I’m skeptical about … testosterone therapy.

“Is it low T?” or just another scam to separate you from your money? The “T,” in this case, stands for testosterone, the male hormone that is responsible for making men, well … men. Sales of this hormone tripled between 2001-2011, thanks largely to an aggressive ad campaign targeting middle-aged men seeking a cure-all for whatever ails them. The natural aging process is associated with decreased libido, diminished sexual performance, loss of muscle mass, decreased strength, and mood swings—all things that testosterone is said to cure. Testosterone levels have been inversely associated with heart disease, stroke, diabetes, obesity, depression, dementia, and all-cause mortality. And this makes a certain amount of sense. For instance, low testosterone is associated with increased abdominal fat stores, a known risk factor for insulin resistance and diabetes. But association is not causation. Does low testosterone lead to abdominal obesity, or does a couch-potato life lead to low testosterone? I’m not surprised that men with lots of health problems have low testosterone levels, but it’s a big stretch to think that supplementing this hormone will cure any of them. There is good evidence that the real culprit for low testosterone is inactivity. One of the best ways to boost levels is through high intensity interval training. This also happens to be one of the best ways to combat obesity. I believe we should be peddling exercise, not AndroGel.

Until recently, supplementing testosterone was difficult, requiring weekly trips to the clinic for injections. It’s a lot easier now that the hormone is available in patch, gel, and buccal preparations. The ads would have you believe that vitality is as close as the nearest drugstore. Given the ease of gel administration combined with healthy doses of direct-to-consumer advertising, it’s no wonder that sales have skyrocketed, projected to reach a cool $4 billion in 2017. But this is not to say that certain specific populations won’t benefit from therapy. Those with hypogonadism related to endocrine disorders certainly warrant treatment, but the market here is small. This post is directed toward the much larger market of middle-aged men seeking a hormone cure for aging.


J. Fam. Practice 2016; 65 (12): 864.

J. Fam. Practice 2016; 65 (12): 864.


By way of reference, male testosterone levels typically range from 230-1000 ng/dL, with peak levels occurring in late adolescence and declining by roughly 1.6% per year after age 30. In the blood, testosterone exists in 3 different forms: biologically active, free testosterone (2-3%); bioavailable, protein-bound testosterone (approximately 50%); and non-bioavailable, testosterone bound to sex hormone binding globulin (approximately 45%). As men age, the proportion of non-bioavailable, protein-bound testosterone increases, meaning that even if the total testosterone level remains constant, there is still less of the stuff available to exert androgen effects. Levels also vary diurnally so testing should be done before 11 a.m., during the time testosterone peaks. At least 2 measurements of less than 280 ng/dL are required for the diagnosis of “low T,” the prevalence of which varies by age. While 12-38% of men over age 45 test low, this range increases to 30-50% by age 70. But that doesn’t mean everyone with low testosterone should be treated. Since the symptoms of “low T” are multifactorial and overlap considerably with those of normal aging, only men with testosterone levels less than 280 ng/dL and symptoms should be considered for treatment. If the testosterone level is greater than 350 ng/dL then “low T” is unlikely to be the source of symptoms. Look instead for another culprit, like diabetes, vascular disease, depression, or medication side-effects.





Most medical societies recommend supplementation to achieve levels equal to the mid-range for men less than 40-years of age (450-500 ng/dL), but this number is completely pulled out of a hat. There is no data to support the notion that a 75-year old man requires testosterone levels of someone half his age to achieve vitality. On the flip side, too much testosterone is clearly harmful, resulting in shrunken testicles, acne, baldness, aggression, and an increased risk of blood clots, stroke, congestive heart failure, and cardiovascular disease. Just take a look at some of the ex-NFL players from the 70s and 80s if you want to see the deleterious, long-term effects of steroid abuse.

The data on cardiovascular disease and testosterone are conflicting, some studies showing benefit, others showing harm. The FDA issued a warning in 2015, after 2 studies reported increases in cardiovascular events associated with TRT (testosterone replacement therapy), noting that: “The benefit and safety of these medications have not been established for the treatment of low testosterone levels due to aging, even if a man’s symptoms seem related to low testosterone.” The most interesting thing to note about the controversy is that the degree of benefit or harm reported with TRT has little to do with testosterone levels and everything to do with who funded the study. A meta-analysis of 27 trials looking at TRT and cardiovascular disease found that, on-average, drug company-sponsored trials reported a 10% reduced risk of adverse events, while non-sponsored trials showed just the opposite, reporting more than twice as many cardiovascular events in TRT patients. And you thought science was pure.

