Vitamin D is associated with bioavailability of androgens in eumenorrheic women with prior pregnancy loss.

Our results show increasing vitamin D concentrations were negatively associated with FAI and positively associated with SHBG in this cohort of eumenorrheic women with history of prior pregnancy loss who were attempting pregnancy. Overall, our results help elucidate the role vitamin D may play in the bioavailability of androgens, a potential factor in the pathophysiology of hyperandrogenism and its related complications for women of reproductive age. The magnitude of the changes in our population of eumenorrheic women was approximately 10%, although the long-term effects of these changes in hormone concentrations on other reproductive outcomes is unknown. To our knowledge, this is the first comprehensive analysis of the relationship between vitamin D and androgens among eumenorrheic, regularly cycling women. Studies of other populations have shown a positive relationship between vitamin D and TT in normal-BMI patients with PCOS 12 and a negative relationship among postmenopausal women. 13 A systematic review and meta-analysis of vitamin D supplementation among a total of 183 women with PCOS found that supplementation with vitamin D significantly reduces TT. 21 Our null results may differ from these studies because our population was, respectively, eumenorrheic, premenopausal, and not being supplemented with vitamin D. With regard to the bioavailability of testosterone, the largest study to date demonstrated a negative relationship between vitamin D and fT among postmenopausal women, 14 which is consistent with our findings. With regard to SHBG, the positive relationship with vitamin D concentration that we observed is consistent with results observed previously among both women with PCOS 22 , 23 and postmenopausal women. 14 , 24 However, our positive finding conflicts with results from the aforementioned meta-analysis; 21 our study’s larger cohort (and therefore greater power) may also explain why we found a relationship between vitamin D and SHBG while the meta-analysis did not. In addition, the present study included a predominantly nonobese population without diagnosed PCOS; it is plausible that differences in obesity prevalence, and accordingly insulin resistance, could perhaps have an impact on the relationship between vitamin D and SHBG. Prior research supports a strong biological rationale for the association between vitamin D and hyperandrogenic states, mediated through vitamin D receptor (VDR) polymorphisms. Different VDR polymorphisms have been linked to an increased risk of having PCOS, 25 increased severity of PCOS, 26 and higher testosterone and androstenedione concentrations than patients with other VDR polymorphisms. 27 VDR polymorphism Bsm-I is specifically associated with lower SHBG concentrations, 28 providing a potential mechanism for vitamin D to affect the bioavailability of androgens as we see in this study. If vitamin D does in fact influence SHBG homoeostasis, then TT could remain unchanged while bioavailability would increase, therefore raising the potential for women to suffer from the physical effects of elevated androgens while having clinically normal total concentrations. Our study is unique in that it evaluated androgen outcomes in eumenorrheic premenopausal women with prior pregnancy loss, whereas prior studies assessed the relationship of vitamin D with either women diagnosed with PCOS or postmenopausal women. Because 92.4% of participants in the trial were taking some sort of vitamin supplement and we do not have data on dietary intake or vitamin contents, it is possible that some patients were consuming supplemental vitamin D for which we were unable to account. However, serum concentrations have been shown to be an accurate biological reflection of vitamin D status, 29 , 30 regardless of source, so we believe these markers should be relevant markers of vitamin D status, given that their intake of vitamins was somewhat consistent over the study. Moreover, we adjusted for the use of any vitamin supplements in a sensitivity analysis and found similar results. It is important to note that our findings are applicable only to vitamin D concentrations with the vitamin D range observed in our population. The cutoffs we used as designated by the Endocrine Society were developed with regard to bonehealth, 18 so they may not be the most appropriate for reproductive outcomes; however, our use of continuous vitamin D concentration models with similar results suggest the relationships are robust, regardless of vitamin D status cutoff. Because our data are cross-sectional in nature, we cannot comment on how changes in vitamin D may affect the bioavailability of androgens over time; prospective studies should be done to further delineate the nature of the relationship between androgen homeostasis and vitamin D. Finally, we do not have information on phenotypic features related to androgen activity such as hirsutism or acne; thus, we can evaluate androgens only biochemically and not clinically in this population. To our knowledge, this is the first study to elucidate the positive relationship of vitamin D with SHBG and negative relationship with the FAI in eumenorrheic, premenopausal women with prior pregnancy loss. Because vitamin D and SHBG interact in other diseases, it is plausible that vitamin D status may affect SHBG concentrations and subsequently the bioavailability of androgens, although we were limited in assessing the directionality of the associations, given the cross-sectional nature of our study. Overall, these findings shed light on a potential pathological process and spurs the need for prospective studies regarding vitamin D and its potential role in addressing androgen excess.

