Wide variability in laboratory reference values for serum testosterone

Introduction. The laboratory determination of testosterone levels consistent with a diagnosis of hypogonadism is complicated by the availability of multiple testosterone assays and varying reference ranges. Aim. To assess current laboratory practices regarding availability of testosterone assays and use of reference values. Methods. A telephone survey of 12 academic, 12 community medical laboratories, and one national laboratory. Main Outcome Measures. Types of androgen assays offered and determination of reference values. Results. All of the academic and eight of the community centers performed total testosterone testing. Free testosterone was performed in-house by six of the 12 academic and one community center. Testing for bioavailable testosterone, free androgen index, and percent free testosterone was performed in-house by no more than two centers. There were eight and four different assays used for total and free testosterone, respectively. One national laboratory offered equilibrium dialysis measurement of free testosterone. Of the 25 labs, there were 17 and 13 different sets of reference values for total and free testosterone, respectively. The low reference value for total testosterone ranged from 130 to 450 ng/dL (350% difference), and the upper value ranged from 486 to 1,593 ng/dL (325% difference). Age-adjusted reference values were applied in four centers for total testosterone and in seven labs for free testosterone. All reference values were based on a standard statistical model without regard for clinical aspects of hypogonadism. Twenty-three of the 25 lab directors responded that clinically relevant testosterone reference ranges would be preferable to current standards. Conclusions. Laboratory reference values for testosterone vary widely, and are established without clinical considerations.