Chapter 100
The Testosterone Studies
There is no universally accepted lower limit for a healthy testosterone level in the blood. Proposed appropriate cutoff values range from less than 200 (American Association of Clinical Endocrinology) up to 350 nanograms/deciliter (European Association of Urology). In a trial in which men were first chemically castrated and then gradually given their testosterone back, however, the researchers found clear changes in sexual desire and sexual function only when levels dropped below 100 nanograms per deciliter. Regardless of the cutoff used, the Endocrine Society guideline states that to diagnose hypogonadism, testosterone must be low on two morning measurements, four weeks apart, with symptoms remaining the same. (Testosterone fluctuates depending on the season, from week to week, from day to day, and even from hour to hour, being higher in the morning and dropping by 30 to 40 percent by the afternoon.)
These guidelines are often ignored. In the United States, a study of hundreds of thousands of men who started taking testosterone found that only 10 percent had gotten the recommended second test. 50 percent had only one test, and the remaining 40 percent apparently did not get tested at all. In as many as 77 percent of older men, testosterone can be below 300 on the first test, but after a second confirmatory test there may be only 18 percent of men left, and then only 3 percent if other recommended criteria are followed, such as drawing the blood in the morning.
The vast majority of men who receive testosterone “replacement” therefore do not actually need it. But that does not necessarily mean it doesn’t benefit them. Maybe they have different thresholds, and the extra testosterone could help them even if no deficiency is found. You can imagine that men might feel better just from the placebo effect of taking testosterone, which is why it is so important to study the effect. Researchers intercepted older men who were standing in line for testosterone because they or their doctors were hoping for relief from symptoms such as reduced energy or libido, and randomized them to receive testosterone gel or a placebo gel. The result? The testosterone worked—but so did the placebo, so in the end there were no significant differences.
Even for sexual symptoms, testosterone did not perform better. But weren’t those precisely the symptoms that went along with low testosterone? Yes, but that does not mean low testosterone is the cause. Instead of low testosterone leading to sexual disinterest, perhaps sexual disinterest leads to less testosterone. When men have sex, testosterone levels in their blood can rise so much that their beards grow faster on days when they have sex. Men who resume having sex after a nonhormonal treatment for their erectile dysfunction—for example, with penis pumps or prostheses—increase their testosterone levels by a substantial 450 ng/dl on average. (By contrast, interestingly, men don’t get a testosterone surge when they masturbate. That may be because testosterone rises with “wins,” as in sports. Although sex is “not usually seen as a competition,” the psychology researchers said, “the mental state after coitus could nevertheless resemble that of a winner,” unlike after masturbation.)
Although the study participants had relatively low testosterone values, the selection criteria for the study were symptoms, not specific cutoff values. No wonder, then, that testosterone proved practically useless, since it was given to men who may already have had enough of it. What about a randomized placebo-controlled double-blind trial of men who have symptoms and meet strict criteria for testosterone deficiency? Enter: the National Institutes of Health–funded Testosterone Trials.
The Testosterone Trials
In 2004, the National Academy of Medicine concluded in an authoritative report that testosterone therapy had no proven benefit for any of the aspects of male health that were examined, and that larger, longer, and better studies would be required to make a definitive determination. In response to this mandate, the NIH funded not one, not two, but seven clinical trials across a dozen academic centers, in which men were randomized to receive testosterone or a placebo for twelve months. Participants had to be at least 65 years old, have a measured and confirmed low testosterone level (< 275 ng/dl), and have a symptom such as reduced vitality or libido. They examined how “correcting” testosterone levels to those of young, healthy men affected seven clinical endpoints: cognition, vitality, physical function, sexual function, anemia, bone health, and cardiovascular health.
The great hope was that testosterone replacement would improve brain function. Population studies had established an association between lower testosterone levels and a higher risk of cognitive impairment and dementia. Prostate cancer patients receiving long-term androgen deprivation therapy appeared to have a higher risk of later dementia. But in the Testosterone Trials, “correcting” testosterone did not improve memory or other cognitive functions, and the same was found by a meta-analysis of more than a dozen other randomized controlled testosterone trials. An editorial in the Journal of the American Medical Association concluded that these “compelling, definitive results confirm that testosterone treatment does not improve cognitive function in older men.”
Physical function and vitality scores were not improved by testosterone either. This is consistent with dozens of other randomized controlled trials that showed little or no impact on physical function, depressive symptoms, energy, or vitality. No wonder that within a year, 80 to 85 percent of men stop taking testosterone. A study of nearly 16 000 patients found that in the first three months, about 50 percent of men stopped treatment with topically applied testosterone, and about 70 percent discontinued injections. That supplemental testosterone does not provide noticeable benefit makes sense if lower testosterone is actually more of a consequence than a cause of obesity, lack of exercise, and chronic disease.
At least the Testosterone Trials found that testosterone improves bone density. However, when you look at the ten existing randomized controlled trials on testosterone therapy and bone health together, unfortunately no overall bone benefit can be demonstrated. For sexual symptoms, though, the opposite may be the case.
Sexual function was temporarily better, but at the end of the year there was no significant difference between the placebo group and those who had received the real stuff. In contrast, most (ten of thirteen) high-quality studies found that testosterone therapy increases sex drive in men with low levels, and seven of twelve studies clearly found that it improves erectile function. The effect, however, was so small that it was described as “marginal.” Testosterone replacement products may help mild cases of erectile dysfunction, but they work only a fraction as strongly as medications such as Viagra. I found it mind-blowing to learn that eunuchs had an active sex life even though they had been castrated as boys. In one experiment, severely hypogonadal men (including those surgically bilaterally castrated) with a low testosterone level of 25 ng/dl not only got erections from an erotic film, but the erections lasted longer than in men in the control group with healthy testicles!
Low libido, however, does seem to truly be a symptom of low testosterone. Men with demonstrably low testosterone who suffer from reduced sexual desire and want to increase their sex drive may therefore be candidates for testosterone treatment after weighing the risks. There are pills, patches, gels, injections, implanted pellets, and even buccal tablets that you stick onto the gums. All of them appear to be similarly effective, but injections are the cheapest at 150 dollars per year, whereas some topical preparations cost more than 2000 dollars. One more note: It can take weeks for testosterone levels to rise, and months for symptoms to reverse, but the placebo effect can set in immediately. What do the downsides look like?