Relationship of Vitamin D status with testosterone levels: a systematic review and meta-analysis

Purpose: Despite a biological plausibility of a direct link between low vitamin D and androgen deficiency, the association remains inconclusive in epidemiological studies. Therefore, this systematic review and meta-analysis of case-control studies aim to assess whether and in what populations such an association can be demonstrated. Methods: A systematic search was performed in PubMed, EMBASE, Cochrane Library, Web of science, Science Direct, and CINAHL. Standardized mean differences (SMDs) and 95% confidence intervals (CIs) in total testosterone (TT) levels between men with 25-hydroxyvitamin D (25(OH)D) <20 and ≥20 ng/mL were combined using random-effects models. Funnel plot and trim-and-fill analysis were used to assess publication bias. Heterogeneity source was explored by a sub-group analysis according to health-related characteristics of the study populations. Results: Eighteen included studies collectively gave information on 9892 men with vitamin D deficiency and 10,675 controls. The pooled SMD revealed a slight, albeit just significant, positive association between 25(OH)D and TT (pooled SMD: −0.23, 95% CI: −0.45 to −0.01; P = 0.04) with a large between-study heterogeneity (I2 = 98%, Pfor heterogeneity < 0.00001). At the sub-group analysis, a significant positive association, along with noticeable decrease in heterogeneity, could only be demonstrated in studies of patients with frailty states (pooled SMD: −0.19; 95% CI: −0.27, −0.10, P < 0.0001; I2 = 51%, Pfor heterogeneity = 0.06). A sensitivity analysis revealed a high stability of the result and the trim-and-fill adjustment for publication bias did not affect pooled estimate. Conclusions: Both hypovitaminosis D and androgen deficiency should be regarded as markers of a poor health status, sharing common underlying aetiologies and risk factors.