Testosterone: Reading the Panel

Most low testosterone in men is secondary, and most secondary hypogonadism has a cause that resolves on its own treatment. Replacing the hormone before identifying that cause suppresses an axis that would often have recovered.

Written for clinicians and health professionals advising on male endocrine health. Scope and prescribing authority vary, and the referral section marks where handing over is the answer.

Free and bioavailable testosterone calculator

Vermeulen calculation from total testosterone, SHBG and albumin — association constants 1.0 × 10⁹ L/mol for SHBG and 3.6 × 10⁴ L/mol for albumin. Use the presets to see how far SHBG moves the usable fraction at a fixed total.

Scope of the calculation

This is arithmetic on three entered values and nothing more.

It does not diagnose. Published equations disagree with one another and reference ranges are laboratory-specific, so a value here is not comparable to one from a different calculator or lab. Interpretation depends on symptoms, draw conditions, gonadotropins and the rest of the panel.

Treat it as a reading aid for a report you already have in front of you.

How much of a total testosterone is actually usable?

Testosterone circulates in three states and only two reach tissue, so a total without SHBG and albumin cannot tell you what the patient can actually use.

Bioavailable testosterone = free + albumin-bound, typically 35–55% of total where SHBG sits mid-range. That share is itself what SHBG moves, so it runs well outside those figures at either edge — the examples below span roughly 30% to 70% on an unchanged total. Because the free fraction is so small, a modest shift in SHBG moves it proportionally far more than it moves the total.

The same total, three different patients

Total testosterone 500 ng/dL and albumin 4.3 g/dL in all three, varying only SHBG. The figures are the calculator’s own output, computed on this page rather than transcribed onto it.

Three identical totals. A 2.4-fold difference in usable hormone. The man at SHBG 70 carries 148 ng/dL bioavailable — the same as a man whose total reads 300 ng/dL with mid-range SHBG, which is the threshold AUA uses. His report says 500.

Measurement hierarchy

• Equilibrium dialysis (or ultrafiltration) is the reference standard for free testosterone. Reserve for cases where the calculated value and the clinical picture disagree.
• Calculated free testosterone from total T, SHBG and albumin is acceptable, cheap and widely available — the Vermeulen equation is the usual implementation, and the calculator above.
• Direct analog free-T immunoassay: do not use. Unreliable across the clinically relevant range and explicitly discouraged in guidance.
• Free androgen index (total T ÷ SHBG) performs poorly in men and is not a substitute for a calculated free testosterone.
• LC-MS/MS for total testosterone where available; immunoassays lose accuracy at exactly the low end that decides management. Male-range estradiol also sits below the reliable range of most immunoassays.

Two caveats for follow-up

Calculated free testosterone is not portable between laboratories. Reference ranges are assay- and equation-specific, and published equations disagree with one another. Follow trends within one lab; a rise that coincided with a change of lab is not a finding.

The calculation inherits its inputs’ errors. A low-end immunoassay total, or an albumin depressed by acute illness, propagates straight through.

When is a low testosterone actually a diagnosis?

Hypogonadism needs consistent symptoms and unequivocally low testosterone on two morning draws. A single low value in someone whose symptoms have another explanation is not a diagnosis, and treating that number is the commonest error in this space.

• Two morning fasted draws, 07:00–10:00, separate days, at least a week apart. A single low value is not a diagnosis.
• Not during acute illness, nor after a hard training block, an acute fast, or sleep deprivation — each produces a transient nadir.
• Order SHBG and albumin with the total, plus LH and FSH on the confirmatory draw.
• Thresholds differ by body: AUA uses total T <300 ng/dL; the Endocrine Society defers to the assay’s own reference range in healthy young men.

A presentation built only from the right-hand column has a low pre-test probability. Depression, sleep debt, iron deficiency, thyroid disease and sustained overwork all produce that list with entirely normal androgens.

Is it primary or secondary?

LH and FSH decide it. Low or normal gonadotropins with low testosterone means secondary — the large majority, and usually reversible. Elevated gonadotropins mean the testis has failed. Everything downstream follows from which of those it is.

Why sequence matters

Exogenous testosterone suppresses LH and FSH within weeks, ending intratesticular testosterone production and spermatogenesis. In someone whose axis was suppressed but intact, that converts a reversible state into an iatrogenic one, with recovery measured in months to over a year and no guarantee of return to baseline.

What causes secondary hypogonadism?

