Comparison of testosterone fractions between Framingham Heart Study participants and Japanese participants.

OBJECTIVES

To determine testosterone fractions in Japanese men and to compare these values with those of Framingham Heart Study participants.

METHODS

We enrolled 498 healthy Japanese men. Total testosterone was assayed by liquid chromatography tandem mass spectrometry, sex hormone-binding globulin was assayed by immunoassay and free testosterone was calculated by a laboratory at the Boston Medical Center. Analog-based free testosterone and immunoassay-based total testosterone were determined by immunoassay. We compared mass spectrometry assay-based total testosterone and calculated free testosterone values in the Japanese participants with values in the American Framingham Heart Study third generation cohort.

RESULTS

The mean serum mass spectrometry assay-based total testosterone, sex hormone-binding globulin, and calculated free testosterone values were 439.4 ± 167 ng/dL, 65.34 ± 30.61 nmol/L, and 58.75 ± 20.0 pg/mL, respectively. The correlation coefficients with age for mass spectrometry assay-based total testosterone, sex hormone-binding globulin, and calculated free testosterone were 0.0010, 0.5041, and -0.496, respectively. There were no age-related changes in mass spectrometry assay-based total testosterone values in healthy men (P = 0.981), whereas sex hormone-binding globulin and calculated free testosterone levels showed similar age-related changes (P < 0.0001). Serum analog-based free testosterone levels (8.24 ± 2.9 pg/mL) showed age-related changes (P < 0.0001) regardless of immunoassay-based total testosterone levels (P = 0.828). Serum immunoassay-based total testosterone values (486.1 ± 162.5 ng/dL) correlated with serum mass spectrometry assay-based total testosterone values (r = 0.740, 95% confidence interval 0.6965-0.7781, P < 0.0001). Similarly, analog-based free testosterone and calculated free testosterone values showed a highly significant correlation (r = 0.706, 95% confidence interval 0.6587-0.7473, P < 0.0001). The analog-based free testosterone values were approximately 10% of the calculated free testosterone values.

CONCLUSIONS

In contrast to the Framingham Heart Study cohort, total testosterone values in Japanese men are not associated with advancing age; thus, they cannot be used to diagnose late-onset hypogonadism in Japan. The analog-based free testosterone value can be considered instead as a suitable biochemical determinant for diagnosing late-onset hypogonadism syndrome.