Vitamin D data were available for 1191 patients. The median vitamin D concentration was 29.1 ng/mL (interquartile range, 23.1—36.0) and concentrations ranged overall from 5.0 to 143.6 ng/mL. With regard to race, a higher proportion of vitamin D—sufficient women were white than vitamin D—insufficient and —deficient women ( Table 1 ). Vitamin D—sufficient participants also had lower BMI and smaller waist circumference, sum of skinfolds, and central-to-peripheral skinfold ratio than their vitamin D—insufficient and —deficient counterparts. Vitamin D was associated with neither TT nor DHEAS ( Table 2 ). However, vitamin D was negatively associated with fT in the unadjusted model (percentage change [95% confidence interval], —4% [—6% to —2%]), but this relationship was attenuated after adjustment for potential confounders (—2% [—4%, 0.4%]). Vitamin D concentrations were negatively associated with the FAI (—5% [—7%, —2%]) and positively associated with SHBG concentrations (4% [2—7%]). These findings remained consistent after additional adjustment for TT, E1G, and insulin (7% [4—10%]). Results using vitamin D status (sufficient, insufficient, deficient) yielded similar results. The restricted cubic spline models yielded no evidence of nonlinearity, and the sensitivity analyses yielded nearly identical results (data not shown). Based on our results in this population, an average woman in this population with insufficient (compared with sufficient) vitamin D concentration would exhibit a 10% higher FAI (eg, 1.18 vs 1.07). Likewise, for women with SHBG values near the average for this population, a 9% lower SHBG attributable to vitamin D insufficiency would equate to approximately 59.5 nmol/L compared with 65.5 nmol/L.

This is a cohort from the Effects of Aspirin in Gestation and Reproduction (EAGeR) trial, which was a multicenter, block-randomized, double-blind, placebo-controlled trial of daily low-dose aspirin (81 mg) in 1228 women recruited from 4 US medical centers from 2007 to 2011. Institutional review board approval was obtained at each study site and the data coordinating center. All participants provided written informed consent. The trial was registered with ClinicalTrials. gov, number NCT00467363 . Full details of the study design, methods, and participant characteristics have been published elsewhere. 16 Women attempting conception, aged 18e40 years, with regular menstrual cycles of 21e42 days in length, with documented confirmation of 1 or 2 prior pregnancy losses, and up to 2 prior live births were eligible for the EAGeR trial. 16 Exclusion criteria for all women included clinical indication for use of anticoagulant therapy or chronic use of nonsteroidal anti-inflammatory drugs; major medical disorders (eg, diabetes, hypertension, etc); and any prior diagnosis of infertility or subfertility including related conditions such as PCOS, endometriosis, or pelvic inflammatory disease. Furthermore, all women must have been off long-acting hormonal contraceptive medication (eg, Depo-Provera, Norplant, intrauterine device) for at least 12 months or off oral contraceptive pills or other exogenous hormones (eg, patch, ring) 3 months prior to enrollment. Participants attended a prerandomization, baseline study visit timed to occur around day 2e4 of their menstrual cycle. During this baseline visit, participants completed questionnaires of reproductive health status, demographic, and lifestyle factors. In addition, anthropometric measures and blood samples were collected. Samples were centrifuged and serum and plasma were aliquoted and frozen within 90 minutes. Samples were stored at e80C until analysis. Primary outcomes for the present analysis were baseline serum concentrations of total testosterone, free testosterone, free androgen index, SHBG, and dehydroepiandrosterone sulfate (DHEAS). Total testosterone concentration (TT; nanograms per deciliter) was determined by liquid chromatography and tandem mass spectrometry using a Shimadzu Prominence liquid chromatogram (Shimadzu Scientific Instruments, Inc, Columbia, MD) with an ABSceix 5500 tandem mass spectrometer (AB SCIEX, Framingham, MA). Interassay coefficients of variance (CVs) were 2.0% at 189.81 and 1.4% at 809.54 ng/mL. Free testosterone (fT) was calculated as 24.00314 × TT/log 10 S — 0.0499 × TT 2 and free androgen index (FAI) as 100 × (TT/SHBG), where TT was measured in nanomoles per liter and SHBG in nanomoles per liter. 