Free Testosterone vs Total Testosterone: What SHBG Changes

When free testosterone changes the answer

On Kantesti AI, we see this mismatch every week: someone uploads a report with total testosterone 290 ng/dL, feels awful, and assumes the answer is simple. Our first pass is to check the collection time, the units, the SHBG blood test, and whether the value sits in the same gray zone described in our borderline lab guide.

When I, Thomas Klein, MD, review a panel showing total testosterone 310 ng/dL with SHBG 72 nmol/L, I worry more than I do about total testosterone 280 ng/dL with SHBG 12 nmol/L. The first pattern can hide a genuinely low free fraction; the second often reflects obesity or insulin resistance lowering SHBG rather than severe androgen failure.

Units trip people up more than most clinicians admit. A total testosterone of 300 ng/dL equals about 10.4 nmol/L, and patients reading international reports on the same day can think two labs disagree when they are actually saying the same thing—our testosterone range by age explainer helps with that.

What total testosterone actually measures—and misses

The 2018 Endocrine Society guideline still frames the diagnosis the same way in 2026: symptoms plus unequivocally and consistently low morning total testosterone on 2 tests (Bhasin et al., 2018). A single isolated low result is not enough, and a total testosterone below 300 ng/dL is a common clinical cutoff rather than a magical biological line.

Total testosterone is useful because it is standardized better than free testosterone and usually enough when the result is clearly low or clearly reassuring. In practice, a repeated value below 200 ng/dL (6.9 nmol/L) carries far more weight than a one-off 295 ng/dL, while values above 500 ng/dL (17.4 nmol/L) make classic androgen deficiency less likely if SHBG is normal.

Age and context still matter. A tired 34-year-old with obesity and snoring deserves a different workup than a healthy 68-year-old who happens to land at 340 ng/dL, which is why I still like broader screening in men reviewing their annual labs in their 30s or later in life rather than chasing one hormone in isolation.

How free testosterone is measured, calculated, and sometimes mismeasured

Equilibrium dialysis is the laboratory reference method for free testosterone, but most community labs do not run it because it is slower, costlier, and technically demanding. That is why many endocrinologists accept a calculated free testosterone derived from total testosterone, SHBG, and albumin, especially when SHBG is abnormal (Vermeulen et al., 1999).

Rosner et al. warned years ago that direct analog free testosterone immunoassays can mislead clinicians, particularly when binding proteins are off balance (Rosner et al., 2007). At Kantesti AI, our method checks the assay language against our clinical validation ruleset because a free testosterone result is only as trustworthy as the way the lab produced it.

Albumin matters less than SHBG most of the time, but it is not irrelevant. If albumin is far below the usual 3.5-5.0 g/dL range—cirrhosis, nephrotic loss, severe illness—the calculated value can drift, and that is one reason our SHBG blood test guide treats albumin as a context marker rather than a footnote.

Which free testosterone method should you trust most?

If you have a choice, ask whether the lab used equilibrium dialysis or a validated calculated free testosterone method. In my experience, a vague direct free testosterone result with no method listed is where many of the strangest borderline panels come from, and clinicians disagree more on universal free testosterone cutoffs than patients realize because the assays are so inconsistent across labs.

How the SHBG blood test reframes a borderline result

SHBG is produced mainly in the liver, and in many adult male labs the reference interval sits roughly around 10-57 nmol/L, though some labs start nearer 18 nmol/L. Values below about 15 nmol/L usually push total testosterone downward, while values above 60 nmol/L can make total testosterone look deceptively comfortable.

Here is the pattern I see in lean older men and endurance athletes: total testosterone 420 ng/dL, SHBG 78 nmol/L, free testosterone low, symptoms real. If liver enzymes, weight, and thyroid function raise suspicion, I look at broader liver test patterns and a full thyroid panel before assuming aging alone explains everything.

Low SHBG creates the opposite illusion. In men with central adiposity or insulin resistance, a total testosterone of 260-320 ng/dL can coexist with a free fraction that is still adequate, and that is one reason Kantesti AI treats SHBG as a decision pivot rather than an optional add-on.

Free testosterone vs total testosterone: the 4 patterns clinicians use

Pattern 1 is low total testosterone + normal free testosterone + low SHBG. That combination often points toward obesity, insulin resistance, sleep apnea, or medication effect rather than primary gonadal failure, and the practical next step is usually metabolic work rather than a rushed prescription.

Pattern 2 is normal total testosterone + low free testosterone + high SHBG. This is the patient who gets told your testosterone is normal despite low libido, low morning energy, and reduced recovery—older age, hyperthyroidism, liver disease, HIV, and oral estrogen are classic settings.

Pattern 3 is low total testosterone + low free testosterone. If the values repeat low on 2 properly timed samples, I take it seriously because both the reservoir and the biologically active fraction are reduced.

Pattern 4 is symptoms + normal total testosterone + normal free testosterone. That is where trend review helps; our trend comparison often shows the hormone was stable while ferritin, thyroid markers, glucose, or sleep-related clues were drifting, and a broader fatigue lab workup is usually smarter than fixating on testosterone.