A second warning, specifically targeting athletes and body-builders, was issued by the FDA in 2016 requiring that testosterone and androgen manufacturers include a list of the adverse health effects associated with steroid abuse on all package inserts. With inconclusive data on both efficacy and safety, the NIH (National Institutes of Health) and NIA (National Institute on Aging) subsidized a series of interrelated trials, known collectively as the “T-Trials,” to determine the efficacy of testosterone when administered to symptomatic men with age-related androgen deficiency. Investigators enrolled a total of 790 symptomatic men aged 65 or older with low serum testosterone (less than 275-300 ng/dL) to receive either a placebo or a testosterone containing gel for 1 year. Participants were enrolled in one or more trials to examine the effects of supplementation on 7 different parameters: bone density, anemia, cognition, cardiovascular disease, physical function, vitality, and sexual function. The results of these trials are now available.


AndroGel (the active ingredient used in the "T Trials" sponsored by the NIH and AbbVie Pharmaceuticals)

AndroGel (the active ingredient used in the “T Trials” sponsored by the NIH, NIA, and AbbVie pharmaceuticals)


TRT demonstrated moderate improvements in bone density as measured by quantitative CT scanning, but the trial’s duration was too short to determine whether this will translate into meaningful patient outcomes like a reduction in falls and fractures. So, too, TRT improved red blood cell counts in patients with anemia, and also resulted in modest increases in the counts of non-anemic patients, a few of whom developed polycythemia (a disorder of increased blood viscosity from too many red blood cells, predisposing to stroke). The number of participants suffering a stroke over the course of the trial was the same in the treatment group and the placebo group (5 of 394 in both). No conclusions can be drawn regarding the long-term stroke risk associated with therapy.

TRT had no meaningful effect on cognition as measured in the trial. Treated patients performed no better than untreated patients on tests of visual or reading memory retention. Vitality was measured via a 13-question survey, each question graded 0-4, based on answers to questions like: “I feel fatigued,” “I feel weak all over,” and “I have trouble finishing things.” On this survey, TRT did not improve outcomes. So much for the vitality claims of the testosterone pushers. If you want more “get-up-and-go,” maybe try a double-shot espresso instead.

As regards to cardiovascular health, TRT resulted in an increase in the amount of plaque deposited along the walls of coronary arteries as measured by specialized CT scanning. And, since we know that the degree of plaque clogging these arteries constitutes a major risk factor for heart attack, these results indicate the potential for harm. It’s important to note that none of the “T Trials” were designed to measure safety. They were, instead, designed to determine efficacy; does the hormone work on real patients in the community? Furthermore, the results of this trial are hard to interpret in light of several observational trials supporting a cardioprotective role for TRT, including those of a trial published earlier this year that found a moderate reduction in cardiac events like heart attack, worsening chest pain, or sudden death in men treated for at least a year with TRT. Unfortunately, observational trials can only show associations—it takes a randomized controlled trial to show cause and effect. The question of cardiovascular safety remains unanswered.


Chemical structure of testosterone.

Chemical structure of testosterone.


How about physical function? Can testosterone make you stronger, thinner, and more agile? This trial didn’t answer those questions, but TRT didn’t have any meaningful effect on the physical performance of deconditioned, elderly men when measured on a 6-minute walk test. Granted that they weren’t a very fit cohort going in—to be eligible required a gait speed of less than 1.2 meters/sec (equating to a 22.4 minute/mile pace). These results are not translatable to highly conditioned athletes where uber-supplementation enhances performance (at a cost), but they do suggest that, at recommended doses, the benefit on physical performance for the average Joe is likely to be minimal.

That leaves the big question: Does testosterone improve sexual function? Being a stud is important, which is why Donald Trump proudly held up his hands during the last election cycle, even releasing his testosterone level to the public (441 ng/dL). Men wanting an improved sex life constitute the overwhelming majority of those requesting treatment. The “T trials” examined 3 components of sexual function: activity, desire, and erectile function. The results were underwhelming. Based on the answers to a series of pre-validated questions, participants related slight improvements in sexual activity (+0.58 on a 0-12 scoring system), sexual desire (+2.93 on a 0-30 scoring system), and erectile function (+2.64 on a 0-33 scoring system). Of note: the improvement in erectile function was less than that achieved with drugs like Cialis and Viagra. At the end of the trial, more of the men in the treatment group reported improvement in sexual desire than in the placebo group. Testosterone works to improve sexual function, but not very well. Unfortunately, there wasn’t a third arm involving exercise. Physical fitness is likely the single most important factor when it comes to sexual desire and erectile function. Want to boost testosterone and improve your sex drive? Try exercising with your partner and showering together afterward. As most women will tell you, a hard man is good to find.

