review of the history is necessary. Between 2001 and 2011, the use of testosterone products in men tripled.  This large increase in prescriptions began to garner a lot of attention.  Many people falsely belief, that since more men die of heart disease than women, it must be linked to men’s higher testosterone levels.

In response to outside pressure, the FDA changed the packaging insert on testosterone products, to include a warning about the risk to cardiovascular health, in 2015.  Many advocate groups like the Public Citizen,  petitioned the FDA for an even stronger warning.  They wanted the FDA to issue a “Black Box Warning” (the most serious warning the FDA can put on a medication), on testosterone products.  The FDA refused to do this, since the research did not show a clear cut link between testosterone therapy and cardiovascular events.

what prompted the fda to put a warning on the box?

Despite a vast amount of available research that indicates testosterone is safe, the FDA essentially relied on only two studies that showed possible harm.  Prior to these two studies, 46 previous studies showed that testosterone therapy provided only a benefit to the cardiovascular system.

Both studies were a observational and retrospective design.  This means instead of giving testosterone to a group of men, then measuring the number of cardiovascular events compared to a control group (called a randomized control trial), they looked back at information in a database of men who were prescribed testosterone.  This type of study presents a few problems: First, it is unknown whether the men actually took the testosterone.  Second, it is difficult to draw reliable conclusions from retrospective studies, especially ones that use lots of fancy statistical analytics.

Lets take a look at the 2 studies the FDA was concerned about, as well as 2 others:

Study #1: The Vigen Study (JAMA)

This is THE study used by the majority clinicians and medical authors to highlight the “dangers” of testosterone therapy.  Many poorly researched articles and one sided opinions were written, without ever considering whether it was even valid.  For an example see this poorly researched article.  The fact that the study appeared in the prestigious JAMA, just added to the hysteria.

Since the journalists were incapable of conducting proper research, let’s break down the study here:

The researchers wanted to examine what happens when testosterone is given to men with existing heart disease. They looked at men, who underwent coronary angiography (a procedure that injects dye into heart vessels to see if there is a blockage).  That means it is highly likely these men already had existing heart disease.  They then compared those who received testosterone replacement therapy to men who did not (non-T group).  Remember that all these men had existing heart disease.

The authors reported that men who received TRT, had a significant increase in the rate of heart attack, stroke and death.  Since this study was posted in JAMA, it was picked up by many large media outlets and they ran with the story, without ever considering any other data to the contrary.

problems with the study:

1.  They Used Questionable Mathematical Modeling to Obtain Results:  Any time a study has to use fancy mathematics to prove their thesis, you should be suspect of the results.  The researchers used a complicated statistical process that had never been used previously, to obtain their results.  The study authors reported that the rate of adverse cardiovascular events was 25.7% in the T (testosterone) group and 19.9% in the no-T group. Again, this was only after they performed fancy mathematical modeling that attempted to adjust for 50 different variables.

A group of doctors at Harvard Medical School, evaluated the actual raw data from the study.  When they looked at the raw data, free from the fancy mathematical model, they saw the exact opposite results!  The actual rate of cardiovascular events was 10.1% in the T-treated group and 21.1% in the non-T group.  So, the raw data showed a lower rate of adverse events in men who received testosterone, the opposite of what the researchers reported!

This and other criticisms prompted the authors of the study to revise their language in a follow up correction to the original study.  Why didn’t the authors completely pull the study all together?  In my opinion, it would be very embarrassing, not only for the researchers but also for the journal in which the study was published.  In this case, it was one of the most prestigious medical journals in the world…JAMA.  It doesn’t look very good, when a well respected journal publishes studies with faulty data and conclusions.

More than 130 doctors and scientists, as well as 7 professional societies called for the study to be retracted.  This never occurred and the study still remains.  Even today, it is often cited by medical writers and doctors, as evidence that testosterone therapy is dangerous, even though the study has been shown to be invalid.