Nearly all of it resolves into a short list, and most of that list is functional and reversible: obesity, energy deficit, opioids, glucocorticoids, prior anabolic use, sleep apnea. Secondary hypogonadism is a finding, not a diagnosis.

Secondary — functional

Secondary — organic

Primary — testicular

Ask about this every time

Prior anabolic steroid, SARM and research-peptide use produces a picture indistinguishable from any other functional secondary hypogonadism, and is frequently not volunteered — including by clinicians, and by people who would not describe what they took as a steroid. Ask directly, and ask about “peptides” and online purchases as a separate question, because people categorize those differently.

Why treat the cause before the hormone?

In functional secondary hypogonadism the cause is the treatment, and it doubles as the diagnostic test — the response confirms the mechanism in a way no single value can.

Reassess at 3 months minimum, 6 preferred — total T, free or bioavailable T, LH, FSH, SHBG, IGF-1. Self-reported intake routinely under-counts a deficit.

What a positive response looks like

Rising LH ahead of rising testosterone is the signature of an axis coming back online, and the most reassuring finding in a follow-up panel.

Because SHBG falls as an energy deficit resolves, part of any rise in total testosterone reflects the binding protein rather than gonadal output. Judge recovery on free or bioavailable testosterone and on LH, not on total alone.

Which treatments preserve the axis, and which replace it?

Organized by the question that decides between them: whether the axis is preserved or replaced. For a confirmed secondary pattern with fertility intact, the gonadotropin-directed agents are the conservative choice and are frequently sufficient.

Axis-preserving

Replacement — accepts axis suppression

If fertility is an active goal

Obtain a baseline semen analysis first, not after. Use hCG or a SERM rather than testosterone; where spermatogenesis needs further support, hCG combined with FSH is the established regimen. Refer to reproductive urology before committing to a long course.

Monitoring on replacement

• Testosterone and hematocrit at 3, 6 and 12 months, then annually. Reduce dose, change formulation or stop if hematocrit exceeds 54%.
• PSA and DRE at baseline, and at 3–12 months if over 40 with a baseline PSA above 0.6 ng/mL.
• Blood pressure each visit — more often on oral undecanoate. Lipids annually; HDL suppression is expected and dose-related.
• Symptoms. No benefit at adequate levels means the diagnosis was wrong — stop rather than escalate the dose.

On cardiovascular safety, TRAVERSE randomized 5,246 men aged 45–80 with hypogonadism and established cardiovascular disease or multiple risk factors, and found testosterone gel non-inferior to placebo for major adverse cardiac events over a mean 22 months. It also found higher rates of atrial fibrillation, pulmonary embolism and acute kidney injury in the treatment arm — which belongs in the consent conversation. Note the population before generalizing: it says nothing about long-term therapy in a healthy 40-year-old.

Set exit criteria before starting

Decide in advance what failure looks like and what happens then. A trial with a defined endpoint is a reasonable clinical act; open-ended therapy started on ambiguous numbers, in someone whose axis was likely recoverable, is how people end up dependent on a treatment they never needed.

What does the evidence not support?

Growth hormone and its secretagogues move lean-mass numbers largely through fluid, and SARMs, peptides and testosterone boosters have no reliable signal. Included because this is where a motivated, well-read client will have encountered the most confident and least reliable information.

• Growth hormone for body composition. In healthy adults it raises lean body mass measures largely through fluid retention, without demonstrated strength or exercise-capacity benefit, and with meaningful rates of edema, arthralgia, carpal tunnel and glucose intolerance. Distributing hGH for non-approved indications is separately prohibited under federal law.
• Growth hormone secretagogues. Ipamorelin, CJC-1295 and oral MK-677 reliably raise GH and IGF-1; the composition change is substantially fluid, insulin resistance and appetite increase are consistent, and functional strength benefit is not established. Tesamorelin is the exception with real data — for HIV-associated lipodystrophy, which doesn’t generalize.
• SARMs. No approved agent in the class. Documented liver injury, marked HDL suppression, and axis suppression reproducing the original problem. Chemical analysis of 44 products sold online found only 52% contained any SARM and 41% matched their label, with undisclosed anabolic steroids in a substantial fraction.
• Peptides. BPC-157 and neighbours have no controlled human efficacy trials. Note the regulatory churn: FDA placed BPC-157 in Category 2 in 2023, removed it in 2026, and a PCAC vote in July 2026 recommended it for the 503A Bulks List — a recommendation that carries no legal effect pending rulemaking, and says nothing about efficacy. Expect it quoted as though it does.
• Testosterone-boosting supplements. Tribulus, D-aspartic acid, fenugreek, ashwagandha — small, inconsistent effects mostly confined to men deficient or stressed at baseline. Correcting a real deficiency works; nothing here raises testosterone in a replete man.