Lastly, a note on my own experience with supplements. A decade ago, at age 48, I was a highly-competitive skater on the in-line marathon circuit struggling to make the podium at national races, mostly due to a lack of top-end speed during the final sprint. I knew from talking to other athletes that supplements alone wouldn’t necessarily make me faster, but they would almost certainly allow me to work harder, longer, and with less recovery time. And maybe that would land me a place on the podium. So, I went to one of those high-priced anti-aging/wellness centers catering to middle-aged men looking to boost their mojo and underwent an executive physical that included a battery of neurocognitive tests, an EKG, a body fat analysis, bone density assessment, and a host of lab tests including a testosterone panel. Since I was there for performance enhancement, I wasn’t surprised when the doctor recommended a course of hGH (human growth hormone) in addition to a bevy of vitamin and nutrition supplements, but I was most definitely surprised when he recommended a testosterone stimulant because my level (at 881 ng/dL) already had me at the very top of my age group. Despite this, he thought I could push it a bit higher still: “If you want to perform like a 20-year old, you need the testosterone of a 20-year old.”  In the end, I declined all supplements, but my point is that if you go to one of these clinics looking for testosterone you’ll get it. That’s how they make their money. Even worse, nearly a quarter of the 11 million men receiving androgen prescriptions haven’t had their testosterone levels screened within the past year. Now, that’s discouraging. Caveat emptor.



  1. Roger Stanworth and Hugh Jones, “Testosterone for the Aging Male; Current Evidence and Recommended Practice,” Clin Interventions in Aging 2008; 3 (1): 25-44.
  2. Lin Xu et al., “Testosterone Therapy and Cardiovascular Events Among Men: A Systematic Review and Meta-Analysis of Placebo-Controlled Randomized Trials,” BMC Med 2013; 11: 108-19.
  3. “FDA Drug Safety Communication: FDA Cautions about Using Testosterone Products for Low Testosterone Due to Aging,” US Food and Drug Administration, Safety Announcement, March 3, 2015, www.fda.gov/Drugs/DrugSafety/ucm436259.htm.
  4. Peter Snyder et al., “The Testosterone Trials: Seven Coordinated Trials of Testosterone Treatment in Elderly Men,” Clin Trials 2014; 11: 362-75.
  5. Peter Snyder et al., “Effect of Testosterone Treatment on Volumetric Bone Density and Strength in Older Men with Low Testosterone: A Controlled Clinical Trial,” JAMA Int Med 2017;177 (4): 471-79.
  6. Cindy Roy et al., “Association of Testosterone Levels with Anemia in Older Men: A Controlled Clinical Trial,” JAMA Int Med 2017; 177 (4): 480-90.
  7. Susan Resnick et al., “Testosterone Treatment and Cognitive Function in Older Men with Low Testosterone and Age-Associated Memory Impairment,” JAMA 2017; 317 (7): 717-27.
  8. Matthew Budoff et al., “Testosterone Treatment and Coronary Artery Plaque Volume in Older Men with Low Testosterone,” JAMA 2017; 317 (7): 708-16.
  9. Craig Cheetham et al., “Association of Testosterone Replacement with Cardiovascular Outcomes Among Men with Androgen Deficiency,” JAMA Int Med 2017; 177 (4): 491-99.
  10. Peter Snyder et al., “Effects of Testosterone Treatment in Older Men,” NEJM 2016; 374 (7): 611-24.
  11. Samantha Huo et al., “Treatment of Men for ‘Low Testosterone’: A Systematic Review,” PLoS One 2016; 11 (9): e0162480.
  12. Andrew Hoover and Jeffrey Kirchner, “Does Your Patient Really Need Testosterone Replacement?” J Fam Practice 2016; 65 (12): 864-75.

4 thoughts on “I’m skeptical about … testosterone therapy.

  1. It’s discouraging to see the medical profession making the same type of mistake with males that they made with female HRT, namely throwing the baby out with the bath water. In females, the optimal HRT treatment likely involves low-dose transdermal estradiol (E2), which gives nearly all of the benefits of higher-dose HRT while minimizing the risks. Similarly, in males, risks can be minimized with relatively moderate doses of testosterone while gaining most of the benefits.

