Does testosterone replacement therapy actually work? What the evidence says about TRT

Testosterone is having a moment. Influencers are telling men that anything below 300 ng/dL means a life of brain fog, joint pain, and despair, and that the fix is a prescription. The actual evidence is narrower, more boring, and more useful than that.

What testosterone actually does

Testosterone is a steroid hormone. Like estrogen and progesterone, it works inside the cell nucleus to change how DNA gets transcribed, which is why it can do dramatic things like turn a non-hair-growing cell into a hair-growing one. It drives the primary and secondary sexual characteristics in men, supports muscle growth, is required for sperm production, stimulates red blood cell production, and plays a clear role in libido and sexual function.

In post-pubescent men, total testosterone usually runs between 300 and 1000 ng/dL. Most clinicians draw the line for "low" at 300. Unlike estrogen in women, which falls off a cliff at menopause, testosterone in men declines slowly and modestly with age. Modeling the NHANES data, the average goes from roughly 400 ng/dL at age 20 to about 300 ng/dL at age 80. There is no manopause. An 80-year-old man can absolutely have the testosterone level of a 30-year-old. An 80-year-old woman not on supplementation cannot have the estrogen level of a 30-year-old.

Who is actually approved for testosterone replacement

The FDA has approved testosterone replacement therapy for two situations: primary hypogonadism (rare genetic conditions that prevent testosterone production) and pituitary hypogonadism (a tumor or stroke knocks out the signal from the brain telling the testes to make testosterone). Less than 1% of men with low testosterone fall into either category. Virtually everyone you hear about getting TRT is getting it off-label. That is legal. It is also a flag that the evidence base for age-related low T is weaker than the marketing suggests.

The measurement problem

Testosterone has a half-life of 10 to 60 minutes in the blood. Most of it is bound to a chaperone protein called sex hormone binding globulin, which keeps it inactive. Only about 2% is free testosterone, the active form. Free testosterone testing is expensive and none of the assays are FDA cleared, so almost everyone is getting total testosterone measured.

Total testosterone has a within-person coefficient of variation of 50% or higher. If your true average is 300, one day you might measure 450 and another 150. It runs about 15% higher in the morning than the afternoon. Endocrinologists recommend not diagnosing low testosterone without at least two measurements, taken on different days, both below 300 ng/dL. This is also why influencer claims like "I raised my testosterone 48% in 21 days" are essentially meaningless. That is within the noise.

One strange wrinkle: current smoking raises sex hormone binding globulin and therefore raises total testosterone, even though it is not raising the active hormone. Population-level testosterone has been declining about 1% per year, and one explanation is simply that fewer people smoke.

What symptoms actually track with low testosterone

A New England Journal of Medicine survey study of 3,369 European men ages 40 to 79 looked at a wide range of symptoms — depression, brain fog, fatigue, sexual function — and correlated them with measured testosterone levels. After a fair amount of statistical work, the only symptoms that consistently tracked with lower testosterone were sexual ones: decreased morning erections, decreased sexual thoughts, and erectile dysfunction. Psychological symptoms did not correlate. The influencer line that being below 300 means you ache, can't focus, and can't lose weight is not what the data show.

Then there is the question of replacement. The Testosterone Trials enrolled 788 men over 65 with total testosterone below 275, randomized against placebo. Men on testosterone had a 40% increase in sexual activity frequency, a 25% increase in libido scores, and a 35% improvement in erectile function. Vitality and energy did not improve. Mood and walking distance improved only slightly. The Reddit version of this is consistent: of men on TRT, roughly 20% report no change, 20% feel transformed, and 60% talk almost exclusively about erections and libido.

The risks at normal replacement doses

Most of what people fear about TRT comes from bodybuilders using 10 to 30 times the replacement dose. At those doses you see infertility, liver failure, roid rage, baldness, and cardiovascular disease. At replacement doses, the picture is different. The TRAVERSE trial randomized more than 5,000 higher-risk older men and found no difference in major adverse coronary events over 33 months of follow-up. There were, however, small absolute increases in atrial fibrillation and blood clots in the testosterone arm.

Other documented effects at replacement doses: decreased sperm production (relevant if you are trying to conceive), acne, hair loss, and benign prostate enlargement (not prostate cancer). Testosterone is converted to estradiol in fat cells via aromatase, which is why men on higher doses, and especially those with higher body fat, can develop gynecomastia.

The myths worth ignoring

Zinc is a cofactor in testosterone synthesis, which influencers have turned into "zinc deficiency is causing your low T." About 8% of Americans are zinc deficient, and they are mostly older people with significant comorbidities. If you are not deficient, more zinc will not raise your testosterone, and excess zinc can interfere with copper metabolism.

The 2D:4D finger ratio, supposedly a marker of prenatal testosterone exposure that predicts adult aggression and hormone levels, is one of those things presented as deeply studied. A meta-analysis on it concludes there is no evidence of a relationship between testosterone levels and digit ratios. It is bunk.

Lifestyle interventions can nudge testosterone. Weight loss, including via GLP-1 medications, produces small increases. The effects are smaller than you might hope.

Bottom line

Testosterone replacement therapy is worth considering if two things are true: your testosterone is consistently below 300 ng/dL on more than one test, and you have sexual symptoms that actually bother you. In that group, the trials show real improvements in libido and erectile function. They do not show meaningful improvements in energy, cognition, or mood. If your level is normal, or if you have a single low value with no symptoms, the case for TRT collapses. The risks at replacement dose are modest but real, including small increases in atrial fibrillation and blood clots, decreased fertility, acne, and prostate enlargement. The influencer pitch — that any number below 300 is a crisis and TRT will transform every aspect of your life — is not what the evidence shows.

Here's the "What's the deal with testosterone?" segment from the episode:

I covered this in depth on Wellness, Actually — listen below.

Wellness, Actually Podcast

"What's the deal with testosterone?" — Listen to the full episode, including the week's health news and listener Q&A.