[Measurement of salivary testosterone by a simple enzyme immunoassay procedure]

The level of serum free testosterone (Se free T) is thought to affect personal emotion and behavior. On the other hand, salivary testosterone (Sa-T) exists mainly in an unbound state and its measurement is noninvasive and repeatable. The aim of this study was to evaluate whether Sa-T can be measured in a clinical setting using a simple enzyme immunoassay kit, and to evaluate whether the level of Sa-T obtained by a simple procedure can be used as a substitute for the level of Se free T by comparing the values obtained with those of Se free T.

Seventy-six volunteers from among patients who visited our outpatient clinic. They were 74 male and 2 female patients between the ages of 28 and 90 (mean 67.2 +/- 12.7 years, median 70 years). Sixteen of 74 male patients were treated with hormone therapy for prostate cancer. All of 76 patients were scheduled for blood examinations because of urological diseases, and gave full informed consent to participate in this study. Saliva and serum samples were collected between 06.00 and 11.00 h. Sa-T, Se free T and the concentration of transferine in saliva was measured with a Salivary Testosterone Enzyme Immunoassay Kit, a Coat-A-Count Free Testosterone and a Blood Contamination Enzyme Immunoassay Kit, respectively. The average of duplicate values was used. Bioavailable testosterone (BT) was assayed by Liquid Chromatography-Mass Spectrometry/Mass Spectrometry. Average intraassay coefficient of variations of all assays were within 15%.

Eleven out of 76 salivary samples were disqualified as unsuitable because of possible blood contamination or falling outside the range of the assay sensitivity. The ranges of Sa-T, Se free T and the transferrin concentration in saliva were 32 to 360 pg/ml (n = 65, mean 176.8 +/- 96.6 pg/ml, median 168.0 pg/ml), 0.15 to 21.0 pg/ml (n = 76, mean 6.6 +/- 4.2 pg/ml, median 7.0 pg/ml), and 0.08 to 6.0 mg/dl (n = 76, mean 0.66 +/- 0.89 pg/ml median 0.42 mg/dl), respectively. In 10 participants randomized chosen out of all, there was significantly correlation between Se free T and BT (r = 0.964, p < 0.01) . There was significant correlation between Se free T and Sa-T (n = 65, r = 0.592, p < 0.01) and estimated Se free T value can be calculated based on the Sa-T level.

Though disqualified samples accounted for 14.5% of the total number of saliva samples, Sa-T can be used as a substitute parameter for biologically active testosterone.