    In females, estradiol levels drop from above 50 pg/ml to <10 with menopause. In this case, a low dose Menostar or Climara estradiol transdermal patch can eliminate nearly all post-menopausal symptoms while also preventing significant bone loss, and with very little if any increased risk of breast and other cancers, and no increased risk of stroke or CVD.

    In males, T decline is more gradual, but eventually results in reduced robustness (frailty, bone and muscle loss), reduced libido, and other effects (one’s maleness relies on relatively high levels of T compared to a female’s reliance on relatively low levels of E2). This makes it more difficult to determine when TRT would be helpful, and also involves relatively larger doses of T to correct a deficiency. However, those are not good reasons to deny TRT to aging males.

    Typically, TRT is initiated in response to low-T related symptoms, plus a measurement showing relatively low serum (free or bioavailable) T levels, and elimination of other possible causes of the symptoms. Under these conditions, if the symptoms are corrected by modest TRT, then this confirms the diagnosis of low T. In fact, in such a case it is irrelevant what you happen to believe is a low T level, or whether you believe that most aging males can do without TRT. The truth is that there are plenty of us aging males who have been shown to benefit from TRT, and your all-or-none approach to the problem is neither wise nor helpful.

    With respect to dosing, I currently use an insulin syringe to inject a small amount (about 4 mg/d T) of testosterone enanthate or cypionate every 3 days. This approach eliminates large fluctuations in T levels, produces a reliable increase in serum T (unlike transdermal products), is safe for my wife (no skin transfer), and is relatively inexpensive. Note also that this (more sensible) approach is not even shown in your table of FDA-approved formulations. This omission exemplifies a huge problem with most TRT studies: it is not clear exactly how much T is getting into the subject’s system, and what the effects are on both T and E2 levels (where E2 is produced from T). If we were talking about any other drug, we would discard much of the research on TRT simply on the basis of not knowing how much of the drug (T) was getting into the subjects, and of inadequate monitoring of resulting blood levels (of both T and E2). But with TRT, all these treatment types and effects get lumped together, and reviewers are left “skeptical about TRT”.

    Unlike ERT in females, TRT affects at least two major hormone levels in males: T and E2. Estradiol in males affects libido, bone density, cardiovascular health, etc., in ways similar to its actions in females. In fact, E2 levels in males are much higher than in postmenopausal females, and are critical to aging male health. However, E2 effects (like many other things) involve a U-shaped risk curve, where both too low and too high doses increase risk. In males, T production is adjusted by the body in response to changes in serum E2, thereby protecting us from too low or too high E2. However, with aging, it is common to find overweight males with too much E2 (stroke risk), or underweight males with too little (frailty risk), since E2 in males is primarily produced via aromatase in fat tissue.

    Adding TRT to this mix will generally raise E2 levels, and also overwhelm the body’s ability to control E2 via adjustments to T production. This means that TRT requires not only monitoring T levels, but also E2 levels, and adjusting T dose (or one’s body fat levels) accordingly. For aging males with relatively high SHBG (which affects free T and E2 levels), E2 serum levels near 30 pg/ml are optimal for protecting bone density (and reducing low-E2 symptoms) without raising stroke risk. (The more common recommendation of 20 pg/ml E2 for males applies to younger males with lower SHBG.) For T, free T levels of 100 pg/ml or more are protective and will reduce most low-T symptoms. My dose of about 4 mg/d T (with normal BMI and moderate fitness), for example, is enough to maintain E2 just above 30 pg/ml and free T just above 100 pg/ml. This also has maintained my bone mass, reduced nighttime joint/muscle pain, and generally kept me more robust, with a healthy libido.

    Having studied the science and worked through the various issues with my own TRT, I’m convinced that some form of moderate dose TRT is probably warranted (at some point) in most aging males. However, I would agree that, with few exceptions, the current medical profession, as exemplified by your post, is not prepared to prescribe TRT, being ill-informed about optimal methods of delivery, the role of estradiol in males, and the importance of measuring and targetting optimal levels of both T and E2.

    • Thanks for taking the time to write and providing insight regarding the interplay between testosterone and estradiol. The purpose of the post, however, was to summarize the results of the “T-Trials.” Having read all of them, I come away less than impressed. Even as regards to sexual functioning, men will likely receive more bang for their buck (pun intended) with Viagra than T supplements.

