Current Diagnostic Criteria for Testosterone Deficiency (Notes)

It is 80 years since testosterone (T) therapy began and ~20 years into the modern era, yet uncertainty remains about diagnosing testosterone deficiency (TD/hypogonadism).
Real-world cases illustrate a “grey zone” where patients have symptoms but no clear threshold below which treatment is indicated.
The lack of a evidence-based probability that a given T value predicts treatment response leads to potential under- or over-treatment.
The core message: diagnostic criteria for TD are inadequate and need rethinking to focus on likelihood of symptomatic response rather than strict normal/abnormal thresholds.

Historically, TD diagnosis relied on clinical presentation; now most TD cases are diagnosed primarily by blood test results (total T).
Thresholds used in guidelines vary:
- US FDA and AUA: threshold around >300\ \text{ng/dl} (>10.4\ \text{nmol/L}).
- European/Urology guidelines: threshold around >348\ \text{ng/dl} (>12.1\ \text{nmol/L}).
- Some experts use >400\ \text{ng/dl} (>13.9\ \text{nmol/L}).
Some studies suggest a threshold near the Framingham 2.5th percentile: $348\ \text{ng/dl}$ ( $12.1\ \text{nmol/L}$ ).
The 2018 Endocrine Society guidelines proposed harmonized reference ranges with a 2.5th percentile of $264\ \text{ng/dl}$ ( $9.2\ \text{nmol/L}$ ).
Clinical utility of these reference ranges is unknown; thresholds remain arbitrary.
EMAS data suggest a combined diagnostic signal: symptoms plus a T value around $11\ \text{nmol/L}$ ( $316\ \text{ng/dl}$ ) and calculated free T (cFT) < $220\ \text{pmol/L}$ ( $64\ \text{pg/mL}$ ).
Lab reference values vary widely across laboratories: in a survey of 25 US labs, 17 provided different reference ranges with about $300\%$ variation; same result can be labeled normal or low.
Guidelines on whether to use free T (FT) are mixed:
- EAU: use FT when discrepancy exists between T and symptoms.
- Endocrine Society: measure FT particularly when SHBG abnormalities are suspected.
- AUA: does not specifically mention FT.
The current system emphasizes total T thresholds despite known limitations.

TD symptoms occur across a wide range of serum T concentrations.
Contributing factors for the variability include:
- Significant inter-individual variation in biology among men with similar T levels.
- Unclear role of the magnitude of T decline over time versus a single absolute value.
- Genetic differences in androgen receptor sensitivity (e.g., CAG repeats).
- Total T is a flawed proxy because SHBG binding affects the readily available free/biologically active testosterone.

Testosterone is lipophilic; 40–60% is bound to SHBG, rendering it unavailable to tissues.
Higher SHBG can yield higher total T even when free T (FT) is low, leading to misleading conclusions about androgen status.
Observational data show that many symptomatic men have low FT despite normal total T, and those with low FT respond to therapy even if total T is above common thresholds.
Symptom burden and treatment response correlate better with FT (or bioavailable T) than with total T.
Practical implications:
- Total T thresholds may misclassify patients; FT should be considered, especially when SHBG is abnormal.
- Guidelines diverge on FT use; the Endocrine Society favors FT assessment when SHBG issues are suspected; EAU suggests FT in discordant cases; AUA omits FT.
Future approach: focus on likelihood of symptomatic response to treatment rather than rigid normal/abnormal thresholds.

Various procedural rules exist (e.g., repeat testing on different days, fasting status, early-morning sampling) but there is no strong evidence that these steps improve diagnostic accuracy or predict treatment response.
Current emphasis should be on whether tests provide better predictive value for symptomatic improvement rather than on rigid procedural defaults.

The core question: can we identify a human biomarker (ideally FT) that reliably indicates androgen status and predicts treatment response?
Recommendations from Morgentaler et al.:
- FT is the most accurate indicator of androgen status; treat symptomatic men with low FT regardless of total T value.
- Calculated FT is a practical measure using total T and SHBG (albumin included in equations; albumin has little influence on cFT).
- Direct FT (analog assay) correlates well with equilibrium dialysis (EqD) and calculated FT, but uses a different scale.
Key proposed study: enroll men with defined TD-like symptoms, provide testosterone treatment regardless of baseline T, and assess the likelihood of symptomatic response as a function of baseline measurements of total T, FT, and bioavailable T (the most predictive metric will drive clinical use).
Goal: move from arbitrary thresholds to evidence-based decisions that estimate the probability of benefit from therapy.
Overall: a shift toward diagnosing TD based on solid evidence of likely benefit rather than adherence to unreliable thresholds.

Historical and guideline sources explore the evolution of thresholds and the role of FT in diagnosis and management.
Data emphasize the variability of laboratory reference ranges and the superior relevance of free testosterone in correlating with symptoms and treatment response.