Genetic evidence linking retinol to birth weight: A two-sample Mendelian randomization study

Abstract

This study aims to clarify the potential causal effects of dietary antioxidant vitamins on pregnancy outcomes (miscarriage, preterm labor, and birth weight) using Mendelian randomization (MR). Our instrumental variables (IVs) were single-nucleotide polymorphisms (SNPs) linked to retinol, vitamin C, carotene, and vitamin E (P < 5×10⁻⁶). The summary statistics for miscarriage, preterm labor, and birth weight were extracted from genome-wide association studies (GWASs), with a sample size of 56,172, 5480, and 261,932, respectively. The present MR study primarily used the inverse-variance weighted method, with additional sensitivity analyses conducted to evaluate the robustness of the findings. We found that retinol was closely related to the birth weight (β=0.091, 95 %CI: 0.009–0.172, P=0.028), and sensitivity analyses showed similar results (MR-RAPS: OR=1.101, 95 %CI: 1.027–1.180; maximum likelihood: OR=1.098, 95 %CI: 1.011–1.194). While the relationship of retinol with miscarriage or preterm labor was not statistically significant. Additionally, our study did not reveal an association between the carotene, vitamin E, and vitamin C and pregnancy-related outcomes, miscarriage, preterm labor, and birth weight (all P > 0.05). In conclusion, our findings indicated a causal effect between retinol and birth weight and suggested that maintaining retinol at normal levels during pregnancy can prevent low birth weight. Therefore, it would be beneficial to measure retinol levels in pregnant women and to supplement with vitamin A in cases of deficiency, as these could be valuable strategies for improving pregnancy outcomes.