2.  They Excluded The Wrong Men:  This it where it gets confusing.  The researchers excluded from the results 1,132 men who received testosterone after they had a heart attack or stroke.  Receiving T after a heart attack is irrelevant for the purposes of this study, but for some reason the researchers excluded these men from the results.  To state this another way, these individuals underwent angiography, never received T therapy, then had a heart attack.  Only after, the heart attack they then received T therapy.  These 1,132 men should have been included in the non-T group, not excluded all together.  This greatly affected the results of the study and artificially lowered the number of men who had adverse events in the non-T group.

3.  They Included Women in the Study :  The final error is bizarre.  In a second correction released by the researchers, they state that 100 women were included in the 1,132 individuals who were excluded from the study.  How 100 women were included in the study, before they were excluded, is baffling.  Many believe this kind of error shows the poor quality of this research and is another reason the study should have been retracted.

Study #2: The Finkle Study

In this study, researchers looked at a health insurance database to examine the rate of nonfatal heart attack (aka. myocardial infarction or MI), after receiving a testosterone prescription.  This is a very commendable undertaking, and could provide very valuable data, if done correctly.

They compared men who received a T prescription and looked at the rate of MI 90 days after the prescription was filled and then compared that to the rate of MI over the previous 12 months (before a T prescription was filled).  They also compared the men who were prescribed T to men who were prescribed PDE5 inhibitors (Viagra, Cialis).  The researchers reported the results in terms of a rate ratio (RR).  A rate ratio of 1.0, means there is no difference between the 2 groups.  A ratio above 1.0 represents increased risk, and below 1.0 decreased risk.

The authors reported that the rate of MI post-prescription compared to pre-prescription was 1.36 (a 36% increase in MI).  In men over 65 the rate of MI with testosterone went up to 2.19 (over double the rate of MI, after a T prescription).  No increased rate of MI was seen in individuals taking PDE5 inhibitors.  From the sounds of this study, taking testosterone is very risky, especially if you are over 65!

Actually, you will often see the recommendation that men over 65 do not use testosterone therapy.  This recommendation is based on this study and study #3 outlined below.  Remember, at the beginning when I said, it can be dangerous to rely on just a few studies to make a recommendation?

Problems with the Finkle Study:

1.  Insurance Data is Often Unreliable:  The most significant criticism of this study was in the design.  Relying on a health insurance database, has some major issues. While we know the men filled a testosterone prescription, we do not know if they actually took the medication.  We also do not know if or how these men were managed.  What were there testosterone levels during therapy? Did their T levels improve?  It typically takes several months to find the appropriate dosage.

2.  Other Factors Played a Role:  Using an insurance database also limited the amount of available information. There was no information regarding cardiovascular risk factors such as history of heart attack, diabetes, high blood pressure, elevated cholesterol levels, smoking history or obesity.  These factors would play a major role in the development of an MI.  We have no idea whether the men who developed an MI, also suffered from conditions that would have put them at risk for MI.  Maybe the men who had an MI were also obese smokers with a long history of high blood pressure and by the way they also were prescribed testosterone.  We have no way of knowing.

These conditions that weren’t accounted for, are called confounding variables. Their presence, if not accounted for, can give you useless results.

3.  Diagnosis Codes are Unreliable:  Since the researchers where only looking at a database, the only way they could ascertain if a man had an MI was to search for a diagnosis code for MI. Just because there is a diagnosis code for MI in the patient’s chart, it does not mean the man actually suffered one.  You are relying on doctors and other providers to code things correctly, which I can tell you, they very often get wrong.  Many times patients get worked up for MI, receive an MI diagnosis code only to later have it proven that they did not actually suffer an MI at all, but the code is still attached to the patient’s chart.

The researchers had no way to verify this.  We cannot say for sure if these men actually had a heart attack.

4.  The MI Risk Was Not Increased Compared to the General Population:  Dr. Morgentaler and colleagues point out that the rate of MI reported in the Finkle study, while sounding alarming, was actually relatively low compared to what is typically seen in the general population.  They state “the observed MI rate among men who received a T prescription was thus approximately one-third the expected rate. In the absence of a control group of men who were untreated, it is impossible to determine whether these reported MI rates were higher, lower, or unchanged in association with a T prescription.”

