Association between serum vitamin D status and early pregnancy outcomes in first in vitro fertilization cycles with high-quality embryo transfer.
TL;DR
Serum vitamin D status was not independently associated with clinical pregnancy in this cohort, but wide CIs suggest that clinically relevant effects cannot be excluded, warranting larger prospective trials.
Abstract
Objective Vitamin D has been widely studied in assisted reproductive technology, but findings remain inconsistent. This study evaluated the association between serum vitamin D levels and early pregnancy outcomes in a strictly controlled cohort designed to minimize confounding. Methods This retrospective cohort study included 129 women undergoing their first in vitro fertilization (IVF) cycle between 2023 and 2025 with transfer of at least 1 high-quality embryo. Patients were categorized according to serum vitamin D status as insufficient (<30 ng/mL) or sufficient (≥30 ng/mL). Multivariable logistic regression was adjusted for maternal age, anti-Müllerian hormone (AMH), embryo transfer day, and seasonal variation. Results In the full cohort (n=129), clinical pregnancy rates were 54.5% (12/22) in the sufficient group and 33.6% (36/107) in the insufficient group (p=0.065). In the complete-case analysis (n=107) with available AMH data, the sufficient group had a higher clinical pregnancy rate than the insufficient group (63.2% [12/19] vs. 36.4% [32/88]; crude odds ratio [OR], 3.00; 95% confidence interval [CI], 1.07 to 8.39; p=0.036). However, in this analysis set, vitamin D sufficiency was not an independent predictor of clinical pregnancy (adjusted OR [aOR], 2.31; 95% CI, 0.77 to 6.98; p=0.137). Sensitivity analysis using a 20 ng/mL threshold showed no significant association (aOR, 1.21; 95% CI, 0.53 to 2.77; p=0.648). Maternal age remained the only statistically significant predictor (aOR, 0.89 per year; 95% CI, 0.79 to 1.00; p=0.041), even after seasonal adjustment. Conclusion Serum vitamin D status was not independently associated with clinical pregnancy in this cohort. However, the wide CIs suggest that clinically relevant effects cannot be excluded. Previously reported positive associations may reflect residual confounding rather than direct biological effects, warranting larger prospective trials.