Vitamin D and Testosterone in Healthy Men: A Randomized Controlled Trial

CONTEXT

Available evidence shows an association of vitamin D with androgen levels in men. However, results from preliminary randomized controlled trials (RCTs) are conflicting.

OBJECTIVE

To evaluate whether vitamin D supplementation increases total testosterone (TT) levels in healthy men.

DESIGN

The Graz Vitamin D&TT-RCT is a single-center, double-blind, randomized, placebo-controlled trial conducted between December 2012 and January 2017.

SETTING

Endocrine outpatient clinic at the Medical University of Graz, Austria.

PARTICIPANTS

Ninety-eight healthy men with TT levels ≥10.4 nmol/L and 25-hydroxyvitamin D [25(OH)D] levels <75 nmol/L completed the study.

INTERVENTION

Subjects were randomly assigned to receive 20,000 IU/wk of vitamin D3 (n = 50) or placebo (n = 50) for 12 weeks.

MAIN OUTCOME MEASURES

Primary outcome was TT measured using mass spectrometry. Secondary outcomes were free testosterone, sex hormone-binding globulin, estradiol, follicle-stimulating hormone, and luteinizing hormone levels; free androgen index; metabolic characteristics; and body composition.

RESULTS

In healthy men [mean values ± standard deviation: age, 39 years (±13 years); 25(OH)D level, 53.3 nmol/L (±18.3 nmol/L); TT, 19.1 nmol/L (±5.6 nmol/l)], no significant treatment effect on TT was found; however, there were significant effects on quantitative insulin sensitivity check index (QUICKI) and a trend toward decreased Matsuda index. In the treatment group, median (interquartile range) changes for TT, QUICKI, and Matsuda index were 0.5 nmol/L (-0.63 to 0.63 nmol/L; P = 0.497), -0.02 (-0.04 to 0.01; P = 0.034), and -0.9 (-3.2 to 0.8; P = 0.051), respectively.

CONCLUSION

Vitamin D treatment had no effect on TT levels in middle-aged healthy men with normal baseline TT, but it significantly decreased QUICKI. Additional studies investigating vitamin D effects on TT and insulin sensitivity in healthy men are required.

Prostate cancers detected by saturation repeat biopsy impairs the Partin tables' accuracy to predict final pathological stage

OBJECTIVE

• To analyse the overall accuracy of Partin tables, with special emphasis to potential limitations resulting from differences between prostate cancers detected by different biopsy schedules.

PATIENTS AND METHODS

• Clinical characteristics from 599 patients treated with radical prostatectomy defined the 2007 Partin probabilities of organ confinement (OC), seminal vesicle invasion (SVI) and extracapsular extension (ECE). Prostate cancers were detected by initial biopsy (IBx) with ≤12 cores in 405 patients (67.6%), by conventional repeat biopsy (CRBx) with ≤12 cores in 99 (16.5%) and by saturation repeat biopsy (SRBx) with ≥20 cores in 95 patients (15.9%). • The area under the curve (AUC) estimated by the receiver operating characteristic curve, assessed the predictive accuracy of the 2007 Partin tables.

RESULTS

• The Partin tables AUC of the IBx, CRBx and the SRBx groups were 0.730 vs 0.701 vs 0.585 for OC, 0.631 vs 0.689 vs 0.547 for ECE, and 0.775 vs 0.755 vs 0.641 for SVI, respectively.

CONCLUSIONS

• The overall accuracy of the 2007 Partin tables was clearly inferior in patients with prostate cancers detected by SRBx. • Prostate cancers detected by SRBx undermine the Partin tables' overall accuracy, and this group of patients may be miscounselled by vague predictions.