17 SHBG concentration was determined by SHBG reagent/sandwich immunoassay method/electrochemiluminescence (Roche Diagnostics, Indianapolis, IN) utilizing a Roche COBAS 6000 chemistry analyzer (Roche Diagnostics). Interassay CVs were 3.0% at 55.64 nmol/L and 3.8% at 19.74 nmol/L. DHEAS was determined by DHEA sulfate reagent/competitive immunoassay method/electrochemiluminescence using a Roche COBAS 6000 chemistry analyzer (Roche Diagnostics). Interassay CVs were 4.6% at 5.43 μ mol/L and 4.9% at 13.01 μ mol/L. Insulin was measured by sandwich immunoassay method using a Roche COBAS 6000 chemistry analyzer (Roche Diagnostics). Interassay CVs were 3.1% at 121.2 pmol/L and 3.1% at 377.9 pmol/L. Urinary estrone-3-glucuronide (E1G) was determined by E1G/PdG Multiplex competitive chemiluminescence assay (Quansys Biosciences, Logan, UT) from daily first-morning urine samples collected at home during the first cycle. The interassay CVs for E1G were 16.9% and 20.2% at mean concentrations of 36.3 ng/mL and 1.9 ng/mL, respectively, for lyophilized manufacturer’s controls and were 14.2% and 15.2% for in-house pooled urine controls at mean concentrations of 23.7 ng/mL and 10.2 ng/mL, respectively. Preconception 25-hydroxyvitamin D (25[OH]D) concentrations were measured in serum using the 25(OH)D ELISA solid-phase sandwich enzyme immunoassay (BioVendor R&D, Ashville, NC). The interassay laboratory CVs were 15.8% and 13.1% at mean concentrations of 15.5 and 41.6 ng/mL, respectively, for lyophilized manufacturer’s controls and 17% for an in-house pooled serum control. The lower limit of detection was 1.6 ng/mL, and all values were above this limit. Participants were determined to be vitamin D sufficient (25[OH]D ≥ 30.0 ng/mL), vitamin D insufficient (20.0 ≤ 25[OH]D < 30.0 ng/mL), or vitamin D deficient (<20.0 ng/mL) using baseline 25(OH)D serum concentrations. 18 Demographic and reproductive histories at baseline were compared by vitamin D status using Student’s t tests and χ 2 tests where appropriate. Both unadjusted and adjusted geometric mean androgen concentrations by vitamin D concentration are presented. Linear egression was used to estimate the associations between vitamin D and TT, fT, FAI, DHEAS, and SHBG. All hormonal markers were natural log transformed for normality, and results are reported as percentage change in the hormone measures per unit change in vitamin D concentrations with 95% confidence intervals. 19 All models were adjusted for age, body mass index (BMI), smoking status, race, income, education, physical activity, and season of blood draw. Models for SHBG were additionally adjusted for TT, mean follicular-phase E1G, and insulin because they are known to affect SHBG concentrations. 20 Models were not adjusted for treatment arm because blood samples were collected before treatment. All analyses were run using both categorical vitamin D status (sufficient vs insufficient vs deficient) and continuous vitamin D concentrations. Assumptions of linearity were tested using restricted cubic splines and sensitivity analyses, excluding patients with vitamin D above the 99th percentile, and controlling for vitamin supplement intake were evaluated. SAS version 9.4 (SAS Institute, Cary, NC) was used for all statistical analysis. This study was funded by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health. The funding source had no role in the study design, data gathering, analysis and interpretation, writing of the report, or the decision to submit the report for publication. All authors had full access to the data, and the corresponding author had the final responsibility to submit the report for publication.