Testosterone: What Your Number Actually Means (and When TRT Is Justified)

There is a version of this conversation that happens thousands of times a day. A man in his forties feels flat — worse energy, less drive, softer body composition. He orders a panel or clicks through a telehealth intake. A number comes back. Somebody tells him it’s low. Within a week there is a vial in his refrigerator.

The problem isn’t that low testosterone is fake. It’s a well-characterized clinical condition with real treatments and real benefits. The problem is that the diagnostic process has been run backward by an industry whose revenue depends on the answer being yes. So it’s worth knowing what your number actually measures — and what the large trials found when they finally asked whether treatment helps.

---

What the Test Actually Measures

Total testosterone — the number on most panels — includes every molecule of testosterone in the sample. But most of it isn’t available to your tissues. Roughly half to two-thirds is bound tightly to SHBG (sex hormone-binding globulin), a transport protein that essentially takes it out of circulation. Most of the rest rides loosely on albumin, and only 1–3% floats completely unbound.

The clinically useful fraction is free testosterone (unbound). And here’s why that matters: SHBG isn’t a constant. It climbs with age, thyroid excess, and liver disease, and it drops with obesity, insulin resistance, and type 2 diabetes.

So two men with identical total testosterone of 380 ng/dL can be in completely different situations. The 70-year-old with high SHBG may have genuinely low free testosterone. The 35-year-old with obesity and low SHBG may have perfectly adequate free testosterone despite an unimpressive total. Ordering total testosterone alone in either man produces a confident-looking number that answers the wrong question.

One caveat: not all free testosterone assays are equal. The cheap direct-analog immunoassays are widely regarded as unreliable. Equilibrium dialysis is the reference standard, and a calculated free testosterone derived from total testosterone, SHBG, and albumin is an acceptable substitute.

---

The Reference Range Is More Contested Than It Looks

For years, every lab published its own range and they disagreed wildly. In 2017, a landmark study in The Journal of Clinical Endocrinology & Metabolism harmonized measurements across four large cohorts covering more than 9,000 men. The normal range for healthy non-obese men aged 19–39 came out at 264–916 ng/dL (the 2.5th to 97.5th percentile).

That’s a nearly fourfold spread among perfectly healthy young men — reason enough to be cautious about declaring a single mid-range number “suboptimal.”

Separately, the American Urological Association settled on 300 ng/dL as a practical treatment threshold — explicitly chosen to balance potential benefit against risk, not because anything biological changes at 299. The Endocrine Society’s guideline is similar in spirit but insists on the point that gets skipped most often: the diagnosis requires both consistently low levels and symptoms consistent with deficiency.

---

Symptoms Are Real but Notoriously Nonspecific

Fatigue, low libido, poor concentration, depressed mood, reduced strength, and increased body fat are all recognized features of testosterone deficiency. They are also features of poor sleep, depression, alcohol overuse, hypothyroidism, anemia, obstructive sleep apnea, and simply being a stressed 45-year-old.

The most specific symptoms — the ones most likely to actually track with testosterone — are the sexual ones: reduced libido, fewer spontaneous morning erections, and erectile difficulty. Fatigue and mood are the least specific, and that ranking turns out to predict exactly what treatment does.

If your levels are low, the next test is LHLH (luteinizing hormone) A pituitary hormone that signals the testicles to make testosterone. Measuring it shows whether low testosterone comes from the testicles themselves or from the signal upstream. and FSHFSH (follicle-stimulating hormone) A pituitary hormone that drives sperm production in men and egg development in women. Measured alongside LH to locate the source of a hormone problem., the pituitary hormones that drive the testicles. High LH with low testosterone points to primary hypogonadismhypogonadism A medical condition in which the body does not produce enough sex hormone — in men, testosterone. "Primary" means the testicles are the problem; "secondary" means the brain signals telling them to work are the problem. — the testicles themselves aren’t producing. Low or inappropriately normal LH points to secondary hypogonadism, a signaling problem upstream, which is where the reversible causes live: obesity, opioids, heavy alcohol use, sleep apnea, anabolic steroid history, high prolactin, and certain medications.

That distinction matters, because secondary hypogonadism is often fixable. A 2013 meta-analysis of 24 trials in European Journal of Endocrinology found weight loss reversed obesity-associated hypogonadism in a dose-dependent way: total testosterone rose by roughly 2.9 nmol/L with low-calorie dieting and 8.7 nmol/L after bariatric surgery, with the size of the weight loss the strongest predictor. For a man whose low testosterone is downstream of adiposity, replacement treats the readout and leaves the cause in place.

