Total vs. Free Testosterone — Which Result Actually Matters?

Understand

Overview

A total testosterone result is the first number nearly every man checking his hormone status looks at — and in most cases, it really is enough for an initial assessment. The problem arises in a specific, but far from rare, situation: when a person's concentration of the transport protein SHBG (sex hormone-binding globulin) deviates significantly from average. In that case, total testosterone — the sum of the protein-bound and free fractions — stops faithfully reflecting how much hormone is actually available to tissues, and interpreting the result without accounting for free testosterone can lead to mistaken conclusions in either direction.

In the blood, testosterone doesn't circulate in a uniform state. Roughly 40–60% is tightly bound to SHBG — a complex that's practically unbreakable under physiological conditions and biologically inactive, since it can't freely cross into cells. Another roughly 33–54% is loosely bound to albumin — a bond weak enough that it dissociates quickly enough in tissue microcirculation for this fraction to be partly biologically available (hence the concept of bioavailable testosterone, the sum of free and albumin-bound). Only a small fraction, usually 1–4% in men, circulates completely free, unbound to any protein. It's this free fraction — and, more broadly, the bioavailable fraction — that's capable of crossing cell membranes and producing a biological effect in target tissues.

Why does this matter in practice? Because SHBG isn't a fixed value — its concentration changes with age, body weight, thyroid function, liver function, and many other factors, and these same factors don't necessarily move total testosterone in the same direction or at the same pace. Obesity and insulin resistance usually lower SHBG, which, with total testosterone unchanged, means more free hormone than the total number alone would suggest — but obesity itself also lowers testosterone production, so the net effect can be unclear. Hyperthyroidism, liver cirrhosis, estrogen use, and simply aging raise SHBG, which can mask a genuine deficiency of biologically active testosterone in a man whose total result looks technically normal. As a result, two men with identical total testosterone — say, 450 ng/dL — can have very different amounts of genuinely available hormone if one has SHBG around 20 nmol/L and the other 80 nmol/L.

So when does free testosterone actually change a clinical decision, and when is it an unnecessary extra cost? In most cases — a young, lean man with an unambiguously normal or unambiguously low total testosterone and consistent symptoms — the total result alone is enough, and SHBG rarely deviates enough to change the picture. Free testosterone (or at least SHBG) has real added value in a handful of specific situations: when the total result is borderline or low-normal and symptoms suggest deficiency; in men who are obese or have type 2 diabetes (lower SHBG that can either overstate real biological availability given a seemingly low total result, or the reverse — mask a deficit); in older men, in whom SHBG systematically rises with age independent of testosterone production itself; with thyroid and liver disease that alter SHBG unpredictably without measuring it; and with medications that affect SHBG (including some anticonvulsants and certain hormone therapies). Endocrine Society guidelines explicitly recommend reaching for free or bioavailable testosterone precisely under these clinical conditions — not routinely for everyone tested.

A separate, equally important thread — and one more often overlooked than the question of 'should you test free testosterone' — is what method was used to measure it. Not all free testosterone results are equal: the gap between a trustworthy and a worthless measurement can be bigger than the gap between two labs measuring total testosterone. This piece focuses on exactly those two questions — when free testosterone genuinely adds something to the clinical picture, and how to check whether the result you're holding can even be interpreted at all.

A good illustration of why a single free-testosterone cutoff can't be used in isolation comes from the European Male Ageing Study (EMAS), a population-based survey of 3,369 men aged 40 to 79 across eight European centers. After testing dozens of candidate symptom-hormone combinations, EMAS investigators found that only three sexual symptoms — reduced frequency of morning erections, reduced frequency of sexual thoughts, and erectile dysfunction — showed a genuine syndromic relationship with declining androgen levels, and that relationship became statistically robust specifically when at least three such symptoms coincided with both a total testosterone below roughly 11 nmol/L (about 320 ng/dL) and a free testosterone below roughly 220 pmol/L (about 64 pg/mL) (Wu et al., 2010). The study's broader, somewhat humbling finding was that most other commonly blamed symptoms of aging men — fatigue, low mood, reduced muscle strength — showed, at the population level, only a weak association with testosterone status once confounders were accounted for. The practical lesson isn't a number to memorize, but confirmation of the logic above: free testosterone earns its place in a work-up specifically when considered jointly with total testosterone and a cluster of consistent symptoms, not as a stand-alone screening test applied to anyone with vague complaints.

A separate complication, upstream of the free-vs-total debate entirely, is that 'normal' itself has historically meant different things depending on which lab ran the assay. Older testosterone reference ranges were often derived from relatively small, inconsistently characterized reference populations using assays that weren't cross-calibrated against each other, so the same blood sample could return meaningfully different 'normal' flags depending on which laboratory processed it. A large harmonization effort pooling four cohort studies from the US and Europe — the Framingham Heart Study, EMAS, and two osteoporosis-focused cohorts, over 9,000 men combined — re-measured a reference subset of samples with a CDC-certified mass-spectrometry method and used the results to standardize values across the original cohorts, deriving an age-stratified reference range: for healthy, non-obese men aged 19 to 39, a harmonized range of roughly 264 to 916 ng/dL (Travison et al., 2017). This matters for the total-vs-free question because any conversation about whether a given SHBG level is high enough to matter only makes sense against a reliably standardized total testosterone value in the first place — a borderline total result measured on an uncalibrated assay can look falsely normal or falsely low regardless of how carefully the free fraction is then calculated.