5.  There Was No Control Group:  For a study to be reliable it needs to have a control group.  A control is a group that is not treated with anything, that would effect the end result.  In this study, there was no real control group since, their “control” received PDE-5 inhibitors (Cialis & Viagra).  These drugs are designed to dilate blood vessels, causing the vessel to become wider.  This allows blood to flow more easily.  This not only occurs in the penile tissue, it also occurs all over the body, including the heart.  Since an MI is a restriction of blood flow to the heart, it would make sense that a drug like Viagra, would have a beneficial effect on the risk of MI.

A real control group, would not have received a treatment that lowers the risk of heart attack.  The two groups should never have been compared.  As Dr. Mortangaler puts it, “this is a classic case of comparing apples and oranges.”

Study #3:  THE BASARIA STUDY (The Study that Started It All)

I am including this study to show the original investigation that started the concern regarding testosterone and cardiovascular events.  The study was not designed to evaluate risk of testosterone replacement on cardiovascular problems.  They were actually investigating whether testosterone gel improved muscle mass, strength and mobility in a group of frail older men, over 65.  The study was stopped early because of an increase in adverse “cardiovascular events” in the men receiving testosterone.  The group receiving T gel had 23 “events”, compared to 5 in the placebo group.

You probably noticed I put “cardiovascular events” in quotes.  This is because this study was not designed to measure cardiovascular events.  They were trying to see whether T gel improved muscular strength and function (which it did).  The adverse cardiovascular events were essentially subjective symptoms or of questionable importance.  These include foot swelling, a feeling of palpitations, and premature heart contractions (a common phenomenon).  Of the 23 events only 4 were actually major adverse events.  There was 1 death, 2 MIs, and 1 stroke.  Regardless, the study’s safety monitoring board recommended ending the trial, which led to even more concern.

If a clinical trial of testosterone had to be stopped, then there must be a big problem!  Well, not really.  The safety board got spooked because they saw such a major increase in “events”, but didn’t stop to think if these events were significant.  Their first priority is always to do no harm.

Although these serious events are concerning, it is difficult to draw definitive conclusions from such a small study of 209 men.  Statistically, these events can occur (especially in a group of frail old men).

The authors of the study even state, “the lack of a consistent pattern in these events and the small number of overall events suggest the possibility that the differences detected between the two trial groups may have been due to chance alone.”

The FDA issued a statement as well:  “The Basaria study does appear to show an empirical dose-dependent association between testosterone and cardiovascular risk, but it was non-conclusive because of the small sample size and small number of events reported in the study, as well as the limitations with respect to confirming the events. The authors of this study have explicitly indicated that the differences between the groups in cardiovascular adverse events might have been due to chance alone.”

Study #4: The Budoff Study

The final study in this review was done in 2017, and did not factor into the FDA’s warning.  In this study the researchers used a special heart scan called CCTA.  This test can measure plaque build up in the arteries of the heart.  The scan can also determinate if the plaques are hard (sealed over with a hard calcium cap) or soft (no calcium cap is present).

This study did use a control group.  The treatment group received testosterone gel over 12 months, while the control group did not get testosterone.  All of the men were older (over 65 with the average age of 71).  Half of the men already had severe coronary artery disease (atherosclerosis) and all of the men had at least some atherosclerosis.

The researches found that both groups had an increase in non-calcified (soft) plaque volume (a normal phenomenon, since heart disease tends to progress), but the increase was larger in the testosterone gel group compared to the control group.  They also found that the total plaque volume increased in the T-gel group, more so than the control group.  Interestingly, they also reported that the coronary artery calcification score (hard plaque) decreased in the T-gel group, but increased in the control group.

Confused yet?

Soft plaques increased in both groups, but more so in the T-gel group.  The hard plaques actually decreased in the T-gel group, but went up in the control group (although both changes were not statistically large enough to be called significant).

