Association Between Plasma 25-OH Vitamin D and Testosterone Levels in Men

Katharina Nimptsch; Elizabeth A. Platz; Walter C. Willett; Edward Giovannucci

Disclosures

Clin Endocrinol. 2012;77(1):106-112. 

In This Article

Results

Using common cut-offs for vitamin D status, 24% (n = 333) of participants had 25(OH)D levels <50 nmol/l (vitamin D deficiency), 44% (n = 601) had 25(OH)D levels between 50 and 74·9 nmol/l (insufficiency), and 31% (n = 428) had 25(OH)D levels ≥75 nmol/l (sufficiency). Men in the highest quintile of 25(OH)D were slightly younger, had a lower BMI and higher physical activity than men in the lowest quintile (Table 1). Circulating 25(OH) vitamin D showed seasonal variation with highest concentrations in summer and fall (peak: August) and lowest in winter and spring (nadir: February), whereas no seasonal variation was seen for total testosterone, free testosterone or oestradiol (Fig. 1).

Figure 1.

Monthly variation of mean (age and batch adjusted) 25-hydroxyvitamin D [25(OH)D], total testosterone, free testosterone and oestradiol. Bars indicate 95% confidence intervals.

Total testosterone concentrations significantly increased across quintiles of circulating 25(OH) vitamin D levels (Table 2) in all adjustment models. The age- and batch-adjusted associations were slightly attenuated by additional adjustment for BMI, and further potentially confounding factors including time and season of blood collection, smoking, geographical region and physical activity. When including only control participants (n = 678), multivariate-adjusted mean testosterone concentrations (95% CI) from the lowest to the highest quintile of 25 (OH) vitamin D were 18·0 (16·9; 19·2), 19·4 (18·1; 20·6), 19·4 (18·2; 20·5), 19·7 (18·6; 20·8) and 20·1 (18·9; 21·3) nmol/l (P-trend = 0·01). A positive association persisted when including only control participants without major morbidities (diabetes, high blood pressure or high blood cholesterol; n = 316; data not shown). Additional adjustment for plasma total cholesterol did not change associations remarkably (data not shown).

Circulating 25(OH) vitamin D levels were significantly positively associated with free testosterone in the age- and batch-adjusted as well as in the multivariate-adjusted model. These findings were not changed when only control participants were included (multivariate-adjusted mean free testosterone concentrations (95% CI) from the lowest to the highest quintile of 94·6 (89·7; 99·5), 96·6 (91·0; 102·2), 96·7 (91·3; 102·1), 100·3 (94·7; 106·0) and 100·8 (95·4; 106·2) (P-trend = 0·02).

When estimating the association between 25(OH)D and free testosterone calculated according to Vermeulen et al.,[12] positive associations of 25(OH) D and free testosterone were observed in all models (data not shown). We observed a significant positive association between 25(OH)D and oestradiol in the multivariate-adjusted model.

The spline models show graphically that the positive association between 25(OH)D and total and free testosterone is linear at lower 25(OH)D levels, below approximately 75–85 nmol/l in the case of total testosterone and 75 nmol/l in the case of free testosterone (Fig. 2a,b), reaching a plateau at higher levels (P for nonlinearity 0·05 for total testosterone and 0·14 for free testosterone).

Figure 2.

Association between 25 (OH) vitamin D and (a) total testosterone (b) free testosterone modelled continuously using splines; 95% confidence interval in grey [model adjusted for age (at blood collection), batch, time of blood collection, season, BMI at blood collection (continuous), smoking status, geographical region and physical activity (MET-h/wk)].

No significant differences in the multivariate-adjusted association between 25(OH)D and total testosterone were observed after stratification by age at blood collection or BMI categories (data not shown). There was some suggestion of differences in the association between 25(OH)D and free testosterone by age at blood collection and BMI categories: a positive association between 25 (OH) D and free testosterone was only seen in men aged 70 years or older (P-trend = 0·002), but not in men below age 70 years (P-trend = 0·53; P for interaction 0·02). The association between 25(OH)D and free testosterone was stronger in men with a BMI ≥30 kg/m2 (P-trend = 0·07) than in lower BMI categories (<23, 23–24·9 and 25–29·9 kg/m2) (P for interaction = 0·03).

When we stratified models by vasectomy, significant positive associations between 25(OH) vitamin D and total and free testosterone were observed in men who had not had a vasectomy (n = 1002) (P-trend = 0·001 for total testosterone and 0·004 for free testosterone), while no significant associations were observed in men who had had a vasectomy (n = 360) (P for interaction 0·24 for total testosterone and 0·02 for free testosterone).

We did not observe a clear geographical trend when stratifying by geographical region: significant positive associations (in multivariate-adjusted models) between 25(OH)D and total testosterone were observed among men residing in the South (n = 381, P-trend = 0·01) and in the Northeast (n = 268, P-trend = 0·03), but not in the West (n = 331, P-trend = 0·34) or Midwest (n = 378, P-trend = 0·13). A significant positive association between 25(OH)D and free testosterone was observed only in men residing in the Northeast (P-trend = 0·04).

Stratification by season of blood collection revealed slightly stronger positive associations between 25(OH)D and testosterone during winter or spring (n = 434, P-trend 0·01) than during summer or fall (n = 910, P-trend 0·06), whereas a significant positive association between 25(OH)D and free testosterone was only observed during summer or fall (P-trend 0·02) but not in winter or spring (P-trend 0·92). However, no significant interactions by season were observed. Associations between 25(OH)D and both total and free testosterone were stronger when blood samples drawn in the afternoon were excluded (data not shown).

When defining hypogonadism based on total testosterone levels <11 nmol/l (n = 248), after multivariate adjustment for age, analysis batch, time and season of blood collection, BMI, smoking, geographical region, and physical activity, comparing participants in the highest vs lowest quintile of 25(OH) vitamin D had a significantly decreased relative risk of hypogonadism of 0·50 (95% CI 0·31–0·93; P-trend = 0·01). This association was stronger when participants whose blood samples were drawn in the afternoon were excluded (OR highest vs lowest quintile of 25(OH)D 0·34, 95% CI 0·18–0·65, P-trend = 0·001).

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....