- Both major guidelines require at least two separate early-morning tests with compatible symptoms; my practice requires both to be fasting before long-term replacement.
- Time of day, acute illness, poor sleep, and ordinary biological variation can each push a single value across a threshold.
- The 300 ng/dL cutoff and the 264 ng/dL reference limit answer different questions; neither is a universal verdict.
- SHBG changes make total and free testosterone disagree, and the free method matters as much as the number.
- A sample drawn during testosterone or steroid use cannot reconstruct the untreated baseline.
The Specimen Matters More Than the Flag
A red-flagged result feels like a verdict, but the thing worth discussing is the tube of blood behind it. Collected mid-afternoon during an illness, after lunch, following several bad nights, it was almost certainly measured accurately. Whether it represents the man's ordinary endocrine state is another matter. Careful testing strips away avoidable variation so that a low value is more likely to reflect persistent physiology than clock time, food, sickness, or assay.
The Endocrine Society and the American Urological Association both require at least two separate early-morning measurements, interpreted alongside compatible symptoms or signs, before anyone is labeled deficient. The Endocrine Society goes further and asks that the samples be fasting; the AUA does not.
My own standard exceeds that minimum, and it is my practice rather than a guideline mandate: before committing a man to long-term replacement, I want two fasting morning values. One clearly low result with convincing symptoms opens the discussion; it does not start the drug. The medication I would be prescribing suppresses the exact axis I am trying to evaluate, so the untreated baseline is my only opportunity to see it. In symptomatic men I also check total and free testosterone together, particularly when the total hovers near the cutoff or contradicts the history. What I will not do is start someone on the strength of one afternoon sample drawn while he was ill.
Sampling a Stable Baseline
Fever, a hospital stay, an operation, an injury, and serious sleep deprivation all transiently suppress reproductive signals. In a small study of hospitalized older men with low testosterone, 62.8 percent were back in the normal range a month after going home. The cohort was tiny and elderly, but the point stands: acute stress is a rival explanation. No one should stop a prescribed medication to improve a test. Recent use of nonprescribed androgens deserves special mention, because it leaves behind a low natural pattern once the product is stopped; a man who conceals that history can walk away with a false label of permanent pituitary disease. The target is a sample that looks like the man on a normal day.
Clock Time and Repeat Draws
Testosterone runs higher early in the waking day and drifts down afterward, most steeply in younger men but at every age. In a study of 66 men aged thirty to eighty, values at about 4 PM were roughly 20 to 25 percent below 8 AM values in men in their thirties and forties, and about 10 percent lower near age seventy. Seventeen men with afternoon results under 300 ng/dL all had morning results above 300. The lasting message is that timing by itself can move a man across a threshold.
Variation between days is the second reason to repeat: among 132 men measured on multiple visits, two samples differed by more than about 18 to 28 percent roughly half the time. Redrawing until the preferred answer shows up is not confirmation; it is selection bias applied to a patient.
Fasting and Its Mixed Evidence
A controlled experiment in 60 healthy men found that a glucose load reduced testosterone by about 100 ng/dL and a mixed meal by about 123. A retrospective study of paired samples from a U.K. hospital saw no meaningful effect of fasting. So the magnitude is contested, fasting remains the Endocrine Society's approach to a consistent baseline, and the AUA accepts a nonfasting morning sample. No man should skip an essential drug or risk hypoglycemia for a hormone test.
Methods, Cutoffs, and SHBG
Every lab's reference range belongs to its own assay. The AUA describes total testosterone under 300 ng/dL as a reasonable cutoff that supports the diagnosis when symptoms and repeat morning testing line up. A harmonization project involving 9,054 men produced a reference range of 264 to 916 ng/dL in healthy, nonobese men aged nineteen to thirty-nine. These do not conflict: one is the floor of a reference interval in healthy young men, the other a supporting cutoff inside a diagnostic framework. The original report, with units and method, is far safer than a number retyped into a forum.
Low SHBG, associated with obesity, diabetes, hypothyroidism, and glucocorticoids, depresses total testosterone while free testosterone drops less. High SHBG, associated with aging, hyperthyroidism, liver disease, and certain anticonvulsants, makes the total look better than the free fraction really is. I never infer an SHBG effect from a man's build; it has to be measured. Free testosterone earns its place when the total sits near the lower limit or SHBG is probably abnormal. Equilibrium dialysis is the reference method, validated formulas approximate it, and direct analog immunoassays match neither.
Interpreting the Combinations
- Two clearly low morning values plus compatible findings make a strong biochemical case; the next question is why.
- Two normal values make persistent deficiency less likely without making the symptoms any less real.
- One low and one normal is uncertainty, not something to average. I repeat under controlled conditions or add a proper free-testosterone measurement.
- Values drawn during testosterone or anabolic steroid use answer a different question; no reference range can reconstruct the untreated baseline.
Plenty of men reach me after starting elsewhere with one pretreatment screenshot and no record of timing or method. There is no universal washout schedule; a supervised pause needs an individualized plan, and when certainty cannot be recovered, the chart should say so.
Common questions
What time of day should testosterone be tested?
Early in the morning. Testosterone runs higher early in the waking day and falls later, most steeply in younger men but at every age. In one study, afternoon values were about 20 to 25 percent below 8 AM values in men in their thirties and forties, and the men who fell under 300 ng/dL in the afternoon were all above 300 in the morning.
Do I need to fast before a testosterone blood test?
The Endocrine Society asks for fasting morning samples, while the AUA requires early-morning timing without fasting. One controlled study showed a meal or glucose load can transiently lower testosterone in some men; a real-world hospital dataset found little difference. My practice is two fasting morning draws. Never skip an essential medication or risk hypoglycemia to get the sample.
Is testosterone under 300 ng/dL low?
The AUA treats total testosterone below 300 ng/dL as a reasonable cutoff that supports the diagnosis when compatible symptoms and repeat early-morning testing are present. That does not make 299 a disease and 301 health. A separate harmonized reference range in healthy young men spans 264 to 916 ng/dL. The number is always tied to the assay, the timing, and the clinical context.