Testosterone Assay Pitfalls: Why One Number Is Not a Diagnosis

Testosterone is not a pass-fail laboratory value: its concentration varies within the same person, most of it circulates bound to proteins, and different assays can disagree near a decision threshold. Symptoms attributed to “low T” are also common in sleep disorders, depression, and medication effects. They are common in obesity, chronic disease, and ordinary aging.

For these reasons, a diagnosis of hypogonadism is a clinical and biochemical conclusion. It is not a single result, an online target range, or a response to a trial of treatment.

Begin with the clinical question#

Testosterone testing is most interpretable when prompted by symptoms or signs that could reflect androgen deficiency. More specific features include reduced sexual desire, fewer spontaneous erections, and impaired sperm production. They include small testes, loss of body hair, hot flashes, or reduced bone density in an appropriate setting. Fatigue, low mood, reduced concentration, weight gain, and diminished exercise capacity are less specific.

The Endocrine Society and American Urological Association both require compatible clinical findings plus consistently low measurements. They differ in some thresholds and details, which is a reminder that a laboratory number works within a diagnostic framework, and population screening of people without relevant symptoms can find low values that reflect variation, illness, or altered binding proteins. When pretest probability is low, repeat results and careful cause-finding become especially important.

Time of day changes the sample#

Testosterone secretion follows a daily rhythm, with concentrations generally higher in the morning and lower later; the rhythm is more pronounced in younger men, but timing remains important because reference ranges and diagnostic evidence usually rely on morning samples.

Shift work and irregular sleep complicate the clock. “Before 10 a.m.” may not reflect the biological morning of someone who sleeps during the day. The clinician and the laboratory need your sleep schedule and the collection time to interpret the result.

Food can lower measured testosterone in some settings, which is why the Endocrine Society specifies fasting morning samples. A meal, strenuous activity, poor sleep, alcohol, and short-term physiologic stress can add within-person variation. Standardizing the conditions of repeat testing reduces noise. One unusual morning should not become a permanent diagnosis. Even with careful collection, a substantial fraction of initially low results return to a non-low range on repetition.

Acute illness can suppress the axis#

Infection, surgery, and trauma can temporarily lower testosterone. So can hospitalization, calorie deficit, and uncontrolled systemic illness. They act through changes in hypothalamic and pituitary signaling, testicular production, binding proteins, and metabolism. Testing during acute illness can identify physiology of illness rather than chronic gonadal failure.

The joint endocrine and laboratory position statement recommends morning fasting testing when the person is not acutely ill. Deferring a nonurgent diagnostic measurement until recovery can prevent unnecessary labeling.

Chronic illness is different. Kidney, liver, and pulmonary conditions may produce persistent changes and symptoms. So may inflammatory, endocrine, and metabolic conditions. The result may still be clinically meaningful, but its cause and the balance of benefit and harm require more than repeating the number.

Total testosterone measures bound and unbound hormone#

Most circulating testosterone is bound to SHBG or albumin. Only a small fraction is free. Total testosterone includes all three pools.

SHBG binds testosterone tightly. Albumin binding is weaker. Changes in SHBG can shift total testosterone even when free hormone changes less. This is why total testosterone is an efficient first test but not always the last test.

Conditions associated with lower SHBG include obesity, insulin resistance or type 2 diabetes, and hypothyroidism. They include nephrotic syndrome and acromegaly. They also include use of glucocorticoids, some progestins, or androgens. A low SHBG can produce a low total result with a free result that is less reduced.

Higher SHBG can occur with aging, hyperthyroidism, or cirrhosis or some other liver disease. It can occur with HIV, estrogen use, and some anticonvulsants. A high SHBG can make total testosterone appear adequate while free testosterone is lower. These associations are not diagnostic by themselves. SHBG is influenced by multiple processes, and the same condition can affect testosterone production as well as binding.

When free testosterone adds information#

The Endocrine Society recommends free-testosterone assessment when total testosterone is near the lower limit or when a condition that alters SHBG is present. It is not a universal add-on to every panel.

Equilibrium dialysis physically separates free from protein-bound hormone and is considered the reference measurement approach when coupled to an accurate quantification method; it is technically demanding and not available in every routine laboratory.

Calculated free testosterone uses total testosterone, SHBG, and albumin. Its quality depends on the accuracy of all inputs and the equation's binding assumptions. Different equations can yield different results, especially at unusual SHBG or albumin concentrations. The report has to say the formula and the reference interval.

Direct analog free-testosterone immunoassays are affected by binding-protein concentrations and do not accurately measure the free fraction. The Endocrine Society advises against using them. The words “free testosterone” on a report are therefore not enough; the method must be checked. “Bioavailable testosterone,” which generally refers to free plus albumin-bound hormone, is a third thing again: methods and reference intervals vary, and it is not interchangeable with free testosterone.

Immunoassay and mass spectrometry are not simple opposites#

Automated immunoassays are widely available, fast, and practical. Their antibodies and calibration can be less accurate at low concentrations or in the presence of cross-reacting compounds; performance differs among platforms, so criticism of one assay cannot be applied to every immunoassay.

Liquid chromatography tandem mass spectrometry, usually written LC-MS/MS, separates molecules before detection and can provide better specificity and low-range performance, and it is often preferred when concentrations are low, such as in women and children, and can improve comparability in men near a threshold.