Creatine monohydrate at 3–5 g/d is the honourable exception among supplements — replicated efficacy for lean mass and strength, excellent safety record. It doesn’t raise testosterone; it works on its own terms.

When should you refer instead of manage?

Referral is decided by which findings require diagnostic workup and prescribing authority — not by how much you know. Scope of practice differs across the professionals who advise on this.

Within scope for most health professionals: ensuring the panel was drawn correctly and completely, recognizing that a normal total with high SHBG deserves a second look, addressing sleep, energy availability, training load and alcohol, and preparing someone to have a better conversation with their clinician.

How do you explain this to a client?

Four translations do most of the work in practice, and each one replaces a number with a mechanism.

The three questions worth leaving someone with: what do you think is causing this, what would we try first, and how will we know whether it worked?

Sources

Each reference is listed against the specific claim it supports rather than as general reading, so a disagreement can be traced to the work it comes from.

The last two disagree with each other about how well this calculation works, and that is left standing rather than resolved. Measured against direct equilibrium dialysis, the Vermeulen result runs roughly 19% high at the median; the model proposed to replace it ran about twice dialysis in the same comparison. The equation above is the best-validated simple estimate available, which is a narrower claim than being correct.

1. A critical evaluation of simple methods for the estimation of free testosterone in serum

   Vermeulen A, Verdonck L, Kaufman JM

   J Clin Endocrinol Metab. 1999;84(10):3666–72

   The equation implemented above, and the association constants it uses — 1.0 × 10⁹ L/mol for SHBG and 3.6 × 10⁴ L/mol for albumin.

   PMID 10523012

2. Testosterone therapy in men with hypogonadism: an Endocrine Society clinical practice guideline

   Bhasin S, Brito JP, Cunningham GR, et al.

   J Clin Endocrinol Metab. 2018;103(5):1715–44

   Free testosterone is worth determining in men whose total sits near the lower limit, or who have a condition that moves SHBG. The conditions named in the panel section are this guideline’s.

   PMID 29562364

3. Evaluation and management of testosterone deficiency: AUA guideline

   Mulhall JP, Trost LW, Brannigan RE, et al.

   J Urol. 2018;200(2):423–32

   The 300 ng/dL figure, as a cut-off in support of a diagnosis rather than a diagnosis in itself; two early-morning measurements before concluding anything; and free testosterone as useful specifically in the equivocal 230–317 ng/dL band.

   PMID 29601923

4. Harmonized reference ranges for circulating testosterone levels in men of four cohort studies in the United States and Europe

   Travison TG, Vesper HW, Orwoll E, et al.

   J Clin Endocrinol Metab. 2017;102(4):1161–73

   Harmonized ranges across 9,054 men: a 2.5th percentile of 264 ng/dL and a 5th of 303. Worth knowing that the widely used 300 threshold lands within a few ng/dL of the 5th percentile of healthy young men.

   PMID 28324103

5. Reassessing free-testosterone calculation by liquid chromatography–tandem mass spectrometry direct equilibrium dialysis

   Fiers T, Wu F, Moghetti P, et al.

   J Clin Endocrinol Metab. 2018;103(6):2167–74

   Measures the calculation against direct equilibrium dialysis: the Vermeulen result ran about 19% high at the median. This is the best available answer to how far off the number above can be.

   PMID 29618085

6. A multi-step, dynamic allosteric model of testosterone’s binding to sex hormone binding globulin

   Zakharov MN, Bhasin S, Travison TG, et al.

   Mol Cell Endocrinol. 2015;399:190–200

   Argues the two-identical-sites assumption underlying the Vermeulen model is wrong, and proposes an allosteric alternative. Included because the disagreement is real and a reader should know it exists.

   PMID 25240469

Professional educational reference, current at the time of writing and based on the guidelines and peer-reviewed research listed under Sources. It is not individualized medical advice and does not establish a clinician–patient relationship. Dose ranges are those reported in published guidelines and the peer-reviewed literature, which this page summarises rather than reproduces; several agents discussed are used off-label or are available only through compounding, and prescribing decisions rest with the treating clinician acting within their own scope and jurisdiction. The calculator performs an arithmetic conversion from values entered and does not diagnose any condition; laboratory reference ranges vary and results are not comparable between laboratories or calculators. Guidelines and compounding regulations are revised — verify current recommendations and product labeling before acting.