If a patient’s total testosterone comes back mid-range and they still feel terrible, the number didn’t tell you what you needed to know. Total testosterone measures everything in circulation, including the fraction that’s bound up and unavailable to tissue. Free testosterone is what the cell can actually use, and sex hormone binding globulin, or SHBG, is what decides how much of it there is. Order all three or you’re guessing.
I talk about this constantly with providers and I want to be clear up front that I didn’t know any of this either when I started. Nobody taught it. We’re all in this boat together.
Two things are going wrong at once, and they have the same root. In women testosterone is the hormone nobody checks. In men it’s the one we assume we’ve already got handled, so it gets ordered, glanced at, and either waved off or answered with the same 200 mg a week and an aromatase inhibitor. My position is that testosterone is still under-prescribed in men, not over-prescribed. Where it does get prescribed, it’s often done poorly, and a lot of guys quit because they never felt what they were told they’d feel.
Young men have it worst. A man in his twenties or thirties with real symptoms can walk into most clinics, ask about testosterone, and get laughed out of the room. That happens because the panel came back inside a range, and nobody looked further than that.
Why is total testosterone not enough on its own?
Total testosterone is probably the least useful number on the panel. It counts every molecule in the blood, and a large share of those molecules aren’t doing anything. Roughly 45 percent is tightly bound to SHBG, and more is bound to albumin. Bound testosterone can’t cross into the cell and can’t act on a receptor, so a large total number can sit on top of a functionally deficient patient.
Here’s the case I walk providers through. A man comes in with a total testosterone of 500 or 600. Solidly mid-range. Nothing on that panel would make a conventional office look twice. But his SHBG is north of 100, which means most of that testosterone is bound and unavailable. His free testosterone, if anyone had ordered it, would come back below range. He has verifiable low testosterone and it never shows up on paper, because the value that would have caught it was never drawn.
That patient gets told his labs are normal, and he walks out with nothing. He isn’t imagining it. We measured the wrong thing.
This is also where the in-range-but-symptomatic patient lives. A retrospective cohort of more than 9,000 men followed through twelve months of testosterone therapy stratified them into classically hypogonadal, biochemically borderline, and technically in-range but symptomatic. Quality of life moved the same direction across all three groups, with energy, life enjoyment, and work and sport performance standing out. The number in the range did not predict who would benefit.
What’s free testosterone, and why does it matter more?
Free testosterone is the unbound fraction, the portion actively available to tissue right now. It’s what your patient’s muscle, brain, and bone are actually seeing. When someone describes the symptoms of deficiency while carrying a normal total, free testosterone is usually where the answer is.
There’s a study I keep coming back to on this. Published in The Journal of Physiology, it followed 27 premenopausal women through twelve weeks of resistance training with muscle biopsies, hormone panels, and strength testing. Physiological total testosterone was not associated with muscle mass or strength, not at baseline and not in response to the training. Bioavailable testosterone and androgen receptor nuclear localisation were. Read the title of that paper and they aren’t hiding anything.
You can measure it directly, though assay quality varies. The more practical route in most clinics is to calculate it from total testosterone, SHBG, and albumin. The calculator I use and teach in the course is the ISSAM free testosterone calculator. It’s free, it takes seconds, and it turns three inexpensive values into the number that actually drives your decision.
What does SHBG actually do?
SHBG is a small protein made mostly in the liver. I describe it to providers as a Pac-Man. It grabs testosterone, binds it up, and holds it hostage where the tissue can’t reach it. The testosterone still exists. It’s just out of play, which is why the total looks reassuring.
This is why SHBG changes the interpretation of everything else on the panel. The same total testosterone of 600 means two completely different things in a patient whose SHBG is low and a patient whose SHBG is running high. In my opinion you can’t interpret a testosterone level without it, and I don’t order one without the other.
What raises and lowers SHBG?
Both directions matter clinically, and the low side gets misread constantly.
- Raises SHBG: thyroid dysfunction, excess estrogen exposure, aging, and oral contraceptives.
- Lowers SHBG: insulin resistance and obesity.