Yet mass spectrometry is not one standardized product. Sample preparation, chromatographic separation, and internal standards determine quality. So do calibration traceability, matrix effects, and interference checks. So do the lower limit of quantification and operator expertise. A poorly validated laboratory-developed LC-MS/MS method can perform worse than a well-standardized immunoassay. Ask whether the method is validated for the concentration and the population you care about, not only which acronym appears on the report.

Standardization makes numbers more comparable#

The CDC Hormone Standardization Program assesses total-testosterone procedures against an accuracy-based reference method. Its April 2026 list identifies currently certified laboratories and assay systems.

Certification means a procedure met defined analytical performance criteria during the certification period. It does not mean every result is error-free, every laboratory using a similarly named platform is certified, or free testosterone is covered by a total-testosterone certification. External quality assessment, daily internal controls, instrument maintenance, reagent-lot checks, and investigation of unexpected results remain necessary, and you can contact the laboratory when a result conflicts sharply with the clinical picture or changes implausibly from a prior value.

Reference intervals are not treatment targets#

A reference interval usually describes the central distribution in a selected population. It depends on age, health criteria, collection time, assay, and statistical method. It does not separate everyone with disease from everyone without disease.

The harmonized reference-range study used CDC-standardized measurements from community cohorts to improve comparability in healthy nonobese young men. That work helps define a lower limit for a specific reference population, but guideline thresholds also incorporate symptoms and evidence about treatment.

The AUA uses total testosterone below 300 ng/dL as a reasonable cutoff supporting diagnosis, obtained on at least two early mornings, and the Endocrine Society emphasizes the lower limit of a rigorously derived assay-specific range. A value near either boundary should not be portrayed as a biological cliff.

Unit errors can add confusion. Laboratories may report ng/dL or nmol/L for total testosterone and several units for free testosterone; conversions must be done correctly, and total and free values cannot be compared numerically without their units and method.

Interference and specimen problems can produce surprises#

Heterophile antibodies, human anti-animal antibodies, and cross-reacting steroids can interfere with immunoassays, depending on platform design. So can supplements containing high-dose biotin and some medications. Hemolysis, lipemia, mislabeling, prolonged storage, and repeated freeze-thaw cycles can also affect laboratory work.

Biotin interference is method-specific and can cause falsely high or low results in susceptible assays. A complete list of your supplements and recent doses helps the laboratory assess risk, and stopping a supplement solely for a test should follow laboratory or clinical instructions, because the required interval varies. An implausible result deserves confirmation with a new specimen, review of collection conditions, and sometimes a different method; repeating the same compromised sample on the same platform may reproduce the problem rather than resolve it.

A low result should lead to cause-finding#

Once consistently low testosterone and compatible findings are established, luteinizing hormone helps distinguish primary testicular dysfunction from a hypothalamic or pituitary pattern. Follicle-stimulating hormone, prolactin, or iron studies may be appropriate based on the pattern. So may pituitary evaluation, genetic assessment, or fertility testing.

Medication review is essential. Opioids, glucocorticoids, and androgenic-anabolic steroids and withdrawal can suppress the axis. So can some cancer therapies and other medicines. Severe obesity, sleep disorders, undernutrition, excessive endurance training, and systemic disease may contribute.

The purpose is not merely to qualify for testosterone. Some causes require their own treatment, some are reversible, and some make testosterone inappropriate or require specialist evaluation. Fertility goals matter because exogenous testosterone can suppress sperm production.

Very low testosterone with low or inappropriately normal gonadotropins, headaches, or visual changes can require prompt endocrine assessment. So can high prolactin, other pituitary-hormone abnormalities, or delayed puberty. Sudden severe headache or acute visual loss is an emergency regardless of a testosterone result.

Diagnosis and monitoring are different measurement problems#

Testing someone who is not on therapy asks whether endogenous production is consistently low, and monitoring prescribed testosterone asks whether the chosen formulation produces an appropriate concentration at a defined time relative to administration.

Peak, trough, and average levels differ by formulation. Applying a morning diagnostic reference range to an arbitrarily timed on-treatment sample can mislead. The prescription, last administration time, and dose schedule belong in the interpretation. So do laboratory method, symptoms, and adverse effects. So do hematocrit and other safety monitoring.

Do not change prescribed testosterone or obtain non-prescribed hormone products in response to one laboratory result. Excess concentrations can increase adverse effects without addressing the original cause of symptoms.

A practical checklist for an unexpected value#

Before you assign meaning to a low, high, or discordant result, ask:

1. Was there a relevant symptom or sign before testing?
2. Was the sample collected on a separate fasting morning and repeated?
3. Was the person acutely ill, sleep deprived, or on an unusual schedule?
4. Which total-testosterone method and reference interval were used?
5. Could SHBG be low or high, and is free testosterone indicated?
6. How was free testosterone measured or calculated?
7. Could medication, supplement, specimen, or assay interference explain the result?
8. What do luteinizing hormone and the broader clinical pattern suggest about cause?

This sequence protects against both missed disease and treatment of a number that does not represent persistent hypogonadism.

References#

1. Endocrine Society testosterone guideline
2. AUA testosterone deficiency guideline
3. CDC certified testosterone assays, updated April 2026
4. Joint endocrine and laboratory position statement
5. Harmonized testosterone reference ranges
6. Endocrine Society position on accurate testosterone measurement

For your own health, talk with your clinician.*