Now that the quintessential male steroid is a $2 billion market - with nearly 3 percent of men over 40 estimated to be taking it in 2011 - there are signs of trouble.
Two recent studies suggest testosterone supplementation raises the risk of heart attack, stroke, and death. The Food and Drug Administration has begun an investigation, and law firms, including one in Philadelphia, have begun filing product-liability suits.
In another echo of the female hormone saga, the federal government has stepped in to do the kind of costly, rigorous, placebo-controlled clinical trial that drug companies have no incentive to do. The $50 million Testosterone Trial is now wrapping up and will present results next year, said the leader, University of Pennsylvania endocrinologist Peter J. Snyder.
But size and duration do matter - in life and in research.
The Women's Health Initiative - with 16,600 women taking estrogen-progestin pills for five years - had what scientists call "statistical power." That's why, despite criticism, the results have held sway.
The T Trial - 788 men on hormone or placebo for a year - is designed to detect benefits, but not risks.
"It is nowhere near large enough to determine any important risk," Snyder said. "Not on prostate cancer, or heart disease."
A government advisory panel said in 2003 that a more definitive trial would be premature, given testosterone's scant research record. Still, if pluses emerge without the proper counterbalance of minuses, could that fuel misguided use of testosterone?
"This is the problem the [advisers] wrestled with," said Mark S. Litwin, a University of California at Los Angeles urologist who was on the panel. "There will never be a randomized controlled trial in every area of medicine where one is needed. Sometimes we have to make do with the evidence we have."
Testosterone is essential for male growth and masculine characteristics such as beards, deep voices, and muscle bulk. For decades, testosterone injections have been vital for men with hypogonadism - little or no testosterone - due to birth defects, chemotherapy, infection, or other causes.
But the concept of age-related testosterone deficiency, or Low T, remains controversial.
Unlike estrogen, testosterone ebbs but doesn't bottom out in middle age. Blood levels fluctuate widely during the day and there is no accepted definition of "low" in older men. Some studies suggest boosting it can increase PSA, the marker of prostate cancer; aggravate sleep apnea; enlarge male breasts; and raise stroke risk by overproducing blood-thickening red blood cells.
Testosterone isn't readily absorbed in pill form and shots are unpleasant, so approval in 2000 of the first rub-on form, Solvay Pharmaceutical's Androgel, was a big deal. (Solvay donated its gel to the T Trial.)
"The introduction of topical gel formulations at the turn of the [21st] century has been a milestone in the evolution of the testosterone replacement therapy market," said Global Industry Analysts Inc. in 2013.
Sales - more than $2 billion last year, according to IMS Health - are forecast to hit $5 billion by 2018.
The FDA says testosterone should be prescribed to men with low levels and symptoms that may be the result of sexual dysfunction or fatigue.
However, an Endocrine Society task force that issued guidelines in 2006 said the level below which symptoms occur "is not known," and even the panel couldn't agree what threshold warranted treatment. (Some on the panel, which included Snyder, said treat below 300 nanograms per deciliter of blood, while others favored 200 ng/dl. The T Trial set a cutoff of 275 ng/dl. A young man would hover at 800 ng/dl.)
One thing the panel agreed on: A diagnosis should be based on "unequivocally low" readings from tests on two separate mornings.
Guidelines, of course, can be ignored.
A recent analysis of a national insurance database found a quarter of testosterone users had no tests of their hormone levels. Between 2001 and 2011, that usage tripled, to 2.9 percent of men 40 and over.
A British analysis of Medicare claims over the last decade found not only lots of testosterone-takers with no tests of their levels, but also 9 percent with tests showing normal or even high ranges.
The authors also found testosterone use climbed steadily in the United States while remaining comparatively flat in Britain. The disparity, they speculated, may reflect the fact that Britain doesn't allow direct-to-consumer marketing of medicines.
Stephen A. Braun, a professional medical writer in Massachusetts, last year ruefully explained how he helped with such marketing by ghostwriting pro-testosterone magazine articles under physicians' bylines, and producing a consumer "education" booklet that lawyers loaded with the wiggle words can and may.
It was, Braun wrote, "an unbalanced presentation of 'facts' that serves primarily to drive people to their physicians seeking the holy grail of 'energy, positive mood, and sexuality' in the form of testosterone."
The thing is, circumstantial evidence suggests testosterone really may be good for some men.
Physicians at the Veterans Affairs Health Care System in Seattle found a beneficial association between testosterone treatment and death over a three-year period ending in 2005: 10 percent of treated men died, compared with 21 percent of untreated men.
Or it might not be good.
A recent analysis that used a national VA database linked testosterone to bad outcomes - namely, more heart attacks, strokes, and deaths.
The FDA cited that analysis in January when it announced a review of the safety of testosterone products.
So did the Philadelphia law firm of Ross Feller Casey, which has so far filed four lawsuits on behalf of men who suffered heart attacks or stroke while taking testosterone.
They are not alone.
"We've screened more than 400 men" who want to sue, said J. Benson Stewart, a Tampa, Fla., lawyer who has filed on behalf of 18. "We're seeing an increase daily just because there's more awareness."
Hormones work, for good or ill, on tissues throughout the body.
In the Women's Health Initiative, the far-ranging effects of hormone pills showed up as slight increases in the risk of heart attacks, strokes, lung blood clots, breast cancer, incontinence, and dementia, and slight decreases in hip fractures and colon cancer.
The T Trial has enrolled men who have trouble with one or more thing that testosterone might improve: walking, vitality, sexual function, memory, bone density, red blood count (anemia), artery hardening.
The men - recruited and evaluated at 12 medical centers across the country - had to have low morning testosterone measurements on two successive tests. To avoid unnecessarily exposing men to possible risks, they were rejected if they had certain illnesses, including prostate cancer, urinary problems, recent heart attack or stroke, or major psychiatric disorders.
"We had to screen over 50,000 men in order to get 788," Snyder said. "We had very strict criteria."
For Snyder, 73, the T Trial is a career capstone - although he originally favored a larger, more definitive study.
He is philosophical about how others may spin the results.
"There are fads in medicine," he said. "When I first went into endocrinology, the fad was thyroid hormone. Everyone who had any problem . . . was treated with thyroid hormone, even if their thyroid was normal.
"Yes, there is a testosterone fad now. The results can be misused, but lots of things in this world are misused. I wish I could control those."