The contraceptive piece deserves its own paragraph, because it is bigger than the training suggests. Panzer and colleagues at Boston University looked at 124 premenopausal women with sexual health complaints. SHBG in the women who stayed on oral contraceptives ran four times higher than in women who had never taken them, 157 nmol/L against 41. That alone would be worth knowing.
Here is the part that changed how I practice. After they stopped the pill, SHBG did not come back down to baseline. Beyond 120 days off, it was still clearly elevated compared to never-users. So the patient who came off her contraceptive six months ago and still has no libido and an undetectable free testosterone is not imagining it, and she is not being impatient. Check her SHBG.
Providers sometimes see a low SHBG and assume it’s good news, because less binding should mean more free testosterone. I love that instinct, and it’s worth pushing on. The problem is what is usually driving it. If SHBG is low because of insulin resistance and excess adiposity, total testosterone is very likely falling right along with it. You don’t get a healthier patient out of that. You get a larger share of a smaller supply.
So a low SHBG isn’t a finding to celebrate. I treat it as a metabolic signal and go looking upstream at insulin resistance and adiposity before I touch the testosterone.
What labs should you order for testosterone?
The minimum panel is total testosterone, free testosterone, and SHBG, plus albumin if you’re calculating free rather than measuring it directly. These are inexpensive and widely available, so there’s no good reason to order total alone.
Once you have them, the sequence is simple. Read the total for context. Read SHBG to understand what’s binding it up. Then read free testosterone, which is the number that reflects what your patient is actually living with day to day. If the free value and the symptoms agree, you have your answer, and the total was never going to give it to you.
What about testosterone in women?
Everything above applies to female patients, and the gap there is wider. Testosterone isn’t a male hormone that women carry a trace of. They have it by design, and it does the same work it does in men: lean body mass and body composition, cognition and mental clarity, drive through a dopaminergic pathway, libido, and bone mineral density through a route that doesn’t depend on estrogen.
The patient you’ll recognize is already on estradiol and progesterone. Her hot flashes eased, her sleep improved, and she still tells you her libido is gone and she can’t hold muscle no matter how hard she trains. That’s frequently a testosterone story, and it goes unnamed because the training never pointed anyone at it.
The two things that stop providers are the male hormone narrative and the fear of androgenic side effects. Those effects are real at supraphysiologic exposure and uncommon at replacement doses handled carefully, and most of the fear traces back to bodybuilding doses that would be high for a man.
Female dosing, concentrations, routes, and the full case for treating are a longer conversation than fits here. I wrote that up separately in my post on testosterone for women.
Why does every female patient look low on free testosterone?
This took me forever to get through my thick skull, and I had some patient mentors. There are two different free testosterone tests and they do not report the same numbers.
The old gold standard is equilibrium dialysis. It physically separates free from bound and reports in ng/dL. Modern commercial labs mostly calculate it instead, from total testosterone, SHBG, and albumin, and report in pg/mL. The conversion is 1 ng/dL to 10 pg/mL. But the methods themselves don’t line up one to one, and calculated results read considerably lower than equilibrium dialysis. Widest gap is in women.
So when a seasoned mentor tells you a young woman should sit around 3 to 6 ng/dL, that is 30 to 60 pg/mL. Then you pull up a modern commercial report and the female reference range tops out around 6.4 pg/mL. Apples to oranges. Your patient is not necessarily deficient. You may be reading a different test than the one the target was built on.
Check which assay your lab runs and what units it reports before you decide anyone is low.
Why can’t the reference range tell you what is optimal?
Reference ranges were built to describe a population, not to define health. Somebody drew blood across a broad age band, plotted the distribution, and drew lines around the middle of it. That’s what normal means. It doesn’t mean optimal, and it never claimed to.
Two things make this worse than it sounds. The population inside the range likely includes people who are symptomatic, so their values are baked into what we now call normal. And the population itself has moved. There’s documented population-level decline in serum testosterone in American men that’s independent of age, along with a more recent decline among adolescent and young adult men. A range recalculated against a declining population keeps redefining deficiency downward.
When I started in this field the top of the male range sat higher than it does now. The range moved. The physiology didn’t.
I want to be straight about the other side of this, because it would be easy to read the above and assume I have a better number to hand you. I don’t. There is no evidence-based optimal reference range for hormones. What we have is an educated guess, built by me and by a lot of people who came before me, from watching where patients actually feel well. That’s it.