      I’m afraid I also don’t agree with your assessment that “if the symptoms are corrected by modest TRT, then this confirms the diagnosis of low T.” There is a huge placebo effect associated with treatment (even more so when the drug is administered via injection as you are doing), which is why the diagnosis requires both the presence of symptoms AND a low serum T. There are definite harms associated with excess T. Doctors and patients both tend to overestimate benefits and underestimate harms associated with treatment of all sorts of conditions.

      I’m also baffled as to why so many people insist that after just a few decades of scientific inquiry, we are now adequately prepared to manipulate the hormonal system of checks and balances that it took nature thousands of years to develop. Before turning to supplements, why not seek to maximize our own potential through high-intensity interval training and a healthy diet? Unfortunately, TRT is being marketed by as a “natural” way to overcome decades of poor health choices. Nothing could be further from the truth.

      Lastly, while I agree that the majority of doctors are “ill-informed about optimal methods of delivery, the role of estradiol in males and the importance of measuring and targeting optimal levels of both T and E2,” I disagree that this is the reason the medical profession “is not prepared to prescribe TRT.” Rather, I believe it is simply due to a lack of evidence of benefit. The results of the T-Trials seem unlikely to convince physicians without a proprietary stake in the game otherwise.

      • The guidelines look for low T plus symptoms to justify TRT, which is then confirmed if the symptoms are addressed by TRT. You don’t get to change the rules and discard everyone’s results by claiming they might be due to a placebo effect!

        In my case, for example, I was dealing with reduced bone density, reduced libido, and nightly pain causing insomnia. This was addressed with modest, inexpensive TRT bringing T and E2 to youthful low-normal levels. Yes, as with postmenopausal women, one could address these symptoms with multiple drugs, but that would be foolish, inviting a host of side effects from the “unnatural” drugs, when a modest dose of “natural” T for someone with low T may address a multitude of issues. Further, if you look at studies measuring the T levels in aging males, lowish T (and/or E2) is strongly associated with lower bone density, higher frailty, more pain and insomnia, lower libido, etc., so it is not surprising that modest TRT can correct a deficiency in some males.

        There is nothing “natural” about what happens to one beyond about age 50 or so. Our distant ancestors rarely lived beyond that age, meaning that older bodily systems (including sex hormone production) degrade in varied and complex ways that were not subject to natural selection. In other words, the decline in T (and E2) production with age is not a clever evolutionary strategy, but more like one of the many things that begin to go wrong with a car driven well beyond its intended life span. A “natural” approach might be to drive the increasingly frail car as little as possible, on the safest and smoothest rodes, in good weather only, etc. But a smarter approach would be to figure out what’s going wrong with the car and fix it, attempting to keep it in peak operating condition for as long as possible.

        We are going to fix the aging human body to extend its vigorous lifespan. Part of that may include modest sex hormone supplementation, or, better yet, a fix to whatever the underlying process is that results in declining T levels and vigorousness with age. At this time, TRT, done correctly, is a cheap way to fix a host of age-related issues in some men. Future approaches are likely to be even more complex and effective, based on a much more detailed understanding of aging and the role of sex hormones. Suggesting that the best approach is to avoid TRT and focus on food and exercise is misleading, and of limited value: some men do benefit now from TRT (including fit older men like myself), and future life extension efforts will overwhelm the modest effects of food and exercise.

        btw, I made it clear that I don’t take “excess T”, but rather an amount that produces youthful low-normal levels known to be associated with healthy males. Your focus on “excess T”, including a large part of your original post, is also misleading. Done properly, TRT does not increase risks significantly. For example, there is no evidence that older males with E2 levels near 30 pg/ml are at increased risk of stroke, and good evidence that such males have higher bone density than those with less than 30 pg/ml E2. Doing TRT properly also means not making changes to blood levels of T and E2 too quickly, which is another mistake made in the TRT studies (i.e., not reporting changes in T and E2 levels over time). As noted before, I don’t think there is much to learn from these “T-Trials” since they are generally so poorly done. On the other hand, if you find just one fit person like myself who is well-monitored and healthier with modest TRT than without, then you’ll need to explain that, or your hypothesis is disproved.

        • Again, thanks for the response. I’m afraid that we will have to agree to disagree. Writing off the entire series of T-Trials because they “are so poorly done” is simply self-serving. They constitute the best available evidence. Although I am glad that you have found a treatment that works well for you, I’m afraid that an anecdotal series with an n of 1 is not evidence of anything. I have no dog in this fight. The name of my blog is Skeptical EP (emergency physician) and when it comes to TRT there is still plenty to be skeptical about.

Leave a Reply

Your email address will not be published. Required fields are marked *