Let’s break this down:

1.  The Study Was Very Small:  Only 88 men were in the T-gel group and 82 men in the control.  Studies of this size need to be interpreted with caution and compared to other similar studies before drawing conclusions.

Imagine flipping a coin 88 times and measuring the number of heads vs. tails.  Then flipping the same coin 88 times and measuring again.  Do you think you would get the same number of heads in each trial?  Sometimes you may get a 50/50 split, other times you may see a significant difference between the two groups.  If you flipped the coin 1,000 times or 10,000 times your result is more likely to be closer to 50/50.  This is called the law of large numbers.  In order to feel confidant in your conclusion, you need to have a large sample size.  Flipping a coin 4 times would give skewed results.  Flipping a coin 1,000 times should provide a more balanced and accurate result.

So, while the results are certainly concerning, they need to be taken with a grain of salt, since the results could be due to chance alone.

2.  The Control Group Was Different:  The T-gel group and the control group were not exactly the same.  The T-gel group had a baseline non-calcified plaque volume of 204 mm³.  The control had a non-calcified plaque volume of 317 mm³.  This is a fairly significant difference.  It would have been better if the T-gel group and control had been closer to each other, in order to draw more accurate conclusions.

3.  There Were Other Factors at Play:  The men in this study had high rates of obesity and other problems (high blood pressure, high cholesterol and diabetes).  These are other well known reasons for plaque progression other than testosterone treatments that should be considered, when interpreting the results.

4.  The Length of the Study was Short: Heart disease is a multi-factorial problem that progress slowly over many many years. The trial length of 12 months was relatively short.  If the study observed these men over a longer period, such as 3-5 years, the difference may have went away.  Maybe it would have been worse?  We can’t really know for sure.

5.  Soft Plaques are Not as Dangerous as We Used to Think:  Newer studies have indicated that hard (calcified) plaques appear to be more dangerous than soft plaques.

Doctors now think that the calcified plaques are more predictive of MI and death, than soft plaques.  More research is underway in this area, but this is a change from how doctors used to think about the two types plaque in our arteries.  The traditional thinking was that hard plaques were more stable and less likely to rupture.  However, we now believe that it is the opposite.  It appears that the soft plaques that are actually more likely to rupture, and cause an MI.

The Take Home

In my opinion, this study is certainly the most concerning of all the studies mentioned here, but it cannot be considered in isolation.  Many, many other studies have shown a positive effect of TRT on heart disease, MI and death.  More research needs to be done in this area.  If you have a known history of heart disease or a significant family history, you may want to proceed with caution when considering testosterone replacement therapy.  No participants in this study had a severe cardiac event, but if the study had been conducted longer we may have seen a difference.

Summary

I have presented you with all the studies that show the case against testosterone replacement therapy.  Certainly, if you only read the study summary, then that will be the impression you have.  This is likely what happened with the media, when they rushed to create a story.  As is the case with many things, the truth is harder to find and is rarely simple.  The FDA has a difficult job to do, and their top priority is to protect those of us who use these medications.  It is up to you and your physician to decide whether the FDA’s warning holds water and if testosterone therapy is right for you.  Interestingly, both the American Association of Clinical Endocrinologists and the American College of Endocrinology, as well as an international expert consensus panel felt the FDA’s warning lacked scientific evidence (2)(3).

I have included some more interesting links below.  Also, if you are wondering if there is any research that shows the other side of the coin, please read my article Is Low Testosterone Bad for Your Heart?.

Interesting links:

Lecture by Dr. Morgentaler about TRT and Heart Disease Risk:

Interview with Dr. Morgentaler regarding his breakdown of the above studies:

Low T: Separating Facts From Frenzy

The Low T Story: Hunting for the Truth

Update: The Low-T Story: Hunting for the Truth, Part 2




Articles from major news outlets regarding the above studies (note the titles):

Don't Ask Your Doctor About 'Low T'

Testosterone Prescriptions Linked to Heart Attack

Popular Testosterone Therapy May Raise the Risk of Heart Attack