Which is exactly why becoming married to a number is the wrong move in both directions. Use the number as a guide, read it next to the symptoms, and treat the person.
Which route of testosterone administration should you use?
All of these routes work. I want to say that plainly before getting into preferences, because people often arrive convinced there’s one right answer and there isn’t.
Intramuscular versus subcutaneous
Both are efficacious. We’ve compared them directly and both performed well. Intramuscular has been the standard for decades and it still performs. What has changed is the hardware. We don’t need the enormous needles anymore. A half inch 27 gauge insulin syringe is thin enough to limit tissue damage and still thick enough to draw and deliver an oil in reasonable time. For patients carrying more subcutaneous fat over the delivery site, a one inch 27 gauge works well. Rotate sites.
Subcutaneous injection has gained real ground. Same 27 gauge half inch needle, into abdominal or gluteal fat. In theory the slower absorption should blunt the peaks and troughs.
Does subcutaneous injection reduce aromatization?
You’ll hear that subcutaneous dosing aromatizes less than intramuscular, and there is some signal for that in the literature. What I want to be careful about is the leap from a lab difference to a clinical one. Seeing higher estrogen on intramuscular dosing doesn’t by itself mean those men are having a worse time of it, and that isn’t something we’ve been able to pin down in terms of actual side effects. Treat it as an open question rather than a reason to switch someone.
How do you manage post-injection pain?
The practical downside of subcutaneous is that an oil deposited in fat tends to sit there. Patients get a red, warm, sometimes firm welt that looks alarming and gets mistaken for cellulitis. It usually isn’t. It’s local irritation, plausibly from the preservatives in the carrier, and it typically resolves on its own over several days.
What helps: a gentle massage around the site for ten to twenty seconds right after the injection to disperse the depot, warmth from a heating pad or a shower, and smaller more frequent doses. Some patients simply won’t tolerate subcutaneous no matter what you do, and those patients should move to intramuscular.
Creams and gels
Compounded topicals are good, and worth distinguishing from retail pharmacy products. The compounded versions let you titrate finely, which is the main practical argument for them. Concentrations differ substantially between male and female prescribing, and that side of it is covered in my post on testosterone for women.
Should you use testosterone pellets?
I’m not a fan of pellets, and I want to be fair about why, because plenty of good providers use them. My general principle is the least invasive effective option that gets the result, and a pellet insertion is a procedure.
There’s also a comparison problem that gets read backwards, where a patient does well on pellets and poorly after switching. What’s usually happening is that the pellets were delivering more than the provider was willing to match, so the switch quietly undertreated the patient. That’s a dosing gap rather than evidence that pellets work better.
Why aim for the maximum effective dose?
This is an HRT University principle and it runs against how we were trained to think about medication.
Minimum effective dosing makes sense for conventional pharmacology. You’re introducing a foreign molecule, it carries a side effect profile that scales with exposure, and you want the smallest amount that gets the job done. Reasonable.
Hormone therapy is a different problem. A bioidentical hormone is structurally identical to what the body already makes. You aren’t adding a foreign compound, you’re restoring a native one. Think about how we dose insulin. Nobody argues for the smallest possible insulin dose that nudges glucose slightly. You use the amount that produces a meaningful result. Same logic here.
The way I define it: a maximal effective dose is the highest dose that is still tolerable against a pre-specified level of side effects. Not the largest dose imaginable. The most useful one you can give without buying trouble.
The study I point people to is Bhasin and colleagues in 1996, which ran 43 men through ten weeks of testosterone enanthate at 600 mg per week against placebo, with training and a high-protein diet. That’s more than four times a typical prescribing dose, and it is worth reading before deciding where the danger actually starts.
So the target is the dose that restores the patient to a hormonal environment where they function well, and that requires clinical judgment rather than a lab cutoff. We monitor. We check labs. But the number is a tool for reasoning, not the verdict. The verdict is whether the patient’s symptoms resolved.
One honest caveat. The word supraphysiologic gets used to shut down conversations, and it’s worth remembering that we’re measuring against a range that was never built to define optimal in the first place. That isn’t an argument for dosing without limits. It’s an argument for knowing what your ceiling is actually based on, and monitoring the patient rather than the range.