Why timing, sleep, illness, and fasting habits can distort testosterone

For most men, the preferred draw is between 7 am and 10 am, and younger men tend to show the biggest morning-to-afternoon drop. Shift workers are the exception; I usually want the sample within about 3 hours of waking, not by the wall clock, which is the same logic we use when teaching cortisol timing.

Acute illness can suppress testosterone transiently by 10% to 30%, sometimes more. I have seen a healthy 38-year-old upload a result of 240 ng/dL two days after a gastrointestinal virus and repeat at 410 ng/dL 3 weeks later without any treatment.

Sleep loss and severe calorie deficit matter more than people think. Thomas Klein, MD, asks patients to repeat borderline results after 2-4 weeks, with a stable sleep pattern, no acute illness, and no heroic fast or post-race lab draw, because that repeat test is often the difference between overdiagnosis and clarity.

Do you need to fast for testosterone?

Testosterone itself does not always require fasting, but if you are pairing it with glucose, insulin, triglycerides, or HOMA-IR, an 8-12 hour fast makes the full panel cleaner. The bigger issue is consistency—use the same lab, the same time window, and ideally the same assay if you are trying to interpret a subtle change.

What drives SHBG up or down

High SHBG is commonly associated with aging, hyperthyroidism, liver disease, HIV infection, anticonvulsants, and oral estrogen. Oral estrogen raises SHBG more strongly than transdermal estrogen in most patients, which is why route of therapy changes how I read the same testosterone number.

Low SHBG is commonly associated with obesity, insulin resistance, type 2 diabetes, hypothyroidism, glucocorticoid exposure, nephrotic protein loss, and androgen use. A male SHBG in the low teens, especially 10-15 nmol/L, often travels with high triglycerides, fatty liver, or elevated fasting insulin rather than isolated testicular disease.

This is where the workup becomes more useful than the label. If SHBG is low and waist circumference, triglycerides, and glucose are all moving the wrong way, I would rather correct the metabolic signal first and calculate HOMA-IR than pretend every low total testosterone needs replacement.

When free testosterone matters more in women and PCOS workups

In women, free testosterone often tells a clearer story than total testosterone because the absolute concentrations are so low that routine assays lose precision. A mildly high free testosterone with low SHBG is a classic biochemical pattern in insulin-resistant PCOS hormone testing, even when total testosterone is only borderline.

SHBG ranges in women are much wider—many labs use something like 18-144 nmol/L—and medications can swing that dramatically. Combined oral contraceptives usually raise SHBG and lower free testosterone, so I read androgen symptoms alongside estradiol ranges and the medication list before I trust the number.

If the question is female hyperandrogenism, ask how the lab measured total testosterone. LC-MS/MS total testosterone plus SHBG is usually more reliable than a routine immunoassay at these low concentrations, and DHEA clues help when the pattern suggests an adrenal source rather than the ovaries.

Which tests belong in the same order set

The minimum panel I prefer for confusing results is total testosterone, SHBG, albumin, LH, and prolactin. LH helps separate a testicular signal from a brain-pituitary signal, and our LH interpretation guide is useful when the testosterone number alone is too blunt.

If prolactin is elevated, the whole story changes because high prolactin can suppress the reproductive axis and lower testosterone. Even a modestly abnormal value deserves context, and marked elevations belong in a dedicated prolactin follow-up rather than a supplement forum.

Context tests save time. Depending on symptoms, I may add TSH, free T4, CBC, ferritin, A1c, ALT, AST, and estradiol, and our biomarker guide helps patients see why hormone interpretation rarely sits in a vacuum.

Kantesti AI groups these markers into patterns instead of isolated flags, and that mirrors how the physicians on our Medical Advisory Board actually reason. A low testosterone result with A1c 6.1%, ALT 58 U/L, and SHBG 11 nmol/L is a different clinical conversation from low testosterone with SHBG 82 nmol/L and unintentional weight loss.

When to expand the panel

If fertility matters, add FSH and usually a semen analysis. If treatment is being considered, add baseline CBC, CMP, and PSA according to age and risk, because therapy decisions are safer when you know the starting hematocrit and liver profile.

What to do next if your labs and symptoms still do not line up

Repeat low values below 150 ng/dL, a prolactin above about 100 ng/mL, or low testosterone with headaches, visual change, or very low or normal LH deserve prompt endocrine review. Men already on therapy who develop a hematocrit above 54% also need medical follow-up sooner rather than later.

If you have the PDF or even a phone photo, our AI blood test platform can normalize units, read the assay context, and flag when free testosterone is likely adding signal beyond total testosterone. You can try a free blood test interpretation if you want the panel translated into plain language before your appointment.

I, Thomas Klein, MD, would rather see 2 careful morning panels than 1 dramatic number interpreted in isolation. The clinical edge comes from patterns, which is why some of our most useful examples in real patient cases involve people whose normal total testosterone masked a low free fraction—or whose scary low total testosterone improved once SHBG and metabolic health were addressed.

Frequently Asked Questions