Total T or Free T: Which is the Better Test for Androgen Deficiency?

—When considering a workup for low testosterone, weigh the utility of different biomarkers.

By Sanjai Sinha, MD
Reviewed by Clifton Jackness, MD, attending physician in Endocrinology, Diabetes, and Metabolism at Lenox Hill Hospital and The Mount Sinai Medical Center in New York City

Age-related decline of testosterone has been confirmed in several cross-sectional and longitudinal studies, and results from dysfunction in the hypothalamic-pituitary-testicular axis. 1-3 While the clinical relevance of this decline is variable, testing for testosterone deficiency has become increasingly common given some possible therapeutic options. Some clinicians start their evaluation for hypogonadism by ordering total testosterone (T), while others order free (bioavailable) T, and others order both. In addition to the operating characteristics of these tests, patient factors play a role in the utility of using each test as a basis for diagnosis. This review will examine recent biomarker data and offer a practical approach to use.

The Bottom Line

Total testosterone (T) measurement is a good initial diagnostic test in the evaluation of hypogonadism, but it has its limitations.
Reserve use of free T for patients whose total T is between 150 and 350 ng/dL, or in obese or older men whose total T may be erroneously low.

Laboratory testing for hypogonadism relies on measuring the products of the testes (testosterone and sperm) and the pituitary hormones that regulate their production, luteinizing hormone (LH) and follicle-stimulating hormone (FSH). The Leydig cells of the testes produce testosterone when stimulated by LH. Sperm production from the seminiferous tubules is stimulated by testosterone primarily, but also by FSH. The feedback loop involves testosterone inhibition of FSH and LH. Primary hypogonadism is the correct biochemical diagnosis when there are decreased sperm in the semen and/or a low serum concentration of total T (usually 4

A recent study examined the accuracy of total T measurement as a test to identify men with normal versus low free T. 5 The investigators studied the charts of 3700 men (mean age, 60; half with obesity) who had both total T and calculated free T checked during a hypogonadism workup. The findings illustrated why testosterone testing can be confusing. At a cutoff of 280 ng/dL (the lower limit of normal in the labs involved in the study), total T was 91% sensitive but only 74% specific. In other words, only 9% of true cases were missed, but 26% of patients with a measurement below 280 ng/dL were falsely identified as having hypogonadism.

Using 350 ng/dL as a cutoff improved sensitivity with further loss of specificity. At a much lower cutoff of 150 ng/dL, total T was only 59% sensitive, but 99% specific. Thus, in this lower range, 41% of biochemical hypogonadism was missed, but all who were identified were truly hypogonadal. What this study implies is that total T is an excellent stand-alone diagnostic test for very low or somewhat high results. At 350 ng/dL it can be reliably ruled out. It is not helpful for results that fall in between these levels. For such results, a free T should be ordered if clinical suspicion for hypogonadism is high.

Consider a few other important points about ordering total T—time and reproducibility. Total T has diurnal variation, especially in younger men, with the highest readings tending to occur in the morning, around 8 AM. 6 As it is easier to see a difference between normal and subnormal when normal is maximal, an 8 AM test is recommended. If the result is borderline low, the test should be repeated before making a definitive diagnosis of androgen deficiency.

J Clin Endocrinol Metab

Clin Endocrinol (Oxf)

J Urol

J Clin Endocrinol Metab