How should SHBG change your dose and frequency?
This is where the whole panel pays off. SHBG doesn’t just tell you how to read the labs, it tells you how to write the prescription.
High SHBG means more of what you give gets bound on arrival, so those patients usually need a bigger dose, and twice a week injections can hold them. Low SHBG patients usually need less drug, and they often do better with smaller amounts given more frequently, sometimes every other day. Same drug, same target, opposite prescriptions, and the only thing that told you which was which was a cheap protein.
Key takeaways
- Total testosterone alone will miss a deficient patient. Order total, free, and SHBG together.
- Free testosterone is what the cell can use. Calculate it if you aren’t measuring it directly.
- SHBG binds testosterone and makes it unavailable. It changes the meaning of every other value on the panel.
- Thyroid dysfunction, excess estrogen, and aging raise SHBG. Insulin resistance, obesity, and contraceptives lower it.
- A low SHBG is a metabolic flag, not good news.
- Reference ranges describe a population that likely includes symptomatic people, and that population has been declining. Normal isn’t optimal.
- Intramuscular and subcutaneous injection are both effective. Post-injection pain is the main reason to move a patient off subcutaneous.
- Aim for the maximum effective dose rather than the smallest one that moves a lab value, because you’re restoring a molecule the body already makes.
- SHBG should inform both dose and dosing frequency.
Frequently asked questions
Is free testosterone or total testosterone more important?
Free testosterone, in almost every case. It represents the fraction available to tissue. Total testosterone is still worth having as context, and you need it to calculate free, but on its own it won’t tell you whether a patient is deficient.
What’s a normal SHBG level?
Lab ranges vary, which is part of the problem. The more useful question is what SHBG is doing to this patient’s free testosterone, and whether the value points at something upstream like thyroid dysfunction or insulin resistance. Read it as context rather than as a pass or fail.
Can you have low testosterone with a normal total testosterone?
Yes, and it’s common. A high SHBG binds up the circulating testosterone and leaves the free fraction below range while the total looks unremarkable. That patient is deficient and will be told he is fine unless free testosterone and SHBG are on the order.
How do you calculate free testosterone?
From total testosterone, SHBG, and albumin. The ISSAM calculator is the one I use and teach. It’s free and takes seconds.
Does a low SHBG mean more usable testosterone?
Mathematically yes, clinically usually no. When SHBG is low because of insulin resistance or obesity, total testosterone is typically falling alongside it. The free fraction is a larger share of a smaller number. Treat the metabolic driver.
What labs should I order before starting testosterone therapy?
At minimum total testosterone, free testosterone, and SHBG, plus albumin if you’re calculating free rather than measuring it. What else belongs on the panel depends on the patient and is covered in depth in the Master Course.
Is subcutaneous or intramuscular testosterone injection better?
Both are efficacious, and my read of the comparison work is that neither route wins on effectiveness. Intramuscular has the longer track record. Subcutaneous is convenient but produces post-injection pain in a subset of patients, and those patients should move to intramuscular.
Why does SHBG change the dose?
Because it changes how much of what you give stays available to tissue. High SHBG usually means a bigger dose and twice a week injections. Low SHBG usually means a smaller dose given more often, sometimes every other day.
Are testosterone reference ranges reliable?
They’re reliable descriptions of a population and unreliable definitions of health. The population sampled includes symptomatic people, and serum testosterone has declined at a population level over decades. Use the range for orientation, then treat the patient.
Do these principles apply to women as well as men?
The interpretation logic is the same in both sexes. The doses, concentrations, and routes aren’t. For the female side specifically, including what testosterone does in female physiology and how to handle the androgenic side effect question, see my post on testosterone for women.
Where can you learn the full reasoning?
Lab interpretation sits in Section 2 of the Male HRT module, and on the female side in Section 1, which covers the menstrual cycle, birth control, and labs. The testosterone section of the Female HRT module then handles benefits, deficiency, administration and dosing, and side effects. If you want the full reasoning rather than the summary, that’s what the Master Course is for. Six modules, 30 CE hours, jointly accredited.