---

What Treatment Actually Delivers

The Testosterone Trials (T-Trials) remain the cleanest look at this. Published in the New England Journal of Medicine in 2016, they randomized 790 men aged 65 and older with testosterone under 275 ng/dL and matching symptoms to testosterone gel or placebo for one year.

The results were split. Sexual function improved consistently and significantly — activity, desire, and erectile function all moved. The two other primary endpoints did not: no improvement in six-minute walk distance, none in vitality. Mood showed modest benefit.

That pattern — sexual symptoms respond, energy and physical function largely don’t — is the most useful expectation-setting fact in this field, and it is almost never what men are told at the point of sale.

The 2024 fracture trial, a prespecified TRAVERSE substudy in the same journal, tested whether testosterone would protect bone in over 5,000 men. It didn’t. Clinical fractures were numerically more common in the testosterone group. Testosterone raises bone density on a scan; that did not translate into fewer broken bones.

---

The Safety Picture in 2026

For a decade, testosterone products carried a boxed warningboxed warning The FDA’s strongest drug warning, printed inside a black border at the top of the label. Reserved for risks the agency judges serious enough that prescribers must see them first. about cardiovascular risk built on weak observational data. TRAVERSE was the FDA-mandated trial designed to settle it: 5,246 men aged 45–80 with existing cardiovascular disease or high risk, symptoms of hypogonadism, and two fasting testosterone levels below 300 ng/dL, randomized to testosterone gel or placebo.

Testosterone was noninferiornoninferior A trial result showing a treatment is not meaningfully worse than the comparison — the standard way safety trials are designed. It demonstrates the absence of harm, not the presence of benefit. to placebo for major adverse cardiac events — no excess of heart attack, stroke, or cardiovascular death. On the strength of that, the FDA issued class-wide labeling changes in February 2025 removing the cardiovascular language from the boxed warning, while keeping the “Limitation of Use” statement that these products are not established as safe and effective for age-related low testosterone alone.

The trial was not entirely clean, though, and honest reporting includes the rest of it. Testosterone groups had more pulmonary embolismpulmonary embolism A blood clot that has traveled to the lungs and blocked an artery there. A medical emergency., more atrial fibrillationatrial fibrillation An irregular, often rapid heart rhythm originating in the upper chambers of the heart. It raises the risk of stroke and is a common reason for blood-thinning medication., and more acute kidney injury. In the same 2025 action, the FDA required new or expanded warnings about increased blood pressure across the class, based on ambulatory blood-pressure monitoring studies.

The other risks are well established:

Fertility. Exogenous testosterone suppresses LH and FSH, and with them sperm production. Counts frequently fall to near zero. Recovery after stopping usually happens but can take many months. This is the single most under-disclosed consequence in direct-to-consumer testosterone prescribing.

Erythrocytosiserythrocytosis An abnormally high red blood cell count, which thickens the blood. It is the most common reason to lower the dose of or stop testosterone therapy.. Testosterone stimulates red blood cell production, and a rising hematocrithematocrit The percentage of your blood volume made up of red blood cells. Rising values on testosterone therapy signal that the blood is getting too thick. is the most common reason to reduce the dose or stop. It needs periodic monitoring, not a one-time check.

Prostate. Testosterone does not appear to cause prostate cancer, but PSAPSA (prostate-specific antigen) A blood protein made by the prostate, tracked as a screening and monitoring marker for prostate enlargement and prostate cancer. and prostate symptoms warrant baseline assessment and monitoring, particularly in older men.

Dependence on the therapy. Once you suppress your own production, coming off means a period — sometimes a long one — of feeling worse than you did before you started. Very few men are counseled on this in advance.

---

Where This Leaves Us

The honest position is neither of the two loud ones. Testosterone deficiency is not a fabrication invented to sell injections, and replacement is not a longevity intervention every man over 40 should be on. It is a specific treatment for a specific, verifiable condition.

Verified properly, it reliably improves the sexual symptoms it’s best suited for and looks safer on cardiovascular endpoints than we believed in 2015. Verified sloppily — one afternoon draw, no free testosterone, no LH, no search for the cause — it is a lifelong commitment built on a number that may never have been low.

If you’re going to test, test correctly. Our guide to reading lab results covers why “in range” and “optimal” are marketing distinctions as often as clinical ones.

---

A note on what isn’t settled: The largest open question is whether men with borderline levels (roughly 300–400 ng/dL) and genuine symptoms benefit from treatment. Nearly all the trial evidence enrolled men well below 300, so the group most commonly treated in the real world is the group we have the least data on.