Testosterone Therapy and Aging: Trial Evidence and Risks

What the US label says

The AndroGel 1.62% label approves testosterone replacement in men with conditions caused by a deficiency or absence of the body's own testosterone, such as testicular failure or pituitary-hypothalamic disease. It states that safety and efficacy in men with "age-related hypogonadism" (also called late-onset hypogonadism) have not been established. [2] So normal age-related decline is not an approved use. Label: approved medicine for diagnosed hypogonadism; not established for age-related decline. For background on low testosterone, see testosterone and low testosterone.

The Testosterone Trials (older men, 1 year)

A set of linked trials assigned 790 men aged 65 or older with testosterone below 275 ng/dL and symptoms to gel or placebo for one year. Raising levels to the mid-normal range of young men gave a moderate benefit for sexual function, some benefit for mood and depressive symptoms, and no benefit for vitality or walking distance. The trial was too small to judge risks. [1]

Companion results from the same program:

• Memory. Among 493 men with age-associated memory impairment, testosterone did not improve memory or other cognitive functions. [3]


• Bone. Spine and hip bone density and estimated bone strength rose; the authors said a larger, longer trial was needed to know whether fractures fall. [4]


• Anemia. Hemoglobin rose in more men with anemia, but in 6 men who were not anemic at the start it rose above 17.5 g/dL. [5]


• Coronary plaque. Noncalcified plaque volume in heart arteries increased more with testosterone than placebo (170 men enrolled); the authors said larger studies were needed to understand the clinical meaning. [6]

TRAVERSE (heart and long-term safety)

TRAVERSE enrolled 5,246 men aged 45 to 80 with symptoms of hypogonadism, two low testosterone readings, and existing or high risk of cardiovascular disease. Major cardiac events (cardiovascular death, heart attack, stroke) occurred in 7.0% with testosterone and 7.3% with placebo (hazard ratio 0.96), meeting the trial's "not worse" target after a mean follow-up of 33 months. More atrial fibrillation, acute kidney injury and pulmonary embolism occurred with testosterone. Funding included AbbVie. [7]

• Fractures. Clinical fractures occurred in 3.50% with testosterone and 2.46% with placebo (hazard ratio 1.43); testosterone did not lower fractures. [8]
• Prostate. High-grade prostate cancer and other prostate events were low and did not differ significantly, in men carefully screened to exclude high prostate cancer risk; PSA rose more with testosterone. [9]

Label overall: mixed. A real but modest benefit for sexual function in men with low levels; no benefit shown for memory or vitality; and heart-rhythm, clot, kidney and fracture signals alongside neutral major cardiac events.

Risks listed on the label

• Blood pressure and clots. The label says the gel can increase blood pressure and that clots in veins and lungs have been reported with testosterone products. [2]
• Women and pregnancy. The label says it is not indicated for women and is contraindicated in pregnancy, including exposure from skin contact. [2]
• Prostate and breast cancer. It is contraindicated in men with breast cancer or known or suspected prostate cancer. [2]
• Thick blood. Red blood cell counts can rise, which the label says may require lowering or stopping treatment. [2]

What to read next

For midlife hormone changes in both sexes, see menopause, midlife hormone changes and healthy aging. For bone, see bone health, osteoporosis and aging. For over-the-counter "boosters", see testosterone boosters claim check. For how hormone tests are used and misused, see hormone and nutrient testing in aging. Questions about your own levels belong with your clinician: talking with a health professional.